11 The Last Common Ancestor of Modern Cells

David Moreira and Purificacion Lopez-Garcıa

11.1 The Last Common Ancestor, the Cenancestor,LUCA: What’s in a Name?

All living beings share a number of essential features pertaining to their bio-chemistry and fundamental processes. Some of these features are so complex (anoutstanding example is the genetic code) that their probability to have appearedseveral times independently is almost negligible. This authorizes the concept ofa common ancestor to all living beings that possessed these traits, and fromwhich diversification occurred, leading to the emergence of the three domainsof life that we recognize today, Archaea, Bacteria, and Eucarya. Nevertheless,the concept of a common ancestor was born well before these universal featureswere even identified, mostly on the basis of purely philosophical and theoreticalconsiderations.

11.1.1 Some Historical Grounds

The idea that all living beings are linked by a natural process has influencedthe scientific thought for a long time. For example in his 1686 Discourse onMetaphysics, the German philosopher Leibniz explicitly states the existence ofintermediate species between those that are distant, implying that all living be-ings are related in a great chain, which is intrinsic to the harmony of the universe.Following this inspiration, many scientists have tried to identify the source ofdiversity and the nature of the process unifying the different species. Darwin’sconcept of descent with modification and natural selection provided the solidgrounds to explain the patterns of similarity and difference among species thathave served as a general framework for biology until present days. In The Originof Species, Darwin concluded that a logical outcome of the premises of descentwith modification and natural selection is that probably all the organic beings,which have ever lived on this earth have descended from some one primordialform, into which life was first breathed (Darwin 1859). The concept of a commonancestor for all life forms was born.

This idea received strong support during the last century, thanks to thespectacular development of biochemistry and, especially, of molecular biology.Bacteria, eukaryotes and the recently discovered archaea (Woese and Fox 1977a)

David Moreira and Purificacion Lopez-Garcıa, The Last Common Ancestor of Modern Cells.In: Muriel Gargaud et al. (Eds.), Lectures in Astrobiology, Vol. II, Adv. Astrobiol. Biogeo-phys., pp. 305–317 (2007)DOI 10.1007/10913314 11 Springer-Verlag Berlin Heidelberg 2007

306 David Moreira, Purificacion Lopez-Garcıa

share the basic structural constituents of the cell and the most fundamentalmetabolic reactions. All life is based on common biochemical themes (Kornberg2000; Pace 2001).

11.1.2 The Hypothesis of a Cenancestor

This hypothetical ancestral organism to all living beings has been baptized withdifferent names according to different authors. The most popular are the lastcommon ancestor, the cenancestor (from the Greek kainos meaning recent andkoinos meaning common) (Fitch and Upper 1987) and the last universal com-mon ancestor or LUCA (Forterre and Philippe 1999). Although all evolutionarybiologists agree that some type of cenancestor existed, its nature is a matter ofintense and highly speculative debates. Most authors favor the idea that the ce-nancestor was a single organism, an individual cell that existed at a given timeand that possessed most of the features (and genes encoding them) that arecommon to all contemporary organisms. From this single ancestor, the differentdomains of life would have diverged (Fig 11.1A–D). Others, on the contrary,envisage a population of cells that, as a whole, possessed all those genes, al-though no single individual did (Kandler 1994; Woese 1998; Woese 2000). Thisimplies that the level of gene exchange and spreading in this population wasvery high. At some point, however, a particular successful combination of genesoccurred in a subpopulation that became isolated and gave rise to a whole lineof descent. Kandler, for instance, proposed in his pre-cell theory that bacteria,archaea, and eukaryotes emerged sequentially in this way (Fig 11.1E) (Kandler1994; Wachtershauser 2003).

Single cell or population, all researchers agree that the cenancestor was al-ready quite complex, having evolved from simpler entities, and that there wasa more or less long evolutionary path from the origin of life to the cenancestorstage. Both, the origin of life and the nature of the cenancestor are differentevolutionary questions. Nonetheless, despite the general agreement that the ce-nancestor was already quite complex, the level of complexity attributed to itvaries depending on the model. For Woese, the cenancestor was a relatively prim-itive entity, which he called a “progenote”, that had not completely evolved thelink between genotype and phenotype (Woese and Fox 1977b). For others, theprogenote state occurred prior to the cenancestor, which was nearly a moderncell (see review in Doolittle 2000a).

