Viruses represent the most abundant biological entities on earth where, they are able to interact with all kingdoms of life. Yet their diversity, ecology and evolutionary aspects are only beginning to be fully elucidated, mainly due to technical limitations. The vast majority of the microbial world remains elusive to culture; more than 90% of genome sequenced viral isolates infect only 5 of the 54 prokaryotic phyla that are currently recognized. In contrast, viral metagenomics bypasses the need for cultures by directly sequencing fragmented genetic material of environmental viral communities. This dissertation uses viral metagenomics by applying well-tested bioinformatic protocols and expanding them to compare and contrast patterns of diversity, richness and specialization of large viral metagenomic datasets, in both local and global scales. First I demonstrate the utility of a functional-based perspective by adopting the protein cluster environment to estimate global viral diversity. Then, I use this PC approach to analyze metagenomes from two ecologically different environments, which by uncovering local gene specialization showcases the adequacy of a gene-centered workflow. Then I continue to expand upon this PC framework to study the Tara Oceans virome analyses of these data reveal patters of diversity that support a seed bank model. Finally, in search of a more meaningful ecological unit, I move from a gene-centered standpoint towards a population-based frame. We adopted a novel metagenomic technique that allowed me to uncover the discontinuity in the genomic sequence space, thus empirically defining a population. This final contribution will allow to sort and count viral communities, the first step to applying ecological and evolutionary theory.

Philosophiae Doctor - PhD / There are approximately 10³¹ bacteriophages in the biosphere, outnumbering bacteria 10:1, hence, the dynamic and specific nature of phage-host interactions exerts significant influence on microbial communities. Bacteriophages also represent the reservoir of the highest known genetic diversity making them a potential source of novel biotechnological products. However, the isolation of novel bacteriophages is limited by the observation that less than 1% of bacterial hosts have been cultured. This study aimed to bypass this problem by developing novel culture independent approaches to improve our ability to isolate novel phage-host pairs. Samples were collected from an abandoned copper prospecting site near the Gobabeb Desert Research and Training Station and a Salt lake located in the Swakopmund region of the Namibian desert. Two approaches were explored in this study namely viral tagging and reverse metaviromics. For viral tagging, fluorescently labelling the environmental phage fraction before challenging the environmental bacterial fraction with tagged phages proved difficult. This was most likely due to the complex interaction of the labelling agent with phages and requires further studies. For the reverse metaviromics approach, total DNA from the environmental phage fractions was extracted, sequenced and analyzed for novel phages. Analysis of the phage diversity showed that the copper site was dominated by tailed viruses as has been shown for other extreme arid environments. However, the saline site was atypical of marine environments, with tailed viruses being the most abundant, suggesting that the diversity present is not only driven by salinity. Using the metaviromic sequence data to guide the selection of potential bacterial hosts, two strategies were employed. In the first, putative hosts were predicted based on similarity of phage sequences to those identified in databases. Media targeting these specific genera were employed, 8 bacterial species were isolated and based on 16S rRNA similarity to the closest known species were identified as Halomonas caseinilytica, Halomonas eurihalina, Halomonas sinaiensis, Idiomarina loihiensis, Marinobacter xestospongiae, Virgibacillus salarius and two Salinivibrio species. The 16S rRNA analysis also suggested that H. sinaiensis, V. salarius and both Salinivibrio species are novel. All 8 isolates were challenged with the environmental phage fraction. A novel phage, SMHB1, was isolated on one of the Salinivibrio spp. and is only the second characterized phage ever described for this genus. SMHB1 is a 32 kb myovirus, with a head diameter of 56 nm, and a tail length of 106 nm. The second approach involved the design of fluorescently labelled probes targeting phages identified from the metaviromic sequence data. In a control E. coli system to detect cloned phage DNA fragments, 87% of the interrogated cells showed significant hybridization of the phage specific probe to the target. The optimized method was applied to a simulated environmental bacterial fraction and a detection limit of 1:100 was observed for the bacteria containing the phage DNA fragment of interest. This study demonstrates the possibility of improving the specificity of isolating phage-host pairs in a culture-independent manner by incorporating sequence data in the experimental design; and contributes to our knowledge of the phage diversity of an understudied extreme environment.

