Les interactions entre les membres d'une communauté microbienne peuvent être un moyen d'adaptation dans l'environnement. Parmi les nombreuses interactions qui ont lieu dans un écosystème et qui joue un rôle majeur sur la diversité et la dynamique des populations microbiennes est celui des virus procaryotes et leurs hôtes. Les virus peuvent également arbitrer le transfert de matériel génétique entre les procaryotes (transduction), qui pourrait être un mécanisme d'adaptation rapide. Afin de déterminer l'impact potentiel des virus et la transduction, nous avons besoin d'une meilleure compréhension de la dynamique des interactions entre virus et leurs hôtes dans l'environnement. Les données sur les virus de l'environnement sont rares, et les méthodes pour le suivi de leurs interactions avec les procaryotes sont nécessaires. Clustered regularly interspaced short palindromic repeats (CRISPRs), qui contiennent des séquences virales dans les génomes bactériens, pourraient aider à documenter l'histoire des interactions virus-hôte dans l'environnement. Ainsi, cette thèse vise à explorer les interactions virus-hôte dans un environnement donné à travers du séquences CRISPR.Les virus de la cryosphère sont considérés comme abondantes, très actif et avec de larges gammes d'hôtes. Ces caractéristiques pourraient faire de la transduction virale, un facteur clé pour l’adaptation microbienne dans ces environnements. Des métagénomes publics créés à partir des environnements avec une gamme de températures différents ont été examinés. De cette manière, certaines dynamiques d'interactions virus-hôte se sont révélées comme ayant une corrélation avec la température. Un flux de travail a ensuite été développé pour créer un réseau reliant les virus et leurs hôtes en utilisant des séquences CRISPR obtenus à partir de données métagénomiques de la glace des glaciers et du sol de l'Arctique. La création de réseaux d'infection à traves du CRISPRs a fourni une nouvelle perspective sur les interactions virus-hôte. En outre, nous avons cherché des événements de transduction dans les données métagénomiques par la recherche de séquences virales contenant de l'ADN microbien. L’analyse indiquée que les bactériophages du Ralstonia pourraient être des agents de transduction dans la glace des glaciers de l'Arctique. / Interactions between the members of a microbial community can be a means of adaptation in the environment. Among the many interactions that take place in an ecosystem and have been seen to play a major role on microbial diversity and population dynamics is that of prokaryotic viruses and their hosts. Viruses can also mediate the transfer of genetic material between prokaryotes (transduction), which could be a mechanism for rapid adaptation. In order to determine the potential impact of viruses and transduction, we need a better understanding of the dynamics of interactions between viruses and their hosts in the environment. Data on environmental viruses are scarce, and methods for tracking their interactions with prokaryotes are needed. Clustered regularly interspaced short palindromic repeats (CRISPRs), which contain viral sequences in bacterial genomes, might help document the history of virus-host interactions in the environment. Thus, this thesis aimed to explore virus-host interactions in a given environment through CRISPRs. Viruses in the cryosphere have been seen to be abundant, highly active and with broad host ranges. These characteristics could make viral transduction a key driver of adaptation in these environments. Public metagenomes created from environments over a range of temperatures were examined through sequence and CRISPR analysis. In this fashion, certain virus-host interaction dynamics were found to have a correlation with temperature. A workflow was then developed to create a network linking viruses and their hosts using CRISPR sequences obtained from metagenomic data from Arctic glacial ice and soil. The creation of CRISPR-based infection networks provided a new perspective on virus-host interactions in glacial ice. Moreover, we searched for transduction events in metagenomic data by looking for viral sequences containing microbial DNA. Further analysis of the viral sequences in the CRISPRs indicated that Ralstonia phages might be agents of transduction in Arctic glacial ice.

