REVIEW Open Access

Is pre-Darwinian evolution plausible?Marc Tessera

Abstract

Background: This essay highlights critical aspects of the plausibility of pre-Darwinian evolution. It is based on acritical review of some better-known open, far-from-equilibrium system-based scenarios supposed to explainprocesses that took place before Darwinian evolution had emerged and that resulted in the origin of the firstsystems capable of Darwinian evolution. The researchers’ responses to eight crucial questions are reviewed. Themajority of the researchers claim that there would have been an evolutionary continuity between chemistry and“biology”. A key question is how did this evolution begin before Darwinian evolution had begun? In other wordsthe question is whether pre-Darwinian evolution is plausible.

Results: Strengths and weaknesses of the reviewed scenarios are presented. They are distinguished betweenmetabolism-first, replicator-first and combined metabolism-replicator models. The metabolism-first scenarios showmajor issues, the worst concerns heredity and chirality. Although the replicator-first scenarios answer the heredityquestion they have their own problems, notably chirality. Among the reviewed combined metabolism-replicatormodels, one shows the fewest issues. In particular, it seems to answer the chiral question, and eventually impliesDarwinian evolution from the very beginning. Its main hypothesis needs to be validated with experimental data.

Conclusion: From this critical review it is that the concept of “pre-Darwinian evolution” appears questionable, inparticular because it is unlikely if not impossible that any evolution in complexity over time may work withoutmultiplication and heritability allowing the emergence of genetically and ecologically diverse lineages on whichnatural selection may operate. Only Darwinian evolution could have led to such an evolution. Thus, Pre-Darwinianevolution is not plausible according to the author. Surely, the answer to the question posed in the title is aprerequisite to the understanding of the origin of Darwinian evolution.

Reviewers: This article was reviewed by Purificacion Lopez-Garcia, Anthony Poole, Doron Lancet, andThomas Dandekar.

Keywords: Metabolism-first, Replicator-first, Combined metabolism-replicator scenarios, Pre-Darwinian evolution,Prebiotic evolution, Darwinian evolution, Origin of life, Origin of Darwinian evolution

BackgroundThis essay highlights critical aspects of the plausibility ofpre-Darwinian evolution. It is based on a critical reviewof some of the better-known open far-from-equilibriumsystem-based scenarios supposed to explain the likelyprocesses that took place in times when Darwinian evo-lution had not yet emerged and resulted in the origin ofthe first systems capable of Darwinian evolution. Amongthe current views about the origin of “life” a phase of“prebiotic” chemical evolution is contemplated beforethe emergence of biological evolution, which the authorprefers to call Darwinian evolution. According to those

researchers that propose “prebiotic evolution”, this de-scribes the chemical processes that took place on the“prebiotic Earth” about 4500 to 3500 mya ago. Theseevents preceded biological evolution, a phase which ledto the appearance of first “living cells” capable ofself-reproduction at the expense of some rudimentarymetabolism [1]. According to this view the study of the“origin of life”’ is an effort to understand the transitionfrom chemistry to biology, a fundamental transition anda lengthy pathway comprising many stages, each ofwhich is the subject of numerous scientific questions [2].This would imply an evolutionary continuity between

chemistry and biology [3–6]. The question then becomeshow such a disordered (high entropy) primordial cellsystem can be re-organized into an evolving “living

Correspondence: marc.tessera@wanadoo.frMeudon, France

© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Tessera Biology Direct (2018) 13:18 https://doi.org/10.1186/s13062-018-0216-7

state”, or whether there are compelling theoretical ortechnical reasons why this is impossible [3].In other words how did evolution begin if the complex

machinery for Darwinian evolution was not in place? [2].When the emergence of Darwinian evolution is consid-ered (often cited as a key aspect of the definition of lifewith good reason, as Darwinian evolution is indeed theunifying characteristic of all of biology) once cells withgenetically encoded advantageous functions existed, clas-sically defined Darwinian evolution had begun accordingto J. Szostak. But what is about the previous steps? [7].Actually the expression “pre-Darwinian evolution” willbe preferred in this article instead of “chemical evolu-tion” or “prebiotic evolution” except when the re-searchers the author cites used them. According to theauthor Darwinian evolution is a mechanism that can beclearly defined while the concept of life is questionable[8–12]. As the concept of pre-Darwinian evolution is de-batable (and questioned in this paper), such an evolutionis only defined in reference to Darwinian evolution, i.e.,as a chemical or metabolic evolution in complexity overtime before Darwinian evolution had emerged. Systemswith the ability of Darwinian evolution should have thefollowing characteristics according to the author (1) beopen far-from-equilibrium systems able to remainfar-from-equilibrium by feeding from their environment;(2) capable of multiplication into similar systems, and(3) of heritability [10].

Pre-Darwinian or Darwinian evolution scenarios?Most authors support the idea that only open far-fro-m-equilibrium systems are able to show “chemical evolu-tion”, pre-Darwinian evolution according to this author’susage. Here, only seeming plausible scenarios or modelswithin the paradigm of open far-from-equilibrium systemsare reviewed. Each model is evaluated according to thefollowing questions: (1) What was the initial chemical sub-strate? (2) Are there experimental data supporting themain hypothesis? (3) What was the energy source? (4) Isthe chiral question solved? (5) What is the ability ofmultiplication of the systems? (6) Is there any hered-ity? (7) Were there plausible site(s) where the initialchemical substrates might arise? (8) Is the evolution-ary path plausible? (Table 1: Strengths and weaknessesof the reviewed scenarios for the origin of Darwinianevolution). The majority of the researchers highlightthe chiral question. For example József Garay claimsthat any theory of origin of life that does not explainthe origin of homochirality in biomolecules is notcomplete [13].It is important to distinguish between metabolism-first,

replicator-first and combined metabolism-replicatormodels. The metabolism-first hypothesis means that or-dered chemical reactions, and not replication was the

property of the initial form that led to “life” and the inter-locked networks of chemical reactions that evolved incomplexity over time. Replicator-first scenarios requireboth replication and at least some heredity. Finally it willbe attempted to specify which initial evolution is hypothe-sized by the authors in their model, i.e., pre-Darwinian orDarwinian.

Metabolism-first scenariosIron-sulfur world modelWächtershäuser proposes what he calls the “FeS Worldtheory”, also called the “iron-sulfur world.” This positsthat life began as an autotrophic metabolism in hotvolcanic-hydrothermal fluids and evolved with organicproducts turning into ligands for transition metal cata-lysts thereby eliciting feedback and feed-forward effects.It postulates a “pioneer organism” at sites of reducingvolcanic exhalations. The pioneer organism is character-ized by a composite structure with an inorganic sub-structure and an organic superstructure. Within thesurfaces of the inorganic substructure iron, cobalt, nickeland other transition metal centres with sulphido, car-bonyl and other ligands were catalytically active and pro-moted the growth of the organic superstructure throughcarbon fixation, driven by the reducing potential of thevolcanic exhalations. This pioneer metabolism would re-produce by an autocatalytic feedback mechanism. Somesynthetic organic products would have exhibited auto-catalytic positive feedback into the synthetic reactionsthat produced them, which equates to reproduction ac-cording to this author. Other organic products served asligands for activating catalytic metal centres fromwhence they arose. In this context Wächtershäuserposits that the three polymers of the genetic machineryessentially coevolved from monomers through oligomersto polymers. According to Wächtershäuser, the unitarystructure–function relationship of the pioneer organismlater gave rise to two major strands of evolution: cellu-larization and the emergence of the genetic machinery[14]. Regarding chirality Wächtershäuser makes severalpostulates: (i) the enzymes for the synthesis of the chiralphosphoglycerol lipids of the pre-cells were notenantio-specific and therefore generated racemic lipids;(ii) the racemic lipid membranes of the pre-cells con-tinuously underwent spontaneous symmetry breaking byspatial lipid segregation into an interdigitatedmicro-pattern of two homochiral lipid domains withineach pre-cell membrane; (iii) the racemic pre-cell mem-brane, while not as stable as a homochiral membrane,was stable enough for maintaining the integrity of hyper-thermophilic pre-cells and for maintaining a definite or-ganism–environment dichotomy over a long period ofevolution, perhaps over hundreds of million years. In

Tessera Biology Direct (2018) 13:18 Page 2 of 18

Table

1Streng

thsandweaknessesof

thereview

edscen

ariosfortheorigin

ofDarwinianevolution

Scen

arios

Initial

Che

mical

Substrate

Expe

rimen

tal

Dataon

main

Hypothe

sis

Energy

Source

ChiralQ

uestion

Multip

lication

abilities

Hered

ityPlausibleSites

Evolutionary

Path

Initial

Evolution

Claim

ed

Metabolism-

First

Iron-sulfur

world

Pion

eer

organism

Lacking

Volcanic-

hydrothe

rmal

fluids

Not

anissue

accordingto

theauthor

Not

truly

presen

tNot

clearly

demon

strated

Volcanic

aquifers

Questionable

Pre-

Darwinian

Vasaset

al.

Lacking

Lacking

Not

specified

Not

addressed

Presen

tAbstract

Not

specified

Plausible

Darwinian

Fernando

&Ro

we

Lipid

vesicles

Lacking

Solar

Not

addressed

Presen

tSome

Not

specified

Com

plex

Pre-

Darwinian

Russell

Inorganic

mem

branes

Prelim

inary

White

smokers

Not

addressed

Unkno

wn

No

White

smokers

Questionable

Pre-Darwinian

Replicator-

First

RNA/Pre-

RNAworld

RNAor

pre-RN

APresen

tNot

specified

Not

addressed

Presen

tPresen

tNot

specified

Plausible

Pre-Darwinian

Com

bine

dDKS

Lacking

Lacking

Solar

Not

addressed

Not

trulypresen

tNot

addressed

Not

clearly

specified

Questionable

Darwinian

Lancet

&Segré

Lipidmicelles

Lacking

Volcanoes,

hydrothe

rmal

vents,solar

Not

anissue

accordingto

theauthor

Presen

tQuestionable

Volcanoe

s,hydrothe

rmal

vents,solar

Questionable

Pre-Darwinian

Tessera

Lipidvesicles

Lacking

White

smokers

Possibly

solved

Presen

tPresen

tWhite

smokers

Possiblyhand

icappe

din

thepo

lymerization

process

Darwinian

Tessera Biology Direct (2018) 13:18 Page 3 of 18

any case, Wächtershäuser does not seem to considerchirality as an issue [14].From a chemical point of view, the theory could be ex-