Given their universality, a ribosome-based protein synthesis (translation),a well-developed transcription machinery for the synthesis of structural and mes-senger RNAs, and an energy-obtaining process based on the generation of a pro-ton gradient across membranes (Gogarten et al. 1989), were among the featuresthat the cenancestor certainly possessed. Other cenancestor properties are, how-ever, much more controversial, such as the existence of a DNA-based genome oreven the possession of lipid-based membranes. The occurrence in the cenancestorof these properties was initially deduced mostly from biochemistry and molecular

11 The Last Common Ancestor of Modern Cells 307

Fig. 11.1. Current hypotheses for the evolution of the three domains of life froma last common ancestor or cenancestor. The cenancestor stage is indicated by a bluesphere. Models A to D envisage that all the properties attributed to the cenancestorco-existed in the same cell, whereas in E, they were present collectively in a populationof primitive cells

308 David Moreira, Purificacion Lopez-Garcıa

biology studies. Obviously, many genes (and proteins) were involved in these an-cestral machineries and processes. At present, the most powerful tools to infer inmore detail which were the genes and proteins already present in the cenances-tor are comparative genomics and molecular phylogeny. This type of analysisis greatly facilitated by the increasing number of complete genome sequencesavailable for very different organisms. Although several problems, in particularhorizontal gene transfer and differential gene loss, make the reconstruction ofthe ancestral gene content troublesome (Koonin 2003). We briefly discuss theseaspects in the following sections.

11.2 How Did the Cenancestor Make Proteins?

Prokaryotic (archaeal and bacterial) species contain approximately between 500and 10 000 genes, whereas eukaryotic species contain approximately between2000 and 30 000 genes. Nevertheless, when the gene content of all availablegenomes is compared, only ∼60 genes are found to be common to all of them.This set of genes is almost entirely integrated by genes encoding ribosomal RNAand ribosomal proteins as well as other proteins involved in translation (espe-cially aminoacyl-tRNA synthetases and translation factors) (Koonin 2003). Thisimplies that these genes are ancestral, and provide strong evidence that the ce-nancestor possessed a fully developed ribosomal-based translation machinery forprotein synthesis that was comparable to the one found in modern organisms.Protein synthesis by ribosomes is, therefore, the most universally conserved pro-cess. Furthermore, the level of conservation appears so high that the process ofprotein synthesis has remained practically unchanged for, likely, more than threebillion years.

A few of the ∼60 genes that conform the universal core encode RNA poly-merase subunits. RNA polymerase is responsible for the synthesis of messengerand other RNAs from genes (DNA templates) (Koonin 2003). As in the case oftranslation, this implies that the cenancestor possessed at least part of the tran-scription machinery found in contemporary organisms. Nevertheless, the degreeof conservation of the transcription machinery is not as high, since several RNApolymerase subunits and transcription factors are not universally distributed.

Transcription and translation are among the most conserved processes incells, and can be traced back to the cenancestor. The fact that both processesdepend primarily on structural and catalytic RNA molecules, e.g., the ribosomalRNAs, indicates that RNA played an essential role since very early in cell evo-lution. This and the recent discovery of ribozymes (small catalytic RNAs) havebeen important elements leading to the proposal of an “RNA!world” (for review,see Joyce 2002). According to this model, there was a very early evolutionarystage when RNA carried both, information-storing and catalytic functions, whichare performed today by DNA and proteins, respectively. While being broadly ac-cepted, this model remains, however, purely hypothetical.

