The first identified giant virus retained by the Chamberland filter, was called 'Mimivirus' in 2003. With a particle of 0.7 micrometer in diameter packing a 1.2 Mb genome encoding 979 proteins, Mimivirus was the first virus overlapping the world of bacteria both in terms of particle size and genome complexity. As we thought we were finally reaching the limit of viral complexity and started to build a new paradigm about the evolution of DNA viruses, the discovery of the Pandoraviruses came ruining this newly built theoretical edifice. With 1.2 micron-long particles packing a genome of 2.5 Mb encoding more than 2,500 proteins, Pandoravirus salinus is now surpassing the complexity of the smallest eukaryotic cells, such as parasitic microsporidia species. As the family expanded with new members from all around the world and diverse environments, clearly represent a class of giant viruses totally unrelated to the Megaviridae. I will also present the discovery of Pithovirus sibericum, isolated from a >30,000-y-old radiocarbon-dated sample of Siberian permafrost that also share the amphora-shaped particles with the Pandoraviruses. This family is also rapidly increasing with now modern members dividing the family in 3 clades. Mollivirus sibericum, isolated from the same permafrost sample, is still the only member of this fourth family of giant viruses. Pandoravirus-like particles may thus be associated with a variety of virus families more diverse than previously envisioned. To conclude, I will briefly present the hypotheses that have been proposed about the origin and evolution of DNA viruses and their possible link with the emergence of eukaryotes.
In the mind of most biologists, a virus remains the most reduced and optimized vehicle to propagate a nucleic acid molecule at the expense of a cellular host, an ultimate parasite at the frontier of the living world. With genome sizes and gene contents larger than many bacteria, as well as particle sizes of the order of half a micron the Megaviridae have clearly made the point that being small and simple should no longer be considered fundamental properties of viruses, nor a testimony of their evolutionary origin. More recently, the discovery of the Pandoraviruses, with amphora shaped virions over a micrometer in length and genome sizes up to 2.8 Mb, surpassing the complexity of the smallest eukaryotic cells, raised a number of fundamental questions about giant viruses' origin and their mode of evolution. Finally, Pithovirus sibericum is an even larger virus in terms of particle size, but despite its amphora shaped particle, this 30,000 years old virus genome only encodes 460 proteins and is much closer to large icosahedral DNA viruses than to the Pandoraviruses. The convergence between the discovery of increasingly reduced parasitic cellular organisms and that of giant viruses exhibiting a widening array of cellular-like functions may ultimately abolish the historical discontinuity between the viral and the cellular world. We will finally discuss the biodiversity of giant DNA viruses in the light of some recent discoveries.
