After the primary hydrogen atmosphere of Earth had escaped into space, the so-called secondary atmosphere built up with volcanic gases; this atmosphere was, most likely, dominated by CO₂, with smaller amounts of N₂, CO, and H₂, similar to that on modern Mars and Venus, where CO₂ still makes up 95% of the atmosphere 2593949596979899100. As the Earth's surface gradually cooled, water vapour started to condense into the first oceans. The atmospheric pressure at the surface of primordial Earth has been estimated to reach several hundred bars; therefore ocean formation could have started when the surface was still very hot 9497100. Zircon data indicate the presence of the first continent(s) by 4.2 Ga 188. Hydrothermal activity of some type would have been established promptly, driven by thermal convection. Most likely, the initial convection systems did not form a continuous chain of mid-ocean ridges as they do now but a pattern of "hot spots" similar to modern volcanic island arcs 9395189. The activity of these hydrothermal/volcanic systems was accompanied by surges of hot hydrothermal fluids to the surface. Owing to the initial atmospheric pressure of ≥ 100 bar (see 94100, and cf. with the pressure of 95 bar at the surface of modern Venus), very hot hydrothermal fluids enriched by dissolved metals could discharge directly to the surface of the first continents. This situation differed fundamentally from the modern one, since today such hot fluids can reach the continental surface only as steam (at the points of volcanic or geyser activity), losing their metal content on the way. Taking into account the slow precipitation of ZnS under high-pressure conditions 183 and the abundance of Zn in the Earth's crust 90, one can expect a major delivery of Zn-enriched hot fluids to the surface of the first continent(s). Zinc could even have been the dominant transition metal in the continental hydrothermal fluids when the atmospheric pressure changed in the range from ca. 100 bar to ca. 10 bar (this would correspond to the temperature of hot fluids reaching 300°C – 200°C, respectively, cf. with the above described situation at modern hydrothermal vents). Hence, it is possible that the large areas of first continent(s) could have been covered by porous ZnS precipitates of hydrothermal origin. These ZnS edifices should have been accessible to solar UV radiation at the surface of continents and in shallow waters surrounding them (see Fig. 1). Hereafter, the term "sub-aerial" is used to denote illuminated settings where the UV-rich solar light could have served as an energy source for primordial syntheses (see also 25114177).

In the absence of an ozone layer, the UV component of solar radiation would have driven the reduction of CO₂ at the ZnS-covered surfaces. Since these surfaces were formed by precipitated ZnS particles (see Fig. 1), similar to those used in the aforementioned experiments with the photosynthetically active "colloidal" ZnS crystals 144145, the reduction of CO₂ may have proceeded with a high quantum yield under the high atmospheric CO₂ pressure. Zinc sulfide is a very powerful photocatalyst that, besides reducing CO₂, is capable of driving diverse reactions of carbon- and nitrogen-containing substrates 144150152190191192. Such substrates could build up in the atmosphere, be generated by photochemical reactions in the water phase 24106109, accompany volcanic extrusions 193 and hydrothermal fluids 123, or be brought by meteorites 194. They could have then participated in further photocatalyzed transformations at the ZnS surfaces.

Electrically charged products of photosynthesis, e.g. negatively charged carbonic acids, would have been attracted by the complementary charges at the ZnS surfaces of mineral compartments. These molecules could have interacted with each other at the catalytic surfaces of continuously operating porous ZnS photoreactors, yielding even more intricate carbon- and nitrogen-containing molecules. The more complex molecules would, generally, have absorbed more light and been more vulnerable to UV quanta. Still, in certain cases, the increase in chemical complexity may have been accompanied by an increase in photostability. Indeed, the destruction of a chemical compound by a light quantum starts with the "trapping" of its energy by a particular chemical bond, followed by an increase in the energy of this bond and its eventual dissociation 195. However, if the absorbed energy is spread over many bonds, then the probability of bond cleavage drops dramatically. Such a spreading of excitation energy occurs in systems that contain conjugated bonds (so-called π-systems with alternating single and double bonds); the spreading is facilitated by a ring-like (aromatic) molecular structure. All nucleobases belong to such ring-like, conjugated systems 159; the lifetime of their excited states (~100 fs 164) is extremely short even for π-systems; this short life time would additionally have decreased the probability of photodestruction. As discussed above, the UV-resistance of π-systems could increase further upon their stacking together and/or adsorption to radiation-absorbing minerals.

Hence, at illuminated ZnS surfaces, the UV-resistant, ring-like compounds could survive as stacked aggregates of "rings joined to rings" (quoted from 1). A relative enrichment in such stacks, as well as their stabilization by covalent linkages, could be driven by a number of factors, namely, the UV-resistance of polymerized and stacked π-systems 112164, the potential to utilize the energy of UV quanta for photopolymerization 1112196, the ability of Zn²⁺ ions to catalyze the polymerization 197198, and the low dielectric permittivity of the surface-adjoining water layers 199 that may have favoured condensation reactions. In addition, a regular mesh of electric charges at the ZnS surface, by attracting reactants and arranging them appropriately relative to each other, may have made the polymerization thermodynamically more favourable than in bulk water 124. Then, however, the resulting polymers should have stayed confined to the surface 200. In the context of a UV-irradiated environment such confinement could be considered a rescue since an eventual detachment of a primordial polymer from the energy-absorbing ZnS surface would have led to faster photodestruction. At the same time, while prevented from detachment from the surface, the molecules would have been able to diffuse along the surface, to interact with each other, and to form aggregates, which is a pre-condition for the abiogenic emergence of increasingly complex structures.

The scenario outlined above should yield two fundamentally different populations of molecules, namely surface-confined, relatively complex structures capable of efficient discarding excess radiation energy (hereafter referred to as zymes), and continuously photosynthesized simpler organic molecules – the future metabolites – that stored the solar energy in their covalent bonds.

The eventual elongation of zymes would promote their entropy-driven folding at the surface. In addition, an increase in their amount would encourage interactions between them. Both factors may have led to the formation of hydrogen bonds within zymes and/or between different zymes. Clustering of zymes may have been favoured by high pressure 201 and periodic drying events, e.g. in tidal regions 18202, resulting in a kind of natural polymerase chain reaction (PCR)-like process 203. The UV-stability of double RNA strands, owing to hydrogen bonding, is much higher than that of single-stranded RNAs 159164204205206. Therefore those π-systems may have been selected – from the initially larger set of compounds – which could make multiple hydrogen bonds with each other. Ultimately, this selection would have led to a relative enrichment of complementary nucleobases, including those that we know now. It is plausible that the π-systems could have been initially linked in various ways. It would appear that one of these constructs, the one with ribose-phosphate units connecting the stacked nucleobases, attained the ability for self-replication and was, because of this, retained by evolution.

Unlike occasional self-replication events, a systematic, accurate replication is thermodynamically demanding 207 and would require special machinery that could consist of several folded polymers resembling modern transfer RNA (tRNA) or ribosomal RNA (rRNA). It is unlikely that we would ever be able to reconstruct all the steps that led to the formation of first replicators (see but refs. 4452545860656772208209 for tentative scenarios). We can, however, try to make guesses on the selective forces underlying their emergence. At least two factors deserve note:

1) Nucleobases quench the UV quanta by converting the energy into heat. The heating, however, is not harmless. Using the approach of Dancshazy and co-workers 210, it is possible to surmise that a single UV quantum should locally heat a 100-unit RNA-like polymer by tens of degrees Celsius. Even if the absorbed energy can be promptly transmitted to a template, local heating of the template can eventually cause ablation of the molecule 211. In the UV-irradiated environment, both outcomes could lead to polymer deterioration. Overheating, however, can be avoided by channelling part of the energy into work. For example, those zymes that could serve as antennas and use the UV energy, e.g., for connecting nucleotides together (as hypothesized by Skulachev 111) had better survival chances. This selective advantage of "working" polymers over the "idle" ones is general in nature: it is applicable not only to the first ribozymes, but also to the first proteins, as discussed in the next section.