11 The Last Common Ancestor of Modern Cells 309

11.3 What Was the Nature of the Genetic Material?

Despite the fact that, in all contemporary cells, DNA is the molecule where ge-netic information is stored, only three out of the ∼60 universal genes are relatedto DNA replication and/or repair. These are one DNA polymerase subunit, oneexonuclease and one topoisomerase (Koonin 2003). From all the genes known tobe involved in DNA replication in organisms of the three domains of life, manyare shared by archaea and eukaryotes but are absent from bacteria. The latterapparently possesses completely unrelated genes encoding the proteins that per-form the equivalent functions. Various hypotheses have been proposed to explainthe profound differences between the bacterial and archaeal/eukaryotic replica-tion machineries. One of them even postulates that the cenancestor did notpossess a DNA genome at all. According to this model, the cenancestor wouldhave had an RNA genome, and DNA replication would have evolved twice inde-pendently, once in the bacterial line of descent and other in the lineage leading toarchaea and eukaryotes (Forterre 2002; Leipe et al. 1999). However, this modelis contested by different lines of evidence that, put together, suggest that thecenancestor already possessed a DNA genome. First, although few, universallyconserved proteins and protein domains involved in DNA metabolism indeedexist (Giraldo 2003). Second, RNA is much more error-prone than DNA due toits higher mutation rate, so that single RNA molecules cannot exceed a certainsize (Eigen limit) without falling into replicative catastrophe (Eigen 1971, 2002).This limit is so small, ∼30–50kb, that an RNA molecule could contain only a fewdozen genes. Recent estimates suggest that the, already quite complex, cenances-tor may have had probably more than 600 genes (Koonin 2003). If its genomewas made of RNA, many RNA molecules would have been required to containall those genes, which poses a serious problem of stability during replication andpartition among daughter cells. The existence of this problem is attested by thecharacteristics of DNA and RNA viruses today. Whereas DNA viruses can havegenomes that reach very large sizes, up to ∼1Mbp (Raoult et al. 2004), RNAviruses’ genomes do not exceed ∼30kb (Domingo and Holland 1997).

The remaining models propose that the cenancestor had a DNA-basedgenome. One of these explains the dichotomy of DNA replication in bacteriaand archaea/eukaryotes by stating that, whereas transcription and translationwere already well developed in the cenancestor, DNA replication was still veryprimitive. DNA replication would have been improved and refined after, or simul-taneously to, the separation of the two lines of descent leading to the bacteria andto the archaea/eukaryotes (Olsen and Woese 1997). A contrasting model couldbe that the cenancestor had a highly complex DNA-based metabolism and con-tained the ancestral versions of the proteins found today in both the bacterialand the archaeal/eukaryotic replication machineries. One set of proteins wouldbe involved in replication, whereas the other would be specialized in DNA repair.During the speciation of the two lines, only one set of proteins would have beenretained in each line of descent. Another proposal suggests that the replicationmachinery was already well developed in the cenancestor, but that it evolved

310 David Moreira, Purificacion Lopez-Garcıa

very fast in one or the two lineages descending from the cenancestor to the pointthat the similarity between the homologous genes in both lines is no longerrecognizable (Moreira 2000; Olsen and Woese 1997). Finally, an additional hy-pothesis suggests that archaea and eukaryotes have retained the ancestral DNAreplication system, whereas the bacteria have seen their machinery replaced bygenes imported from viruses (Forterre 1999). However, there is no evidence forthis hypothesis but, on the contrary, many examples of the opposite, i.e., numer-ous acquisitions of replication-related genes by viruses from the genome of theircellular hosts (Moreira 2000). In conclusion, although still a matter of debate,it appears likely that the cenancestor had a DNA genome. However, the mech-anism for its replication and the early divergence of the replication machineriesin the bacteria and the archaea/eukaryotes remain a mystery.

11.4 What Did the Cellular Metabolism Look Like?

The question of how metabolism looked in the cenancestor is a difficult andcomplex one, and hence, is generally put aside in descriptions of a putativeancestor. The theoretical proposal of a cenancestor is fundamentally based onconserved genes related to the storage, expression and transmission of the ge-netic information present in all contemporary organisms (for recent reviews, seeDoolittle 2000a, 2000b; Koonin 2003). However, genes involved in energy andcarbon metabolism not only display a patchy distribution in organisms of thethree domains of life, but they very often belong to large multigenic familieswhose members have been recruited for different functions in various metabolicpathways. Furthermore, horizontal gene transfer is known to affect preferentiallymetabolic genes, since they may confer an immediate adaptive advantage. There-fore, the reconstruction of ancestral metabolic pathways is frequently masked bya complex history of gene duplication and differential enzyme recruitment (Cas-tresana and Moreira 1999).