Viruses are common causative agents of ocular infection among humans. Epidemic keratoconjuntivitis (EKC) is a severe and contagious ocular disease with reported outbreaks worldwide. It is estimated that this disease affects 20-40 million individuals every year, which leads to huge socioeconomic costs for the affected countries. EKC is characterized by keratitis and conjunctivitis but is also associated with pain, edema, lacrimation, and decreased vision that can prolong for months after the infection and in rare cases years. This disease is caused by human adenoviruses (HAdVs), which belong to the family of Adenoviridae. Currently, there is no available treatment against EKC. EKC is mainly caused by HAdV-8, HAdV-19, HAdV-37, HAdV-53, HAdV-54, and HAdV-56, which belong to species D HAdVs. HAdV-8, HAdV-19 and HAdV-37 have previously been shown to use sialic acid (SA)-containing glycans as cellular receptors to bind to and infect human corneal epithelial (HCE) cells. To characterize the receptor in more detail, we performed a glycan array, which included SA-containing glycans. A branched hexasaccharide terminating with SA in each arm was identified as a candidate receptor. This glycan corresponds to the glycan motif found on a ganglioside, GD1a. By performing a series of biological and biochemical experiments we confirmed the function of the GD1a glycan as a cellular receptor for EKC-causing HAdVs. However, the glycan used as a receptor was linked to plasma membrane protein(s) through O-glycosidic bonds, rather than to a lipid (as in the ganglioside). X-ray crystallography analysis showed that the two terminal SA:s interacted with two of the three previously identified SA-binding sites on the knob domain of the HAdV-37 capsid protein known as the fiber. Based on the structural features of the GD1a:HAdV-37 knob interaction, we assumed that a three-armed molecule with each arm terminating with SA would be an efficient inhibitor. Such molecules were designed, synthesized and found to efficiently prevent HAdV-37 binding to and infection of corneal cells. These results indicate that trisialic acids-containing compounds may be used for treatment of EKC. After binding to its primary receptor, most HAdVs have been shown to interact with αVβ3 and αVβ5 integrins to enter human cells. This interaction occurs through the RGD (arginine-alanine-aspartic acid) motif in the capsid protein known as the penton base. However, it was not clear if corneal epithelial cells express αVβ3 and αVβ5 integrins. Thus, to better understand additional early steps of infection by EKC-causing HAdVs, we performed binding and infection competition experiments using human corneal epithelial cells and siRNA, integrin specific antibodies, peptides and RGD-containing ligands indicating that α3, αV, β1 affected HAdV-37 infection of but not binding to HCE cells. We could also see that HAdV-37 co-localize with α3 and αV at after entry into HCE cells. In situ histochemistry confirmed that the expression of α3 and αV in human corneal tissue. Overall, our results suggest that αV and α3 integrins are important for HAdV-37 infection of corneal cells. Altogether, these results provide further insight into the biology of HAdVs and open up for development of novel antiviral drugs.

<div>Viruses are a group of contagious microbes that have compact structures, containing a nucleic acid core and a protein shell. The replication of viruses requires assistance from hosts which can be almost any cellular organism. Viral infections are often associated with diseases and have been a major threat to the human race. To cope with viral diseases, we need to understand viruses, including their structures, life cycle, pathogenesis and interactions with their hosts. The first structure of a human virus was determined by the Rossmann lab in 1985 using X-ray crystallography.</div><div>Thanks to the recent advances in both hardware and software, cryo-electron microscopy (cryo-EM) has emerged as a powerful tool to study virus structures. Cryo-EM allows structural determination for a wide range of specimens to high resolution comparable to what can be achieved by X-ray crystallography. Currently two techniques of cryo-EM are commonly used in structural virology: single particles analysis (SPA) and electron tomography (ET). </div><div>Single particle analysis has been used to determine the structures of viruses complexed with host factors in three studies that are to be discussed with more details in chapters 2-4. </div><div>The structure of B19 parvovirus complexed with Fabs of a neutralizing human antibody was determined to 3.2 Å resolution. This structure showed that amino acids from three neighboring VP2 proteins form a quaternary structure epitope. In addition, the structure of human rhinovirus-C (RV-C) complexed with its cellular receptor, CDHR3, was determined to 3.9 Å resolution. Despite the low occupancy of the receptors, a “powerful” localized 3D classification procedure helped to select viral particles that had more bound receptors. Furthermore, structures were determined to 10 Å resolution of bacteriophage ΦX174 bound to lipopolysaccharide (LPS) bilayers, before and after genome ejection. These structures showed a series of conformational changes that occurred when a phage penetrated the bacterial membranes. These studies are good examples of applying cryo-EM to investigate virus-host interactions.</div><div>However, single particle analysis requires samples to be isolated, homogenous and monodispersed. On the contrary, tomography allows in situ studies and is applicable to samples with more flexibility and more heterogeneity. In the case of ΦX174, the structural changes that are involved in the assembly of the H-tube during infection remains a huge mystery. To provide an environment that is more similar to the surface of a bacterial cell, LPS-containing liposomes were mixed with ΦX174 viruses. It was then observed that the ΦX174 particles bound to these liposomes in a very compact manner which was impossible interpret with single particle analysis. Using cryo-ET, 3D volumes of liposome-ΦX174 complexes were reconstructed and structural details were visualized by sub-tomogram classification and averaging.</div><div>The emergence of cryo-EM has not only made high-resolution structural studies possible but also broadened the scope of samples with which virologists could work. Moreover, studies on flexible and heterogeneous complexes between viruses and host factors are now possible using either single particle analysis or electron tomography. These techniques will help us to understand virus-host relationships and finally, to develop effective anti-viral therapies.</div>

No description available.

