A Universal Geochemical Scenario for Formamide Condensation And

DOI:10.1002/chem.201803889 Review

& Prebiotic Chemistry AUniversalGeochemical Scenario for Formamide Condensation and Prebiotic Chemistry Raffaele Saladino,*[a] Ernesto Di Mauro,[a] and Juan Manuel García-Ruiz*[b]

Chem. Eur.J.2019, 25,3181 –3189 3181 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review

Abstract: The condensation of formamide has been shown tures forming under alkaline silica-rich solutions are excellent to be arobust chemical pathway affording molecules neces- catalysts for the condensation of formamide with respect to sary for the origin of life. It has been experimentally demon- other minerals. We also propose thatthese structures were strated that condensation reactions of formamide are cata- likely forming as earlyas4.4 billion years ago when the lyzed by anumber of minerals, including silicates, phos- whole earth surface was areactor,aglobalscale factory, re- phates, sulfides, zirconia, and borates, and by cosmicdusts leasing large amounts of organiccompounds. Our experi- and meteorites. However,acritical discussion of the catalytic mental resultssuggestthat the conditions required for the power of the tested minerals, and the geochemical condi- synthesis of the molecular bricks from which life self-assem- tions under whichthe condensation would occur,isstill bles, rather than being local and bizarre, appears to be uni- missing. We show here that mineralself-assembled struc- versal and geologically rather conventional.

1. Introduction marizedbythe aphorism:“metabolism-first or genetics-first?”) might thus consist on thepresence and on the productivein- Origin of life lacks at present an uncontroversial scenario and teractions of elements of both pre-genetics and pre-metabo- unequivocal chemical and physical–chemical frames of refer- lism. It is difficult to imagine that metabolic processes of any ence. Recent progress, however,allow some mote of opti- level of complexitycould develop and evolve in the absence mism. of aform of information transmission and that information- Life is arobust phenomenon. These considerations suggest containing molecules could form and replicate in the absence that its first chemicalsteps must have been as robust and effi- of some form of energy harnessing and control. Hence the ne- cient. Indications that the core chemical frame of the early pro- cessity for prebiotically fertile syntheses to be as rich and varie- cesses was centered on the properties of hydrogen cyanide gate as possible, necessarily affording compounds of different HCN and, more diffusely and more abundantly,onthose of its type and purport. hydrolyzed but still highly reactive derivativeformamide Given the elemental compositionofbiological molecules,

NH2COH have been collected during the last twenty years. largely consisting of H, O, Cand Natoms,the starting com- Here we discuss recent experimental evidences that the geo- pound(s) should reflect such composition. One-carbon atom chemicalprocesses of the early earth provided the necessary, compounds as formaldehyde, formic acid, hydrogen cyanide, efficient, and selectivecatalytic setupfor the transition from and formamide abound in circum-and interstellar space.[1–6] one-carbon atom compounds to nucleic bases,amino acids, Thus, the conditions under which these syntheses may occur sugars,and carboxylic acids. The processeswhich were at the are multiform and potentially diffused at universal scale.[7–13] basis of the geochemical evolutionofthis planetprovided at Reactions from these compounds leadingtohigher chemical the same time the niche for the formation and the environ- complexity occur under avariety of energy sources, of differ- ment for the evolutionofbiogenic compounds. ent chemical mixtures, of catalysts, and of varying physical- Biogenicprocesses were necessarily preceded by the forma- chemicalconditions.[14] tion of prebiotic compounds.Occamian logic (“the simplest is The logic for individuation of potentially biogenic com- the likeliest”) suggests that the conditions in which these earli- pounds works top-down: amolecule or aprocess is prebiotic est processes occurred were not rare, occasional, exotic, or fas- and biogenic if it is aposterioripossible to recognize its bio- tidious. The solutiontothe bias of the origin of the prebiotic logical role or relevance. In this respect, HCN andNH2COH compounds and to the start of their fertile interactions (sum- have shown their worth. HCN and NH2COH chemistries are related,[15–17] as predicted and shown by an in-depth computa- tional study of the formationand decomposition reaction [a] Prof. R. Saladino, Prof. E. Di Mauro channels of NH COH by meansofabinitio molecular dynamics, Dipartimento di Scienze Ecologiche eBiologiche 2 Università della Tuscia,Via San Camillo De Lellis, 01100 Viterbo (Italy) allowing the study of reactions both in gas and in liquid [18] E-mail:[email protected] phase. This analysishas shown asimilarstabilityofNH2COH [b] Prof. J. M. García-Ruiz and HCN in solution as well as their relatively facile intercon- Laboratorio de Estudios Cristalogrµficos [19] version. NH2COH was shown to form in the classical Urey– Instituto Andaluz de Ciencias de la Tierra, CSIC-Universidad de Granada Miller condensation experiment,[7] alongwith formic acid but Av.Delas Palmeras 4, Armilla (Granada) (Spain) E-mail:[email protected] not with HCN. Formamide has in particular shown its versatili- [20] Supporting information and the ORCID identification number(s) for the ty owing to the mentioned ubiquity and to its facile forma- author(s) of this articlecan be found under: tion and possibility of concentration.[21,22] https://doi.org/10.1002/chem.201803889. The geochemicalscenario taken into consideration has been 2018 The Authors. Published by Wiley-VCH Verlag GmbH&Co. KGaA. chosen based on its geochemicallikelihood and its evidence- This is an open access article under the terms of Creative Commons Attri- based consistence. This scenario has been experimentally chal- bution NonCommercialLicense, whichpermitsuse, distributionand repro- duction in any medium, provided the original work is properlycited and is lengedwith NH2COH synthetic properties (as described below), not used for commercial purposes. because of its widespread presence andthe experimentally ob-

