Looking for LUCA

(the Last Universal Common Ancestor)

Patrick Forterre

Infos / Organisation Meeting

French Version

Reférences Bibliographiques / References

Infos Séquençages / Sequencing info

Contact Participants / Reach Attendies

Crédits / Credits

   For the first time, scientists coming from different perspectives have gathered themselves together to reflect on the nature of the last common ancestor of all beings living today. The Treille Foundation has enabled the organization of a colloquium on this theme going beyond the usual framework for scientific conferences. While numerous participants presented several papers, some among them improvised. Very quickly, the atmosphere of conviviality appropriate to "Les Treilles" gave rise among all the members of our little group (many of whom had never met before) a sentiment of belonging to one community, at the dawn of a new scientific adventure. Within a field (the first step of life on earth) where the confrontation of ideas (and now and then of personalities) could be enlivened, the luminous atmosphere of Provence permitted the discussion of even the more contradictory hypotheses in complete serenity.

  The first objective of the colloquium was to assemble the researchers appearing to come from different scientific 'tribes'. Each tribe has its own rites, in particular the large conferences organized at regular intervals, its journals of predilection and of great reputation. The members of these different clans rarely mix , and only some of the more audacious sometimes are accepted in two or three different tribes. At "Les Treilles", at least five from among these were represented: those researchers obsessed by the problem of the origin of life, the large international meeting on which will be held at Orleans, that of the archaeo-microbiologists, specialists of the third group of living beings on earth, the archaebacteria or archea, that of the 'thermophilists' , who study microbial life at very high temperature ( up to 110 °C) , that of the molecular evolutionists, who try to retrieve the genealogy of all living beings (the universal tree) by comparing the sequences of their macromolecules, and finally the genomicists, a new tribe in formation, the aim of which is the exhaustive analysis of genomes, rendered possible thanks to large programs of systematic sequencing of DNA.

  Some participants appear in none of the tribes mentioned, they have been chosen by reason of their competence recognized in a domain the importance of which appears crucial to confront the problem of the last common ancestor. Some were confronted for the first time with the great questions posed by the evolution of life, and they were not any less enthusiastic.

  Most of the researchers present were experimentalists, they have behind them a long tradition of prudence, rigor, going sometime to the point of opposition to all too speculative ideas. In particular, reductionism of molecular biology and the division which has been created since fifty years ago between the 'pure and hard' biochemists and the evolutionists are major obstacles in a field where the imagination plays a determining role. From this point of view, Christian De Duve had played an irreplaceable role in the course of these days. A Nobel prize winner, doyen of our colloquium, and recognized for his work in cell biology, he has proved to have a fertile imagination, playing a particularly active role in our discussions. Our hopes are that the seminar at Treilles will be the foundation of a new theme of wholly separate research: research on LUCA (the Last Universal Common Ancestor)

  LUCA was baptised at Treilles, yet we don't know if this name imposed itself, but it was not unanimous between us. Other terms have been proposed over the years: the name progenote, put forward in 1977 by Carl Woese (the creator of the concept of archaebacteria) had its hour of glory, but it seems on the decline. The notion of 'progenote' is in effect associated with the conception of a particularly primitive ancestor, much more simple than actual cells, but which does not correspond any more with the thought of many specialists. The majority vision today is more that of a common ancestor resembling either bacteria or archaea, and for some , that of a creature intermediate between prokaryotes (cells without nuclei) and eukaryotes (cells with nuclei).

  The term cenancestor (from the Greek root cen. together) proposed by Fitch in 1987, has the favour of purists. For others, it presents the inconvenience of being incomprehensible to ordinary mortals, and even to the ordinary among biologists. The appelation 'Last Common Ancestor' is more and more used in the literature. José Castresana, remarked during our seminar that it is also too vague. The term 'last common ancestor' could be used (and is in effect) for all groups of organisms.

  LUCA is a compromise between LCA and LUA (last universal ancestor) proposed at this colloquium by Christos Ouzounis. It could be popular with the media, since 'from LUCA to LUCY' encapsulates or replays the trajectory of the evolution of life. It defines a sympathetic entity, one which seeded our planet- we are all its descendants.s

  Why focus our attention on this key personage in our history? Why a colloquium on this theme today? For a long while research on the origins of life has been uniquely devoted to understanding how the first molecules of life appeared on our planet. Once they were present, the consequences went of themselves( the rest followed). There was no place for LUCA in these researches. On the other hand, the cellular and molecular biologists (with rare exceptions) did not want to waste their time (very precious time) speculating on the ancient unknowable, when the inexhaustible richness of model studies offered themselves to them, those more real (in the form of proteins and nucleic acids).