2) A population of RNA-like polymers adsorbed at a ZnS template could gather light and therefore enhance the yield of the photocatalysis. Since the hydrothermal ZnS structures are highly porous 179180181, the pores/compartments that contained efficient replicators and, hence, an increasing number of UV-absorbing zymes could produce more metabolites, which, in turn, could be used to build new replicators, resulting in a kind of positive feedback.

As first pointed out by Horowitz 212, and as our simulations showed (112; see also the section on energetics of the Zn world below), an abiogenic formation of complex molecules (e.g. long polymers) would imply the presence of an overwhelmingly larger amount of simpler molecules of the same type (shorter polymers). The first replicators could therefore utilize the available shorter fragments; with time, they could acquire the ability to attain building blocks by cleaving the (phosphor)ester bonds in other surface-adsorbed zymes. Owing to this development: (i) a coupling mechanism could be established that, with some modifications, has been in use ever since – the polymerization of RNA and DNA is still driven by cleavage of the phosphoester bonds in ATP; and (ii) the genuine fight for survival could begin, since each replicator became a potential prey for others.

Generally, the thermodynamics of primordial syntheses of polymers, and in particular of polypeptides, has been deemed a riddle since the respective condensation reactions, which are accompanied by the release of water molecules, should be unfavourable in water (see e.g. refs. 83213 for further discussion of this point). A closer inspection of modern metabolic chains offers a way out of this conundrum. Although biological syntheses are indeed accompanied by the release of water molecules, they never proceed "for free", but are coupled to thermodynamically favourable, exergonic reactions, e.g. of ATP hydrolysis, which require water. Hence, if (i) there were substrates that could be hydrolyzed and (ii) a thermodynamic coupling between syntheses and hydrolyses could be established – then the primordial syntheses could proceed without violating the laws of thermodynamics.

In this framework, the emergence of new synthetic pathways could proceed in two steps, as follows. First, a (ribo)zyme capable of cleaving a new type of chemical bond – in a thermodynamically favourable reaction – would have emerged. Then, the ability to make this particular type of chemical bond could develop as a reversal of the new catalytic pathway, provided that coupling with some exergonic reaction (e.g. a phosphate group transfer or a hydrolysis of a phosphoester bond) could be established.

In the ZnS settings, photosynthetically produced polycarbonic compounds, inorganic polyphosphates, RNA-like oligomers, and diverse low-molecular-weight phosphorylated derivates could serve as cleavage substrates for the first synthases. The reducing equivalents that would be needed for some synthetic reactions were continuously generated at the illuminated ZnS surfaces; nucleotide-containing redox cofactors, such as NADH, NADPH, FAD, and FMN 22214215216, could have emerged as mediators that picked up the photoexcited electrons from photoactive ZnS surfaces and delivered them to the respective ribozymes.

After the emergence of the first replicators capable of connecting amino acids by peptide bonds, the synthesis of first, random polypeptide sequences at nucleotide and/or ZnS templates could begin. Without considering the elusive chemical details, it seems fitting, as in the previous section, to focus on selective factors that could drive the emergence of the first polypeptides. They may have been recruited to perform functions that did not require a particular amino acid sequence but only the ability of a polypeptide to bind to a replicator, e.g. via the protein backbone groups 209. Such a binding, for example, could protect the backbone of replicators from hydrolysis or cleavage. In addition, the bound polypeptides could absorb heat into which the energy of UV quanta was converted, therefore protecting the replicators from thermal damage. It is noteworthy that, if a protecting polypeptide could eventually discard excess energy by funnelling it into a chemical reaction, then the probability of denaturation would have become lower. Those replicators that could synthesize proteins capable of supporting catalysis by (ribo)zymes 67 or catalyzing useful chemical reactions themselves may have got an advantage. Eventually such a selection could establish a correspondence between the nucleotide and amino acid sequences, see 464753586773217218 for tentative scenarios on the evolution of genetic code and protein synthesis.

The ZnS precipitates would have attenuated the UV component of solar light, thus providing shade for the inhabitants of the lower-lying compartments. Hence a stratified system could be established with the illuminated upper layers accounting for maximal photosynthesis, and lower, less productive, but more protected layers providing shelter for the first replicating entities in their pores. The porosity of the ZnS precipitates 179180 would have enabled metabolite transport between the layers. Moreover, the sponge-like inner structure could eventually enable variable hydration of the compartments that may have served as an additional selective factor. Both the gradient of light and the interlayer metabolite exchange are typical of modern stratified, seashore phototrophic communities (e.g. stromatolites 2593139).

Those consortia of replicators that were able to couple useful syntheses with exergonic chemical reactions could have became relatively independent of upper, photosynthesizing layers and could penetrate the depths of their porous habitats. This penetration can be considered the first wave of the Earth's colonization by living organisms that eventually evolved from replicating zymes. This colonization would have been supported by exchanges of metabolites between the illuminated strata and those that were deeper and darker. In the darkness, direct contact between the RNA-like "bodies" of replicators and the radiation-absorbing ZnS template was no longer crucial, so that the replicators could evolve into life forms that were enclosed in protein envelopes resembling modern viruses 6270219.

The energetics of these communities consisted in the interplay between the continuous ZnS-mediated photosynthesis and the increasingly complex heterotrophy of the first organisms. This heterotrophy may have been based on coupling the exergonic breakdown of photosynthetically produced metabolites with endergonic (bio)synthetic reactions. This same interplay of photosynthesis and heterotrophy still drives the majority of terrestrial communities today.

Upon further decrease in the amount of CO₂ in the atmosphere and, accordingly, in atmospheric pressure, the delivery of Zn-rich fluids to the surface of continents would have gradually ceased, so that fresh, photosynthetically active ZnS surfaces could no longer form in sub-aerial settings. After that, ZnS-rich hydrothermal edifices could persist only deeply at the sea floor, ultimately clustering around the mid-ocean ridges. The organisms that remained confined to the sub-aerial settings would have found alternative ways to reduce CO₂. Moreover, in the absence of a zinc supply, they were forced to confront unfamiliar minerals, in particular those containing iron. At that time the dominating transition metal ion in sea water would have been Fe²⁺ 90220221. Iron, unlike zinc, is redox-active and can generate harmful hydroxyl radicals 222223224225. It would seem that the expatriates of the Zn world had to be full-fledged organisms with reliable replication machinery, robust metabolism, and protective envelopes. The story of how they could populate the Earth is beyond the problem of the origin of life proper and hence remains out of the scope of this communication; this topic, however, is addressed in the accompanying article 226.

First forms minute, unseen by spheric glass,

Move on the mud, or pierce the watery mass;

These, as successive generations bloom

New powers acquire and larger limbs assume.

