LUCA, ten years after
Colloque organisé par Céline Brochier, Patrick Forterre et Simonetta Gribaldo du 4 au 9 septembre 2006
Bastien Boussau, Céline Brochier, Diego Cortez, Patrick Forterre, John Fuerst, Nicolas Glansdorff, Nigel Goldenfeld, Manolo Gouy, Simonetta Gribaldo, Henri Grosjean, Eugene Koonin, Antonio Lazcano, Paola Londei, Purification Lopez-Garcia, Arcadi Mushegian, Patrick O’Donoghue, Christos Ouzounis, Anthony Poole, Didier Raoult, Scott Roy, Chloe Terras
Molecular biology has permitted to unify the living world by showing that all organisms are the descendants of a same common ancestor. During the past 30 years, organism classification has witnessed a real revolution, with the crowing of Darwin’s dream of a natural classification of life based on the evolutionary relationships among organisms. In fact, the seminal work of Carl Woese in the seventies has provided the basis for such a natural classification, starting what may be called the « Woesian revolution », the proposal of a tripartite vision of Life that has received full legitimacy by the subsequent findings of the genomics era. Indeed, all cellular organisms belonging to the three domains of life share a unique genetic code and similar machineries for protein synthesis, indicating that they descend from a unique ancestor. What did the last ancestor of Bacteria, Archaea, and Eukarya look like? Since the 70s, Carl Woese has been stressing that this is one of the most important questions that a biologist can ask. In fact, had we the answer, we could more easily trace back the steps that preceded the apparition of this ancestor, and get closer to the origin of life. We may also better understand how the three domains of life originated and evolved, in particular the one we belong to, the eukaryotes.
To try obtaining more insights into this issue, one among us co-organized in 1996 a meeting at the Treilles Foundation that was entitled The Last Common Ancestor, and beyond. During this meeting our ancestor received a new name. In the stream of discussion it became LUCA, « The Last Universal Common Ancestor« . The term LUCA was rapidly adopted by the community of biologists, as well as largely diffused in the scientific press (the formula « from LUCA to Lucy » easily summing up most of biological evolution). Today, the term LUCA appears in countless scientific articles and reviews.
The meeting that has taken place in Les Treilles from September 4 to 9 2006 was meant to celebrate the tenth anniversary of the term LUCA, and make the point of what has been done and remains to be done on one of the most fascinating issues in Evolution. An important difference between the previous 1996 meeting and the present one was immediately evident: the 1996 meeting coincided with the publication of the very first completely sequenced genomes, those of the bacteria Haemophilus influenzae and Mycoplasma genitalium (1995), of the archaeon Methanococcus jannaschii (1996), and of the eukaryote Saccharomyces cerevisiae (1996).
In contrast, 10 years after, over 350 genomes from the three domains of life have been sequenced. This impressive data set has permitted the rapid development of comparative genomics and phylogenomics (the study of the evolution of cellular systems by combining molecular phylogeny and comparative genomics techniques), and have enhanced our understanding of the nature of LUCA.
The meeting gathered a number of scientists from all around the world with different expertise and evolutionary thinking in order to try answering a number of important questions around the LUCA such as: Was LUCA simple or complex? Was the genome of LUCA based on DNA or RNA? Was LUCA hyperthermophilic? What was the role of viruses in cellular evolution?
At the introduction to the meeting, Patrick Forterre (France) pointed out that there exist two logical approaches for studying LUCA: bottom up and top down. The two approaches are far from being in opposition, but instead provide important complementary data to dig into such ancient evolutionary events. Bottom up approaches try to reconstruct the conditions of early Earth and the possible pathways from prebiotic chemistry to the first cells.
Following this logic, Antonio Lazcano (Mexico) discussed important questions concerning the early steps of evolution that led to the first cells, such as how and why L-ribose was selected over all abiotically produced sugars at the origin of the RNA world, or whether alternative molecules may have harbored catalytic activity in the pre-RNA world and could them be used as a genetic information support before RNA.