Chem. Eur.J.2019, 25,3181 –3189 www.chemeurj.org 3182 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review served flexibility of its chemistry.This approach does not ex- biotic mixture is usually richinproductsirrespective of the clude the eventual integration with the possible contribution nature of the specific mineral present during the condensation. of other reactants as HCN and/orHCOOH. However, one On the other hand, some selectivityoccurs dependingonthe should consider that the more complex is the prebiotic reac- mineralogical origin and elemental composition of the catalyst. tion mixture, the higheristhe possibility of encounteringthe For example, in the silicatefamily the synthesis of pyrimidine “asphalt problem”, as pointed out by S. Bennerand co-au- nucleobases is enhanced by the presence of either magnesium thors.[23] and iron (Table 1, entries 7–9).[28] In the case of sulfur-contain- We report in Part 2, Figure 1, the biologically mostrepresen- ing minerals, sulfur–copper–iron minerals are generally more tative compoundsthat are obtained by reactingNH2CHO in active than simple pyrite and pyrrhotine, pyrite being more ef- the presence of various minerals, as fueled by differentenergy fectivethan pyrrhotine (Table 1, entries 10–17).[29] Again, in the sources. Among the variousclassifications possible, the com- case of borates,the so called hydrates containing hydroxyl pounds obtained are listed as afunctionofthe role that they groups (group A, Dana’s classification), and anhydrous borates assume in biological systems(i.e.,nucleic bases, amino acids, (groupB,Dana’s classification), are more efficient in the forma- carboxylic acids, sugars, etc.), rather than more formal chemical tion of nucleobases than borosilicatesand borocarbonates [30] principles (for example, chemical composition;number of (Table 1, entries 22–29). Finally,the NH2CHO condensation carbon atoms, etc.). works well also in the presence of water,thermalwater being The result of this series of syntheses boils down to the con- areaction medium better than sea water when the condensa- sideration that NH2CHO yields mixtures of prebiotic relevant tion is performed in the presence of the same type of miner- compounds in every energy and catalyst combinationtested. Thus, in order to understand the scenariothat allowed the biogenic processes on this planet, the question becomes:are the geochemical conditions that bona fide characterizedthe earli- RaffaeleSaladino is President of the Italian est times of this planetcompatible with the production of bio- Society of Astrobiology. He obtained aMaster genic molecules?The answer to this question is positive, as degree in Chemistry at University “Sapienza’’ described in Part 3. (Rome, Italy), and aPh.D. in Chemical Scien- ces at the same University.After post-doctoral training at the University of Montreal 2. The Effect of Minerals on Formamide (Canada)and teaching and researchposition at the University of Tuscia, Viterbo (Italy), he is Condensation now Full Professor in Organic Chemistry at Mineralsplay akey role in the condensation of NH CHO. In the the same University. He is involved in studies 2 on prebiotic chemistry,green catalysis and ox- absence of minerals, NH2CHO yields only avery limitedpanel idativetransformationsofnatural substances of compounds, including simple purine, associated with very mainly focused on bioactive compounds. low amountsofN,N-diethyl formamide,3-hydroxypyridine, pyr- uvic acid and 2-propanol, under radiation conditions, and urea, Ernesto Di Mauro wasborn in Valmontone lactic acid, andpyruvic acid, under thermal condition at high (Italy), in 1945. In 1967 he obtained his pH values.[24] degree in Biological Sciences from “Sapienza’’ As reportedinFigure 1, the number and variety of conden- University of Rome(Italy). In 1969 he joined the Department of Genetics(Seattle, USA),as sation products increase enormously in the presence of miner- apostdoctoral fellow. Appointed in 1978 as als, encompassing nucleic acid components with different de- an associateprofessor of Enzymology at the grees of structuralcomplexity (nucleobases,acyclonucleosides University of Rome, he has been aprofessor and nucleosides),[25] amino acids, sugars, carboxylic acids, and of Molecular Biology since 1987. His research interests are centered on gene regulation, some organic condensing agents(e.g.,carbodiimide),which DNA and chromatin structure and topology are essential for the formation of oligomers and polymers. and, at present, on the various aspectsofthe originoflife. Table 1summarizes aselected panel of reactions of NH2CHO performed in the presence of minerals reasonably present in the geochemical scenario of the primitive earth.Simple metal Juan Manuel García-Ruiz is Research Professor at the National Research CouncilinGranada oxides (Al2O3,SiO2,TiO2), clays, silicates(olivine, forsterite, faya- (Spain). Founder of the Laboratory of Crystal- lite, and their derivatives), sulfur–iron and sulfur–copper–iron lographic Studies, and of the Crystallization [26] compounds, zirconia minerals and borate minerals, are all ef- Factory,heisaninternationalexpertinthe ficient catalysts in the synthesis of biomolecules from NH2CHO. crystallization of minerals,drugs and proteins, In the selected cases, the reactions have been performed including self-assembly in natural and syn- theticmaterials. He is currently leading the from neat NH CHO under thermalconditions, mimicking vol- 2 EuropeanResearch Council project “Prome- canic or hydrothermalscenarios. Acomprehensive table, in- theus”onearly Earth. Prof. García-Ruiz is the cluding extraterrestrial minerals (meteorites, cosmicdust sam- author of the book “The Mystery of the Giant ples),[27] as wellasenergy conditions related to space condi- Crystals”. In 2006 he initiatedthe governmen- tal Program Explora. tions (proton beam, high energy heavy atom beam) is report- ed in the Supporting Information. As ageneral trend, the pre-