The presence of pathogenic viruses in coastal environments is an important tool in evaluating water quality and health risks. Millions of viruses are excreted in fecal matter and bacterial indicators do not correlate with the presence of pathogenic viruses. Enteroviruses have been used to identify fecal pollution in the environment; however, other viruses shed in fecal matter could be used to indicate fecal pollution. The purpose of this research is to develop a baseline understanding of the diversity of viruses found in raw sewage and to assess their presence in the marine environment. PCR was used to detect adenoviruses, herpesviruses, hepatitis B viruses, morbilliviruses, noroviruses, papillomaviruses, pepper mild mottle viruses, picobirnaviruses, reoviruses, rotaviruses, and sapporoviruses in raw sewage collected from throughout the United States and from five marine environments ranging in their proximity to dense human populations. Adenoviruses, noroviruses, pepper mild mottle viruses, and picobirnaviruses were detected in raw sewage but absent in the marine environment, making these viruses potential indicators of fecal pollution in marine environments. These viruses were also found in many of the final effluent samples. Pepper mild mottle viruses may be useful for source tracking fecal contamination since it was consistently found in human sewage and is not expected in the feces of other animals due to its dietary origin. Furthermore, this research uncovered previously unknown sequence diversity in human picobirnaviruses. This baseline understanding of viruses in raw sewage and the marine environment will enable educated decisions to be made regarding the use of viruses in water quality assessments.

A água doce no Brasil, assim como o seu consumo é extremamente importante para as diversas atividades criadas pelo ser humano. Por esta razão o consumo deste bem é muito grande e consequentemente, provocando o seu impacto. Os mananciais são normalmente usados para abastecimento doméstico, comercial, industrial e outros fins. Os estudos na área de ecologia de micro-organismos em águas (mananciais e esgoto) vêm sendo realizados com mais intensidade nos últimos anos. Nas últimas décadas foi introduzido o conceito de virioplâncton com base na abundância e diversidade de partículas virais presentes no ambiente aquático. O virioplâncton influencia muitos processos ecológicos e biogeoquímicos, como ciclagem de nutriente, taxa de sedimentação de partículas, diversidade e distribuição de espécies de algas e bactérias, controle de florações de fitoplâncton e transferência genética horizontal. Os estudos nesta área da Virologia molecular ainda estão muito restritos no país, bem como muito pouco se conhece sobre a diversidade viral na água no Brasil. / Freshwater in Brazil, as well as its consumption is extremely important for the various activities created by human being. For this reason the consumption of this good is very great and consequently, causing its impact. The sources are usually used for domestic, commercial, industrial and other purposes. Studies on the ecology of microorganisms in waters (freshwater and sewage) have been carried out more intensively in recent years. In recent decades the concept of virioplankton has been introduced based on the abundance and diversity of viral particles present in the aquatic environment. Virioplankton influences many ecological and biogeochemical processes, such as nutrient cycling, particle sedimentation rate, diversity and distribution of algal and bacterial species, control of phytoplankton blooms and horizontal gene transfer. Studies in this area of molecular Virology are still very restricted in the country, and very little is known about viral diversity in water in Brazil.

