Metals in Prebiotic Catalysis: A Possible Evolutionary Pathway for the Emergence of Metalloproteins

Abstract

Proteinaceous catalysts found in extant biology are products of life that were potentially derived through prolonged periods of evolution. Given their complexity, it is reasonable to assume that they were not accessible to prebiotic chemistry as such. Nevertheless, the dependence of many enzymes on metal ions or metal–ligand cores suggests that catalysis relevant to biology could also be possible with just the metal centers. Given their availability on the Hadean/Archean Earth, it is fair to conjecture that metal ions could have constituted the first forms of catalysts. A slow increase of complexity that was facilitated through the provision of organic ligands and amino acids/peptides possibly allowed for further evolution and diversification, eventually demarcating them into specific functions. Herein, we summarize some key experimental developments and observations that support the possible roles of metal catalysts in shaping the origins of life. Further, we also discuss how they could have evolved into modern-day enzymes, with some suggestions for what could be the imminent next steps that researchers can pursue, to delineate the putative sequence of catalyst evolution during the early stages of life.

Introduction: Starting Simple to Kickstart Life

Life, as we know it, is an out-of-equilibrium system¹ maintained by a huge supporting network of physical and chemical processes.² However, the processes of life are only a small subset of all possibilities allowed by the laws of physical chemistry.³ This selectivity and out-of-equilibrium character necessitates the continuous supply of energy from the environment and mechanisms to accelerate specific processes or reactions over the others.^1,3,5 This will maintain an unmistakable identity for the constituents of a living system, in terms of the molecules and their concentrations, and their physical/chemical organization. How does life achieve this?

Catalysis is life’s answer to protection from thermodynamic equilibrium or death, and in extant biology this is achieved mostly through proteins, which are highly sequence-specific polymers.² For example, ATP hydrolysis is progress toward a thermodynamic equilibrium, but the cellular machinery acts against it and continuously regenerates ATP by utilizing reduced carbon substrates. This process is dependent on an extensive metabolic network that relies on enzymes acting in concert, with the product of one enzyme being utilized as the substrate for a downstream reaction. The modularity of the metabolic network allows channeling the reducing equivalents obtained from different sources into generating a pH gradient across the membrane, which ultimately drives the regeneration of ATP. For life to have originated on the prebiotic earth, catalysis would have been necessary to accelerate specific reaction networks out of a vast background of possibilities.^1,5 However, given protein synthesis’ dependence in extant biology on an elaborate translation machinery,⁵ proteins may not have been the catalysts that would have kick-started the origins of life itself. Life perhaps started off simpler and later, through evolutionary changes, reinventing its catalyst repertoire. Alternatively, this could have been facilitated by building on prebiotic catalysts, to subsequently make highly specialized modern proteins. Conservation of protein structures in the core enzymes of cellular metabolism,⁶ redundancy seen in pathways,⁷ and also the promiscuity of enzymes toward substrates⁸ possibly supports the latter argument. What then could have been these relatively simpler catalysts?

It is now well-known that nearly one-third of all enzymes^3,9—and perhaps all enzymes of the core metabolism (C/N-fixation or bioenergetics)—have an obligatory dependence on metals, with the metal center being catalytically active and directly involved in catalysis. Nitrogenases and hydrogenases are common examples of such enzymes. Given this, could metal centers have catalyzed reactions by themselves or with the support of minimalistic ligand environments to form the biomolecular inventory, starting with CO₂ and abiotic nitrogen sources? What properties or underlying principles could have guided their selection into different roles in extant biology? What pathway(s) in evolution might have guided their development into the modern proteins?

Given their ubiquitous nature, we are inclined to believe that metal ions and minerals that were available on the early Earth were the first catalysts. They were potentially superseded where necessary and possible, by metal–ligand complexes and later scaffolded by short peptide motifs (Figure 1¹⁰⁻¹³). These steps probably laid the groundwork for the evolution of complex proteins with highly specific functions, the kinds that we know in extant biology.

Figure 1.

Possible pathway for the emergence of proteinaceous catalysts. Metal ions and minerals (not shown) were readily available and provided by the prebiotic earth to catalyze the earliest reaction (networks). As organic molecules slowly accumulated over time, they could coordinate with available metal centers to result in metal–ligand based cofactors—for example, cysteine or other thiols may have allowed the formation of iron–sulfur clusters. Over time, small peptides evolved which could bind the metal centers. Finally, these peptides and peptide motifs may have been the basis for the development of modern proteins, chiral and sequence-specific. The iron–sulfur clusters, the metallopeptide segment, and the metalloprotein depicted belong to bacterial ferredoxin (PDB ID 1H98,^10,11 protein structure visualized using UCSF ChimeraX v1.5rc202210241843^12,13).

