Unité Mixte
de Recherche

Biologie et Génétique
des Interactions Plante-Parasite

Campus International
de Baillarguet
34398 Montpellier Cedex 5


Copyright © CIRAD 2009
Group 2 : Interaction Virus Insecte Plant (VIP) )

Leader : Stéphane Blanc
Tél : 04 99 62 48 04
Secretary : Michel Amphoux & Marie-Carmen Martinez

Staff back to top of page
Technical staff
Bastien Cayrol
Marie Ducousso
Sophie Le Blaye
Jean-Luc Macia
Alexandre Morisset
Elodie Pirolles
Michel Yvon
Edwige Berthelot (2015-2018)
Maëlle Deshoux (2016-2019)
Faustine Ryckebush (2016-2019)

Research interest back to top of page

The research group Virus-Insect-Plant (VIP) is interested in the impact of the mechanisms of host-to-host vector transmission on the evolution of viral populations. On several chosen virus-host-vector models, our research programs investigate the mechanisms of virus transmission, those of virus evolution, or are positioned at the interface between these two scientific fields.

We implement a multidisciplinary approach including structural biology and biochemistry, cell biology, and evolutionary biology.

The group is currently organized as five interconnected but autonomous programs, each developed by one permanent researcher:

Molecular interactions between viruses and their insect vectors   Cellular virus-host interaction regulating the uptake of the virus by its insect vector
Responsible: Marilyne Uzest   Responsible: Martin Drucker
All non-circulative viruses transmitted by aphids are retained in their insect mouthparts (stylets). This project aims at identifying Caulimo-, Poty- and Cucumovirus receptors in the stylets to thoroughly characterize virus/vector interactions. A second objective is to decipher the role of the acrostyle – organ which bears some receptors – in plant-insect interactions.
  Many parasites form during their life cycle specific cellular structures (morphs) that are dedicated to transmission. This project studies such morphs, formed by plant viruses in response to arrival, installation and feeding of their arthropod vectors on infected plants. [more] .

Transmission and evolution of geminiviruses

  The strange biology of multipartite viruses
Responsible: Michel Peterschmitt   Responsible: Stéphane Blanc
Many geminiviruses are ssDNA emerging viruses transmitted by hemipteran insect vectors. This project studies the transmission and evolution of these viruses, from the within host level to the field and landscape [more].   The way of life of multipartite viruses is poorly understood. Using Nanoviruses as biological models, we address the basic biology of multipartite viruses and experimentally investigate the possible benefits and costs in such biological systems [more].

Impact of abiotic stresses on the rate of vector-transmission of plant viruses    
Responsibles : Manuella van Munster & Stéphane Blanc    
Abiotic stresses affect the development of viral infection within host plants. The aim of this project is to evaluate how abiotic stresses associated to global changes may impact on the rate of vector transmission and thus on epidemiology of viruses [more].    

Publications back to top of page

refer to each project above

PhD back to top of page

Faustine Ryckebusch (2016-2019) - Caractérisation de la transmission de l'Alfalfa leaf curl virus, le premier géminivirus transmis par le puceron