Discussions and controversies about early metabolism do take place, but theyconcern the very first metabolism at the time when life arose rather than themetabolic traits in a more evolved cenancestor. These discussions are model-dependent. If the first living beings were heterotrophs feeding upon an organicprebiotic soup as proposed classically (Oparin 1938), then fermentation, whichis mechanistically simple, may have been the first way of gaining energy (Broda1970), while cell building blocks were directly uptaken from the soup. If the firstliving beings were chemolithoautotrophs initially developed on pyritic (FeS2)surfaces (Wachtershausser 1988) or in iron monosulfide (FeS) tridimensionalcompartments (Russell and Hall 1997), energy was derived from redox reactionsinvolving inorganic molecules such as H2S, H2 and FeS. In this case, organicmolecules were synthesized de novo from CO2 or CO by an ancestral, yet un-determined, metabolic pathway. Today, four different pathways of autotrophiccarbon fixation are known (revised in Pereto et al. 1999). The Calvin–Benson

11 The Last Common Ancestor of Modern Cells 311

cycle (reductive pentose-phosphate pathway) is found in oxygenic photosynthe-sizers such as cyanobacteria and plants, but also in other autotrophic prokary-otes. The Arnon cycle (reductive citric acid pathway) is found in several bacte-ria and archaea. The Wood–Ljundahl cycle (reductive acetyl-CoA pathway) isfound in acetogenic and sulfate-reducing bacteria and methanogenic archaea. Fi-nally, the hydroxypropionate pathway is found in the green non-sulfur bacteriumChloroflexus sp. and a few thermophilic archaea. There is general agreement toconsider that the Calvin–Benson cycle appeared relatively late during bacterialevolution. However, which of the other three is older is a matter of debate. TheArnon cycle is proposed to be older by Wachtershauser and other adherents ofthe pyrite-based chemoautotrophic origin of life because of its wide distributionin bacteria and archaea (Wachtershauser 1990). Nevertheless, a reductive acetyl-CoA pathway is proposed to be the primordial autotrophic pathway because ofits higher simplicity (acetate, a two-carbon molecule is synthesized) and its exer-gonic (energy-releasing) nature under certain hydrothermal conditions (Russelland Martin 2004). Finally, although less known, the hydroxypropionate pathwayis also very simple and has been proposed to be the fist autotrophic pathway atleast in phototrophic bacteria (Pereto et al. 1999).

Although the metabolic properties of the cenancestor are not generally dis-cussed, if any or several of these pathways had developed in earlier times, thecenancestor must have possessed them. In any case, the presence of a universalhighly conserved membrane-bound ATPase indicates that the cenancestor wasable to produce energy in the form of ATP by generating a proton gradient acrossthe cell membrane. What is unclear is the type of electron donors and acceptorsrequired to generate this proton gradient, although it might have been able touse a variety of oxidized inorganic molecules as electron acceptors. It is also likelythat the cenancestor was able to carry out a simple heterotrophic metabolism,at least some type of fermentation, but whether it was a full autotroph or notremains an open question.

11.5 Was the Cenancestor Membrane-Bounded?

All contemporary cells are surrounded by a plasma membrane that is madeout of phospholipids generally organized in bilayers. However, as in the caseof replication, there exist profound differences between, this time, the mem-brane lipids of archaea and the membrane lipids of bacteria and eukaryotes.In archaea, phospholipids are made out of generally isoprenoid lateral chainsthat are bounded by ether linkages to glycerol-1-phosphate, whereas in bacte-ria and eukaryotes phospholipids are made out of fatty acids bounded by esterlinkages to glycerol-3-phosphate. From these differences, the most fundamentalis the opposite stereochemistry of the glycerol-phosphate, since some bacteriaand eukaryotes make ether linkages under certain circumstances, and since ar-chaea do make fatty acid ether phospholipids as well (Pereto et al. 2004). The