  The situation has much changed today, precisely thanks to the progress of molecular biology. One has become aware , in comparing the molecules of life of all organisms (from bacteria to humans) that they display common points, a fossilized record of their ancestral background. These are the researches which have been conducive to the discovery of the presence on earth of a group of prokaryotes, the archaea, as distant from the bacteria as from the eukaryotes.

  In comparing the fundamental mechanisms of life in the three Domains (bacteria, archaea, and eukaryotes) it then becomes possible to define " the smallest common denominator of the three": LUCA.

  There remains a division between the biologists loving evolution, who also reconcile themselves with first origins in going back to the past, and the pioneers of researches on the origins of life, who try since the founder experiment of Stanley Miller in 1953 to start from point zero (the primitive soup or dawn of stone) to advance towards the present. The time has come to gather together all the tribes.

Some questions debated at "Les Treilles"

Is it possible to draw up an 'Identikit' portrait of LUCA ?

  The comparison of genes present in the three domains of life is one of the favoured approaches for achieving knowledge of LUCA. This necessitates as a preliminary step the complete inventory of genes present in each domain. This work will benefit from the large number of genomes which will be completely sequenced in the years to come. A certain number of these have already been completed. The colloquium at Treillles is being held within three months after the publication of the complete sequence of the genome of the yeast of the bakery (baker's yeast) , Saccharomyces cerevisiae, and one month before the publication in Science of the first genome of an archaea, Methanococcus jannaschii. Odile Ozier-Kalogeropoulos has therefore presented the results obtained with the complete analysis of S. cerevisiae, while Nikos Kyrpides gave us a hint in advance of those obtained in analysing the genome of Methanococcus jannaschii.

  Christos Ouzounis and Nikos Kyrpides have presented a first preliminary work comparing between genomes for organisms belonging to all three domains. This work should open on the inventory of proteins common to the three domains, that is to say present similarities of sequence which one could reasonably suppose to all be derived from the same common ancestor (In technical terms, these proteins are said to be homologs). These homologous proteins have a strong chance of having been present in the LUCA.

  Right now, the first results suggest a degree of complexity unexpected in LUCA. That organism would without doubt have already possessed many thousands of genes (and therefore of different proteins). It seems that all the large systems which permit the maintenance and expression of genetic material were already present, as were also many metabolic capacities. José Castresana contributed solid arguments to advance the iconoclastic idea according to which the molecular mechanisms of oxygen respiration were already present in LUCA, despite an environment apparently impoverished in oxygen! Was oxygen already present on our planet in localized niches, or alternatively did the mechanisms of oxidative respiration have another function in the epoch of LUCA ?

  One question which is more general remains to be posed: is the presence of the same gene in the three domains at the present day always synonymous with the presence in LUCA ? Was the transfer of genes which distributed or could distribute throughout the living world an invention which only appeared more lately in one of the three domains? In particular, Jim Brown and Hervé Philippe have shown that many protein phylogenies suggest the transfer of a very large number of genes implicated in the metabolism of bacteria to the eukaryotes, without doubt by the intermediation of mitochondria, cellular organelles which have evoloved in the eukaryote cells starting from ancient endosymbiotic bacteria.

Communion around LUCA

  To understand the origin and evolution of real genomes with the object of going back in time, it is equally necessary to understand the mechanisms of their recent evolution. The comparative analysis of genomes of organisms relative to the evolutionary plan is very instructive from this point of view. Odile Ozier-Kalogeropoulos has also compared a part of the genome of the yeast Kluyveromyces lactis with that of S. cerevisiae, while Renaud de Rosa has compared the genome of two bacteria, the coccibacillus Escherichia coli , of which more than 60% of the sequence is known, and the pathogen Haemophilus influenzae, of which the sequence has been completely determined in the last year. These works put forward evidence for very important differences in the rate of evolution of one gene compared to another (and thus of one protein to another) , of phenomena of rapid and massive loss (relative to evolution) of certain genes, and finally a large number of proteins called paralogs, that is to say having diverged by duplication since the time of a common ancestral gene.