Maëlle Deshoux (2016-2019) - Acrostyle dans les stylets de puceron

Edwige Berthelot (2015-2018) -

Déborah Conflon (2012-2015) - Assessing the possible maintenance of TYLCV-satellite association
Résumé : Begomoviruses (family Geminiviridae) are frequently detected with half genome size defective DNAs, and for some of them with alpha- and betasatellite DNAs of similar size. Except some begomoviruses, like Cotton leaf curl virus (CLCuV) which rely on a betasatellite for a full symptomatic infection in their natural host, most of the begomoviruses which were frequently detected with satellites do not rely on them for infectivity. The alphasatellite was rarely proved to have an impact on the helper virus but the betasatellite was often shown to increase its virulence. The association of Tomato yellow leaf curl virus (TYLCV) with satellites was only reported for a TYLCV strain isolated from Oman but the Mediterranean IL and Mld strains of TYLCV were experimentally proved to transreplicate readily satellite DNAs. As betasatellites dramatically increase the virulence of TYLCV, the potential economic impact of betasatellites on tomato was assessed by testing various factors involved in the potential maintenance of TYLCV-satellite association: (i) the relative intra-plant accumulation of TYLCV and the satellites, (ii) the frequency of host cells co-infected with TYLCV and the satellites, and (iii) the transmission efficiency of satellites by the natural whitefly vector of TYLCV, Bemisia tabaci. Three satellites previously isolated from cotton in Burkina Faso, were shown here to be transreplicated by TYLCV in tomato plants: Cotton leaf curl Gezira betasatellite (CLCuGB), Cotton leaf curl Gezira alphasatellite (CLCuGA) and Okra leaf curl Burkina Faso alphasatellite (OLCBFA). The dynamic of TYLCV and satellite DNAs monitored overtime by quantitative PCR (qPCR) revealed contrasted virus-satellite interactions. Satellites were maintained in all the test plants up to the latest sampling at 150 days post inoculation (dpi). At 32 dpi, the proportion of cells in which TYLCV and satellites were co-detected by FISH was above 50% of the total number of cells infected with TYLCV and/or satellites. The transmission of CLCuGB and CLCuGA with B. tabaci was tested and both of them were readily transmitted from source plants at 32 dpi. As the ratio of virus and satellite DNAs is evolving over time, the transmission efficiency was tested also with 150 dpi source plants. The most prospective objective of the thesis was to determine for the first time the multiplicity of infection (MOI) of a begomovirus, which, according to its level, might be conducive to the maintenance of defective or satellite DNAs. Finally, a generic detection test of betasatellites was designed to be used as a diagnostic tool by the French Agency for Food, Environmental and Occupational Health & Safety (ANSES). Besides the specific question of the possible maintenance of satellites with TYLCV, the results of our study are expected to provide a new insight on the interactions between begomovirus and satellites, at the plant, cellular and molecular levels.