312 David Moreira, Purificacion Lopez-Garcıa

enzymes responsible for the synthesis of glycerol-1-phosphate and glycerol-3-phosphate are not homologous in archaea and bacteria/eukaryotes, but belongto two different enzymatic families. This difference is so hard to explain thatsome authors have proposed that the cenancestor had not yet reached a cellu-lar stage. Membrane lipids, and cells simultaneously, would have evolved inde-pendently from a non-enzymatic lipid synthesis pathway to generate bacteriaand archaea (Koga et al. 1998). Other authors propose that the cenancestordid not possess a lipid membrane at all, and that membrane lipids (and theirbiosynthetic pathways) were invented twice, in each one of the lineages that di-verged from the cenancestor leading to the bacteria and to the archaea. Insteadof lipids, the cenancestor would have been endowed with a mineral membranemade out of iron monosulfide; cells would be mineral compartments in a partic-ular kind of hydrothermal chimneys (Martin and Russell 2003). Finally, another,less radical option, is that the cenancestor had a lipid membrane, but that it washeterochiral, i.e., it was composed by a mixture of lipids built upon glycerol-1-phosphate and glycerol-3-phosphate (Wachtershauser 2003). A subsequent spe-cialization of the biosynthetic pathways to yield the two types of homochiralmembranes would have accompanied the speciation of archaea and bacteria. Re-cent phylogenetic analyses of the enzymes involved in the synthesis of glycerolphosphate strongly suggest that the cenancestor possessed a non-stereospecificpathway of phospholipid biosynthesis and was therefore endowed with heterochi-ral membranes (Pereto et al. 2004). A cenancestor endowed with a lipid mem-brane would be in agreement with the occurrence of several membrane-boundproteins extremely well conserved, notably the proton-pump ATPases (Gogar-ten et al. 1989) and the signal recognition particle, SRP (Gribaldo and Cam-marano 1998).

11.6 Other Unresolved Questions

As we have seen above, some properties of the hypothetical cenancestor appearmore or less well-defined, such as the possession of a quite modern transcrip-tion and translation machinery and, most likely, the existence of phospholipidmembranes. Other features remain obscure, such as the type of carbon and en-ergy metabolism or the replication of the genetic material. However, there areadditional questions that remain open and that have been the subject of livelydebates. Among these, whether the cenancestor was hyperthermophilic or not,and whether it was simple or complex.

The proposal of a hyperthermophilic cenancestor is linked to the discov-ery of hyperthermophilic bacteria and archaea growing optimally at >80◦C,and to the first proposals of a hot, autotrophic origin of life in a warmer earlyEarth with extended hydrothermal activity (Achenbach-Richter et al. 1987; Pace1991). The most important argument used was that hyperthermophilic prokary-otes branched at the most basal positions in phylogenetic trees (Stetter 1996).The first criticisms to a hot origin of life derived from the fact that RNA and

11 The Last Common Ancestor of Modern Cells 313

other important biomolecules have relatively short life-times at high temper-atures (Lazcano and Miller 1996). However, as mentioned before, the originof life and the cenancestor were separated in time, and might have occurredin different environmental conditions (Arrhenius et al. 1999). Thus, some au-thors propose that the cenancestor was hyperthermophilic, and that the originof life might have taken place at much lower temperatures, but only hyper-thermophiles could survive the late heavy meteorite bombardment ∼3.9Ga ago(Gogarten-Boekels et al. 1995). However, the most important criticism to a hy-pothetical hyperthermophilic ancestor derives from recent phylogenetic anal-yses. On the one hand, computer reconstruction of ancestral ribosomal RNAsequences suggests that the content of guanine and cytosine in the cenances-tor’s RNA was incompatible with life at >80◦C (Galtier et al. 1999). How-ever, the analysis of the same datasets using other methods questions this con-clusion favoring, on the contrary, a hyperthermophilic cenancestor (Di Giulio2000). On the other hand, refined phylogenetic analyses of ribosomal RNA se-quences suggest that the basal emergence of hyperthermophilic bacteria wasan artifact of phylogenetic tree reconstruction and favor the idea that theyadapted secondarily to hyperthermophily (Brochier and Philippe 2002). How-ever, as in the previous case, the use of other phylogenetic methods comesup with opposite results (Di Giulio 2003). At any rate, although the situa-tion in bacteria is highly controversial, there is general agreement that theancestor of the archaea was indeed hyperthermophilic (Forterre et al. 2002).Therefore, if the bacterial ancestor was also a hyperthermophile, then themost parsimonious conclusion is that the cenancestor was hyperthermophilic.If the bacterial ancestor was not hyperthermophilic, then it will be very dif-ficult to infer the type of environmental conditions in which the cenances-tor thrived. At any rate, all the current analyses and proposals appear com-patible with the occurrence of a thermophilic (60–80◦C) cenancestor (Lopez-Garcıa 1999).