Experiments in prebiotic chemistry and organometallic chemistry¹⁴ alike have shown the potential of metal ions and metal–ligand complexes in catalyzing a vast array of reactions, making them a near-indispensable part in these areas of research. These experiments and the importance of metalloenzymes to life have together motivated increasing research on the role of metals in setting up biochemistry from prebiotic chemistry. In this review, we try to summarize developments in the aforementioned aspects, while also trying to rationalize the role of metals. We have focused on key examples from prebiotic chemistry research, to highlight the developments and possibilities in studying biocatalyst evolution. Some suggestions for what areas are incomplete and what could be explored further are laid along the way, with the hope of motivating research targeted into the suggested areas.

Role of Metals and Metal Ions

Even though metal ions represent the lowest rung in the ladder of complexity of catalysts, they are also the rational choices, as they are simpler (minimal) alternatives for a plethora of proteinaceous catalysts. The Archean ocean is believed to have been enriched in different metal ions, particularly Fe²⁺,¹⁵ as the planet was far more reducing then.^16,17 While there were other metal ions as well, the high abundance of Fe²⁺ on the prebiotic earth, alongside their ability to switch by one oxidation state, perhaps explains their prevalence in extant metalloproteins. In addition to this, Mg²⁺, Mn²⁺, Co²⁺, Ni²⁺, Cu²⁺, and Zn²⁺ are all metal ions very often found in enzymes of extant biology⁹ and were part of the early Earth niches.¹⁵ What catalysis does one expect from these species?

The RNA world hypothesis for the origins of life posits the existence of a simultaneously catalytic and genetic RNA molecule. Mg²⁺ ions have been shown to be essential in almost every stage of the formation and functioning of such a molecule. Mg²⁺ not only coordinates with and stabilizes ATP and ADP but also facilitates nucleophilic attack on the α-phosphate, leading to polymerization.¹⁸ Ribozymes, the conjectured pioneer catalysts of the RNA world, rely on Mg²⁺ to stabilize the folded and catalytically active structure. In ligases, which catalyze formation of longer RNAs¹⁹ as well as self-cleaving ribozymes,²⁰ Mg²⁺ also participates directly in the transition state stabilization. Such importance of Mg²⁺ to the RNA world is reminiscent of extant biology and its role in DNA/RNA/ribosome processes.^2,5 These functions are attributed to the size of Mg²⁺ that allows an optimum in the coordination parameters with the oxyanions on the phosphate group.¹⁸ Given its centrality, is Mg²⁺ irreplaceable in its role then? Interestingly, density functional calculations show similarity in RNA binding with Mg²⁺ and Fe²⁺, and experiments have demonstrated cases of enhanced activity of ribozymes with Fe²⁺ in comparison to Mg²⁺.²¹ Why might have biology then chosen Mg²⁺ over Fe²⁺?

As the oxygen fugacity of the Earth increased with time, Fe²⁺ decreased in abundance so that Mg²⁺ remained the possible alternative.^16,17 Yet, a rigorous understanding of why Mg²⁺ got selected for RNA catalysis is unclear, as there is insufficient literature comparing the performance of other metal ions in similar roles. The answers are possibly related to the binding affinities of metal ions toward different ligand groups and the resulting coordination geometry. A combination of experiments and computation, as done by Athavale et al.,²¹ might explain the rationale in the choices of biological catalysis.

Calcium ions, particularly the hydroxylated species CaOH⁺, catalyze the formose reaction that produces a whole array of sugars starting with formaldehyde.²² On the downside, the same metal ion also leads to the decomposition of the sugar products.²³ However, in a prebiotic chemistry setting, these problems might have been mitigated by stabilization through silicate²⁴ and borate²⁵ ions, as has been previously reported. Fe²⁺ and Fe³⁺ are shown to expand and accelerate an otherwise less-diverse and slow sugar–phosphate reaction network, bringing semblance to the biological pentose phosphate (Scheme 1) and glycolysis networks.²⁶ Interestingly, parasitic reactions are minimized and selectivity for ribose-5-phosphate, which is the precursor to RNA/DNA synthesis, is enhanced by Fe²⁺ in comparison to Fe³⁺. A strong pH dependence of these catalyzed reactions which translates to different populations of Fe²⁺ and Fe³⁺ also implies a stark effect of metal oxidation states on these reactions.²⁷

Scheme 1. General Network of the Pentose Phosphate Pathway, as Depicted in Keller et al.²⁶.

The network as a whole potentially comprises many reaction types, leaving room for mechanistic exploration that can throw light on some unexplored possibilities for metal-ion catalysis. Schematic made using ChemDraw Professional v21.0.0.28.

The acetyl-CoA (Ac-CoA) pathway²⁸ and rTCA cycle⁷ (Figure 2²⁹) are posited to be the most ancient carbon-fixation routes and have been argued to be the foundation stones for the origins of life.^8,30,31 Enzymes in extant biology that are involved in these pathways are heavily metal dependent,³² which has added to them being considered as very ancient pathways, besides motivating experimental research in this direction.²⁹ Metallic iron, cobalt, nickel, molybdenum, and tungsten, as their respective nanoparticles, are able to reduce CO₂ to reduced organic molecules.³³ These organic molecules mostly belong to the Acetyl-CoA pathway, with the product repertoire being dependent on the metal and the reaction conditions involved. The oxidized states of these metals are found at the cores of the Ac-CoA pathway enzymes, and akin to the biological pathway, the intermediates are suggested to be metal-center bound.³³ The reactions likely proceed through the reduction of the organic species using electrons from the metal center (suggested to also involve hydrogen production). The electron transfer is thought to be facilitated by surface attachment of the organic intermediates.