perimentally tested on its predictive power, i.e. its powerto predict hitherto unknown chemical reactions. Wäch-tershäuser admits that the reactions demonstrated so farmay well be the tip of an iceberg of synthetic reac-tions for a pioneer organism that has yet to be dis-covered experimentally. Thus, the main hypothesis ofthe model, i.e., the possible existence of the so-called“pioneer organism” has not yet been confirmed by ex-perimental data. Another issue is, as noticed byWächtershäuser, that fluid flow through a localizedvolcanic-hydrothermal flow duct is of a short dur-ation, too short to cover a long period of evolution.To overcome the issue the author hypothesizes thatintense ejection and scattering of crustal materials,and their settlement in renewed flow beds would havecaused the process to be resumed time and timeagain with ever more advanced catalytic endowmentsand cellular organization [14]. JF Allen points out aproblem with autocatalysis. The product of a series ofreactions is also their substrate. Autocatalysis be-comes more plausible if the autocatalytic cycle takesplace in a confined space, so that the concentrationof its components can build up. Surface catalysisalone cannot easily explain how a reactant becomesconcentrated, since the volume of the solute is, for allpractical purposes, infinite, being, one supposes, thevolume of water in all the oceans of the World. On aflat surface, autocatalysis will tend to be prevented, orquenched, by dilution [14]. Wächtershäuser arguesthat inorganic compartments exist, and can also beproduced, apparently, in the laboratory, as a result ofmixing two solutions, which have solutes that react toform a ‘froth’ of insoluble precipitates, citing Russell’shypothesis that such structures, consisting of vesiclesbounded by sulphides of iron and nickel, served asthe first ‘incubators’ of life by permitting, containing,and sustaining chemoautotrophic synthesis [15].Wächtershäuser claims that there is reproduction be-cause some of the synthetic organic products exhibitan autocatalytic positive feedback into the syntheticreactions whence they arise. This is questionable as atrue reproduction is not really the multiplication ofspecific chemical compounds but rather the multipli-cation of individual systems leading to a populationof similar systems. The author adds that evolution oc-curs because the autocatalytic feedback effect exhibitsvariations and the existence of an autocatalytic feed-back cycle signifies the historic fact of its previous denovo ignition and in this sense, each autocatalyticfeedback effect constitutes an instance of inheritanceor a memory effect [14]. Such a speech is obscure to

me and does not sound very much scientific. Hence,Wächtershäuser’s claim that his model is able toevolve is questionable, at best. According to the au-thor the first stage of cellularization is the formationof acetyl thioester as evolutionary precursor ofacetyl-CoA. Then, there would have been a very com-plex chemical evolutionary path leading to the forma-tion of the first lipids followed by a transition fromsurface lipophilization to a surface-supported bilayermembrane. Another complex evolutionary path wouldhave followed leading to semi-cellular structures,which would have constituted the beginning of indi-viduation. Finally, again another complex evolutionarypath would have led to a full cellularization [14]. Thisis a nice story, but more a fairy tale than a scientificreasoning.

Non-templated systems of replicators modelSome authors notice that there are two camps in the ori-gin of life. The metabolism- first camp advocates con-sider improbable that RNA-like self-replicating polymersappeared before natural selection had operated onchemical networks, whereas genetics-first supportersfind implausible the idea that molecular networks with-out genetic control could have undergone Darwinianevolution. These authors think that a solution to theconundrum can be found in general evolutionary princi-ples shared by some chemical and biological systems[16]. The origin of the first hereditary replicators is stillan unsolved problem. On its own, that transition is notevolutionary because, without hereditary replicators, noDarwinian evolution is possible. There should be a grayzone where chemistry and evolution had the first overlap[17]. A scenario should be based on the evolutionary dy-namics of populations of replicator molecules because ithas the advantage of providing a potential solution totwo fundamental problems faced by prebiotic systems atthe same time: (i) the physiological problem of overcom-ing the degradation tendency of any complex molecule,like an oligomer or a polymer; and (ii) the evolutionaryproblem of transmitting the selective advantages of thatcomplex molecule to the offspring [18]. Thus, a schemeis proposed for how Darwinian evolution could have oc-curred prior to template replication. The authors use anabstract chemistry based on autocatalytic chemical net-works with a particular structure and existing in net-works of compartments. These simulations did notconfirm previous claims that autocatalytic sets of organicpolymer molecules could undergo evolution in any inter-esting sense by themselves. It was discovered that if gen-eral conditions are satisfied, the accumulation ofadaptations in chemical reaction networks could occur.These conditions are the existence of rare reactions pro-ducing viable cores (analogous to a genotype) that

Tessera Biology Direct (2018) 13:18 Page 4 of 18

sustains a molecular periphery (analogous to a pheno-type) [16]. More recently they state that they believethey have shown limited heredity of a few types hencelimited evolvability. They also admit that the system sizeis unknown. They claim that the most exciting form ofevolvability is indefinite, open-ended on going evolution,possibly leading to an increase in complexity, at least incertain lineages [19].The model in Vasas et al.’s paper traditionally belongs

to the metabolism-first scenarios granted they show thatthere are replicators in their system, but replicator in thestrict sense only means an entity capable of autocatalyticgrowth. Thus, this model has been considered asmetabolism-first. William Martin and Eugene Kooninnoted the abstractness of the model, which far-removedfrom real chemistry and real metabolism. The authorsalso overlook the possibility of preexisting inorganic cat-alysts that are not part of their reaction products andthe possibility that things chemical fall into place morealong the lines of thermodynamics than along the linesof mathematics. Moreover, the model requires independ-ent viable autocatalytic cores embedded in a large mo-lecular network as units of evolution [16]. EugeneKoonin argued that it is difficult to understand how andwhere such large complex molecular networks couldhave emerged on early Earth and how such a putativenon-templated system of replicators could have beenstable in real time [16]. In addition, the authors specifiedthat their model would only shown limited evolvabilitywhile they claim that the most exciting form of evolva-bility is indefinite, open-ended on going evolution. Be-sides the authors don’t address the chiral questioncompatible with their model.

Autocatalytic cycles in liposomes modelVasas et al. say, in terms of “real” chemistry they havebeen inspired by the concept of autocatalytic proteinnetworks but they mention the model by Fernando andRowe that shows a similar evolutionary mechanismbased on random biomolecular rearrangements ratherthan ligation and cleavage reactions [16]. Fernando andRowe quote Maynard-Smith, “Natural selection is an al-gorithm that operates in populations of entities capableof multiplication, variation and heredity” [20]. The ques-tion they address is “what is the simplest ‘machine’ cap-able of implementing the natural selection algorithm,and how likely was this implementation to arise spontan-eously?” Thus, they opt for an individual- orcompartment-first hypothesis based on a population oflipid aggregates. They hypothesize that a geophysicalprocess would have produced such lipid aggregates with-out specifying the process. The lipid aggregates couldhave undergone division by externally-imposed agitationand the authors assume that specific intra-lipid

aggregate chemical reactions are not required for a baserate of lipid aggregate replication. Then a novel specieshas one probability of being in the lipid phase, soremaining within the lipid aggregate, and another prob-ability of being in the water phase, so being extrudedfrom the lipid aggregate. During division approximatelyhalf of the lipophilic constituents of the lipid aggregateare inherited by each daughter cell. The authors makethe assumption that a potential autotrophic reaction ex-ists that is capable of utilizing light energy to drive anotherwise non-spontaneous reaction. They claim segre-gation of the population into lipid aggregates allowsharmful avalanches to be isolated to a compartment. Fi-nally the autocatalytic cycles would have evolved in theliposomes because they were necessary for the mainten-ance of chemical reaction networks that were beneficialat the liposome level. According to the authors, the sys-tem should have shown natural selection at both thelevel of autocatalytic molecules and lipid aggregates [21].This model would have been able firstly to show pre-Darwinian evolution and latter on Darwinian evolution.Surely, their model is less abstract than Vasas et al.’s

because it is based on lipid aggregates. While there areindeed possible processes that lead to symmetry break-ing the authors do not specify how these processeswould have operated in their model and have thussolved the chiral question. In addition, they are unableto specify where on early Earth there would have been ageophysical process-producing lipid aggregates with apotential autotrophic reaction capable of utilizing lightenergy.

Inorganic membrane in alkaline hydrothermal vent modelIn 1989 Russell et al. had already predicted the proper-ties of deep-ocean alkaline hydrothermal vents morethan a decade before their discovery [22]. Then, theysuggested the prime exergonic reaction underpinningthe synthesis of organic molecules would have been thereduction of CO2 itself with electrons stemming fromH2. The H2 driving CO2 reduction stems in turn fromserpentinization. The fixation of CO2 required energywhich could have been readily supplied by H2 itself, aug-mented by an ambient chemiosmotic gradient actingacross the inorganic membrane and/or the catalyticcoupling of exergonic redox reactions to those yieldingformate/formyl [23]. At the time of early Earth atmos-pheric CO2 concentrations were much higher than todayand molecular oxygen was absent giving very differentocean chemistry from today. High CO2 made the oceansmildly acidic (pH 5.5–6) compared with pH 8 today,which, in the absence of O2, allowed reduced transitionmetals, most significantly Fe2

+ and Ni2+, to accumulate

in the early oceans. These metals exhaled from volcanicvents (possibly nearby), gave rise to mineral precipitates

Tessera Biology Direct (2018) 13:18 Page 5 of 18

at alkaline vents such as silicates, clays, carbonates, andsulfides. For example, at a Lost-City-type vent in anearly-Earth setting, this chemistry delivers catalyticFe(Ni)S minerals. [24]. This model would have followedchemical evolution (i.e., pre-Darwinian evolution) [23].In favor of this model are the first results of recent ex-

periments with a simple electrochemical reactor tosimulate conditions in alkaline hydrothermal vents,allowing that abiotic vent chemistry could prefigure theorigins of biochemistry. The precipitation of thin-walled,inorganic structures containing nickel-doped mackina-wite, a catalytic Fe(Ni)S mineral, under prebiotic oceanconditions showed the generation of low yields of simpleorganics while synthetic microporous matrices were ableto concentrate organics by thermophoresis over severalorders of magnitude under continuous open-flow ventconditions. These experiments provide empirical evi-dence that simple organics can be generated and con-centrated under mild alkaline hydrothermal conditionsfrom H2 and CO2 using Fe(Ni)S catalysts transected bynatural proton gradients in microporous matrices [25].It is questionable whether these processes could lead tosomething else than simple inactivated organics. More-over, there are severe problems for the natural pH gradi-ent hypothesis. Firstly, there is the difficulty ofinterposing a suitable, stable, thin inorganic membranebetween the alkaline vent fluid and the acidic ocean thatcould hold a sharp pH gradient. At Lost City, there is lit-tle or no evidence for the existence of stable inorganicmembranes, let alone of membranes holding sharp pHgradients. The second and even more difficult problemis that it is hard to accept that a molecular machine cap-able of abstracting useful energy from the natural Δ pHwas assembled within the inorganic membrane bychance on the prebiotic earth, that is before the adventof gene encoded proteins and the great power of naturalselection. Thirdly, a simple, small, Δ pH-consuming mo-lecular machine, even if by chance adopting an as yetunknown chemical structure, and a set of barely sus-pected, facilitating, mechanistic principles, will be unableto function in an inorganic membrane of a thicknessdeemed acceptable by Russell et al. [26]. Even when theauthors supposed that inorganic molecular machinesmight assemble by chance in the precipitate membranes,and be capable of using the Δ pH to drive unfavourablereduction of CO2 by H2 to formate and formaldehyde(and indeed, these workers detected both of these com-pounds in their origin-of-life reaction vessel and con-tended that was proof of principle for their hypothesis)JB Jackson demonstrated by a straightforward calcula-tion that the formate produced was only that reached onapproach to equilibrium without any driving force fromΔ pH. Jackson therefore concluded that the reaction wasfacilitated by isotropic catalysts in the precipitate