Chem. Eur.J.2019, 25,3181 –3189 www.chemeurj.org 3183 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review

Figure 1. The biologically more relevant compounds obtained by reacting NH2CHO in the presence of various minerals. A: nucleobases. B: nucleosidesand acyclonucleosides.C:sugars. D: carboxylic acids. E: amino acids, condensing agents, and HCN oligomers. al.[31] At the moment,the highcomplexity of heterogeneous yield of each compound rarely exceeds 0.1%. This might catalysis, associated to the specificityofthe mineral surface appear as amajor limitation of the system if considered in the chemistry,does not allow to delineate detailed structure–activi- logics of preparative chemistry.Inthe prebiotic perspective, ty relationships able to correlate aspecific mineral (and an this fact is only apparently adrawback;given the large diversi- energy source) to adefined NH2CHO reactionpathway.The ty of compounds formed together (nucleic bases, carboxylic

Chem. Eur.J.2019, 25,3181–3189 www.chemeurj.org 3184 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review

Table 1. Aselected panelofreactions of NH2CHO performed in the presence of mineralsreasonably present in the geochemical scenario of the primitive earth.

Mineral Purine[a] Pyrimidine[a] Carboxylic Amino Condensing nucleobases nucleobase acids[a] C1/C6 acids[a] agents[a] metal oxides 1alumina 40.45.7 –––

2SiO2 4.9 5.7 –––

3TiO2 28.31.1 ––– clays 4kaolin 51.9. 2.0 ––– 5zeolite 83.14.4 ––– silicates

6MgFeSiO4 –5.5 ––0.2

7Fe2SiO4 –5.2 –––

8Mg2SiO4 –0.13 ––– sulfur iron and sulfur copper/iron minerals 9FeS 0.01 0.18 0.01 –0.01 [b] 10 pyrrhotineFe1 xS 0.01 0.13 0.03 –0.04 À 11 FeS2 0.26 1.05 ––0.04 [c] 12 pyrite FeS2 0.28 0.73 ––0.21 [c] 13 chalcopyrite FeCUS2 0.20 6.56 ––6.80 [c] 14 bornite FeCu5S4 1.80 1.71 0.50 –0.05 15 tetrahedrite (Fe,Cu,Sb)S[d] 0.19 3.07 ––0.42 16 covelliteCuS[d] –5.86 0.50 –0.32 zirconica minerals [e] 17 ZrO2 0.08 0.83 0.08 –0.16

18 ZrSiO4 0.01 0.21 0.04 –0.07

19 Li2ZrO3 0.0001 0.110.01 –0.05

20 BaZrO3 0.05 1.06 0.14 –0.63 borate minerals [f] 21 borax Na2B4O7·5H2O – <0.01 0.03 < 0.01 – [f] 22 CaMgB6O8(OH)6·6 H2O –0.04 0.03 < 0.01 0.18 [f] 23 MgB3O3(OH)5·5 H2O –0.02 0.03 0.05 0.01 [g] 24 Mn3B7O13Cl 0.02 <0.01 0.20 0.09 – 3+ [g] 25 Mg2Fe BO5 0.02 <0.01 0.61 < 0.01 – 2 + 3+ [g] 26 Fe2 Fe BO5 0.66 0.02 3.93 < 0.01 0.21 [h] 27 Al6.9(BO3)(SiO4)3O2.5(OH)0.5 – <0.01 0.08 < 0.01 – 2 + [h] 28 NaFe Al6(BO3)Si6O18(OH)4 –0.04 0.41 < 0.01 – chemical gardens

29 ZnCl2 ––<0.01 (0.001) traces traces

30 FeCl2 ––<0.01 (0.008) –0.01

31 CuCl2 ––<0.01 (0.0002) – <0.01 (0.04)

32 Fe2(SO4)3 ––<0.01 (0.002) < 0.01 (0.0003) <0.01 (0.003)

33 MgSO4 ––<0.01 (0.001) < 0.01 (0.0001) <0.01 (0.0003)

34 ZnCl2 –traces traces 0.01 traces

35 FeCl2 –traces traces < 0.01 (0.005) traces

36 CuCl2 traces 0.08 <0.01 (0.004) –0.02

37 MnCl2 – <0.01 (0.002) ––<0.01 (0.0005)

38 Fe2(SO4)3 traces <0.01 (0.004) <0.01 (0.005)0.01 <0.01 (0.001)

39 MgSO4 traces <0.01 (0.003) <0.01 (0.005)0.01 <0.01 (0.006)

40 CuN2O6 traces 0.02 <0.01 (0.007) – <0.01 (0.003) silica vesicles

41 ZnCl2 0.02 0.10 0.46 0.007 0.04

42 FeCl2 –0.00016.6 0.00001–

43 CuCl2 –7.94 14.74 –5.0

44 MnCl2 2.24 11.62 11.90 0.000012.67

45 Fe2(SO4)3 5.8 29.08 6.28 4.64 1.22

46 MgSO4 2.8 41.35.26 3.82 0.42

47 CuN2O6 0.94 7.98 11.96 0.32 1.64

[a] Yield definedasmgofproduct per gram of starting formamide. [b] Highly pure commercially available Fe1 xS, in which0

acids, amino acids, etc.), the yield of each compound adds up tial further interactions during molecular evolution. Anyhow, to the amounts of the other compounds, favoring their poten- generaltrends are observed that require additional analysis