  The rapid evolution of the sequence of certain proteins renders difficult to determine evidence in support of their relationship to one domain or to another. This could be in part explained by the presence in the genomes completely sequenced of many genes coding for proteins which have no detectable homologs in the other two domains. In certain cases, it is possible to find evidence for resemblance of one domain to the other at the level of 3-dimensional protein structure. Chris Sanders has presented a work of structural analysis which announces the putting in place of a systematic strategy for identifying the very large number of genes and functions possible in entire genome sequences. The number of genes possessing homologs in the three domains (thus likely to be present already in LUCA ) will also become augmented in years to come.

What root for the universal tree

  What can be made of genes which are present only in two domains? Were they already present in LUCA and in this case, have they been later (eventually) lost only in one of the three domains, or have they really appeared in one branch common to the two domains? the problem is complex; in effect certain genes are present only in Bacteria or Archaebacteria, others in archaebacteria and eukaryotes, others again in Bacteria and the eukaryotes.

S. Miller & C. de Duve

  If one puts aside the problem of sampling, it will be a priori easier to answer the question posed if one knows the placement of the root of the universal tree which connects the three Domains (between them). That is a question which is very controversial. Many authors actually situate this root in the Bacterial branch (the eukaryotes and archaebacteria are in this case brother groups). Jim Brown has presented some results along this line, based on some universal trees founded on proteins having diverged since the time of duplication before the separation of the three domains.

Location of LUCA and Hyperthermophiles

This result could explain the presence of numerous characters of the "eukaryote" type within the archaea, as well as protein phylogenies which associate archaea and eukaryotes. It exists along with other phylogenies associating Archaea and Bacteria. Jim Brown interprets those in terms of the transfer of genes from one Domain to the other.

  The validity of these results is however contested by Patrick Forterre and Hervé Philippe, on the basis of studies on the variability of the evolution of different positions of amino acids in the same protein. According to them, there no longer remains exploitable information permitting the rooting of trees based on sequences having diverged so long ago. For Hervé Philippe, aberrant trees between proteins of three domains (for example those which do not permit finding division into three of the living world) are the norm, and it is the "coherent" trees which demand to be explained. He suggests that the appearance of "monophyletic" groups in a molecular tree indicates an important switch in the structure/function of the molecule studied.

It is important to note that the position of the root of the universal tree in the bacterial branch favors the idea by which LUCA resembles actual prokaryotes. Nevertheless, as long as the position of the root is not known with certainty, the question remains open. In particular , if the root is situated in the branch of the eukaryotes, LUCA could in fact resemble a cell of this type!

The eukaryote/prokaryote transition (or vice-versa?)

  For Christian de Duve, LUCA very much resembles a real bacterium ( or archaeon). He presents an original scenario which envisages the transition of prokaryotes to eukaryotes in a quasi-solid milieu in which loss of the bacterial wall is followed by the expansion of the cytoplasmic membrane to permit phagocytosis. This expansion leads to the formation of intracellular membrane networks characteristic of eukaryote cells.

G. Ourisson & N. Glansdorff

  Other participants are in contrast in favor of a LUCA in which the genome was more of the eukaryote type. For Nicolas Glansdorff, the operons appeared slowly. perhaps in a lineage common to archaea and to bacteria. For Rudiger Cerff, the genome of LUCA was fragmented into numerous chromosomes the genes of which contained introns. Patrick Forterre thinks that LUCA was neither prokaryote nor a eukaryote , but an organism of intermediate type , the descendants of which have opted for opposite adaptive strategies; adaptation towards miniaturization and the maximum replication rate having led to prokaryotes, while feeding on other organisms by predation led to the birth of eukaryotes.

  For many participants, LUCA was not an organism, but a collection of diverse organisms exchanging their genes more or less without constraint. In this respect was found debate of the same type as that concerning the problem of the African Eve. Without doubt a population geneticist would have been needed to relate to our reflections. We regretted the absence of Miroslav Radman, prevented at the last minute from participating in our colloquium. His lab studies the mechanisms of speciation at a molecular level, and it seems that certain of these are common to both prokaryotes and eukaryotes, suggesting that the notions of species and species barrier existed already during the epoch of LUCA ?

  Another essential question remains to be posed- what decisive invention , appearing with LUCA, gave it such a predominance over its competitors of the epoch that only its descendants today populate our planet?

Was LUCA a hyperthermophile?