Figure 2.

(a) Acetyl-CoA pathway depicted with the general scheme and metalloenzymes involved. Apparently this pathway is heavily dependent on metalloenzymes, one of the reasons it is posited to be among the most ancient carbon fixation pathways.²⁹ (b) General mechanism of the rTCA cycle showing all the intermediates and the inputs at various steps. Metals come in handy, particularly in conversion of acetyl-CoA to pyruvate and succinyl-CoA to α-ketoglutarate, where iron–sulfur clusters provide electrons. Schematic made using ChemDraw Professional v21.0.0.28, based on the depiction in ref (29).

Analyzing such results in light of the free energies of the formation of different organic molecules and possible transition states, and the redox potentials of the metals involved, may provide an understanding of the partitioning of different metal ions into enzymes catalyzing the different steps.

Utilizing metallic Fe in combination with Zn²⁺ and Cr³⁺, Moran’s group achieved all the reduction–dehydration–hydration reactions relevant to the rTCA cycle, which represent two of three reaction sequences.³⁴ What is interesting is that the catalysts maintain selectivity of the products within the cycle. Parasitic reactions leading to off-cycle dead ends are possible on multiple intermediates; however, the catalyst is selective to the reactions of the cycle. The suggested mechanism points to the metal ions perhaps acting as Lewis acids, which coordinate the negative charge centers of the substrate and polarize bonds to facilitate dehydration–rehydration steps. Alternatively, Ni⁰ (as well as Rh⁰) on a SiO₂/Al₂O₃ support is able to drive three reactions of the rTCA cycle: the conversion of oxaloacetate to malate, malate to fumarate, and fumarate to succinate.³⁵ Two of these are reduction reactions, and the role of Ni here is similar to its popular use as a hydrogenation catalyst. Utilizing glyoxylate and pyruvate as starting materials in the presence of Fe⁰/Fe²⁺, the same group recreated an extensive metabolic network encompassing all five universal precursors of biosynthesis, and a variety of reaction types.³⁶ While unoxidized metals are generally not considered for prebiotic chemistry, they are nonetheless known to form in alkaline hydrothermal vent-like systems³⁷ under extreme reducing conditions. Although the all-important carbon-fixation steps in the rTCA cycle are metal-dependent, these have currently not been recreated nonenzymatically in a convincing way. A major challenge here is perhaps the energetics involved in the carboxylation steps or requirement of cofactors like thiamine, for which the primordial alternatives have not yet been worked out. Two of these steps depend on high-potential electrons from ferredoxins and the other two steps depend on ATP.

Are high-energy molecules such as ATP plausible for synthesis in a prebiotic environment? Fe³⁺ has been shown to promote the formation of ATP from ADP and a high-energy acetyl-phosphate molecule.³⁸ The ions, however, remain potentially coordinated to the product, precluding catalytic cycles by themselves. Addition of Mg²⁺ or Ca²⁺ can alleviate this problem by liberating Fe³⁺ from coordination with the ATP and making it available for subsequent cycles.³⁹ Computational studies have shown the coordination of Fe³⁺ with the β-phosphate and acetyl-phosphate during catalysis. NADH is another high-energy molecule used in biological redox, generated using hydride transfers from reduced substrates.⁴⁰ Recently, it was also shown that native metals (Fe, Co, and Ni) are capable of performing hydride transfers to NAD⁺ to generate NADH.⁴¹ While Fe and Co act as “proto-hydrogenases” as well as reductants themselves, Ni acts exclusively as a hydrogenase, transferring hydrides without undergoing a redox change itself.

Various metal ions are known to drive reactions of interest to amino acid metabolism and nucleobase synthesis. Cu²⁺, Ni²⁺, Co²⁺, and V⁵⁺ have been shown to be capable of driving transamination reactions between different amino acid and α-keto acid substrates.⁴² Even though the suggested intermediates bear some semblance to those in extant biology, the role of metal ions is starkly different from that of pyridoxal phosphate.⁴³ The metal ions act exclusively as catalysts in these reactions, not undergoing redox changes themselves. In contrast, the reports utilizing Fe⁰ potentially involved the metal center as a source of electrons to drive the reductive transamination.^34,36 Stubbs et al. previously also demonstrated that Al³⁺ catalyzes the reductive transamination of oxoglutarate using glycine, producing glutamate.⁴⁴ Metal ions such as VO²⁺ and VO₃^– also assist in the transamination of α-keto acids from organic amine donors,⁴⁵ potentially implying the positive role of metals in supporting the origins of life even after the evolution of the first organic cofactors.

