AIX-MARSEILLE UNIVERSITE FACULTE DE MEDECINE DE MARSEILLE ECOLE DOCTORALE DES SCIENCES DE LA VIE ET DE LA SANTE

THESE DE DOCTORAT Présentée par Morgan GAÏA Né le 24 Octobre 1987 à Aubagne, France

Pour obtenir le grade de DOCTEUR de l’UNIVERSITE AIX -MARSEILLE SPECIALITE : Pathologie Humaine, Maladies Infectieuses

Les de The Mimiviridae virophages

Présentée et publiquement soutenue devant la FACULTE DE MEDECINE de MARSEILLE le 10 décembre 2013

Membres du jury de la thèse :

Pr. Bernard La Scola Directeur de thèse Pr. Jean -Marc Rolain Président du jury Pr. Bruno Pozzetto Rapporteur Dr. Hervé Lecoq Rapporteur

Faculté de Médecine, 13385 Marseille Cedex 05, France URMITE, UM63, CNRS 7278, IRD 198, Inserm 1095 Directeur : Pr. Didier RAOULT Avant-propos

Le format de présentation de cette thèse correspond à une recommandation de la spécialité Maladies Infectieuses et Microbiologie, à l’intérieur du Master des Sciences de la Vie et de la Santé qui dépend de l’Ecole Doctorale des Sciences de la Vie de Marseille. Le candidat est amené à respecter des règles qui lui sont imposées et qui comportent un format de thèse utilisé dans le Nord de l’Europe permettant un meilleur rangement que les thèses traditionnelles. Par ailleurs, la partie introduction et bibliographie est remplacée par une revue envoyée dans un journal afin de permettre une évaluation extérieure de la qualité de la revue et de permettre à l’étudiant de commencer le plus tôt possible une bibliographie exhaustive sur le domaine de cette thèse. Par ailleurs, la thèse est présentée sur article publié, accepté ou soumis associé d’un bref commentaire donnant le sens général du travail. Cette forme de présentation a paru plus en adéquation avec les exigences de la compétition internationale et permet de se concentrer sur des travaux qui béné ficieront d’une diffusion internationale.

Pr. Didier Raoult

 Résumé

Les virophages sont des petits à ADN possédant une capside icosaédrique de 50- 60 nm et un génome de 17 à 26 Kb codant potentiellement pour une vingtaine de gènes. Ils ont été découverts associés à des grands virus à ADN appartenant à l’ordre des 0HJDYLUDOHV , pour lesquels leur présence serait délétère. Le Sputnik est ainsi associé aux Mimiviridae , à Cafeteria roenbergensis virus, et les pourraient également avoir leurs virophages. Il a été suggéré que leur répartition dans l’environnement est largement sous- estimée et qu’ils pourraient y jouer un rôle majeur dans la régulation des populations virales. Si leurs principales caractéristiques sont largement acceptées, leur position dans le monde viral, ainsi que les interactions qu’ils pourraient avoir avec leur virus hôtes sont encore discutées.

Le premier projet du travail de thèse a été de faire le bilan des propriétés connues des virophages au travers d’une revue. Les virus hôtes, la structure de leurs particules, ainsi que leur génome ou encore leurs effets potentiels sur leurs hôtes sont abordés dans ce travail. La deuxième partie correspond à un bilan des avancées en matière d’isolement de virus géants dans les amibes – hôtes naturels des Mimiviridae –, la découverte de virophages étant indissociable de celle de leurs hôtes. La troisième section se focalise sur la réplication des virophages Sputnik avec différents virus parmi les Mimiviridae , lesquels sont répartis en trois groupes – A, B et C – sur la base de leur séquence du gène de la polymérase B. Il a été montré que Sputnik, initialement isolé avec des Mimiviridae du groupe A, pouvait se multiplier en association avec tous les groupes de Mimiviridae , capacité ayant permis l’isolement d’une nouvelle souche de Sputnik sans son hôte natif par l’utilisation d’un Mimiviridae en tant que virus rapporteur. Dans ce travail, l’impact délétère du virophage sur le virus géant qui l’accompagne a également été d émontré à nouveau. La dernière partie est enfin basée sur l’identification d’un nouveau virophage – Zamilon – isolé en association avec un Mimiviridae du groupe C. Proche de Sputnik, ce virophage présente toutefois des caractéristiques propres, notamment une spécificité d’hôtes limitée aux Mimiviridae des groupes B et C. Un gène prédit en particulier pourrait être un acteur majeur de cette spécificité : par analyse phylogénétique, celui-ci rapprocherait Zamilon des Mimiviridae des groupes B et C, quand Sputnik serait rattaché au groupe A.

Les résultats présentés dans cette thèse démontrent une certaine complexité des interactions entre les virophages et leurs hôtes. Au sein d’une même famille d’hôtes, certains virophages possèdent un large spectre de spécificité, alors que d’autres ne peuvent se multiplier qu’avec certains d’entre eux, comme cela a déjà été observé chez les bactériophages. Compte-tenu de leur impact potentiel sur les virus géants, ces résultats soutiennent l’hypothèse d’une régulation des populations virales environnementales par les virophages.

Mots-clés : Mimiviridae , 0HJDYLUDOHV , virophages, Sputnik, amibes.

 Abstract

Virophages are small DNA with a 50-60 nm width icosahedral encompassing a 17 to 26 Kb , putatively coding approximately 20 . They have been discovered in association with large DNA virus belonging to the order of the 0HJDYLUDOHV . The Sputnik virophage is thus associated to Mimiviridae , the Mavirus virophage with Cafeteria roenbergensis virus, and the giant Phycodnaviridae could have their own virophages too. They are thought to be more prevalent in the environment than initially expected, and they could be major players in the regulation of viral populations. Yet, if their mean characteristics are widely recognized, their place in the viral world as well as their interactions they have with their host viruses are still a matter of debate.

The first project of this thesis work was to recapitulate in a review the known features of the virophages. Their hosts, the virion structure, their and their potential impacts on their host viruses are discussed herein. The second part corresponds to a summary of the advances in the field of giant viruses isolation in – the common hosts of Mimiviridae –, for the discovery of virophages is closely related to that of their giant hosts. The third section is focused on the replication of the Sputnik virophages with viruses belonging to the Mimiviridae , which are divided in three lineages – A, B and C – based on their polB sequences. It has been shown that Sputnik, initially isolated with Mimiviridae of the group A, can replicate associated to all the groups of Mimiviridae , feature that allowed isolating a new Sputnik strain with a Mimiviridae reporter instead of with its natural viral host. In this work, the detrimental impact of the virophage over its host was confirmed. Finally, the last part is based on the identification of a new virophage – Zamilon – isolated in combination with a group C Mimiviridae . Closely related to Sputnik, this virophage nevertheless has unique features, particularly a host specificity limited to viruses from the groups B and C of the Mimiviridae . One particular predicted gene could be an important actor of this specificity: its phylogenetic analysis clusters Zamilon more related to groups B and C Mimiviridae than to the group A and Sputnik virophages.

The results described herein show the complexity of the interactions between virophages and their giant hosts viruses. Within the same host family, some virophages have a broad-range host spectrum whereas others are limited to some viruses, a feature already described for . Regarding the potential impact of the virophages over their host viruses, these results support the hypothesis of a virophages’ major role in a regulation of viral populations in environment.

Keywords : Mimiviridae , 0HJDYLUDOHV , virophages, Sputnik, amoebae.

 Remerciements

Je tiens à remercier le Professeur Didier Raoult de m’avoir accueilli au sein de l’URMITE, me permettant de réaliser mon projet dans des conditions stimulantes. Mes remerciements vont également au Professeur Bernard La Scola pour m’avoir offert l’opportunité d’intégrer son équipe. Durant mes nombreuses années ici, j’ai pu grandir et apprendre beaucoup plus que je ne l’aurais cru.

Je tiens à exprimer toute ma gratitude envers le Professeur Jean-Marc Rolain, qui a accepté de présider le jury de cette thèse, ainsi qu’envers les rapporteurs dont la participation dans l’évaluation de mon travail m’honore.

Merci à Lina, qui m’a montré qu’il était possible de conserver sa passion sans concession, et qui m’a toujours soutenu. J’ai appris énormément de choses grâce à toi, tu resteras ma marraine de cœur.

Les mots, et la place, ne suffiraient pas à exprimer toute ma reconnaissance envers tous ceux qui m’ont permis d’en arriv er là, à travers les bons moments comme les mauvais. Merci à ma famille et à ceux qui en font partie à mes yeux pour leur soutien de chaque instant. Merci à celles et ceux qui par leurs encouragements et leur présence me permettent aujourd’hui d’écrire ces lignes.

De trop nombreuses personnes ont marqué ma thèse par bien des façons, ponctuellement ou dans la durée, et toutes les lister serait impossible, aussi j’espère que ceux - ci comprendront en lisant ceci que même s’il n’y pas d’encre pour en témoigner, je les remercie sincèrement.

 Table des matières

Avant-propos  Résumé  Abstract  Remerciements  Table des matières  Liste des abréviations 

Chapitre Un – Introduction  1.1 – Les Megavirales  1.2 – Les virophages  1.3 – Objectifs de thèse 

Chapitre Deux – Le concept de virophage  2.1 – Revue de la littérature : le concept de virophage  Virophage Concept, The 

Chapitre Trois – L’isolement des virus géants d’amibes  3.1 – Article : une décennie d’avancées dans l’isolement de  Mimiviridae et de dans les amibes A Decade of Improvements in Mimiviridae and Marseilleviridae  Isolation from

Chapitre Quatre – La spécificité d’hôtes des virophages Sputnik  4.1 – Article : le large spectre d’hôtes des virophages de  Mimiviridae permet leur isolement en utilisant un rapporteur

 Broad Spectrum of Mimiviridae Virophage Allows Its Isolation  Using a Mimivirus Reporter

Chapitre Cinq – Le virophage Zamilon  5.1 – Article : une séquence d’un nouveau virophage associée à  la spécificité d’hôtes Mimiviridae A Novel Virophage Sequence is Associated With Mimiviridae  Host Specificity

Chapitre Six – Conclusions et perspectives 

Bibliographie ǂ 

ǂ La section « Bibliographie » liste les références des articles cités dans les chapitres Un et Six. Les références concernant les autres chapitres sont détaillées à la fin de chaque article correspondant.

 Liste des abréviations

ADN : Acide désoxyribonucléique APM : polyphaga mimivirus ARN : Acide ribonucléique Kb : kilo-bases Mb : méga-bases NCLDV : grand virus à ADN (Nucleo-Cytoplasmic Large DNA Viruses) nm : nanomètres OLV : Organic Lake virophage PCR : Réaction de polymérisation en chaine (Polymerase Chain Reaction) YLV : Yellowstone Lake virophage

 Chapitre Un

Introduction

Les virophages identifiés jusqu’à présent sont tous associés à des virus géants. Comme les virus satellites, ils sont incapables de se multiplier sans co- infection. Les relations qu’ils partagent avec leurs virus dits hôtes sont toutefois plus complexes, puisque la présence d’un virophage pourrait être délétère pour le virus hôte. Au vu de ces interactions, et compte tenu du fait que la découverte des virophages est intimement liée à celles des virus géants, il convient de faire un bref bilan de la composit ion de l’ordre des Megavirales, regroupant ces virus géants à ADN, avant d’évoquer la découverte des virophages pour finalement en venir aux objectifs de la thèse présentée ici.

1.1 – Les Megavirales

Découvert en 2003, Acanthamoeba polyphaga Mimivirus (APM) a révolutionné la vision du monde viral [1]. Avec une capside à symétrie icosaédrique d’un diamètre d’environ 500 nm entourée de fibrilles longues de 150 nm, ce virus d’amibes est plus grand que certaines bactéries, et a d’ailleurs été considéré à tort comme étant un bacille Gram négatif pendant près d’une décennie. Son matériel génétique n’est pas en reste, puisqu’avec un génome à ADN double- brin d’environ 1,2 Mb présentant 1018 gènes – dont certains codant des ARNs de transfert –, il fait figure de géant dans la virosphère [2,3]. Suite à cette découverte, d’autres souches de Mimivirus furent isolées, identifiées, et regroupées au sein de la famille nouvellement constituée des Mimiviridae [4,5] . Cette famille s’inscrit dans le groupe des grands virus à ADN, les NCLDVs (Nucleo-Cytoplasmic Large DNA Virus) [6 – 10]. Ce groupe inclut également les , infectant des insectes et des vertébrés [11], les Asfarviridae , infectant le porc [12], les , infectant des invertébrés et des vertébrés poïkilothermes [13], les , infectant des insectes [14], les Phycodnaviridae , qui infectent des algues [15,16], et depuis peu les Marseilleviridae qui infectent, comme les Mimiviridae , les amibes du genre Acanthamoeba [17,18]. Il a été proposé de remplacer l ’appellation NCLDV par un ordre, celui des Megavirales, reflétant

 davantage la réalité des homologies génomiques entre ses différents membres [19,20]. Les derniers entrants dans cet ordre sont les récemment décrits , qui avec une taille avoisinant le micromètre et un génome de 2,5 Mb, repoussent encore une fois les limites de la définition couramment admise des virus [21].

1.2 – Les virophages

L’impact important de Mimivirus au sein de la communauté scientifique a été renouvelé quelques années plus tard, quand une autre découverte majeure eut lieu dans le sillage de celle d’APM. En 2008 a en effet été isolé un petit virus associé à une souche de Mimivirus nouvellement identifiée [22] . Ce petit virus, disposant d’une capside icosaédrique de 50 nm de diamètre et d’un ADN double -brin de 18 Kb, fut dénommé Sputnik [2,23]. Incapable de se répliquer dans les amibes si celles-ci ne sont pas infectées par un Mimiviridae , sa nature le rapproche des virus satellites. Néanmoins, l’influence de Sputnik sur le virus géant qu’il accompagne l’écarte de cette classification. En effet, Sputnik se multiplie dans l’usine à virus – compartiment cytoplasmique formé dans l’amibe consécutivement à l’infection par un Mimiviridae et dans lequel il se réplique – du virus géant, et des co- infections d’amibes par Mimivirus et Sputnik ont montré que la présence du petit virus n’était pas sans conséquence pour le virus géant. Le taux de particules virales anormales et abortives de Mimivirus est ainsi significativement augmenté, et le virus géant perd également en infectivité. Le virus géant apparait alors comme étant un virus hôte : il a été proposé de classifier Sputnik comme étant un « virophage », par analogie avec les bactériophages [24 – 26].

Malgré le vif débat encore en cours au sujet de cette classification [27 –30] , d’autres virophages furent identifiés, comme Mavirus qui a été isolé en association avec Cafeteria roenbergensis virus, virus géant parmi les 0HJDYLUDOHV qui infecte un protozoaire bi-flagellé marin [31,32]. La construction de génomes de virophages à partir de données métagénomiques de divers environnements aquatiques – un lac méromictique de l’Antarctique pour Organic Lake virophage [33], et un lac du parc Yellowstone pour Yellowstone Lake virophage [34] – semble également indiquer une prévalence plus importante dans ces milieux qu’initialement envisagé. Un hypo thétique rôle des virophages dans

 l’équilibre écologique de ces environnements est d’ailleurs envisageable, avec une influence sur les populations virales [35,36].

1.3 – Objectifs de thèse

Afin d’éclairer au mieux la nature des virophages, le premier projet de cette thèse à été de faire le bilan de leurs caractéristiques au sein d’une revue de la littérature, pouvant servir d’introduction avancée au concept de virophage.

La deuxième partie correspond au bilan des résultats obtenus par l’ensemble de l’équipe travaillant sur les virus géants dans les amibes. Ceci fait suite à l’objectif continu de chacun d’améliorer les techniques et approches basées sur la co -culture d’amibes, permett ant d’isoler des virus géants, et donc des virophages.

Le troisème projet de la thèse a été d’étudier les capacités réplicatives des virophages Sputnik avec différentes souches de Mimiviridae . Les Mimiviridae sont en effet divisés en trois lignées – les groupes A, B et C – sur la base de la comparaison de leurs séquences du gène codant pour la polymérase B. Les deux souches de Sputnik alors disponibles avaient été auparavant isolées en association avec des virus faisant partie du groupe A. Ce travail a été réalisé en se basant sur des co- infections sur des cultures d’amibes, des PCRs en temps réel, et des observations en microscopie électronique. Les résultats obtenus, démontrant un large spectre d’hôtes, ont permis d’élargir les objectifs à l’isolement d’un e nouvelle souche du virophage Sputnik sans son virus hôte, en utilisant un Mimiviridae en tant que virus rapporteur. Le dernier objectif de cette partie a été de confirmer, de manière visible si possible, l’impact du virophage sur son virus hôte.

La dernière partie est basée sur l’identification d’un nouveau virophage – Zamilon – associé à un Mimiviridae du groupe C. Les objectifs étaient donc de caractériser ce virophage via des co-infections avec les trois groupes de Mimiviridae , des PCRs en temps réel et des observations en microscopie électronique. Une fois le génome entièrement séquencé, ce travail s’est poursuivi par son analyse de façon plus poussée afin de déterminer des indices concernant l’origine de la spécificité d’hôtes limitée de Zamilon.

 Chapitre Deux

Le concept de virophage

 2.1 – Revue de la littérature : le concept de virophage

Virophage Concept, The. Morgan Gaia, Philippe Colson, Christelle Desnues, Bernard La Scola.

Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes (URMITE), Aix Marseille Université, Marseille cedex, France

Ǧ

Published online: May 2013 In: eLS. John Wiley & Sons, Ltd: Chichester.

DOI: 10.1002/9780470015902.a0024410

Erratum : une erreur, signalée à l’éditeur, est présente sur l’article disponible. A la page  , à la place de “However, one such virus is the PRD1 –adenovirus lineage group viruses with different recognised – and accepted – cycles, and thus this membership does not threaten the virophage concept. ”, veuillez considérer “However, the PRD1 – adenovirus lineage groups viruses with different recognised – and accepted – life cycles, and thus this membership does not threaten the virophage concept.”

 Avant-propos

Le premier projet de cette thèse correspond à une approche globale des connaissances actuelles concernant les virophages, regroupées dans une revue de la littérature. Leur hôte, leur génome, la structure de leur capside sont ainsi détaillés, de même que leur influence potentielle sur les virus géants. Un comparatif y est également réalisé avec les virus satellites afin de mettre en exergue les caractéristiques qui leur sont propres. Cette revue permet donc de mieux cerner les virophages et le concept qu’ils sous-tendent, et a fait l’objet d’une publication dans l’ Encyclopedia of Life Sciences (eLS) de Wiley & Sons. Elle permet également de mieux appréhender les travaux décrits dans les chapitres suivants et leurs apports.

 Virophage Concept, The Advanced article

Morgan Gaia, Unite´ de Recherche sur les Maladies Infectieuses et Tropicales Emergentes Article Contents . (URMITE), Aix-Marseille Universite´, Marseille cedex, France Introduction . From Giant Viruses to Small Virophages ´ Philippe Colson, Unite de Recherche sur les Maladies Infectieuses et Tropicales Emergentes . Structure of the Virophage

(URMITE), Aix-Marseille Universite´, Marseille cedex, France . The Life Cycle of the Virophage Christelle Desnues, Unite´ de Recherche sur les Maladies Infectieuses et Tropicales Emergentes . Genomic . (URMITE), Aix-Marseille Universite´, Marseille cedex, France Comparison with Viruses

th Bernard La Scola, Unite´ de Recherche sur les Maladies Infectieuses et Tropicales Emergentes Online posting date: 15 May 2013 (URMITE), Aix-Marseille Universite´, Marseille cedex, France

The existence of small viruses that depend on the coinfec- families distributed in nine orders, for a total of approxi- tion of their host cells by another virus has been known for mately 2500 species. Some viruses are not only dependent a long time. These viruses are considered to be satellites of on the cellular machinery of their host but also on the their helper viruses. The discovery of Sputnik, Mavirus and presence of another virus. These viruses used to be organ- ised into another group, independent of the rest of the the Organic Lake virophage shook this vision of subviral virosphere: the subviral agents. This group includes the agents. Indeed, these new viruses are not only dependent , , virus-dependent nucleic acids and satellite on but are also noxious for their so-called host viruses. The latter are small viruses encoding their own viruses, leading to sick particles. This surprising capability coating but lacking some of the genes necessary for established the concept of a virophage, which is a virus their replication, thus depending on the coinfection of their infecting a virus. Studies on the morphology, life cycle and host with another virus, known as the . The genomes of virophages have enriched this concept by satellite virus group includes the Mimivirus-associated highlighting unexpected features. Thus, the question of satellite virus, also called Sputnik, the virophage. the classification of the virophages among the satellite The name ‘virophage’ was given to indicate homology viruses within the virosphere has been brought to light. with the ‘’, as the viruses they are associated with not only help them to replicate but also appear to be their target, introducing the concept of ‘viruses of viruses’ (La Scola et al ., 2008) and the notion of ‘host viruses’. Introduction According to the definition, virophages have no nuclear phase during their infection cycle but multiply instead in The virosphere, the group of all viruses, has been under the virion factory of their host virus, leading to a reduction debate for decades not only for its possible links with the in the production and infectivity of the host virus with the three-branch tree of life (the Eukarya, the and appearance of abortive particles, and depend on enzymes the ) but also regarding its own internal structure. from their host virus rather than from the infected The virosphere is usually arranged according to different (Claverie and Abergel, 2009). characteristics, such as the nature of the cell hosts – for Although the features of the virophages clearly define example, vertebrate or invertebrate viruses, viruses them as new, never-before-seen viruses, their dependence on and bacterial viruses – and, principally, the nature of their coinfection of their host cell with another virus brings them genomes: deoxyribonucleic acid (DNA) or ribonucleic acid close to satellite viruses. However, even if the virophages (RNA) viruses, single- (negative or positive sense) and share some similarities with satellite viruses, they also exhibit double-stranded, reverse transcribing DNA and RNA some differences. The aim of this review is to highlight the viruses. The Virus : Ninth Report of the Inter- concept of virophages, including all of their features, specific national Committee on Taxonomy of Viruses (ICTV) or not, and to describe them in parallel with their counter- (King et al ., 2012), released in November 2011, shows 94 parts among the satellite viruses to enable the readers to form their own ideas about their correlations. eLS subject area:

How to cite: From Giant Viruses to Small Gaia, Morgan; Colson, Philippe; Desnues, Christelle; and La Scola, Virophages Bernard (May 2013) Virophage Concept, The. In: eLS. John Wiley & Sons, Ltd: Chichester. To better understand the virophages, they should be DOI: 10.1002/9780470015902.a0024410 observed with their host viruses because they are also

eLS & 2013, John Wiley & Sons, Ltd. www.els.net 1

 Virophage Concept, The curiosities within the virosphere. This section includes brief the amoebal host but also mainly from bacteria. It has been descriptions of the virophages described thus far in add- hypothesised that amoeba could be a sort of melting pot of ition to the viruses they depend on. genes from the that can survive within the amoeba (Moreira and Brochier-Armanet, 2008; Raoult Sputnik and acanthamoeba polyphaga and Boyer, 2010; Thomas and Greub, 2010). Mimivirus mimivirus also has characteristic viral features, such as a non- negligible fraction of its ORFs having no homologues in Sputnik was the first, and thus is currently the best any database (ORFans) (Boyer et al ., 2010), and a viral described, virophage, and it is the most represented, with eclipse phase (Raoult et al ., 2004). After entering the three isolation strains that have been reported (La Scola amoeba Acanthamoeba polyphaga, APMV replicates et al ., 2008, 2010; Gaia et al ., in press) and three more within the cytoplasm in a structure that appears a few hours currently under analysis in our laboratory. Sputnik is a after infection (generally at approximately 4 h p.i.), known small icosahedral, nonenveloped virus of approximately as the virion factory, as do the Poxviridae. The newly 50–60 nm in diameter (see Figure 1 ). Its host viruses are formed virions are produced from this viral factory, in giant viruses from the Mimiviridae family, of which which replication and occur. Mimivirus and are the best representatives. Following this discovery, other Mimivirus-like viruses Acanthamoeba polyphaga Mimivirus (APMV) was the first were isolated and identified (La Scola et al ., 2010) as member of this family to be discovered (La Scola et al ., members of the new Mimiviridae family. All of the viruses 2003). Initially thought to be a Gram-positive, amoeba- within this family infect amoeba from the genus Acantha- associated bacterium, the ‘Bradford coccus’ – as it was moeba (species polyphaga and/or castellanii). The Mimi- called – could not be identified by classical bacterial viridae are part of the nucleocytoplasmic large DNA protocols, such as the 16S rRNA sequencing. It was only viruses (NCLDVs) in addition to the Iridoviridae, Asfar- described as a virus more than a decade after its isolation viridae, Ascoviridae, Poxviridae, Phycodnaviridae and (La Scola et al ., 2003; Raoult et al ., 2004). The first iden- (Boyer et al ., 2009). All the NCLDVs shared tification mistake is easily understood by the unusual a set of nine ORFs relatively conserved among them, as characteristics of Mimivirus (for mimicking microbes). well as sets of ORFs shared by some families. See also: This nonenveloped virus has an icosahedral capsid with a Nucleo-cytoplasmic Large DNA Viruses (NCLDV) of diameter of approximately 500 nm surrounded by 150-nm- long fibres, defining it as the largest virus ever observed. Its The first Sputnik was isolated alongside Mamavirus, a genome, a dsDNA of 1.2 Mb, is also larger than some different strain of APMV that is closely related to Mimi- bacterial genomes. With 986 open reading frames (ORFs) virus (albeit slightly larger) by its shape and its genome, (Raoult et al ., 2004; Legendre et al ., 2010), some of which whereas the second strain, Sputnik2, was isolated with potentially encode components of the translational Lentillevirus, another APMV strain (La Scola et al ., 2010). apparatus (a surprise as viruses are thought to be entirely Finally, Sputnik3 was isolated without its host virus, using dependent on the host machinery), APMV is also one of the a new protocol based on Mamavirus-infected amoeba most complex viruses. Many of the Mimivirus genes with an coculture, but can grow – as do Sputnik and Sputnik2 – inferred function (24% of the total) (Colson and Raoult, with almost any of the Mimiviridae (though at different 2010) are supposed to have a foreign origin, not only from levels) (Gaia et al ., in press).

