Positive-strand RNAviruses invariably assemble their viral replication complexes (VRCs) by remodeling host intracellular membranes. How viral replication proteins are targeted to specific organelle membranes to initiate VRC assembly remains elusive. Brome mosaic virus (BMV), whose replication can be recapitulated in Saccharomyces cerevisiae, assembles its VRCs by invaginating the outer perinuclear endoplasmic reticulum (ER) membrane. Remarkably, BMV replication protein 1a (BMV 1a) is the only viral protein required for such membrane remodeling. We show that ER-vesicle protein of 14 kD (Erv14), a cargo receptor of coat protein complex II (COPII), interacts with BMV 1a. Moreover, the perinuclear ER localization of BMV 1a is disrupted in cells lacking ERV14 or expressing dysfunctional COPII coat components (Sec13, Sec24 or Sec31). The requirement of Erv14 for the localization of BMV 1a is bypassed by addition of a Sec24-recognizable sorting signal to BMV 1a or by overexpressing Sec24, suggesting a coordinated effort by both Erv14 and Sec24 for the proper localization of BMV 1a. The COPII pathway is well known for being involved in protein secretion; our data suggest that a subset of COPII coat proteins have an unrecognized role in targeting proteins to the perinuclear ER membrane.

Arenaviruses are enveloped negative-strand RNA viruses that cause significant human disease. Despite decades of research, it is still unclear how these viruses establish a lifelong, asymptomatic infection in their rodent hosts while infection of humans often results in severe disease. Unable to enter a state of bona fide latency, the transcription and replication of the viral genomic RNA is likely highly regulated in time and subcellular space. Moreover, we hypothesize that the viral nucleoprotein (NP), responsible for the encapsidation of the viral RNA and the most highly expressed viral gene product, plays a key role in the regulation of the viral gene expression program. Further, exploring host-virus interactions may elucidate the basic aspects of arenavirus biology and how they cause such severe disease in humans. To explore these questions in greater detail, this dissertation has pursued three main avenues. First, to better understand lymphocytic choriomeningitis mammarenavirus (LCMV) genome replication and transcription at the single-cell level, we established a high-throughput, single-molecule (sm)FISH image acquisition and analysis pipeline and followed viral RNA species from viral entry through the late stages of persistent infection in vitro. This work provided support for a cyclical model of persistence where individual cells are initially transiently infected, clear active infection, and become re-infected from neighboring reservoir cells within the population. Second, we used FISH to visualize viral genomic RNA to describe the subcellular sites where LCMV RNAs localize during infection. We observed that, viral RNA concentrates in large subcellular structures located near the cellular microtubule organizing center and colocalizes with the early endosomal marker Rab5c and the viral glycoprotein in a proportion of infected cells. We propose that the virus is using the surface of a cellular membrane bound organelle as a site for the pre-assembly of viral components including genomic RNA and viral glycoprotein prior to their transport to the plasma membrane where new particles will bud. Last, we used mass spectrometry to identify human proteins that interact with the NPs of LCMV and Junín mammareanavirus (JUNV) strain Candid #1. We provided a detailed map of the host machinery engaged by arenavirus NPs, and in particular, showed that NP associates with the double-stranded RNA (dsRNA)-activated protein kinase (PKR), a well-characterized antiviral protein that inhibits cap-dependent protein translation initiation via phosphorylation of eIF2α. We demonstrated that JUNV antagonizes the antiviral activity of PKR completely, effectively abrogating the antiviral activity of this surveillance pathway. In sum, the work composing this dissertation has given us fresh insight into how arenaviruses establish and maintain persistence; the nature of the subcellular site where viral genomic RNA is transcribed, replicated, and assembled with other viral components; and a global view of the cellular machinery hijacked by the viral nucleoprotein. This work improves our basic understanding of the arenavirus life cycle and may suggest novel antiviral therapeutic targets that could be exploited in the future.

Les virus sont des parasites intracellulaires obligatoires recrutant des cofacteurs cellulaires afin de détourner les différents processus biologiques leur permettant notamment de répliquer leur génome et de former d'autres particules virales. Si des cofacteurs cellulaires de la réplication du virus Semliki Forest ont été récemment identifiés, très peu d'études ont permis de révéler des partenaires de la réplication du proche Alphavirus Chikungunya (CHIKV). Nous avons découvert, au cours de cette étude, un recrutement d'Hélicases à domaine DExD/H au niveau de sites de réplication du CHIKV. Parmi elles, DHX9 ou RNA Helicase A (RHA), grâce à ses propriétés de liaison et de modulation de structures des ARNs ou de complexes de Ribonucléoprotéines, est impliquée dans diverses fonctions depuis la transcription, la traduction, la réplication de génomes et jusqu'à la production de particules infectieuses de nombreux virus. Dans le cas du virus Chikungunya, nous avons caractérisé une fonction provirale dans la traduction de protéines non-structurales et une fonction antivirale dans la réplication du génome. Cette double fonction opposée est manipulée par le CHIKV afin d'assurer une production de protéines non-structurales composant le complexe de réplication tout en maintenant sa réplication. Ces travaux révèlent un nouveau mécanisme de régulation de la traduction d'ARN génomique de CHIKV et apportent des éléments de compréhension dans la dynamique de passage du phénomène de traduction à l'étape de réplication du génome CHIKV. / Viruses are obligate intracellular parasites recruiting cellular cofactors to divert different biological processes enabling them to replicate their genome and to form other viral particles. If cellular cofactors of Semliki Forest virus replication have recently been identified, very few studies have revealed the replication partners of the very close Alphavirus Chikungunya (CHIKV). During this study, We have discovered recruitments of several DExD/H Box Helicases at the CHIKV replication sites. Among them, DHX9 or RNA Helicase A (RHA) through its RNA binding properties and in modulating RNA secondary structures or Ribonucleoproteins complexes, is involved in various functions from transcription, translation, replication of genomes and up to production of infectious particles of many viruses. In the case of Chikungunya virus, we have characterized a proviral function in the translation of non-structural proteins and an antiviral function in the genome replication. These opposite functions are manipulated by CHIKV to ensure production nonstructural proteins, components of the CHIKV replication complex while maintaining its replication. These works reveal a new translation regulation mechanism of CHIKV genomic RNA and bring some knowledge on the passage from the translation stage to the replication step of CHIKV genome.