membrane but not by an anisotropic ΔpH-driven mo-lecular machine [27]. Moreover, the problem of the tran-sition from inorganic to organic membranes is faced.According to Russell et al. the organic takeover from themineral precursors would have been facilitated by thesurface structure of amyloidal peptides between 6 and10 amino acids long that behave as nests for inorganicclusters. Together these would do the work ofproto-enzymes [28]. The authors do not specify how ho-mochiral amino acids would have been produced in theirmodel and about the way their metabolic systems couldreplicate and transmit their characteristics to the pro-geny. Therefore the evolutionary path from the mineralprecursors to amyloidal peptides is unlikely withoutsolving the chiral issue. Moreover, the evolutionary pathfrom vesicles with membranes composed of amyloidalpeptides to the first anaerobic chemolithotrophy organ-isms using Earth’s inorganic geochemicals is not straight-forward without the existence of some kind ofDarwinian evolution and in particular of heredity.

Conclusions on metabolism-first scenariosThe four metabolism-first scenarios discussed above aresupposed to show pre-Darwinian evolution. They showmajor issues as follows: no practical chemical substrateis known (Vasas et al.); experimental data supporting themain hypothesis on which the model is based are lacking(iron-sulfur world, Fernando & Rowe, Vasas et al.) orpreliminary (Russell); the source of energy is question-able (Fernando & Rowe’s) or not specified (Vasas et al.);the chiral question is not addressed (Fernando & Rowe,Russell, Vasas et al.) or not considered as an issue (iron--sulfur world); multiplication abilities are not trulypresent (iron-sulfur world) or unknown (Russell), hered-ity is not clearly demonstrated (iron-sulfur world) or ab-sent (Russell), a plausible site on early Earth is unknownor not specified (Fernando and Rowe and Vasas et al.);the system is unstable (Vasas et al); the evolutionarypath seems either complex (Fernando and Rowe,) orquestionable (iron-sulfur world, Russell).

Replicator-first scenarios: RNA and pre-RNAworldsThe RNA world hypothesis describes an early Earth withself-replicating and catalytic RNA but no DNA or pro-teins. The possibility of an RNA world was suggestedwell before Gilbert coined the name in 1986 [29]. Benneret al. have well summarized both the arguments in favorof the hypothesis of an RNA world and the main pitfalls[30]. They noticed that the most direct evidence for anRNA World, crystallographic evidence that the RNA inthe ribosome catalyzes peptide bond, is applied within ahistorical argument, one that infers that all ribosomes inbiology use RNA to synthesize peptide bonds, and

Tessera Biology Direct (2018) 13:18 Page 6 of 18

therefore the last common ancestor of all ribosomesused RNA to synthesize peptide bonds. In the modelthere would be first pre-Darwinian evolution until theemergence of RNA and then Darwinian evolution.The idea of an RNA world was actually built upon the

diverse roles of RNA in contemporary metabolism andnot upon the limited catalytic role of these natural ribo-zymes. Anyway a major question is how do the firstself-replicating ribozymes emerge in the absence of tem-plate-directed information replication? [28]. Admittedly,the group of Sutherland shows that activated pyrimidineribonucleotides can be formed in a short sequence thatbypasses free ribose and the nucleobases, and instead pro-ceeds through arabinose amino-oxazoline and anhydronu-cleoside intermediates [31]. Even among those searchingfor a prebiotic origin of RNA, Sutherland’s approach isnot universally regarded as satisfactory and critics of thattype of prebiotic chemistry find that an excessive Deus exmachina, but also note that if amino-oxazole is to beformed in a different locale than glyceraldehyde, the effortto find pH neutral conditions to form amino-oxazole (rec-ognizing the instability of glyceraldehyde at high pH) wasunnecessary. If this were not sufficient, the final step re-quires yet another change of environments, to urea orformamide as a reaction mixture, not water [30]. More re-cently the same team provided experimental evidence insupport of the hypothesis that precursors of ribonucleo-tides, of amino acids and of lipids have a common chem-ical origin [32]. In their experiments precursors ofribonucleotides, amino acids and lipids can all be derivedby the reductive homologation of hydrogen cyanide andsome of its derivatives, and thus that all the cellular sub-systems could have arisen simultaneously through com-mon chemistry [33]. In a very recent article Powner et al.demonstrated a divergent, prebiotically plausible reactionstrategy for the synthesis of both pyrimidine and purine ri-bonucleotides on a single oxazoline scaffold [34]. Oneshould nevertheless accept that an important gap remainsin our understanding of the processes this could havespontaneously supported the emergence, maintenanceand evolutionary potential of such a precellular world.Various chemical reactions traditionally involved in theartificial organic syntheses of RNA precursors are thermo-dynamically uphill or show high kinetic barriers in the ab-sence of activating molecules or catalysts, respectively[35]. All these experiments use enantiomers in theirchemical reactions and thus the chiral question comesback. The origin of life, especially the origin of RNAworld, cannot be disclosed without explaining the originof chiral homogeneity of biomolecules [13]. There isample research on the chirality of RNA. In particular, J.Garay’s hypothesis is the construction of complexes by thechemo-autotrophic metabolism on mineral surfaces, e.g.,the iron-sulfur world proposed by Wächtershäuser [36].

He calls these “MAR (Metal-Amino-Ribo) complexes”: aMAR-complex is constituted of “a metal ion surroundedby co-enzymes (for example, amino acids or short peptides,etc.), and short RNAs”. His hypothesis does not reallylead to a model as such, as the author does not ad-dress questions such as: are there any experimentaldata supporting his hypothesis? What was the energysource? Were there plausible site(s) where the prim-ordial chemical substrates might arise? (Table:“Strengths and weaknesses of the reviewed scenariosfor the origin of life”). Secondly, the initial chemicalsubstrate of his hypothesis is what he calls a “MAR--complex” that seems to answer the chiral question.The author only answers the question if there wereMAR-complexes at the very beginning. He likewiseanswers the questions about replication, heredity, andthe evolutionary path from the same hypothesis. Theinescapable question is: how could the abiotic emer-gence of MAR-complexes be explained? Specificallyhow could the abiotic synthesis of RNA and peptidesbe explained? While Garay agrees that the origin ofthe RNA world has not been explained satisfactorilyas yet he now supposes that RNA and peptides couldhave been synthesized abiotically, in particular beforethe emergence of chirality. From a theoretical per-spective, M. Stich et al. say, the current state ofknowledge strongly suggests that the emergence ofchirality must be based on reactions leading to spon-taneous mirror symmetry breaking (SMSB). SMSB aretransformations yielding chiral outcomes asnon-thermodynamic final stable states, and in the ab-sence of any chiral polarization or external chiralphysical forces. This is provided by enantioselectiveautocatalysis, but not by the simple linear asymmetricinduction reactions on which past discussions on de-terministic or chance phenomena were based for thejustification of biological homochirality. These authorsconcluded that the racemic outcome results evenwhen the autocatalytic cycles are driven irreversiblyby external reagents, in manifestly non-equilibriumconditions. The stability of the thermodynamic limitproves that the racemic outcome is the unique stablestate for strictly irreversible externally driven auto-catalysis [37]. These theoretical results are clearly infavor of the claim that chiral question is not a trivialquestion. P. Joshi et al. studied the selective adsorp-tion on the platelets of montmorillonite. They dem-onstrate that catalytic montmorillonite platelets serveas selective templates for the formation of RNA oligo-mers. In their experiments the authors start frommonomers of nucleotides. Thus, there remains thequestion of the abiotic synthesis of the monomers[38]. In their experiments for a cross-chiral RNApolymerase J. Sczepanski and G. Joyce began with a

Tessera Biology Direct (2018) 13:18 Page 7 of 18

population of 1015 random-sequence D-RNAs. Intheir experiments the authors used RNA molecules asingredients [39]. Hence, again, the question of theabiotic synthesis of RNA is still there. It can be con-cluded, at least from these publications, that thechiral question is not yet clearly answered. Among in-surmountable problems of chemical and informationalnature are the unreliability of the synthesis of initialcomponents, the instability of the molecules whichincreases with chain elongation, the catastrophicallylow probability of meaningful sequences, the lackingmechanism of the formation of membrane-bound ves-icles permeable for the nitrogenous bases and otherRNA precursors as well as able to divide on a regularbasis [40]. The difficulty in synthesizing nucleotidemonomers, not to mention the idea that RNA itself istoo complex to have been the first genetic molecule,has focused researchers to search for non-standardmolecules as earlier replicators and catalysts and onother means of transitioning to standard nucleicacids. For example, it was proposed the idea that pep-tide nucleic acids (PNAs) might have preceded RNAnucleotides [41]. This would have been the pre-RNAworld. The difficulties remain open in synthesizingthe initial components within the paradigm of openfar from thermodynamic equilibrium systems and es-pecially because of the chiral question. Thus, it isunderstandable that Benner et al. note that it is notsurprising that Joyce and Orgel called RNA a “pre-biotic chemist’s nightmare”. Moreover, there is a para-dox that all the efforts which have focused on makingRNA under a gene first model for the origin of lifecuriously have come as close as any to providing aworking example of a cycle reminiscent of modelswhich put metabolism first. Another limitation of theRNA world has been hypothesized to be the paucityof building blocks in the RNA biopolymer. Certainly,proteins, with 20 amino acids, have a richer diversity offunctionality than RNA, and this functionality is useful forbinding and catalysis [30]. Finally, as Benner et al. specifyit, some have abandoned all genetic biopolymers, suggest-ing instead that something like Darwinian evolution musthave been supported by a set of small organic moleculesdissipating free energy in a cycle that through its oper-ation can adapt to changing conditions and evolve in aDarwinian sense [30].