(a) (b)

Figure 1 Sputnik particles observed by TEM (a, scale bar 200 nm; b, scale bar 500 nm).

2 eLS & 2013, John Wiley & Sons, Ltd. www.els.net

 Virophage Concept, The

Table 1 Gene homologies between Sputnik, Mavirus and Organic Lake virophage (OLV) and their hypothetical functions Percentage Percentage Percentage identity identity identity between between between Sputnik (S) S and M Mavirus (M) M and O OLV (O) O and S L3: putative DNA 33.64 MV15: FtsK-HerA 35.45 OLV4: hypothetical 39.11 packaging protein family ATPase protein R6: hypothetical protein OLV13: collagen-like 43.42 protein R9: hypothetical protein 32.76 MV16: putative 25.6 OLV7: hypothetical 34.91 cysteine protease protein R13: putative DNA OLV25: putative DNA 32.35 replication protein primase/polymerase R18: putative minor virion No match MV17: hypothetical 31.65 OLV8: putative minor 26.03 protein protein capsid protein R20: putative major capsid 26.52 MV18: putative major 23.39 OLV9: putative major 26.64 protein capsid protein capsid protein R21: hypothetical protein No match MV14: hypothetical 23.37 OLV5: hypothetical 42.31 protein protein MV6: hypothetical 44.44 OLV1: hypothetical protein protein MV13: hypothetical 25 OLV12: hypothetical protein protein

Best homologies using basic local alignment search tool for protein (BLASTp). The data in the table concerned alignment with e-value lower than 0.001 only.

Mavirus and Cafeteria roenbergensis virus samples from Organic Lake, a hypersaline meromictic lake (i.e. in which the layers of water remain unmixed for years, Mavirus was isolated and identified more recently (Fischer decades or more) in Antarctica. This circular genome of and Suttle, 2011) and is still under investigation. This vir- 26 kb (compared with the 18 kb of the Sputnik genome, ophage exhibits 60-nm-diameter spherical virions and also circular) encodes 26 predicted , some having multiplies only with the Cafeteria roenbergensis virus homologues in the Sputnik proteins (notably the major (CroV). This , isolated nearly 20 years ago capsid protein (MCP)) with 27–42% amino acid identities (Garza and Suttle, 1995), has an isometric capsid of ( ). Added to these arguments, Sputnik-like particles  Table 1 300 nm in diameter and infects the marine zooplankton (i.e. spherical 50-nm diameter particles) were observed by C. roenbergensis , a heterotrophic flagellate. Thus, CroV is transmission electron microscopy (TEM). smaller than Mimivirus, which is also true at the genomic Similarly, the near-complete genome of a phycodnavirus level with a dsDNA genome of 730 kb (544 ORFs) (Fischer (algae virus) was obtained from these samples, and par- et al ., 2010). However, phylogenetic analyses showed that ticles with characteristic morphological features of phy- CroV is actually a new subfamily of the Mimiviridae codnaviruses were observed by TEM (particle diameters of (Fischer et al ., 2010; Colson et al ., 2011) and thus a member 150–200 nm), suggesting that OLV could be the virophage of the NCLDV. Indeed, approximately 30% of the CroV of a member of the Phycodnaviridae. This family, which genome has homologues in the Mimivirus genome. As for includes , , , Prym- Mimiviridae, CroV possesses potential genes that are usu- nesiovirus, and , belongs to the ally absent in viral genomes, such as the predicted proteins NCLDV, as do the Mimiviridae and CroV. The obser- implicated in DNA replication, transcription, vation that all of the host viruses of the virophages and repair. CroV appears to share another feature with described thus far are members of the NCLDV group Mimivirus, as a part of its genome is also thought to have a suggests that the virophages require proteins that are bacterial origin. expressed only in these giant viruses.

Organic Lake virophage (OLV) The case of the OLV is slightly different from the other Structure of the Virophage virophages as it has not been isolated but only detected by (Yau et al ., 2011). Indeed, the OLV genome Structural studies of the virophages have only been con- was constructed from metaproteogenomic analysis on ducted on Sputnik, and thus this section discusses this

eLS & 2013, John Wiley & Sons, Ltd. www.els.net 3

 Virophage Concept, The

T=27 icosahedral capsid 260 pseudohexameric capsomers 12 pentameric capsomers

Three parallel density layers of approximately 2.5 nm-thick that could be The capsomers are approximately 7.5 nm semiordered or ordered dsDNA in thickness and 7.4-8.5 nm in diameter

The pseudohexameric capsomers are formed by the trimerization of the MCP (encoded by ORF20)

5.5 −10 nm-long mushroom-like fibers protrude from the center of each Each pentameric capsomer, formed from pseudohexameric capsomer the minor capsid protein, has a central cavity that could be a gate for DNA exit

50 −60 nm

Figure 2 Diagram summarising the general structure of the Sputnik particle. virophage only. However, even if structural studies of approximately 3.6 Â 10 4nm 3 (thus containing the 18 343 bp Mavirus are still ongoing, the initial sequence analysis genome of Sputnik, with a predicted density of suggests that it presents a similar profile (Fischer and 1.966 Â 10 23nm 3/bp) (Sun et al ., 2010). A recent cryo-EM Suttle, 2011). study with a higher resolution showed, in fact, that Sputnik lacks a viral membrane (Zhang et al ., 2012) and harbours three parallel density layers of approximately 2.5 nm in General structure thickness inside the capsid. These density layers are prob- The three-dimensional reconstruction of Sputnik was ably (partially) ordered dsDNA. Figure 2 summarises the performed using cryo-EM, showing that its MCP and its general structure of Sputnik. minor capsid protein (produced from ORF18 and 19) are organised in a hexagonal surface lattice to form the capsid (Sun et al ., 2010; Zhang et al ., 2012). The MCP of Sputnik The double appears to be encoded by ORF 20 and represents the most abundant protein in the particles (La Scola et al ., 2008). All the data suggest that the MCP of Sputnik has a double The hexagonal lattice has a T=27 triangulation number jelly roll fold, a characteristic structure of the bacteriophage (h=3 and k=3) and is composed of 260 pseudohexameric PRD1–adenovirus lineage (Sun et al ., 2010). This fold, and 12 pentameric capsomers. Each of the pseudohexa- formed from two consecutive antiparallel b barrels (each meric capsomer is formed by trimerisation of the MCP and composed of eight b strands), is shared by all the members of has a thickness and diameter of approximately 7.5 nm. this lineage, which groups structurally related viruses Fibres with a triangular head and a length of 5.5–10 nm infecting all domains of life (Bamford et al ., 2005; Krupovic protrude from the centre of each pseudohexameric cap- and Bamford, 2008), including human adenovirus ( Adeno- somer without a clear function (they have been hypothe- viridae family, infecting the Eukarya), the bacteriophage sised to help to stabilise the capsomers or to play a role in tectiviruses PRD1 ( Tectiviridae family) and PM2 ( Corti- the adhesion of the Sputnik particles to the giant viruses for coviridae family), Sulfolobus turreted icosahedral virus (not their entry in the cell). Instead of protrusions, pentameric assigned to a family, infecting the Archaea) and the capsomers, formed from the minor capsid protein, have Eukarya-infecting NCLDV. All these viruses present an central cavities that could serve as a portal for DNA entry icosahedral capsid surrounding a linear or circular dsDNA or exit (Zauberman et al ., 2008; Xiao et al ., 2009), as has genome. This icosahedral capsid is formed by triangular been previously observed for some bacteriophages (Cher- plates composed of capsomers arrayed in a hexagonal rier et al ., 2009). An initial cryo-EM suggested the presence shape, which is obtained by the trimerisation of the MCP. of a lipid bilayer inside the capsid, enclosing a volume of The capsomers are approximately 7.5 nm in thickness and

4 eLS & 2013, John Wiley & Sons, Ltd. www.els.net

 Virophage Concept, The

7.4–8.5 nm in diameter. Sputnik, as described above, fits this particle without these fibres cannot propagate the Sputnik description. virophage (Boyer et al ., 2011). Serial subcultures of This lineage classification regrouped the viruses Mimivirus led to the emergence of a bald form of the virus according to their viral capsid proteins, virion structure that lacked the surface fibres associated with a dramatic and even the nature of the genome (Bamford et al ., 2005). genome reduction involving genes encoding the proteins of Thus, members of a lineage could be very different as these fibres. sequence similarities, hosts or replication cycles are not Once inside the amoeba, the initial journey of Sputnik considered. For example, PRD1 has an internal membrane and Mimivirus remains to be identified. It has been sug- enclosing its genome whereas adenoviruses – and Sputnik gested that the release of the Sputnik genome could be – does not have such a membrane inside the capsid (Zhang associated to pH reduction or another stress resulting in et al ., 2012). This classification relies on the idea that viral penton loss (Zhang et al ., 2012). In contrast, the release of coat proteins are precisely what define a virus as a member Mimivirus genome is now well understood – its internal of the virosphere and not as any other infective particles viral membrane merges with the endocytic membrane, with (e.g. satellite nucleic acids). Then, these proteins should be an opening of the vertex of the gate for DNA exit, dubbed transmitted vertically, and thus members of the same lin- the ‘stargate’ (Zauberman et al ., 2008). A few hours fol- eage should have a common ancestor (Krupovic and lowing coinfection, replications of APMV and Sputnik Bamford, 2008). However, it is also conceivable that these occur in the viral factory, a dense cytoplasmic region sep- coat proteins, if they represent an advantage, have been arate from the nucleus of the amoeba. This viral factory, independently acquired in different viruses. It is important observed by Hemacolor # staining or immunofluorescence to note that although Sputnik (and certainly Mavirus and (IFF) with DAPI (4 ’,6 ’-diamidino-2-phenylindole, which OLV, given that the capsid proteins of these three vir- stains nucleic acids) after 4 h postinfection (p.i.), is often ophages are related to each other) appears to be a member observed in the NCLDV group and can also be obtained of the PRD1 – adenovirus lineage with respect to its without the virophage. After an eclipse phase, Sputnik structure, no significant similarity was observed between particles are detected at one pole of the factory between 6 h the sequence of the MCP of Sputnik and those of other p.i. and 8 h p.i., followed by the APMV particles, as members of this lineage (Sun et al ., 2010). observed by IFF using mouse anti-Sputnik and anti- Mimivirus antibodies (La Scola et al ., 2008). At 16 h p.i., all of the infected cells are filled with Sputnik and APMV The Life Cycle of the Virophage particles, and at 24 h p.i., more than two-thirds of the coinfected cells have been lysed, thus liberating all the new, The life cycle of the virophage, including its replication and ready-to-infect virophage particles. TEM performed on a its effects on the host virus, has only been studied in Sputnik culture of Cafeteria roenbergensis infected by CroV and with Mimivirus or Mamavirus, which will be referred to Mavirus shows CroV and Mavirus particles outgoing from herein as APMV for either of the two strains. the virion factory at 24 h p.i., consistent with the idea that Sputnik and Mavirus belong to the same group of viruses Intracellular localisation (Fischer and Suttle, 2011; see Figure 3). The mechanism by which the virophage enters the infected Effects on host virus and host cells cell remains unknown. It had been hypothesised that Sputnik could be cointernalised with its host virus (Des- The coinfection of amoeba by a giant Mimiviridae and its nues and Raoult, 2010). Mimivirus enters macrophages by virophage leads to several consequences, not only for the phagocytosis (Ghigo et al ., 2008), a mechanism usually cell hosting all the viral production but also for the giant used by amoeba, the classic cell hosts of Mimiviridae, to virus (justifying their ‘host’ denomination) (La Scola et al ., feed on bacteria larger than 0.4–0.5 mm. Thus, Mimiviridae 2008). Indeed, Sputnik coinfection is associated with a are likely to enter their host by this relatively nonspecific decrease of approximately 70% in the yield of infective method. As observed many times by TEM, Sputnik par- Mamavirus particles. Moreover, there is a 13% decrease in ticles appear to attach to the APMV surface; APMV strains amoeba lysis at 24 h p.i. when Sputnik is present with are covered by a forest of  150-nm-long fibre composed of Mamavirus (79% of the cells lysed vs 92% in the absence of glycosylated proteins (Kuznetsov et al ., 2010). These pro- Sputnik). Even if Sputnik can replicate in any of the APMV tein-rich fibres, rather than solely protecting the virus, strains, different rates of the decrease in infectivity and could also provide APMV with a bacteria-like surface amoeba lysis are obtained for each strain as the growth of constitution (APMV can be stained by both Gram and the virophage is not the same from one host virus to Gimenez stains, usually used to stain bacteria), allowing its another (Gaia et al ., in press). Sputnik not only impacts the internalisation by amoeba (La Scola et al ., 2003; Raoult infectivity and cytopathogenic effect of the Mamavirus but et al ., 2004). The Sputnik particles may bind to the surface also induces deformity of the virions (La Scola et al ., 2008; of APMV and thereby benefit from this feature of Mimi- see Figure 4 ). Instead, the presence of Sputnik leads to an viridae to enter the cell at the same time. As a supporting approximately 10% increase in the production of abnor- argument for this hypothesis, we observed that an APMV mal Mamavirus particles: the virions can have a capsid

eLS & 2013, John Wiley & Sons, Ltd. www.els.net 5

 Virophage Concept, The

6

5 1 3

2 VF

4

N

(a) (b)

DAPI IF sputnik Merge DAPI IF sputnik Merge (c) (d)

Figure 3 The intracellular life cycle of Sputnik: (a) Scheme summarising the life cycle of Sputnik, with the different steps of coinfection of an amoeba together with the host virus. In (1), the giant virus and the virophage enter the amoeba, and then are internalised in vacuoles (2). The virion factory (VF) appears in (3), followed by the production of virophage’s particles (4) and then by the particles of the giant virus (5). Finally, the amoeba is lysed and the two viruses are liberated (6). ‘N’ refers to the nucleus. (b) TEM of steps 4 and 5, (c) and (d) IFF (performed using mouse anti-Sputnik and anti-Mimivirus antibodies and DAPI staining) of steps 4 and 5 with Mimivirus (c) and Courdo11 (d), another Mimiviridae strain. layer with a thickness of approximately 240 nm compared organisms’ (REO). In fact, according to the recent inputs with the 40-nm-thickness of the normal particle capsid. from the Mimivirus genome to the general understanding Other particles present an accumulation of several capsid of viruses, the absence of genes encoding ribosomal pro- layers at one pole in an asymmetrical onion shape, as teins appears to be the main difference between viruses and observed by TEM, with occasional fibrils on the normal cells. Thus, the concept of a ‘virocell’, an infected cell, side. In a few cases, Sputnik virions are encapsidated within implies the existence of a ‘ribocell’, a noninfected cell, and a the Mamavirus particles. ‘ribovirocell’, a cell with an integrated (lysogenic state of the phage leading to the cohabitation of a CEO with an REO). The observation that Sputnik, and appar- The ‘virocell’ concept ently Mavirus and OLV, hijacks the internal processes of replication of their host viruses – thus creating the The nature of the virophage is consistent with a particular virocells – rather than the machinery of the host cells – the vision of the viral world, based on the ‘virocell’ concept ribocells – adds support for this view. (Forterre, 2010, 2011). This concept describes the cellular phase of viruses as an entity on its own. These viruses modify the internal structures and regular functions of the cells to match their own needs; an example of such an Genomic activity is the formation of the virion factories by APMV infections. This concept relies on the difference between Organization ‘capsid-encoding organisms’ (CEO), that is, viruses, with the virion as the infective phase of the CEO encompassed The genomes of the three strains of Sputnik have been fully by its own coating proteins, and ‘ribosome-encoding sequenced and are available in the GenBank database

6 eLS & 2013, John Wiley & Sons, Ltd. www.els.net

 Virophage Concept, The

without known functions (Table 1). Figure 5 shows the gen- eral organisation of the genomes of Sputnik2, Mavirus and OLV.

Replication and transcription Although the mechanism of genome replication of the (b) virophages is still not known, some elements appear to be the features of these viruses of viruses. All of their genomes contain genes encoding proteins implicated in a DNA replication system. Sputnik ORF13, for example, encodes a minimal domain of a transposon virus-polymerase (TV- pol) fused to a D5-like helicase (Iyer et al ., 2008). The TV- pol family is related to phage T7-like DNA polymerase and (a) (c) likely functions as primases and polymerases. This Sputnik ORF has a homologue in OLV; however, even if the heli- Figure 4 TEM of a normal particle of Mamavirus (a, scale bar 200 nm) case domain is similar, the polymerase domain in OLV is compared with abnormal particles. A Mamavirus particle with an more similar to a protein in a prasinophyte (Yau et al ., accumulation of several capsid layers (b, scale bar 200 nm) and an empty 2011). The Mavirus genome, meanwhile, contains an ORF particle of Mamavirus (c, scale bar 500 nm). (the ORF3) encoding a predicted protein-primed DNA polymerase B (Fischer and Suttle, 2011). Thus, all the vir- ophages have proteins involved in a DNA replication (Sputnik=NC_011132; Sputnik2=JN603369; Sput- system. nik3=JN603370). These strains have circular dsDNA Despite this degree of autonomy, virophages are deeply genomes of 18.434 bp, 18.338 bp and 18.338 bp, respect- dependent on the transcription system of their host viruses, ively, with a 99% identity among each other. Sputnik as shown by the presence of specific promoters and poly- possesses 21 genes encoding proteins (La Scola et al ., 2008), adenylation signals shared by the virophages and their with an overlap of 22 nucleotides between ORF18 and giant virus hosts. Indeed, late promoters observed on ORF19, whereas these two ORFs are merged in the two mRNA transcripts of Mimivirus (Legendre et al ., 2010) are other strains (Gaia et al ., in press). A recent study identified also present upstream of 12 ORFs of Sputnik, and all the a sequencing error in the original Sputnik genome sequence ORFs of Mavirus are preceded by a conserved motif highly and confirmed the fusion between ORF18 and 19 (Zhang similar to a putative late promoter motif in the CroV gen- et al. , 2012). Seventeen ORFs of the genome are located on ome (Fischer et al. , 2010). In addition, the Sputnik genome the positive strand, and 13 predicted genes do not have any exhibits 16 Mimivirus-like hairpin structures (Claverie and homologues in sequence databases (ORFans). The three Abergel, 2009). These structures, created by the pairing of strains of Sputnik have a high A+T content (greater than at least 13 nucleotides due to palindromic sequences, were 70%), similar to the APMV strains (see Figure 5). initially observed at the 3 ’ends of more than 80% of the Mavirus has a circular dsDNA genome of 19.063 bp Mimivirus mature transcripts and are actually poly- encoding 20 ORFs (Fischer and Suttle, 2011), which is adenylation signals. The observation that this signal is available in the GenBank database (NC_015230). With unique to Mimivirus and clearly absent in the amoeba cell regard to Sputnik, its genome has a nearly 70% A+T host – and thus specific to the transcription system of the content. Fourteen ORFs are actually ORFans according to giant virus – added to the significance of the presence of a comparison with current sequence databases. Four pre- these signals in the Sputnik genome (14 out of 16 in inter- dicted genes have homologues in the Sputnik genome – a genic regions and the remaining two in intragenic regions), DNA-pumping ATPase, an endonuclease/zinc-ribbon and is consistent with the hypothesis that the Sputnik genes domain protein, a cysteine protease and the MCP ( Table 1 ). are transcribed by the Mimivirus transcription machinery The OLV has a circular DNA genome of 26.421 bp and therefore inside the virion factory (Claverie and (GenBank access number: HQ704801) with an approxi- Abergel, 2009; Desnues et al ., 2012a, b). mately 70% A+T content (Yau et al ., 2011). It possesses 21 ORFs, seven of which have homologues in the Sputnik Phylogeny genome. The predicted proteins seemingly shared by the two virophages are quite distinct as they share 27–42% The phylogenetic analyses showed diverse origins of the amino acid identities. The MCP is among these predicted predicted genes (La Scola et al ., 2008; Desnues et al ., 2012a, proteins and thus appears to be found in all the virophages b). For example, ORF6, ORF12 and ORF13 are seemingly known thus far. The same ATPase predicted in the Sputnik derived from its APMV hosts, whereas ORF10 may come and Mavirus genomes is also one of the homologues found from archaea viruses and ORF13 from a bacteriophage or in OLV. Sputnik and OLV also share an ORF encoding a transposon (Iyer et al ., 2008). ORF3 is also found (though predicted DNA polymerase/primase and four other ORFs with limited similarity) in all NCLDVs and in many

eLS & 2013, John Wiley & Sons, Ltd. www.els.net 7

 Virophage Concept, The

R1 MV1 R2 MV20 R21 L3 1 R4 MV19 1 R5 MV2 R6 MV3 R20 15 000 MV18 15 000 Sputnik2 R7 Mavirus 18 338 bp 19 063 bp 5000 R18 5000 R8 MV4 L17 MV17 R9 MV5 L16 MV16 MV6 L15 MV7 L10 MV15 MV8 R14 10 000 R11 MV14 10 000 MV9 MV13 MV10 R12 MV11 R13 MV12

OLV26 Sputnik2 OLV1 Total length: 18 338 bp OLV25 OLV2 1 GC content: 27% OLV24 25 000 OLV3 OLV23 GenBank access: JN603369 OLV4 OLV22 OLV5 OLV21 5000 OLV6 Mavirus OLV20 OLV7 Total length: 19 063 bp 20 000 OLV OLV8 GC content: 30% 26421 bp GenBank access: NC_015230 OLV9

OLV19 10 000 Organic Lake Virophage 15 000 OLV10 Total lenght: 26 421 bp OLV18 OLV11 OLV17 OLV12 GC content: 37% OLV13 GenBank access: HQ704801 OLV14 OLV16 OLV15

Figure 5 Illustration of the genomes of Sputnik2, Mavirus and OLV.

bacteriophages. The presence of APMV genes in the observed from Sputnik2 purified-Lentillevirus particles Sputnik genome suggests a inside suggesting that it could integrate inside its giant viral host the coinfected amoeba (La Scola et al ., 2008). genome. Deep sequencing analysis confirmed a provir- According to the genomic analysis, Mavirus appears to ophage form of Sputnik2 that could probably insert in be related to the Maverick or transposable elem- multiple sites (Desnues et al ., 2012a, b). In addition, a novel ents (MP TEs) (Fischer and Suttle, 2011). MP TEs are 9– mobile genetic element called ‘’ was found 22 kb long DNA sequences encoding up to 20 genes and can associated with Lentillevirus, Sputnik2 and other giant be found in many eukaryotes. Mavirus and MP TEs share 7 viruses. The may also integrate into the homologous ORFs, including the predicted protein- genome of giant viruses and then contribute to interviral primed DNA polymerase B (ORF2). Furthermore, gene transfer. Mavirus and MP TEs have similar genome lengths: 19 kb for Mavirus and 15–20 kb for MP TEs. The structures of the genomes also present some similarities. Comparison with Satellite Viruses

Integration in APMV genome Description The identification of Sputnik2 led to another surprise. This Satellite viruses were first described in 1962 (Kassanis 1962) virophage was found together with Lentillevirus (a Mimi- with the satellite tobacco necrosis virus (STNV), although it virus-like strain), isolated from the contact lens fluid of a had been observed previously (Bawden and Pirie, 1942). patient suffering from keratitis (La Scola et al ., 2010; This virus could only multiply in the presence of a larger Desnues et al ., 2012a, b). Sputnik2 presents the same main helper virus, tobacco necrosis virus (TNV), which infects features of Sputnik1 except one major difference with and provokes necrotic lesions. Other satellite viruses respect to its relationship with its host virus. After amoeba were then identified and defined as viruses that are ‘‘unable infection, a spontaneous production of the virophage was to multiply in cells without the assistance of a ‘specific’