  Certain authors think that life appeared at very high temperature and that LUCA itself was a hyperthermophile (an organism living between 80 and 110 °C). However, Stanley Miller and Antonio Lazcano put us on guard against the danger of extrapolating the supposed conditions of life of LUCA to that of the origins (of life). We have seen in effect that LUCA was already a very sophisticated organism, its appearance has therefore been preceded by a long period of evolution. Stanley Miller insists equally on the instability of numerous prebiotic compounds at very high temperature. Stanley Miller and Patrick Forterre e out the stress on the contradiction which exists between the idea of a primordial evolution which be entirely produced at high temperature (of the origin of LUCA ) and the hypothesis of an RNA world, taking into account the instability of this molecule at temperatures in the neighbourhood of the boiling point of water. Piero Cammarano nevertheless presents some new facts on the evolution of elongation factors (which participate in protein synthesis) in favour of the ancientness (antiquity) of hyperthermophilic bacteria. There again, the viability of molecular phylogenetic analysis is at the centre of the debate. Patrick Forterre remarked that the results obtained with ribosomal RNA , which are generally interpreted in favour of the antiquity of hyperthermophiles (their branches declare themselves by a reduced length). could be tainted with error, given their higher GC content, which has the effect of artificially reducing the rate of evolution.

  Nicolas Glansdorff thinks that the formation of operons could have been produced in response to adaptation of prokaryotes to thermophilic conditions, whereas Purificacion Lopez-Garcia presented a new scenario which envisages how an ancestral thermophile was modified in graduated steps towards the bacteria and archaea. This scenario is based on the study of enzymes which introduce supercoils into the DNA molecule (gyrase and reverse gyrase) and on the hypothesis of a precise geometry of this molecule necessary for its function. This geometry will have been initially adapted to temperatures in the region of 50 to 70 °C: the gyrase and reverse gyrase will have then appeared in order to permit he adaptation of organisms to lower and higher temperatures respectively. In effect, the action of these enzymes permits counteracting the effects of changes of temperature on the DNA double helix (opening at high temperature, compression at low temperature).

  Whether or not hyperthermophiles were near LUCA, their study is a particularly fascinating field of research. Franck Robb presented some experiments which contribute evidence on the mechanisms permitting proteins of hyperthermophilic archaea to function at temperatures in the neighbourhood of the boiling point of water. One of the strategies utilized is the formation of a network of ionic interactions at the surface of the protein. He furthermore presented the first results showing that the hyperthermophilic archaebacteria are particularly resistant to ionizing radiation. These organisms should possess DNA repair mechanisms of extremely good efficiency.

The age of LUCA.

  Some months before holding our colloquium, an article published in the American review Science rejuvenated LUCA. It would have been two billion years old, not 3.5 or 4 as one thought previously according to examination of microfossils found in some sedimentary rocks of this epoch which are still detectable. Hervé Philippe has subjected the facts of this work to a thorough critical analysis from which he deduces (sets forth) that the date advanced of 2 billion years rests on facts of very bad quality. The protein phylogenetic trees used gave for the most part aberrant results and the calibration of the molecular clock is itself in question.

M. Fontecave & O. Ozier-Kalogeropoulos

Could we go back in time, beyond the epoch where LUCA lived?

  The complexity of LUCA suggests that a long period (at least in terms of evolutionary changes) preceded its emergence. The first steps of life on our planet had been the appearance and assembly of the molecular constituents of life until the appearance of he first cells. The genome of these first cells was without doubt constituted by molecules of RNA and not DNA. In contrast to DNA, RNA could in effect play at the time the role of enzyme and of genetic material. Stanley Miller and Christian de Duve think that the appearance of RNA was itself a late event , in effect, this molecule does not seem to be able to be synthesized by simple methods of prebiotic chemistry. RNA would have therefore been preceded by molecules the exact nature of which we will never know.

  If until LUCA the first cells were RNA based , the road ought to have been very long, always in terms of evolutionary changes. Patrick Forterre suggested dividing this evolution into several periods: the two ages of the RNA world (before and after the invention by RNA of a genuine system for synthesis of proteins) and the first age of the DNA world (from the first DNA cell until the emergence of LUCA. . How to obtain information about these different stages? According to him; the RNA virus and certain DNA viruses could be the descendants of cellular organisms having lived before LUCA. They would only have survived the domination of the latter as parasites of its descendants. The study of the molecular biology of viruses could contribute information on the stages of evolution before LUCA.