Conclusions on the RNA and pre-RNA world scenariosThe only replicator-first scenarios discussed aboveshould show pre-Darwinian evolution, actually, firstlypre-Darwinian evolution and secondly Darwinian.There are crucial issues: the source of energy is notspecified, the chiral question is not addressed, and aplausible site on early Earth is unknown. In addition,

it has its specific issues (see above). An alternativeshould be therefore found.

Combined metabolism-replicator scenariosDynamic autocatalytic systems within the dynamic kineticstability concept (DKS)Some researchers claim that a chemical environmentheld in a non-equilibrium state by kinetic barriers con-stitutes a prerequisite for the development of very spe-cific dynamic chemical systems of constituents andcatalysts capable of circumventing these kinetic barriers.They support the concept of the dynamic kinetic stabil-ity, DKS [42–44]. According to these authors, the stabil-ity kind is the stability associated with entities able tomake copies of them at a rate that results in anon-equilibrium steady-state population of replicatingentities being maintained over time. The replicated en-tity, the replicator, is a specific molecule of the autocata-lytic metabolic system but not the system per se. Theauthors claim that when there are several autocatalyticmetabolic systems they will tend from less stable (per-sistent) to more stable (persistent) forms because of a lo-gical law of nature they term the persistence principle,i.e., only the most persistent systems will remain after acertain time. The model would be capable of Darwinianevolution according to the authors.When a metabolic system is very much persistent it

doesn’t mean that its composition is more complexand diversified. It would not be unlikely that simplermetabolic systems were more persistent. The evolu-tion of a given population of metabolic systems to-ward a population of increasingly complex anddiversified systems is not the most likely. Accordingto R. Pascal the solar radiation, with energy max-imum in the visible domain precisely correspondingto the requirement for self-organization would be thesource of energy for DKS systems [43]. Even if theaccumulated evidence suggests that photosynthesisbegan early in Earth’s history it was probably not oneof the earliest metabolisms [45]. The author does notaddress the chiral question. The authors also foundthat the replicated entity, the replicator, is a specificmolecule of the autocatalytic metabolic system butnot the system per se. Hence, it is questionable toclaim there is a true reproduction because, as inWächtershäuser’s model, the system does not multi-ply. In addition A. Pross and R. Pascal say nothingabout heredity [42–44]. Finally, R. Pascal admits thatno practical example has yet been identified of anautocatalytic network capable of both reproducing it-self, fed by light energy, and carrying information in away possibly leading to selection and accumulation ofgenetic information [43].

Tessera Biology Direct (2018) 13:18 Page 8 of 18

Compositional genomes in the lipid world modelAt Weizmann in Israel D. Lancet conceptualized amodel based on heterogeneous assemblies of diverselipid-like mutually catalytic amphiphilic molecules [46].The so-called “graded autocatalysis replication domain(GARD) model” utilizes chemical kinetics to simulate thebehavior of mutually catalytic sets as an alternative toalphabet-based inheritance. A basic feature in GARD isthat non-covalent, micelle-like molecular assemblies arecapable of growing homeostatically, i.e. catalyticallymaintaining the assembly’s composition as it grows (dy-namic buffering). The maintained compositional infor-mation can be propagated to progeny assemblies uponoccasional fission [47]. Energy sources available to drivethe synthesis reactions leading to the amphiphilic mole-cules would range from volcanoes and hydrothermalvents to solar photochemistry and pyrite-dependent re-duction [48]. According to the authors their GARDmodel would have had some capacity for pre-Darwinianevolution [46].Recently, it has been argued that collectively autocata-

lytic metabolic networks, such as the GARD, do notallow for fitter compositional genomes to be maintainedby selection [49]. The authors have replied, based on alarge number of simulations, that when quasi-stationarycomposomes rather than arbitrary compositions serve asselection targets, GARD networks are capable of a sig-nificant response to selection. Importantly, this can hap-pen chiefly when a high proportion of mutual catalysis ispresent in a GARD network [47]. However, the authors’hypothesis seems rather unlikely, i.e., that, in the condi-tions of the prebiotic earth, a high proportion of mutualcatalysis could have emerged. In addition, it is question-able that “compositional information” could have beentransferred to bilayer membrane lipid vesicles whenthese would have taken over the “micelle-like molecularassemblies”. In experiments looking at transitions frommicelles to vesicles the fatty acid in micelles are just in-corporated into the preformed vesicles without any pos-sibility to transmit a “compositional information” presentin the micelles [50–52]. Against this argument, accord-ing to D. Lancet (see his comment as a reviewer of thisarticle), is that the GARD model would not invoke acapacity of a small micelle to confer its composition ontoa much larger vesicular assembly when fusing with it. Ra-ther, it would invoke the possibility that homeostaticgrowth via single molecule accretion could gradually leadto increasingly larger assemblies having a similar compos-ition to that of the original micelle. Regarding the chiralquestion the authors described an extension of theirGraded Autocatalysis Replication Domain (GARD) model.The new model called Chiral-GARD (or C-GARD) is in-troduced to explain the symmetry breaking between left-or right-handed biomolecules in the modern biosphere.

Their main conclusion is that a strong symmetry breakingmay result from a relatively modest asymmetry in mutual(auto)catalytic activity: that is D-enantiomers which aremore likely to catalyze other D-enantiomers thanL-enantiomers, while L-enantiomers preferentiallycatalyze their L-brothers and sisters. The C-GARD modelwould highlight the possibility that chiral selection is a re-sult of, rather than a prerequisite for early life-like pro-cesses and thus would not have been an issue [53].Actually, the authors assume that all molecules in theC-GARD simulation are asymmetric as they claim that forsufficiently complex molecular structures it is justified toassume that essentially all molecules are chiral. Thus, theydisregard non-chiral isomers in their simulations. The au-thors’ assertion seems unrealistic, i.e., that, in the prebioticenvironment, there would have been sufficiently complexmolecular structures allowing to assume that all moleculeswere chiral.

Lipid vesicles in alkaline hydrothermal field modelIn accordance with the hypothesis that something likeDarwinian evolution must have been supported by a setof small organic molecules dissipating free energy in acycle that through its operation can adapt to changingconditions and evolve in a Darwinian sense [31] a modelbased on lipid vesicles with a bilayer membrane wasimagined. Considering the general importance and thevarious roles of compartments in contemporary cells(both in terms of general cell identity and processes) it isvery likely that the formation of some type of compart-ment already occurred in early pre-Darwinian times.Furthermore, most popular are vesicular structures,which currently attract considerable attention as primor-dial cell compartment models [32]. Such vesicles with amembrane composed of potentially abiotic lipidic am-phiphiles could have been produced in alkaline hydro-thermal fields akin to a Lost-City-type vent in anearly-Earth setting [54]. Their formation could have oc-curred from abiogenic production of short-chain hydro-carbons [55]. As provided for in abiotic chemicalreactions which typically produce very heterogeneousmixtures of compounds the vesicle bilayer membraneswould have been composed of mixtures of amphiphilesin such a way that the vesicle membranes become mo-lecularly, compositionally and organizationally highlycomplex, similarly to the lipid matrix of biological mem-branes [56, 57]. When hypothesizing a bi-layer mem-brane composed of a mixture of various distinctamphiphilic compounds there is the opportunity to gen-erate by chance a huge number of theoretically possiblecombinations of the amphiphiles leading to specific ar-rangements. Among the combinations, a specific localarrangement of amphiphiles could have appeared in theinner part of the membrane able to catalyze the

Tessera Biology Direct (2018) 13:18 Page 9 of 18

formation of larger products from small carbon-basedmolecules. Small molecules could easily penetrate themembrane while larger products would have tended tobe retained because of difficulty of the larger producedmolecules crossing back out of the membrane. Amonglarger products there were specific compounds able tocatalyze the transformation of the local membrane ar-rangement into a stabilized membrane site by the forma-tion of covalent bonds (instead of intermolecularH-bonds) most likely between the hydrophilic poles ofadjacent amphiphiles. Although the reaction ofstabilization was theoretically reversible, practically itwas not. This was because there was a very low likeli-hood of another interaction of the intra-vesicular solublechemical catalyst with the membrane site due to the di-lution of the former in the inner volume of the vesicle.Hence, a hypercycle according to the terminology byManfred Eigen [58] would have been formed. It was ac-tually a positive feedback produced by mutually catalyticsoluble catalyst/membrane site pairs. In the meantimethe increasing chemical catalyst concentration in thevesicle would have allowed the formation of new mem-brane sites by its interaction with the same kind of localmembrane arrangements and the sites would have beenmore or less randomly distributed over the inner surfaceof the vesicle membrane. When the vesicle divided intodaughter vesicles [59] the chemical catalyst and the mem-brane sites or only one element of the soluble catalyst/membrane site pair was likely to be present in the daugh-ter vesicles leading to the transmission of the hypercycle[10]. Chirality would have occurred by chance when com-pounds with a chiral centre carbon atome were synthe-sized from specific small carbon-based molecules. Inaddition, the model would have had the potential to favorthe polymerization of carbon-based molecules, for ex-ample leading to the synthesis of peptides [9]. As a re-minder, the kind of catalysis in the model mediated solelyby small organic molecules is comparable to organocataly-sis. Organocatalysis or “small-molecule catalysis” is a fieldrecently rediscovered in chemistry that in the past decadehas become a thriving area of general concepts and widelyapplicable asymmetric reactions [60]. C. Barbas believedthat this chemistry not only provides for fascinating andefficient syntheses of chiral molecules but also may serveto explain the emergence of homochirality in the prebioticworld [61]. This is one of our main hypotheses already de-veloped in a previous paper [9]. When the chemical cata-lyst is chiral it is possible to say that once a single chiralityis developed, a proto-organism could grow this way, anddivide, so propagating the chirality [62]. In the model thepolymerization process of the carbon-based molecules(e.g., amino-acids) needs the migration of the sites in themembrane to allow the formation of bi-sites, tri-sites etc.and thus polymerization [9]. When vesicles divide a few

bi-sites, tri-sites etc. may be transmitted to the daughtervesicles but cannot be replicated as such. Migration of thesites in the daughter vesicles is needed again to form otherbi-sites, tri-sites etc. Hence, this particular point should betested in a research programme [10]. Darwinian evolutionhas been implemented in the model since the beginning.The only pre-Darwinian processes at work were the for-mation of lipid vesicles with multiplication abilities andthe selection of the most viable. Before the emergence ofhypercycles genetically and ecologically distinct lineagescould not yet appear on which natural election might op-erate and the evolution over time of the composition andcomplexity of the vesicles was not yet possible. The mainissue of the model is the absence of experimental datatoday supporting the hypothesis that mutual catalysis ispossible in lipid vesicles with heterogeneous membranesin the environment of alkaline hydrothermal fields. A re-search programme has been proposed to confirm theplausibility of the model [10].

