1.

Overview Increasingly large amounts are known about the phylogenetic history of life back to the Last Universal Common Ancestor (LUCA) of all living organisms (Lane, Allen and Martin, 2010). However, LUCA was already a complex organism that was the product of a long pathway of evolution. Reconstructing the pathway that led to LUCA is challenging (e.g. see Nitschke and Russell, 2010). The first section of this paper provides a critical analysis of some of essential features of living cells, present in LUCA, and the challenges that had to be surmounted for LUCA to emerge. A logical analysis of these challenges reveals that the hydrothermal iron sulfur theories of Mike Russell and colleagues (Russell and Hall, 2006) provide the most plausible context for the origin of life on earth. In the second section a consideration of the nature of pre-biological evolution, allows the proposal of an ordering of key events in the emergence of life within this environment. This provides further support for the Russell hypothesis and opens up the possibility of testing aspects of the theory.

2. An Introduction: The First Cells were the Product of Evolutionary Processes

Living organisms are discrete, autocatalytic entities that are dependent on an ongoing energy source capable of fueling redox chemistry. With that as a starting point, two questions spring to mind: why are the living cells of earthly life the way they are, and how did they come to be that way? All living cells are complex structures. Understanding how the apparently irreducible complexity of even the earliest cells arose is the biggest challenge for origin of life theories. As many have commented, the possibility of all these elements coming together spontaneously in a single event is statistically implausible. The solution to this conundrum is that life cannot have emerged as the result of a single chance event in an undifferentiated “primordial soup”. Like all the elaborations of life that have emerged after LUCA, the complexity of original life is the product of evolutionary processes. Only by taking pre-LUCA evolution seriously can we hope to understand the origin of life. Furthermore, an understanding of these early evolutionary processes will shed more light on the reasons for life’s emergence that will inform our search for extra-terrestrial life.

3. Taking Evolution Seriously: Cells are Archaeological Artefacts

One of the key messages of Darwin’s Origin of Species is that complexity emerges from evolutionary processes. Another critical insight of the Origin of Species is the continuity of life manifest in Darwin’s law of common descent. Common descent implies that all living cells carry with them a continuous legacy back to the earliest forms of life. In short, all cells are archaeological sites, bearing the imprint of evolutionary history. From this perspective, several lines of evidence about life’s past can be unearthed. Firstly, contemporary cells are morphological fossils: as Lyn Margulis championed, the organelles of eukaryotic cells are tangible evidence of symbiotic events that produced them (Margulis, 1996). This morphological record can be traced back, via authenticated fossil microbes, to approximately 3 billion years ago, although a good deal of caution has to be used in dealing with the claims of possible biogenic remains from earlier periods up to the first billion years of earth’s history (Brasier, McLoughlin, Green and Wacey, 2006). Whilst the morphological fossil record is limited, the notion that the sequences of biological macromolecules of contemporary cells are themselves molecular fossils (Zuckerkandl and Pauling, 1965) has given rise to molecular phylogenetic analysis that has provided much evidence about life’s history. Furthermore, distinctive biomarkers have been used to provide complementary chemical insights into early life (Brocks, Buick, Summons and Logan, 2003).

The distribution of biochemicals within contemporary cells also provides insights into early cells. In particular, the reducing conditions within the cytoplasm of cells presumably reflect the fact that life was born in an anaerobic world predating the rise of free oxygen in the atmosphere. Life in such anaerobic environments makes widespread use of chemicals, notably iron-sulfur species, which are much rarer in the contemporary aerobic surface world due to oxidative damage and precipitation. The central role of iron-sulfur clusters within the suite of electron-transfer proteins in the cell membranes that mediate the energy generating processes of both chemotrophic and phototrophic organisms is another testament to the emergence of life in an anaerobic environment much richer in iron sulfides that the oceans of the past 2 billion years (Fraústo da Silva and Williams, 2001). Finally the pH gradients associated with all living cells, with a relatively alkaline interior and acidic exterior may reflect pH gradients in the environment that begat life (Lane, Allen and Martin, 2010).