Erasmus Darwin, The Temple of Nature, 1802 1

The search for a solution to the origin of life puzzle is hindered by the impossibility of providing an ultimate experimental proof, namely to (re)produce life from scratch de novo (see e.g. 58218). In view of this obstacle, Wächtershäuser 227 has suggested that the hypotheses on the origin of life could be validated by rigorous application of Karl Popper's principles of testing scientific theories 228229. Popper wrote: "We may if we like distinguish four different lines along which the testing of a theory could be carried out. First there is the logical comparison of the conclusions among themselves, by which the internal consistency of the system is tested. Secondly, there is the investigation of the logical form of the theory, with the object of determining whether it has the character of an empirical or scientific theory, or whether it is, for example, tautological. Thirdly, there is the comparison with other theories, chiefly with the aim of determining whether the theory would constitute a scientific advance should it survive our various tests. And finally, there is the testing of the theory by way of empirical applications of the conclusions which can be derived from it" (quoted from ref. 229). The core of Popper's approach is the idea of falsification, i.e. putting a hypothesis to the test by making testable predictions and checking them. The falsification tests of the Zn world hypothesis, because of their key importance, are described in a separate accompanying article 226. Here I consider only the empirical supporting evidence, the internal consistency of the hypothesis, and the relation of the posited concept to other hypotheses on the origin of life.

Below, the supporting data are organized along the steps of the above-presented evolutionary scenario. One outstanding feature, however, deserves preferential treatment since it is crucial for almost every one of these steps. This feature is the unique ability of ZnS crystals to store radiation energy (see 144230231232233 for reviews). This property manifests itself in phosphorescence (afterglow), so that ZnS – widely known as "phosphor" – is used in numerous devices, from various types of displays to 'glow in the dark' items. ZnS is a broad-band n-type semiconductor with a gap band energy of 3.2–3.9 eV (depending on the particular crystal structure). When radiation strikes such a semiconductor, it may excite an electron and consequently leave a "hole" (see Fig. 1). If the energy of the radiation quantum is larger than the band gap energy, the electron reaches the so-called conduction zone and can move inside the crystal. Ultimately, the electron can recombine with the hole. However, if the electron gets into an energy trap (see Fig. 1), then the recombination can proceed only slowly, under the condition of thermal activation 234. In pure ZnS, the atoms at the surface can trap electrons 234235, particularly if the semiconductor interacts with a polar solvent such as water. At the surface, a photoexcited electron loses part of its energy owing to the interaction with the molecules of the solvent and, accordingly, is prevented from returning into the bulk of ZnS. In addition, recombination may be prevented by prompt replenishing the hole with an electron coming from a potent electron donor ("hole scavenger"). The trapped electron then remains confined to the surface until an external electron acceptor, e.g. a molecule of CO₂, picks the electron up, see 144190234236 for reviews.

Small nanocrystals of ZnS and of its chemical "twin" CdS have been found to behave as quantum dots – systems with physical properties intermediate between those of bulk materials and single molecules 144234237238239240241242. Accordingly, these nanoparticles have drawn much attention as promising fluorescent labels for biology, so that the interaction of proteins and nucleic acids with the CdS and ZnS quantum dots has been intensively studied (see 242243244245246247248 for reviews). It is noteworthy that the ZnS particles obtained from hydrothermal vent plumes could pass through 200 nm filters 249 and, hence, can be reasonably categorized as nanoparticles.

A scenario for the evolution of a complex system must consist of plausible elementary steps, each conferring a distinct advantage (Darwinian continuity principle; see also ref. 2767218). Furthermore, these steps also have to be physically plausible, which implies, in addition to the correspondence with physical laws, the continuity of underlying mechanisms and driving forces. Below, the interplay between the Darwinian and physical continuities in the Zn world is considered in more detail.

The suggested scenario invokes elements from different, even seemingly contradicting, hypotheses on the origin of life. This should not be surprising, since each of these hypotheses builds upon empirical data and/or certain theoretical bases. Accordingly, these empirical observations and theoretical considerations have to be accounted for by any scenario that claims to serve as a platform for further discussions in the field.

Oparin's hypothesis of a heterotrophic origin of life implied that the initially established metabolism could have resembled fermentation 34. The rationale behind this supposition (see 26282934 for further development of this view) was straightforward: fermentation is a simple mode of metabolism that is inherent in the most primitive anaerobic organisms. We now know that catalytic chemistry beyond the breakage or formation of covalent bonds is indeed the least structurally demanding; in most cases only acid-base catalysis is involved, often enhanced by metal atoms 368369370371. Not surprisingly, reactions of this type can be catalyzed even by ribozymes, unlike many other chemical transformations that can be performed only by proteins. The Zn world scenario builds on this rationale and suggests that primordial metabolism, as catalyzed by the first ribozymes and enzymes, constituted a set of chemically simple transformations where reactions of bond breakage were coupled with reactions of bond formation, perhaps being mediated by group transfer events. Oparin, however, did not treat primordial energetics explicitly 34. The formation of the first polymers was suggested to proceed spontaneously, without coupling to energy flow. Without such coupling, however, an unsupported formation of long polymers would contradict the laws of thermodynamics, as routinely noted by critics of Oparin's idea (see e.g. 83). In contrast, the Zn world scenario invokes solar UV light as an energy source both for prebiological syntheses and for (photo)selection.

Haldane, in addition to a fermentation-like metabolism, which he postulated independently of Oparin, suggested that "the first living ... things were probably large molecules synthesized under the influence of Sun radiation" (quoted from 5). Hence, on this point Oparin's and Haldane's hypotheses differ. In addition, Haldane argued that the complexity of the first organisms should be compatible to that of bacteriophages, thus anticipating some aforementioned modern views related to the RNA world concept 6162219. On all these points, there is an agreement between the Zn world concept and Haldane's hypothesis. Nonetheless, Haldane suggested that life emerged in the homogenous phase (it was he who mentioned the "hot, dilute soup"). In contrast, the Zn world concept considers primeval life as a surface-confined phenomenon.

The Zn world concept is in accord with the RNA world (RNA life) hypothesis 37383940414243444546474849505152535455565758596061626364656667686970717273; the RNA-based life forms may have inhabited the photosynthesizing, porous ZnS edifices that could help to (photo)select first RNA organisms, provide a shelter and nourish them. RNA organisms most likely inhabited the Zn world only in the beginning, being followed by RNA/protein life forms. Concerning RNA research, the involvement of ZnS surfaces suggested here implies that, upon modelling or simulating primeval events, electrically charged ZnS surfaces may be explicitly invoked as supporting (photo)catalytic templates (see also 414). With their assistance, replication could have been carried out by relatively simple aggregates of RNA-like polymers. In addition, the first systems of protein synthesis, owing to supporting ZnS templates, may have been fundamentally simpler than their modern counterparts (see also 73). The gradual decoupling from the ZnS surfaces and the emergence of enclosed organisms would have required more complex devices, such as primeval ribosomes, which, however, had time to evolve gradually from more simple predecessors – by recruiting new parts to functionally replace the ZnS baseplates.