Zineb Delabess (2010-2015) - Success story of a non-canonical Tomato yellow leaf curl virus recombinant in Morocco
Résumé : Although viruses have a clonal multiplication, they readily generate genetic diversity, particularly through error prone polymerases and by recombination. The generated diversity is potentially useful for adaptation to new environments as highlighted here with the study of Tomato yellow leaf curl (Tylc) disease-associated begomoviruses (family Geminiviridae) in Morocco. Two Tylc-associated virus species known to be highly recombinogenic were previously reported from Morocco, Tomato yellow leaf curl virus (TYLCV) and Tomato yellow leaf curl Sardinia virus (TYLCSV). The first TYLCV-TYLCSV recombinant from Morocco was isolated from a plant sample collected in 2010 from a tomato cultivar resistant to Tylc-associated viruses. It exhibited a non-canonical recombination profile characterized by an unusually short 76-nt TYLCSV genome fragment from which it was named IS76. As the previously reported detection tools would misidentify IS76 as a parental TYLCV virus, a new multiplex PCR tool had to be designed for identifying in routine recombinant and parental genomes in individual plants. The prevalence of IS76 was monitored in time and space with 879 plants sampled throughout Morocco between 1998 and 2014. IS76 recombinant was not detected in the samples collected before 2004 but was detected from 2010 onwards. The frequency of IS76 increased up to the complete replacement of parental viruses in Southern Morocco (Souss) in the years 2012-2014; the replacement was only partial in non-Souss regions. The date of the recombination event leading to IS76 was inferred with a Bayesian analysis of the sequences of 51 TYLCV genomes available in GenBank and of 5 TYLCV and 14 IS76 genomes cloned in this study. The date estimated between 1994 and 1999 (95% HPD 1986–2004) is compatible with the first observation of the parental viruses in Morocco by the end of the 1990s. It is supposed that IS76 was not frequent enough before 2004 to be detected in our samples. Its subsequent tremendous frequency increase coincides with the period after 2002 when the susceptible cultivars were replaced with cultivars bearing the Ty1 resistant gene. The possible selective advantage of IS76 in Ty1-bearing resistant cultivars was tested by comparing its fitness to those of representatives of parental virus species which were selected among those exhibiting the highest nucleotide identity with the TYLCV and TYLCSV derived fragments of IS76. Viral DNA accumulation measured with real time PCR (qPCR) was used as a proxy of fitness. The accumulation of IS76 was significantly higher than that of each of the parents in Ty1-bearing resistant plants, both in competition or agroinoculated individually. The selective advantage of IS76 on resistant cultivars did not pay any fitness cost on susceptible non-Ty1-bearing plants in which IS76 accumulation was similar or higher than that of each of the parents. Interestingly the selective advantage of IS76 was associated with a dramatic negative impact on TYLCV accumulation, the parent exhibiting the highest accumulation in mono-infected plants. On the contrary, IS76 has a positive impact on TYLCSV accumulation. In plants where TYLCV and IS76 were co-inoculated, TYLCV accumulation was about 100 times lower than TYLCV accumulation in mono-infected plants at 30 dpi, and felt below the detection level at 120 dpi. As the agroinfectious IS76 clone differ from the agroinfectious TYLCV clone not only by the recombined region but also by 27 mutations, the relative contribution of mutations and recombination on the selective advantages of IS76 was not known. To focus solely on the effect of the recombination, an artificial IS76’ recombinant was engineered with the TYLCV fragment derived from the agroinfectious TYLCV clone and the TYLCSV fragment derived from IS76. IS76’ was proved to have a negative impact on TYLCV-DNA accumulation in Ty-1 bearing plants and it was not significantly different from that of IS76. Although the recombination event account solely for the dramatic effect on the accumulation of the TYLCV parent, mutations probably contributed to the full selective advantage of IS76 and particularly for an efficient onset of plant infection. For example, when each was co-inoculated with TYLCV, the accumulation of IS76’ was below that of IS76 at 10 dpi but their accumulation was similar at 30 dpi. As IS76 was not reported outside of Morocco and was not detected in experimental conditions, the recombination breakpoint at locus 76 seems to be an extremely rare event. Consistently our phylogeny studies suggest a unique IS76 emergence in the South of Morocco. To further understand the peculiarity of locus 76 recombination breakpoint, we tried to generate it in Ty1-bearing and TY1-non-bearing tomato plants co-infected with representatives of the parental species. The locus 76 breakpoint was detected but the frequency of the recombinant allele remained extremely low in both cultivars up to the last sampling time at 240 dpi. Although the Ty1-bearing cultivars were most likely conducive of the IS76 emergence at the field and regional level, they did apparently not provide any advantage for its emergence at the plant level; the IS76 recombinant might have reached its pre-emergence level in a non-tomato host. Finally, it is interesting to note how the study of unexpected Tylc-associated symptoms on resistant tomatoes in Southern Morocco, has finally pointed to the importance of a short genome region of TYLCV for which hardly no functional analysis were reported.

Anne Sicard (2011-2014) - Fonctionnement des populations de virus multipartites de plantes au cours des différentes étapes de leur cycle de vie
Résumé : Les virus multipartites sont caractérisés par un génome divisé entre au moins deux segments, chacun étant encapsidé individuellement. Bien que ces virus représentent près de 30% des genres de virus de plantes connus, les coûts et avantages engendrés pas cette organisation génomique restent mal connus. Basée sur l’étude d’un Nanovirus, le Faba bean necrotic stunt virus (FBNSV), ma thèse vise à mieux comprendre les dynamiques des populations d’un virus multipartite au cours de l’infection de son hôte et de sa transmission par pucerons. Dans un premier temps, nous avons démontré que chacun des huit segments composant le génome du FBNSV s’accumulait de manière reproductible à une fréquence relative spécifique, certains segments représentant près de 30% de la population virale au sein de la fève et d’autres n’excédant pas 2%. En changeant de plante hôte, nous avons pu montrer que les fréquences relatives des différents segments étaient hôte-spécifiques. Nous avons par ailleurs observé des changements de fréquences relatives au sein des pucerons vecteurs, changements révélant des interactions intimes entre le FBNSV –précédemment perçu comme traversant simplement les barrières cellulaires des pucerons- et son vecteur.
Cette plasticité et l’ajustement correspondant de la composition génétique de la population virale au sein de l’hôte pourrait permettre aux virus multipartites de s’adapter rapidement à un nouvel environnement grâce à un changement spécifique du nombre de copies de ses gènes. Cependant, ces différences de fréquences relatives pourraient également constituer un coût additionnel pour le virus, si comme il est généralement admis, tous les segments ont besoin de pénétrer au sein de la même cellule pour garantir l’infection. En tenant compte des segments rares au sein des populations virales, ce coût apparaît énorme et nous avons donc décidé de vérifier l’existence d’un tel coût expérimentalement. Une étude de la distribution des différents segments génomiques au sein des cellules végétales est donc en cours et ce, dans le but de regarder si les huit segments sont toujours ensemble au sein des cellules ou s’ils peuvent être physiquement séparés. Cette question est d’importance majeure non seulement en ce qu’elle permet de réévaluer le coût potentiel des virus multipartites mais aussi en ce qu’elle permet de tester le concept de cycle de réplication de cellule autonome chez les virus.

Aurélie Bak (2010-2013) - Etude des mécanismes cellulaires de la transmission du Cauliflower mosaic virus
Résumé : La majorité des phytovirus utilise des vecteurs pour être transmis d’une plante à une autre, et les pucerons sont de loin les vecteurs les plus importants. Alors que les interactions moléculaires entre le virus et son vecteur font l’objet de nombreuses études, les phénomènes intracellulaires qui précèdent l’acquisition du virus par le vecteur dans la cellule végétale ont été peu étudiés. Le Cauliflower mosaic virus (CaMV), développe dans les cellules de la plante hôte une structure spécialisée et indispensable à la transmission : le corps à transmission ou CT. Le CT contient la protéine P2 qui est essentielle à la transmission car elle constitue un lien entre la particule virale et un récepteur localisé au niveau de la pointe des stylets de l’insecte. Les particules virales, quant à elles, sont massivement stockées dans un autre type de corps d’inclusion: les usines virales. Cette localisation différentielle des composants majeurs du complexe transmissible implique qu’un mécanisme réunisse P2 et les particules virales lors de l'acquisition, c'est-à-dire au moment des piqûres tests du vecteur dans les cellules du parenchyme. Au cours de ma thèse, nous nous sommes focalisés sur les événements cellulaires qui se produisent lors de l’acquisition du CaMV par le puceron. Les résultats montrent que la piqûre du puceron est un stimulus qui déclenche deux événements : 1. le CT se désintègre quasi instantanément et la P2 qu’il contient est relocalisée sur les microtubules dans toute la cellule ; 2. en parallèle, les usines virales libèrent des particules virales, qui se distribuent sur le réseau microtubulaire, en s'associant à la protéine P2. Ainsi, un très grand nombre de complexes transmissibles se forment dans une configuration facilement accessible au vecteur partout dans le cytoplasme. De manière surprenante, ce remaniement des composants viraux au sein de la cellule est totalement réversible : P2 reforme un CT, et les particules virales sont ré-absorbées par les usines virales ; l’ensemble est ainsi disponible pour un nouveau « cycle d’acquisition ». Ces résultats indiquent que le CaMV manipule l'hôte au travers de corps d’inclusion aux fonctions multiples qui i) permettent la « perception du puceron » par le virus, et ii) mettent en œuvre une réponse immédiate qui favorise les chances d’acquisition du virus. Ces résultats supposent que le CaMV détourne une ou des voies de perception et transduction du « signal puceron » de la plante. Nous avons initié la caractérisation de cette/ces voies de signalisation par plusieurs approches, dont les résultats préliminaires sont présentés en détail.