4. Key Features of Life

All living organisms on earth, are reproducing entities comprised of a variety of organic compounds, including complex macromolecules, and water. The suite of ubiquitous biochemicals is one of the key indicators of a common origin of all life on earth. As outlined above, phylogenetic analysis of the sequences of nucleic acids from a diverse range of contemporary organisms supports the common descent of all life from a putative Last Universal Common Ancestor (LUCA). Whilst many of the details are conjectural there is convergent evidence that LUCA (either a single organism, or a dynamic microbial ecosystem) included many of the features of contemporary life on earth (Glansdorff, Xu and Labedan, 2008; Lane, Allen and Martin, 2010): a nucleic acid-based genetic code that encoded proteins that were built from L-amino acids; hundreds of protein-based enzymes that mediated metabolic transformations, including the manipulation of D-sugar phosphates; a cytoplasm that was bounded by some sort of membrane; ATP as an energy source that was generated from both substrate-level phosphorylation and chemiosmosis. Whilst much can be inferred about the core metabolites within LUCA, the cell membranes of eubacteria and archaebacteria utilize different lipids that have diverged since LUCA, and so the precise nature of the membranes of LUCA are still contested.

In short, LUCA is already a fully-fledged form of life that contains the basic features of cells (Gánti, 2003) namely: (i) a metabolic network of aqueous reactions that harness energy and building blocks for the reproduction of the cell; (ii) membrane-based compartmentalisation that allows the emergence of cellular complexity via evolution; and (iii) a genetic informational system, in the form of nucleic acids, that is capable of carrying complex information needed for cellular biochemistry. In order to generate an evolutionary account of the origin of life (see Section II) we need to understand the key challenges that were overcome en route; this is addressed in Section I.

5. Section I: The Intrinsic Constraints for Earthly Life

In this section the key features of earthly cells, metabolism and compartmentalisation, will be considered in turn. This overview sets out to establish the constraints that core cellular features place on the evolutionary trajectory that gave birth to life. This aids in the development of a tentative chronology to underpin subsequent evolutionary proposals, in Section II.

5. (A) Metabolites and Metabolism

Metabolism is the cornerstone of cellular activity, serving multiple purposes: the production of key metabolites from available raw materials; and the harnessing of chemical redox energy for biochemical functions, notably in the provision of dehydrating power which is used to produce the key biochemical macromolecules. The macromolecular metabolites are used for diverse purposes in cells, including the catalysis of the organized network of reactions that is metabolism. A sub-set of macromolecules, DNA, has a specialized role in the transmission of genetic information.

(a) Metabolites: Biochemical Building Blocks. There are good reasons why earthly life is based on aqueous mixtures of organic compounds. Carbon is unique in forming strong bonds to itself and to many other elements. This allows the generation of complex molecular structures bearing a diverse range of functionality. This molecular complexity underpins the higher levels of complexity found in living organisms. In general, the organic molecules of life are not the most thermodynamically stable, but they are kinetically stable: they have a finite lifetime once made. In general, life operates away from equilibrium and exploits such kinetic stability as the basis for short term, local order.

That water is a highly ordered liquid of moderate chemical reactivity is critical for life. Unlike gases (where particles are dispersed and cannot interact effectively) or solids (where particles are in close proximity, but with restricted motion), liquids alone offer the opportunity for both intimate chemical mixing and dynamic organised structures. Under any combination of temperature and pressure there are limited options for chemicals to be in the liquid phase since this requires intermolecular forces of attraction to be sufficiently strong to encourage close-packing, but sufficiently weak to allow mobility. The acid-base properties of water are exploited by biochemistry, but it is the strong hydrogen bonding of water that is most significant for life. Since water forms strong hydrogenbonds to itself, other chemicals will only freely dissolve in water if they, too, interact strongly with water, either by hydrogen bonding, or related ionic interactions.