8 eLS & 2013, John Wiley & Sons, Ltd. www.els.net

 Virophage Concept, The helper virus, (_), not necessary for the multiplication of approximately 5 kb. The particles of their helpers range the helper virus, and (without) appreciable sequence from 90 to 120 nm in diameter and have an icosahedric homology with the helper virus genome’’ (Murant and capsid without an envelope but surrounded by fibrils. In Mayo, 1982). The last statement about satellite viruses some cases, autonomous replication of AAVs has been from the International Committee on Taxonomy of Viruses , observed without any helper viruses (Meyers et al ., 2000). regarding new knowledge, described these viruses as Finally, the last subgroup, according to the latest ICTV ‘‘subviral agents lacking genes that could encode functions classification, is the Mimivirus-associated satellite viruses, needed for replication and depending on the coinfection of that is, Sputnik and now its relations, Mavirus and OLV. a host cell with a helper virus for their multiplication’’ and The virophage is currently the most complex satellite virus as encoding their own coating-proteins that encapsidate described thus far. Its particles of diameter 50 nm and its their genome, in contrast to satellite nucleic acids (King genome of 18–26 kb, encoding 20 or more proteins, are et al. , 2012). unrivalled among the satellite viruses. Moreover, it is the only satellite virus with a dsDNA genome. Classification The Hepatitis delta virus (HDV) is also commonly con- sidered as a satellite virus, even though it is classified among The satellite viruses are currently classified into five sub- the negative sense ssRNA viruses. Its 40-nm diameter vir- groups. The first subgroup comprises ssRNA viruses ions have an envelope provided by its helper the infecting insects and is represented by the chronic bee- virus (HBV), surrounding a nucleocapsid composed of the paralysis satellite virus (CBPSV). This satellite virus has an RNA genome of HDV and approximately 70 copies of isometric virion with a diameter of 17 nm encapsidating the Hepatitis delta antigen (the only protein encoded by the three fragments of  1100 nucleotides each (Bailey et al. , HDV genome). Its helper virus, which has particles of 1980), and it multiplies only in the presence of the Chronic 40–45 nm in diameter with a 32–36 nm nucleocapsid inside bee-paralysis virus (CBPV, not yet classified). This helper and a 3.2 kbp genome, belongs to the virus, which infects honeybees, has a large range of particle family (reverse transcribing DNA and RNA viruses). shapes and sizes. Most are ellipsoidal with a modal length Despite the HDV dependence on the presence of its helper of 30–65 nm and a width of approximately 20 nm (Ribie ` re HBV, it is not classified as a subviral agent because of its et al. , 2010), but rings, figure of eights, branching rods and specific features. lengths of up to 640 nm have been observed (Bailey and Ball, 1991). Controversy with virophage The second subgroup includes satellite viruses resem- bling STNV and that infect plants. In addition to STNV Regarding the similarity between the so-called virophages and its helper TNV ( Necrovirus TNV), this and satellite viruses, the ICTV regrouped them into the subgroup comprises satellite tobacco mosaic virus (STMV) same group in their last statement, the Ninth Report of the and TMV ( TMV), satellite maize International Committee on Taxonomy of Viruses for white line mosaic virus (SMWLMV) with maize white line Virus Taxonomy (2012). They are also considered to be mosaic virus (Tombusviridae MWLMV) and satellite viruses, inside the satellite group that also includes satellite panicum mosaic virus (SPMV) with Sobemovirus virus-dependent nucleic acids. There is no superior classi- PMV. All of these satellite viruses have small isometric fication as they are simply placed into the subviral category, particles of approximately 17 nm with ssRNA genomes of together with viroids and prions, which is a group without 800–1200 bp, whereas their helpers have different sizes and any relationship with the rest of the virosphere. This clas- shapes. For example, TNV is a spherical virus with a sification is a point of debate as some researchers agree with nonenveloped icosahedral capsid with a diameter of 30 nm it, commenting that the features of Sputnik, Mavirus and (Murant and Mayo, 1982), whereas TMV is a rod-shaped, OLV are not new and can be found in other satellite viruses nonenveloped virus that is approximately 300 nm long with (Krupovic and Cvirkaite-Krupovic, 2011a, b, 2012). Other an average diameter of 20 nm (Dodds, 1998). See also: researchers are not satisfied with the status quo, stating that Tombusviridae the virophages have characteristics that exceed the limits of The extra small virus (XSV) satellite virus represents the the satellite virus definition (Fischer, 2011; Desnues and third subgroup, which also includes the plant-infecting Raoult, 2012). viruses. Its particles are approximately 15 nm in diameter One of the arguments concerns the intracellular local- and coat an  800 bp ssRNA genome. Its helper virus is isation of the virus. The replication of Sputnik occurs in (or Macrobrachium rosenbergii nodavirus (MrNV). Despite in close proximity to) the virion factory produced by the evidence of a clear relationship with , this virus APMV within the cytoplasm of the infected cell, without has not yet been classified (Qian et al ., 2003). any phase in the nucleus of the cell (as it would be expected The fourth subgroup is composed of adeno-associated for a dsDNA virus); this feature is considered to be char- satellite viruses (AAV), which have 20-nm diameter par- acteristic of the virophage concept. However, it has been ticles with a nonenveloped T=1 isometric capsid and are noted that the satellite viruses just follow their helper vir- associated with many human and domestic ade- uses within the cell wherever they might go; for example, noviruses (Hoggan et al ., 1968). Their ssDNA genomes are AAVs replicate in the nucleus of the cell, as do their helpers,

eLS & 2013, John Wiley & Sons, Ltd. www.els.net 9

 Virophage Concept, The and RNA satellite viruses (such as STNV) replicate in the being closer to ‘classic’ autonomous viruses. However, cytoplasm, as do their associated viruses (Kassanis et al ., their dependence on the coinfection of their host cells with 1970). However, considering the nature of the genome, other viruses fits the main point in the definition of a sat- AAVs, with their DNA genomes, are expected to have a ellite virus. Given all the differences and similarities, other phase in the nucleus, and the RNA satellite viruses are also classifications have been made in an attempt to embrace the expected to replicate in the cytoplasm. specificities of the virophages within the same group, such The polyadenylation signals and the promoters shared as that of the CEOs and REOs (see ‘The life cycle of the between the virophages and their host viruses are also virophage’ and ‘The ‘virocell’ concept’). Here, satellite debated because RNA signals appear to be shared between viruses and virophages would be regrouped within a group some viruses and their associated satellite viruses (Murant called ‘viruses of viruses’ (Desnues and Raoult, 2012). and Mayo, 1982). These signals, however, only prove the Krupovic and Cvirkaite-Krupovic (2012) proposed the relative dependence of the virophages/satellite viruses on conservation of the current classification while organising the replication and transcription machineries of their host/ this group using family level taxa. In this view, the satellite helper viruses. In fact, the virophage concept is supported virus group would be regrouped with respect to the nature by the association of these elements with the supposed of their genome. Thus, the virophages would become the intracellular localisation, which together are consistent first dsDNA satellite viruses. As the problem does not with the particular life cycle of the virophages. Moreover, appeared to be regrouping satellite viruses and virophages Sputnik, Mavirus and OLV apparently encode proteins together regardless of the type of classification, semantics involved in a DNA replication system (see ‘Genomic’), are just an expression of the need for a new definition for suggesting that they have a degree of autonomy, in contrast satellites to include the specific features of the virophages, with their recent classification as subviral agents. notably the degree of autonomy and complexity not Another point is the origin of the satellite viruses and observed in subviruses. Until then, Sputnik, Mavirus and virophages. Indeed, the membership of Sputnik (and sup- OLV (and maybe others in the future) remain curiosities of posedly of Mavirus and OLV) in the bacteriophage PRD1 the virosphere, regrouped into a concept, that of the – adenovirus lineage based on the structure of its virions virophage. and the nature and size of its genome predicts a common ancestor. Therefore, the virophages would be related to autonomous viruses from which they may have emerged (Krupovic and Cvirkaite-Krupovic, 2011a, b). However, References one such virus is the PRD1–adenovirus lineage group Bailey L, Ball BV, Carpenter JM and Woods RD (1980) Small viruses with different recognised – and accepted – life virus-like particles in honey bees associated with chronic par- cycles, and thus this membership does not threaten the alysis virus and with a previously undescribed disease. Journal virophage concept. Another interesting observation is the of General Virology 46 : 149–155. absence of significant similarity between the sequences of Bailey L and Ball BV (1991) Honey Bee Pathology, 2nd edn. the virophage MCPs with those of the other members in London, UK: Academic Press. this lineage. Bamford DH, Grimes JM and Stuart DI (2005) What does Last but not least, the effects of the virophages on their structure tell us about virus evolution? Current Opinion in cellular and viral hosts were also used to compare them Structural Biology 15 (6): 655–663. with satellite viruses. Some of the latter actually have an Bawden FC and Pirie NW (1942) A preliminary description of adverse effect on their helper viruses. For example, STNV some of the viruses causing tobacco necrosis. British Journal of in excess during coinfection can reduce the production of Experimental Pathology 23 : 314. TNV to nondetectable levels, and AAV can do the same to Boyer M, Azza S, Barrassi L et al . (2011) Mimivirus shows dra- adenovirus production. Additionally, the cytopathic matic genome reduction after intraamoebal culture. Proceed- 108 effects on the host cell could be altered by the presence of a ings of the National Academy of Sciences of the USA (25): satellite virus. Coinfection of STNV and TNV can produce 10296–10301. fewer and smaller necrotic lesions than infection with TNV Boyer M, Gimenez G, Suzan-Monti M and Raoult D (2010) alone (in some cases, TNV can have an increased infectivity Classification and determination of possible origins of ORFans through analysis of nucleocytoplasmic large DNA viruses. when inoculated with its associate satellite virus) (Kassa- Intervirology 53 (5): 310–320. nis, 1962; Murant and Mayo, 1982). These effects are Boyer M, Yutin N, Pagnier I et al . (2009) Giant Marseillevirus indeed similar to those induced by a virophage on its host, highlights the role of amoebae as a melting pot in emergence of with a major difference. Sputnik–APMV chimeric microorganisms. Proceedings of the National Academy result in the increased formation of abnormal APMV of Sciences of the USA 106(51): 21848–21853. particles, with the virophage occasionally found within the Cherrier MV, Kostyuchenko VA, Xiao C et al . (2009) An icosa- giant virus shell. These ‘diseased’ forms of the host/helper hedral algal virus has a complex unique vertex decorated by a virus have never been observed in satellite virus infections. spike. Proceedings of the National Academy of Sciences of the See also: Satellites USA 106(27): 11085–11089. All these points demonstrate the important differences Claverie JM and Abergel C (2009) Mimivirus and its virophage. between virophages and satellite viruses, with the former Annual Review of Genetics 43 : 49–66.

10 eLS & 2013, John Wiley & Sons, Ltd. www.els.net

 Virophage Concept, The

Colson P, Gimenez G, Boyer M, Fournous G and Raoult D (2011) Krupovic M and Bamford DH (2008) Virus evolution: how far The giant Cafeteria roenbergensis virus that infects a wide- does the double beta-barrel viral lineage extend? Nature spread marine phagocytic is a new member of the fourth Reviews 6(12): 941–948. domain of Life. PLoS One 6(4): e18935. Krupovic M and Cvirkaite-Krupovic V (2011a) Virophages Colson P and Raoult D (2010) Gene repertoire of amoeba-asso- or satellite viruses? Nature Reviews Microbiology 9(11): ciated giant viruses. Intervirology 53 (5): 330–343. 762–763. Desnues C, Boyer M and Raoult D (2012a) Sputnik, a virophage Krupovic M and Cvirkaite-Krupovic V (2011b) Sputnik and infecting the viral domain of life. Advances in Virus Research 82 : Mavirus: not more than satellite viruses. Nature Reviews 63–89. Microbiology 10 : 78. Desnues C, La Scola B, Yutin N et al . (2012b) Provirophages and Krupovic M and Cvirkaite-Krupovic V (2012) Towards a more transpovirons as the diverse of giant viruses. Pro- comprehensive classification of satellite viruses. Nature Reviews ceedings of the National Academy of Sciences of the USA Microbiology 10 : 234. 109(44): 18078–18083. Kuznetsov YG, Xiao C, Sun S et al . (2010) Atomic force micro- Desnues C and Raoult D (2010) Inside the lifestyle of the vir- scopy investigation of the giant mimivirus. Virology 404(1): ophage. Intervirology 53 (5): 293–303. 127–137. Desnues C and Raoult D (2012) Virophages question the existence La Scola B, Audic S, Robert C et al . (2003) A giant virus in of satellites. Nature Reviews Microbiology 10 (3): 234. amoebae. Science 299(5615): 2033. Dodds JA (1998) Satellite tobacco mosaic virus. Annual Review of La Scola B, Campocasso A, N’Dong R et al . (2010) Tentative Phytopathology 36 : 295–310. characterization of new environmental giant viruses by Fischer MG (2011) Sputnik and Mavirus: more than just satellite MALDI-TOF mass spectrometry. Intervirology 53 (3): 344– viruses. Nature Reviews Microbiology 10 (1): 78. 353. Fischer MG, Allen MJ, Wilson WH and Suttle CA (2010) Giant La Scola B, Desnues C, Pagnier I et al . (2008) The viro- virus with a remarkable complement of genes infects marine phage as a unique parasite of the giant mimivirus. Nature 455: zooplankton. Proceedings of the National Academy of Sciences 100–104. of the USA 107(45): 19508–19513. Legendre M, Audic S, Poirot O et al . (2010) mRNA deep Fischer MG and Suttle CA (2011) A virophage at the origin of sequencing reveals 75 new genes and a complex tran- large DNA transposons. Science 332(6026): 231–234. scriptional landscape in Mimivirus. Genome Research 20 (5): Forterre P (2010) Giant viruses: conflicts in revisiting the virus 664–674. concept. Intervirology 53 (5): 362–378. Meyers C, Mane M, Kokorina N, Alam S and Hermonat PL Forterre P (2011) Manipulation of cellular syntheses and the (2000) Ubiquitous human adeno-associated virus type 2 nature of viruses: the virocell concept. Comptes Rendus Chimie autonomously replicates in differentiating keratinocytes of a 14 : 392–399. normal skin model. Virology 272: 338–346. Gaia M, Pagnier I, Campocasso A et al . (in press) Broad spectrum Moreira D and Brochier-Armanet C (2008) Giant viruses, giant of Mimiviridae virophage allows its isolation using a Mimivirus chimeras: the multiple evolutionary histories of Mimivirus reporter. PLoS One genes. BMC Evolutionary Biology 8: 12. Garza A and Suttle CA (1995) Large double-stranded DNA vir- Murant AF and Mayo MA (1982) Satellites of plant viruses. uses which cause the lysis of a marine heterotrophic nano- Annual Review of Phytopathology 20 : 49–70. flagellate (Bodo sp.) occur in natural marine viral communities. Qian D, Shi Z, Zhang S et al . (2003) Extra small virus-like particles Aquatic Microbial Ecology 9: 203–210. (XSV) and nodavirus associated with whitish muscle disease in Ghigo E, Kartenbeck J, Lien P et al . (2008) Ameobal pathogen the giant freshwater prawn, Macrobrachium rosenbergii. Jour- mimivirus infects macrophages through phagocytosis. PLoS nal of Fish Diseases 26 (9): 521–527. Pathogens 4(6): e1000087. Raoult D, Audic S, Robert C et al . (2004) The 1.2-mega- Hoggan MD, Shatkin AJ, Blacklow NR, Koczot F and base genome sequence of Mimivirus. Science 306(5700): Rose JA (1968) Helper-dependent infectious deoxyribonucleic 1344–1350. acid from adenovirus-associated virus. Journal of Virology 2(8): Raoult D and Boyer M (2010) Amoebae as genitors and reservoirs 850–851. of giant viruses. Intervirology 53 (5): 321–329. Iyer LM, Abhiman S and Aravind L (2008) A new family of Ribie` re M, Olivier V and Blanchard P (2010) Chronic bee par- polymerases related to superfamily A DNA polymerases and alysis: a disease and a virus like no other? Journal of Invertebrate T7-like DNA-dependent RNA polymerases. Biology Direct Pathology 103: 120–131. 3: 39. Sun S, La Scola B, Bowman VD et al . (2010) Structural studies of Kassanis B (1962) Properties and behaviour of a virus depending the Sputnik virophage. Journal of Virology 84 : 894–897. for its multiplication on another. Journal of General Micro- Thomas V and Greub G (2010) Amoeba/amoebal symbiont biology 27 : 477–488. genetic transfers: lessons from giant virus neighbours. Inter- Kassanis B, Vince DA and Woods RD (1970) Light and electron virology 53 (5): 254–267. microscopy of cells infected with tobacco necrosis and satellite Xiao C, Kuznetsov YG, Sun S et al . (2009) Structural studies of viruses. Journal of General Virology 7(2): 143–151. the giant mimivirus. PLoS Biology 7(4): e92. King AMQ, Adams MJ, Carstens EB and Lefkowitz EJ Yau S, Lauro FM, DeMaere MZ et al . (2011) Virophage (2012) Virus Taxonomy. Ninth Report of the International control of antarctic algal host-virus dynamics. Proceedings Committee on Taxonomy of Viruses . London: Elsevier of the National Academy of Sciences of the USA 108(15): Academic. 6163–6168.

eLS & 2013, John Wiley & Sons, Ltd. www.els.net 11

 Virophage Concept, The

Zauberman N, Mutsafi Y, Halevy DB et al . (2008) Distinct DNA Web Links exit and packaging portals in the virus Acanthamoeba polyphaga mimivirus. PLoS Biology 6(5): e114. Fleming N (2008) http://www.newscientist.com/article/dn14480- Zhang X, Sun S, Xiang Y et al . (2012) Structure of Sputnik, a first-virophage-could-take-the-fight-to-viruses.html virophage, at 3.5-A ˚ resolution. Proceedings of the National Yong E (2012) http://www.the-scientist.com/?articles.view/ Academy of Sciences of the USA 109(45): 18431–18436. articleNo/32840/title/A-Parasite-s-Parasites/

Further Reading

Pearson H (2008) ‘Virophage’ suggests viruses are alive. Nature 454: 677.

12 eLS & 2013, John Wiley & Sons, Ltd. www.els.net

 Chapitre Trois

L’isolement des virus géants d’amibes

 3.1 – Article : une décennie d’avancées dans l’isolement de Mimiviridae et de Marseilleviridae dans les amibes

A Decade of Improvements in Mimiviridae and Marseilleviridae Isolation from Amoeba. Isabelle Pagnier, Dorine-Gaelle Ikanga Reteno, Hanene Saadi, Mondher Boughalmi, Morgan Gaia, Meriem Slimani, Tatsiana Ngounga, Meriem Bekliz, Philippe Colson, Didier Raoult Bernard La Scola*.

Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes (URMITE), Aix- Marseille Université, Marseille, France.

* Corresponding author: [email protected]

Published October 17, 2013.

Intervirology2013;56:354 –363 DOI: 10.1159/000354556

 Avant-propos

La découverte de Mimivirus en 2003, bien que fortuite, a profondément changé la vision du monde viral. Le protocole qui était alors utilisé visait l’identification d’agents pathogènes d’amibes ressemblant à des légionnelles, en se basant sur la co -culture d’amibes. A la suite de la découverte du virus géant, cette technique initialement destinée à isoler des bactéries a été modifiée pour s’adapter à la recherche de virus. Au fil des années, de plus en plus de Mimiviridae , ainsi que des Marseillevirus , ont p u être isolé. D’une part par un échantillonnage de plus en plus large et varié, mais surtout par les améliorations successives apportées aux protocoles d’isolement des virus géants. Ces techniques sont non seulement à la base de l’isolement des Mimiviridae , mais ont également permis la découverte des virophages qui y sont associés. Le bilan de ces techniques est présenté dans la partie ci-après.

 Intervirology 2013;56:354–363 Published online: October 17, 2013 DOI: 10.1159/000354556

A Decade of Improvements in Mimiviridae and Marseilleviridae Isolation from Amoeba

a a a Isabelle Pagnier Dorine-Gaelle Ikanga Reteno Hanene Saadi a a a a Mondher Boughalmi Morgan Gaia Meriem Slimani Tatsiana Ngounga a a a, b a Meriem Bekliz Philippe Colson Didier Raoult Bernard La Scola a Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes (URMITE), Aix-Marseille Université, b Marseille , France; Special Infectious Agents Unit, King Abdulaziz University, Jeddah, Saudi Arabia

Key Words throughput isolation of new isolates to improve the record Mimivirus · Marseillevirus · · Megavirales · of giant virus distribution in the environment and the deter- Virophage · Acanthamoeba mination of their pangenome. © 2013 S. Karger AG, Basel

Abstract Since the isolation of the first giant virus, the Mimivirus, by Introduction T.J. Rowbotham in a cooling tower in Bradford, UK, and after its characterisation by our group in 2003, we have continued Mimivirus, the first giant virus identified, was isolated to develop novel strategies to isolate additional strains. By from an amoeba in Bradford, UK, in the 1980s during the first focusing on cooling towers using our original time-con- investigation of a pneumonia outbreak by T.J. Rowbo- suming procedure, we were able to isolate a new lineage of tham while he was trying to isolate Legionella -like bacte- giant virus called Marseillevirus and a new Mimivirus strain rial pathogens that infect amoebae [1, 2] . Acanthamoeba called Mamavirus. In the following years, we have accumu- polyphaga Marseillevirus was isolated in France 5 years lated the world’s largest unique collection of giant viruses by later during work to isolate other strains of improving the use of antibiotic combinations to avoid bacte- [3] . Those two viruses are the founding members of two rial contamination of amoeba, developing strategies of pre- new viral families. Members of the two new viral families liminary screening of samples by molecular methods, and are the largest known viruses based on the sizes of their using a high-throughput isolation method developed by our capsid and genome. Mamavirus, another member of the group. Based on the inoculation of nearly 7,000 samples, our Mimiviridae, was later isolated and found to harbour a collection currently contains 43 strains of Mimiviridae (14 in virophage [4] , and other members of the Marseilleviridae lineage A, 6 in lineage B, and 23 in lineage C) and 17 strains have been recovered from the Seine River [5] and from of Marseilleviridae isolated from various environments, in- human stool [6] . The isolation of several strains of the gi- cluding 3 of human origin. This study details the procedures ant viruses made it possible to classify them as members used to build this collection and paves the way for the high- of a new order of Megavirales [7] . This order is divided

© 2013 S. Karger AG, Basel Prof. Bernard La Scola 0300–5526/13/0566–0354$38.00/0 Unité de Recherche sur les Maladies Infectieuses et Tropicales Emergentes (URMITE) UM63, CNRS 7278, IRD 198, Inserm 1095, Faculté de Médecine, Aix-Marseille Université E-Mail [email protected] & is is an Open Access article licensed under the terms of the 27, boulevard Jean-Moulin, FR–13385 Marseille Cedex 05 (France) www.karger.com/int Creative Commons Attribution-NonCommercial 3.0 Un- E-Mail bernard.la-scola @ univ-amu.fr ported license (CC BY-NC) (www.karger.com/OA-license),  applicable to the online version of the article only. Distribu- tion permitted for non-commercial purposes only. PM 7:47:35 - 10/20/2013 88.183.11.77 into two distinct groups. The first group contains the internal organs from the rest of the body. The digestive tract and Mimiviridae family, which is divided into 3 lineages (A, internal organs were crushed together with 3 ml of PAS buffer, and the suspension was homogenised. A solution consisting of 4 anti- B and C), and the Marseilleviridae family [8] . The second biotics (ciprofloxacin at 4 mg/l, vancomycin at 4 mg/l, colimycin is represented by the single Cafeteria roenbergensis virus at 500 IU/l and rifampicin at 4 mg/l) and 1 antifungal (Fungizone CroV, which is found in a unicellular marine biflagellate at 100 mg/l) was added to the suspensions to prevent bacterial and [9] . However, the first three isolates of Megavirales were fungal contamination. The homogenised suspensions were then obtained from fastidious amoeba coculture procedures washed in PAS buffer to remove traces of the antimicrobial solu- tion. The pellets were resuspended in PAS buffer and used for co- on water samples from cooling towers. For many years, culture with amoebae [15] . we have tried to improve our isolation procedures and have tested many biotopes to investigate the distribution C ocultures of giant viruses. We report herein the cumulative results Primary Procedure. The amoebae A. polyphaga (strain Linc AP- of this work, which have enabled the isolation of 18 strains 1) or A. castellanii were cultivated in PYG medium (proteose pep- tone, yeast extract, glucose) and subcultured every 2 days until con- of Marseilleviridae and 45 strains of Mimiviridae from fluent cell monolayers were obtained. For inoculation, a culture of diverse environments, including human samples. amoeba was rinsed in PAS buffer using successive low-speed cen- trifugations at 2,000 rpm for 10 min. The amoebae were counted in counting slides and were adjusted to 5 × 105 amoebae/ml in PAS. Methods The amoebal suspensions were distributed in 12-well microplates and 100 μl of the sample suspensions were inoculated into the wells. Sampling and Preparation of Samples Amoebal microplates were screened with an inverted microscope A variety of samples were used when seeking amoeba-associated to detect amoebal lysis. The cocultures were then subcultured onto giant viruses, including water, soil, insect, stool or human respira- a fresh amoebal microplate suspension and the subcultures were tory samples. The different types of samples analysed for the pres- also screened for amoebal lysis. The lysed cocultures were shaken ence of giant viruses are summarised in figure 1 . In the original pro- to suspend the remaining amoebae, and 100 μl of the suspension cedure, 500- to 1,000-ml water samples stored in sterile bottles were was cytocentrifuged at 800 rpm for 10 min. The slides were stained kept at 4° before processing. Samples were then filtered through a using Gram and Gimenez stains. When the presence of viruses was 0.22-μm-pore filter to concentrate all microorganisms larger than suspected, bacteria were removed from the culture using an appro- this size, and the filters were vortexed in 1–2 ml of sterile Page’s priate antimicrobial agent or by filtration through 0.8-µm-pore fil- amoebal saline (PAS) buffer [1, 3, 10] . Next, to isolate ‘small’ giant ters. The presence of the giant virus was assessed by electron mi- viruses, 0.22-µm-pore membranes were replaced by 0.1-µm-pore croscopic observation of cultures negatively stained with a 1% am- membranes [6] . Other samples from tap water and biofilms were monium molybdate suspension. The giant virus was then measured, sampled using sterile swabs and were directly vortexed in 1–2 ml of and the size enabled primary classification into the Mimiviridae sterile PAS [11] . The latest improvement for simplifying the water group (capsid size of approximately 450 nm) or the Marseilleviri- sampling procedure involved sampling 10–50 ml of water followed dae group (capsid size of approximately 200 nm). Following clas- by a high-speed centrifugation step (15,000 rpm for 10 min). This sification, the single virus amoeba culture was subcultured for end- procedural modification concentrated the microorganisms and was point dilution cloning and further analyses. Nearly all cultures re- faster than the original filtration procedure. quired antibiotics to eliminate bacterial contamination. Due to this For soil samples, each 15- to 100-gram sample was mixed with necessity, the method was improved by the systematic use of anti- 50–150 ml of sterile water. Decantation was performed for 24– biotics in the early stages of the procedure. This primary procedure 48 h at 4° followed by filtration through Whatman paper and then has been described previously [3, 10] . through a 0.1-µm membrane. The membranes were then added to Antibiotic-Directed Procedure. Amoebal microplates were pre- 1 ml of PAS and vortexed following the same procedure as that pared with amoebae under the same conditions as described above. used for the water samples [11–13] . Each sample was inoculated onto amoebal microplates with both The same principle used for the soil samples was applied to the colistin at 500 UI/l and vancomycin at 10 mg/l. The cocultures processing of stool samples. Briefly, 50 ml of water was added to were subcultured onto a fresh amoebal microplate suspension. The approximately 1 g of stool sample, decantation was then carried primary cultures and subcultures were screened daily for cyto- out for 24–48 h at 4°, and the supernatants were filtered through a pathogenic effects using an inverted microscope and 100-µl sam- 0.1-µm membrane. The membranes were then added to 1 ml of ples of resuspended amoebae were cytocentrifuged. The slides PAS and vortexed [6] . were stained with Gimenez and Gram stains followed by addition- For the respiratory samples such as sputum or bronchoalveolar al Hemacolor staining (Merck, Darmstadt, Germany) if obvious lavage, mechanical breakdown of the cells was performed by pass- viral factories inside the amoebae were noticed. On day 7 of the ing the samples through syringes of 0.33 mm diameter (bioMéri- culture, 50 µl of each coculture was systematically subcultured eux, Marcy l’Etoile, France). The samples were then inoculated onto axenic media, buffered charcoal yeast extract agar and Co- directly onto the amoebal monolayer for giant virus research [14] . lumbia sheep blood agar plates to evaluate and eliminate residual Other types of samples, such as samples from insect larvae or bacterial contamination. The antimicrobial susceptibility of the leeches, required more specific treatments. The were first isolates was tested using a disk diffusion assay with gentamicin, rinsed with 96% ethanol to sterilise their exterior, then washed cotrimoxazole, erythromycin, rifampicin, doxycycline and cipro- with sterile PAS and dissected to separate the digestive tract and floxacin. If bacterial overgrowth was moderate, the antibiotics that

Improvements in Mimiviridae and Intervirology 2013;56:354–363 355 Marseilleviridae Isolation from Amoeba DOI: 10.1159/000354556

 Downloaded by: Downloaded PM 7:47:35 - 10/20/2013 88.183.11.77 Fig. 1. A collection of giant viruses isolated in our laboratory from 6,989 samples according to the origin of the samples and grouped by species/genotype.

proved to be effective on the isolated bacteria were added to the at 4 mg/l and amphotericin B at 100 mg/l. Agar plates were pre- subcultures. If bacterial overgrowth was massive and destroyed the pared by adding 15 g of agar (Research Organics, Cleveland, amoeba monolayer, the sample was re-inoculated and supple- Ohio, USA) to a 1-litre solution of PAS medium followed by ster- mented with one or more of the effective antibiotics. In cases in ilisation in an autoclave. The agar medium was supplemented which the bacterial overgrowth was due to bacteria that did not before solidification with the same antibiotic cocktail as de- grow on agar plates, different antibiotics were tested on subcul- scribed above and 50-ml volumes were distributed into square tures or re-inoculations until complete decontamination was Petri dishes (23.5 × 23.5 cm; Dominic Deutscher, Brumath, achieved. When bacterial decontamination was evident, the cul- France). After solidification, the agar was coated with a mono- tures were treated as previously described with the single isolated layer of amoebae diluted to 2 × 106 amoeba/ml, and a drop of virus. This antibiotic-directed procedure has been described previ- viral enrichment supernatant was inoculated on the monolayer. ously [11] . After incubation, as a virus multiplied, a lysis plaque could be High-Throughput Procedure . As described above, the initial visualised with the naked eye ( fig. 2 ). The lysis plaques were mea- enrichment step consisted of inoculating 100 µl of the samples sured, and the agar under the plaque was cut and divided into onto an amoeba monolayer in 12-well microplates without the small pieces, resuspended in 1 ml of PAS, vortexed and filtered addition of antibiotics. A subculture was made after 3 days on a through a 1.2-µm-pore filter before inoculation with fresh amoe- fresh amoebal culture with the addition of antibiotics. The anti- bae in PAS buffer. After this step, single viruses in suspension biotic cocktail used was improved and consisted of ciprofloxacin were treated as previously described. This high-throughput pro- at 4 mg/l, vancomycin at 4 mg/l, colimycin at 500 IU/l, rifampicin cedure has been described previously [12] .