Toward nucleobase synthesis, which is important for the formation of RNA and many cofactors, Cu²⁺ catalyzes conversion of carbamoyl aspartate to orotate and uracil.⁴⁶ This happens via dihydroorotate as in biology (or hydantoin-5-acetate), utilizing Mn⁴⁺ as an oxidant, although Cu²⁺ is detrimental to the stability of carbamoyl phosphate, one of the precursors to carbamoyl aspartate.

What stands out through these examples is that metal ions can catalyze diverse reaction types, with some patterns of functional segregation being apparent in some cases. While a great deal of research has gone into characterizing the possible role of metal ions in prebiotic catalysis, a major lacuna persists in discerning the specific mechanisms involved in catalysis, which have yet to be fully explored. This, we feel, will potentially allow developing a novel hypothesis on the possible role of metal ions as catalysts or substrates in hitherto unexplored reactions. Nevertheless, catalysis with metal ions alone is limited by their core properties and their solvent environment, both of which are not amenable to direct changes in the context of prebiotic chemistry. Particularly, in their native state, metals are also experimentally observed to lack the property of catalysis in the strict sense.⁴⁷ They serve as reagents (specifically, as reductants) and are consumed over the course of reactions. Their role is generally found in redox reaction catalysis, wherein they undergo oxidation. Interestingly, in synthetic organic chemistry, there are known examples wherein native metals provide robust catalysis. Environmental reductants able to reduce metal ions like Fe²⁺ into their native state (Fe⁰) are practically unknown. From a biological standpoint, or from the perspective of protocell evolution, this poses a problem, as continuous replenishment of metals is impractical. In biology, the metal ions in catalysis undergo cyclical redox changes so that a large supply of native metals is avoided. Combined with the issue of how such native metals can be solubilized and mobilized, we see a clear requirement for an alternative to be necessary at least for native metals. How then did the functionality and availability of metal-based catalysts increase during evolution?

Minerals in Prebiotic Catalysis

Besides the widespread occurrence of different mineral species in almost all geochemical niches, why should one expect something more novel from minerals given that metal ions already bring in a diverse array of catalysis? Metal ions by themselves have limited (albeit widely applicable) properties. However, due to the presence of coprecipitating ions and multiple cationic species in the same mineral, we think that minerals explore a larger space of catalytic possibilities by bringing multiple properties and characteristics to the same catalytic center. Additionally, mineral catalysis may often work through adsorption-interaction, allowing increased local concentrations of substrates to improve reaction kinetics and selectivity. In some cases, hypothesized mechanisms also suggest that organic intermediates bound to the metal center akin to organometallic structures allow relatively facile group transfers. Not surprisingly, many origins-of-life experiments have shown the production of biomolecules and recapitulation of biological pathways through the use of minerals.⁴⁸

Hydrothermal minerals like magnetite/hematite/awaruite, whose structures bear semblance to enzymatic metal centers, have been shown to catalyze CO₂ fixation into some end products of the Ac-CoA pathway/methanogenesis, using molecular hydrogen.⁴⁹ FeNiS minerals that are suggested to have been formed in hydrothermal vents⁵⁰ are capable of producing high-energy thioesters analogous to the ancient Ac-CoA pathway.⁵¹ Thioesters are high-energy molecules able to drive ATP generation downstream⁵² or participate in further biosynthesis, as in lipid synthesis² or reductive carboxylation.²⁹ Interestingly, FeS alone is incapable of the aforementioned catalysis while NiS is catalytically active; the coprecipitated mineral has been shown to have higher catalytic activity. The requirement of Ni centers, while not explored mechanistically here, might potentially be explained by drawing parallels to extant biology,³² where the Ni center in CODH/ACS enzyme acts to transfer the methyl group to the CO and undergoes a 2e^– redox cycle. The 2e^– redox might be less feasible with Fe, leading to this selectivity.

Alternatively, the same FeNiS mineral, when sustaining a pH gradient across its walls, has been shown to catalyze the fixation of CO₂ into formate by enabling a transmembrane electron transport, albeit requiring large pH gradients.⁵³ The mineral acts like an electrical wire to channel the electrons obtained from hydrogen on one side of the membrane to CO₂ on the other side. Perhaps the ability of iron to switch between Fe²⁺ and Fe³⁺ is of importance here in electron transport. In the presence of amino acids (and thiols), this mineral can use the energy from oxidation of CO to CO₂ into driving a “peptide cycle”. In this cycle, the system has been demonstrated to drive continuous oligomerization⁵⁴ of amino acids into peptides and hydrolysis of these peptides back into amino acids.⁵⁵