Chem. Eur.J.2019, 25,3181 –3189 www.chemeurj.org 3185 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review using advanced methods of multivariate statistical treatment cess of the texture of the membrane lasting more than one of experimental data. day.[37] Third, during the first half of that time, aconsiderable electrochemical potentialdifference across the tubular precipi- tates occurs, which enhances the catalytic properties of the 3. AGeological Setting for Formamide membrane.[40] Condensation 3.1. Mineral self-assembly 3.2. Serpentinization The solubility of silica in water strongly depends on pH.[32] It seems reasonable to expect that the loci for prebiotic From pH 2topH8.3, the solubility of silica is almostconstant chemistry in alifeless planetshould be close to the geological and varies from 120 ppm to 180 ppm depending on the solid settinginwhich organic compounds form from raw mineralre- phase at equilibrium, either quartz (or any of the other poly- actions.One of the few geochemical reactions known to morphsofsilicon oxide) or amorphous silica. Within that range create areduced environmentand to form molecular hydro- of pH, dissolved silica is present mainly as silicic acid. Above gen is the so-called serpentinization[41] (Figure 2). This reaction pH 8.3, the solubility of silica increases dramatically with pH is triggered when water interactswith olivine, amineral sili- due to the first and the second deprotonation of the silicic cate, which is asolid solution between two end members, acid. Thus, at pH 12, the concentration of silica could reach fayalite (the iron silicate) and forsterite (the magnesium sili- values about 4000 ppm. Interestingly,the precipitation of silica cate). The reactionproduces magnetite, molecular hydrogen from alkaline solutions may couple chemically with the precipi- and aclay mineral called serpentine [Eq. ()]: tation of metal carbonate and hydroxide creating self-assembled nanocrystalline materials with interesting physical and chemicalproperties.[33] We can distinguish two main regions of 6Mg1:5Fe0:5 SiO4 7H2O 3Mg3Si2O5 OH 4 Fe3O4 H2 ðÞþ seeped water ! ðÞþmagnetite þ chemicalcoupling. In the range of pH from 9.5 to 11.5, a olivine "#serpentine chemicalcoupling of silica and alkaline-earth carbonate precip- 1 ð Þ itation (specifically barium, strontium or calcium carbonate) occurs yielding aparticular type of nanocrystalline materials The reactionalso produces an environmentoflow oxygen that develop non-crystallographic morphologies through a fugacitythat favors the formation of carbon-based organic self-assembly mechanism. These materials, named “silica/car- compound. Thus, upon reaction of H2 and CO2,the so-called bonatebiomorphs”, closely resemble typical biologically in- Sabatier reaction produces methane (CH4). When the reaction duced mineral texturesand shapes, and have been proposed occursatrelatively high temperature (3008C) andpressure as an alternative explanation to some of the oldest putative (500 bars) several high-carbon number hydrocarbonsare pro- remnants of life.[34] At higher pH, when high concentrated duced by the so-called Fischer–Tropsch-type(FTT) synthesis,[42] silica solutions are mixed with metal solutions, atotally differ- which is catalyzed by magnetite and aweurite (Ni3Fe) [Eq. (2)]: ent phenomenon appears, namely the spontaneous formation of tubular structures known for many years as chemical gardens.[35] The tubulesare the results of the interaction of a chemicalreaction with osmotic and buoyantforces. Thewall of the tubulesconsists of ametal–[hydroxide/silica(te)] membrane[36] that separates two solutionswith fundamentally different compositions. The kinetics of the overall process varies depending on the type of metal cation used, essentially due to the different solubilities of the hydroxides,oxyhydroxides, or hydroxyanions forming during the precipitation of the tubular membranes.[37] The phenomenon of achemical gardenhas been unconvinc- ingly proposed almost two hundred years ago to be related with the origin of life[38] on the basis of its biomimetic appear- ance:tubular shapes, membrane formation,and osmotic forces. Nevertheless,three main characteristics of these structures made them particularly interesting in the framework of prebiotic chemistry.First, the membrane has atextural and Figure 2. 4400 Million years ago,water condensed on the first solid crust of the planet, which was made of peridotites and ultramafic rocks. The reaction compositional gradient, the outer part being made of aregion of water with the mineral olivine triggers serpentinization reaction at large rich in amorphous silica and the inner part made of nano- and scale and creates the reduced environment required for the formation of or- microcrystallinemetal oxyhydroxides. Second, while the forma- ganic compounds of geologicalorigin. Alkaline pH and subsequent enrich- tion of the tubular structures occurs almostimmediately after ment insilica allowsthe formation of mineralself-organized structures, such as silica/carbonate biomorphs andmineral vesicles, which are illustrated. mixing,there is an experimentally demonstrated diffusion of Scale bar of silica/carbonate biomorphs:25microns;Scale bar of mineral ions across the membrane[39] that triggers areorganization pro- vesicles:300 microns.