Another controversial issue concerns the level of complexity that the ce-nancestor possessed. As stated above, it is clear that it was already verycomplex. Even the authors that call it simple, envisage that its genome con-tained several hundred genes (∼600–1000), which is in the range of the sim-plest present-day prokaryotes (Koonin 2003). Most authors imagine an ances-tor that was also structurally simple, i.e., with a cellular organization resem-bling that of today’s prokaryotes. An ancestor of this type would have thegenetic material directly immersed in the cytoplasm, where the replication,transcription and translation would take place. This type of ancestor is sup-ported by the widely accepted bacterial rooting of the tree of life, i.e., that theroot of the tree of life lies between the bacteria and the archaea/eukaryotes(Fig. 11.1A) (Woese et al. 1990), but would be also compatible with two al-ternative tree topologies (Fig. 11.1B,C). However, this has been challenged byauthors that propose a eukaryotic rooting of the tree of life, i.e., that the rootlies in between the eukaryotes and a branch leading to the two prokaryotic

314 David Moreira, Purificacion Lopez-Garcıa

domains (Fig. 11.1D) (Brinkmann and Philippe 1999; Philippe and Forterre1999). This eukaryotic rooting would still be compatible with a prokaryote-like cenancestor, but it opens the possibility that the cenancestor had someof the traits that characterize modern eukaryotes, in particular the presence ofa membrane-bound nucleus and of many small RNA molecules claimed to berelics of a hypothetical RNA world (Poole et al. 1999). In this model, prokary-otes would be derived by a reductive process from more complex eukaryotic-likeancestors. However, although the position of the root is indeed an open ques-tion, models proposing a eukaryotic-like cenancestor do not explain how sucha complex entity was built from the prebiotic world. In this sense, a simpler,prokaryotic-like cenancestor appears much more parsimonious in evolutionaryterms.

11.7 Perspectives

Obviously many questions have to be answered before achieving an accuratepicture of the putative cenancestor. Fortunately, it seems likely that a number ofthese questions can be addressed thanks to the increasing amount of data derivedfrom comparative genomics. Nonetheless, a crucial issue will be to determinethe impact of horizontal gene transfer in evolution because, if horizontal genetransfer has been rampant all along life history as some authors suggest (Doolittle2000b), it would then be very difficult or even impossible to reconstruct anyancestor. In a scenario of frequent or even massive horizontal gene transfer,a single cenancestor that contained all the genes ancestral to those shared amongthe three domains of life did not likely exist. Rather, these ancestral genes wereprobably present in different organisms and at different times (Zhaxybayeva andGogarten 2004). The application of population genetics methods to the studyof gene emergence and extinction over long timescales may shed some light intothis problem.

Another open question concerns the origin and evolution of viruses, and howthese have affected cellular evolution. For instance, they may have contributedsignificantly to horizontal gene transfer serving as vehicles of gene exchange.They may have also helped to increase the evolutionary rate of many genes thuscontributing to the development of novel gene functions.

An alternative source of information to reconstruct a model portrait of thecenancestor may come from the generation of more accurate models of theearly Earth. This could help to delineate the environmental conditions wherelife arose and first evolved, giving clues, for instance, as to the putative hy-perthermophilic nature of the cenancestor, and/or to the most likely metabolicpathways it utilized. Similarly, resolving the controversies that currently existon the earliest microfossils and increasing the number of unambiguous fossildata will be most helpful to establish a likely chronology of early life evolu-tion.

11 The Last Common Ancestor of Modern Cells 315

References

Achenbach-Richter, L., R. Gupta, K.O. Stetter, and C.R. Woese. 1987. Were the orig-inal Eubacteria thermophiles? Syst. Appl. Microbiol. 9: 34–39.

Arrhenius, G., J.L. Bada, G.F. Joyce, A. Lazcano, S. Miller, and L.E. Orgel. 1999.Origin and ancestor: separate environments. Science 283: 792.

Brinkmann, H., and H. Philippe. 1999. Archaea sister group of Bacteria? Indicationsfrom tree reconstruction artifacts in ancient phylogenies. Mol. Biol. Evol. 16: 817–825.

Brochier, C., and H. Philippe. 2002. Phylogeny: A non-hyperthermophilic ancestor forBacteria. Nature 417: 244.

Broda, E. 1970. The evolution of bioenergetic processes. Prog. Biophys. Mol. Biol. 21:143–208.