N. Goldenfeld (USA) showed that relics of early life, preceding even the root of the universal tree, are present in the structure of the modern day canonical genetic code. He presented a model explaining the establishment of an optimal universal genetic code that minimizes errors (i.e. replaces an amino acid by a similar amino acid when nucleotide substitutions occur). His model favoured the hypothesis that horizontal gene transfer and genetic drift were the main forces driving the fixation of the genetic code.
Top down approaches try to reconstruct the characteristics of the LUCA by those present in present-day organisms. These approaches are mainly based on comparative genomics, a rapidly expanding discipline with the growing number of available genomic data. One of the main questions concerns the nature of the genome of LUCA. This is a very important question that was addressed as soon as the first complete genomes were published.
Strong discussions turned around the presence and the type of membranes of LUCA, for instance how membrane lipid composition may have evolved before and after LUCA, and how to relate this important aspect of cellular evolution to the emergence of the three domains (P. Lopez-Garcia, Spain and N. Glansdorff, Belgium).
By contrast, E. Koonin (USA) presented a quite different and provocative hypothesis in which LUCA would have been acellular. However, other evidences such as the likely presence of enzymes exploiting and generating membrane potential seems to favour the first hypothesis (as for example shown by a phylogenetic analysis of respiratory chain terminal oxidases presented by C. Brochier and S. Gribaldo (France). Finally, another provocative hypothesis of a complex eukaryote-like LUCA with membrane-bounded cell compartments was put forward by J. Fuerst (Australia). This hypothesis is based on the discovery of eukaryote-like features in members of different phyla of the domain Bacteria, making the concept of evolution of a compartmentalized LUCA at least conceivable without the need for symbiotic fusions between pre-existing cell types.
D. Cortez (France) reported a comparative genomics analysis showing that archaea and bacteria likely harbor a very similar system for the segregation of sister chromosomes at the end of replication, a process which is intimately linked to cell division, providing another argument in favor of a cellular LUCA.
Another important subject of discussion centred on the informational systems of LUCA. A. Poole (Sweden) presented evidence suggesting that LUCA could have had an RNA genome, such as the recently shown proofreading ability of RNA polymerases. However, arguments in favour of a LUCA with a DNA genome were put forward (for instance it is not clear whether RNA genomes could have been compatible with a large number of genes). In fact, comparative genomics points to an already quite complex LUCA. Since the publication of the first genomes ten years ago, many studies dealing with this question have been published. However, most of the evolutionary models based on shared gene content between complete genomes from the three domains of life were too simple and led often to an underestimation of real gene content of LUCA.
C. Ouzounis (Greece) presented new results based on a comparative approach of 153 genomes using phylogenetic criteria rather than gene content only, giving an estimation of more than 1000 genes in LUCA. P. Londei (Italy) presented the most recent data on archaeal translation and their comparison with Bacteria and Eukarya, from which she draw a model of the ancestral machinery for translation initiation in LUCA.
H. Grosjean (France) presented the evolutionary history of the enzymes involved in tRNA modification, and suggested that of a number of these enzymes appeared before LUCA supporting the idea of a rather sophisticated LUCA.
P. O’Donoghue (USA) presented a study of the evolutionary history of aminoacyl tRNA synthases suggesting that two different cysteine tRNA synthetases were present in LUCA and that their phylogenetic analysis can help to study the root of the tree of life. Another long-standing question is whether LUCA was a hyperthermophile or not. B. Boussau (France) presented an interesting model based on the estimation of the GC content of the rRNA of LUCA and of the ancestors of each domain. The results were quite surprising since they suggest a mesophilic LUCA with subsequent independent adaptations to hyperthermophily in Bacteria and Archaea.
Finally, a novelty of the meeting was that, for the first time, the role of viruses in early evolution has been considered. E. Koonin (USA) discussed the role of viruses and selfish elements all along of the evolutionary pathway towards LUCA. According to his models, viruses are very ancient lineages that could have predated LUCA. Supporting the possible antiquity of certain viruses’ lineages, D. Raoult (France) presented data on large modern viruses such as the Mimivirus, which was recently discovered.
To conclude the meeting, C. Terras (France) presented a historical and epistemological analysis of the scientific career and research of C. Woese in the frame of the concept of scientific revolutions.