356 Intervirology 2013;56:354–363 Pagnier et al. DOI: 10.1159/000354556

 Downloaded by: Downloaded PM 7:47:35 - 10/20/2013 88.183.11.77 b

Fig. 2. Illustration of plaque lysis using the high-throughput method of stained agar plates according to Gaia et al. [13] . a The cross indicates the sites of inoculation and the squares 1 and 2 indicate the area of lysis for 2 samples, whereas positive controls are located in area 3. Magnification of lysis ob- tained by Mimivirus ( b ) and Marseillevirus a ( c ). c

Additional Enrichment by Blind Subculture Combined with the probe systems. The tentative detection of mimiviruses from group High-Throughput Procedure. This method uses ciprofloxacin at 20 C was performed on 156 soil samples. The primer-probe system mg/l, vancomycin at 10 mg/l, imipenem at 10 mg/l and thiabenda- was designed to target a gene encoding a hypothetical protein from zole at 50 mg/l (Sigma-Aldrich, Saint-Quentin Fallavier, France). the group C Mimiviridae system CE7-1675721 (table 1 ). DNA ex- Thiabendazole was used to avoid fungal contamination, which of- traction was performed as described above, and real-time PCR was ten occurs with soil samples. Thiabendazole replaced amphoteri- performed by adding 5 µl of DNA to an amplification mix contain- cin B because it showed a higher toxicity for amoebae. The cocul- ing 12.5 µl of 2× QuantiTect Probe PCR Master Mix, 0.5 µl of Taq- ture procedure consisted of a primary coculture step using the new Man probe and 0.5 µl (0.2 µmol) of both reverse and forward prim- antibiotic cocktail for 3 days at 32°. After the 3-day antibiotic treat- ers in a final reaction volume of 25 µl. Amplifications were per- ment step, filtration through 0.8-µm-pore membranes was used to formed on a thermocycler Light Cycler (Roche, Meylan Cedex, reduce fungal contamination when thiabendazole treatment was France) with an initial enzyme activation step (95°, 10 min) fol- not sufficient. Plates were then subcultured onto a fresh amoebal lowed by 40 cycles of denaturation (95°, 30 s) and hybridisation/ monolayer with the same antimicrobial agent cocktail with the elongation (60°, 1 min). More recently, we developed primer-probe omission of thiabendazole. An additional enrichment step was systems to extend the preliminary molecular detection to all Mimi- added to the previous procedure by subculturing onto a fresh viridae (groups A, B and C) and Marseilleviridae. DNA extraction amoeba monolayer containing the same antibiotic cocktail. After was performed with the automated extraction system EZ1 Virus this final enrichment step, the culture supernatants were inocu- MiniKit v.2 (Qiagen GmbH) following the manufacturer’s instruc- lated onto amoebal monolayers deposited on agar plates, as de- tions. Real-time PCR was performed by adding 5 µl of DNA to the scribed for the previous high-throughput method. amplification mix containing 12.5 µl of 2× QuantiTect Probe PCR Master Mix, 0.5 µl of TaqMan probe, and 0.5 µl (0.2 µmol) of both Molecular Screening of Samples reverse and forward primers in a final reaction volume of 25 µl. The In several studies, we intended to improve the efficiency of the reactions were run on a CFX96TM thermocycler (BioRad Labora- culture yield by selecting the samples to be inoculated based on an tories Inc., Hercules, Calif., USA) with an enzyme activation step initial molecular screening step. In an initial study [13] , 90 soil sam- (95°, 15 min) followed by 44 cycles of denaturation (95°, 30 s) and ples were screened by standard PCR using primer pairs specifically hybridisation/elongation (60°, 1 min). targeting the three lineages (A, B and C) of Mimiviridae. The prim- ers used were targeted to the beta subunit of the DNA polymerase PCR Primary Characterisation (polB) based on the genome sequencing of Mimivirus and Terra2 As explained above, the initial identification of giant viruses was for group A, Moumouvirus for group B, and Courdo11 and Terra1 first based on the direct observation of the culture suspension. First, for group C (table 1 ). The viral DNA was extracted from the sam- samples stained with Gram or Hemacolor stains were observed un- ples using a QIAGEN© QIAmp Mini Kit (Qiagen GmbH, Hilde, der a light microscope at 100× magnification. Second, negative Germany) according to the manufacturer’s instructions. Standard staining was performed using a 1% solution of ammonium molyb- PCR amplification was performed and the PCR products were vi- date and observation under an electron microscope. Because of the sualised under UV light after migration on an agar gel stained with design and application of the specific primer-probe systems listed ethidium bromide. In a second study, we modified the detection of in table 1 , amplification and sequencing of viral genes allows a more giant viruses using real-time PCR with TaqMan specific primer- precise preliminary classification of the newly isolated viruses into

Improvements in Mimiviridae and Intervirology 2013;56:354–363 357 Marseilleviridae Isolation from Amoeba DOI: 10.1159/000354556

 Downloaded by: Downloaded PM 7:47:35 - 10/20/2013 88.183.11.77 Table 1. Sequences of different primers and probes used to amplify specific genes of Mimiviridae groups A, B and C, and Marseillevirus

Group Primer/probe Sequence (5 –3 ) Viral group Target gene

A Mimi-TJA01F 5 -GCAGCCCTTTGACACTT-3Ԣ Ԣ mimiviridae A polB Mimi-TJA01R 5 -CATGCGGGAGTTGGAGA-3 mimiviridae A Mimi-TJA02F 5Ԣ-GAAAATGGTGAAGAGAAAACTGA-3Ԣ mimiviridae A Mimi-TJA02R 5Ԣ-ACCAGGATAAATGGATGCAA-3Ԣ mimiviridae A CE11-TE1-01F 5Ԣ-AGTTACCCAACCACAAGAAGA-3 Ԣ mimiviridae C CE11-TE1-01R 5Ԣ-CAGAAGGACTAACAAAAGAACCA-3Ԣ mimiviridae C CE11-TE1-01F 5Ԣ-AAAATATTGGGGACGTTGGTG-3Ԣ mimiviridae C CE11-TE1-01R 5Ԣ-ATGGAAGACTGGCTGTTGAAA-3 Ԣ mimiviridae C VA10-01F 5Ԣ-AAGGGGACAAGGAGTTAAAATAT-3Ԣ mimiviridae B VA10-01R 5Ԣ-TAGATATACGTTTGGTTTTGGAGTGA-3Ԣ mimiviridae B Ԣ Ԣ C A865F2 5Ԣ-TGGATACATTGATGGTTGATAA-3 Ԣ mimiviridae A hypothetical A865R1 5 -TTTCGACTTTACACTTGGGATTG-3 mimiviridae A protein A865 Prb2 FAM-TTATGAAAAACCTAATCCAGAAGATT-TAMRAԢ Ԣ mimiviridae A Ԣ Ԣ C groupeB_for 5 -GAGCTATAATTGGGGCAACG-3 mimiviridae B intergenic groupeB_rev 5 -TCTTATTAAAAGATTCCTGTTTGACA-3 mimiviridae B region Mimi_groupeB_FAM_MGB FAM-AATTTATTTAATCCTTTACCAAAACCA-MGBԢ Ԣ mimiviridae B Ԣ Ԣ B/C CE7-1675721FPr2 5 -TAATTTTATATTCAACACCAAGG-3 mimiviridae C hypothetical CE7-1675721RPr1 5 -CCAATGACCTATCGTTGG-3 mimiviridae C protein CE7-1675721 Prb1 FAM-CTTGGTCTAACAACCAAACACTA-TAMRAԢ Ԣ mimiviridae C Ԣ Ԣ C Mars_Fwd1 5 -TCTGGGAGTGGGCTTTATCT-3 marseilleviridae hypothetical Mars_Rev1 5 -AGGGTAATGACCTCGGGTA-3 marseilleviridae protein Mars_Pr1 FAM-AGGATTGAACCTTCGCTGTTAC-TAMRAԢ Ԣ marseilleviridae Ԣ Ԣ

the Mimiviridae groups A, B or C, or Marseilleviridae group. DNA tower [4] . In addition, the first Marseillevirus strain was was extracted from the positive culture samples using the automat- also found in the water of a cooling tower in 2009 [3] . In ed extraction system EZ1 Virus MiniKit v.2 (Qiagen GmbH). Real- 2010, La Scola et al. [11] isolated 3 new strains of Mimi- time PCR was performed as described above using a CFX96TM ther- mocycler (BioRad Laboratories Inc.), and sequencing was conduct- virus (Moumouvirus, Monve and Bus) and 2 new strains ed using the same primers that were used for amplification. of Marseillevirus (Cannes8 and Cannes9) from 39 cool- ing water samples. During the same study, other types of environmental water was investigated, and the authors Results were able to isolate 8 strains of Mimiviridae and 1 strain of Marseilleviridae (Saintcharles) from 53 freshwater Number of Isolated Strains samples, such as fountains, lakes, rivers and hospital wa- Since the isolation of the first giant virus, Mimivirus, ter. Moreover, La Scola et al. [11] tested 2 soil samples and methodological improvements have led to the isolation of 10 seawater samples. Both soil samples were positive for 43 strains of Mimiviridae (14 from lineage A, 6 from lin- Mimiviridae (Terra1 and Terra2), and 2 out of the 10 sea- eage B and 23 from lineage C) and 17 strains of Marseil- water samples were positive for Mimiviridae (Pointe- leviridae ( fig. 1 ). They are all classified within group I of rouge1 and Pointerouge2). This was the first time that Megavirales, which includes the Marseilleviridae and giant viruses were isolated from an ecological system oth- Mimiviridae lineages A, B and C [7] . er than from the water of a cooling tower. Based on the observation that giant viruses can also be found in other Nature of the Positive Samples environments, a study was performed focusing on hyper- After the isolation of the first Mimivirus in 2003 from saline water and soil samples from Tunisia [12] , and an- the water of an air conditioning system, the second strain, other study focused on diverse soil samples collected named A. castellanii Mamavirus, was found in a cooling around Marseille, in the south of France [13] . Both stud-

358 Intervirology 2013;56:354–363 Pagnier et al. DOI: 10.1159/000354556

 88.183.11.77 - 10/20/2013 7:47:35 PM 7:47:35 - 10/20/2013 88.183.11.77 Table 2. Main features of Mimiviruses and Marseilleviruses isolated in our laboratory

Family Name Source Country/region Capsid GenBank Genome Refer- size accession size (nt) ence (nm) no.

Mimiviridae (n = 43) Group A (n = 14) A. polyphaga Cooling tower water UK (Bradford) 400 NC_014649 1,181,549 [1, 2] Mimivirus A. castellanii Cooling tower water France (Paris) 450 JF801956 1,191,693 [4] Mamavirus

Terra2 Soil France (Marseille) 370 – 1,170,000 [11] Pointe-Rouge2 Seawater France (Marseille) 500 – 1,160,000 [11] Cher Rivers and lakes France (Tours) 420 – – [11] Fauteuil Hospital water France (Marseille) 600 – 1,180,000 [11] Longchamps Decorative fountain France (Marseille) 450 – 1,103,000 [11] water

Lactours Rivers and lakes France (Tours) 450 – 1,180,000 [11] Pointe-Rouge1 Seawater France (Marseille) 390 – 1,146,000 [11] Lentille Lens liquid France (Marseille) 500 JF979182 1,220,000 [11]

Marais Swamp France (Aubagne) – – 1,197,000 NP Univirus Compost France (Marseille) – – 1,087,000 NP Hirudovirus Leech France (Marseille – – – NP Montadette2 Soil France (Martigues) – – – NP Group B (n = 6) Moumouvirus Cooling tower water France (Rousset) 420 JX962719 1,021,421 [11] Monve Cooling tower water France (Puget sur Argens) 390 JN885994- 1,015,033b [11] JN886001

Ochan Compost France (Marseille) – – 1,026,000 NP Goulette Seawater Tunisia (Tunis) – – 1,026,000 [10] Istres Soil France (Istres) – – – NP Cassis49 Soil France (Cassis) – – – NP Group C (n = 23) Courdo7 Rivers and lakes France (Saint-Raphaël) 400 JN885990- 1,170,106b [11] JN885993

Terra1 Soil France (Marseille) 420 – 1,230,000 [11] Montpellier Decorative fountain France 370 – 1,225,000 [11] water (Montpellier)

Courdo11 Rivers and lakes France (Saint-Raphaël) 450 – 1,245,674 [11] Courdo5 Rivers and lakes France (Marseille) 400 – – [11] Bus Cooling tower water France (Marseille) 400 – 1,227,000 [11] Mont1 Soil (mountain) Tunisia (Tunis) – – – [10] LBA111 Broncholalveolar lavage Tunisia – – 1,230,519 [13] Avenue9 Soil Tunisia (Tunis) – – 1,214,000 [10] Afrovirus Soil France (Aubagne) – – – NP Montadette1 Soil France (Martigues) – – – NP Balcon Soil France (Marseille) – – – NP Terrain en Soil France (Marseille) – – – NP construction

Boug1 Chott (hypersaline soil) Tunisia (Gafsa) – – – [10] Shan Stool Tunisia (Tunis) – – – NP Cornil Soil France (Marseille) – – – NP Saint Pierre Stagnant water France (Marseille) – – – NP Borély Stagnant water France (Marseille) – – – NP Capucin Stagnant water France (Marseille) – – – NP Potager Soil France (Marseille) – – – NP Feuillage Soil France (Martigues) – – – NP Luminy43 Water France (Marseille) – – – NP Sète Soil France (Sète) – – – NP

Improvements in Mimiviridae and Intervirology 2013;56:354–363 359 Marseilleviridae Isolation from Amoeba DOI: 10.1159/000354556

 Downloaded by: Downloaded PM 7:47:35 - 10/20/2013 88.183.11.77 Table 2 (continued)

Family Name Source Country/region Capsid GenBank Genome Refer- size accession size (nt) ence (nm) no.

Marseilleviridae (n = 17) Marseillevirus (n = 2) Marseillevirus Cooling tower water France (Cannes) 190 JF979175.1 374,000 [3]

Senegalvirus Stool Senegal – JF909596- 386,000 [9] JF909602

Tunisvirus (n = 1) Fontaine2 Fountain water Tunisia (Ariana) – JX484143 382,000 [10] Unassigned (n = 14) Cannes8 Cooling tower water France (Cannes) 190 JF979175.1 374,000 [11]

Cannes9 Cooling tower water France (Cannes) 150 – – [11] Saint-Charles Decorative fountain France (Marseille) 230 – 376,000 [11] water

Seb1 eau Sebkha (hypersaline water) Tunisia (Tunis) – – – [10] Seb1 sol Soil (hypersaline soil) Tunisia (Tunis) – – – [10] Seb6 sol Soil (hypersaline soil) Tunisia (Tunis) – – – [10] Seb2 sol Soil (hypersaline soil) Tunisia (Tunis) – – – [10] Oued1 River Tunisia (Bézert) – – – [10] Cité1 Soil Tunisia (Kef) – – – [10] Rivière1 River (Majerda) Tunisia (Kef) – – – [10] Puit1 Well water Tunisia (Cap Bon) – – – [10] Hammam1 Hammam water Tunisia (Tunis) – – – [10] Sidi thabet Soil Tunisia (Ariana) – – – [10] Insectomime Diptere larvae Tunisia – – – [14] NP = Unpublished data.

ies resulted in the isolation of Mimiviridae and Marseil- lanii, and an 18% positive identification rate was achieved. leviridae in all of the environments that were sampled. In 2012, the development of high-throughput methods Moreover, both studies led to the isolation of virophages: helped to process a greater number of samples. Of the Sputnik 3 [13] and Sputnik 4 [12] . Later, we explored the 1,000 samples that were screened, 15 were positive (1.5% possibility that giant viruses were eventually found in hu- positive) for giant viruses (4 Mimiviridae and 11 Marseil- man samples (respiratory and stool samples) and animals leviridae) [12] . (i.e. insects and leeches), which further expanded the known ecological niches of giant viruses [14, 15] . Culture P rocedure with Preliminary Molecular Detection D irect Culture Procedure The first study using the culture procedure with pre- The first isolation of giant viruses was conducted by liminary molecular detection was conducted with stan- directly inoculating a sample onto amoeba without a pre- dard PCR using primer pairs designed for Mimiviridae liminary detection step. The isolation of Mimivirus in lineages A, B and C. This initial study led to 9 positive 2003 was performed following this protocol; later, the iso- identifications from 90 soil samples (8 from lineage A and lation of Mamavirus and Marseillevirus resulted from the 1 from lineage C); a 10% positive hit rate. Three of the direct inoculation of a large sample volume of cooling positive samples could be isolated after culture; 2 of these tower water that was concentrated by filtration. After the belonged to lineage A (Univirus, Marais) and 1 belonged direct inoculation findings, we began to set up prospec- to lineage B (Ochan). More recently, the primary molecu- tive studies of giant viruses using an antibiotic-directed lar detection system was improved by using a primer- procedure. The first prospective study occurred in 2010 probe system that targets lineage C of Mimiviridae. Using [11] and led to the isolation of 19 giant viruses (16 Mimi- this detection strategy, 11 positive samples were detected viridae and 3 Marseilleviridae) by cocultivating 105 envi- in 156 soil samples (17% positive hit rate), and 8 out of 11 ronmental samples, including cooling tower water, fresh positive samples were cultivated (Potager, Montadette1, environmental water and soil with the amoeba A. castel- Montadette2, Balcon, Sete, Terrain, Feuillage and Istres;

360 Intervirology 2013;56:354–363 Pagnier et al. DOI: 10.1159/000354556

 Downloaded by: Downloaded PM 7:47:35 - 10/20/2013 88.183.11.77 table 2 ). Surprisingly, the molecular characterisation of lated, A. polyphaga Mimivirus, was found in the water of the 8 cultivated viruses using the specific primer-probe an air conditioning system. Cooling towers represent a systems listed in the table 1 showed that Montadette2 be- very specific ecological system and are mostly closed sys- longs to lineage A and Istres to lineage B of Mimiviridae. tems that only slowly renew circulating water. In addition, The last molecular detection system using the primer- the temperatures are favourable for several microorgan- probe system included targets for lineages A and B of isms, particularly protozoa such as amoebae [16] . More- Mimiviridae and for Marseilleviridae. The completed over, in those systems, amoebae and other microorganisms molecular detection system detected 4 positive samples have the capacity to form biofilms, which are ideal for mi- from 68 samples (5.9% positive hit rate), and all positive crobial development and are difficult to remove. The spec- samples were in Mimiviridae lineage C. All 4 could be ificity of the cooling tower ecosystem could result in the cultivated (Cornil, Saintpierre, Borely and Capucin). In best environment for the development of giant viruses be- our last series of 96 environmental samples, we tested the cause of their specific link to amoebae. Therefore, the first same primer-probe systems for detection in parallel and studies of Mimiviridae were performed on cooling tower direct cultivation using isolation with additional enrich- water systems [3, 4] . However, the first giant virus pros- ment by blind subculture. No sample was positive by pecting study focused not only on cooling tower water but PCR, but 15 (15.6%) were positive by culture; all were also on other types of environmental samples, such as Mimiviridae and included 13 from lineage A, 1 from lin- freshwater, seawater and soil [11] . Based on this study, we eage B (Cassis49) and 1 from lineage C (Luminy43). noticed that giant viruses are ubiquitously distributed in all the environments studied (18% positive samples). The first procedure used to isolate the viruses was empirical and Discussion based on the use of antibiotics specifically adapted to the bacterial contamination encountered. The first step lead- In Marseille, during the last decade, the first 8 giant ing to an improvement in the giant virus isolation proce- virus isolates were isolated from cooling tower water dure was to decontaminate the coculture using adapted (5 mimiviruses and 3 marseilleviruses). Later, screening antibiotic cocktails, including antifungal agents. These of diverse environmental systems led to virus isolation cocktails were improved to exhibit better antimicrobial ef- from freshwater (13 Mimiviridae, 7 Marseilleviridae) and ficiency and to decrease toxicity against amoebae. common soil (18 Mimivirus, 2 Marseillevirus). However, A second improvement was used to increase the num- more unexpected environments led to the isolation of gi- ber of samples that could be tested in parallel by establish- ant viruses, in particular seawater, hypersaline water and ing a high-throughput system [12] . This adaptation of the soils (5 Mimiviridae and 4 Marseilleviridae). In parallel, agar plate method was based on the observation that the during a culturomic study of the human gut in 2012, the presence of viruses can be detected by the naked eye first giant virus isolated from a human sample was found through direct observation of a lysis plaque around the in the stool of a healthy Senegalese patient (Senegalvirus, inoculation point of the enriched culture on an amoebal belonging to the Marseilleviridae) [6] . Another study per- monolayer. This phenomenon had already been observed formed in 2013 on respiratory samples of patients with for chlorella viruses on agar plates coated with monocel- pulmonary infection resulted in the isolation of the first lular algae [17] , and the method could be adapted to strain of Mimiviridae in a human sample (LBA111) [14] , amoebae infected with giant viruses. The combination of and this study used the high-throughput method de- high-throughput screening and the visual plaque assay scribed by Boughalmi et al. [12] . With the same method, led to the rapid isolation of 15 positive samples out of another giant virus (Shanvirus) was isolated from a stool 1,000 total samples. The difference in efficiency com- sample and was classified in lineage C of Mimiviridae. A pared with the previous study (18 vs. 1.5% positive sam- Mimiviridae lineage A (Lentille) was found in contact ples) can be explained by the fact that the samples were lens washing solution, which is related to the human en- largely taken from extreme environments, such as hyper- vironment. More recently, two strains of giant viruses iso- saline water or soil. The higher amount of Marseilleviri- lated from animals were found in a leech (Hirudovirus, dae can also be explained by the use of preliminary filtra- Mimiviridae lineage A) and in the larvae of the Diptera tion through 0.8-µm-pore filters, which is the same meth- Eristalis tenax (Insectomime, Marseilleviridae) [15] . od that led to the isolation of Senegalvirus, the first giant The first investigation of giant viruses focused on the virus isolated from a human stool sample [6] . However, water in cooling towers because the first giant virus iso- the low rate of positive samples in cultures spurred the use