Pyrite minerals, which are at the heart of the iron–sulfur world hypothesis,⁵⁶ drive the production of most intermediates of the TCA cycle and the glyoxylate shunt in the presence of peroxydisulfate radicals, whose availability is questionable.⁵⁷ Preoxidized pyrite surfaces have been shown to drive reduction of N₂ into ammonium ions, although the use of atmospheric oxygen to drive this process as done in the study is questionable, especially in the context of origins of life.⁵⁸ That said, alternative oxidants on the prebiotic Earth may have substituted for the role of oxygen. Pyrite minerals can catalyze phosphorylation of inorganic phosphate into pyrophosphate using acetyl phosphate,⁵⁹ analogous to the formation of polyphosphates in acetogens.⁵² While the sulfide mineral leads to a lesser rate of phosphorylation compared to Fe²⁺ coprecipitating with acetyl phosphate and P_i, it also provides enhanced stability to the product. Formation of pyrite from FeS minerals also drives transamination between α-keto acids and NH₃⁶⁰ or simple organic amine donors.⁶¹

Recently, chemical gardens have attracted a great deal of interest within the origins of life community due to their dynamic nature, their potential as a mineral membrane capable of encapsulating biomolecules, and their catalytic properties.⁶² Mineral membranes composed of various metal silicates can catalyze the synthesis of a large number of nucleobases (including the RNA canonical nucleobases), amino acids, and carboxylic acids starting with formamide.⁶³ However, they show different product selectivities on the inner and outer sides of such membranes.⁶⁴ Iron oxyhydroxide based chemical membranes catalyze reductive amination of α-keto acids or reduction into α-hydroxy acids, depending on the Fe²⁺:Fe³⁺ ratio.⁶⁵ What makes these chemical garden minerals so versatile in their catalytic property has yet to be elucidated in mechanistic terms.

Semiconductor minerals like TiO₂, CdS, ZnO, FeS₂, and CdSe can catalyze photocatalytic redox processes that can reduce NAD⁺ into NADH nonenzymatically⁶⁶ (Figure 3). They can do this even when they are coupled to an appropriate protocellular setup, while simultaneously generating a proton gradient. ZnS colloids can act as photocatalysts (active around 360 nm UV light), driving three out of the attempted five reactions of the rTCA cycle.⁶⁷ However, it has been shown that the efficiency is low in these reactions including the carbon fixation steps. In these minerals, the energy available from sunlight excites the electrons of the valence band into the conduction band, and this excited-state species is able to transfer its electrons to an appropriate acceptor. The same photocatalytic mineral system is also potent in driving transamination reactions using NH₄⁺ as the amine donor and photoelectrons as reductants.⁶⁸

Figure 3.

Photocatalytic transmembrane electron transfer using photocatalytic minerals. Image reproduced from Dalai and Sahai.⁶⁶ Polycyclic aromatic hydrocarbons act as membrane channels for electron transfer. They receive electrons from outside the membrane, from the photocatalyst (depicted using the orange hexagons), and pass it to the Ru³⁺ complex. The resulting Ru⁺ complex then passes this electron to NAD⁺, resulting in NADH. Meanwhile, the photocatalyst quenches its hole (h⁺) using electrons from a sacrificial donor. Adapted with permission from ref (66). Copyright 2020 American Chemical Society.

The repertoire of minerals on the prebiotic Earth was reasonably extensive, and they were potentially present in all relevant geochemical niches.⁴⁸ In principle then, one can envisage that a large number of processes of interest to origins of life could have been driven by minerals during those early stages. Nevertheless, extant life is not a mineral-surface-bound life, whose processes occur in the aqueous medium of cells. How might have prebiotic evolution mobilized similar functions to the solution phase and facilitated the emergence of such functional protocells?

Metal–Ligand Complexes and Metalloproteins

Given that metal ions and minerals can facilitate catalytic functions, it is imperative to wonder how and when organic ligands and proteins emerged for this same purpose during the course of evolution. Nonprotein ligands most often tune the redox properties of the metal center by stabilizing particular oxidation states. As an example, [FeFe]-hydrogenases (Scheme 2) that are involved in stripping out the electrons from H₂ in core metabolic pathways have an unusual choice of ligands in CO and CN^–.⁶⁹ These allow the stabilization of the iron in the +1 oxidation state, which results when the iron centers accept two electrons from H₂. Coordination to soluble ligands is also necessary in the evolutionary pathway, to allow mobilizing the catalysts of life into the solution phase, which is central to life’s (proto)cellular architecture.

Scheme 2. General Metal Active Structure of [FeFe]-hydrogenase,⁶⁹.

Coordination with CO and CN^–, both of which likely originate from amino acids, allow the iron to switch between +1 and +3 oxidation states. Presumably, the back-bonding character that comes with the cyanide and carbon monoxide ligands reduces the electron density on the metal centre by shifting it on to the ligands. The sixth and open site on the iron site can provide coordination with solvent or substrate. The exact nature of X is unknown, although it is likely to be CH₂, NH/NH⁺, or O. Schematic made using ChemDraw Professional v21.0.0.28.

Further in the evolutionary developments, proteins bring several important advancements to the catalytic repertoire through their elaborate 3D structure and geometry (Figure 4^12,13,70,71). Proteins provide selective rate enhancements by binding very tightly and mostly specifically to their target substrates, thereby also minimizing stray reactions. In the context of metalloproteins, additional benefits come from the protein providing a coordination environment necessary to protect against degradation. They also optimally orient the metal center against the substrate (e.g., Mg²⁺ in the polymerization of nucleoside triphosphates into nucleic acid polymers/FeS clusters in aconitase²), optimize distances between consecutive metallocofactor units (e.g., the distances between cofactors in ETC to allow electron transfer⁷²), and tune redox potentials to enhance electron transfer rates and directions.