Les HPV sont les agents d'infections latentes, d'hyperplasies bénignes ou encore de cancer. Afin de mieux comprendre leur pathogenèse, nous avons cartographié les réseaux d'interactions de protéines de régulation virale E2 pour 12 génotypes HPV. Par double hybride, nous avons procédé à l'identification des interacteurs des protéines E2 suivi par une validation en cellule de mammifère par une méthode basée sur la reconstitution d'une luciferase. Le regroupement des profiles d'interaction montre une corrélation avec la phylogénie, établissant ainsi la contribution de E2 dans la pathogénie associée aux HPV. L'étude des réseaux d'interaction a révélé le ciblage préférentiel de protéines cellulaires hautement connectées, impliquées dans 5 catégories fonctionnelles récapitulant les principales fonctions de E2 mais aussi ouvrant de nouvelles perspectives quant au rôle de cette protéine virale dans les mécanismes d'infection. Dans un deuxième temps, ce travail s'est focalisé sur l'étude d'une interaction impliquant spécifiquement la protéine E2 d'HPV16, le virus le plus représenté dans les cancers cervicaux, et une protéine cellulaire, CCHCR1. En identifiant la surface d'interaction de CCHCR1 sur E2, il s'est avéré qu'elle induisait une compétition avec BRD4, un interacteur majeur de E2, se traduisant par une diminution des capacités transcriptionelle de E2. De même, nous avons montré que CCHCR1 induisait la délocalisation de E2 du noyau vers le cytoplasme. Enfin, nos résultats indiquent qu'en présence de CCHCR1, HPV16 E2 est moins apte à induire la différentiation précoce des kératinocytes, ce qui pourrait potentiellement avoir un effet important sur le cycle viral.

Les virus infectant les archées présentent des morphotypes inhabituels et des génomes extrêmement divers. Leur isolation a permis de développer nos connaissances sur la diversité de la virosphère et demeure une piste de recherche primordiale. Durant ma thèse, j’ai isolé et caractérisé de nouveaux virus d'archées et étudié les interactions virus-hôte dans un système modèle bien établi. Des cultures d'enrichissement et des analyses bioinformatiques nous ont permis de décrire de nouveaux virus de crénarchées hyperthermophiles infectant les espèces du genre Pyrobaculum et ainsi de mieux comprendre la diversité architecturale des virus filamenteux. De plus, un provirus a été identifié chez P. oguniense et l’étude de sa réplication a révélé par microscopie électronique des nanostructures pyramidales à la surface des cellules ressemblant à des structures connues de sortie des virions chez des virus de crénarchées. Deux études de protéomique nous ont fourni un aperçu de la dynamique du protéome de Sulfolobus islandicus : l’analyse du protéome de cellules de S. islandicus non infectées a révélé de nombreuses modifications post-traductionnelles, et l’analyse de la régulation des protéines dans des cellules de S. islandicus infectées par le virus SIRV2 a révélé 136 protéines de l'hôte présentant une régulation temporelle significative. L’analyse structurale de SIRV2 a permis d'étudier la résistance des virus crénarchées à des conditions extrêmes et a révélé pour la première fois que la forme A de l'ADN est biologiquement pertinente. Ces résultats ont contribué au développement des connaissances sur la diversité de la virosphère et sur l'évolution des virus d'archées. / Viruses infecting Archaea display unusual morphotypes and highly diverse genomes. Several virus-host systems have emerged enabling the detailed characterization of virus-host interplay in archaea. However, isolation of new archaeal viruses proved to be instrumental for expanding the knowledge on the diversity of the Earth’s virosphere. Therefore, I have focused on two major lines of research: isolation of new archaeal viruses and characterization of the virus-host interactions in a well-established model system. A new Pyrobaculum virus with a unique architecture among DNA viruses was described, expanding our knowledge on the diversity of architectural solutions explored by filamentous viruses. Furthermore, attempts to trigger the replication of a provirus in P. oguninese led to the development of six-fold pyramidal nanostructures on the cell surface, resembling known virion egress structures of archeal viruses. Finally, I focused on the interplay between Sulfolobus islandicus and the rod-shaped virus SIRV2. Two proteomic studies yielded insights into the dynamics and posttranslational modifications (PTMs) of the Sulfolobus proteome. Sulfolobus proteins were found to carry a high degree of PTMs on functionally diverse proteins. The global analysis of the regulation of viral and host proteins in SIRV2-infected S. islandicus cells yielded insights the temporal regulation of host and virus proteins. Structural studies on SIRV2 virion have resulted in the first ever description of A-form DNA being a biologically relevant form of DNA. Together, these results contribute to the knowledge on the diversity of the extant virosphere, and the biology and evolution of archaeal viruses.