  An event essential for all this evolution had been the invention of DNA. Today the synthesis of DNA depends on a key enzyme, ribonucleotide reductase (RNR), which makes the precursors of RNA (reduction of ribose of RNA by elimination of oxygen to give a deoxyribose). Marc Fontecave has made the point of our knowledge on the 3 classes of RNR known at present and on their distribution in the living world. He presented the characterization by his team of the first RNR of archaebacteria, in collaboration with Franck Robb. This work has permitted the demonstration for the first time that the three classes of RNR derive from the same ancestral enzyme. Marc Fontecave and Franck Robb have taken advantage of their common presence at Treilles to write up an article reporting this discovery. The existence of an RNR homologue in the three domains of life strengthens the idea according to which the genome of LUCA was already composed of DNA.

  To establish the 'identikit'-portrait of predecessors of LUCA, Antonio Lazcano has proposed using the paralog proteins the origin of which (by gene duplication) is before the cell of LUCA. Arturo Becerro presented an example in a work-in-progress of this strategy in the study of the evolution of genes implicated in certain metabolic pathways. Another example of comparative analysis of metabolic pathways in the three domains was presented by Nicolas Glansdorff. Again, the problem is that of distinguishing duplications which may have occurred before or after LUCA , only the first can be permitted in reconstituting the history of genes before LUCA. Bernard Labedan and Renaud de Rosa have shown in effect that recent duplications are numerous within the bacteria and Odile Ozier-Kalogeropoulos has reported results of the same type in yeast.

P. Lopez Garcia, B. Labedan & R. deRosa

  Bernard Labedan summarized work in collaboration with Monica Riley on the protein families of E. coli which suggested that many ancestral proteins had already been formed from large modules the shape and function of which were close to that of genuine proteins. To rediscover the most ancient homologies, we must without doubt make an analysis of proteins in modular terms and perform structural analysis of the type proposed by Chris Sander.. The problem will then be to distinguish between homology (shared structural heritage from a common ancestor) and analogy (shared structure acquired by convergent evolution). As a matter of fact, everyone agreed on the necessity of combining genomic studies (comparison of sequences) and structural facts which could be more informative on the level of functions, for the purpose of analysing evolution of he latter, which are the true targets of natural selection.

  Finally the question poses itself: is it possible to extrapolate from present-day metabolism to ancestral metabolism? Guy Ourisson has presented an attempt of this genre in the case of cellular membranes which he considers as ought to have formed (themselves) very early, perhaps before all other processes. He has shown how the very wide variety of amphiphilic membranes of the present day , often uncovered from sediments in the form of their "molecular fossils", form an evolutionary series. By retrograde analysis, it can be deduced that the first membranes were probably formed by phosphates of polyprenyl, which he has obtained by abiotic reactions- and therefore probably prebiotic.

  From his perspective, Christian de Duve pointed out that present day metabolism up to protometabolism preceded the RNA world (the world of thioesters which he popularised in his book "Blueprints of a Cell". He pitted himself against the absolutist idea of a world of omnipotent RNA. According to him, all present enzymes have not been preceded by the corresponding ribozyme, and he would imagine a world of peptide catalysts which were efficacious and very early diversified. Stanley Miller, who has made the point on the present state of research in the field of origins of life in his introductory discussion, proposed in conclusion a true programme of experimental research for connections between prebiotic and contemporary metabolism and advanced many new avenues.

P. Cammarano, P. Lopez Garcia, A. Lazecano & A. Becerra

  In conclusion, Christian de Duve emphasized the quality and intensity of the discussions. In appearance, each had defended his point of view on the hotly debated questions, but it was evident that the arguments brought to bear by one and another will contribute to the enrichment in the months to come of many reflections, reposing of questions , or deepening (profound study) of diverse hypotheses. The majority are going away only to meet again for the editing of a series of articles to appear in a special number of the international review "Journal of Molecular Evolution" dedicated to this meeting. In the opinion of many participants, it is one of the best scientific meetings at which they have had the occasion to participate (the setting of Treilles very much contributed to this appreciation), and we all hope to renew this experience in the years to come.



Several papers coming out of this exciting workshop (including new collaboration initiated at "Les Treilles" ) have been published recently in a special issue of "Journal of Molecular Evolution" Vol 49, Oct 1999.

Last revision and translation on Tuesday September 11 2001