Conclusions on combined metabolism-replicatorscenariosCombined metabolism-replicator scenarios show majorissues too: no practical chemical substrate is known(DKS), experimental data supporting the main hypoth-esis on which the model is based are lacking (all scenar-ios); the source of energy is questionable (DKS), thechiral question is not addressed (DKS) or considered asnot essential (Lancet and Segré), replication is disputed(Lancet and Segré), a plausible site on early Earth is notclearly specified (DKS), the evolutionary path is ques-tionable (DKS) or unlikely (Lancet and Segré) or wouldhave been handicapped in case of a tediouspolymerization (Tessera).

DiscussionThe definition of the concept of “life” should be dis-cussed when considering the question of prebiotic evolu-tion from relatively simple chemistry to life. The firstmajor difficulty encountered when facing the problem ofhow life originated is that there is not a general consen-sus on what life is and on how it should be defined. Bycontrast, it is clear, since the pioneering work by CR.Woese and GE. Fox, that all the current biodiversity isthe outcome of Darwinian evolution from a primitivecellular species, the so-called last universal common an-cestor or LUCA [63]. Is it possible to define life scientif-ically, with strictly chemical and biological criteria? [64].The same authors find that although finding consensusabout the nature and definition of life is a very difficultissue, and will remain as a subject of debate probably fora long time, there is nowadays relatively widespreadagreement on which features should be shared by thesimplest “living systems”. They must possess a genetic

Tessera Biology Direct (2018) 13:18 Page 10 of 18

apparatus able to store and transmit information to theirprogeny, some sort of metabolism for gathering nutri-ents and energy from the environment, and a selectivelypermeable boundary that separates and distinguishesthem from this environment. They deduce that it isnecessary to develop chemistries which enable the syn-thesis of information-bearing polymers, protometabolicnetworks, and protocellular compartments under com-patible prebiotic conditions to explain how the first or-ganisms might have appeared on Earth, or elsewhere. Ifthere is a consensus that the first “living systems” shouldhave possessed a genetic apparatus able to store andtransmit information to their progeny, i.e., systems withhereditary property, the issue is how protometabolicnetworks without heredity would have had the capabilityto evolve towards systems with heredity while Darwinianevolution had not started operating yet [27]. Reviewedscenarios were divided between metabolism-first,replicator-first and combined metabolic-replicator. Allscenarios show major issues either because of the lack inthe model of any practical example, experimental data,plausible sources of energy, or a convincing explanationof the emergence of chirality. There are indeed possibleprocesses to lead to symmetry breaking. For example,counted among the most noteworthy findings of the lastdecade in asymmetric catalysis research must certainlybe Soai and coworkers’ discovery of asymmetric amplifi-cation in the autocatalytic alkylation of pyrimidyl alde-hydes with dialkylzincs [65]. In his exciting articleDonna Blackmond confirmed the Soai reaction. Amplifi-cation of enantiomeric excess (ee) can only result if ameans exists to suppress the catalytic action of the“wrong” hand of the catalyst. Experimental studies of theSoai reaction reveal that statistical formation of dimercatalyst species coupled with lower activity of the het-erochiral dimer is sufficient to rationalize the evolutionof high ee from a tiny initial imbalance. This generalmechanism could be effective in a world of simple or-ganic molecules such as those likely to have beenpresent in the prebiotic world. Blackmond might agreethat the Soai system represents a “triumph of reduction-ism” in helping us to understand the chemical origin oflife in molecular terms [66]. In particular, Blackmondand her team notably highlight that the Soai reactionhas inspired a wealth of studies – experimental, theoret-ical, and computational – aimed at understanding theprocesses of symmetry breaking and asymmetric amplifi-cation. They nevertheless notice that this most prominentexperimental example of asymmetric autocatalysis exhibitssignificant constraints on substrates and reaction condi-tions. Finally, the search for further examples, in particularfor reactions exhibiting chemistry of greater prebioticrelevance, continues apace but has proved inconclusiveto date [67]. There are problems too regarding the

evolutionary path, information on the mechanism ofreplication (either inaccuracy or absence of replication),heredity, stability of the systems, existence of plausiblesites on early Earth favoring the model emergence, reli-ability of the synthesis of primordial ingredients, in-stability of the molecules which increases with chainelongation, catastrophically low probability of meaning-ful sequences etc.Regarding the metabolism-first scenarios there are

specific issues well emphasized by Abel and Trevors [64]:

� Only physicodynamic bias reduces through self-ordering tendencies the vast sequence spaces neededfor prebiotic molecular evolution. Even if vast se-quence spaces had been available in a theoreticalprimordial soup, no known mechanism exists for theprebiotic selection of prescriptive sequences;

� In prebiotic molecular-evolution environment, nodifferential survival or differential reproduction ex-ists yet. Natural selection in the context of Darwin-ian evolution does not exist yet in a prebioticenvironment;

� The wheel would have to be reinvented with eachnew generation. Homeostatic metabolism isstatistically prohibitive enough as it is as a one-timeevent;

� No metabolism-first model can be sustained withoutrapid incorporation of any minimal successes into arecorded, integrated, heritable, cybernetic scheme.

Moreover, a theoretical argument plays against theplausibility of metabolism-first scenarios. The term “evo-lution” is a generic term that can be understood as thechange over time of dynamical systems. Thermodynam-ics drive change (i.e., evolution) in any physical systems,including the diversity of life on earth. There are actuallydifferent levels of evolution, from simple to more com-plex, taking into account the level of interaction betweensystems and populations of systems (i.e., agents) andtheir local environments. It is possible to describe fourfundamental levels of evolution with the most basic levelcorresponding to the monotonic degradation of isolatedsystems, which excludes any interaction with the exter-nal environment. When systems interact with the localenvironment through inputs and outputs they are ableto oppose degradation through self-organizing dynamics.Self-organization may be identified as the second level ofevolutionary dynamics, because the complexity ofself-organization confers much greater persistence onthe system. Self-organizing systems tend to grow as theyreceive environmental inputs, as long as the inputs out-weigh the outputs, but growth ultimately destabilizessystems as internal dynamics become uncoordinated.This results in large systems splitting into several smaller

Tessera Biology Direct (2018) 13:18 Page 11 of 18

systems. Multiplication of self-organizing systems maybe identified as the third level of evolution because itleads to populations of systems, which further extendspersistence of the system’s dynamical cascade. Only thefourth fundamental level, i.e., Darwinian evolution, isable to promote the emergence of genetically and eco-logically distinct lineages on which natural selection mayoperate [68]. Metabolism-first systems may reach thethird level of evolution at best but not the fourth be-cause heredity is lacking. Systems reaching the thirdlevel of evolution (i.e., able to multiply themselves) mayimprove their lifetime but still fully depend on their en-vironment for the evolution over time of their compos-ition and complexity. Without a mechanism of aphenotypic inheritance, which is independent from vari-ation in the local environment, metabolism-first systemscannot sustain any newly acquired composition or com-plexity. Szathmáry and co-workers notice that stablehereditary propagation is not possible without effectiveautocatalysis in contrast to reflexively autocatalytic net-works, complemented by rare uncatalyzed reactions andcompartmentation. The latter networks, resting on thecreation and breakage of chemical bonds, can generatenovel (‘mutant’) autocatalytic loops from a given set ofenvironmentally available compounds. Real chemical re-actions, which make or break covalent bonds, ratherthan mere incorporation of components, are necessaryfor open-ended evolvability. They conclude that themost exciting form of evolvability is indefinite,open-ended on going evolution, possibly leading to anincrease in complexity, at least in certain lineages. Theythink the following conditions are required: (1) a richchemical combinatorics, (2) digital inheritance based ontemplate replication, (3) an environment made morecomplex by evolution itself, and (4) the fact that we can-not pre-state in general the possible preadaptations.How all this could have emerged in early evolution is anut hard to crack [28]. In the lipid vesicles in alkalinehydrothermal field model the only pre-Darwinian pro-cesses at work were the formation of lipid vesicles withmultiplication abilities and the selection of the most vi-able. Without the emergence of genetically and ecologic-ally diverse lineages on which natural election mightoperate the evolution over time of the composition andcomplexity of the vesicles was not possible. Only afterDarwinian evolution would have emerged by chance inone step with the occurrence of specific arrangements ofthe amphiphiles among a huge number of combinationsin the inner part of the bilayer membrane and thus lead-ing to the emergence of hypercycles. The emergence ofDarwinian evolution was impossible as long as multipli-cation and heritability were lacking. Only Darwinianevolution could have led to an evolution in complexityover time.

The replicator-first scenario RNA/Pre-RNA worldseems to solve the previous issues, as multiplication andheritability are present in its more advanced state. How-ever, it encounters its own insurmountable difficultiesand remains a “prebiotic chemist’s nightmare” even ifsome of these difficulties seem to have been solved [30,31, 33, 34]. In addition, all the efforts which have fo-cused on making RNA under a gene first model for theorigin of life curiously have come as close as any to pro-viding a working example of a cycle reminiscent ofmodels which put metabolism first. If the RNA andPre-RNA world were rather metabolism-first models attheir very beginning they would have encountered thesame problems. As a combined metabolism-replicatorscenario the DKS model would have been able to evolvethrough Darwinian evolution since the beginning ac-cording to the authors. The replicator was a specificmolecule of the autocatalytic metabolic system but notthe system per se. It is questionable to claim there was atrue reproduction, as the system did not multiply. Inaddition Pross and Pascal say nothing about heredity.The two other combined metabolism-replicator sce-narios have also their own problems (see above andTable 1). The last scenario has the fewest issues. It onlyrequired small carbon-based molecules, i.e., withoutneeding complex molecules like proteins and/or nucleo-tides at the beginning. It would also have taken the op-portunity of a combinatorial of all the possible localmolecular arrangements of a heterogeneous lipid mem-brane composed of a mixture of amphiphiles. Such acombinatorial of the possible local molecular arrange-ments would have allowed the emergence of a hyper-cycle based on mutual catalysis and been a source ofvariation. The lipid vesicles of the model are openfar-from-equilibrium systems generated by a continu-ously supplied flow of matter and energy from the alka-line hydrothermal fields, may multiply, and havehereditary properties. The model would have allowedDarwinian evolution to start operating from the begin-ning. Its catalyzing properties would have solved theissue of chirality. However, its evolutionary path of thesystems would have been handicapped if the poly-merization was tedious. The main issue of the model isthe present lack of experiments supporting the hypoth-esis that a hypercycle by mutual catalysis might emergein lipid vesicles with a heterogeneous membrane in theconditions of a relatively high temperature (90 °C) andhigh pressures in alkaline hydrothermal vents in saltyseawater. This is why a research program has been re-cently proposed to implement the different steps tosupport this new direction of research in the field.In their scenario the authors appeal to either

pre-Darwinian or Darwinian evolution to support anevolution over time. Pre-Darwinian scenarios do not

Tessera Biology Direct (2018) 13:18 Page 12 of 18

seem the most plausible. They may reach the third levelof evolution (i.e., able to multiply themselves in additionto involving self-organizing dynamics) but not the fourthlevel that requires the emergence of genetically and eco-logically distinct lineages on which selection may oper-ate. Among the scenarios that appeals to Darwinianevolution from the beginning, one seems to show thebest profile although it has its own shortcomings.