Chem. Eur.J.2019, 25,3181 –3189 www.chemeurj.org 3186 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review

FTTS CO2 aq 2 m=2n H2 1=n CnHm 2H2O 2 that the planet shouldalready have athin crust of ultramafic ðÞþ½ þ ðÞ Fe Ni!alloyðÞ þ À ð Þ rocks, that is, rocks mostly composed of olivine and pyroxenes. Therefore, the water that startedtocondense on the crust of Interestingly,intermediate stages of the serpentinization re- the planet 4.4 billions years ago would interact, mostly,with action produce brucite, amineral composed of magnesium hy- the rocks required to trigger serpentinization. Note that we are droxide that increases the pH and leads to alkaline aquatic en- talking of aserpentinization process at globalscale, at the vironments. Nowadays, watersknown to derive from serpenti- scale of the planet. The serpentinization reaction would release nizationprocesses may contain ahighconcentration of OHÀ at ahuge amount of hydrogen that upon reactionwith CO de- which pH reaches values higher than 12. In fact, serpentiniza- 2 gassed from the inner of the planetwould form methane tion is awidespread process and there are several places in through the Sabatier reaction. Thus, areduced atmosphere the world where it can be found at work. As shown by the would set in for along time until the productionofH be- above equations, serpentinizationdoes not releasesilica, be- 2 comes smaller than the flow of CO released by the inner cause the silica dissolvedfrom the olivine at high pH values 2 core.[48] Under these reduced conditions formamide would will be as silicate species and will quicklyprecipitate as serpen- easily form from HCN. As the serpentinizationprocess pro- tine. However,when high pH watersderived from serpentiniza- gresses, it would provoke adistinct mechanical difference be- tion infiltrate through silica-rich rocks, such as granite, they will tween the ultramafic rocks and the clay minerals formed produce silica-rich solutions. Oneofthe contemporary sites during the reaction. This in turn would facilitaterock alteration, where serpentinization waters are enriched in silica is located triggering crack formation and the infiltrationofwater to in the Cascades Range, in the contact between the trinity deeper and hotterrocks, which could be the origin of the very ophiolite complex and atertiary basaltic–andesiticvolcanic de- first and simplest hydrothermal systems. The chemistry of posits, in Northern California. In these mountains there are 1 these hydrothermalwaterswould be basically derived from some springs with waters of pH 12 and up to 4grLÀ of fluid/olivine interactions at temperatures hot enough for Fisch- silica.[43] We collected waters from Aqua de Ney spring (Siskiyou er–Tropsch-type synthesis. Therefore, we can envisage the ear- country,California,USA) and demonstrated that mixing these liest crust of the planetatthat time as ahuge factory recycling waters with metallic solutions can produce all the three main an extremelyalkaline fluid full of organic compounds. It is im- types of mineral self-assembly induced by silica. These experi- portanttonote that contemporary serpentinization is framed ments confirmthat silica/carbonate biomorphs, silica/metal by an atmospheric compositionrich in CO that buffers the pH oxyhydroxide mineral vesicles, and calcite mesocrystals with 2 of serpentinization waters as soonasthey reach the surface. biomimetictextures are geochemically plausible phenom- An atmosphere rich in methane stabilizes high pH values de- ena.[44] rived from serpentinizationreaction and makes the kinetics of silica releasefaster than the production of clays. Further rock 3.3. Serpentinization in the earliest earth differentiation would connectthese waters with mineralssuch Today,there are few geological sites with alkaline water and a as quartz and plagioclase that would be dissolved by the high silica concentration higherthan the one required to trigger pH alkaline waters from serpentinization, increasing their con- self-organization, that is, 300 ppm forsilica biomorphs and bio- centration in silica. Then, the chemical conditions for the for- mimeticcalcite and 900 ppm for mineral vesicles, depending mation of biomorphs (that requirelower pH values and smaller on pH, salinity and temperature.[33b] In addition to the springs silica concentrations) but also of metal silicate hydrate mem- in the Cascades Range mentioned above, most of the so-called branesand mineral vesicles would be available.[49] Fatty acids, soda lakes in the Rift Valley fulfill these conditions.[45] Even polycyclic aromatic hydrocarbons, and other organic com- thoughwehave to admit that the geological scenarios for the pounds pumped by the very first hydrothermal vents would in- chemicalcoupling of silica with carbonate and hydroxides are teract with the product of the condensation of NH2CHO. We today rare and bizarre,these alkaline geochemical scenarios predict that the textural properties of the membranes forming could have been widespreadduring the first billion years of self-assembled mineral vesicles and the small but continuous [46] this planet. energy generated during their formation will catalyze NH2CHO The earth is 4.5 billion years old. The first five hundred mil- condensation and carbon-based reactions to form organic lion years are called the Hadeanera, the Greek word for the compounds of increasing diversity and complexity. hell, because it wasthought until recently that during that time the planetwas full of volcanic activity,underwent huge 3.4. Formamide condensation enhancedbymineralvesicles meteoricbombardment and high ultraviolet radiation, and had avery hot surface, that is, conditions that would not only steri- To prove the proposed catalytic effect of silica-inducedmineral lize the planet, but also hinder any prebiotic relevant chemis- self-organization, we have tested two experimental set-ups. try.However,recent studies have shown that water condensed First we used classical chemical gardens made by inserting pel- on the rocks of the crust 4.4 Ga ago, that is, just after the for- lets of soluble metallic salts within sodium silicate solutions at mation of the moon.[47] At such an early time, it is currently pH 12.[50] The osmotic-driven membranous structures catalyzed thought that the planet was stillinthe first stage of differentia- the condensation of NH2CHO to yield nucleobases, amino tion of asolid crust.[42a] Thereare no rock remnants from that acids, and carboxylic acids. Membranes formed from sodium earliest earth, but geochemicaland petrological models sustain silicatesolutions in the presence of magnesium and iron salts