Le virus chikungunya est un arbovirus, transmis par les moustiques du genre Aedes, qui provoque des arthralgies invalidantes et parfois des rhumatismes chroniques. Dans une première partie nous avons décrit les aspects ambulatoires cliniques, biologiques et virologiques du chikungunya (CHIK) de la phase aiguë jusqu'au 300ème jour lors de l’épidémie de la Réunion en 2006. Des scores d’aide au diagnostic ont été élaboré et une étude de la diversité virale intra-hôte a été réalisée. Pour compléter nos premiers résultats nous avons étudié une épidémie survenue en République du Congo en 2011. La description clinique était similaire à celle identifiée lors de l’épidémie de la Réunion. L’évaluation du score clinique ne permettait pas de le proposer comme outil diagnostique à l’échelle individuelle mais apparaissait comme un bon marqueur pour le suivi de la courbe épidémique. Une étude de séroprévalence et une analyse phylogénétique complètent ce travail. Le dernier travail porte sur l’utilisation de la chloroquine à la phase aiguë du CHIK lors d’une prise prophylactique chez le singe et lors d’un essai clinique chez l’homme. Le principal effet de ce type de traitement semble lié son action immuno-modulatrice ; en prise préventive il provoque une exacerbation de la symptomatologie aiguë tandis qu’en prise à la phase précoce de la maladie il augmente le risque d’évolution vers des arthralgies chroniques. En conclusion nous avons réalisé une description des formes ambulatoires du CHIK, identifié des facteurs de risques de formes chroniques, proposé des scores d’aide au diagnostic et argumenté la contre-indication de l’utilisation de la chloroquine à la phase aiguë du CHIK. / Chikungunya virus (CHIKV) is an arthropod-borne virus transmitted by Aedes mosquitoes that cause debilitating arthralgia and possible chronic rheumatism. In the first part we describe the clinical, biological and virological presentation of outpatients with chikungunya disease (CHIK) from the acute stage to the chronic stage at day 300, during the outbreak in the Reunion Island in 2006. We elaborated scores for CHIK diagnosis and we also analysed the intra-host genetic diversity.To complete our first results, we investigated a CHIKV outbreak, which occurred in the Republic of Congo in 2011. The clinical presentation was similar to the first description of the Reunion island outbreak. We assessed the clinical score which appeared to be unusable at the individual level but was still relevant to follow the epidemic curve. This work was completed by seroprevalence and phylogenetic analyses.The last study presented in this thesis focused on the use of chloroquine during the acute stage of CHIK in a non-human primate (NHP) model (prophylactic use) and during a clinical trial (therapeutic use). The main effect of chloroquine treatment at the acute stage of CHIK appeared to be related to its immuno-modulatory action; in prophylactic taking, chloroquine exacerbated acute symptoms while treatment during the early stages of the disease increased the risk of acquiring chronic arthralgia.In conclusion, we provide a detailed description of CHIK outpatients and identify risk factors for the chronic stage of the disease. We propose tentative diagnostic scores and we firmly establish that the use of chloroquine at the acute phase of CHIK is contraindicated.

La grippe est une infection respiratoire responsable de complications respiratoires ou neurologiques nécessitant une prise en charge rapide et adaptée. L’émergence des technologies de séquençage à haut débit (NGS) permet l’étude des communautés microbiennes résidentes ainsi qu’une étude approfondie du génome des pathogènes impliqués. Cette thèse a pour objectif de caractériser le microbiome respiratoire et la diversité génomique virale des patients infectés par les virus grippaux, en corrélant les données clinicobiologiques recueillies. Après recueil des prélèvements respiratoires d’enfants hospitalisés entre 2010 et 2014, le séquençage de leur microbiome respiratoire a mis en évidence une augmentation de la diversité microbienne ainsi qu’une signature microbienne différentielle entre formes cliniques. Une répartition différentielle de taxons (OTU) permet la prédiction de complications chez les enfants infectés. L’étude d’échantillons respiratoires de patients adultes permettra de compléter la signature prédictive. Après validation des processus analytiques et bioinformatiques par reconstitution artificielles de quasi espèces et recueil de 125 prélèvements cliniques respiratoires, le séquençage du génome entier par NGS des virus grippaux permet de différencier les diversités initiales en fonction de la nature du virus infectant et de la complication. En comparaison du prélèvement initial précoce les échantillons prélevés successivement mettent en évidence une diversification différentielle entre les différents segments des virus grippaux infectant les patients, que ce soit chez les patients immunocompétents ou chez un patient immunodéprimé à l’excrétion prolongé / Influenza is a respiratory infection responsible for respiratory or neurological complications and require rapid and adapted management. The