Castresana, J., and D. Moreira. 1999. Respiratory chains in the last common ancestorof living organisms. J. Mol. Evol. 49: 453–460.

Darwin, C. 1859. The Origin of Species by Means of Natural Selection. J Murray,London.

Di Giulio, M. 2000. The universal ancestor lived in a thermophilic or hyperthermophilicenvironment. J. Theor. Biol. 203: 203–213.

Di Giulio, M. 2003. The ancestor of the Bacteria domain was a hyperthermophile.J. Theor. Biol. 224: 277–283.

Domingo, E., and J.J. Holland. 1997. RNA virus mutations and fitness for survival.Ann. Rev. Microbiol. 51: 151–178.

Doolittle, R.F. 2000a. Searching for the common ancestor. Res. Microbiol. 151: 85–89.Doolittle, W.F. 2000b. The nature of the universal ancestor and the evolution of the

proteome. Curr. Opin. Struct. Biol. 10: 355–358.Eigen, M. 1971. Selforganization of matter and the evolution of biological macro-

molecules. Naturwissenschaften 58: 465–523.Eigen, M. 2002. Error catastrophe and antiviral strategy. Proc. Natl. Acad. Sci. USA

99: 13374–13376.Fitch, W.M., and K. Upper. 1987. The phylogeny of tRNA sequences provides evidence

for ambiguity reduction in the origin of the genetic code. Cold Spring Harbor Symp.Quant. Biol. 52: 759–767.

Forterre, P. 1999. Displacement of cellular proteins by functional analogues from plas-mids or viruses could explain puzzling phylogenies of many DNA informationalproteins. Mol. Microbiol. 33: 457–465.

Forterre, P. 2002. The origin of DNA genomes and DNA replication proteins. Curr.Opin. Microbiol. 5: 525–532.

Forterre, P., C. Brochier, and H. Philippe. 2002. Evolution of the Archaea. Theor. Pop.Biol. 61: 409–422.

Forterre, P., and H. Philippe. 1999. The last universal common ancestor (LUCA),simple or complex? Biol. Bull. 196: 373–375.

Galtier, N., N. Tourasse, and M. Gouy. 1999. A nonhyperthermophilic common ancestorto extant life forms. Sciences 283: 220–221.

Giraldo, R. 2003. Common domains in the initiators of DNA replication in Bacteria,Archaea and Eukarya: combined structural, functional and phylogenetic perspec-tives. FEMS Microbiol. Rev. 26: 533–554.

316 David Moreira, Purificacion Lopez-Garcıa

Gogarten, J.P., H. Kibak, P. Dittrich, L. Taiz, E.J. Bowman, B.J. Bowman,M.F. Manolson, R.J. Poole, T. Date, T. Oshima, J. Konishi, K. Denda, and M.Yoshida. 1989. Evolution of the vacuolar H+-ATPase: implications for the originof eukaryotes Proc. Nat. Acad. Sci. USA 86: 6661–6665.

Gogarten-Boekels, M., E. Hilario, and J.P. Gogarten. 1995. The effects of heavy mete-orite bombardment on the early evolution–the emergence of the three domains oflife. Orig. Life Evol. Biosph. 25: 251–264.

Gribaldo, S., and P. Cammarano. 1998. The root of the universal tree of life inferredfrom anciently duplicated genes encoding components of the protein-targeting ma-chinery. J. Mol. Evol. 47: 508–516.

Joyce, G.F. 2002. The antiquity of RNA-based evolution. Nature 418: 214–221.Kandler, O. 1994. The early diversification of life. In: Bengston, S. (ed.) Early Life on

Earth, pp. 152–160, Columbia University Press, New York.Koga, Y., T. Kyuragi, M. Nishihara, and N. Sone. 1998. Did archaeal and bacterial

cells arise independently from noncellular precursors? A hypothesis stating that theadvent of membrane phospholipid with enantiomeric glycerophosphate backbonescaused the separation of the two lines of descent. J. Mol. Evol. 46: 54–63.

Koonin, E.V. 2003. Comparative genomics, minimal gene-sets and the last universalcommon ancestor. Nat. Rev. Microbiol. 1: 127–136.

Kornberg, A. 2000. Ten commandments: lessons from the enzymology of DNA replica-tion. J. Bacteriol. 182: 3613–3618.