The interactions between organic compounds and water provide a further basis for order exploited by biochemistry (Dobson, Gerrard and Pratt, 2002). Organic molecules festooned with polar groups that can hydrogen bond with water are hydrophilic (“water-loving”) and dissolve freely; many amino acids and sugar phosphates fit into this category. Organic molecules bereft of polar groups are hydrophobic and unable to interact strongly with water. These play only peripheral roles in biochemistry. Organic molecules possessing both polar and non-polar surfaces are, in many ways, the most interesting class of organic molecules for life since such amphipathic molecules spontaneously adopt ordered shapes in water to minimise the portion of non-polar molecular surface exposed to water. The complex architectures of cell membranes (based on amphipathic lipids), proteins (with both polar and non-polar residues in well defined sequence) and nucleic acids (with helical structures that bury the low polarity bases whilst exposing the polar sugar phosphate backbone) all arise spontaneously from this so-called hydrophobic effect (Figure 1).

Amphipathic Molecules and Complex Structures.

(b) Macromolecular metabolites: proteins and nucleic acids. A distinctive feature of cells is the high levels of control possible via sophisticated macromolecular structures, notably proteins and nucleic acids. All condensation polymers, including these and polysaccharides, are built by dehydrative coupling of monomeric building blocks. Joining these monomers in a well-defined order, to produce amphipathic macromolecules, can result in complex, functional architectures. The helical structures of DNA and RNA result from the regularity of the amphipathic building blocks (purine and pyrimidine ribose phosphates are identical in charge and similar in size and shape). The hydrophobic effect forces the buried bases to be adjacent, thereby establishing an interface for the Watson-Crick base-pairing that underpins genetic information (Figure 1).

The much greater variety of protein structures that underpins their diverse functional roles within cells is a consequence of the diversity of the approximately twenty amino acid building blocks. The starting point for adopting proteins as functional polymers is the availability of amino acid building blocks and access to dehydrating power for peptide coupling. With these in hand there are two subsequent challenges: the ability to generate structures which fold into well-defined three-dimensional shapes; and the ability to order amino acid building blocks in well-defined sequences to optimise their structure-function properties.

The three dimensional architecture of proteins is due to their amphipathic nature that results from combining monomers with either polar, or non-polar side chains (Figure 1). The availability of hydrophobic branched chain amino acids was a critical development in the evolution of proteins since these components are required for the generation of hydrophobic cores which control the structures of contemporary proteins. In the absence of a hydrophobic core, other mechanisms are needed to generate well-defined structures. Section II describes how metal ion templating of oligopeptides was a likely fore-runner to proteins, allowing the construction of small oligomers with ordered, functional structures.

The functional structures of proteins and nucleic acids are the result of evolutionary processes occurring spontaneously at the molecular level. The ability to link organic building blocks together via dehydration in water is a critical discovery in the emergence of life. Since the resulting condensation polymers are unstable with respect to hydrolysis; they spontaneously ‘die’. Only polymers that are made faster than they decompose can proliferate. This provides a form of molecular evolution. The differential survival of oligomers sets in train a form of pre-Darwinian evolutionary selection that underpins the emergence of functional macromolecules (see Section II).

(c) Biochemical Energy. One of the core functions of metabolism is the generation of useful chemical energy. Biochemical energy is often equated with ATP, since ATP can act as a water-compatible dehydrating agent: it is capable of providing the driving force for the creation of biochemical macromolecules as highlighted in (b). ATP, and other polyphosphates fulfil this role because they are thermodynamically unstable with respect to hydrolysis, but kinetically stable – reacting only slowly with water, and other nucleophiles, in the absence of catalysts.

There are actually two classes of water-compatible dehydrating agents in metabolism:

polyphosphates, such as ATP, and thioesters e.g. acetyl coenzyme A. Thioesters are generated directly in core metabolism and their dehydrating power can be interconverted with that of polyphosphates via substrate level phosphorylation. The other route to the generation of ATP in metabolism is the indirect route of chemiosmosis, whereby a proton gradient across the cell membrane is harnessed to yield polyphosphates. The latter mechanism is much more productive, but requires the maintenance of appropriate concentration gradients and appropriate enzymes to convert concentration gradients into polyphosphate energy. In contemporary metabolism the latter is generally accomplished via an elaborate F-type ATPase enzyme for ATP generation, although a simpler enzyme system, a proton translocating pyrophosphatase, is known for the conversion of proton gradients into pyrophosphate (Baltscheffsky, 1997; Lane, Allen and Martin, 2010).