Le virus de la fièvre catarrhale ovine (Bluetongue virus, BTV) est l’agent étiologique de la maladie du même nom, une arbovirose non contagieuse transmise aux ruminants domestiques et sauvages par l’intermédiaire de morsures de moucherons hématophages du genre Culicoides. Il existe actuellement 27 sérotypes décrits de BTV à travers le monde qui se distinguent par les pathologies qu’ils induisent et leur capacité à infecter et se propager chez leur(s) hôte(s) mammifère(s). Le premier objectif de mon projet de thèse visait à identifier les interactions cellulaires spécifiques des sérotypes 8 et 27 pour identifier des facteurs de pathogénicité/virulence et/ou de franchissement de barrière d’espèces. Pour atteindre cet objectif, l’ensemble des protéines virales du BTV a été criblé par la méthode du double-hybride en levure contre deux banques d’ADN complémentaires, l’une d’origine bovine et l’autre d’origine culicoïde. A l’issue de 70 cribles, une centaine de nouvelles interactions virus-hôtes a été mise en évidence et révèle un enrichissement pour quatre processus cellulaires : l’épissage des ARNm, les ribosomes, la SUMOylation et l’apoptose. Cette étude nous a ainsi permis de réaliser le premier interactome pour le BTV qui se poursuit au travers de multiples validations biochimiques et fonctionnelles des interactions identifiées. En parallèle de ce travail de protéomique, le second objectif de mon projet de thèse a été de déterminer l’impact du BTV sur la voie MAPK/ERK, une voie cellulaire essentielle à la prolifération et différenciation cellulaire et classiquement modulée lors d’infections virales. En plus de son rôle antagoniste sur la voie des interférons de type I, nous avons démontré la capacité de la protéine NS3 de BTV à activer la voie MAPK/ERK. En effet, nous avons démontré que NS3 a la capacité d’augmenter le niveau de phosphorylation des protéines kinases ERK1/2 mais également du facteur de traduction eIF4E. Cette fonction, qui semble être spécifique au BTV par rapport aux autres orbivirus, implique l’interaction de NS3 avec la protéine cellulaire BRAF, une protéine MAP3 kinase jouant un rôle majeur dans l’activation de la voie MAPK/ERK. L’activation cette voie par NS3 pourrait être un mécanisme de détournement de la traduction cellulaire au profit de celle du virus mais aussi constituer un élément de réponse pour expliquer l’hyper-inflammation observée dans le cas d’une infection par ce virus / Bluetongue virus (BTV) is the etiological agent of the bluetongue (BT) disease, a non-contagious arbovirus that affects a wide range of wild and domestic ruminants. It is transmitted by blood-feeding midges of the genus Culicoides. There are currently 27 serotypes described of BTV in the world that are distinguished by their differences in term of pathology/virulence and their capacity to infect and disseminate in their mammalian host(s). The first objective of my thesis project was to identify specific cellular interactions of serotype 8 and 27 to reveal new factors of pathogenicity/virulence and/or cross species barrier. To reach this goal, all the proteins encoded by BTV were used as baits to screen, by a high-throughput yeast two-hybrid (Y2H) approach, two complementary DNA libraries originating from hosts naturally infected by BTV : Culicoides and cattle. Therefore, 70 screens were performed to identify a hundred of new virus-host interactions and reveal an enrichment for four cellular processes : mRNA splicing, ribosomes, SUMOylation and apoptosis. This study allowed us to build the first interactome of BTV which continues through multiple biochemical and functional validations of the identified interactions. In parallel to this proteomics work, my second objective was to determine the impact of BTV on the MAPK/ERK pathway, a cellular pathway essential for cell proliferation and differentiation usually modulated during viral infections. In addition to its antagonist role on the type I interferon pathway, we have demonstrated the ability of BTV-NS3 to activate the MAPK/ERK pathway. Indeed, we have demonstrated that NS3 has the ability to increase the level of phosphorylation of ERK1/2 protein and the eIF4E translation factor. This function, which seems to be specific to BTV compared to other orbiviruses, involves the interaction of NS3 with BRAF cellular protein, a MAP3 kinase protein that plays a major role in the regulation of the MAPK/ERK pathway. These results could provide a better understanding of the molecular basis underlying the hijacking of the translation machinery to support virus replication but also constitute a hypothesis to explain the hyperinflammation observed in the BTV infection context