The Zn world scenario incorporates elements from some "metabolism first" concepts, in particular from the "Pyrite world/Pioneer organism" concept of Wächtershäuser, who put forward a detailed concept of a two-dimensional primordial life on iron disulfide (FeS₂, pyrite) surfaces at the sea floor 124125126127128129130. The Zn world scenario exploits the suggestion of Wächtershäuser that the mineral surfaces may have promoted the abiogenic syntheses of the first polymeric molecules. In addition, the Zn world model builds upon the idea of Russell and co-workers on porous chimneys of the deep-sea hydrothermal vents as incubators for the first life forms 115116117118119120121122123. In contrast with these two concepts, the Zn world scenario considers zinc sulphide – instead of iron sulfides, which can generate potentially harmful hydroxyl radicals 223224 – as the "mineral of life", and locates the first life settings at the illuminated surface of Earth. Moreover, the suggested metabolism in the Zn world, which combines abiogenic photosynthesis with the primitive heterotrophy of the first replicators, is fundamentally different from the (bio)chemical mechanisms that were postulated for the two "iron-sulfide" worlds and based upon iron-catalyzed redox reactions (see the accompanying article 226 for more details on this point).

Kuhn and Waser have noted that the pores of primordial rocks could have served as compartments upon the evolution of the first RNA molecules 45. A related attempt to merge the FeS compartments with the RNA life has recently been performed by Martin and co-workers 117219, who have suggested that "within a hydrothermally formed system of contiguous iron-sulfide (FeS) compartments, populations of virus-like RNA molecules, which eventually encoded one or a few proteins each, became the agents of both variation and selection" (quoted from 219). Experimental evidence of polynucleotide accumulation in simulated pore systems has been provided by Braun and co-workers 415. The idea of porous natural settings as incubators for the first RNA-like life forms is used in the Zn world scenario. It, however, suggests that the function of "honeycombed" ZnS settings was not limited to compartmentalization but also included photosynthesis of metabolites, (photo)selection of most fit polymers, and (photo)catalysis of diverse reactions. In addition, the Zn concept implies that colonization of these settings by first life forms proceeded from the illuminated, photosynthesizing top to the bottom, in contrast to the scheme of Martin and co-workers in which life propagated from the bottom, "secure" part of the deep-sea hydrothermal FeS vent to the surface.

A few authors 19416417 have stressed the potential importance of high radioactivity on primordial Earth for the origin of life, albeit without suggesting how the energy of ionizing radiation could be used for biochemically relevant syntheses. The Zn world scenario incorporates such a mechanism: ZnS crystals can trap and transiently store, in a form of charge-separated states, diverse kinds of radiation energy, namely X-rays, electrons (as in displays), α-particles (ZnS was introduced as the first inorganic scintillator by Sir William Crookes in 1903), and so on. The excited electron(s) could then drive the CO₂ reduction at the surface of ZnS crystals. Hence, ZnS crystals show a unique ability to convert diverse kinds of energy into the chemical energy of organic compounds. Taking into account the high radioactivity level on primordial Earth, the ZnS edifices would have worked as natural converters of radioactive energy into reduced carbon and nitrogen-containing metabolites. The high-energy radiation could have penetrated deeper into the ZnS edifices than the solar UV light and, in addition, could have supported the ZnS photoreactors during the night hours.

Nonetheless, it is necessary to emphasize that the major source of energy for primordial life was solar light, since its energy, according to available estimates, was many orders of magnitude larger than that of ionizing radiation (see e.g. 192034109). In addition, the unique photostability of polynucleotides indicates that they were shaped by evolution under the selective pressure of solar light.

Arcady R. Mushegian, Stowers Institute for Medical Research, Kansas City, MO, and Department of Microbiology, Molecular Genetics, and Immunology, University of Kansas Medical Center, Kansas City, KS, USA

This is the first of two extremely interesting, thought-provoking manuscripts that discuss the hypothesis of the "Zn world" in which Life on Earth may have emerged. Both papers are elaborately argued and well (or, at times, beautifully) written.

The first paper leans heavily on (bio)physical and geological data, which is the area outside of my expertise. I have nominated Simon Silver to review it and have only brief remarks of my own. I have reviewed the second paper in much more detail.

p. 8 ln 14–17 "The evolutionary development from simplest prokaryotes to mammals is accompanied by a comparable increase in complexity, see e.g. 30399. This increase is routinely explained by Darwin's natural selection mechanism: more complex organisms emerge owing to their selective advantage over simpler predecessors 516792404."

-- this whole discussion is not central to the theme, but, since the authors mentions it, there are certainly recent fundamental contributions to the topic that provide non-selective explanation for several crucial transitions in the process, most notably Michael Lynch's alternatives.

Author's response: The sentence has been deleted upon the streamlining of the manuscript. Still I fully agree with this comment of the Reviewer.

p. 20 "It is noteworthy that if a protecting polypeptide could eventually discard excess energy by channeling it into a chemical reaction, then the probability of its own heat denaturation became lower; accordingly, the survival chances of the host replicator (that produced this peptide by using its own strand as a template) should increase."

-- replicator already able to translate RNA into peptides – is it not much later in evolution?

Author's response: The phrase in parentheses has been deleted.

Manuscript as a whole: There is no discussion of the effect of various concentrations of the Zn ion on the stability of the phosphoester bond. (Fe²⁺ ions are mentioned on pg 17 of the second paper, but not Zn²⁺ ions). Even if the effect is negligible, this should be reviewed.

Author's response: This is a very important point. Since Zn²⁺ ions can catalyze the formation of proper 3'-5' linkages 198349, they should also catalyze the cleavage of these linkages. Indeed, Zn²⁺ ions have been shown to catalyze the cleavage of phosphoester bonds, particularly at high temperatures 367418419. It is necessary to emphasize that the Zn world concept does not imply that primeval oligomerization was driven by Zn²⁺ ions as one-way catalysts – these cannot exist. Instead, the Zn world concept suggests, following the ideas of Wächtershäuser 124, that the accumulation of primeval RNA-like oligonucleotides could proceed because it was thermodynamically favourable. This could be due to a combination of several factors, namely electrostatic matching, high atmospheric pressure, the potential to utilize the energy of light, photoselection, and so on. Therefore, Zn²⁺ ions could accelerate the reactions but not affect their direction. However, the "life span" of primordial RNA-like polymers could have been increased if, after formation, they gained protection from the Zn²⁺ catalyzed cleavage. In the revised manuscript, I discuss how the first polypeptides, by blocking the 2'-OH groups of ribose entities, may have prevented the binding of Zn²⁺ ions to these groups and, hence, the chain cleavage.

The mechanism of polynucleotide cleavage by iron is quite different. Here, the predominant damage comes from harmful hydroxyl radicals that could be produced in the presence of redox-active Fe²⁺/Fe³⁺ ions 222223225. Because Fe²⁺ ions are photoactive 24109, the danger from hydroxyl radicals should be especially high in illuminated settings. Bearing in mind the expected high solubility of Fe²⁺ ions in the primeval waters of 10^-5–10^-6 M 90221420, we believe that Fe-rich, illuminated settings would have been hostile for pre-cellular life forms that, unlike modern organisms, had no protection from hydroxyl radicals (see also the accompanying article 226).