Lazcano, A., and S.L. Miller. 1996. The origin and early evolution of life: prebioticchemistry, the pre-RNA world, and time. Cell 85: 793–798.

Leipe, D.D., L. Aravind, and E.V. Koonin. 1999. Did DNA replication evolve twiceindependently? Nucleic Acids Res. 27: 3389–3401.

Lopez-Garcıa, P. 1999. DNA supercoiling and temperature adaptation: a clue to earlydiversification of life? J. Mol. Evol. 49: 439–452.

Martin, W., and M.J. Russell. 2003. On the origins of cells: a hypothesis for the evo-lutionary transitions from abiotic geochemistry to chemoautotrophic prokaryotes,and from prokaryotes to nucleated cells. Philos. Trans. R. Soc. Lond. B. Biol. Sci.358: 59–83.

Moreira, D. 2000. Multiple independent horizontal transfers of informational genes frombacteria to plasmids and phages: implications for the origin of bacterial replicationmachinery. Mol. Microbiol. 35: 1–5.

Olsen, G.J., and C.R. Woese. 1997. Archaeal genomics: an overview. Cell 89: 991–994.Oparin, A.I. 1938. The Origin of Life. Mac Millan, New York.Pace, N.R. 1991. Origin of life–facing up to the physical setting. Cell 65: 531–533.Pace, N.R. 2001. The universal nature of biochemistry. Proc. Natl. Acad. Sci. USA 98:

805–808.Pereto, J., P. Lopez-Garcia, and D. Moreira. 2004. Ancestral lipid biosynthesis and

early membrane evolution. Trends Biochem. Sci. 29: 469–477.Pereto, J.G., A.M. Velasco, A. Becerra, and A. Lazcano. 1999. Comparative biochem-

istry of CO2 fixation and the evolution of autotrophy. Int. Microbiol. 2: 3–10.Philippe, H., and P. Forterre. 1999. The rooting of the universal tree of life is not

reliable. J. Mol. Evol. 49: 509–523.Poole, A., D. Jeffares, and D. Penny. 1999. Early evolution: prokaryotes, the new kids

on the block. Bioessays 21: 880–889.

11 The Last Common Ancestor of Modern Cells 317

Raoult, D., S. Audic, C. Robert, C. Abergel, P. Renesto, H. Ogata, B. La Scola,M. Suzan, and J.M. Claverie. 2004. The 1.2-megabase genome sequence ofMimivirus. Science 306: 1344–1350.

Russell, M., and A.J. Hall. 1997. The emergence of life from iron monosulfide bubblesat a submarine hydrothermal redox and pH front. J. Geol. Soc. Lond. 154: 377–402.

Russell, M.J., and W. Martin. 2004. The rocky roots of the acetyl-CoA pathway. TrendsBiochem. Sci. 29: 358–363.

Stetter, K.O. 1996. Hyperthermophilic prokaryotes. FEMS Microbiol. Rev. 18: 149–158.Wachtershauser, G. 1990. Evolution of the first metabolic cycles. Proc. Natl. Acad. Sci.

USA 87: 200–204.Wachtershauser, G. 2003. From pre-cells to Eukarya–a tale of two lipids. Mol. Micro-

biol. 47: 13–22.Wachtershausser, G. 1988. Before enzymes and templates: theory of surface

metabolism. Microbiol. Rev. 52: 452–484.Woese, C. 1998. The universal ancestor. Proc. Natl. Acad. Sci. USA 95: 6854–6859.Woese, C.R. 2000. Interpreting the universal phylogenetic tree. Proc. Natl. Acad. Sci.

USA 97: 8392–8396.Woese, C.R., and G.E. Fox. 1977a. Phylogenetic structure of the prokaryotic domain:

the primary kingdoms. Proc. Natl. Acad. Sci. USA 74: 5088–5090.Woese, C.R., and G.E. Fox. 1977b. The concept of cellular evolution. J. Mol. Evol. 10:

1–6.Woese, C.R., O. Kandler, and M.L. Wheelis. 1990. Towards a natural system of organ-

isms: proposal for the domains Archaea, Bacteria, and Eucarya. Proc. Natl. Acad.Sci. USA 87: 4576–4579.

Zhaxybayeva, O., and J.P. Gogarten. 2004. Cladogenesis, coalescence and the evolutionof the three domains of life. Trends Genet. 20: 182–187.