5. (B) Membranes and Compartmentalisation

All cells are bounded by cell membranes that provide a semi-permeable barrier between the contents of the cell and the outside world. The enclosed cytoplasm is enriched in high concentrations of organic metabolites. The maintenance of distinctive localized chemistry is required for cells to be subject to Darwinian evolution: one of the defining features of life. All cell membranes are comprised of lipid bilayers (Figure 1), but vary in the precise nature of the component lipids. Fatty acids and glycerol are common components of the lipids of all cells. However, these components are linked via ether links in archaebacteria and via ester links in eubacterial lipids. These membranes are highly conserved and form part of the inheritable information of cells that is passed down during reproduction. Since LUCA is believed to predate the divergence of archaebacteria and eubacteria, the details of the lipid constituents of cell membranes in this ancestral life remain unclear.

Contemporary cell membranes are complex with many proteins associated with the lipids. Specific transport proteins embedded within the membrane control the flow of material into and out of cell: food influx, waste efflux, and the generation and of functional concentration gradients. Other membrane-bound proteins mediate the key electron transfer and related events associated with chemiosmotic energy generation. Proteins also mediate communication between the cell and the outside world.

Although amphipathic lipids spontaneously aggregate to form organized structures, such as micelles and vesicles (Figure 1), several challenges confront the development of the lipid bilayer membranes that are now ubiquitous for life. In particular, there is a trade-off between ease of division and mechanical strength. Simple lipid vesicles have low mechanical integrity. Whilst this aids in fusion with other lipids, and in division to daughter vesicles, it leaves them at risk of rupture from changes in the external environment, e.g. changes in osmotic pressure. If early cells were solely bounded by such simple membranes they could only emerge in an environment which provided a stable homeostatic backdrop with limited osmotic pressure gradients. Exo- and endo-skeletons, such as cross-linked polymeric cell walls have evolved to cope with these structural demands, but these, in turn, require complex systems for restructuring cells during division.

The second trade-off intrinsic to cell membranes is in the exchange of material with the outside world. Rudimentary membranes allow significant diffusion of materials. Furthermore, they can fuse with related structures and thereby undergo facile exchange of materials. These processes allow the influx and efflux of raw materials and waste products respectively. However, they compromise the inheritable information retained within any single compartment and the ability to sustain concentration gradients. Selective control over the flux of materials across the membrane requires complex macromolecular structures. Of the two possibilities, proteins and nucleic acids, only the former can form hydrophobic surfaces suitable for transmembrane structures. It is likely that genetically encoded proteins are a pre-requisite for proto-cells with robust biological membranes. Once invented, the combination of physically robust membranes and selective macromolecular transport provides many advantages, notably the ability to exploit concentration gradients. It is possible that the asymmetry of cell membranes and resulting vectorial flow is the result of environmental gradients (Lane, Allen and Martin, 2010). Finally, robust, selective, extracellular boundaries open up the possibility of full-blown Darwinian evolution.

6. Conclusions I: Constraints and the Trajectory to Life

The first conclusion from the above analysis is that the earliest stages of life involved the availability of appropriate carbon-based molecules in an aqueous environment with an ongoing source of chemical energy. Life is a chemical phenomenon and is therefore fuelled by the favourable relocation of electrons from electron donors to electron acceptors, i.e. the energy source involves a favourable redox potential. The second feature is that this chemical energy is converted into a biochemically usable form, i.e. the ongoing generation of water-compatible dehydrating agents. These are needed to underpin the development of the macromolecules that are distinctive features of biochemistry. A combination of organic building blocks and a way of transiently joining them, allows the generation of condensation macromolecules. By making such polymers transiently, functional polymers can be selected by evolutionary processes. Because such dehydrating agents predate the formation of complex structures they must have been produced by simple, direct processes, rather than by the complex systems of chemiosmosis and/or photosynthesis (Baltscheffsky, 1997). This strongly suggests that the chemistry associated with substrate level phosphorylation, namely the interconversion of thioesters and polyphosphates via acyl phosphate intermediates, is an early foundation for protometabolism that predates biological macromolecules.

With a dynamic turnover of organic molecules, amphipathic molecules played an important role in the emerging complexity of life. Individual amphipathic molecules can function as lipids, whilst ordered oligomers, that adopt well defined structures, can undertake more complex functions, including catalysis. The final piece in the jigsaw is the increasing informational demands for reproduction of the resulting complex macromolecules and associated proto-metabolic networks. This implies that specialized mechanisms for reproduction of proto-metabolic information are a late stage in the journey to life, emerging as the complexity of the proto-cells increased.

The constraints on life’s origin provide evidence about the environment in which life arose. The environment must have provided access to required nutrients: the chemical composition of cells is based on the availability of raw materials with selection for function. The fact that life is based on chemical disequilibrium means that there is a requirement for an ongoing energy input in the form of a redox gradient that can be harnessed in the form of useful chemistry. Given the traumatic nature of external stresses on the early earth, e.g. impact damage and high intensity UV radiation, the environment must have provided resistance to external shocks. The requirement for relatively high concentrations of functional chemicals and the potential fragility of early proto-cells indicates that the local environment was one that could allow local accumulation of chemicals against a backdrop of relatively stable osmotic pressure.

The original energy source for life was chemical and not photochemical. The environment provided a constant input of chemical energy that could be directly harnessed for dehydration chemistry in water without the complex apparatus required for membrane-based energy generation from the manipulation of concentration gradients that is the basis of photosynthesis and chemiosmosis. Once proteins and robust membranes were available, the ability to tap into environmental energy courses more efficiently provided the driving force for the emergence of these membrane-based energy mechanisms.

Open bodies of water fail to satisfy these criteria since they are beset by a dilution problem (cells are based on relatively concentrated levels of key chemicals) and were subject to the ravages of environmental damage. Nor can they provide an out-of-equilibrium situation to fuel life. The only conclusion is that life was not born by spontaneous aggregation of chemicals in a primordial soup (Lane, Allen and Martin, 2010). If our carbon-based life arose in an environment sheltered from impact damage, this further constrains the environmental context for life’s origins. Some of the challenges inherent in using organic compounds are the availability of suitable raw materials, the requirements to produce specific functional molecular structures, and the stability of functional structures in the face of environmental damage. Although external deposition, e.g. from carbonaceous chondrites, is an important source of carbon on the early earth, such deposition is unlikely to be the direct source of the molecules of life in an environment secluded from such sources. Instead, the organic compounds incorporated into these early living systems are likely to have been produced in situ within the geochemical environment, rather than by external deposition. Fixation of oxides of carbon, both CO₂ and CO, either geochemically and/or biochemically into more complex organic compounds is therefore foundational for life.

Of all the many scenarios that have been proposed for life’s origins only one, the proposal that life emerged from proto-metabolic autocatalytic networks in iron-sulfur rich hydrothermal mounds (Russell and Hall, 1997 and 2006; Wächtershäuser, 1988), can accommodate all the constraints described above. The version of this model that fits the preceding analysis utilizes the following concepts.

(i) Life arose because of intrinsic redox gradient of the planet (Russell and Kanik, 2010) where the electron-rich core is segregated from electron acceptors at the surface of the planet preventing facile dissipation of this thermodynamic resource. This situation provides a niche for life to emerge, as an entropy-maximising phenomenon that can catalyse the dissipation of the intrinsic geochemical redox gradient.

(ii) Life arose in hydrothermal systems because of ongoing flux of chemical redox energy provided by the leaching of electron-rich species from the crust into the aqueous environment. Geochemically generated hydrogen is likely to be an important carrier of electrons. This environment also provides protection from external shocks. Furthermore, local compartmentalisation was facilitated by the porous nature of the mineral deposits.

(iii) Life is based on organic compounds because complex structures, offering high levels of control, can be generated that are thermodynamically unstable but kinetically stable.

(iv) The initial step towards life is the generation of autocatalytic systems that form the basis for metabolism.

(v) With organic building blocks and dehydrating power in place, functional macromolecules emerged.

(vi) Once the apparatus to produce complex macromolecules became reliable complex catalytic networks and sophisticated membranes, incorporating proteins, could emerge. The resulting distinct cells, capable of high fidelity reproduction and Darwinian evolution, became the first life.

7. Section II: Evolutionary Pathways to Life

Armed with a tentative trajectory to life, and a likely environment for life’s emergence, the next stage is to draft a plausible scheme by which the three core features of life (metabolism, compartmentalisation and genetic information) emerge as a coherent whole as an end product of a continuous evolutionary process. This involves analysing the nature of the pre-biological evolutionary processes that can occur spontaneously in physical systems and the selection mechanisms that can drive the formation of cells.

In charting the pre-biological evolution of the first cells it is important that selection pressures exist that can plausibly drive the complexity of living cells. The following over-arching hypotheses guide the subsequent analysis.

(i) A key element in the development of life was the transfer of electrons from mineral sources (notably iron (II), sulfide and hydrogen) to electron acceptors leading to the fixation of carbon oxides (CO and CO₂) that both dissipated the redox gradient and simultaneously created organic compounds. As contemporary carbon fixation illustrates, the direct reduction of carbon dioxide is harder than indirect reduction of carboxylated derivatives thus providing a driving force for the generation of more complex structures.

(ii) Even in hydrothermal mixing zones there are kinetic barriers to the dissipation of the redox gradient in this way that require catalysis. Furthermore, simple catalysts were less efficient than networks of catalysts in mediating the appropriate redox chemistry, thereby providing a selection pressure for increasingly complex networks of catalysts.

(iii) Macromolecules were produced via dehydration chemistry and their transient stability in water provided a mechanism for molecular evolution that allowed the selection of functional polymers.

(iv) Mobilisation of the resulting autocatalytic networks within membrane-bounded vesicles allowed them to continue to access raw materials and reproduce.

(v) The increasing information demands for fidelity of reproduction of the mobilising, reproducing autocatalytic networks provided a driving force for the emergence of a specialized genetic information system. This was the step that completed the origin of life since the resulting entities became capable of full Darwinian evolution.

8. An Environment Poised for Life: Porous Hydrothermal Systems Rich in Metal Sulfides

As indicated in section I, the most likely environment for life’s origin are porous hydrothermal mounds receiving an influx of hydrothermal fluid rich in metal sulfides, notably iron sulfide, along with potential feedstocks for organic compounds such as hydrogen and carbon oxides. Such porous structures provide a sheltered environment and a multitude of discrete compartmentalised environments. There is the opportunity for parallel testing of innumerable variants of chemical systems. Individual pores with distinctive chemistry provide alternative chemical opportunities, (at least) one of which resulted in the generation of living systems.s

Iron sulfide minerals (Rickard and Luther III, 2007) have diverse chemistry, including the ability to catalyse varied chemical processes. This chemical capability has been harnessed by biochemistry where iron sulfur clusters at the active sites of enzymes mediate a wide range of biochemical transformations, including electron-transfer and acid-based chemistry (Beinert, Holm, and Münck, 1997; Flint and Allen, 1996). The catalytic capacity of individual iron sulfur species is controlled by surface-bound molecules (e.g. iron sulfur clusters fully coordinated by sulfur ligands are typically restricted to electron-transfer chemistry, whereas clusters with bound carboxylates can mediate acidbase chemistry) providing the opportunity for feedback loops between organic proto-metabolites and iron sulfur catalysts.