emergence of next-generation sequencing (NGS) allows the study of resident microbial communities as well as an in-depth study of the genome of the pathogens. This thesis aimed to characterize the respiratory microbiome and the viral genomic diversity of influenza virus infected patients, correlating these data to the collected clinical data. After sampling of respiratory specimens from hospitalized children between 2010 and 2014, the sequencing of their respiratory microbiome revealed an increase in microbial diversity and a differential microbial signature between clinical forms. A differential taxon distribution (OTU) allows the prediction of complications in infected children. The study of adult respiratory samples will complete the predictive signature.After validation of the analytical and bioinformatic processes by artificial reconstitution of quasi-species and collection of 125 respiratory clinical specimens, the sequencing of the whole genome by NGS of the influenza viruses allow to differentiate the initial diversities according to the nature of the infecting virus and the complication. Compared to early samples, specimen sampled successively show a differential diversification between the different segments of influenza viruses, whether in immunocompetent patients or in an immunocompromised patient with prolonged excretion

La primo-infection est un moment critique de l’établissement du réservoir justifiant de l’initiation d’un traitement précoce. Nous avons initié un essai randomisé évaluant l’impact de deux ans d’un traitement antirétroviral intense (essai ANRS147 OPTIPRIM, trithérapie versus pentathérapie) sur le réservoir et avons initié des études physiopathologiques au cours de cet essai. Nous montrons ainsi la faible diversité génétique des virus en primo-infection dans les compartiments sanguins et rectaux. Le réservoir s’établit dès le premier mois de l’infection par diffusion d’un cluster viral homogène au sein des lymphocytaires T CD4 naïfs (TN) et mémoires centrales (TCM), transitionnelles (TTM), effectrices (TEM) quiescents. Il en résulte une perturbation de l’homéostasie lymphocytaire associée à une faible contribution au réservoir des cellules peu différenciées à longue demi-vie, TN et TCM. Par ailleurs nous montrons que la majorité des patients au moment de leur primo-infection n’ont pas la capacité de développer des réponses T CD8 à même de supprimer la réplication virale comme chez les patients HIV Controllers. Après deux ans de traitement, nous observons que la diversité virale n’a pas évolué, par contre la taille du réservoir est fortement réduite. Les anomalies de l’homéostasie lymphocytaire T CD4 persistent, par contre le traitement très précoce a permis de protéger les TN et TCM. Il n’y a pas de bénéfice additionnel d’une pentathérapie mais nous avons validé le concept qu’un traitement précoce permet d’induire un contrôle virologique au long cours après arrêt de traitement. Nos résultats indiquent qu’un traitement plus long que deux ans permettrait de renforcer la diminution du réservoir. Ces résultats seront à prendre en compte pour l’élaboration de futurs essais en primo-infection visant à réduire le réservoir pour une rémission au long cours. / HIV primary infection is a critical period in the establishment of the reservoirs that justifies the initiation of an early treatment. We started a randomised trial to assess the impact of a two-year intense HAART (ANRS147 OPTIPRIM trial: five-drug therapy versus. three-drug therapy) on the blood reservoir; within this this trial, we included some pathophysiological studies. Thus, we show that during the primary infection, viruses have a low genetic diversity in blood and rectal compartments. The reservoir establishes itself as early as the first month of the infection by spreading a homogeneous viral cluster in CD4 T cells subsets, naive T cells (TN), central memories (TCM), transitional memories (TTM), effector memories (TEM), and resting T cells. This results in a disruption of the lymphocyte homeostasis, linked to the low contribution to the reservoir of TN and TCM, which are little differentiated cells with long half-lives. Moreover, we show that, at the time of the primary infection, the majority of patients do not have the ability to develop CD8 T cells responses that could suppress the viral replication, as HIV Controllers patients do. After two years of treatment, we observe that there is no evolution of the viral diversity, but the size of the reservoir is significantly reduced. The abnormalities of the CD4 T cells lymphocyte homeostasis remain, but the very early treatment was able to protect the TN and TCM. The five-drug therapy does not have any additional benefit, but we confirm the idea that early treatment can induce long-term virological control after the discontinuation of the treatment. Our results show that a treatment lasting more than two years would be able to reinforce the reduction of the reservoir. These results should be taken into account in the development of future trials aiming to reduce the reservoir in patients treated at the time of primary infection for a sustainable remission.