Hepatitis C virus (HCV) is a single-stranded plus-sense RNA virus that is transmitted by blood-to-blood contact, and infects the human liver. HCV has a unique dependence on the liver-specific microRNA miR-122, where miR-122 binds the 5´ un-translated region of the viral RNA at two tandem sites and increases viral RNA abundance. The mechanisms of augmentation are not yet fully understood, but the interaction is known to stabilize the viral RNA, increase translation from the viral internal ribosomal entry site (IRES), and result in increased viral yield. In an attempt to create a small animal model for HCV, we added miR-122 to mouse cell lines previously thought non-permissive to HCV, which rendered these cells permissive to the virus, additionally showing that miR-122 is one of the major determinants of HCV hepatotropism. We found that some wild-type and knockout mouse cell lines – NCoA6 and PKR knockout embryonic fibroblasts – could be rendered permissive to transient HCV sub-genomic, but not full-length, RNA replication upon addition of miR-122, and that other wild-type and knockout cell lines cannot be rendered permissive to HCV replication by addition of miR-122. These knockout cell lines demonstrated varying permissiveness phenotypes between passages and isolates and eventually completely lost permissiveness, and we were unable to achieve sub-genomic RNA replication in PKR knockout primary hepatocytes. Knockdown of NCoA6 and PKR in Huh7.5 cells did not substantially impact sub-genomic replication, leading us to conclude that there are additional factors within the cell lines that affect their permissiveness for HCV replication such as epigenetic regulation during passage or transformation and immortalization. We also added miR-122 to Hep3B cells, a human hepatoma cell line lacking expression of miR-122 and previously thought to be non-permissive to HCV replication. Added miR-122 rendered the cells as highly permissive to HCV replication as the Huh7-derived cell lines commonly used to study the virus. In these cells, we were also able to observe miR-122-independent replication of sub-genomic HCV RNA. This was verified by use of a miR-122 antagonist that had no impact on the putative miR-122-independent replication, and by mutating the miR-122 binding sites to make them dependent on a single nucleotide-substituted microRNA. This replication in the absence of miR-122 was not detected in full-length HCV RNA, but was detectable using a bi-cistronic full-length genomic replicon, suggesting that the addition of a second IRES in sub-genomic and full-genomic replicons altered replication dynamics enough to allow detectable RNA replication without miR-122 binding. Because miR-122 has been implicated in protecting the viral RNA from destabilization and degradation by Xrn1, the main cytoplasmic 5´ to 3´ RNA exonuclease, we employed our miR-122-independent system to test this miR-122-mediated protection. We verified that miR-122 functions to protect the viral RNA from Xrn1, but this was insufficient to account for the overall impact of miR-122 on replication, meaning that miR-122 has further functions in the virus’ life cycle. We showed that the effect of miR-122 on translation is due to stabilization of the RNA by protecting it from Xrn1, through binding at both sites. We further evaluated the role of each miR-122 binding site (S1 and S2) in the virus life cycle, and found that binding at each site contributes equally to increasing viral RNA replication, while binding at both sites exerts a co-operative effect. Finally, we determined that binding of miR-122 at site S2 is more important for protection from Xrn1, suggesting that miR-122 binding at S1 is more important for the additional functions of miR-122 in enhancing HCV RNA accumulation. Altogether, we have shown that miR-122 is partially responsible for the hepatotropic nature of Hepatitis C virus, and that supplementation with this microRNA can render non-permissive cells permissive to viral replication. We have also identified and confirmed replication of both sub-genomic and full-length HCV RNA in the absence of miR-122. Finally, we have characterized the impact of the host RNA exonuclease Xrn1 on the HCV life cycle, and determined the roles of each miR-122 binding site in shielding the viral RNA from this host restriction factor.