Improvements in Mimiviridae and Intervirology 2013;56:354–363 361 Marseilleviridae Isolation from Amoeba DOI: 10.1159/000354556

 Downloaded by: Downloaded PM 7:47:35 - 10/20/2013 88.183.11.77 of a second improvement to the method: the addition of week) was an improvement, we believe that a higher- a preliminary molecular detection step. Indeed, only the throughput procedure (i.e. testing up to 500 samples per PCR-positive samples were cultivated, and in nearly all week) cannot be achieved using the current technique, cases, the detected virus could be isolated. The first PCR which is time consuming and not amenable to automa- system focused on Mimiviridae and used standard PCR tion. For future work, we believe that the combination of with primer pairs based on the sequences of the beta sub- liquid culture of amoebae in microplates with one or unit of the Mimiviridae DNA polymerase (polB) from more blind enrichment steps and the detection of amoe- each lineage. The couple Mimivirus-Terra2, Moumouvi- bal lysis by an automated flow cytometer will be the best rus, and the couple Courdo11-Terra1 were used to design strategy for optimal high-throughput analysis. Converse- primer pairs for lineages A, B and C, respectively. From ly, the use of ethanol to decontaminate samples without this study, 9 PCR-positive samples from the initial 90 killing viruses, especially mimiviruses, should be a suit- samples (10%) were tested in culture, and 3 of them led able alternative to the use of antibiotics in situations to the isolation of a giant virus (Univirus, Marais and where cultures need to be performed on non-axenic pro- Ochan). However, the specificity of those primer pairs for tozoa [22, 23] . A unique proposed alternative was pub- the targeted group was not optimal and led to the design lished in a report related to the isolation of Megavirus of primer-TaqMan probes targeting all three lineages of chilensis [24] . In this work, the authors incubated their Mimiviridae as well as the Marseilleviridae. The first test- liquid sample for 1 month in the dark with a source of ed primer-probe targeted lineage C and led to the isola- carbon, thus allowing heterotrophic bacteria to multiply. tion of 8 Mimiviridae in 11 PCR-positive samples from In this system, the heterotrophic bacteria serve as the food 156 environmental samples. We expected that those 8 vi- source for the protozoa present in the sample, thus ex- ruses would belong to lineage C, but, surprisingly, 1 was panding the virus population. This technique will have to a lineage A isolate (Montadette2) and 1 was a lineage B be compared to blind enrichment, which is easier and isolate (Istres). It is possible that several viruses were pres- quicker to perform. However, we believe that the best ent in the same samples and that molecular detection am- strategy for isolating these giant viruses would be to sam- plified one of them, whereas culture led to the isolation of ple biofilms, where the amoeba hosts are concentrated, another. The next step in culture improvement was to test rather than free water. the three other primer-probe systems (Mimiviridae lin- eage A and B, and Marseillevirus). Using all four primer- probe systems, a series of 68 samples were analysed, and Disclosure Statement 4 were PCR positive for lineage C of Mimiviridae and resulted in cultivation of the correct lineage C viruses The authors declare that there is no potential conflict of interest or financial disclosure. (Cornil, Saintpierre, Borely and Capucin). Currently, it is difficult to draw conclusions concerning the actual distri- bution of giant viruses in the environment. Our studies confirm the results of metagenomic data that has identi- References 1 La Scola B, Audic S, Robert C, Jungang L, de fied sequences related to these viruses in many environ- Lamballerie X, Drancourt M, Birtles R, Cla- ments [18, 19] , including humans [20] . Moreover, we re- verie JM, Raoult D: A giant virus in amoebae. Science 2003; 299: 2033. cently isolated a giant virus closely related to Marseillevi- 2 Raoult D, Audic S, Robert C, Abergel C, Re- rus from a human sample that did not grow on amoebae nesto P, Ogata H, La Scola B, Suzan M, Cla- but did grow on cells, which demonstrates that the panel verie JM: The 1.2-megabase genome sequence of Mimivirus. Science 2004; 306: 1344–1350. of hosts is likely not limited to Acanthamoeba , the only 3 Boyer M, Yutin N, Pagnier I, Barrassi L, Four- host utilised to isolate giant viruses up to this point [21] . nous G, Espinosa L, Robert C, Azza S, Sun S, Rossmann MG, Suzan-Monti M, La Scola B, Giant viruses are likely ubiquitous, as are their amoe- Koonin EV, Raoult D: Giant Marseillevirus bal hosts, throughout a variety of environments. This is highlights the role of amoebae as a melting because it was possible to isolate giant viruses from all of pot in emergence of chimeric microorgan- isms. Proc Natl Acad Sci USA 2009; 106: the biotopes we tested, including extreme environments. 21848–21853. In the future, it is likely that improvements in the spe- 4 La Scola B, Desnues C, Pagnier I, Robert C, cific antibiotic cocktail and antibiotic concentration will Barrassi L, Fournous G, Merchat M, Suzan- Monti M, Forterre P, Koonin E, Raoult D: The increase efficiency. However, although the high-through- virophage as a unique parasite of the giant put procedure (allowing 50–100 samples to be tested per mimivirus. Nature 2008; 455: 100–104.

362 Intervirology 2013;56:354–363 Pagnier et al. DOI: 10.1159/000354556

 88.183.11.77 - 10/20/2013 7:47:35 PM 7:47:35 - 10/20/2013 88.183.11.77 5 Thomas V, Bertelli C, Collyn F, Casson N, 12 Boughalmi M, Saadi H, Pagnier I, Colson P, 18 Kristensen DM, Mushegian AR, Dolja VV, Telenti A, Goesmann A, Croxatto A, Greub Fournous G, Raoult D, La Scola B: High- Koonin EV: New dimensions of the virus G: Lausannevirus, a giant amoebal virus en- throughput isolation of giant viruses of the world discovered through metagenomics.

coding histone doublets. Environ Microbiol Mimiviridae and Marseilleviridae families in Trends Microbiol 2010; 18: 11–19.

2011; 13: 1454–1466. the Tunisian environment. Environ Micro- 19 Steward GF, Preston CM: Analysis of a viral

6 Lagier JC, Armougom F, Million M, Hugon P, biol 2013; 15: 2000–2007. metagenomic library from 200 m depth in Pagnier I, Robert C, Bittar F, Fournous G, 13 Gaia M, Pagnier I, Campocasso A, Fournous Monterey Bay, California constructed by di-

Gimenez G, Maraninchi M, Trape JF, Koonin G, Raoult D, La Scola B: Broad spectrum of rect shotgun cloning. Virol J 2011; 8: 287. EV, La Scola B, Raoult D: Microbial cul- Mimiviridae virophage allows its isolation us- 20 Colson P, Fancello L, Gimenez G, Armougom

turomics: paradigm shift in the human gut ing a Mimivirus reporter. PLoS One 2013; F, Desnues C, Fournous G, Yoosuf N, Million microbiome study. Clin Microbiol Infect 8:e61912. M, La Scola B, Raoult D: Evidence of the

2012; 18: 1185–1193. 14 Saadi H, Pagnier I, Colson P, Cherif JK, Beji megavirome in humans. J Clin Virol 2013; 57: 7 Colson P, de Lamballerie X, Fournous G, M, Boughalmi M, Azza S, Armstrong N, Rob- 191–200. Raoult D: Reclassification of giant viruses com- ert C, Fournous G, La Scola B, Raoult D: First 21 Popgeorgiev N, Boyer M, Fancello L, Monteil posing a fourth domain of life in the new order isolation of Mimivirus in a patient with pneu- S, Robert C, Rivet R, Nappez C, Azza S, Chia-

Megavirales . Intervirology 2012; 55: 321–332. monia. Clin Infect Dis 2013; 57:e127–e134. roni J, Raoult D, Desnues C: Giant Blood 8 Colson P, Pagnier I, Yoosuf N, Fournous G, 15 Boughalmi M, Pagnier I, Aherfi S, Colson P, Marseillevirus recovered from asymptomatic La Scola B, Raoult D: ‘Marseilleviridae’, a new Raoult D, La Scola B: First isolation of Mar- blood donors. J Infect Dis 2013, E-pub ahead family of giant viruses infecting amoebae. seillevirus in the Diptera Syrphidae Eristalis of print.

Arch Virol 2013; 158: 915–920. tenax . Intervirology 2013; 56: 386–394. 22 Campos RK, Andrade KR, Ferreira PC, Bon- 9 Fischer MG, Allen MJ, Wilson WH, Suttle CA: 16 Pagnier I, Merchat M, La Scola B: Potentially jardim CA, La Scola B, Kroon EG, Abrahao JS: Giant virus with a remarkable complement of pathogenic amoeba-associated microorgan- Virucidal activity of chemical biocides against genes infects marine zooplankton. Proc Natl isms in cooling towers and their control. Fu- mimivirus, a putative pneumonia agent. J

Acad Sci USA 2010; 107: 19508–19513. ture Microbiol 2009; 4: 615–629. Clin Virol 2012; 55: 323–328. 10 La Scola B, Barrassi L, Raoult D: Isolation of 17 Fitzgerald LA, Wu PK, Gurnon JR, J Biffinger 23 Slimani M, Pagnier I, Boughalmi M, Raoult D, new fastidious α-Proteobacteria and Afipia fe- JC, Ringeisen BR, Van Etten JL: Isolation of La Scola B: Alcohol disinfection procedure for lis from hospital water supplies by direct plat- the phycodnavirus PBCV-1 by biological la- isolating giant viruses from contaminated

ing and amoebal co-culture procedures. ser printing. J Virol Methods 2010; 167: 223– samples. Intervirology 2013; 56: 434–440.

FEMS Microbiol Ecol 2000; 34: 129–137. 225. 24 Arslan D, LeGendre M, Seltzer V, Abergel C, 11 La Scola B, Campocasso A, N’Dong R, Four- Claverie JM: Distant Mimivirus relative with nous G, Barrassi L, Flaudrops C, Raoult D: a larger genome highlights the fundamental Tentative characterization of new environ- features of Megaviridae. Proc Natl Acad Sci

mental giant viruses by MALDI-TOF mass USA 2011; 108: 17486–17491.

spectrometry. Intervirology 2010; 53: 344–353.

Improvements in Mimiviridae and Intervirology 2013;56:354–363 363 Marseilleviridae Isolation from Amoeba DOI: 10.1159/000354556

 Downloaded by: Downloaded PM 7:47:35 - 10/20/2013 88.183.11.77 Chapitre Quatre

La spécificité d’hôtes des virophages Sputnik

 4.1 – Article : le large spectre d’hôtes des virophages de Mimiviridae permet leur isolement en utilisant un Mimivirus rapporteur

Broad Spectrum of Mimiviridae Virophage Allows Its Isolation Using a Mimivirus Reporter. Morgan Gaia, Isabelle Pagnier, Ange´ lique Campocasso, Ghislain Fournous, Didier Raoult, Bernard La Scola*.

URMITE, UM63, CNRS 7278, IRD 198, Inserm 1095, Aix Marseille Universite, Marseille, France

* Corresponding author: [email protected]

Published April 15, 2013. PLoS ONE 8(4): e61912. doi:10.1371/journal.pone.0061912

 Avant-propos

Comme décrit dans le chapitre Deux, le virophage Sputnik a été isolé avec une souche de Mimiviridae , Mamavirus. Un deuxième Sputnik a été isolé deux années plus tard en association avec Lentillevirus. A ce moment, le nombre de Mimiviridae isolés dépassait déjà la vingtaine, avec des différences nucléotidiques relativement significatives : sur la base de la comparaison de leur séquence du gène codant pour l’ADN polymérase B, ils sont d’ailleurs partagés en trois groupes, le A, le B et le C. Mamavirus et Lentillevirus, avec lesquels les virophages Sputnik avaient été isolé, appartiennent tous deux au groupe A. Une fois les virophages Sputnik purifiés et séparés de leur virus hôte, la question de leur spécificité au sein des Mimiviridae fut naturellement soulevée. Le fait que cette spécificité soit large au sein des Mimiviridae a permis par la suite d’établir un protocole d’isolement de vi rophage de type Sputnik sans son hôte natif, par l’utilisation d’un virus géant rapporteur. Cette approche a abouti à l’isolement de Sputnik 3.

 Broad Spectrum of Mimiviridae Virophage Allows Its Isolation Using a Mimivirus Reporter

Morgan Gaia, Isabelle Pagnier, Ange ´lique Campocasso, Ghislain Fournous, Didier Raoult, Bernard La Scola* URMITE, UM63, CNRS 7278, IRD 198, Inserm 1095, Aix Marseille Universite, Marseille, France

Abstract The giant virus Mimiviridae family includes 3 groups of viruses: group A (includes Acanthamoeba polyphaga Mimivirus), group B (includes Moumouvirus) and group C (includes Megavirus chilensis). Virophages have been isolated with both group A Mimiviridae (the Mamavirus strain) and the related Cafeteria roenbergensis virus, and they have also been described by bioinformatic analysis of the Phycodnavirus. Here, we found that the first two strains of virophages isolated with group A Mimiviridae can multiply easily in groups B and C and play a role in gene transfer among these virus subgroups. To isolate new virophages and their Mimiviridae host in the environment, we used PCR to identify a sample with a virophage and a group C Mimiviridae that failed to grow on amoeba. Moreover, we showed that virophages reduce the pathogenic effect of Mimivirus (plaque formation), establishing its parasitic role on Mimivirus. We therefore developed a co-culture procedure using Acanthamoeba polyphaga and Mimivirus to recover the detected virophage and then sequenced the virophage’s genome. We present this technique as a novel approach to isolating virophages. We demonstrated that the newly identified virophages replicate in the viral factories of all three groups of Mimiviridae, suggesting that the spectrum of virophages is not limited to their initial host.

Citation: Gaia M, Pagnier I, Campocasso A, Fournous G, Raoult D, et al. (2013) Broad Spectrum of Mimiviridae Virophage Allows Its Isolation Using a Mimivirus Reporter. PLoS ONE 8(4): e61912. doi:10.1371/journal.pone.0061912 Editor: John Parkinson, Hospital for Sick Children, Canada Received August 14, 2012; Accepted March 18, 2013; Published April 15, 2013 Copyright: ß 2013 Gaia et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: These authors have no support or funding to report. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction Mimivirus growth. Later, we identified an isolate (Lentille) in our APM collection that was associated with a new strain of virophage; Free-living amoebas are ubiquitous protozoa that feed on we named the new virophage strain Sputnik 2 [9]. More recently, microorganisms in their environment through phagocytosis. another virophage was isolated with Cafeteria roenbergensis virus, a However, some microorganisms are able to resist digestion by giant virus that infects a marine phagocytic protist [10]. Using this predator after phagocytosis [1]. Thus, amoebas have been metaproteogenomic analysis, another virophage was detected in a used as a tool for the isolation of digestion-resistant environmental sample of a hypersaline meromictic lake in Antarctica; this bacteria, such as Legionella sp. [2]. This technique allowed the virophage is found in association with Phycodnaviridae, a giant fortuitous discovery of the first giant virus, Acanthamoeba polyphaga virus that infects algae [11]. The concept of virophages, small Mimivirus (APM) [3]. APM is a large icosahedral virus with a viruses that infect giant virus factories and lead to population 500 nm capsid [4] that is covered with surrounding fibrils and control of their hosts, is controversial [12–14]. However, this contains a 1.2 Mbp genome [5]. It belongs to the nucleocytoplas- emerging field justifies the exploration of the spectrum of mic large DNA virus (NCLDV) family in the recently proposed virophages in possible hosts and the design of new tools to order Megavirales [6], which also includes Iridoviridae, Phycod- complete their repertoire. naviridae, Asfarviridae, Ascoviridae and Poxviridae [7]. A second The first aim of our study was to investigate the replication closely related but slightly larger APM was later isolated, and the capacity of Sputnik 1 and Sputnik 2 in an A. polyphaga co-culture strain was called Mamavirus [8]. Since the isolation of the first two with each APM in our collection using quantitative real-time PCR giant viruses in our laboratory, we have improved the co-culture and, more notably, to investigate if virophages previously isolated technique, established a collection of APM and reported 18 with group A APM can infect group B or group C viruses. Noting isolates [9]. We have established a preliminary phylogenetic tree the inability to co-culture a group C virus with its virophage that based on partial polB sequences; the tree shows a repartition into was detectable by PCR, we decided to inoculate the sample on three groups: group A (includes the Mimivirus and Mamavirus), amoeba and Mimivirus to recover the virophage. group B (includes the Moumouvirus) and group C (includes the recently described Megavirus chilensis) [6]. Materials and Methods We isolated a small virus that co-isolates with Mamavirus and infects the Mamavirus virus factory. We named this new virus APM and virophage production Sputnik and classified it as the first virophage [8]. This icosahedral, Each virus in our collection (Table 1) was produced by the 50 nm virus with an 18 kb DNA genome is deleterious for inoculation of 10 ml of PYG medium with 1 ml each of the

PLOS ONE | www.plosone.org 1 April 2013 | Volume 8 | Issue 4 | e61912  Mimiviridae Virophage individual viral suspension calibrated at 10 6 particles/ml and the Real-time PCR assay for evaluation of Sputnik 5 Acanthamoeba polyphaga amoebal suspension at 5 610 cells/ml multiplication (strain Linc AP-1). After the complete lysis of the amoebas, each After incubation, 200 ml of each co-culture was sampled for co-culture was centrifuged at 2000 rpm for 10 min to pellet the DNA extraction and real-time PCR at H1, H0 and Day 3 (to remaining fragments. The supernatant was filtrated through a observe the complete lysis of amoebas). The DNA extraction was 0.8 mm pore filter to remove residual amoebas and cysts. The performed with a Qiamp DNA extraction kit (Qiagen, Hilden, 2 u supernatants were frozen at 80 C before being used for viral co- Germany) according to the manufacturer’s instructions. Real-time culture inoculation. PCR was performed with the LightCycler 480 SYBR Green I Sputnik 1 [8] and Sputnik 2 [9] were produced in co-culture Master (Roche) according to the manufacturer’s instructions. with their natural hosts, Mamavirus and Lentille virus, respec- Sputnik was detected and quantified by using a primer pair tively, in 10 ml of PYG medium containing 1 ml of the amoeba 5 targeting the ORF20 of Sputnik encoding the major capsid protein suspension at 5 610 cells/ml, 1 ml of giant virus suspension and (Forward primer 59-GAGATGCTGATGGAGCCAAT-39, Re- 1 ml of each Sputnik strain. After the complete lysis of the verse primer 59-CATCCCACAAGAAAGGAGGA-39). For each amoeba, the purification was performed as described above, but sample, the cycle threshold (Ct) was correlated to a reference scale the filtration was executed on three successive filters of 0.8-, 0.45- to allow the evaluation of the Sputnik concentration. A total of and 0.22 mm pore sizes to obtain a pure Sputnik suspension. This 200 ml of each co-culture was analyzed by real-time PCR directly suspension was prepared before each inoculation assay. These tests after inoculation to assess the quantities of Sputnik 1 and Sputnik were performed in triplicate. 2.

Sputnik co-culture with APM collection Microscopic observations Each giant virus was inoculated separately into 10 ml of PYG m 5 After incubation, 200 l of each co-culture at H0 and H6 was medium containing 1 ml of amoeba suspension at 5 610 cells/ml added to a Cytospin single chamber (Thermo), cytocentrifuged at and 1 ml of each Sputnik suspension. After 1 hour of incubation at 800 rpm for 10 min and fixed with methanol. Indirect immuno- u 32 C, the supernatant was gently removed to eliminate the fluorescence microscopy was performed with mouse anti-Sputnik Sputnik particles that were not internalized by the amoebas, and serum as described previously. To label the virus factories and the 10 ml of fresh PYG medium was added. This time point was cell nucleus, we used 5 ml of DAPI (4 9, 6 9-diamino-2-phenylindole, defined as H0. The Lentille virus, naturally infected by Sputnik 2, Molecular Probes). The viable trophozoites were counted using was used as a positive control. trypan blue (Oxoid) to assess the speed of the lysis of the amoebas by each virus. The preparation of selected samples for transmission electron microscopy was performed as described previously [15] to confirm the results of the quantitative PCR and immunofluores- cence assays.

Table 1. The list of giant viruses and their classification into Pol B gene sequence and phylogeny genotype groups. The complete Pol B gene sequences for all the viruses used in this study were available from unfinished APM genome sequenc- ing in progress in our laboratory. The sequences of DNA Name Group GenBank polymerase B were aligned using MUSCLE and trimmed by APM A HQ336222 TrimAL after visual editing. The phylogenetic tree was built with a Mamavirus A JF979171 maximum likelihood using Phyml software. Lentille A JF979182 Environmental detection of APM and virophages Fauteuil A JF979168 Tentative detection of APM and virophages was performed on Pointerouge1 A JF979167 90 soil samples collected in Marseille and the surrounding areas Cher A JF979166 (south of France). To each 15 g sample, 150 mL of sterile water Longchamps A JF979169 was added. Decantation occurred overnight at 4 uC, and then the Terra2 A GU265562 samples were filtered, first through Wattman’s paper and then Lactour A JF979173 through 0.1 mm membrane. Membranes were then put in 1 ml of Page’s Amoeba Saline buffer (PAS) and vortexed. The membranes Courdo5 A JF979179 were removed, and the suspension was kept at 280 uC prior to use. Pointerouge2 A JF979161 DNA was extracted with the QIAGENß QIAmp Mini Kit Monve B JF979181 following the manufacturer’s protocol. A combination of several Moumou B GU265560 primer pairs was required because the low polB gene conservation Terra1 C GU265563 in the available fragment [9] prevented the design of broad range Courdo7 C JF979172 primers (Table 2). We have defined the viruses belonging to the Mimivirus group as group A, the viruses belonging to Moumou- Courdo11 C GU265561 virus as group B and the viruses belonging to Courdo11 as group Bus C JF979178 C. Group C includes the recently described Megavirus chilensis [6]. Montpellier C JF979165 Sputnik was detected using the primer pair targeting the ORF20 Megavirus chilensis* C NC_016072 of Sputnik, which encodes the major capsid protein described above. The giant virus labeled with * is not part of our collection. doi:10.1371/journal.pone.0061912.t001

PLOS ONE | www.plosone.org 2 April 2013 | Volume 8 | Issue 4 | e61912  Mimiviridae Virophage

4 9 Table 2. Sequences and Tm of the different primers pairs 10 pfu/mL. Sputnik3 was quantified at 10 particules by mL used in the study to amplify partial polB gene of Mimiviridae, using a virus-titer estimation protocol with negative staining of group A (Mimi-TJA 01, Mimi-TJA 02), group B (CE11-TE1 01, observed with electron microscopy. Fifteen-microliter deposits CE11-TE1 02) and group C (VA10 01). were made on amoeba plates, with 25 deposits per plate. Thirty- four deposits were made with 7.5 mL of Mimivirus at 10 9 pfu/mL and 7.5 mL of PAS, 34 with 7.5 mL of Mimivirus at 10 9 pfu/mL Primerspair Sequence Tm and 7.5 mL of Sputnik3 suspension. Deposits of PAS-only and Sputnik3-only were used as negative controls. The plates were Mimi-TJA 01 F 5 9-GCAGCCCTTTGACACTT-3 9 52 uC checked every day for lysis plaque formation. As the plaques were R 5 9-CATGCGGGAGTTGGAGA-3 9 circular, diameters were measured to calculate mean areas. Mimi-TJA 02 F 5 9-GAAAATGGTGAAGAGAAAACTGA-3 9 50 uC Statistical comparison was performed by Student’s t-Test, using R 5 9-ACCAGGATAAATGGATGCAA-3 9 R software (package stats version 2.15.0). CE11-TE101 F5 9-AGTTACCCAACCACAAGAAGA-3 9 45 uC Results R 5 9-CAGAAGGACTAACAAAAGAACCA-3 9 CE11-TE102 F5 9-AAAATATTGGGGACGTTGGTG-3 9 45 uC Pol B gene sequence and phylogeny R 5 9-ATGGAAGACTGGCTGTTGAAA-3 9 The sequences of pol B gene of all the giant viruses available in VA1001 F5 9-AAGGGGACAAGGAGTTAAAATAT-3 9 45 uC our collection (Table 1) were compared with each other. At the protein level, these sequences were identical. At the nucleotide R 5 9-TAGATATACGTTTGGTTTTGGAGTGA-3 9 level, the phylogenetic tree, which was built according to the full, doi:10.1371/journal.pone.0061912.t002 5-kbp sequences of the viral pol B genes, shows a repartition into three groups (Figure 1). The majority of the giant viruses in our collection are clustered in group A. Tentative isolation of virophages To date, virophage cultures have always been obtained by co- isolation with their viral hosts. However, in our study, all attempts Sputnik growth in the 3 groups of Mimiviridae to isolate the giant viruses with their virophages in the 5 samples Each of the 17 Mimiviridae in our collection (10 in group A, 2 that tested positive using amoeba co-cultures were unsuccessful. in group B and 5 in group C) was inoculated with Sputnik. Using real-time PCR quantification, we demonstrated a 10- to 30-fold Therefore, we designed a new protocol to isolate virophages by increase in Sputnik concentration in most viruses (Figure 1). inoculating samples on a co-culture of Acanthamoeba polyphaga and However, in 2 group C viruses (Bus and Montpellier), we observed Mamavirus (used as a reporter virus) [8]. We performed daily only a 5- to 10-fold increase in Sputnik concentration. blind subcultures and quantitative PCR to detect growth. The five Immunofluorescence observations using Sputnik antibodies and samples detected as positive by PCR were inoculated on 1 mL of DAPI confirmed the results obtained with real-time PCR. We PYG with 100 mL of fresh A. polyphaga amoebae at a concentration could observe clearly the viral factories (VF) and the production of of 5 610 5 cells/mL and 100 mL of Mamavirus (previously cured of Sputnik virions at one pole of the VF (Figure 2) as previously its virophage) at a concentration of 10 6 pfu/mL. The Mamavirus described [15]. For Bus and Montpellier, we observed a weak, was confirmed free of virophages by PCR amplification (using the diffuse staining of Sputnik (unpublished data) and fewer amoebas primers described above) before inoculation. Every day, 750 mL of (Figure S1). For these two viruses, the number of amoebas declined the suspension was mixed with fresh A. polyphaga and Mamavirus at significantly and more rapidly between H15 and H18 compared the same concentrations. Putative growth of virophages was with the 15 other isolates, whether co-infected with Sputnik or not, monitored daily by quantitative PCR. and reached levels similar to those observed between H36 and H40 with the viruses of the APM group (Figure S1). The Lysis plaque assay with Mimivirus and Sputnik3 production of Sputnik by these two isolates was checked by TEM Non-nutritive plates containing PAS with 1.5% Agar and 5% of at 6 and 12 h post-infection, but no Sputnik virions were observed. Merckß HemacolorH solution 3 in square Petri dishes of This result is consistent with the diffuse staining observed using approximately 520 cm 2 were used. These plates were inoculated immunofluorescence. The more rapid multiplication of these with 30 mL of fresh A. polyphaga (5 610 6 cells/mL) that was strains may explain the lower Sputnik production. previously rinsed in PAS. After one hour of sedimentation, the liquid and surplus of amoeba were removed. Before inoculation, Detection of Mimiviridae and virophages in the 9 rinsed Mimivirus was suspended in PAS and quantified at 10 environment infectious particles using an end-point dilution assay. The We tested 90 soil samples collected in Marseille and surround- Sputnik3 aliquots used for inoculation were diluted in PAS and ing areas (south of France) for Mimiviridae and virophage using previously filtered through a 0.2 mm pore-sized filter. Sputnik3 was PCR. We detected 9 Mimiviridae: 8 group A strains and 1 group quantified by inoculation of 50 mL of dilutions ranging from the C strain. We detected 5 virophage sequences, one of which was 215 5 original tube to 10 in 500 mL of fresh A. polyphaga (5 610 cells/ associated with the group C virus mentioned above. Detection of 9 mL) rinsed in PAS, infected by Mimivirus (50 mL at 10 pfu/mL). the other 4 virophages was not associated with the detection of a Five daily subcultures were performed in Mimivirus-infected A. giant virus. The available partial sequence together with both the polyphaga, then 7.5 mL of each well were mixed with 7.5 mL of sequences of the giant viruses that we studied previously and the 9 Mimivirus at 10 pfu/mL and dropped off on plates. Observations recently described Megavirus chilensis was used to build a three days after the deposits showed smaller diameters of the lysis phylogenetic tree (Figure S2) [6,9]. Our efforts to grow the group plaque formations of the originally Sputnik3 not diluted to the C virus potentially associated with a virophage failed, and we 24 10 dilution, and almost the same diameters for the other concluded that our amoeba support was ineffective. Therefore, we dilutions of Sputnik3 and the Mimivirus-PAS controls: according tried to isolate the virophage alone on a co-culture of amoeba and to these results, we estimated the concentration of Sputnik3 at Mamavirus.