It is nice to acknowledge here L.E. Orgel (1927–2007), who worked and thought productively and long on these questions. Orgel helped establish reluctantly and unintentionally that there never has been agreement of among the (accurately said by Mulkidjanian to be) "many scenarios" for the Origin of Life. It is implausible and unlikely to think that this new additional model of "A Zinc World" will add to or replace the currently popular RNA World 17 or that it will simplify and make the current alternatives fewer. Mulkidjanian is also wrong if he thinks one can experimentally "falsify" (as Karl Popper hopefully, but erroneously influenced a generation of mostly-British and German scientists) what happened nearly 4 billion years ago by current efforts mostly of analysis and argumentation, and not direct experimentation. The situation concerning understanding of the Origin of Life seems more like Thomas Kuhn taught us about other realms of science: ideas disappear only when the people who like them die and are replaced by younger people with differing ideas. A contemporary of L.E. Orgel, H. Morowitz together with younger colleagues 36 has a new popular report that lays out the issues nicely and also demonstrates that the current two efforts by Mulkidjanian and Galperin will not succeed in simplifying the alterative ideas. Trefil, Morowitz, and Smith 36 agree with Mulkidjanian and Galperin that metabolism came first, before hereditary polymer replication (this is where we primarily disagree) and they quote Albert Szent-Gyorgi (1893–1986) that Life is nothing but an electron looking for a place to rest. Redox chemistry and electron transport along pathways that conserve energy are essential to the Origin of Life as to contemporary life: on that we all agree. But the rest seems far indeed for consensus thinking.

There are major scientific groups and gatherings about these topics, so there really are "establishment views" (http://www.google.com is excellent and see the International Society for the Study of the Origin of Life and http://www.panspermia.org). But there is no consensus. There is instead wide range of interest and a wide range of ideas to which The Zn World is one more useful model.

Author's response: The suggested concept of a Zn world complements the RNA world model by suggesting a physically and geologically plausible habitat for RNA-based life forms. Furthermore, the recent data of Sutherland and co-workers on the elevated photostability of natural nucleotides 17support, in fact, the Zn world concept. It seems necessary to emphasize here that the "replication first" models, such as the RNA world, leave biologists happy but cannot completely satisfy those with a background in physics or physical chemistry who realize that RNA life would be thermodynamically implausible unless coupled to some utilizable energy source. In contrast, the "metabolism first" schemes are favoured by some physicists, chemists, and geologists, but these cannot convince the majority of biologists who simply cannot accept life without replication. By combining these currently duelling models, the Zn world concept may, hopefully, satisfy physicists, chemists, and geologists as well as biologists.

I am not the first to apply a Popperian approach to the origin of life problem (see e.g. 58227). This approach is based not only on falsification tests but also on experimentation. Therefore in the current article, I cite a wealth of experimental data in support of the suggestion that ZnS-mediated primordial photosynthesis could have powered the emergence of life. In the accompanying article 226, a separate section is devoted to the analysis of potentially feasible experimental tests of the Zn world concept.

I appreciate the reference to the review of Morowitz and co-workers 36; this review is cited in the revised manuscript.

Immediately after this beginning, the author makes a very common erroneous confusion between reproduction and replication. While he quotes later on in the paper Freeman Dyson, he should have read (and followed) his demonstration that these are not at all the same, and that reproduction must have predated replication. This is, in fact, a major point for any scenario of the origin of life. I suspect, in fact, that quite a few of the references quoted in this « review » part of the paper have not been read carefully by the author. Perhaps should he restrict his references to papers he read in-depth. This would certainly make his review of models of scenarios of the origin of life less fuzzy and more accurate and incisive. In fact, rather than presenting a historical account of some scenarios (which may be interesting for philosophy of sciences, but not for a scientific article), the author should make a rational construct of all those theories. Once again, I think that the way things have been presented by Freeman Dyson is a model of what could (should) be done. This is a model of clarity.

Instead the author states: « We have argued that the "replication first" and "metabolism first" concepts complement rather than contradict each other and have suggested that the life on Earth started from a "metabolism-driven replication" ». I am afraid that this lacks proper rationality, and I would suggest him again to go back to Dyson, and re-think his scenario.

Author's response: In the revised version of the article, this review section has been deleted. The section in which the Zn world concept was compared/related to other concepts on the origin of life has been somewhat expanded and moved to the end of the article. The relation between reproduction, metabolism, and replication is now considered in more depth.

Of course I would have liked to have also suggested the author to read my books " L'œuf et la poule, histoires du code génétique, Fayard 1983 – The Chicken and the Egg, stories of the genetic code " (translated into Japanese and Portuguese, not English) and « Une Aurore de Pierres. Aux origines de la vie, Le Seuil, 1990 – A dawn on stones. At the origins of life. » but it is in French (and translated into Portuguese, but not, unfortunately in English). Too bad, but this is life. So, I will have to summarise my arguments here.

Author's response: The book Une Aurore de Pierres. Aux origines de la vie 23proved to be very useful indeed and is now cited in the text.

The problem of the scenarios, as remarked by some authors (in particular Graham Cairns-Smith, in his Genetic Takeover) is that they have to decide about initial conditions. The soup cannot be a good one, as it is a poisonous soup (there is a need for selecting relevant compounds), and furthermore, as it lack essential components: coenzymes first (as they are the basis of metabolic pathways), then nucleotides (nucleobases are meaningless), basic amino acids, and long chain fatty acids, to chose a few of the essential ingredients. It must also avoid containing molecules that are too close to the ones retained later on. No scenario which does not take these constraints at their onset are or significant interest. Typically, the very fact that the author is not aware of these constraints (which, by the way, place metabolism first, for a reason complementary to that given in Dyson's demonstration) is shown in the fact that, for example, he writes about « adenine monophosphate » (p.10, l.5). It is absolutely not trivial to make a nucleotide, in particular because ribose is an extremely unstable molecule. And this implies a steady state process to make it...

Author's response: "Adenine monophosphate" has been replaced by "adenosine monophosphate".

Concerning the stability of nucleotides: Sutherland and co-workers have recently demonstrated the potential to make pyrimidine nucleotides from cyanamide, cyanoacetylene, glycolaldehyde, glyceraldehyde, and inorganic phosphate by circumventing the stages of single nucleobases and ribose 17. Their synthesis yielded activated ribonucleotide β-ribocytidine-2',3'-cyclic phosphate as a major product and several co-products. Prolonged irradiation of this mixture by 254 nm light caused the destruction of various co-products and the partial conversion of ribonucleotide β-ribocytidine-2',3'-cyclic phosphate into ribonucleotide β-ribouridine-2',3'-cyclic phosphate. The authors concluded that there must be some (photo)protective mechanism functioning with the two natural nucleotides but not with other pyrimidine nucleosides and nucleotides. They write that "the mechanism provides a means whereby ...the two activated pyrimidine ribonucleotides needed for RNA synthesis, can be enriched relative to other end products of the assembly process" 17. They have even claimed that "the prebiotic synthesis of activated pyrimidine nucleotides should be viewed as predisposed" 17. This result is a clear-cut evidence that photoselection was a potential driving force behind the emergence of the first biological (macro)molecules.

Concerning the steady-state processes: I do not question their primacy. On the contrary, I argue that ZnS-mediated photosynthesis was the primary, steady-state process that supported the emergence of life. Unlike the postulated reactions at pyrite surfaces 124, which remain elusive, ZnS-mediated photosynthesis has been repeatedly shown to proceed with high yield at moderate temperatures 144149272.

However, I do not share the view of Dyson and others who consider the systems of interacting steady-state flows as alive. As we have argued elsewhere 85, it would seem more reasonable to date the origin of life to the establishment of the first replicating systems.

Now, a second part of any scenario, needs to state where we start, and from what data. If genetic takeover happened, then we are witnessing palimpsests (as suggested by Steve Benner) and there is not much hope, except for purely educated guesses. In contrast, if we still can have archives, then we could start from extant metabolism. This was the idea of Sam Granick and later on of Gunter Wächtershäuser (who ignored Granick, which made difficult uncovering his remarkable work: it took me several years to stumble upon that work). Here, the idea is to remark (which the author apparently ignores) that anabolism is very special in that it is essentially made of molecules carrying negative charges, often phosphate. Sulfur is also an essential component (as it remains for the author), associated to iron and to the idea of electron transfers (a neat idea with the prevailing conditions on Earth, when oxygen was not there, so that reversible electron transfers could be quite efficient).

Author's response: Indeed, many metabolites carry negative charges, and we have extensively discussed this point in our previous article 85. In the revised manuscript, it is now explicitly mentioned that photosynthetic reduction of CO₂ yields negatively charged molecules that could remain confined to the ZnS surfaces.

Now, curiously (in fact I do not exactly understand why, as this would fit quite well with a model where zinc would play the role I think iron played) the author tries to discard the idea of surface metabolism. This is however, to my view, one of the most indisputable scenarios explaining how life could have led to a constructive metabolism. Granick, Wächtershäuser and I should include myself, argue very strongly for that step as the basis of selection (much needed to avoid the poisonous broth) and or polymerisation. True that Miller and De Duve (for reasons which are obscure to me, except that surfaces led to much less interest in sparks and the like) tried to discard the idea of surfaces by stating that chemical reactions on a surface would simply preclude separation from the surface of the polymers: this is a completely far-fetched argument, ignoring (purposedly?) that the polymerisation is not driven, in most cases by enthalpy, but by entropy (because of elimination of a water molecule, which goes in the solution: exactly the opposite of what would happen in the solution, where polymers would tend to be unstable unless the medium is highly compartmentalised, with a role on water activity). Their argument cannot be retained. Worse, I think that De Duve should have, on the contrary, built up his theory (which is excellent) of the role of thioesters, on surfaces: 4-phosphopantetheine would be the swinging arm to make many complex compounds including the much wanted long fatty acids !!! This is what I tried to demonstrate in my books (and in the short summary I published in English, with a conjecture on the synthesis of nucleotides 74).

Author's response: The idea of surface chemistry is not discarded. In contrast, I build upon the works of Wächtershäuser in this respect and argue that the syntheses of polymers could proceed at ZnS surfaces where the first RNA-like polymers were protected from photodamage. In the Zn world, therefore, the selection was not a one-step process but a two-step process: first, electrostatically preferred molecules adsorbed at the ZnS surfaces and could participate in diverse photochemical reactions, and next, those product molecules that were particularly photostable (e.g. oligonucleotides) were photoselected. As noted above, a photoselection of natural nucleotides has just been experimentally demonstrated by Sutherland and co-workers 17. Thus, I consider the surface confinement of first polymers as a pre-condition of their survival in a UV-irradiated environment.

I do not think that it is possible to rebuff the argumentation of De Duve and Miller 200, because their reasoning was based on a straightforward application of energy conservation law. The mechanism of the membrane ATP synthase – as is considered in the article – exemplifies that, if the free energy (ΔG) of a particular synthetic reaction differs in two environments, then a portion of free energy – equal to the difference between the two respective ΔG values – should be "invested" to transfer the reaction product between the two environments. But, and this is the point that I have tried to emphasize, the displacements

within

So much for these points: if I go on, I will write a new book! Just one final point: the author writes " The living organisms are evolving, dissipative systems that can exist only being supported by energy flow ", this is either trite or wrong (or both) and this rests on a completely wrong view of the way entropy fluxes are involved in Nature. The author (and I can send him the paper) should read the remarkable paper of Rolf Landauer " Inadequacy of entropy and entropy derivatives in characterizing the steady state " published in Physical Review A, in august 1975. He would certainly think twice, delete this sentence (which does not bring anything more than fuzzy words to his demonstration) and omit any reference to Prigogine anyway. This would not harm his work, quite the contrary.

Author's response: It would seem that there is some misunderstanding on this point. If I understand properly, the Reviewer means that Prigogine was wrong while suggesting that dissipative structures tend to reach states with minimal entropy production. Landauer showed in his paper that this rule is not general and is valid only in some cases that are close to equilibrium 422. The living systems are far from equilibrium, and therefore this rule is most likely not applicable to them in any case. I have been long aware of this controversy between Prigogine and his opponents, and therefore these ideas of Prigogine were not even mentioned in the original manuscript. I refer to Prigogine in another relation. Prigogine has repeatedly argued, in his many articles and books, that the emergence of ordered complexity in nature requires continuous supporting energy flow and out-of-equilibrium settings. Since some hypotheses on the origin of life simply ignore this point, I find it worthy to emphasize that life could not emerge either "for free" or in an equiponderated primordial soup – and cite Prigogine in this respect. Nevertheless, I have somewhat changed the respective wording to avoid any misinterpretations.

Now, I come to his opinion about zinc as central to the origin of life. This is based on an idea, worth investigating, placing light at the centre of energy management. Why not, this certainly happened with cyanobacteria and before. Why not very early on. I have no major objection to the generality of the idea. However I do not think that the way the author involves nucleobases is convincing. Light needs to be, indeed, coupled to energy, and electron transfers are cases in point. This may place heme or related molecules in the limelight, but also ... iron. Certainly not nucleic acids. In contrast, I would certainly accept ZnS (but I did not have access to the second paper, where falsification tests are proposed). Photosynthesis in a Zn world, why not, and this apparently would happen as a mineral surface metabolism (so that I did not really understand why the author was so much against the idea). But I would not easily accept nucleobases as primitive: as I stated above I think nucleotides must be made more or less directly, and I proposed a scenario which sees them as a leak of a general process of nitrogen fixation. This might well work in a Zn world.

Author's response: While being quite happy that the Reviewer has no major objection to the idea of abiogenic ZnS-mediated photosynthesis, I would like to reiterate that I am not against the idea of surface "metabolism" (see above).

I readily agree that it is difficult even to imagine a spontaneous formation of nucleotides in the "primeval broth" from occasionally formed nucleobases, ribose molecules, and phosphate groups. The Zn world concept, however, does not require the formation of single nucleotides in solution. The concept implies that polynucleotides could directly assemble at the ZnS surfaces from simpler building blocks. In the aforementioned work of Sutherland and co-workers, the viability of such syntheses has been experimentally demonstrated 17. In this way, the step of mononucleotide synthesis could be bypassed. The survival of the entire polymeric constructs could then be guaranteed by their extreme photostability. In this framework, single nucleotides could have first appeared as cleavage products of the polynucleotides.

Second, I would urge the author to think of replication as secondary to metabolic reproduction, where peptides were present first (this is what I tried to demonstrate in my "The chicken and the egg" book (see above)). Also, the author, once again, needs to think of the origin of coenzymes! This can certainly be included in his scenario. Nucleic acids are latecomers, they came first as surface substitutes, then discovered that rather than being substrates they could become templates. There I am interested by the idea that Zn favors formation of 3'-5' bonds (which is indeed a very difficult feat, normally favoured by the presence of peptides).

Author's response: In fact, I think (and write) about replication as a phenomenon that was secondary to the reproducing, steady-state flows of energy and matter (although I think the word "metabolism" might be misleading in relation to abiogenic processes).

The "metabolism first" models, however, have certain difficulties in explaining the transition to the first replicating entities. For example, Dyson has suggested a multistep model with the reproducing but not replicating cells emerging first and followed by nucleotide-based replicators initially as parasites of these cells and then as their symbionts 398. However, neither the (bio)chemical origin of the replicators nor the selection factors that could favour their emergence were addressed. In contrast, the Zn world model considers photoselection as a force that, by driving the emergence of photostable entities built of stacked and paired nucleotides, could have paved the way to the first self-replicating systems.