Chem. Eur.J.2019, 25,3181 –3189 www.chemeurj.org 3187 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review

[MgSO4 and Fe2(SO4)3·9H2O] showedthe highest efficiency However,the mechanism by which some of the molecules of with respecttocopper-, zinc- and manganese-based com- the organic mixture self-assembled tocreate complex mole- pounds (CuCl2·2H2O, ZnCl2,and MnCl2·4H2O). It is worth noting cules able to self-reproduce is stillunknown. The geological that the collections of compounds formed insideand outside niche proposed here for the transitionfrom inorganic to or- the membrane were clearly site-specific, nucleobases prevail- ganic geochemistry wassettled most likely as early as 4.4 Ga, ing inside and carboxylic acids prevailingoutside, demonstrat- that is, almost one billion years earlier than the oldest putative ing that the mineral self-assembled membranes simultaneously remnantsoflifeonour planet. Once more, time shows to be provide space compartmentalization and selectivecatalysis of again akey parameter for origin of life studies. the synthesis of relevant compounds. In asecond set of experi- [51] ments we used mineral vesicles as catalysts. The micro vesi- Acknowledgements cles were synthesized by mixing microdrops of asolutionof the metallicsalts with solution of sodiumsilicate.They can be We acknowledge funding from the European Research Council made either by pouring the drop on top of the silicate solution under the European Union’s Seventh Framework Programme or immersing the drop into the silicate solutions. In both cases (FP7/2007–2013)/European Research Council grant agreement metal/silicate membranes formed at the interfaceofthe acidic no. 340863(Prometheus). Spanish Ministerio de Economia y or neutral drop of the metallic solution and the surrounding al- Competitividad is acknowledged for Project CGL2016-78971-P, kaline silicate solution. Then, after few minutes, the initial AEI/FEDER. MIUR Ministero dell’Istruzione, dell’Università della quasi-spherical membranous vesicle starts to corrugate and it Ricerca and Scuola Normale Superiore (Pisa, Italy), project PRIN bursts as aresult of the osmotic forces. The vesicle is observed 2015 STARS in the CAOS—SimulationTools for Astrochemical to break expelling ajet of the inner solution into the outer sili- Reactivityand Spectroscopy in the Cyberinfrastructure for As- cate solution, creating new membranousbulbs or tubulesthat trochemical OrganicSpecies, cod. 2015F59J3R,isacknowl- severelymodified their shape. Thus, NH CHO condensation 2 edged.This work was supported by COST Action TD 1308. occurs during the time at which the membrane formation and reorganization is going on. Our experiments clearly show that the four nucleobases required forRNA synthesis,three amino Conflict of interest acids (glycine, alanine, and N-formylglycine), and six carboxylic acids can be synthesized from NH2CHO in asingle geochemical The authors declare no conflict of interest. scenario. In particular, the condensation of NH2CHO is catalyzed during the formation of silica-oxyhydroxidemembranes Keywords: biomorphs · formamide · geochemistry · origin of of different metals (iron,magnesium, manganese,and copper) life · prebiotic chemistry · serpentinization that are common components of the ultramafic and komaititic rocks of the earliest crust of the planet, able to interact with [1] R. H. Rubin,G.W.Swenson,R.C.Benson, H. L. Tigelaar,W.H.Flygare, As- water to start the serpentinization process.[42a] Moreover,min- trophys. J. 1971, 169,L39. eral vesicles are more efficient in term of both yieldingand di- [2] J. E. Geach, P. P. Papadopoulos, Astrophys. J. 2012, 757,156. versity than classical chemical gardens. This is due to several [3] G. R. Adande, N. J. Woolf, L. M. Ziurys, Astrobiology 2013, 13,439 –453. [4] N. Biver,D.BockelØe-Morvan, V. Debout, J. Crovisier, J. Boissier,D.C.Lis, factors, but mostly to the fact that they wereactive during N. D. Russo, R. Moreno,P.Colom,G.Paubert, R. Vervack, H. A. Weaver, NH2CHO condensation and that the crystallinity of the mem- Astron.Astrophys. 2014, 566,L5. brane of mineral vesicles—as seen by X-ray powder diffrac- [5] A. López-Sepulcre, A. A. Jaber,E.Mendoza, B. Lefloch, C. Ceccarelli, C. tion—isdemonstrated to be smaller than the crystallinity of Vastel, R. Bachiller, J. Cernicaro,C.Codella, C. Kahane, M. Kama, M. Tafal- la, Mon. Not. R. Astron. Soc. 2015, 449,2438–2458. the garden membranes, thus increasingcatalytic efficiency. [6] V. S. Airapetian, A. Glocer,G.Gronoff, E. HØbrard, W. Danchi, Nat. Geosci. 2016, 9,452 –455. [7] S. L. Miller, H. Urey, Science 1959, 130,245 –251. 4. Outlook [8] C. Chyba, C. Sagan, Nature 1992, 355,125 –132. [9] S. A. Benner,A.Ricardo,M.A.Carrigan, Curr.Opin. Chem. Biol. 2004, 8, The main conclusion of our ongoing investigation is that the 672–689. chemicalmolecules that we know are the bricks of the life can [10] S. A. Benner, Origins LifeEvol. Biospheres 2014, 44,339 –343. be easily formed under geological conditions that were widely [11] P. Schmitt-Kopplin, Z. Gabelica, R. D. Gougeon, A. Fekete, B. Kanawati, distributed in the primitive earth. We propose that the crust of M. Harir,I.Gebefuegi, G. Eckel, N. Hertkorn, Proc. Natl. Acad.Sci. USA 2010, 107,2763 –2768. the early earth was afactoryoforganic compounds working at [12] A. S. Burton, J. C. Stern, J. E. Elsila, D. P. Glavin, J. P. Dworkin, Chem. Soc. planetary scale, as much as 4.4 billion years ago. This geo- Rev. 2012, 41,5459 –5472. chemicalniche should also exist, or have existed, on earth-like [13] E. Carota, G. Botta, L. Rotelli, E. Di Mauro, R. Saladino, Curr.Org. Chem. planets (on which water condensed after the first differentia- 2015, 19,1963 –1979. [14] R. Saladino, G. Botta, S. Pino, G. Costanzo, E. Di Mauro, Chem.Soc. Rev. tion of an ultramafic crust),onmeteorite parentbodies, on 2012, 41,5526 –5565. comets, and on the interstellar dusts made of olivine that are [15] M. Ferus, F. Pietrucci, A. M. Saitta,A.Knizek, P. Kubelik, O. Wanek, V. and were in contact with enriched regionsofformamide in the Shestivskµ„S.Civisˇ, Proc. Natl. Acad. Sci. USA 2017, 114,4306 –4311. ˇ ˇ universe. Thus, we claim that the transition from inorganic ge- [16] M. Ferus, D. Nesvorny´,J.Sponer,P.Kubelík, R. Michalcíkovµ,V.Shes- tivskµ,J.E.Sˇponer,S.Civisˇ, Proc. Natl. Acad.Sci. USA 2015, 112,657 – ochemistry to organic geochemistry and even to prebiotic ge- 662. ochemistry was an unavoidable and universal phenomenon. [17] J. Kua, K.L. Thrush, J. Phys. Chem. B 2016, 120,8175 –8185.