ConclusionPre-Darwinian evolution is here defined as an evolutionarycontinuity between chemistry and Darwinian evolution. Itis meant to describe the initial chemical steps before theemergence of Darwinian evolution. Some of thebetter-known open far-from-equilibrium system-based sce-narios are reviewed. No metabolism-first scenario is con-vincing enough to make such a pre-Darwinian evolution isplausible. Alternatives to metabolism-first scenarios are ei-ther replicator-first or combined metabolism-replicator sce-narios. The only so-called replicator-first scenarios, theRNA and the Pre-RNA world scenarios, remain a “prebioticchemist’s nightmare”. Other approaches are combinedmetabolism-replicator scenarios implementing eitherpre-Darwinian evolution or Darwinian evolution. Theynevertheless have their own issues. Among the reviewedmodels, one shows the least, although it has shortcomings.In particular, its main hypothesis needs to be validated byexperimental data. Darwinian evolution is at work from thevery beginning in this scenario. From this critical review itis inferred that the concept of “pre-Darwinian evolution”appears questionable, in particular because it is unlikely ifnot impossible that any evolution in complexity over timemay work without multiplication and heritability allowingthe emergence of genetically and ecologically distinct line-ages on which natural selection may operate. Only Darwin-ian evolution could have led to such an evolution. Thus,Pre-Darwinian evolution is not plausible according to theauthor. Surely, the answer to the question posed in the titleis a prerequisite to the understanding of the origin of Dar-winian evolution.

Reviewers’ commentsReview #1: P. Lopez-Garcia, Centre National de laRecherche Scientifique, FranceThe manuscript has some originality and comments onimportant aspects on the transition from chemistry to bio-logy. I find the style a bit complex and not very clear evenif the ideas are interesting (though not necessarily novel). Itdeserves publication, the literary style could be improved.

I thank the reviewer for her comments, particularlywhen she finds that the manuscript has someoriginality. Regarding the literary style I tried toimprove it.

In this manuscript, M. Tessera critically examines vari-ous models on the origin of life and more specifically inter-rogates about whether these models involve pre-Darwinian(chemical) evolution prior to the Darwinian evolutioncharacteristic of open far-from-equilibrium systems thatare considered alive. The subset of chosen models areclassed in three categories: metabolism-first, replicator-firstand coupled metabolism-replicator models. Many of thecriticisms and concerns highlighted by Tessera have beenalready raised by previous authors; he revisits them in thecontext of a transition from chemical to Darwinian (bio-logical) evolution. Overall, the ideas summarized in thiscritical review are stimulating for research on thechemistry-biology transition. I have, however, some com-ments: - My major comment is that definitions of Darwin-ian and pre-Darwinian evolution are somewhat fuzzy andthis affects whether a system is considered to evolve byDarwinian or pre-Darwinian evolution. The definition ofDarwinian evolution is more obvious, this corresponds toencoded (genetically inherited) variation upon whichnatural selection acts. Pre-Darwinian evolution in-volves chemical evolution. But does a system wheregenotype and phenotype are not (yet) coupled (i.e.,containing a replicator plus not-encoded components)evolve via pre-Darwinian or Darwinian evolution? Forinstance, Tessera claims that lipid vesicles producedin the surroundings of alkaline vents and acting aschemical reactors that eventually include replicatorsdisplay Darwinian evolution since the beginning.However, these initial lipid-vesicle reactors are notencoded (even if there are replicators inside), so inprinciple, there should equally represent apre-Darwinian to Darwinian evolution transition as inother models. - I personally find amphiphile-vesiclemodels involving co-evolving metabolism and replica-tor systems as the most practically plausible for theorigin of life on Earth. However, even if these systemswould show Darwinian evolution from the beginning(which I questioned above), does this mere fact (Dar-winian evolution from the beginning) qualify them asmore likely than models where a pre-Darwinian/Dar-winian evolution transition is required? If so, why?

As I specify it in the “Background” of the manuscript Iprefer to use the expression “pre-Darwinian evolution”instead of “prebiotic evolution” because the concept oflife is very much debatable according to me,eventually questionable [8–12] while the mechanismof Darwinian evolution can be well defined. Thus Ichoose not to use words like “alive”, “life”, “livingorganisms”, “biotic”, “prebiotic” etc. except when theresearchers I cite use them. I would have preferred touse the term “level-4 evolution” instead of “Darwinianevolution” in accordance with the claim that there are

Tessera Biology Direct (2018) 13:18 Page 13 of 18

four fundamental levels of evolution [67].Unfortunately, the scientific community to refer to itdoes not yet accept this view. I find questionable the“pre-Darwinian evolution” concept. Consequently, Icannot find an accurate definition of it. According tome, “Pre-Darwinian evolution” may only be defined inreference to the definition of “Darwinian evolution”. Iagree with the reviewer that genotype and phenotypeshould be coupled. For instance, genotype isrepresented by the membrane sites in my model andphenotype by the carbon-based molecules catalyzed bythe latter as they may impact the structure and thefunctions of the membrane. Genotype and phenotypeare clearly coupled as they form a hypercycle. Whenvesicles multiplied the daughter vesicles inherited bothgenotype (i.e. membrane sites) and phenotype (i.e.,carbon-based molecules) either directly when thehypercycle was transmitted to the daughter vesicles orindirectly when only one element was transmitted butable to reconstruct the hypercycle [10]. Thus, separatelineages formed on which natural selection might haveoperated. Once lipid vesicles with multiplicationabilities formed Darwinian evolution would haveemerged in one step with the occurrence of specificarrangements of the amphiphiles among a hugenumber of combinations in the inner part of thebilayer membrane. The only Pre-Darwinian processesat work were the formation of lipid vesicles with multi-plication abilities and the selection of the most viable.Finally I think only Darwinian evolution could haveled to an evolution in complexity over time. To avoidany ambiguity I now clearly answer the question of theplausibility of pre-Darwinian evolution. Of course, Iconfirm in the manuscript that the answer to the ques-tion posed in the title is a prerequisite to the under-standing of the origin of Darwinian evolution.

Why not including Wächtershäuser’s iron-sulfur worldin the comparison? I understand that it has severe prob-lems, notably in the transition to cellularization. None-theless, this is one of the most influential models andnot less problematic than the Russell’s chemical garden.At least, some of its chemical predictions were proved. –Chirality.

Wächtershäuser’s iron-sulfur world model is now ana-lyzed in the manuscript.

The fact that many models do not address the issue ofchirality does not imply that they may not accommodatean explanation for chirality. The absence of proof is notthe proof of absence. I guess that in many of thesemodels, chirality is simply seen as some kind of conse-quence of chance. Once you start incorporating one

particular isomer, the choice was selectively main-tained. - This brings me to my last general comment.The role of chance is ignored in this review. Withinthe realm of biology, Darwinian evolution is not a fullsynonym of biological evolution because in additionto selective processes, there is genetic drift. Whatabout pre-genetic drift? This is not trivial becauseeven if one gets experimental evidence for a particu-lar model, this does not mean that historically lifeoriginated that way. This would only provide an argu-mental basis not to discard a particular model.

For sure chirality has emerged luckily. In my modelchance would have been at work when mutualcatalysis emerged for the first time in lipid vesicleswith heterogeneous membranes. Even if the emergenceof a mutual catalysis was allowed by the structure ofthe vesicle membrane composed of a mixture ofamphiphiles chance would have played its part. Thisoccurred when a specific arrangement of amphiphilesappeared among the huge number of possiblecombinations of arrangements. It was able to catalyzethe synthesis of a specific carbon-based molecule. Thislatter soluble compound had the property of catalyzingthe transformation of the local membrane arrange-ment into a stabilized membrane site. Chance wouldhave operated again when specific small carbon-basedmolecules led to the synthesis of a bigger molecule witha chiral centre carbon atome. Surely, in the othermodels chance may have played its part too to makechirality appear but the researchers should present arational and plausible explanation. I agree that gen-etic drift should have played its part. There is no par-ticular reason why it would not have occurred in mymodel. I do not understand what the reviewer meansby “pre-genetic drift”.

Review #2: A. Poole, Stockholm University, SwedenThis manuscript has a promising title. Unfortunately themanuscript seems instead to be more focused on givinga potted critique of the shortcomings of some of thebetter-known models for the origin of life. The main is-sues the author highlights are around whether themodels adequately address issues like chirality or Dar-winian evolution. A more extensive review of the variousmodels would be helpful; as presented, this review partis a bit too uneven, and needs clearer explanations ofthe proposed models before launching into critique. Thesection that discusses the question posed in the title is toobrief - it is only a few lines on page 13 (lines 1–25), wherethe author presents four ‘levels’ of evolution. This is notreferenced, but it does note that ‘evolution’ is broader than‘Darwinian evolution’. However, the question posed in the

Tessera Biology Direct (2018) 13:18 Page 14 of 18

title is not really discussed in any great depth, and, for me,did not expand the existing discussion around this inter-esting question. The impression I got from reading thearticle was that the author’s answer to the question heposed was, ‘yes, but it’s not important’.

Thank to the reviewer I realized that my view onthe plausibility of pre-Darwinian evolution was notso clear as there was a misunderstanding in the re-viewer’s impression of my opinion about it. Thisessay highlights critical aspects of the plausibility ofpre-Darwinian evolution. It is based on a criticalreview of some better-known open, far-from-equilibrium system-based scenarios supposed to ex-plain processes that took place before Darwinianevolution had emerged and that resulted in the ori-gin of the first systems capable of Darwinian evolu-tion. Each model was evaluated according to theresearchers’ answers to eight crucial questions thatshould be addressed (Table 1). I tried to summarizethe models as well as possible by using theresearchers’ wording as far as possible but, surely,my reports cannot be fully exhaustive and unbiased.I appreciate the reviewer’s citation of ourproposition, G. Hoelzer and I, that there are fourfundamental levels of evolution [67]. In accordancewith our claim I would have preferred to use theterm “level-4 evolution” instead of “Darwinianevolution”. Unfortunately, the scientific communityto refer to it does not yet accept this view. I findthe concept of “pre-Darwinian evolution”questionable. It is unlikely if not impossible thatany evolution in complexity over time could haveworked without multiplication and heritability.Only Darwinian evolution would have led to suchan evolution. By the way the only pre-Darwinianprocesses at work in the model I propose were theformation of lipid vesicles with multiplication abil-ities and the selection of the most viable. Only afterDarwinian evolution would have emerged by chancein one step with the occurrence of specific arrange-ments of the amphiphiles among a huge number ofcombinations in the inner part of the bilayer mem-brane. To avoid any ambiguity I now clearly answerthe question of the plausibility of pre-Darwinianevolution. Of course, I confirm in the manuscriptthat the answer to the question posed in the title isa prerequisite to the understanding of the origin ofDarwinian evolution.

Review #3: D. Lancet, Weizmann Institute of Science, IsraelAs instructed, I am checking whether the original refereecomments have been addressed to satisfactory standards.

My own comments refer to the revised version (R1). Re-viewer 1 Points 1,3, are satisfactorily addressed by theauthor. Point 2 (beginning “In this manuscript, M. Tes-sera critically examines various models on the origin oflife…”) is valid, and has not been fully addressed. In theabstract, the authors present the following conclusion:“From this critical review it is (inferred) that the conceptof ‘pre-Darwinian evolution’ appears questionable, inparticular because it is unlikely if not impossible thatany evolution in complexity over time may work withoutmultiplication and heritability. Only Darwinian evolutioncould have led to such an evolution. Thus,Pre-Darwinian evolution is not plausible according tothe author”. How then can the attribute “Pre-Darwinian”appear for any model in the last column of the table, acolumn entitled “initial evolution”? I read the author’sconclusion above as implying that whatever chemicalprocesses that took place prior to the advent of Darwin-ian evolution cannot be called evolution at all. This isbecause the author applies the same necessary criteria(multiplication and heritability leading to complexifica-tion) to both pre-Darwinian and Darwinian evolution.Thus if the criteria are not fulfilled, we have neither Dar-winian nor pre-Darwinian evolution. But then, to avoidthe confusion on which both reviewer 1 and I agree, thetitle of the paper should be “what chemical processes ledto Darwinian evolution”. Point 4 beginning with “Thefact that many models do not address the issue of chiral-ity does not imply that they may not accommodate anexplanation for chirality”. I fully agree with this com-ment and feel that it has not been adequately addressed.The many models which have the value “not an issue”and “not addressed” in the column entitles “Chiralityissue” of the paper’s Table, attest to the idea that homo-chirality should not serve as a yardstick for judging evo-lution of any kind. This is, in fact, supported by theauthor’s inference based on a paper of ours (Ref 52):“The C-GARD model would highlight the possibility thatchiral selection is a result of, rather than a prerequisitefor early life-like processes and thus would not havebeen an issue”. I strongly suggest that the an issue criter-ion for evolution be eliminated altogether. Reviewer 2 Iagree with this reviewer’s comment: “A more extensivereview of the various models would be helpful; as pre-sented, this review part is a bit too uneven, and needsclearer explanations of the proposed models beforelaunching into critique”. I wish to support this commentby addressing the example of my own model (GARD),pointing to necessary corrections. The author shouldplease re-check that the description of other modelsmight not have been similarly afflicted. Here are the pointsthat need to be corrected in the description of the GARDmodel: 1) The author’s statement: “The model is alsobased on the view that non-equilibrium self-organizing

Tessera Biology Direct (2018) 13:18 Page 15 of 18

systems have dynamic properties that exist in a state closeto chaotic behavior allowing the emergence of autocata-lytic cycles…” is not correct. In the GARD model (as inother similar models) the emergence of mutually catalyticnetworks (not “autocatalytic cycles”, which is a restrictiveterm) is afforded merely by the nature of the molecules,i.e. their capacity to exert catalysis on each other, irre-spective of chaotic behavior. 2) The author says: “…mutu-ally catalytic sets as an alternative to alphabet-basedinheritance”. The GARD model has a form ofalphabet-based inheritance. The crucial difference be-tween GARD and a templating biopolymer model is thatthe former accumulates and reproduces compositional in-formation (counts of chemical “alphabet” letters), whilethe latter encompasses sequence information (order ofchemical “alphabet” letters). 3) The current text proclaims:“A basic feature in GARD is that non-covalent,micelle-like molecular assemblies capable of growinghomeostatically (i.e., buffered enough as to maintain sta-bility) according to the assembly’s constitution store com-positional information can be propagated after occasionalfission (i.e., assembly splitting)”. This confusing sentenceshould better read: “A basic feature in GARD is thatnon-covalent, micelle-like molecular assemblies are cap-able of growing homeostatically, i.e. catalytically maintain-ing the assembly’s composition as it grows (dynamicbuffering). The maintained compositional information canbe propagated to progeny assemblies upon occasional fis-sion”. 4) The author states, based on ref. 47, that “Regard-ing the evolvability of the system it has recently beendemonstrated that replication of compositional informa-tion (in GARD) is so inaccurate that fitter compositionalgenomes cannot be maintained by selection”. This criti-cism is hotly disputed, as exemplified in one of our papers[PMID: 22662913], and it would be fair to make this state-ment less unqualified and quote the alternative view. 5)The author says (based on Refs 49–51): “Moreover, thereis no reason why ‘compositional information’ should havebeen transferred to bilayer membrane lipid vesicles whenthese would have taken over the ‘micelle-like molecularassemblies’.” This statement is based on a misunderstand-ing, and should be omitted. The GARD model does notinvoke a capacity of a small micelle to confer its compos-ition onto a much larger vesicular assembly when fusingwith it. Rather, it invokes the possibility that homeostaticgrowth via single molecule accretion could gradually leadto increasingly larger assemblies having a similar compos-ition to that of the original micelle. Returning to Reviewer 2comments: A sweeping negativity of this reviewer is mani-fested in the statement “…the question posed in the title isnot really discussed in any great depth, and, for me, did notexpand the existing discussion around this interesting ques-tion”. This, in my opinion, is overstated. I feel there is valuein this review, warranting publication in Biology Direct.

I thank the reviewer for his helpful comments.Regarding the GARD model I modified the sentencewhen I agreed with the reviewer that it was incorrect(e.g., the basic feature in GARD). When a question isdisputed I still mention it with the arguments, i.e., thecriticisms, and the counter-arguments, i.e., thereviewer’s reply (e.g., the question of evolvability and ofthe transfer of the compositional information). Withregards to the chiral question I cannot agree thathomochirality should not serve as a yardstick forjudging evolution in complexity over time. While it isnot addressed in most models it does not mean thatthe chiral question is not crucial. As the reviewernoticed, I present the results of his simulations tendingto support the reviewer’s view that chiral selection is aresult of, rather than a prerequisite for early life-likeprocesses and thus would not have been an issue.However, I also observe that the authors’ assertion sup-porting the relevance of these simulations seems un-realistic, i.e., that, in the prebiotic environment, therewould have been sufficiently complex molecularstructures allowing them to assume that all moleculeswere chiral. Finally, I believe that the new title of themanuscript the reviewer proposes does not suit theaim of the manuscript, i.e., to support the view thatPre-Darwinian evolution is not plausible. According tothis view, the likely processes that would have takenplace and allowed evolution in complexity over timebefore Darwinian evolution emerged are not satisfac-tory. In my model, the only pre-Darwinian processes atwork are the formation of lipid vesicles with multipli-cation abilities and the selection of the most viable.These processes are not sufficient to allow an evolutionin complexity over time.

Review #4: T. Dandekar, Department of Bioinformatics,University of Wuerzburg, GermanyThe paper makes a case that purely pre-Darwinian evo-lution does not exist, looking at different examples ofchemical evolution and the different major aspects theycover. This table presents some new comparative results.The conclusion is that even in chemical evolutionmodels there is some hereditary element involved, so ac-cording to the author, some Darwinian evolution. 2. Ithink it is worthwhile to stress this point and the com-parison with the chemical models stresses this point ap-propriately and that justifies a publication 3. What couldbe added would be some more implications, for instanceif any type of evolution always needs some element ofheredity, does this (more information is passed to thenext generation) enhance then always the speed of evo-lution? 3b. Does then determining whether there is evo-lution always boil down to identify some heritable

Tessera Biology Direct (2018) 13:18 Page 16 of 18

element (or information storage)? 4. Of course, choosingthe suitable definition you can always be right as an au-thor, but probably by such a definition of evolution youlargely ignore the more general and bigger class ofself-organizing processes, right? 4b. So my main worry isthat by defining terms and things as you do, you may getrid of pre-Darwinian evolution (as you claim that thereis always heredity necessary), however, you then becomeblind for the large, interesting and important class of selforganizing phenomena in physics and chemistry whohappen without any gene storage or any other such dir-ect storage.

I agree with the reviewer that it would be lessconstraining for the purposes of the search of the originof life if pre-Darwinian evolution was plausible. Itwould open the search to include self organizing phe-nomena in physics and chemistry which happen with-out any gene storage or any other such direct storage. Idon’t think it is a question of definition, like, for ex-ample the definition of life. Darwinian evolution is amechanism. It works as it allows evolution in complex-ity over time, It is based, amongst other things, on nat-ural selection on distinct lineages. Withoutinformation storage and the possibility of transfer tothe progeny, no lineage may emerge. Thus, natural se-lection cannot operate and allow evolution in com-plexity over time.

AbbreviationsDKS: Dynamic kinetic stability; DNA: Deoxyribonucleic acid; D-RNA: Dextro-ribonucleic acid; ee: Enantiomeric excess; GARD: Graded autocatalysisreplication domain; LUCA: Last universal common ancestor; MAR-complex: Metal-Amino-Ribo complexes; RNA: Ribonucleic acid;SMSB: Spontaneous mirror symmetry breaking

AcknowledgmentsI thank Peter Walde for his comments, which allow me to improve the initialdrafts. I thank Kepa Ruiz-Mirazo for the interesting discussions we have hadon the topic.

FundingI confirm that I am an independent researcher and that I have not receivedany funds.

Ethics approval and consent to participateNot applicable.

Consent for publicationNot applicable.

Competing interestsThe author declares that he has no competing interests.

Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.