Therefore, I cannot agree that nucleic acids were latecomers. In the literature, many arguments are in favour of RNA-first scenarios, the majority referring to the ability of RNA molecules, but not proteins, to replicate; I do not want to repeat this argumentation here. One more argument, based on selection criteria, can still be invoked. One can presume that nucleobases, as well as amino acids, could be selected for their particular tasks because of specific properties that were useful under given circumstances and were shared by all nucleobases and amino acids, respectively. The common property of nucleobases is their extreme photostability. Hence it has been repeatedly hypothesized that they were selected because of this property (see 1785159165 and references therein). The respective selective factor, solar UV irradiation, is abiotic, so that the selection could have started before the origin of life. The common property of amino acids is the presence of amino and acidic groups that allow them to join into polypeptide chains. The selection for this property, however, implies the existence of devices that could connect amino acids by peptide bonds. In modern organisms, this function is routinely performed by RNA-based machines (ribosomes); it is even now becoming clear how these machines could evolve from simpler RNA constructs already capable of peptide bond formation (see 73 and references cited therein). This consideration, in my opinion, strongly supports the RNA-first schemes.

Finally, a word about the quotation by the author of the work of Rolf Landauer. It happens that I was lucky enough to meet him before he passed away, and to discuss much of the questions involving entropy and information (he showed to me how Prigogine was completely wrong, for example, see his paper cited above). The important point here is to see that if creation of information is not a real problem, accumulation of information is. And this requires an energy source. It is most likely, to my view, that this was originally performed by a mineral such as polyphosphate, which is quite metastable, and later on perhaps by nucleotides and polynucleotides (such as what is seen in the management of energy in the degradosome today, see my Phylogenetic view of ribonucleases 423).

Author's response: The energetics of the accumulation of information is a very important point. Since polynucleotides are not eternal, replication per se can be considered as a means to preserve (accumulate) information before an information carrier – a polynucleotide strand – eventually breaks down because of thermal fluctuation or (photo)cleavage. In the manuscript, several tentative energy sources for replication are considered. Initially, the replication could be supported by the thermodynamically coupled cleavage of available polynucleotides – "failed" information carriers, in the given context. The cleavage of inorganic polyphosphates, as suggested by the Reviewer, could provide energy as well; chemically, this reaction is similar to the cleavage of polynucleotides. This possibility is now mentioned.

Well. To complete this review, I think that the author should reorganise it completely, shorten some of his points and take into account really seriously the reflection of Freeman Dyson, which he places only at the end of his text.

Of course as much of what I wrote is opinion (doxa), I have no strong reluctance for the publication of this work.

Author's response: The text has been streamlined, and the discussion of Dyson's views has been expanded. In the revised manuscript, photoselection is explicitly considered as a possible bridge between reproduction and replication.

First, for a hypothesis paper, the manuscript does a good job in collecting literature for a specific scenario, namely UV-triggered biosynthesis on Zn-surfaces. Also the focus on possible physical boundary conditions is the way to go in origin of life research in my viewpoint. The paper's discussion will lead to new experiments, which is a good thing.

That said, I think the author is a bit 'pushy' on his proposal – calling it Zn-World as if Zinc would be a common biological molecular construction. Also saying in the introduction that the Zn world proposal "... resolves the conflict between the "metabolism first" and the "replication first" concepts of abiogenesis, which are suggested to reflect the two facets of the Zn world, namely the continuous abiogenic photosynthesis of metabolites and their further conversion by the ZnS-confined replicating entities." may be correct on the first, but sketchy to say the least on the second.

Author's response: The tag "Zn world" was chosen to emphasize that the (photo)chemical properties of Zn²⁺ ions and Zn-containing substances could have shaped the first life. While the photochemistry of ZnS crystals could have governed the nature of photosynthesized compounds and that of their photo-derivates, the catalytic properties of Zn²⁺ ions may have determined the particular traits of the first (bio)polymers, e.g. the choice of 3'-5' linkages for RNA strands.

Why the author chooses to make far-fetched connections to very old literature (Erasmus, Darwin's pond, to give two examples), I do not understand. Literature from times where the molecular basis of biology or thermodynamics was not understood can only be lucky to make a proper proposal, but is not helpful to be repeated in current literature.

Author's response: I quote the older literature to express my deep respect for scientists who could make suggestions that were correct from the viewpoint of thermodynamics in "times where the molecular basis of biology or thermodynamics was not understood". Ironically, some hypotheses on the origin of life that were put forward during the last decades appear shaky from the thermodynamic viewpoint.

I am missing a discussion on whether clouds have obscured early earth or whether land mass and shallow waters have existed at all on the early earth. This would substantiate two fundamentals of the concept.

Author's response: Clouds, if there were any at that time, should have consisted of water. Water is transparent to UV beams (of wavelength > 200 nm 425). Therefore, one can become tanned even on a cloudy day.

The presence of the first continents already at 4.2 Ga 188implies a shoreline around them and water patches at their surface, in particular at sites of volcanic or geothermal activity.

Another question I am missing in the manuscript is in which geometry and location the author sees the UV radiation not being absorbed by porous rock or sedimentation layers. As it stands, the argument is basically restricted to surface water some meters deep and rock structures some micrometer thin.

Author's response: Taking into account the expected interplay between the continuous precipitation of ZnS and photocorrosion, I would instead estimate the thickness of the habitable layer as being several millimetres (similar to that of modern microbial mats built up of phototrophic prokaryotes).

In addition, I would like to note that the ZnS-covered areas could spread over hundreds of square kilometres – owing to the high thermal activity of the young Earth.

Above relates to the puzzle I have how a Zn world should accumulate and trap their molecules by a nonequilibrium process and at the same time be fully exposed to UV radiation. I guess, exposure to UV inherently means that the molecules synthesized diffuse then out into the open waters?

Author's response: As noted in the revised manuscript, the photosynthesized molecules should be electrically charged and could adsorb at the surface of ZnS (see also 85). In addition, the porous structure of ZnS precipitates should constrain the diffusion of photosynthesized molecules out into the open waters (see the structure of modern ZnS precipitates in 179180181). Moreover, as discussed in relation to the works of Wächtershäuser 124126, the molecules that were polymerized at the surface would remain confined to the surface.

The author takes "Mauserall has introduced a major additional constraint by noting that the energy requirements of the first living beings had to be compatible with that of modern organisms 109. He argued that "the ur-cell would be simpler, but it would also be less efficient" " to the conclusion "More rigorously speaking, the intensity of the energy flux(es) that supported the emergence of life should be either comparable with the intensity of modern life-supporting energy flows or even stronger.". I do not see such a strict logic connection, since it assumes that the biomass production should have been equal. It would be helpful here if the authors would clarify his understanding of "intensity" (i.e. number of photons) as opposed to for example "wavelength" (i.e. the energy gap).

Author's response: Here I would just like to quote David Mauzerall: "The ur-cell would be simpler, but it would also be less efficient. A crude number for this energy flux is 10 mwatts per gram of biocarbon. In anticipation of the stress on photochemistry, this flux can be normalized to energy flux per area, using the area of a 'typical' prokaryote cell ~10 μ² and its weight of carbon, 0.1 ng. This number is 0.01 mwatt per cm². Table 1 shows the energy sources of the primitive earth. It is clear that only solar photon energy is sufficient for the continuing evolution of life." The quotation is taken from 109, in which the table referred to can be found.