Chem. Eur.J.2019, 25,3181 –3189 www.chemeurj.org 3188 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim Review

[18] F. Pietrucci, A. M. Saitta, Proc. Natl. Acad. Sci. USA 2015, 112,15030 – [38] a) S. Leduc in The Mechanism of Life,Rebman, London, 1911;b)A.L. 15035. Herrera, Science 1942, 96,14; c) H. J. Cleaves, A. Lazcano, I. Ledesma [19] A. M. Saitta, F. Saija, Proc. Natl. Acad. Sci. USA 2014, 111,13768 –13773. Mateos,A.Negrón-Mendoza, J. Peretó,E.Silva, Herrera’s in Plasmogenia, [20] R. Saladino, C. Crestini, S. Pino, G. Costanzo, E. Di Mauro, Phys. Life Rev. Collected Works,Springer,New York, 2014. 2012, 9,84–104. [39] F. Glaab,J.Rieder,J.M.García-Ruiz, W. Kunz, M. Kellermeier, Phys. Chem. [21] D. Niether,D.Afanasenkan, J. K. G. Dhont, S. Wiegand, Proc. Natl. Acad. Chem. Phys. 2016, 18,24850 –24858. Sci. USA 2016, 113,4272–4277. [40] a) F. Glaab, M. Kellermeier,W.Kunz,E.Morallon, J. M. García-Ruiz, [22] D. Niether,S.Wiegand, Entropy 2017, 19,33. Angew.Chem. Int. Ed. 2012, 51,4317 –4321; Angew.Chem. 2012, 124, [23] S. A. Benner,H.-J. Kim, M. A. Carrigan, Acc.Chem. Res. 2012, 45,2025 – 4393 –4397;b)L.M.Barge, Y. Abedian, M. J. Russell, I. J. Doloboff, J. H. E. 2034. Cartwright,R.D.Kidd,I.Kanik, Angew.Chem. Int. Ed. 2015, 54,8184 – [24] R. Saladino, E. Carota, G. Botta, M. Kapralov,G.N.Timoshenko, A. Y. Ro- 8187; Angew.Chem. 2015, 127,8302 –8305. zanov,E.Krasavin, E. Di Mauro, Proc. Natl. Acad. Sci. USA 2015, 112, [41] a) N. H. Sleep,A.Meibom,T.Fridriksson, R. G. Coleman, D. K. Bird, Proc. E2746 –E2755. Natl. Acad. Sci. USA 2004, 101,12818 –12823;b)A.Neubeck,N.T.Duc, [25] R. Saladino, B. M. Bizzarri, L. Botta, J. Sˇponer,J.E.Sˇponer,T.Georgelin, D. Bastviken,P.Crill,N.G.Holm, Geochem. Trans. 2011, 12,6. M. Jaber,B.Rigaud, M. Kapralov,G.N.Timoshenko, A. Rozanov,E.Krasa- [42] a) D. I. Foustoukos, W. E. Seyfried, Science 2004, 304,1002 –1005; vin, A. M. Timperio, E. Di Mauro, Sci. Rep. 2017, 7,14709. b) M. E. Berndt, D. E. Allen, W. E. Seyfried,Jr. , Geology 1996, 24,351 – [26] R. Saladino, V. Neri, C. Crestini, G. Costanzo, M. Graciotti, E. Di Mauro, J. 354;c)J.Horita, M. E. Berndt, Science 1999, 285,1055 –1057;d)F.Klein, Mol. Evol. 2010, 71,100 –110. W. Bach, N. Jçns, T. McCollom,B.Moskowitz, Geochim. Cosmochim. Acta [27] R. Saladino, G. Botta, M. Delfino, E. Di Mauro, Chem. Eur.J.2013, 19, 2009, 73,6868 –6893. 16916 –16922. [43] a) J. H. Feth, S. M. Roogers, C. E. Roberson, Geochim. Cosmochim. Acta [28] R. Saladino, C. Crestini, V. Neri, J. R. Brucato, L. Colangeli, F. Ciciriello,E. 1961, 22,75–76;b)T.Boschetti, L. Toscani, P. Iacumin, E. Selmo, Aquat. Di Mauro, G. Costanzo, ChemBioChem 2005, 6,1368–1374. Geochem. 