Figure 4.

Portion of the tungsten-based formate dehydrogenase dimer (PDB ID 5T61,^70,71 protein structure visualized using UCSF ChimeraX v1.5rc202210241843^12,13), which shows some of the benefits to metallocofactors in having proteinaceous scaffolding. The protein environment protects the iron–sulfur clusters (yellow-brown) against hydrolytic scission. It maintains a <15 Å distance between the metal centers (W-active site marked in dark green and the iron–sulfur clusters) to form a continuous wire to allow facile electron transfer.⁷² The redox potentials of individual [4Fe-4S] clusters—though not measured for every individual cluster—are known to be modulated by protein environments to ensure unidirectional electron transfer.

Recently, our group demonstrated that complexing with porphyrin macrocycles endows metal ions with enhanced oxidizing capabilities, which is not seen in the metal ion or the macrocycle itself.⁷³ Even simple aggregation through ionic interactions between the metal and the macrocycle was shown to enhance the oxidizing capability of Fe³⁺. Porphyrins as ligands are capable of complexing with a wide variety of metals, leading to a plethora of functions, some of them very ancient in their origins. For example, the Ni-porphyrin complex (Cofactor F430) is involved in the terminal methyl group transfer in the methanogenesis pathway.³² Given the plausible prebiotic availability of porphyrin derivatives,⁷⁴⁻⁷⁶ we think a natural next step in this direction is to study the possible participation of the different metal-porphyrin complexes in different nonenzymatic pathway analogues and uncover the underlying chemistry behind the selectivity of metals.

Ralser’s group demonstrated the catalytic ability of Fe²⁺ in driving the reactions of a primordial glycolysis/PPP when the metal ion complexes with cysteine.⁷⁷ However, the enhancement is not very different from the combined catalytic performance of the metal ion and the cysteine in isolation. Nevertheless, these complexes show enhanced product selectivity for ribose-5-phosphate as the temperature decreases. A remarkable demonstration came from the group of Mansy, wherein mononuclear iron sulfur peptides (a single iron unit coordinated to four glutathione thiolates, as in rubredoxins) were shown to catalyze electron transfer from NADH to H₂O₂.⁷⁸ This happened via the membrane soluble electron carrier ubiquinone, generating a proton gradient across a protocell membrane. Previous work has shown the possibility of nonenzymatic reduction of NAD⁺ to NADH coupled to decarboxylation of α-keto acids⁷⁹ or through photocatalysis. Coupled to the oxidation of NADH using FeS peptides⁷⁸ and the computational work elucidating the possible origins and evolution of the different cofactors in the electron transport chain,⁸⁰ we can envisage a functional prebiotic version of the ETC and some roadmap on how it might have developed into its extant version.

Iron–sulfur clusters are known to be among the most ancient cofactors involved in redox reactions and electron transport, with participation in a huge number of pathways where they mediate 1e^– transfers.^81,82 Different clusters like rubredoxins, [2Fe-2S], [3Fe-4S], and [4Fe-4S], are associated with different redox potentials,⁸¹ making them as a whole suitable for electron transport across a wide range of substrates. The formation of iron–sulfur clusters in a prebiotic context has been shown, starting with Fe²⁺/Fe³⁺,^83,84 sulfide, and different thiolate ligands, along with their redox stability through techniques like cyclic voltammetry. However, experiments showing their involvement in possible reactions of prebiotic chemistry have been limited. Schrauzer’s (iron–sulfur) complexes can drive CO₂ fixation and amino acid synthesis using acyl-thiol substrates and organic pyridoxamine⁸⁵ or ammonia⁸⁶ as nitrogen sources. Although such experiments demonstrate the catalytic power of such complexes/clusters, these are nonbiological clusters⁸⁷ and utilized organic solvents for the reaction. Similar experiments in a more prebiotically realistic environment are necessary to prove the relevance of such reactions to the origins of life. Given the involvement of the iron–sulfur clusters in central metabolic processes like the acetyl-CoA pathway³² and the electron transport chain,² we think it is a fruitful direction to pursue the nonenzymatic catalysis facilitated by iron–sulfur clusters, starting with a detailed characterization through model reactions and studies, followed by their application in such reactions of prebiotic chemistry.

Apart from the aforementioned studies, there are not many examples where the application of metal–organic ligand complexes has been demonstrated in driving the reactions of prebiotic chemistry. In our opinion, the area of catalysis with metal–ligand complexes has a vast potential for exploration in the context of prebiotic chemistry. This is especially pertinent considering the utilization of such complexes in biology in very ancient pathways and the vast amount of discovery, both application and principle based, which has happened in the field of organic/organometallic chemistry. Porphyrins, nucleotides, amino acids, and thiolates may represent some of the starting choices for ligands based on their biological utilization and different affinities for metal coordination. On the other hand, element fixation, redox, and biomolecular polymerization could be possible reactions of interest one could explore based on their importance in prebiotic chemistry.