PLOS ONE | www.plosone.org 3 April 2013 | Volume 8 | Issue 4 | e61912  Mimiviridae Virophage

Figure 1. Sputnik 1 and 2 growth in different giant viruses. A histogram of Sputnik 1 and 2 growth in different giant viruses according to their phylogenetic position in the groups A, B and C of Mimiviridae based on Pol B gene sequences. The growth, measured by real-time PCR quantification, was calculated between day 0 and day 3 and corresponded with complete amoebal lysis. doi:10.1371/journal.pone.0061912.g001

Isolation of virophage by co-culture with Mamavirus and Sputnik2 (GenBank: JN603369) (Figure 3d and Figure S3). amoebae Sputnik2 and Sputnik3 show four differences compared with the Prepared soil samples were inoculated on the co-culture, and we first Sputnik identified: insertions-deletions (in-dels) of an adenine performed daily blind subcultures and quantitative PCR to detect base at positions 877 and 7949/7951 and in-dels of a thymine base growth (Figure 3a). Of the 5 samples in which a virophage was at positions 12936 and14047 (the Sputnik genome was used as the detected after 5 blind subcultures, only the soil sample containing reference for nucleotide numbers). The adenine in-dels are located the group C virus sequence demonstrated increasing concentra- between ORFs and therefore do not affect the coding sequence. tion of virophage DNA as determined by real-time quantitative However, the first in-del of a thymine is located in gp17, the gene PCR (data not shown). This was confirmed by electron encoding a putative IS3 family transposase A protein; the change microscopy (Figure 3 b and c). As soon as growth of this potential causes a shift in the reading frame and is associated with an virophage was detected, production was performed on A. polyphaga extended gp17 in Sputnik2 and Sputnik3 (88 amino acids in newly infected by Mamavirus in PYG (1:1:10 ratio in volume, Sputnik, and 187 in Sputnik2 and Sputnik3) (Figure S4). The respectively). Daily subculture was performed at a 1:2 ratio (a second in-del of a thymine occurs at a crossing point between gp18 volume of culture in the double of fresh amoeba infected by and gp19 and results in a single ORF instead of gp18 and gp19. Mamavirus). Later, we were able to propagate this virophage on However this appeared to be actually an initial mistake from the Courdo11, a group C strain, and on Moumouvirus, a group B sequencing of Sputnik as recently proved (Zhang et al , 2012 [16]), strain. The complete genome sequencing of Sputnik3 (GenBank and thus gp18 and gp19 are Sputnik3 has three substitutions accession number: JN603370) allowed a comparison with the compared with Sputnik and Sputnik2. The first substitution, genomes of virophages Sputnik (GenBank: EU606015) and adenine to thymine, is located between gp12 and gp13 at position

PLOS ONE | www.plosone.org 4 April 2013 | Volume 8 | Issue 4 | e61912  Mimiviridae Virophage

Figure 2. Sputnik within viruses of the different groups of Mimiviridae. (A) A DAPI and immunofluorescence labeling of Sputnik 1 within viruses of the groups A (Mimivirus), B (Moumouvirus) and C (Courdo11) of Mimiviridae at 6 h post infection. This figure shows the Sputnik particles labeled with mouse antibody serum (green), and the nucleic acids are indicated by DAPI stain (blue/purple). The virus factories are especially visible by the abundant green stain. ( B) Aspect of Sputnik3 virophage produced in virus factory of group A Mimivirus (on the left; scale bar 1 mm), group B Moumouvirus (at the center; scale bar 200 nm) and group C Courdo11 (on the right; scale bar 2 mm). doi:10.1371/journal.pone.0061912.g002

8986/8991. A guanine to adenine change is located at position ORF20 of Sputnik 3 also differed by 3 nucleotides compared to 16098 in ORF20, which encodes the putative major capsid that obtained directly by PCR on the soil sample. protein. This substitution changes a nonpolar alanine to a polar threonine and could affect the function of the putative protein or Suspicion that Sputnik3 virophage is a pathogen for its structure, since the mutation occurs in a predicted beta strand. Mimivirus The last substitution, a guanine to adenine change at position Within the first day after inoculation, cythopathogenic effects 17666 in ORF21, induces a change from an aspartic acid to the associated with Mimivirus lysis on amoeba appeared on culture polar amino acid asparagine, potentially affecting the putative plates as circular clear areas or lysing plaques. The plaques were function of the unknown protein encoded by this ORF. The almost identical for all deposits (Sputnik3-infected Mimivirus and

PLOS ONE | www.plosone.org 5 April 2013 | Volume 8 | Issue 4 | e61912  Mimiviridae Virophage

Figure 3. Sputnik 3 virophage isolation. (A) Schematic summarizing the protocol used in this study for the isolation of virophage. Observations of the virophage isolated in this study by ( B) transmission electron microscopy of negatively stained particles (A, scale bar 200 nm) and ( C) transmission electron microscopy of thin section of culture (B, scale bar 500 nm). ( D) Genome of Sputnik 3. doi:10.1371/journal.pone.0061912.g003 control Mimivirus with PAS) and extended until the third day may reflect the production of sputnik proteins without the (Figure S5a–b). No plaques appeared on negative controls (PAS- formation of mature particles. only and Sputnik3-only), even after 7 days. The plaque surfaces for The isolation of an additional virophage, Sputnik 3, using an the control Mimivirus-PAS deposits ranged from 0.63 cm 2 to original procedure confirms the capability of virophages to be 1.21 cm 2 with a mean area of 0.88 cm 2 (60.15 cm 2), whereas the cultured in viruses of the 3 groups of APM. We speculated that we plaques for the Sputnik3-infected Mimivirus deposits ranged from were not able to isolate the virus detected in the sample from 0.22 cm 2 to 0.72 cm 2 with a mean of 0.56 cm 2 (60.12 cm 2) which we isolated the virophage because the amoeba we used was (Figure S5c). These differences were significant according to the not susceptible to this virus [19]. We were able to propagate this 2 Student’s t-Test (p-value 4.49e 15 ). Cross dilutions of Mimivirus virophage on strains of group B and C viruses, demonstrating that and Sputnik3 revealed that plaque size was not dependent on Sputnik 3 is able to infect strains from all 3 groups of Mimiviridae. Sputnik3 concentration, but the deposits with 10 5 pfu/mL of Sputnik 3 is the first virophage recovered from group C, which Mimivirus plaques were smaller and difficult to measure. The contains Megavirus chilensis. The growth of Sputnik 3 in Moumou- effect of Sputnik3 on Mimivirus was confirmed by real-time PCR virus, the third group of Mimiviruses, further demonstrates that, on co-culture (Figure S6). like Sputnik 1 and 2, Sputnik 3 is able to infect the 3 groups of Mimivirus. This result supports the hypothesis of the ability of Discussion virophages to drive gene transfer, potentially contributing to the mosaicism of giant viruses’ genomes. This fuels the debate on the Our study shows that Sputnik 1 and Sputnik 2 are able to real nature of virophages[14]. The highly suspected deleterious replicate in all of the APM virus factories. Immunofluorescence effect of sputnik 3 on amoeba lysis associated with Mimivirus analyses confirmed the polar production of Sputnik viruses in the infection (Figures S5 and S6) argues against the theory of a simple periphery of the giant virus factory as described previously [15]. satellite. Our data suggest that the host range of virophages is The quantitative PCR results confirm these observations, although wider than previously thought. we observed a lower multiplication in the group C viruses Finally, the procedure reported here, which uses a cultivable Montpellier and Bus. However, virophages multiplied efficiently in helper giant virus, paves the way for the isolation and discovery of the group C viruses Courdo7, Courdo 11 and Terra1. We new virophages without the isolation of their giant virus host. hypothesized that the small Sputnik viral particles, trapped Isolating virophages using a reporter giant virus is significant between the surrounding fibers, are phagocytosed along with the because diverse virophage signatures have been identified in giant virus, as observed previously [15] because amoebas have the nearly all types of aquatic systems by searching the GOS data base capacity to phagocytize particles of a size greater than 0.5 mm, and thus are likely to play a key role in the ecosystem regulation of including latex beads [17]. In a recent study, we observed that aquatic environments by regulating host-virus interactions [15]. after 150 passages in its amoeba host, the genome of APM shifted Isolating other virophages using a Phycodnaviridae-algae co- dramatically from 1.2 to 0.993 Mbp and was associated with culture to grow those detected by metagenomic analysis only [11] deletions of genes encoding two proteins (R135 and L725) will facilitate our understanding of their biology, including their associated with wild-type virus fibrils. This APM clone lacking developmental cycle and host specificity. surrounding fibrils was not able to propagate the Sputnik virophages [18]. Using TEM, Montpellier and Bus were observed Supporting Information to have fibrils comparable to that of other APMs, including other group C APMs [9]. Thus, a difference in fibrils is not responsible Figure S1 Number of amoebas in PYG medium. A : The for the lower level of multiplication. The lack of efficient virophage number of viable amoebas present during the infection by giant multiplication in Montpellier and Bus seems associated with the viruses. B: The number of viable amoebas present during co- rapid amoeba lysis observed with these viruses, which leads to infection with giant viruses and Sputnik 1. C: The number of limited multiplication. The observed diffuse immunofluorescence axenic amoebas compared with the number of amoeba infected by Mamavirus (Green: Montpellier, Blue: Bus, Orange: Courdo 5,

PLOS ONE | www.plosone.org 6 April 2013 | Volume 8 | Issue 4 | e61912  Mimiviridae Virophage

Red: Pointerouge 1, Purple: Mamavirus, Black: control uninfected Bank at www.rcsb.org, ID: 3J26; [16]), green boxes indicate beta amoeba). Statistical analysis was performed according to the strand and blue boxes indicate alpha helix. Wilcoxon Rank Sum test performed with R software using the (TIF) package stats version 2.15.0 5 [Hollander, M. & Wolfe, D. A. Figure S5 Lysis plaque assay with Mimivirus and (1973) Nonparametric Statistical Methods (John Wiley and Sons). Sputnik3. (A) Scan of a colored lysis plaques with A polyphaga Bauer, F. B. (1972) Journal of the American Statistical Association 67, monolayer inoculated with Mimivirus (4 right spots, 3 to 6) and 27–33]. Mimivirus and Sputnik3 (2 left spots, 1 and 2) 3 days after (TIF) inoculation; (B) magnification of a right spot; ( C) difference of lysis Figure S2 Position of giant viruses detected in the soil plaques means measured on colored plates 3 days following compared to currently known Mimiviridae. Phylogenetic inoculation, between 34 deposits of Mimivirus and 34 deposits of trees based on partial polB gene sequence showing the position of Mimivirus and Sputnik (Spt3). giant viruses detected in the soil (T17,T16,T12, T32, T39, T71, (TIF) C5 and C6) compared to currently known members of the Mimiviridae family presented in our former study [La Scola B, et Figure S6 Quantification of Mimivirus and amoebal al. (2010) Tentative characterization of new environmental giant lysis. Quantification of Mimivirus by real-time PCR (curves), viruses by MALDI-TOF mass spectrometry. Intervirology 53:344– with Sputnik3 (red triangle) and without (blue square) from H0 to 53]. H24 post-infection, from co-culture in Acanthamoeba polyphaga in (TIF) PAS (non-infected amoebas were used as negative control and provided no amplification). The bars represent the number of Figure S3 Alignment of the genomes of the 3 isolates of amoebas for each time: Mimivirus-infected amoebas in blue, Sputnik. Mimivirus/Sputnik3-infected amoebas in red and non-infected (TIF) amoebas in green. Figure S4 Sequences of ORF17 (Gp17), ORF20 (Gp20) (TIF) and ORF21 (Gp21) in Sputnik, Sputnik2 and Sputnik3. The black line indicates that the sequence is the same that the one Author Contributions just above. The amino acids in red are those that are different Conceived and designed the experiments: BL. Performed the experiments: between the three Sputnik. In the Gp21 sequence, for which MG AC. Analyzed the data: MG GF DR BL. Contributed reagents/ predicted secondary structures are available (RCSB Protein Data materials/analysis tools: MG AC. Wrote the paper: MG IP BL.

References 1. Pagnier I, Raoult D, La Scola B (2008) Isolation and identification of amoeba- 10. Fischer MG, Suttle CA (2011) A virophage at the origin of large DNA resisting bacteria from water in human environment by using an Acanthamoeba transposons. Science 332: 231–234. polyphaga co-culture procedure. Environ Microbiol 10: 1135–1144. 11. Yau S, Lauro FM, DeMaere MZ, Brown MV, Thomas T et al. (2011) 2. Rowbotham TJ (1983) Isolation of Legionella pneumophila from clinical specimens Virophage control of antarctic algal host-virus dynamics. Proc Natl Acad via amoebae, and the interaction of those and other isolates with amoebae. J Clin Sci U S A 108: 6163–6168. Pathol 36: 978–986. 12. Krupovic M, Cvirkaite-Krupovic V (2012) Sputnik and Mavirus: not more than 3. La Scola B, Audic S, Robert C, Jungang L, de Lamballerie X et al. (2003) A satellite viruses. Nat Rev Microbiol 10: 78. doi:10.1038/nrmicro2676-c2. giant virus in amoebae. Science 299: 2033. 13. Fischer MG (2012) Sputnik and Mavirus: more than just satellite viruses. Nat 4. Xiao C, Chipman PR, Battisti AJ, Bowman VD, Renesto P et al. (2005) Cryo- Rev Microbiol 10: 78. doi:10.1038/nrmicro2676-c1. electron Microscopy of the Giant Mimivirus. J Mol Biol 353: 493–496. 14. Desnues C, Raoult D (2012) Virophages question the existence of satellites. Nat 5. Raoult D, Audic S, Robert C, Abergel C, Renesto P et al. (2004) The 1.2- Rev Microbiol 10: 234. megabase genome sequence of Mimivirus. Science 306: 1344–1350. 15. Desnues C, Raoult D (2010) Inside the lifestyle of the virophage. Intervirology 6. Arslan D, LeGendre M, Seltzer V, Abergel C, Claverie JM (2011) Distant 53: 293–303. Mimivirus relative with a larger genome highlights the fundamental features of 16. Zhang X, Sun S, Xiang Y, Wong J, Klose T et al. (2012) Structure of Sputnik, a Megaviridae. Proc Natl Acad Sci U S A 108(42): 17486–91. virophage, at 3.5-A ˚ resolution. Proc Natl Acad Sci U S A 45(109): 18431–18436. 7. Iyer LM, Balaji S, Koonin EV, Aravind L (2006) Evolutionary genomics of 17. Raoult D, Boyer M (2010) Amoebae as genitors and reservoirs of giant viruses. nucleo-cytoplasmic large DNA viruses. Virus Res 117: 156–184. Intervirology 53: 321–329. 8. La Scola B, Desnues C, Pagnier I, Robert C, Barrassi L et al. (2008) The 18. Boyer M, Azza S, Barrassi L, Klose T, Campocasso A et al. (2011) Mimivirus virophage as a unique parasite of the giant mimivirus. Nature 455: 100–104. shows dramatic genome reduction after intraamoebal culture. Proc Natl Acad 9. La Scola B, Campocasso A, N’Dong R, Fournous G, Barrassi L et al. (2010) Sci U S A 108: 10296–10301. Tentative characterization of new environmental giant viruses by MALDI-TOF 19. Suzan-Monti M, La Scola B, Raoult D (2006) Genomic and evolutionary aspects mass spectrometry. Intervirology 53: 344–353. of Mimivirus. Virus Res 117: 145–155.

PLOS ONE | www.plosone.org 7 April 2013 | Volume 8 | Issue 4 | e61912  Figure S1

 Figure S2

 Figure S3

 Figure S4

 Figure S5

 Figure S6

 Chapitre Cinq

Le virophage Zamilon

 5.1 – Article : une séquence d’un nouveau virophage associée à la spécificité d’hôtes Mimiviridae

A Novel Virophage Sequence is Associated With Mimiviridae Host Specificity. Morgan Gaia, Samia Benamar, Mondher Boughalmi, Isabelle Pagnier, Olivier Croce, Philippe Colson, Didier Raoult, Bernard La Scola*.

Pôle des Maladies Infectieuses, Assistance Publique-Hôpitaux de Marseille and URMITE UMR CNRS-IRD 7278, IFR48, Faculté de Médecine, Université de la Méditerranée, Marseille, France.

* Corresponding author: [email protected]

Submitted to PLoS ONE.

 Avant-propos

Les Mimiviridae ont été isolé à partir d’échantillons de différentes origines, qu’il s’agisse de lacs, de fontaines, de cours d’eau, et même de terre. Dans le but d’accroitre la collection de Mimiviridae , de nombreux échantillons sont couramment testés. Lorsqu’il y a forte suspicion de présence d’un virus géant – notable en culture par l’observation de lyse amibienne et par des colorations de GRAM et de Gimenez – des observations en microscopie électronique avec coloration négative sont réalisées. C’est dans ce cadre que f urent observées des particules de type Sputnik associées à un virus géant, à partir d’un échantillon de terre récupéré en Tunisie. Le virus géant, Mont1, fut identifié comme étant un Mimiviridae du groupe C. En revanche, les amorces basées sur la séquence du gène de la capside de Sputnik ne donnèrent aucune amplification par PCR, suggérant que le petit virus soit divergent. Un des objectifs de la thèse présentée ici a donc été de purifier et de caractériser ce petit virus. Apparenté au Sputnik, le virophage Zamilon a ainsi été décrit au travers de sa spécificité d’hôtes relativement restreinte, et de son génome unique.

 A Novel Virophage Sequence is Associated With Mimiviridae Host Specificity

Morgan Gaia, Samia Benamar, Mondher Boughalmi, Isabelle Pagnier, Olivier Croce,

Philippe Colson, Didier Raoult, Bernard La Scola

Pôle des Maladies Infectieuses, Assistance Publique-Hôpitaux de Marseille and URMITE

UMR CNRS-IRD 7278, IFR48, Faculté de Médecine, Université de la Méditerranée,

Marseille, France

Corresponding author: Prof. Bernard La Scola, Pôle des Maladies Infectieuses, Assistance

Publique-Hôpitaux de Marseille and URMITE UMR CNRS-IRD 7278, IFR48, Faculté de

Médecine, Université de la Méditerranée, Marseille, France. Tel: 33 4 91 32 43 75 Fax: 33 4

91 38 77 72 E-mail: [email protected]

Keywords: Giant viruses, virophages, Sputnik, APM, amoebae.

 ABSTRACT

Virophages, which are potentially important ecological regulators, have been discovered in association with members of the order Megavirales . Sputnik virophages target the

Mimiviridae , Mavirus was identified with the Cafeteria roenbergensis virus, and virophage genomes reconstructed by metagenomic analyses may be associated with the

Phycodnaviridae . Despite the fact that the Sputnik virophages were isolated with viruses belonging to group A of the Mimiviridae , they can grow in amoebae infected by Mimiviridae from groups A, B or C. In this study we describe Zamilon, the first virophage isolated with a member of group C of the Mimiviridae family. By co-culturing amoebae with purified

Zamilon, we found that the virophage is able to multiply with members of groups B and C of the Mimiviridae family but not with viruses from group A. Zamilon has a 17,276 bp DNA genome that potentially encoding 20 genes. Most of these genes are closely related to genes from the Sputnik virophage, but a phylogenetic analysis based on ORF19 clustered Zamilon together with groups B and C Mimiviridae and distant from the group A Mimiviridae and the related Sputnik virophages. This sequence could be a key feature of the specificity of this novel virophage by playing a critical but as yet unknown role in infection.

 INTRODUCTION

For over a decade, giant viruses of the nucleo-cytoplasmic large DNA virus (NCLDV) group have been extensively investigated [1–4]. This group is composed of the Poxviridae , which infect insects and vertebrates, the Asfarviridae , which infect swine, the Iridoviridae , which infect invertebrates and poikilothermic vertebrates, the closely related Ascoviridae , which infect insects, the Phycodnaviridae , which infect algae, and the putative

Marseilleviridae family, which infect amoebae [5]. The Acanthamoeba polyphaga Mimivirus

(APM) that was discovered in 2003 [6], as well as related viruses, cluster with the

Mimiviridae family [7] within the NCLDV group. It has been proposed that the viruses of the

NCLDV group should constitute a new viral order called Megavirales [8,9]. The Mimiviridae family consists of 2 groups: the first group includes Mimivirus -like viruses that infect amoebae (Group I), while the second group is composed solely of the Cafeteria roenbergensis virus (CroV), which infects a marine heterotrophic bi-flagellate [10 ]. The Mimiviridae group currently consists of more than 40 Mimivirus -like viruses that infect the widespread amoeba genus Acanthamoeba and have been arranged in three lineages based on their pol B gene sequences [11 ,12 ]. These lineages correspond to group A (which includes Mimivirus and

Mamavirus), group B (Moumouvirus) [13 ], and group C ( Megavirus chiliensis ) [14 ]. The exploration of new giant viruses recently led to the identification of [15 ], which infect amoebae. Pandoraviruses form particles of approximately 1 µm that contain a

2.5 Mb-long DNA genome, in contrast to Mimivirus , which has 0.7 µm particles and a 1.2 Mb genome [16 ]. Pandoraviruses seem to cluster with the NCLDVs.

In 2008, a small virus with 50 nm icosahedral virions and an 18 kb genome was co- isolated with Mamavirus. This small virus infects the virus factory in which the giant virus replicates [17 ]. Due to its negative impact on the giant virus host, which is characterized by an increased production of abnormal particles and a decrease in infectivity and lytic ability, this

 small virus was dubbed the Sputnik virophage [18 ]. A second Sputnik strain, named Sputnik

2, was later isolated with the giant Lentillevirus, which, like Mamavirus, belongs to group A of the Mimiviridae [7,19 ]. A third strain, Sputnik 3, was isolated with a Mimivirus reporter instead of with its natural viral host [11 ]. The three Sputnik virophages, which share more than 99% identity, have a broad host spectrum among the Mimiviridae and can replicate with viruses belonging to groups A, B and C [11 ]. In 2011, a new virophage was isolated in association with Cafeteria roenbergensis virus [20 ]. The detection from environmental datasets of abundant virophage genomes associated with Phycodnaviridae or phytoplankton- infecting viruses led to the hypothesis that virophages play an important ecological role in the regulation of viral populations [21 –25 ]. However, the nature of the virophages is still a matter of debate [26 ].

The aim of this study was to identify and characterize a new Sputnik-like virophage that was isolated previously [27 ]. This virophage was discovered in association with the

Mont1 virus. Based on the Mont1 pol B gene sequence, this Mimiviridae family member belongs to group C. The virophage is thus the first described associated to this group.

Amoebal cultures were co-infected with Zamilon and viruses from groups A, B and C to investigate the fitness of the virophage by transmission electron microscopy and real-time

PCR. The virophage genome was then sequenced and analyzed.

MATERIAL & METHODS

Isolation of Mont1 and Zamilon

The giant virus Mont1 and its virophage were isolated from a soil sample collected in Tunisia using a high-throughput protocol, as described previously [27 ]. Mont1 was identified a

Mimivirus-like virus, with particles of approximately 500 nm in diameter surrounded by fibrils. The virus was classified as a group C Mimiviridae based on a partial sequence of the

 polymerase B gene (GenBank Accession No. JX484142). The virophage was Sputnik-like, i.e., had a spherical particle with a diameter of approximately 60 nm, and thus was initially named “Sputnik 4” . Based on its features, the virophage was named Zamilon.

Purification

To purify the giant virus, supernatants from amoebal co-cultures in PAS (Page’s amoeba saline buffer) containing both Mont1 and Zamilon viruses were heat-inactivated. After 2 hours at 65°C, the absence of virophage particles was verified by negative staining electron microscopy, and the suspension was serially diluted up to 10 -10 in PAS. End-point dilutions were performed in fresh Acanthamoeba polyphaga (strain Linc AP-1) cultures at a concentration of 5x10 5 cells/mL. The most dilute sample that induced lysis was sub-cultured with fresh amoebae. The absence of virophage particles was verified again by negative staining electron microscopy, and the purified cultures containing only the giant virus Mont1 in A. polyphaga were stored at -80°C.

The Zamilon virophage was purified from a large volume (approximately 1.5 L) of supernatant from amoebae co-cultured with Mont1 and Zamilon in PYG. The supernatant was successively filtered through 0.8-, 0.45- and 0.22-µm membranes. The filtrate was then concentrated by ultracentrifugation at 25,000 rpm for 1.5 h. The pellet was resuspended in 1 mL of PAS and then purified over a 15% sucrose gradient by centrifuging at 25,000 rpm for

1.5 h. The highly concentrated pellet was resuspended in 1 mL of PAS and stored at -80°C after the absence of Mont1 particles was confirmed by negative staining electron microscopy.

Co-culture of the virophage

A. polyphaga was inoculated with the Zamilon virophage and several giant viruses that are representative of the 3 groups of Mimiviridae (APM and Mamavirus for group A,

Moumouvirus and Monve virus for group B, and Courdo11 virus and Terra1 virus for group

 C). Zamilon was also co-cultured with its native host, Mont1 (group C Mimiviridae ). The co- cultures were performed as previously described [11 ]. Briefly, 1 mL of filtered Mimiviridae and 100 µL of the purified virophage diluted 10-fold were added to 10 mL of fresh amoebae in PAS (5x10 5 cells/mL). After 1 h of incubation at 32°C, the supernatant was removed, and the pellet was resuspended in 10 mL of PAS. The culture flasks were then incubated at 32°C.

This time point was defined as H0. A 1 mL sample of each co-culture was removed at time point H16 (i.e., after 16 h of incubation) for transmission electron microscopy analysis, except for Mont1 and Mamavirus, for which samples were taken at H6, H8, H12, H16, H24 and H30.

The number of amoebae was quantified at each time point using a KOVA Glasstic Slide

(Hycor Biomedical Inc., Garden Grove, California, USA).

Plaque assay

Plaque assays were performed as previously described [11 ] with suspensions of Mont1 or

Mamavirus with or without the Zamilon virophage. The virophage used in these experiments was diluted 10-fold from the frozen purified stocks. For each assay, 7.5 µL of the giant virus at a concentration of 10 9 pfu/mL was combined with 7.5 µL of the diluted virophage or 7.5

µL of PAS and added to the plates. The plates were monitored daily for plaque formation, and diameters of the plaques were measured with calipers.

Real-time PCR

DNA extractions and real-time PCR were performed using 200 µL of each co-culture taken at

H0 and H16. Additional samples were taken from the co-cultures of Zamilon with Mont1 or

Mamavirus at H6, H8, H12, H24 and H30. The EZ1 DNA Tissue Kit (Qiagen, Hilden,

Germany) was used for DNA extraction according to the manufacturer’s instructions. A

LightCycler 480 SYBR Green I Master (Roche Applied Science) was used to perform the real-time PCR according to the manufacturer’s instructions. The following primers were used

 to detect the Zamilon virophage: f orward primer 5’ -GGGATGAACATCAAGCTGGT- 3’ and reverse primer 5’ - GGGTTGTTGGAAGCTGACAT- 3’.

Sequencing and bioinformatics analysis

The Zamilon genome sequence was obtained using a MiSeq sequencer (Illumina), the MIRA assembler [28 ] and CLC Genomics Workbench version 4.9 (CLC BIO Aarhus, Denmark).

Gene predictions were performed using GeneMarkS [29 ] and Prodigal [30 ] software with default parameters. The genome was annotated manually based on protein homology using

BLASTp searches (Basic Local Alignment Search Tool) against the non-redundant protein collection in the NCBI database (http://http://blast.ncbi.nlm.nih.gov/Blast.cgi) and conserved domains were predicted using BLASTp, PSI-BLAST (Position-Specific Iterated BLAST) and

InterProScan [31 ]. Nucleotide sequence comparisons were made using BLASTn searches against the nucleotide collection in the NCBI database. The genome architecture of the virophages and the Mimiviridae family members were compared using MUMmer [32 ].

Multiple sequence alignments were performed using MUSCLE [33 ] and curated using

Gblocks [34 ]. Phylogenetic trees were constructed using the PhyML Maximum Likelihood algorithm [35 ]. The trees were visualized using MEGA v5 [36 ].

RESULTS

Selective growth of the virophage in Mimiviridae

Seven viruses from our laboratory’s collection of the giant viruses were co-cultured with the

Zamilon virophage: 2 viruses belonging to group A, 2 group B viruses, and 3 viruses from group C of the Mimiviridae . Based on real-time PCR and transmission electron microscopy analysis, Zamilon grew well with all of the viruses from group B and group C, but not with those belonging to group A of the Mimiviridae (Figure 1). Electron microscopy revealed that the virophage particles were spherical, with a diameter of approximately 50 to 60 nm. The

 particles appeared in the cytoplasm of the amoebae and were produced from the virus factory when the amoebae were co-infected with group B and group C Mimiviridae . Relative quantification by real-time PCR showed that the rate of virophage multiplication depends on the giant virus, with the higher titers observed when the virophage was co-cultured with viruses from group C.

The Zamilon genome

We sequenced the 17,276 bp circular genome of the Zamilon virophage (EMBL-EBI ID

HG531932.1). The GC content was 29.7%. Analysis of the whole genome at the nucleotide level showed that it is most similar to the Sputnik virophage (76% identity, 75% coverage). A total of 20 ORFs (Open Reading Frame) were identified by gene prediction and ranged from

222 bp to 2337 bp in length (Figure 2). Most of these ORFs had significant homology to predicted Sputnik virophage protein sequences (Table 1). However, a genomic dot plot of

Zamilon and the Sputnik virophage showed that Zamilon constitutes a new virophage and that its genome contains a reversed portion from approximately 6,000 bp to the end

(Supplementary Figure 1A). Reversed nucleotide sequences are also evident in the genome of a group A Mimiviridae family member ( Mimivirus ) compared to the genomes of virophages from groups B and C (Moumouvirus and Megavirus chiliensis , respectively) (Supplementary

Figure 1).

Fifteen different ORFs (ORF4-ORF7, ORF9-ORF18 and ORF20) showed between 40 and

80% homology in amino acids to predicted genes from the Sputnik virophage. Some of the predicted proteins have functions, including a putative transposase, capsid-forming proteins, a collagen-like protein, a helicase, an integrase and an ATPase (Supplementary Figure 2). The closest homolog to ORF12 was the Sputnik V9 gene, which encodes an unidentified protein.

However, BLAST alignments showed that this ORF was also related to a putative cysteine

 protease protein from the Mavirus virophage (32% identity and 83%coverage, E-value 4 -17 ;

GenBank Accession No. YP_004300284.1). Similarly, ORF17 is related to the uncharacterized Sputnik V4 gene and also to a zinc-finger C2H2-type domain-containing protein. Phylogenetic analysis of ORF11 and ORF18 confirmed that they are closely related to Sputnik genes that potentially encode an integrase and a DNA-packaging protein with a putative ATPase domain, respectively (Figure 2). The predicted protein sequence encoded by

ORF9, which encodes a putative helicase, also shows homology to the Organic Lake

Virophage putative DNA primase/polymerase (GenBank Accession No. ADX05784.1) and to the putative DNA primase from the virophage associated with Phaeocystis globosa (GenBank

Accession No. YP_008059889.1) (Figure 2).

The Zamilon ORF3 shows significant homology to the Megavirus chiliensis mg3 gene.

Annotation of ORF19 showed homology to the Megavirus chiliensis mg664 gene, and this homology was further confirmed by phylogenetic analysis (Figure 2). This ORF clustered closer to the group B and C Mimiviridae viruses than to the group A Mimiviridae viruses and their associated Sputnik virophages. Significant homology was detected between the Zamilon

ORF8 protein sequence and the hypothetical protein tv_L8 from the transpoviron

Moumouvirus Monve, which does not have a predicted function.

ORF1 and ORF2 did not exhibit significant similarity or homology to any entries in the NCBI databases and were therefore aligned with the Sputnik virophage genome (GenBank

Accession No. NC_011132.1). Zamilon ORF1 and ORF2 showed some homology to the

Sputnik V15 and V2 genes, respectively ( ≥ 30% identity ; E-values 0.081 and 7 -06 , respectively). The predicted protein sequence encoded by ORF1 contained a putative conserved protein domain related to a transmembrane domain from cytochrome c oxydase subunit II ( E-value 2.3e-4; EMBL-EBI ID: IPR011759).

 The impact of the Zamilon virophage on its host

There was no significant difference in the diameters of the plaques formed by the giant viruses Mont1 and Mamavirus whether they were co-inoculated with or without the Zamilon virophage. However, Mont1 formed sun-like plaques, while Mamavirus formed rounded plaques (Supplementary Figure 3). Transmission electron microscopy revealed a high proportion of abnormal Mont1 particles when the virus was co-cultured with the Zamilon virophage (Figure 3A). However, a similar number of abnormal particles were observed when

Mont1 was cultured alone (Figure 3B). When the Zamilon virophage was co-cultured with other members of Mimiviridae groups B and C, the proportion of abnormal giant particles produced remained unchanged. Moreover, co-culture with Zamilon had no effect the ability of

Mont1 or Mamavirus to induce lysis in the amoebal host (Figure 3C).

DISCUSSION

In this study, we report the identification and characterization of a novel virophage,

Zamilon, that is associated with giant viruses from the Mimiviridae family. This virophage is closely related to the Sputnik virophages and is the first virophage isolated together with a member of Mimiviridae group C.

The Zamilon virophage particles are spherical, with a diameter of 50 to 60 nm, similar to Sputnik, Mavirus, and Organic Lake virophage particles [17 ,20 ,21 ]. The Zamilon genome is also similar in size to known virophage genomes, as it is approximately 17 kb and contains

20 putative ORFs. In comparison, the Sputnik genome is 18 kb long and contains 20 ORFs, and the Mavirus genome is 19 kb long and encodes 20 ORFs [17 ,20 ]. Virophage genomes detected by metagenomic analysis of environmental samples exhibit similar characteristics.

The Organic Lake virophage has a 26 kb genome that contains 26 predicted genes, and the 19 kb genome of the Phaeocystis globosa virus virophage encodes 16 proteins [21 ,24 ]. Other

 virophage genomes constructed from metagenomic datasets range from 17 to 27 kb and contain 21 to 26 ORFs [22 ].

Although the Zamilon virophage is related to the Sputnik virophage, there are significant genomic differences. In particular, a significant portion of the Zamilon genome is inverted compared to the Sputnik genome. A similar inversion is also evident when comparing the genomes of group A Mimiviridae viruses to groups B and C. The conservation of this pattern between groups B and C viruses and their associated virophages suggests that virophages are as closely related to each other as they are as to their respective hosts.

Several Zamilon virophage ORFs encode proteins that share sequence homology or conserved domains with predicted proteins from several virophages, such as ATPase, helicase, integrase, transposase and capsid proteins [11 ,17 ,20 ,21 ] (Supplementary Figure 2).

These genes could therefore represent a set of core functions for these viruses. In addition, most of the predicted Zamilon virophage proteins exhibited moderate to high homology to predicted Sputnik virophage proteins, as well as proteins encoded by Megavirus chiliensis and the Moumou monve transpoviron. While the predicted protein homolog to a protein of the

Moumou monve transpoviron shows similarity to the Sputnik virophage genome, the

Megavirus chiliensis homologs do not have such similarity with other virophages.

In contrast to the Sputnik virophages [11 ], Zamilon does not seem to have a significant impact on the giant virus host. Indeed, the Sputnik virophage increases the rate of abnormal giant virus particles and decreases the lytic and infective capacity of the giant virus [17 ].

Zamilon did not affect the rate of abnormal particle formation, nor did it affect the ability of the giant virus to lyse infected amoebae. The potential impact of Zamilon on giant virus infectivity remains to be investigated.

 The three Sputnik virophages that were previously described have a broad host spectrum and can replicate with Mimiviridae from groups A, B and C [11 ]. The Sputnik virophage and Sputnik 2 were isolated with Mamavirus and Lentillevirus, respectively, both of which are group A Mimiviridae [7,17 ]. Sputnik 3, however, was isolated alone, without a giant virus, and was presumed to be associated with a member of the group C Mimiviridae

[11 ]. Despite these differences, the 3 Sputnik strains share more than 99% similarity. In contrast, the Zamilon virophage was isolated with Mont1, a group C Mimiviridae , and is thus the first virophage known to be associated with this group [27 ]. This novel virophage is un able to grow in association with group A Mimiviridae (as assessed by transmission electron microscopy and real-time PCR), despite its similarity to Sputnik virophages. The Zamilon virophage contains predicted genes that are related to Megavirus chiliensis , a group C

Mimiviridae , but are not present in the Sputnik virophages. These genes could explain this novel host specificity. In particular, Zamilon ORF19 appears to be closely related to

Megavirus chiliensis mg664. Phylogenetic analysis of this ORF showed that it clusters closer to members of Mimiviridae group B and C than to the Sputnik virophages associated with group A. We did not identify a putative function or any conserved protein domains for this

ORF. Thus, it is difficult to evaluate the potential role that this gene plays in host virus genotype specificity, although we speculate that it is likely to be a major factor in this selectivity. This host selectivity has also been described in bacteriophages, as some bacteriophages are specific to a single bacterial species within a microbial community or even only a few strains within a single species [37 –39 ]. Changes in bacteriophages host ranges could arise due to nucleotide or protein mutations [40 ,41 ]. These mutations could induce changes in the balance between phage-infectivity and host-resistance, which could lead to host specificity [42 ,43 ]. Infectivity requires several steps that are shared by all viruses, including virophages, from recognition of the host, to entry and transport to the replication

 compartment, to replication itself. We hypothesize that the Zamilon virophage ORF19, which clusters with the group B and C Mimiviridae , plays a role in one of these stages of infection.

The mechanism and timing of viral host selection remains unknown. The Sputnik virophages presumably take advantage of the phagocytosis of their giant virus hosts to enter the amoebal host [44 ]. Indeed, amoebae from the Acanthamoeba genus can internalize even particles greater than 0.5 µm in diameter, including latex beads, and it has been hypothesized that the Sputnik virophages penetrate the amoeba by attaching themselves to Mimiviridae fibrils during phagocytosis [44 ,45 ]. Structural studies have revealed fibers protruding from the surface of Sputnik that do not have a clear function and may be associated with this host virus recognition [46 ,47 ]. The host virus genotype specificity exhibited by Zamilon may involve recognition of a specific pattern on the surface of the giant virus. Once internalized, the

Sputnik virophages multiply in the viral factory formed by the associated giant viruses

[17 ,20 ,26 ]. However, as the interactions between functional Mimiviridae proteins and the virophages during replication are not clearly identified, specific sequence recognition cannot be ruled out.

Virophages are suspected to be key players in the regulation of environmental virus populations [21 ]. They may reduce the infectivity, and thus the reproductive fitness, of viruses, thus decreasing host mortality [22 ,23 ,25 ]. This regulation of global ecology through virus-induced cell lysis suggests that more virophage lineages remain to be discovered that target viruses implicated in environmental ecologies. Our results show that, even within a single lineage, virophages are more complex than initially thought and can target specific genotypes within in a virus family.

 The host-specificity of the Zamilon virophage supports the distinction between satellite viruses (opportunistic entities associated with a virus) and virophages, which target specific hosts.

REFERENCES

1. Iyer LM, Aravind L, Koonin EV (2001) Common origin of four diverse families of large eukaryotic DNA viruses. J Virol 75: 11720-11734. 10.1128/JVI.75.23.11720-11734.2001 [doi].

2. Yutin N, Wolf YI, Raoult D, Koonin EV (2009) Eukaryotic large nucleo-cytoplasmic DNA viruses: clusters of orthologous genes and reconstruction of viral genome evolution. Virol J 6: 223. 1743-422X-6-223 [pii];10.1186/1743-422X-6-223 [doi].

3. Van Etten JL, Lane LC, Dunigan DD (2010) DNA viruses: the really big ones (giruses). Annu Rev Microbiol 64: 83-99. 10.1146/annurev.micro.112408.134338 [doi].

4. Boyer M, Madoui MA, Gimenez G, La Scola B, Raoult D (2010) Phylogenetic and phyletic studies of informational genes in genomes highlight existence of a 4 domain of life including giant viruses. PLoS One 5: e15530. 10.1371/journal.pone.0015530 [doi].

5. Colson P, Pagnier I, Yoosuf N, Fournous G, La Scola B, Raoult D (2013) "Marseilleviridae", a new family of giant viruses infecting amoebae. Arch Virol 158: 915-920. 10.1007/s00705-012-1537-y [doi].

6. La Scola B, Audic S, Robert C, Jungang L, de Lamballerie X, Drancourt M, Birtles R, Claverie JM, Raoult D (2003) A giant virus in amoebae. Science 299: 2033. 10.1126/science.1081867 [doi];299/5615/2033 [pii].

7. La Scola B, Campocasso A, N'Dong R, Fournous G, Barrassi L, Flaudrops C, Raoult D (2010) Tentative characterization of new environmental giant viruses by MALDI-TOF mass spectrometry. Intervirology 53: 344-353. 000312919 [pii];10.1159/000312919 [doi].

8. Legendre M, Arslan D, Abergel C, Claverie JM (2012) Genomics of Megavirus and the elusive fourth domain of Life. Commun Integr Biol 5: 102-106.

9. Colson P, de Lamballerie X, Fournous G, Raoult D (2012) Reclassification of giant viruses composing a fourth domain of life in the new order Megavirales. Intervirology 55: 321-332. 000336562 [pii];10.1159/000336562 [doi].

 10. Fischer MG, Allen MJ, Wilson WH, Suttle CA (2010) Giant virus with a remarkable complement of genes infects marine zooplankton. Proc Natl Acad Sci U S A 107: 19508-19513. 1007615107 [pii];10.1073/pnas.1007615107 [doi].

11. Gaia M, Pagnier I, Campocasso A, Fournous G, Raoult D, La Scola B (2013) Broad spectrum of mimiviridae virophage allows its isolation using a mimivirus reporter. PLoS One 8: e61912. 10.1371/journal.pone.0061912 [doi];PONE-D- 12-24503 [pii].

12. Pagnier I, Ikanga Reteno DG, Saadi H, Boughalmi M, Gaia M, Slimani M, Ngounga T, Bekliz M, Colson P, Raoult D, La Scola B (2013) A decade of improvements in Mimiviridae and Marseilleviridae isolation from amoeba. Intervirology 56: In press.

13. Yoosuf N, Yutin N, Colson P, Shabalina SA, Pagnier I, Robert C, Azza S, Klose T, Wong J, Rossmann MG, La Scola B, Raoult D, Koonin EV (2012) Related giant viruses in distant locations and different habitats: Acanthamoeba polyphaga moumouvirus represents a third lineage of the Mimiviridae that is close to the megavirus lineage. Genome Biol Evol 4: 1324-1330. evs109 [pii];10.1093/gbe/evs109 [doi].

14. Arslan D, Legendre M, Seltzer V, Abergel C, Claverie JM (2011) Distant Mimivirus relative with a larger genome highlights the fundamental features of Megaviridae. Proc Natl Acad Sci U S A 108: 17486-17491. 1110889108 [pii];10.1073/pnas.1110889108 [doi].

15. Philippe N, Legendre M, Doutre G, Coute Y, Poirot O, Lescot M, Arslan D, Seltzer V, Bertaux L, Bruley C, Garin J, Claverie JM, Abergel C (2013) Pandoraviruses: amoeba viruses with genomes up to 2.5 Mb reaching that of parasitic eukaryotes. Science 341: 281-286. 341/6143/281 [pii];10.1126/science.1239181 [doi].

16. Suzan-Monti M, La Scola B, Raoult D (2006) Genomic and evolutionary aspects of Mimivirus. Virus Res 117: 145-155. S0168-1702(05)00237-6 [pii];10.1016/j.virusres.2005.07.011 [doi].

17. La Scola B, Desnues C, Pagnier I, Robert C, Barrassi L, Fournous G, Merchat M, Suzan-Monti M, Forterre P, Koonin E, Raoult D (2008) The virophage as a unique parasite of the giant mimivirus. Nature 455: 100-104. nature07218 [pii];10.1038/nature07218 [doi].

18. Claverie JM, Abergel C (2009) Mimivirus and its virophage. Annu Rev Genet 43: 49- 66. 10.1146/annurev-genet-102108-134255 [doi].

19. Desnues C, La Scola B, Yutin N, Fournous G, Robert C, Azza S, Jardot P, Monteil S, Campocasso A, Koonin EV, Raoult D (2012) Provirophages and transpovirons as the diverse mobilome of giant viruses. Proc Natl Acad Sci U S A 109: 18078-18083. 1208835109 [pii];10.1073/pnas.1208835109 [doi].

20. Fischer MG, Suttle CA (2011) A virophage at the origin of large DNA transposons. Science 332: 231-234. science.1199412 [pii];10.1126/science.1199412 [doi].

 21. Yau S, Lauro FM, Demaere MZ, Brown MV, Thomas T, Raftery MJ, Andrews- Pfannkoch C, Lewis M, Hoffman JM, Gibson JA, Cavicchioli R (2011) Virophage control of antarctic algal host-virus dynamics. Proc Natl Acad Sci U S A 108: 6163-6168. 1018221108 [pii];10.1073/pnas.1018221108 [doi].

22. Zhou J, Zhang W, Yan S, Xiao J, Zhang Y, Li B, Pan Y, Wang Y (2013) Diversity of virophages in metagenomic data sets. J Virol 87: 4225-4236. JVI.03398-12 [pii];10.1128/JVI.03398-12 [doi].

23. Wodarz D (2013) Evolutionary dynamics of giant viruses and their virophages. Ecol Evol 3: 2103-2115. 10.1002/ece3.600 [doi].

24. Santini S, Jeudy S, Bartoli J, Poirot O, Lescot M, Abergel C, Barbe V, Wommack KE, Noordeloos AA, Brussaard CP, Claverie JM (2013) Genome of Phaeocystis globosa virus PgV-16T highlights the common ancestry of the largest known DNA viruses infecting eukaryotes. Proc Natl Acad Sci U S A 110: 10800- 10805. 1303251110 [pii];10.1073/pnas.1303251110 [doi].

25. Jassim SA, Limoges RG (2013) Impact of external forces on cyanophage-host interactions in aquatic ecosystems. World J Microbiol Biotechnol 29: 1751- 1762. 10.1007/s11274-013-1358 -5 [doi].

26. Gaia M, Colson P, Desnues C, La Scola B (2013) Virophage concept, the. eLS . 10.1002/9780470015902.a0024410.

27. Boughalmi M, Saadi H, Pagnier I, Colson P, Fournous G, Raoult D, La Scola B (2013) High-throughput isolation of giant viruses of the Mimiviridae and Marseilleviridae families in the Tunisian environment. Environ Microbiol 15: 2000-2007. 10.1111/1462-2920.12068 [doi].

28. Chevreux B, Wetter T, Suhai S (1999) Genome sequence assembly using trace signals and additional sequence information. Computer Science and Biology: Proceedings of the German Conference on Bioinformatics (GCB) 99: 45-56.

29. Besemer J, Lomsadze A, Borodovsky M (2001) GeneMarkS: a self-training method for prediction of gene starts in microbial genomes. Implications for finding sequence motifs in regulatory regions. Nucleic Acids Res 29: 2607-2618.

30. Hyatt D, Chen GL, Locascio PF, Land ML, Larimer FW, Hauser LJ (2010) Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics 11: 119. 1471-2105-11-119 [pii];10.1186/1471-2105-11-119 [doi].

31. Quevillon E, Silventoinen V, Pillai S, Harte N, Mulder N, Apweiler R, Lopez R (2005) InterProScan: protein domains identifier. Nucleic Acids Res 33: W116- W120. 33/suppl_2/W116 [pii];10.1093/nar/gki442 [doi].

32. Kurtz S, Phillippy A, Delcher AL, Smoot M, Shumway M, Antonescu C, Salzberg SL (2004) Versatile and open software for comparing large genomes. Genome Biol 5: R12. 10.1186/gb-2004-5-2-r12 [doi];gb-2004-5-2-r12 [pii].

 33. Edgar RC (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res 32: 1792-1797. 10.1093/nar/gkh340 [doi];32/5/1792 [pii].

34. Talavera G, Castresana J (2007) Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments. Syst Biol 56: 564-577. 780704285 [pii];10.1080/10635150701472164 [doi].

35. Guindon S, Delsuc F, Dufayard JF, Gascuel O (2009) Estimating maximum likelihood phylogenies with PhyML. Methods Mol Biol 537: 113-137. 10.1007/978-1- 59745-251-9_6 [doi].

36. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S (2011) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol Biol Evol 28: 2731-2739. msr121 [pii];10.1093/molbev/msr121 [doi].

37. Hyman P, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70: 217-248. S0065-2164(10)70007-1 [pii];10.1016/S0065- 2164(10)70007-1 [doi].

38. Duplessis M, Moineau S (2001) Identification of a genetic determinant responsible for host specificity in Streptococcus thermophilus bacteriophages. Mol Microbiol 41: 325-336. mmi2521 [pii].

39. Koskella B, Meaden S (2013) Understanding bacteriophage specificity in natural microbial communities. Viruses 5: 806-823. v5030806 [pii];10.3390/v5030806 [doi].

40. Duffy S, Burch CL, Turner PE (2007) Evolution of host specificity drives reproductive isolation among RNA viruses. Evolution 61: 2614-2622. EVO226 [pii];10.1111/j.1558-5646.2007.00226.x [doi].

41. Scanlan PD, Hall AR, Lopez-Pascua LD, Buckling A (2011) Genetic basis of infectivity evolution in a bacteriophage. Mol Ecol 20: 981-989. 10.1111/j.1365-294X.2010.04903.x [doi].

42. Riede I, Degen M, Henning U (1985) The receptor specificity of bacteriophages can be determined by a tail fiber modifying protein. EMBO J 4: 2343-2346.

43. Rakhuba DV, Kolomiets EI, Dey ES, Novik GI (2010) Bacteriophage receptors, mechanisms of phage adsorption and penetration into host cell. Pol J Microbiol 59: 145-155.

44. Desnues C, Boyer M, Raoult D (2012) Sputnik, a virophage infecting the viral domain of life. Adv Virus Res 82: 63-89. B978-0-12-394621-8.00013-3 [pii];10.1016/B978-0-12-394621-8.00013-3 [doi].

45. Ghigo E, Kartenbeck J, Lien P, Pelkmans L, Capo C, Mege JL, Raoult D (2008) Ameobal pathogen mimivirus infects macrophages through phagocytosis. PLoS Pathog 4: e1000087. 10.1371/journal.ppat.1000087 [doi].

 46. Sun S, La Scola B, Bowman VD, Ryan CM, Whitelegge JP, Raoult D, Rossmann MG (2010) Structural studies of the Sputnik virophage. J Virol 84: 894-897. JVI.01957-09 [pii];10.1128/JVI.01957-09 [doi].

47. Zhang X, Sun S, Xiang Y, Wong J, Klose T, Raoult D, Rossmann MG (2012) Structure of Sputnik, a virophage, at 3.5-A resolution. Proc Natl Acad Sci U S A 109: 18431-18436. 1211702109 [pii];10.1073/pnas.1211702109 [doi].

Figure 1. Zamilon growth in Mimiviridae . (A) Histogram of Zamilon growth in group A, B and C Mimiviridae family members, measured by real-time PCR. The difference in the Cycle threshold (Ct) between time points H0 and H16 is shown. ( B-E) Transmission electron microscopy images of the virus factory in amoebae co-infected with Zamilon and

Mimiviridae . No virophage particles were detected in the presence of Mimivirus ( B; scale bar

0.1 µm), unlike Moumouvirus ( C; scale bar 0.1 µm), Terra1 ( D; scale bar 0.1 µm) and Mont1

(E; scale bar 0.1 µm).

Figure 2. The Zamilon genome. The Zamilon genome, with predicted coding sequences on the forward strand (blue) and the reverse strand (red). Phylogenetic analyses of ORF6, 9, 11,

12, 18 and 19 are included with bootstrap values indicated (cutoff ≥ 50). * indicates the best hit ( E-values: 0, 0, 8 -80 , 2-90 , 5-147 , and 5 -34 , respectively).

Figure 3. Impact of the Zamilon virophage. (A-B) Transmission electron microscopy images of abnormal Mont1 virus particles (arrows) produced from the virus factory with ( A; scale bar 0.1 µm) and without ( B; scale bar 0.1 µm) Zamilon. ( C) Kinetics of survival of amoebae infected with Mont1 or Mamavirus, with or without the Zamilon virophage (Blue:

Mont1, Red: Mont1 and Zamilon, Green: Mamavirus, Purple: Mamavirus and Zamilon,

 Turquoise: negative control). The x-axis shows the time points, and the y-axis shows the number of amoebae per milliliter (x10 5 cells/mL).

Table 1. Closest homologs of the Zamilon open reading frames (ORFs). Best hit for

Zamilon’s ORFs obtained with BlastP against the non-redundant (nr) NCBI database.

 Figure 1

 Figure 2

 Table 1

ORF (size in Closest homolog in GenBank nr (accession no.) Identity / E-value Predicted function amino acids) alignment length ORF1 (111)* Sputnik virophageV15 (YP_002122376.1)* 32% / 97 0.081* hypothetical protein* ORF2 (73)* Sputnik virophage V2 (YP_002122363.1)* 31% / 62 7e -6* hypothetical protein* ORF3 (135) Megavirus chiliensis mg3 gene product (YP_004894054.1) 67% / 52 9e -14 hypothetical protein ORF4 (221) Sputnik virophage 2 putative IS3 family transposase A protein 40% / 48 0.003 putative transposase (AFH75271.1) ORF5 (376) Sputnik virophage 2 minor virion protein (3J26_N) 66% / 375 2e -178 minor virion protein ORF6 (609) Sputnik virophage putative capsid protein V20 86% / 609 0 capsid protein (YP_002122381.1) ORF7 (442) Sputnik virophage V21 (YP_002122382.1) 70% / 442 0 hypothetical protein ORF8 (81) Moumouvirus Monve hypothetical protein tv_L8 (AEY99266.1) 72% / 53 4e -18 hypothetical protein ORF9 (778) Sputnik virophage V13 (YP_002122374.1) 67% / 778 0 putative helicase ORF10 (168) Sputnik virophage V11 (YP_002122372.1) 53% / 165 6e -44 hypothetical protein ORF11 (247) Sputnik virophage V10 (YP_002122371.1) 58% / 217 8e -80 putative integrase ORF12 (175) Sputnik virophage V9 (YP_002122370.1) 77% / 175 2e -90 hypothetical protein ORF13 (184) Sputnik virophage V8 (YP_002122369.1) 71% / 184 5e -82 structural protein ORF14 (241) Sputnik virophage V7 (YP_002122368.1) 80% / 241 3e -120 hypothetical protein ORF15 (305) Sputnik virophage V6 (YP_002122367.1) 75% / 314 9e -136 collagen -like protein ORF16 (121) Sputnik virophage V5 (YP_002122366.1) 59% / 86 1e -31 hypothetical protein ORF17 (133) Sputnik virophage V4 (YP_002122365.1) 55% / 143 5e -44 hypothetical protein ORF18 (245) Sputnik virophage V3 (YP_002122364.1) 81% / 245 5e -147 DNA packaging - ATPase ORF19 (147) Megavirus chiliensis mg664 gene product (YP_004894715.1) 50% / 129 5e -34 hypothetical protein ORF20 (147) Sputnik virophage V1 (YP_002122362.1) 60% / 126 2e -18 hypothetical protein

Hypothetical functions were determined by homology and conservation of protein domains. * indicates ORFs with no significant homology in the nr database. These ORFs were aligned directly to the Sputnik virophage.

 Figure 3

 SUPPLEMENTARY MATERIAL

Supplementary figure 1. Comparisons of virophages and Mimiviridae genomes. (A)

Comparison of the Zamilon genome to the Sputnik genome. ( B-D) Comparisons of

Mimiviridae genomes depending on the group they belong to: group A Mimivirus compared to group B Moumouvirus (B), group A Mimivirus compared to group C Megavirus chiliensis

(C), and group B Moumouvirus compared to group C Megavirus chiliensis (D).

Supplementary figure 2. Putative functions in virophages. Genes encoding hypothetical and putative functions shared among the Zamilon, Sputnik, Mavirus, Phaeocystis globosa virus (PgVV) and Organic Lake (OLV) virophages are shown in the same color. Function predictions were made according to homologies between virophages or to nr NCBI collection, or regarding conservation of protein domains.

Supplementary figure 3. Lysis plaque assay with Mont1 and Mama. Scan of colored lysis plaques with A. polyphaga monolayer inoculated with Mont1 ( A) and Mamavirus ( B) 3 days after inoculation. Magnification of a Mont1 spot ( C).

 Supplementary figure 1

 Supplementary figure 2

 Supplementary figure 3

 Chapitre Six

Conclusions et perspectives

Le travail de thèse présenté ici permet une meilleure compréhension des virophages en faisant le point sur les connaissances actuelles, mais également en fournissant de nouvelles données comme par exemple les caractéristiques du premier virophage associé à un Mimiviridae du groupe C. Les virophages sont des entités virales aux caractéristiques proches de celles des virus satellites : ils sont regroupés au sein de la même classification, celle des agents sous- viraux, ce qui est d’ailleurs source de débats [30]. Les virophages sont associés à un virus dit hôte, sans lequel la multiplication est impossible, et pour lesquels leur présence est délétère [22]. Bien que les interactions entre les virophages et les virus hôtes soient peu connues, ils pourraient partager entre autres certains motifs comme des promoteurs ou des signaux de polyadénylation, qui permettraient l’utilisation de certaines protéines du virus géant pour le compte de la réplication du virophage [24,32,37]. La dépendance des virophages vis-à-vis de leurs hôtes reste cependant à étudier, car en effet les génomes des virophages contiennent des gènes pouvant coder pour des protéines impliquées dans la réplication de l’ADN [22,32,33] . Cette fonction fait d’ailleurs partie d’un ensemble de fonctions potentielles que les viroph ages décrits jusqu’à présent pourraient plus ou moins partager, y compris le virophage Zamilon dernièrement isolé. Parmi ces fonctions se trouvent également des protéines formant la capside, une hélicase, ou encore une possible protéase. Les derniers génomes de virophages construits à partir de données métagénomiques provenant d’un lac du parc Yellowstone – les Yellowstone Lake virophages – les possèdent aussi [33,34]. En plus de ce noyau fonctionnel que pourraient partager les virophages, il semblerait qu’ils possèdent tous des virions et des génomes de taille semblable. Ils présentent néanmoins de nombreuses divergences, notamment au niveau des hôtes : Mavirus est associé à Cafeteria roenbergensis virus, et OLV et YLV pourraient l’être à un Phycodnaviridae , bien que ceci reste à vérifier [38]. Même Sputnik et Zamilon, qui sont tous deux associés à des Mimiviridae et sont relativement proches du point de vue de leur génome, n’ont pas le même spectre d’hôtes possibles. Jusqu’à présent, tous les viropha ges identifiés semblent toutefois associés à des 0HJDYLUDOHV , ou du moins à des Mimiviridae – Cafeteria roenbergensis virus correspond à une branche des Mimiviridae à l’écart des trois groupes régulièrement évoqués dans cette thèse, et les virus hôtes de OLV et YLV restent hypothétiques – soulevant inévitablement la

 question de l’existence possible de virophages associés à d’autres ordres viraux . Et si tel n’est pas le cas, peut-on trouver des virophages associés aux 0HJDYLUDOHV en raison des caractéristiques remarquables de ces derniers, ou peut-on encore trouver des 0HJDYLUDOHV parce qu’il existe des virophages pouvant potentiellement freiner leur évolution.

Les virophages Sputnik sont capables de se multiplier dans les usines à virus de tous les Mimiviridae testés, quel qu’en soit le groupe, bien qu’ayant été isolés avec des virus du groupe A [39] . Ce large spectre d’hôtes a d’ailleurs été confirmé par l’isolement de Sputnik 3 sans son virus géant natif, en utilisant un Mimivirus comme virus rapporteur. Par la suite, il a également été montré que Sputnik 3, qui partage plus de 99% d’identité avec l es deux autres souches de Sputnik, pouvait se répliquer en cas de co-infection avec des Mimiviridae des groupes A, B et C. Cette capacité pourrait être un des moteurs de la variabilité nucléotidique importante observée entre les génomes des Mimiviridae , via des transferts de gènes. D’ailleurs, des séquences de Sputnik peuvent être intégrées dans le génome des Mimiviridae [40,41] . Contrairement à Sputnik, Zamilon a une spécificité d’hôte plus restreinte puisqu’il est incapable de se multiplier en association avec les Mimiviridae du groupe A. En cherchant à comprendre l’origine de cette spécificité, il est apparu qu’un gène potentiel, l’ORF19, pouvait y jouer un rôle. Bien qu’aucune fonction connue n’ait pu lui être associé, il présente une homologie protéique significative avec Megavirus chiliensis , un membre du groupe C [42], tandis que la plupart des autres gènes seraient davantage proches de Sputnik. La construction phylogénétique réalisée pour cet ORF a en outre montré que Zamilon serait plus proche des virus des groupes B et C, tandis que les virophages Sputnik seraient quant à eux rattachés au groupe A. Ce gène pourrait être responsable de la spécificité d’hôte, bien que ce point reste à approfondir. Un autre point pourrait également être un acteur de cette spécificité : la structure du génome. En effet, Zamilon possède une grande portion de son génome inversé en comparaison avec le génome de Sputnik. De façon intéressante, cette particularité est également remarquée lorsque l’on compare les génomes de Mimiviridae des groupes B ou C avec ceux de virus du groupe A. Il se peut donc également qu’un motif structurel partagé soit reconnu, comme ce pourrait être le cas pour certains motifs fonctionnels (promoteurs et signaux de polyadénylation) [24,32]. Cette complexité des interactions entre les virophages et leurs virus hôtes détonne de leur classification commune avec les virus satellites, témoignant de relations ciblées plus qu’opportunistes. Par ce point, un parallèle peut d’ailleurs être tracé avec les bactériophages, lesquels peuvent présenter le même type de variation de spécificités

 pour leurs hôtes [43 –45]. Les interactions entre virophages et virus hôtes demeurent ainsi encore un mystère qu’il reste à étudier pour comprendre pleinement leur nature.

Ces études pourront bénéficier des apports successifs faits aux techniques employées au fil du temps. Initialement, les Mimiviridae étaient ainsi cherchés dans des eaux de tours aéroréfrigérantes, mais l’élargissement des origines des échantillons a permis d’en isoler davantage, et de sources très diverses, témoignant d’une relative abondance des virus géants dans l’environnement [1,4] . Cela avait déjà été supposé d’après la détection significativement importante de séquences apparentées à Mimivirus dans différents environnements marins [46 – 48]. Toutefois, l’accélération du nombre de virus isolés a surtout été le fait de l’amélioration des méthodes utilisées : d’une approche empirique, nécessitant d’adapter les cocktails d’antibiotiques utilisés dans les co -cultures d’amibes, à des protocoles pouvant être réalisé s à l’aveugle et en routine, permettant l’analyse de plus en plus d’échantillons simultanément [4,5,49] . Si ces méthodes ont permis d’isoler de nombreux Mimiviridae , elles sont également à la base de l’identification des virophag es y étant associés [4,22]. Ces techniques sont néanmoins dépendantes des caractéristiques des virophages : bien que le virophage Sputnik 3 ait pu être isolé en utilisant un virus rapporteur, ceci n’a été réalisable que par le large spectre d’hôtes de Sputnik [39] . La découverte récente de Zamilon et de sa spécificité d’hôtes présente les limites de cette approche. De même, la détection de séquences de virophages par PCR est gênée par l’ absence de séquences conservées : même les protéines qui pourraient avoir des fonctions partagées par les virophages n’ont que de relativement faibles similarités nucléotidiques.

Bien que relativement récente, la métagénomique a déjà apporté des informations intéressantes concernant les virophages [33,34]. Outre les constructions de génomes complets de virophages à partir de données métagénomiques provenant de lacs, ces études ont également montré l’abondance relativement importante de séquences pouvant provenir de virophages dans l’environnement. Ceci concorde avec l’abondance estimée de virus géants dans l’environnement [46 –48], et a mené à l’hypothèse d’un rôle des virophages dans les écosystèmes aquatiques, en régulant les populations virales environnementales [35,36]. Les virophages permettraient de diminuer la mortalité globale des populations ciblées par les virus, qu’ils s’agissent d’algues ou de protozoaires, et impacteraient donc toute une chaine d’organismes. Une telle régulation, si elle existe, ne pourrait être le fruit que d’interactions complexes, ce qui est appuyé par l’existence de virophages de ayant des spécificités d’hôtes différentes. Malgré les apports nombreux fournis par la métagénomique, celle-ci ne permet

 pas de travailler directement sur le virus, et donc d’en apprendre plus sur ses caractéristiques. De fait, c’est la combinaison entre métagénomique et techniques de culture qui présente le plus de possibilité pour accroître prochainement les connaissances dans le domaine des virophages.

 Bibliographie

1. La Scola B, Audic S, Robert C, Jungang L, de Lamballerie X, Drancourt M, Birtles R, Claverie JM, Raoult D (2003) A giant virus in amoebae. Science 299: 2033.

2. Suzan-Monti M, La Scola B, Raoult D (2006) Genomic and evolutionary aspects of Mimivirus. Virus Res 117: 145-155.

3. Legendre M, Santini S, Rico A, Abergel C, Claverie JM (2011) Breaking the 1000- gene barrier for Mimivirus using ultra-deep genome and transcriptome sequencing. Virol J 8: 99.

4. La Scola B, Campocasso A, N'Dong R, Fournous G, Barrassi L, Flaudrops C, Raoult D (2010) Tentative characterization of new environmental giant viruses by MALDI-TOF mass spectrometry. Intervirology 53: 344-353.

5. Pagnier I, Ikanga Reteno DG, Saadi H, Boughalmi M, Gaia M, Slimani M, Ngounga T, Bekliz M, Colson P, Raoult D, La Scola B (2013) A decade of improvements in Mimiviridae and Marseilleviridae isolation from amoeba. Intervirology 56:6.

6. Iyer LM, Aravind L, Koonin EV (2001) Common origin of four diverse families of large eukaryotic DNA viruses. J Virol 75: 11720-11734.

7. Yutin N, Wolf YI, Raoult D, Koonin EV (2009) Eukaryotic large nucleo-cytoplasmic DNA viruses: clusters of orthologous genes and reconstruction of viral genome evolution. Virol J 6: 223.

8. Van Etten JL, Lane LC, Dunigan DD (2010) DNA viruses: the really big ones (giruses). Annu Rev Microbiol 64: 83-99.

9. Boyer M, Madoui MA, Gimenez G, La Scola B, Raoult D (2010) Phylogenetic and phyletic studies of informational genes in genomes highlight existence of a 4 domain of life including giant viruses. PLoS One 5: e15530.

10. Koonin EV, Yutin N (2010) Origin and evolution of eukaryotic large nucleo- cytoplasmic DNA viruses. Intervirology 53: 284-292.

11. Lefkowitz EJ, Wang C, Upton C (2006) Poxviruses: past, present and future. Virus Res 117: 105-118.

12. Dixon LK, Abrams CC, Bowick G, Goatley LC, Kay-Jackson PC, Chapman D, Liverani E, Nix R, Silk R, Zhang F (2004) African swine fever virus proteins involved in evading host defence systems. Vet Immunol Immunopathol 100: 117-134.

13. Williams T (1996) The iridoviruses. Adv Virus Res 46: 345-412.

 14. Federici BA, Vlak JM, Hamm JJ (1990) Comparative study of virion structure, protein composition and genomic DNA of three ascovirus isolates. J Gen Virol 71 ( Pt 8): 1661-1668.

15. Larsen JB, Larsen A, Bratbak G, Sandaa RA (2008) Phylogenetic analysis of members of the Phycodnaviridae virus family, using amplified fragments of the major capsid protein gene. Appl Environ Microbiol 74: 3048-3057.

16. Wilson WH, Van Etten JL, Allen MJ (2009) The Phycodnaviridae: the story of how tiny giants rule the world. Curr Top Microbiol Immunol 328: 1-42.

17. Boyer M, Yutin N, Pagnier I, Barrassi L, Fournous G, Espinosa L, Robert C, Azza S, Sun S, Rossmann MG, Suzan-Monti M, La Scola B, Koonin EV, Raoult D (2009) Giant Marseillevirus highlights the role of amoebae as a melting pot in emergence of chimeric microorganisms. Proc Natl Acad Sci U S A 106: 21848-21853.

18. Colson P, Pagnier I, Yoosuf N, Fournous G, La Scola B, Raoult D (2013) "Marseilleviridae", a new family of giant viruses infecting amoebae. Arch Virol 158: 915-920.

19. Legendre M, Arslan D, Abergel C, Claverie JM (2012) Genomics of Megavirus and the elusive fourth domain of Life. Commun Integr Biol 5: 102-106.

20. Colson P, de Lamballerie X, Fournous G, Raoult D (2012) Reclassification of giant viruses composing a fourth domain of life in the new order Megavirales. Intervirology 55: 321-332.

21. Philippe N, Legendre M, Doutre G, Coute Y, Poirot O, Lescot M, Arslan D, Seltzer V, Bertaux L, Bruley C, Garin J, Claverie JM, Abergel C (2013) Pandoraviruses: amoeba viruses with genomes up to 2.5 Mb reaching that of parasitic eukaryotes. Science 341: 281-286.

22. La Scola B, Desnues C, Pagnier I, Robert C, Barrassi L, Fournous G, Merchat M, Suzan-Monti M, Forterre P, Koonin E, Raoult D (2008) The virophage as a unique parasite of the giant mimivirus. Nature 455: 100-104.

23. Sun S, La Scola B, Bowman VD, Ryan CM, Whitelegge JP, Raoult D, Rossmann MG (2010) Structural studies of the Sputnik virophage. J Virol 84: 894-897.

24. Claverie JM, Abergel C (2009) Mimivirus and its virophage. Annu Rev Genet 43: 49- 66.

25. Desnues C, Raoult D (2010) Inside the lifestyle of the virophage. Intervirology 53: 293-303.

26. Desnues C, Boyer M, Raoult D (2012) Sputnik, a virophage infecting the viral domain of life. Adv Virus Res 82: 63-89.

27. Krupovic M, Cvirkaite-Krupovic V (2011) Virophages or satellite viruses? Nat Rev Microbiol 9: 762-763.

 28. Fischer MG (2012) Sputnik and Mavirus: more than just satellite viruses. Nat Rev Microbiol 10: 78.

29. Desnues C, Raoult D (2012) Virophages question the existence of satellites. Nat Rev Microbiol 10: 234.

30. Gaia M, Colson P, Desnues C, La Scola B (2013) Virophage concept, the. eLS.

31. Fischer MG, Allen MJ, Wilson WH, Suttle CA (2010) Giant virus with a remarkable complement of genes infects marine zooplankton. Proc Natl Acad Sci U S A 107: 19508-19513.

32. Fischer MG, Suttle CA (2011) A virophage at the origin of large DNA transposons. Science 332: 231-234.

33. Yau S, Lauro FM, Demaere MZ, Brown MV, Thomas T, Raftery MJ, Andrews- Pfannkoch C, Lewis M, Hoffman JM, Gibson JA, Cavicchioli R (2011) Virophage control of antarctic algal host-virus dynamics. Proc Natl Acad Sci U S A 108: 6163-6168.

34. Zhou J, Zhang W, Yan S, Xiao J, Zhang Y, Li B, Pan Y, Wang Y (2013) Diversity of virophages in metagenomic data sets. J Virol 87: 4225-4236.

35. Jassim SA, Limoges RG (2013) Impact of external forces on cyanophage-host interactions in aquatic ecosystems. World J Microbiol Biotechnol 29: 1751- 1762.

36. Wodarz D (2013) Evolutionary dynamics of giant viruses and their virophages. Ecol Evol 3: 2103-2115.

37. Legendre M, Audic S, Poirot O, Hingamp P, Seltzer V, Byrne D, Lartigue A, Lescot M, Bernadac A, Poulain J, Abergel C, Claverie JM (2010) mRNA deep sequencing reveals 75 new genes and a complex transcriptional landscape in Mimivirus. Genome Res 20: 664-674.

38. Yutin N, Colson P, Raoult D, Koonin EV (2013) Mimiviridae: clusters of orthologous genes, reconstruction of gene repertoire evolution and proposed expansion of the giant virus family. Virol J 10: 106.

39. Gaia M, Pagnier I, Campocasso A, Fournous G, Raoult D, La SB (2013) Broad spectrum of mimiviridae virophage allows its isolation using a mimivirus reporter. PLoS One 8: e61912.

40. Desnues C, La Scola B, Yutin N, Fournous G, Robert C, Azza S, Jardot P, Monteil S, Campocasso A, Koonin EV, Raoult D (2012) Provirophages and transpovirons as the diverse mobilome of giant viruses. Proc Natl Acad Sci U S A 109: 18078-18083.

41. Yutin N, Raoult D, Koonin EV (2013) Virophages, , and transpovirons: a complex evolutionary network of diverse selfish genetic elements with different reproduction strategies. Virol J 10: 158.

 42. Arslan D, Legendre M, Seltzer V, Abergel C, Claverie JM (2011) Distant Mimivirus relative with a larger genome highlights the fundamental features of Megaviridae. Proc Natl Acad Sci U S A 108: 17486-17491.

43. Hyman P, Abedon ST (2010) Bacteriophage host range and bacterial resistance. Adv Appl Microbiol 70: 217-248.

44. Duplessis M, Moineau S (2001) Identification of a genetic determinant responsible for host specificity in Streptococcus thermophilus bacteriophages. Mol Microbiol 41: 325-336.

45. Koskella B, Meaden S (2013) Understanding bacteriophage specificity in natural microbial communities. Viruses 5: 806-823.

46. Ghedin E, Claverie JM (2005) Mimivirus relatives in the Sargasso sea. Virol J 2: 62.

47. Monier A, Claverie JM, Ogata H (2008) Taxonomic distribution of large DNA viruses in the sea. Genome Biol 9: R106.

48. Monier A, Larsen JB, Sandaa RA, Bratbak G, Claverie JM, Ogata H (2008) Marine mimivirus relatives are probably large algal viruses. Virol J 5: 12.

49. Boughalmi M, Saadi H, Pagnier I, Colson P, Fournous G, Raoult D, La Scola B (2013) High-throughput isolation of giant viruses of the Mimiviridae and Marseilleviridae families in the Tunisian environment. Environ Microbiol 15: 2000-2007.