2017, 23,299–313. [29] R. Saladino, V. Neri, C. Crestini, G. Costanzo, M. Graciotti, E. Di Mauro, J. [44] J. M. García-Ruiz, E. Nakouzi, E. Kotopoulou,L.Tamborrino, O. Steinbock, Am. Chem. Soc. 2008, 130,15512–15518. Sci. Adv. 2017, 3,e1602285. [30] R. Saladino, M. Barontini, C. Cossetti, E. Di Mauro,C.Crestini, Origins Life [45] J. M. García-Ruiz, Geology 1998, 26,843–846. Evol. Biospheres 2011, 41,317 –330. [46] “Carbonate Sedimentation and Diagenesis in the Evolving Precambrian [31] L. Rotelli, J. M. Trigo-Rodríguez,C.E.Moyano-Cambero, E. Carota, L. World”: J. M. García-Ruiz in SEPM Special Publication, Vol. 67 (Eds.:J.P. Botta, E. Di Mauro, R. Saladino, Sci. Rep. 2016, 6,38888. Grotzinger,N.P.James), Society for Sedimentary Geology, Tulsa, 2000, [32] a) G. B. Alexander,W.Heston,R.K.Iler, J. Phys. Chem. 1954, 58,453 – pp. 75 –89. 455;b)H.E.Bergna in The Colloid Chemistry of Silica, Vol. 234,American [47] O. Nebel, R. P. Rapp, G. M. Yaxley, Lithos 2014, 190,313–327. Chemical Society, Washington, DC, 1994,p.724. [48] N. G. Holm, E. M. Andersson in The Molecular OriginsofLife:Assembling [33] a) M. Kellermeier,H.Cçlfen, J. M. García-Ruiz, Eur.J.Inorg. Chem. 2012, Pieces of the Puzzle (Ed.:A.Brack), Cambridge University Press, Cam- 5123 –5144;b)M.Kellermeier,F.Glaab,E.Melero-Garcia, aJ.M.García- bridge, 1998,pp. 86–99. Ruiz in Methods in EnzymologyVol. 532 (Ed.:J.J.DeYoreo),Elsevier, [49] H. Satoh, K. Tsukamoto, J. M. García-Ruiz, Eur.J.Mineral. 2014, 26,415 – Oxford, 2013,p.45; c) E. Nakouzi,O.Steinbock, Sci. Adv. 2016, 2, 426. e1601144. [50] R. Saladino, G. Botta, B. M. Bizzarri, E. Di Mauro, J. M. García-Ruiz, Bio- [34] J. M. García-Ruiz, S. T. Hyde,A.M.Carnerup, A. G. Christy,M.J.Van Kra- chemistry 2016, 55,2806 –2811. nendonk, N. J. Welham, Science 2003, 302,1194–1197. [51] B. M. Bizarri, L. Botta, M. I. PØrez-Valverde, R. Saladino, E. DiMauro, J. M. [35] a) T. H. Hazlehurst, J. Chem.Educ. 1941, 18,286 –289;b)J.M.García- García-Ruiz, Chem. Eur.J.2018, 24,8126 –8132. Ruiz, A. Carnerup, A. G. Christy,N.J.Welham, S. T. Hyde, Astrobiology 2002, 2,353 –369. [36] J. J. Pagano, S. Thouvenel-Romans,O.Steinbock, Phys. Chem. Chem. Manuscript received:July 30, 2018 Phys. 2007, 9,110–116. Revised manuscript received:September 14, 2018 [37] F. Glaab, J. Rieder,R.Klein, D. Choquesillo-Lazarte, E. Melero-Garcia, J. M. García-Ruiz, W. Kunz,M.Kellermeier, ChemPhysChem 2017, 18,338 – Accepted manuscript online:September 19, 2018 345. Version of record online:December 27, 2018

Chem. Eur.J.2019, 25,3181 –3189 www.chemeurj.org 3189 2019 The Authors. Published by Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim