ewr

3. The three hypotheses for virus origin revisited The hypothesis that viruses were already present before the emergence of modern DNA-cells opens new perspectives about their origin and suggests we revisit the three classical hypotheses in this new context. 3.1. The virus-first hypothesis The virus-first hypothesis has been revived in the last decade by Wolfram Zillig who suggested that viruses originated in the prebiotic word, using the primitive soup as a host (Prangishvili et al., 2001). Such hypothesis is in line with the view, advocated by several molecular biologists, that the formation of cells occurred relatively late in the evolution of life. It was common for some time to imagine the RNA world as a community of free molecules competing with each others (Gilbert, 1986). Recently, some authors have even proposed that the Last Universal Common Ancestor (LUCA) was not a cellular entity (Kandler, 1998; Martin and Russell, 2003; Koga et al., 1998). In particular, they suggested that cellular membranes originated independently after the divergence of Archaea and Bacteria, in order to explain why archaeal lipids are so different from eukaryotic/bacterial ones (with opposite stereochemistry and different carbon chains). If all life evolution from the very beginning up to LUCA occurred in an acellular context, it is indeed possible to imagine that viruses first emerged as individual entities in a world of competing proteins and nucleic acids, either bathing in a !primitive soup” or located on a mineral platform. However, the hypothesis of a LUCA without membrane is contradicted by the existence of homologous proteins functioning at the membrane level that are encoded by all sequenced genomes from the three domains of life, strongly suggesting that these proteins (hence a membrane) were already present in LUCA (Pereto et al., 2004). In fact, the first cells probably arose well before the emergence of LUCA. For instance, it appears unlikely that a world of free molecules could have evolved to such an extent to produce a ribozyme capable of synthesizing proteins (the ancestor of present-day ribosomes). Even in the early RNA world, a primitive metabolism should have produced at least precursors for RNA and lipid syntheses, as well as the energy required to perform these reactions. It is difficult to imagine the emergence of such a metabolism without Darwinian selection, and this requires the competition between well-defined individual entities (at least proto-cells). Since modern viruses contain proteins, they should have originated after the emergence of the ancestral ribosome, i.e. well after the apparition of primitive RNA-cells (in the second age of the RNA world, sensu Forterre, 2005). Accordingly, in my opinion, the virus first theory can be rejected for present-day viruses (even viroids would need a suitable cellular environment providing nucleotides for their emergence). In this case, one is presently left with only two possibilities: either the first RNA viruses originated from RNA cells by regressive evolution (a new version of the reduction theory), or from RNA fragments that escaped from RNA cells (a new version of the escape theory). 3.2. The escape hypothesis The traditional hypothesis viewing viruses as elements of cell genomes that escaped from their cellular environment, becoming autonomous and infectious selfish elements, is easier to defend in the context of a pre-LUCA scenario for virus origin. In this context, one does not expect anymore any specific relationship between proteins encoded by viruses and those encoded by their hosts, since viruses now derive from genome fragments escaped from cells predating LUCA. Furthermore, one can reasonably assume that it was easier for a genome fragment to become autonomous in ancient RNA cells, since the different molecular mechanisms operating in these proto-cells were probably much simpler and less integrated than in modern DNA cells (Woese, 2002). In particular, it has been often argued that the genomes of ancestral RNA cells were fragmented (as in the case of modern double-stranded RNA viruses). These genomes could have been composed of semi-autonomous chromosomes (possibly formed by a few RNA genes) that were replicated independently and transferred randomly from cells to cells (Woese, 1987 ; Poole et al., 1998). Some RNA chromosomes could have encoded a coat protein that helps the proto-virus to be transferred, finally becoming infectious (Fig. 2). This process would be reminiscent of the moron theory proposed by Hendrix and co-workers for the origin of virus (Hendrix et al., 2000). Although this theory was inferred from the observation that modern DNA viruses can easily

acquire more DNA by illegitimate recombination it can be easily extrapolated for the origin of RNA viruses in the RNA world. 3.3. The reduction hypothesis The transformation of a cellular organism into a viral one could have been also much easier in a world of RNA cells, again because these cells were much simpler than modern ones. Just as modern parasites can loose part of their metabolic channels, an RNA-cell living as a parasitic endosymbiont in another RNA cell could have lost its own machinery for protein synthesis and for energy production, using instead those of the host (Fig. 2) (Forterre, 2005). In this model, viral capsids could have originated from the envelopes of RNA cells composed of identical proteins, resembling the S-layer of modern prokaryotes. The reduction hypothesis might have been driven by the harsh competition that most likely occurred between RNA cells all along the evolution of the RNA world. As a consequence of this competition, early life evolution probably went through several bottlenecks each time a crucial new molecular mechanism was invented. At each of these bottlenecks, the descendents of the individual endorsed with such a great selective advantage (the winners) would have eliminated all other lineages of proto-cells that previously coexisted with them (the losers). The only chance of survival for the losers was to become parasites of the winners. In this hypothesis, viruses evolved by parasitic reduction from ancient lineages of RNA cells that were out-competed in the Darwinian selection process, and thus could only survive by parasiting the winner of this competition (Forterre, 1992).

Fig. 2. Two alternative hypotheses for the origin of viruses in the second age of the RNA world (after invention of protein synthesis), black circles correspond to translation machinery (e.g. ancestral ribosomes) and lines to linear RNA chromosomes. Upper panel, the escape hypothesis: unequal cell division produces minicells with single chromosome (a or b) but no translation apparatus. The chromosome a will be eliminated but chromosome b will survive because it is associated with a proteins coat that allows its transfer into a new RNA cell, it becomes a virus. Lower panel, the reduction hypothesis: a small RNA cell became an endosymbiont of a larger RNA cell. It looses its translation apparatus but continue to replicate autonomously and become infectious (similar to some pathogenic bacteria in eukaryotic cells).