L’épidémiologie moléculaire des virus VIH-1 en France est caractérisée par une augmentation constante de la diversité virologique et de la fréquence des virus de sous-types non-B chez les patients en primo-infection. Entre 1997 et 2007, 28.4% des 591 patients suivis étaient infectés par des virus de sous-types non-B. De plus, 49 patients (8.3%) étaient infectés par des souches de sous-types différents sur les gènes pol et env, témoignant d’évènements de recombinaisons entre ces gènes. Ces virus recombinants étaient isolés à la fois chez des patients originaires d’Afrique sub-saharienne (28.3%) et des sujets caucasiens (6.3%). Ces résultats témoignent des échanges de souches virales entre les populations d'origine africaine et caucasienne, contribuant encore à augmenter la diversité virologique dans ces deux populations. Parmi 131 virus de sous-types non-B, 12.2% étaient classés comme ayant un tropisme CXCR4 par méthode génotypique, mais seulement 0.8% par méthode phénotypique, indiquant d’une part la faible proportion de virus non-B de tropisme CXCR4 en primo-infection en France, et d’autre part le manque de spécificité des méthodes génotypiques de détermination du tropisme pour ces sous-types, rendant nécessaire la mise au point d’autres algorithmes spécifiques pour ces virus.L’analyse de 987 virus isolés dans la cohorte entre 1999 et 2010 a mis en évidence que 12.7% d’entre eux étaient regroupés en "clusters" de transmission. Les patients en primo-infection contribuent donc de façon significative à la propagation de l’épidémie de VIH en France, particulièrement les hommes homo/bi-sexuels, avec une fréquence augmentant au cours de la période récente (2006-2010).La comparaison des quasi-espèces virales circulant concomitamment chez 8 patients en primo-infection (« receveurs ») et leurs 8 partenaires sexuels respectifs (« donneurs ») a révélé dans tous les cas la transmission d’un virus unique, présent de façon minoritaire parmi les sous-populations virales du donneur. La transmission virale muqueuse implique donc une sélection génétique drastique. / High genetic diversity is a major characteristics of HIV-1. In France, although subtype B strains are still predominant, the proportion of non-B viruses isolated in patients at the time of primary HIV-1 (PHI) infection increases over time. Between 1997 and 2007, 28.4% of patients were infected with non-B subtypes strains. Forty-nine viruses showed different phylogenies between the pol and env genes, indicating that recombinations have occurred in 8.3% of cases. These recombinants were isolated both in patients from Sub-Saharan Africa (28.3%) and in white subjects (6.3%).The phenotypic analysis of viral tropism of 131 non-B strains showed a very low (0.8%) proportion of CXCR4-tropic strains (X4 strains) at the time of PHI. Compared to phenotypic tests, genotypic predictions can overestimate (12.2% versus 0.8%) the proportion of X4 strains in non-B subtypes.The phylogenetic analysis of 987 strains isolated in 1999-2010 showed that 12.7% of PHI cosegregated into 56 transmission chains. PHIs are a significant source of onward transmission, especially in men having sex with men, with increasing frequency during the recent years (10.2% in 1999-2006 versus 15.2% in 2006-2010, p=0.02).The comparison of the viral quasispecies isolated in plasma and PBMC samples from 8 patients at the time of PHI ("recipients") and their transmitting partners ("donors") suggested that a severe genetic bottleneck occurrs during HIV-1 heterosexual and homosexual transmission. Indeed, we observed in all cases the transmission of a single variant, which was derived from an infrequent variant population within the blood of the donor. The proportion of X4 quasispecies in donors were higher in case of X4 versus CCR5-tropic viral transmission, suggesting that X4 transmission may be associated with a threshold of X4 circulating quasispecies in donors.