In biocatalysis, irrespective of the metal having distinct organic ligands, they all end up in the highest rung in the ladder of complexity being a part of proteins, scaffolded by a protein backbone (directly or over its coordinating ligands).⁹ Therefore, the next step in deciphering the evolution of metalloproteins is studying simpler-than-biology metallopeptides. Studying the evolution of metallopeptides, however, presents a formidable challenge in itself. This is because sampling all possibilities of peptide sequences of a reasonable length is practically not feasible. So how does one go about addressing this task—what metallopeptides are perhaps the most imminent targets?

Bromberg et al. recently performed a computational analysis of 4500+ protein microenvironments present around the metal sites of different metalloproteins, using metrics of structural alignment, sequence alignment, and cofactor separations.⁸⁸ Through their study, they conclude that the most ancient metalloproteins are oxidoreductases. This finding makes a lot of sense when we view the involvement of oxidoreductases in “fixation” of different elements into forms relevant to biosynthesis and the importance of such pathways in sustaining biology with abiotic inputs. The problem of proteins being very sequence-specific, long chains of amino acids, is alleviated to some degree when we consider the following two important aspects.

• 1.The metal center essentially depends on a small region of the protein around itself for its function,⁸⁹ with the regions outside required for tight binding of substrates or interaction with other proteins/membranes. This is reasonable considering that the protein components in the immediate vicinity directly coordinate the metal center and those further away stabilize structures and shield the metal center.
• 2.To a large extent, the binding motif requires retention of structure and not necessarily the entire amino acid sequence so that some sequence alterations can be tolerated.⁸⁸⁻⁹⁰ Such a toleration is obviously less permitted in the amino acid positions involved in metal coordination. However, the present understanding has not sufficiently mapped a sequence-to-structure understanding; thus, we are limited in predicting from first principles all the sequences that can fold into a specific structure.

Computational analysis based on sequence profile alignments and structural comparisons suggest independent origins of peptide domains binding different oxidoreductase metal centers.⁹⁰ It indicated that the peptides binding [4Fe-4S], [2Fe-2S], and heme are perhaps the most ancient and most widespread in extant biology. Further, the domains binding [2Fe-2S] are the likely ancestors of those binding heme groups involved in electron transport. Evolutionary analysis using structural data and metal center proximity suggests a gradual evolution of peptides for the more oxidizing cofactors.⁹¹ The most ancient metalloproteins consist of the well-known “ferredoxin folds” binding to iron–sulfur clusters. This possibly has implications for their order of incorporation during early biological evolution. To simplify matters, a computational analysis of metalloprotein microenvironments suggests that as few as two to six short peptide motifs, some of which likely predate the most ancient protein folds, can explain a vast fraction of the protein structure in extant biology, metal binding or otherwise.⁸⁸ Interestingly, the different metalloproteins involved in energy conservation are all said to be constructed through a mix-and-match of different peptide motifs of the “redox protein construction kit”.⁶ These proteins are, in fact, among the most ancient proteins participating in certain pathways that are central to biology.

Experimental demonstrations in this direction, however, have been limited. Kim et al. demonstrated the successful reconstruction of [4Fe-4S] clusters using a minimal dodecapeptide with the ferredoxin binding motif CXXCXXCXXC, along with its redox stability, for hundreds of cycles through cyclic voltammetry.⁹² Duplication of the glutathione tripeptide (into a hexapeptide or a dodecapeptide) leads to the well-known ferredoxin binding motif.⁹³ While the tripeptide, hexapeptide, and dodecapeptide all bind the [2Fe-2S] cluster, the hexa- and dodecapeptides improve the half-life of the cluster against hydrolysis by up to 5-fold in comparison to the tripeptide. We believe that studies of this kind project very interesting meanings to some evolutionary trajectories. The extensive requirement of FeS clusters ([4Fe-4S] or [2Fe-2S]) and their labile nature toward hydrolysis impose strong selection pressure favoring the evolution of such peptides that can provide stability against hydrolysis while allowing incorporation of these clusters as catalytic motifs. This potentially could explain why such cluster-scaffolding peptide motifs were the first ones to evolve.⁹⁰

While the requirement of specific chirality in such peptides makes the abundance of such sequences in prebiotic chemistry a genuine concern,^92,93 the dependence on structure over sequence perhaps could provide some respite here.⁸⁸⁻⁹¹ How can we possibly progress the research in this area? Computational studies utilizing tools such as molecular dynamics can possibly be of help in advancing the study of metallopeptide evolution. Keeping in mind the most primitive metal binding motifs or domains, one can use MD/Monte Carlo simulations to look at what set of amino acid sequences (homochiral or heterochiral) could have acquired such a motif structure. Computational resources are an obvious limiting factor here. Nonetheless, such searches may also be guided through phylogenetic studies or sequence analysis from extant proteins, to shed light through the differential conservation of different amino acid positions. Using these approaches, one can narrow down on a representative set of sequences, which could lead to the reconstituted motif. Subsequently, rigorous experimental studies of the function and stability of such metallopeptide complexes will enable us to rationalize their evolutionary pathways in a systematic and experimentally verifiable manner.