Received: 29 December 2017 Accepted: 15 June 2018

References1. Follmann H, Brownson C. Darwin’s warm little pond revisited: from

molecules to the origin of life. Naturwissenschaften. 2009;96:1265–92.2. Cronin L, Walker SI. Beyond prebiotic chemistry What dynamic network

properties allow the emergence of life? Science. 2016;352:1174–5.3. Spitzer J, Pielak GJ, Poolman B. Emergence of life: Physical chemistry

changes the paradigm. Biol Direct. 2015;10:1–15.4. Krishnamurthy R. On the Emergence of RNA. Isr J Chem. 2015;55:837–50.5. Sahai N, Kaddour H, Dalai P. The transition from geochemistry to

biogeochemistry. Elements. 2016;12:389–94.6. Mayer C, Schreiber U, Dávila MJ. Selection of prebiotic molecules in

amphiphilic environments. Life. 2017;7:3–13.7. Szostak JW. Attempts to define life do not help to understand the origin of

life. J Biomol Struct Dyn. 2012;29:599–600.8. Tessera M. Life began when evolution began: A lipidic vesicle-based

scenario. Orig Life Evol Biosph. 2009;39:559–64. https://doi.org/10.1007/s11084-009-9175-4.

9. Tessera M. Origin of evolution versus origin of life: a shift of paradigm. Int JMol Sci. 2011;12:445–58.

10. Tessera M. Research program for a search of the origin of Darwinianevolution. Orig Life Evol Biosph. 2017;47:57–68.

11. Tessera M. Is a n+1 definition of life useful? J Biomol Struct Dyn.2012;29:635–6.

12. Gayon J. Defining life: Synthesis and conclusions. Orig Life Evol Biosph.2010;40:231–44.

13. Garay J. Active centrum hypothesis: The origin of chiral homogeneity andthe RNA-world. Biosystems. 2011;103:1–12.

14. Wächtershäuser G. From volcanic origins of chemoautotrophic life toBacteria, Archaea and Eukarya. Philos Trans R Soc Lond Ser B Biol Sci. 2006;36:11787–806.

15. Martin W, Russell MJ. On the origin of cells: an hypothesis for theevolutionary transitions from abiotic geochemistry to chemoautotrophicprokaryotes, and from prokaryotes to nucleated cells. Philos Trans R Soc B.2003;358:59–85.

16. Vasas V, Fernando C, Santos M, Kauffman S, Szathmáry E. Evolution beforegenes. Biol Direct. 2012;7:1–14.

17. Szathmáry E. Toward major evolutionary transitions theory 2.0. PNAS. 2015;112:10104–11.

18. Ruiz-Mirazo K, Briones, de la Escosura A. Prebiotic systems chemistry: Newperspectives for the origins of life. Chem Rev. 2014;114:285–366.

19. Vasas V, Fernando C, Szilágyi A, Zachár I, Santos M, Szathmáry E. Primordialevolvability: impasses and challenges. Theor Biol. 2015;381:29–38.

20. Maynard-Smith J. The Problems of Biology. Oxford: Oxford UniversityPress; 1986.

21. Fernando C, Rowe J. The origin of autonomous agents by natural selection.Biosystems. 2008;91:355–37.

22. Russell MJ, Hall AJ, Turner D. In vitro growth of iron sulphide chimneys:possible culture chambers for origin-of-life experiments. Terra Nova.1989;1:238–41.

23. Russell MJ, Hall AJ, Martin W. Serpentinization as a source of energy at theorigin of life. Geobiology. 2010;8:355–71.

24. Lane N, Martin WF. The origin of membrane bioenergetics. Cell. 2012;151:1406–16.

25. Herschy B, Whicher A, Camprubi E, Watson C, Dartnell L, Ward J, Evans JRG,Lane N. An origin-of-life reactor to simulate alkaline hydrothermal vents. JMol Evol. 2014;79:213–27.

26. Jackson JB. Natural pH gradients in hydrothermal alkali vents were unlikelyto have played a role in the origin of life. J Mol Evol. 2016;83:1–11.

27. Jackson JB. The “origin-of-life reactor” and reduction of CO2 by H2 ininorganic precipitates. J Mol Evol. 2017;85:1–7.

28. Russell MJ, et al. The drive to life on wet and icy worlds. Astrobiology.2014;14:308–43.

29. Kun A, Szilágyi A, Könnyú B, Boza G, Zachar I, Szathmáry E. The dynamics ofthe RNA world: insights and challenges. Ann N Y Acad Sci. 2015;1341:75–95.

30. Benner SA, Kim HJ, Yang Z. Setting the Stage: The history, chemistry, andgeobiology behind RNA. Cold Spring Harb Perspect Biol. 2012;4:a003541.

31. Powner MW, Gerland B, Sutherland JD. Synthesis of activated pyrimidineribonucleotides in prebiotically plausible conditions. Nature. 2009;459:239–42.

Tessera Biology Direct (2018) 13:18 Page 17 of 18

32. Monnard PA, Walde P. Current ideas about prebiologicalcompartmentalization. Life. 2015;5:1239–63.

33. Patel BH, Percivalle C, Ritson DJ, Duffy CD, Sutherland JD. Common originsof RNA, protein and lipid precursors in a cyanosulfidic protometabolism. NatChem. 2015;7:301–7.

34. Stairs S, Nikmal A, Bučar DK, Zheng SL, Szostak JW, Powner MW. Divergentprebiotic synthesis of pyrimidine and 8-oxo-purine ribonucleotides. NatCommun. 2017;8:15270–81.

35. Ruiz-Mirazo K, Briones C, de la Escosura A. Chemical roots of biologicalevolution: the origins of life as a process of development of autonomousfunctional systems. Open Biol. 2017;7:170050.

36. Wächtershäuser G. Groundworks for an evolutionary biochemistry: theiron–sulphur world. Prog Biophys Mol Biol. 1992;58:85–201.

37. Stich M, Ribó JM, Blackmond DG, Hochberg D. Necessary conditions for theemergence of homochirality via autocatalytic self-replication. J Chem Phys.2016;145:074111.

38. Joshi PC, Aldersley MF, Ferris JP. Homochiral selectivity in RNA synthesis:Montmorillonite-ed quaternary reactions of d, l-purine with d, l- pyrimidinenucleotides. Orig Life Evol Biosph. 2011;41:213–36.

39. Sczepanski JT, Joyce GF. A cross-chiral RNA polymerase ribozyme. Nature.2014;515:440–2.

40. Bregestovski PD. “RNA World”, a highly improbable scenario of the originand early evolution of life on earth. J Evol Biochem Physiol. 2015;51:72–84.

41. Maurel MC, Leclerc F. From foundation stones to life: Concepts and results.Elements. 2016;12:407–12.

42. Pascal R, Pross A. Stability and its manifestation in the chemical andbiological worlds. Chem Commun. 2015;51:16160.

43. Pascal R. Kinetic barriers and the self-organization of life. Isr J Chem. 2015;55:1–11. https://doi.org/10.1002/ijch.201400193.

44. Pross A, Pascal R. How and why kinetics, thermodynamics, and chemistryinduce the logic of biological evolution. Beilstein J Org Chem. 2017;13:665–74.

45. Blankenship RE. Early evolution of photosynthesis. Plant Physiol.2010;154:434–8.

46. Segré D, Ben-Eli D, Lancet D. Compositional genomes: Prebioticinformation transfer in mutually catalytic noncovalent assemblies. PNASUSA. 2000;97:4112–7.

47. Markovitch O, Lancet D. Excess Mutual Catalysis Is Required for EffectiveEvolvability. Artif Life. 2012;18:243–66.

48. Segré D, Ben-Eli D, Deamer DW, Lancet D. The lipid world. Orig Life EvolBiosph. 2001;31:119–45.

49. Vasas V, Szathmary E, Santos M. Lack of evolvability in self-sustainingautocatalytic networks constraints metabolism-first scenarios for the originof life. PNAS. 2010;107:1470–5.

50. Söderman O, Herrington KL, Kaler EW, Miller DD. Transition from micelles tovesicles in aqueous mixtures of anionic and cationic surfactants. Langmuir.1997;13:5531–8.

51. Hanczyc MM, Fujikawa SM, Szostak JW. Experimental models of primitivecellular compartments: Encapsulation, growth, and division. Science. 2003;302:618–22.

52. Zhai L, Zhang J, Shi Q, Chen W, Zhao M. Transition from micelle to vesiclein aqueous mixtures of anionic/zwitterionic surfactants studied byfluorescence, conductivity, and turbidity methods. J Colloid Interface Sci.2005;284:698–703.

53. Kafri R, Markovitch O, Lancet D. Spontaneous chiral symmetry breaking inearly molecular networks. Biol Direct. 2010;5:38.

54. Budin I, Bruckner RJ, Szostak JW. Formation of protocell-like vesicles in athermal diffusion column. J Am Chem Soc. 2009;131:9628–9.

55. Proskurowski G, Lilley MD, Seewald JS, Fruh-Green GL, Olson EJ, Lupton JE,Sylva SP, Kelley DS. Abiogenic hydrocarbon production at lost cityhydrothermal field. Science. 2008;319:604–7.

56. Chen IA, Walde P. From self-assembled vesicles to protocells. Cold SpringHarb Perspect Biol. 2010;2:a002170.

57. Hanczyc MM, Monnard PA. The origin of life and the potential role of soaps.Lipid Technol. 2016;28:88–92.

58. Eigen M, Schuster P. The hypercycle. A principle of natural self-organizationPart A: Emergence of the hypercycle. Naturwissenschaften. 1977;64:541–65.

59. Szostak JW. An optimal degree of physical and chemical heterogeneity forthe origin of life? Philos Trans R Soc B. 2011;366:2894–901.

60. Mitchinson A, Finkelstein J. Small-molecule catalysis. Nature. 2008;455:303–49.

61. Barbas CF. Organocatalysis Lost: Modern chemistry, ancient chemistry, andan unseen biosynthetic apparatus. Angew Chem Int Ed. 2007;46:2–8.

62. Woolf NJ. A Hypothesis About the Origin of Biology. Orig Life Evol Biosph.2015;45:257–74.

63. Woese CR, Fox GE. Phylogenetic structure of the prokaryotic domain: Theprimary kingdoms. PNAS. 1977;74:5088.

64. Abel DL, Trevors JT. Self-organization vs. self-ordering events in life-originmodels. Phys Life Rev. 2006;3:211–28.

65. Soai K, Shibata T, Morioka H, Choji K. Asymmetric autocatalysis andamplification of enantiomeric excess of a chiral molecule. Nature. 1995;378:767–8.

66. Blackmond DG. Asymmetric autocatalysis and its implications for the originof homochirality. PNAS. 2004;101:5732–6.

67. Hein JE, Gherase D, Blackmond DG. Chemical and Physical Models for theEmergence of Biological Homochirality. Top Curr Chem. 2013;333:83–108.

68. Tessera M, Hoelzer GA. On the thermodynamics of multilevel evolution.Biosystems. 2013;113:140–3.

Tessera Biology Direct (2018) 13:18 Page 18 of 18