Bien que le Virus Respiratoire Syncytial, responsable de la bronchiolite du nourrisson, soit aujourd’hui un problème de santé publique majeur, il n’existe encore aucun vaccin ou antiviral curatif contre ce pathogène. Le manque de données sur les étapes clés du cycle viral et sur les interactions virus-cellule freine le développement de nouvelles molécules antivirales.Nous avons étudié l’interactome de deux protéines virales : la polymérase L et le facteur de transcription M2-1. Dans ce but, nous avons mis au point un crible s’appuyant à la fois sur des critères d’interactomique et sur des critères fonctionnels.La première étape consistait à identifier des partenaires potentiels de M2-1 et L par des co-immunoprécipitations couplées à une approche de protéomique quantitative. Pour plus de pertinence, ce crible a été réalisé sur cellules infectées, grâce des virus recombinants produits par génétique inverse. Ceci nous a permis d’identifier 45 et 137 partenaires potentiels de L et M2-1 respectivement. Une étude systématique de l’impact de l’inhibition de 15 partenaires potentiels de M2-1 sur la multiplication virale a mis en avant trois candidats : ILF2, PABPN1 et PABPC1.Nous nous sommes par la suite concentrés sur PABPC1. L’inhibition de l’expression de PABPC1 altère la multiplication virale, mais nous n’avons pas pu mettre en évidence un effet spécifique sur la transcription ou la traduction virale. Son interaction avec M2-1 a été confirmée, et le domaine MLLE de PABPC1 a été identifié comme le site de liaison à M2-1. L’interaction entre M2-1 et PABPC1 a été observée à la fois dans le cytoplasme et dans les IBAGs, des sous-structures concentrant les ARNm viraux au sein des corps d’inclusion viraux. Nous avons formulé l’hypothèse que M2-1, liée à PABPC1, accompagne les ARNm viraux après leur sortie des corps d’inclusion. Ceci suggère un rôle de M2-1 dans le devenir des ARNm viraux en aval de leur transcription. / Although the Respiratory Syncytial Virus, responsible of bronchiolitis in infants, represents a major public health problem, there are currently no vaccine or curative antiviral directed against it. The lack of information on key steps of its viral cycle and on virus-cell interactions hinders the development of new antiviral molecules.We chose to study the interactome of two viral proteins: the polymerase L and the transcription factor M2-1. To do so, we developed a screen based on interactomic and functional criteria.The first step consisted in identifying potential binding partners of M2-1 and L by co-immunoprecipitations coupled to quantitative proteomics. For better relevance, this screen was realised on infected cells, thanks to recombinant viruses produced by reverse genetics. 45 and 137 potential binding partners of M2-1 and L respectively were thus identified. A systematic study of the inhibition of 15 potential partners of M2-1 and its impact on viral multiplication enabled the selection of three candidates: ILF2, PABPN1 and PABPC1.We chose to concentrate on PABPC1. The inhibition of PABPC1’s expression reduces viral multiplication, but no specific effect on viral transcription or translation was brought to light. Its interaction with M2-1 was confirmed, and the MLLE domain of PABPC1 was identified as the M2-1 binding site. The interaction between M2-1 and PABPC1 was observed both in the cytoplasm and in IBAGs, substructures of viral inclusion bodies where viral mRNA accumulate. We formulated the hypothesis that M2-1, with PABPC1, stays with viral mRNA after leaving inclusion bodies and during their translation. This suggests a role for M2-1 in the fate of viral mRNA downstream of transcription.