9. A Starting Point for Proto-Metabolism: Wood-Ljungdahl Chemistrys

The Wood-Ljungdahl pathway (Figure 2) is one of the few routes to carbon fixation in anaerobic organisms. At the heart of this pathway lies a bifunctional enzyme that reduces carbon dioxide at one active site (CO dehydrogenase, CODH) and transfers the resulting carbon monoxide to an adjacent active site (acetyl CoA synthase, ACS) where it carbonylates a methyl species, ultimately resulting in acetyl coenzyme A after interception of a nickel acetyl intermediate by a thiol (Grahame, 2003; Hegg, 2004; Russell and Martin, 2004). Both reactions are mediated by iron, nickel, sulfur centres within the enzyme (Volbeda and Fontecilla-Camps, 2005). Simple iron, nickel sulfide has been shown to mediate a biomimetic analogue of this chemistry (Figure 2) in which carbon monoxide and methane thiol combine to form methyl thioacetate (Huber and Wächtershäuser, 1997). This observation suggests that this (inefficient) chemistry will occur spontaneously within simple geochemical systems containing iron, nickel and sulfide, which are expected to feature in a number of hydrothermal environments.

There are several interesting features about this chemistry that are suggestive of a foundational role in the origin of proto-metabolism. Firstly, it involves the production of organic molecules from simple inorganic materials via direct redox chemistry in a plausible and potentially fertile geochemical context. Thus the generation of organic compounds in an environment rich in iron sulfur species provides varied opportunities for further catalytic chemistry as described above. This is augmented by the simultaneous generation of ligands, such as carboxylates, that can bind to – and modify – the catalytic chemistry of iron sulfur species. Secondly, the immediate organic product of the reaction is a thioester, an important class of reactive intermediates in biochemistry. Amongst other things, as noted in Section I, thioesters are one of the two main classes of water-compatible biochemical dehydrating agents and can be directly equated with biochemical energy. Substrate-level phosphorylation allows the direct interconversion of the dehydrating power of thioesters and polyphosphates (such as ATP). Hence, the Wood-Ljungdahl pathway supplies both fixed carbon and biochemical energy to an organism. For these reasons, it has been proposed that Wood-Ljungdahl chemistry might have played a central role in early metabolism (Ferry and House, 2006).

The thioester product of Wood-Ljungdahl chemistry provides a key building block for the generation of a core family of proto-metabolites. Iron-sulfur chemistry is known to mediate the reductive carboxylation of thioesters to α-keto acids, e.g. pyruvate (Cody et al., 2000). Iron sulfide can catalyse the biomimetic fixation of nitrogen and/or nitrate to ammonia (Dörr et al., 2003) and reductive amination of α-keto acids leads to α-amino acids (Huber and Wächtershäuser, 2003). All of these organic species are ligands for metals and can coordinate to surface iron ions, thereby modifying their catalytic properties (Figure 3).

10. Oligopeptides and Molecular Evolution

The generation of amino acids in the presence of dehydrating power (thioesters) will lead to the formation of oligopeptides (Huber and Wächtershäuser, 1998) which can also bind to the mineral surfaces associated with catalysis. As indicated in section I, these oligopeptides will spontaneously, but slowly, hydrolyse. Only oligopeptides that are formed at a faster rate than they hydrolyse can accumulate. Oligopeptides that increase the catalytic capability of the local chemical system will lead to an increase in the rate of their own formation. This cycle of oligopeptide formation and selective hydrolysis provides one of the first mechanisms of molecular evolution.

The generation of oligopeptides by iron sulfur species can take place, at least in part, via templated synthesis, whereby individual amino acids are coupled in proximity to metal ions and the oligopeptides are formed ready bound to metal centres (Costisor and Linert, 2004). This is of great significance for structure-stability issues of the oligopeptides produced. One major challenge to the evolution of functional proteins is that small peptides are generally unstructured. Folding into well-defined threedimensional structures provides functional properties and decreases hydrolytic lability. As indicated in Section I, the formation of such folded protein structures in contemporary biochemistry requires the availability of branched chain amino acids of low polarity to allow the formation of a hydrophobic core. Such branched chain amino acids present synthetic challenges, as exemplified by the specialized biosynthetic routes to these compounds in contemporary metabolism. Templating of oligopeptides around metal centres provides a route to folded functional oligopeptides in advance of having the building blocks necessary to produce stable hydrophobic cores to proteins. At a subsequent point, when amino acid hydrophobic side chains become available, there is a driving force to produce larger, more complex proteins, whose architecture is based on hydrophobic cores. Of course, the full exploitation of such technology (to incorporate a greater range of amino acids, with more diverse functionality and with more precise control) requires that the monomers are linked together in a well defined order, i.e. that some form of genetic information is available. This is one of many ways in which a ratchetting selection pressure for reproductive fidelity drives the emergence of complex cells.