At the core of the argument is the finding that DNA/RNA can get rid of harder light quanta by the hydrogen bonding of the bases. However why then are the bases constructed such that they absorb UV radiation with good efficiency at all? The absorption cross section is so large that it is even used for DNA quantification. I would guess that one could probably readily design molecules which would combine both a lower UV absorption and – if they do still absorb one – get rid of the energy with similar mechanisms. Am I missing something here?

Author's response: Lower UV absorption of a chemical compound means that a passing UV quantum is captured not by the first molecule on its way but, say, by the tenth molecule. But then this tenth molecule still has to cope with the entire energy of the UV quantum. As long as organic molecules are considered, deactivation of a UV quantum requires prompt spreading of its energy over many bonds (therefore conjugated bonds are needed) and an efficient sink for energy (here several diverse mechanisms might be involved 164165). In the absence of conjugated bonds, the energy of UV quanta will "gather" at the "weakest" bond and eventually break it. As a result, the quantum yield of photodamage might be much higher for a compound with low absorption of UV than for a compound with high UV absorption and many conjugated bonds. Goosen and Kloosterboer 166have shown that the release of phosphate in response to 254 nm light was 10 times slower in the case of adenosine monophosphate (AMP, a highly UV-absorbing compound) than in the case of glycerol 2-phosphate (which barely absorbs at 254 nm). It would seem that, in a kind of trade-off, the increased absorption cross section was over-compensated by increased photostabilty in the case of AMP. In this framework, all means that increased the spreading of excitation energy without increasing the absorption cross section would be strongly favoured by evolution. The stacking of nucleotides, as well as their pairing, would result in the spreading of excitation energy over a larger number of chemical bonds without increasing the absorption cross section.

It would be useful if the author could also comment on the "Lost City" type vents as his arguments on page 10 apparently only applies to so called "black smokers".

Author's response: Lost City-type vents were found at depths of ~700–800 m (see 102for a review). Their hydrothermal fluids were cooler than those of black and white smokers (45°C–75°C versus 250°C-350°C). Not surprisingly, these fluids were depleted of any transition metals and were low in H₂S. It is not clear how metabolism of any type could emerge under such conditions in the absence of transition metals as catalysts.

Beginning with about page 15 the manuscript tends to become highly repetitive or very vague in its argumentation. I would like to point to "low density" parts of the manuscript in the following. Shortening there would make the manuscript much more accessible.

The repetitive sections of the manuscript start at page 15. I will give a short list where I see strong repetitions:

- 15.22–16.2: Spreading excitations were mentioned before

- 16.3–16.11: highly repetitive

- 22.17–23.15: I do not see why the paper needs to cite Popper here.

- 23.18–24.28: I see this as repetitive review of thoughts formulated before

- 25.23–27.15: Many repetitions, things of review character and not adding much to the hypothesis proposed

- 27.16–28.8.: Highly repetitive

- 29.21–29.27 was mentioned before

- 34.20–35.13 is quite repetitive

- 37.4–37.11 was mentioned before

- 39.3–40.15: I do not understand why we need a review of Oparin and Haldane here.

- 42.13–46.14 is both repetitive and vague

- 46.23–47.18 was said before quite often

- 48.23–49.16 is a review and does not add much to the discussion

- 50.5–50.12 is repetitive

Author's response: Hopefully, the repetitions are less frequent in the revised manuscript.

There are rather vague ideas:

- 17.7–17.17: The conclusion that dsRNA-selection leads to replications systems is not well documented.

Author's response: Regretfully, I cannot document better the properties of the first replicating entities that emerged more than 3.5 Ga ago. However, the self-replicating RNA system, as presented recently by Lincoln and Joyce 72, consists of ribozymes that are built of paired RNA strands.

- 18.1–18.12: The should-heating of a hundred 'unit' RNA-like polymer should be complemented by more data, for example by the time scale and photon rate over which the tens of degrees are applied. I would think there is a difference between bleaching and thermal damage. The subsequent conclusions in the manuscript are in my view far fetched.

Author's response: The estimate from 210, which I have used, was calculated for an ensemble of bacteriorhodopsin molecules, each being hit by one quantum of visible light. I would strongly encourage a corresponding quantitative analysis for RNA molecules and UV light quanta.

- 18.19–22.15: Much of what is written here is rather hopeful speculation and uses imprecise language.

Author's response: Indeed, the story of how the first proteins could have emerged is hopeful speculation. However, the field is developing fast. In the revised manuscript, I invoke the just-published seminal work of Bokov and Steinberg 73 and use more precise language.

- 33.23–34.2: I cannot see the biological evidence which supports the connection between dye sensitization on ZnS surfaces with "zyms" helping in adding the yield of metabolite production.

Author's response: Although the photochemical interactions between polynucleotides and ZnS/CdS quantum dots have been actively studied, the light-gathering function of polynucleotides in such systems, to the best of my knowledge, has not yet been quantified. Therefore, I have used as my basis quantitative data obtained with ZnS crystals and adsorbed organic dyes 329.

- 37.20–37.24: It is not clear to me what colonization waves have to do with the "Zinc" world.

Author's response: From a biological viewpoint, the transition from the Zn world to modern life has to be envisioned in an evolutionary scenario.

- 48.4–48.22: Both is the hypothesis presented here vague and I also do not see a connection to the Zn-World.

Author's response: The Zn world concept has two major constituents, namely ZnS-catalyzed photosynthesis and the UV selection of most photostable polymers 112. Accordingly, the emergence of the first information-carrying polymers as a result of photoselection is a legitimate topic of the Zn world scenario.

Also, a strange reference is "in accordance with the "rings joined to rings" scenario of Erasmus " on page 16.4 which can hardly be taken seriously.

Author's response: In my opinion, the expression "rings joined to rings", taken from the work of Erasmus Darwin 1, adequately describes the emergence of the first RNA-like polymers. Their formation should have invoked both the stacking of ring-like nucleobases and their Watson-Crick pairing.

Much more detail would be in my opinion necessary to relate the presented scheme of UV absorption to X-rays. Much more reaction surface could be gained at the subsurface. However we only read "Taking into account the high radiation level ... converters of radioactive energy into reduced carbon and nitrogen containing metabolites." Some more elaboration on this point would be very helpful.

Author's response: Radiation chemistry is not my field of expertise. A more detailed analysis of the "radio-synthesis" of organic compounds within ZnS edifices might be a worthwhile task.

Last but not least I find the last sentence of the paper rather revealing: what could the aesthetics of minerals to do with a scientific argument on the origin of life?

Author's response: Aesthetic criteria are of great importance in scientific research, see e.g. 426. For example, my initial opposition to the idea of abiogenesis at the floor of the Hadean ocean 117, when I first heard about it, was purely aesthetic. I simply did not like the idea of the origin of life being in complete darkness. Only later on I realized the (bio)chemical shortages of this scenario and, at the same time, learned to appreciate the seminal idea of Russell and co-workers on inorganic metal-sulfide compartments as incubators for the first life forms. As a result, initial aesthetic opposition could be transformed into a scientific context.

I know that the open refereeing process makes it risky to raise the voice (the scientific exchange has many possible for punishment) – but without accepting criticism there is no progress in science.

Author's response: I fully agree with this statement by the Reviewer. A major advantage of the open review procedure is that referees are free to express their own opinions on the subject. As a result, the open exchange of ideas between authors and reviewers has its own value.