Closing Remarks

Life sustains through catalysis, and it is only reasonable to think that it came into being due (in part) to primordial catalysis. Before biology, the catalysts that brought catalysis into existence were surely not the complex protein machines that keep it alive today. Understanding and discovering what these primordial catalysts were comprised of is key to understanding the emergence of life, perhaps even the structure of life.

What were these catalysts? Biology has preserved potential answers to this question in the form of the crucial and conserved core of its catalytic repertoire. The most central pathways and processes of life are dependent on metalloenzymes, with the metal or metal complexes constituting its catalytic core. The reasons are not very hard to decipher, after all: metals provide a plethora of properties in isolation or through simple complexation, many of which are not accessible to simple organic molecules. Pertinently, these metals that are present as ions and in minerals predate the first organic molecules and hence have always been available.

Metal ions are capable of catalyzing diverse reactions, as evidenced by their place in experiments of prebiotic and synthetic organic chemistry. Nonetheless, their properties are limited by their electronic structures and the aqueous environment where they function. Coprecipitation with anions and other cationic species into minerals renders new catalytic properties, some of them being an emergent property of the mineral assemblage and its constituents. These include surface catalysis that stems from the result of formation of surfaces, which allows for adsorption and concentration of solutes, while in some cases the catalysis is an outcome of the formation of a direct bond with the reaction intermediates, to facilitate chemical transformations.

Enter organic ligands, and the possibilities for catalysis just explodes! There are various possibilities for how the first organic ligands came into being, including metal ion/mineral catalyzed synthesis, oligomerization of building blocks like cyanide, and exogenous delivery via meteorites and comets. Organic ligands are amenable to fine-tuning in their properties since there are possibilities of many functional groups within a given molecular framework. This affects the steric as well as electronic properties of the ligands. Coordination with the metal center tunes reactivity at these metal sites analogously as well. Coordination with metal ions is a two-way street. The ligands affect the stable states of the metal center and could allow access to otherwise unstable states, facilitate solubilization of insoluble species in water (example: Fe³⁺), and most importantly, allow independence from being bound to mineral surfaces, facilitating progress toward a protocellular architecture. Conversely, it is also reasonable to expect that metal binding tunes properties of the organic framework relevant to catalysis. The ultimate stage in the evolution of catalysts for biology is observed in metallopeptides. These provide stability against environmental stresses, fine-tuning of properties for specific reactions and pathways, ability to localize into specific environments, etc. Thus, we have a reasonable but discrete pathway that still needs to be systematically discerned, suggesting a route for the origins of metal-based proteinaceous catalysts.

Despite such promising avenues, the developments in simple metal-cofactor catalysis especially in the context of prebiotic chemistry have been limited. One reason could be the conundrum of choice: what metal–ligand combination does one target for a specific process and what process does one target, to begin with? Looking to biology might help narrow down these possibilities. Consider a hypothetical example of the Mo-/W-dependent formate dehydrogenase family,⁹⁴ which catalyze the reduction of CO₂ into formate in bacteria as part of the acetyl-CoA pathway. The metal center (W⁶⁺/Mo⁶⁺) is ligated by two metallopterin cofactors, a sulfide and a cysteine/selenocysteine. The metallopterin cofactors are very complex in structure and are not ones envisaged to form readily in the absence of enzymes. A simpler replacement could be cyclic 1,2-thiols. Studying the minimal complexity of such model complexes that is needed to enable catalysis, and the progressive enhancement as one synthetically evolves the coordination environment, may help answer the plausibility of such catalysts, the choice of the coordinating ligands, and their evolutionary refinements too.

The field can surely benefit from a more principle-based approach. Even though hypothetical at this point, targeting metal ions or ligands based on specific properties necessary for interactions and catalysis might be a more meaningful direction to take. Such an approach will itself benefit from a mechanistic understanding of catalytic reaction centers in metalloenzymes or metal–ligand complexes. Answering questions on the choice of metals or ligands for different requirements, particularly in terms of their properties, will allow for a rational screening of catalysts. The field of organometallic chemistry has made significant advances in the specific use of metal–ligand structures as catalysts for many reactions and the underlying principles of chemistry driving the reactions. Given this, it is reasonable to also expect these advances to guide screening of catalysts for prebiotic chemistry.

In conclusion, understanding the origins of metalloproteins will shed light on the course of early evolution, some of which might potentially even predate geological records. Further, it is expected that these revelations will also guide rational choices in designing catalysts for synthetic biology and other kinds of molecular synthesis. With catalysts being among the crucial supporting pillars of extant life, it is only befitting to think that life’s origins as well were founded on simpler versions of catalysts.

The authors declare no competing financial interest.