Juliette Doumayrou (2008-2011) - Virulence, transmission et accumulation intra-hôte et leurs conséquences sur l'évolution virale: exemple du Cauliflower mosaic virus infectant deux Brassicaceae
Résumé : Identifier les pressions de sélection imposées aux parasites par leurs hôtes et les conséquences que celles-ci peuvent avoir sur l’évolution de ces parasites sont des points essentiels pour comprendre le développement des maladies émergentes. L’aptitude pour un parasite à faire évoluer sa virulence constitue un paramètre clé pour appréhender ces phénomènes d’émergences. Cette aptitude est notamment expliquée chez les parasites par l’hypothèse adaptative du trade-off virulence - transmission. Dans ce contexte, nous avons étudié les relations entre la virulence, la transmission et l’accumulation intra-hôte d’un virus de plante, le Cauliflower mosaic virus (CaMV), afin (i) de tester, à partir de plusieurs isolats de CaMV, l’hypothèse du trade-off virulence - transmission dans leur hôte natif (Brassica rapa) et dans un nouvel hôte (Arabidopsis thaliana) et (ii) de voir si la virulence et la transmission sur l’hôte natif permettent de prédire celles sur un nouvel hôte. Les résultats montrent que la virulence et la transmission dans l’hôte natif, comme dans le nouvel hôte, sont positivement corrélées. La mesure de l'accumulation virale intra-hôte a révélée l'existence inattendue de deux groupes d'isolats, différant d'environ un ordre de grandeur dans leur accumulation. Bien qu'à l'intérieur de chacun des groupes la transmission et la virulence semblent positivement corrélées à l'accumulation intra-hôte, l'existence de ces deux groupes ne permet pas de mettre en évidence ces relations sur l'ensemble des isolats. Les performances des isolats de CaMV sur l’hôte natif (virulence, transmission et accumulation virale) ne permettent pas de prédire leurs performances sur le nouvel hôte. Ce travail souligne l'importance d’étudier différents traits d’histoire de vie du virus pour comprendre les interactions hôte-parasite. Une étude des mécanismes cellulaires impliqués dans l’adaptation à l’hôte et dans la transmission virale permettra, à terme, d’imaginer des moyens de lutte contre les émergences virales.

Florence Vuillaume (2007-2010) - Analysis of the phenotypic effects of the recombination between two viral species
Résumé : There is an immense literature on recombination in viruses and the apparent impact of recombination on virus evolution and emergence of new pathotypes. Fragmentary data on the subject have promoted the implicit view that most viral recombinant genomes may be deleterious or lethal, especially when the nucleotide identity of parental sequences decreases below 90%. However, the actual distribution of the phenotypic effects caused by genome-wide random homologous recombination has never been formally investigated. Moreover, no experimental data are available on the evaluation of potential emerging recombinant plant viruses. The aim of this study was to make up for this lack of information. To do so, we decided to create a bank of near-random recombinants between two viral species of the genus Begomovirus (Family Geminiviridae). The parental virus chosen exhibit 82% nucleotide identity and were the Tomato yellow leaf curl virus (TYLCV), one of the most damaging pathogens of tomato crop and the Tomato yellow leaf curl Mayotte virus (ToLCYTV) a South West Indies Ocean begomovirus. The bank was created with DNA-shufflingTM (Proteus, Nîmes), a technology initially designed for random shuffling of individual genes, and here implemented for the first time to shuffle full-length viral genomes. For 47 recombinant sequenced clones randomly extracted from this bank, we tested two fitness traits at two different times describing the reproductive capacity: the infectivity of the recombinants and accumulation in planta. For 27 out of these 47 recombinants, the virulence of the clone and its relationship to reproductive capacity were characterized. Finally, 18 out of those 27 recombinants were tested for their transmissibility and its relationship to reproductive capacity. Surprisingly in our biological model, genome-wide random homologous recombination on the distribution of the phenotypic effects resulted in a majority of intermediate parental phenotypes with very few deleterious and no lethal phenotypes. Finally, a good correlation between reproductive capacity and virulence but not between reproductive capacity and transmission was found. Such traits might be important for understanding potential emergence of viral recombinants onto tomato crops or new hosts and are discussed. As all the recombinants were found to be at least as virulent to their parents and because especially two of them were much more virulent, we think there is a high risk of emergence of recombinants between the TYLCV and the SWIO begomovirus.


asques et ascospores de Magnaporthe orizae - copyright : JL Notteghem spores Magnaporthe oryzae - copyright : JL Notteghem bactéries Xanthomonas pseudoalbilineans (gauche) et Xanthomonas albilineans (droite). Les deux produisent l'antibiotique albicidine (structure en haut de la photo - copyright : S. Cociancich/A. Mainz
  champignon Magnaporthe (vert) en train d'attaquer une feuille de riz - copyright : A. Delteil/JB Morel test d'anticorps sur puceron (Mysus persicae) - copyright : MS Vernerey/M. van Munster/M. Uzest