Skip to main content

A second metagenomics generation to better understand plant virus ecology and evolution

Final Report Summary - GEOMETAGENOMICS (A second metagenomics generation to better understand plant virus ecology and evolution)

GEOMETAGENOMICS: A second metagenomics generation to better understand plant virus
ecology and évolution

Agricultural encroachment into natural environments have strong potential to promote novel host-pathogen interactions and the emergence of new crop diseases. Understanding virus dynamics across agro-ecological interfaces is thus critical for forecasting and managing viral diseases. Deep-sequencing and bioinformatics approaches now give us the power to detect both novel and known viruses in ecological vegetation samples at the landscape scale.

Our over-arching hypothesis of the Geometagenomics project was that the emergence of socially/economically relevant viral diseases is in many cases linked with ecological disturbances caused by human intrusions into natural ecosystems. Specifically, we proposed at the beginning of the project that the displacement of natural plant species by intensive agriculture is an important subset of these intrusions that represent fundamental disturbances in the diversity, demographics and evolutionary dynamics of viral communities inhabiting natural ecosystems. This hypothesis was tested along the project by acquiring fine-scale “geo-referenced-metagenomic” data on contemporary viruses present within preserved areas, and within contexts of fynbos-edging agro-ecosystems (Objective1) and on “ancient” viruses present within fynbos plant specimens stored in French and South African herbaria (Objective2). In addition, we have inferred and revived ancestral viruses to retrace the emergence and spread over the past 100 years of African geminivirus species (Objective 3).

Objective1 aimed at exploring contemporary plant viral dynamics within a defined spatiotemporal framework. During the outgoing phase of the Geometagenomics project, we have completed the experimental tasks (sampling and molecular analysis, Photo 1) and improved the computational analysis of the geometagenomics (i.e. spatially informed metagenomics approach) dataset to assess the spatial and temporal distributions of plant viruses at the landscape scale within two Mediterranean-climate ecosystems: the Western Cape region of South Africa and the Rhône delta river region of France. This extensive study has led to a major publication in September 2017 in ISME Journal, which title is “Geometagenomics illuminates the impact of agriculture on the distribution and prevalence of plant viruses at the ecosystem scale”. In this study, we have found within just a small fraction (~40 Km2) of the Earth’s vegetated area, encompassing farmland and natural environments that (1) virus prevalence is significantly greater in cultivated areas in both countries (2) some virus families show strong associations with agriculture; and (3) the novel viruses identified (94 putative species) are primarily from uncultivated plants. This study represents the first effort to systematically evaluate plant-associated viromes across broad agro-ecological interfaces. Our findings emphasize how little is currently known about the diversity of plant-associated viruses and provide insight into the impacts that agriculture has had on their distribution and prevalence.

Objective2 aimed at using preserved herbarium specimens to analyze past viral diversity. Methodological studies have been carried out in 2016. We have tested several RNA extraction methods and estimated the minimum weight of leaf material needed for obtaining a reliable detection of plant viruses. Based on these methodological studies, we have started the inventory of RNA and DNA plant viruses contained within French and South African herbarium plant specimen samples (Photo 2). Noteworthy, we have succeeded in detecting traces of a DNA virus (geminivirus) from a 25-years old wild spurge (Euphorbia caput-medusae) and sequences of RNA viruses (partitivirus, amalgavirus, endornavirus and betaflexivirus) from 30-50 years old West African grasses. The characterization of the partial genomes of these viruses is underway. Meanwhile, we have started using viral metagenomics-based approaches for detecting sequences of RNA and DNA viruses from 386 herbarium specimens collected at the Bolus herbarium of UCT. I have also started in 2016 building a consortium of virologists, botanists, and bioinformaticians from Europe, USA and South Africa. Curators have been contacted and have agreed on participating to this consortium: University of Montpellier, National Museum of Natural History and the University of la Réunion (France), University of Mauritius (Mauritius), the Bolus herbarium, University of Cape Town (South Africa), Michigan State University (USA), Arizona State University (USA). Herbarium specimen samples have been collected from the majority of these herbaria and sent to Montpellier for metagenomics analyses.

Objective3 aimed at Using the geminivirus model to study virus adaptation and emergence. Since September 2015, we have coupled computational inferences with ancestral virus synthesis and phenotypic characterization to investigate the evolution of Maize streak virus (MSV) pathogenicity through time and space. All experimental tasks have been fully achieved in 2015/2016: (i) production of infectious clones of MSV genomes, (ii) reconstruction of six ancestral viral genomes, (iii) synthesis of infectious clones of these six ancestral viral genomes and (iv) production of data on disease symptoms and virus replicative fitness for infectious clones in three different maize host genotypes. The computational part of the project, which is still underway, is likely to shed light on Maize streak virus adaptation and virulence.

Besides indicating that agriculture substantially influences plant virus distributions and highlighting the extent of current ignorance about the diversity and roles of viruses in nature, this project had paved the way towards using infectious full genome clones of either actual or ancient genome sequences or computationally inferred ancient genome sequences to determine when and where over the past few centuries major changes in virus virulence have occurred.