11. Solubility and the Mobility of Proto-Life

Solubility issues are important in the development of a soluble cytosolic metabolism. In addition, the ability to relocate autocatalytic networks en masse is one of the driving forces underpinning the development of discrete reproducing cellular structures. Autocatalytic systems can develop within individual pores of a hydrothermal mineral deposit. They will, however, have a finite lifetime if constrained in space, since they can only continue to grow whilst there is a constant input of new raw materials. The ability of autocatalytic networks to migrate, and hence extend their lifetime, is dependent on solubility issues. Chemicals that are freely soluble in the hydrothermal environment are more likely to relocate. This is a purification mechanism if it results in side-products being lost; however, it is a reproduction mechanism if it allows the controlled migration of species.

The manipulation of free metal ion concentrations is an important control mechanism in the development of proto-metabolism, since the availability of other chemical species, notably phosphates, are compromised by precipitation in the presence of significant concentrations of multi-valent metal ions. Encapsulation of metal ions, such as iron, by oligopeptides can modify the reactivity and the solubility of the metal species. Essentially all the iron in living cells is sequestered within proteins in this way.

Phosphate plays a critical role in metabolism (Westheimer, 1987), but it is often a limiting macronutrient by virtue of its propensity to precipitate in the presence of multivalent metal ions, such as iron or calcium. This precipitation chemistry limits the opportunity for phosphate utilisation by proto-metabolic networks rich in multi-valent metal ions. Indeed the level of free soluble phosphate in hydrothermal systems is vanishingly small. This predisposition towards precipitation actually acts as a concentration mechanism providing enriched local amounts of this precious resource on the surfaces of hydrothermal pores. Furthermore, precipitated phosphate can undergo chemistry on mineral surfaces that mimics substrate level phosphorylation and can lead to the accumulation of polyphosphates (de Zwart, Meade and Pratt, 2004). However, this potential energy resource can only be tapped as part of a soluble metabolism if it can be relocated into aqueous solution.

Solubilisation of phosphates becomes possible with two features of proto-metabolism: iron encapsulation and the conversion of inorganic phosphates into organic derivatives. Both of these changes can be accomplished by the proto-metabolites. In competition for precipitation with limiting amounts on iron (II), inorganic phosphate precipitate preferentially compared to sugar phosphates (Pratt, Golovko and Toombs-Ruane, 2009), thus providing a selection mechanism for the incorporation of sugar phosphates as a subsequent stage in the developing iron/oligopeptide proto-metabolic system.

12. Vesicles and the Reproduction of Complex Autocatalytic Networks

Once amphipathic molecules, be they oligopeptide complexes and/or fatty acid based lipids, accumulate sufficiently they spontaneously generate vesicular structures (Deamer et al., 2006. Hydrothermal concentration mechanisms (Baaske et al., 2007) facilitate this process (Budin, Bruckner and Szostak, 2009). Such vesicles can relocate populations of molecules to adjacent pores in accordance with autocatalytic network theories such as the GARD model (Shenhav, Oz and Lancet, 2007). In this way, protocells based on oligopeptides, iron and sugar phosphates can prosper. However, there are limits to the complexity of proto-metabolic information that can be relocated in this manner. Szathmáry and co-workers have also pointed out that such systems are incapable of true Darwinian evolution (Vasas, Szathmáry and Santos, 2010). The development of the final level of complexity associated with cellular life depended on the discovery of specialised information sub-systems in the form of nucleic acids.

13. Fidelity of Information Reproduction and Nucleic Acids

As highlighted by Eigen and Schuster (1977, 1978a, 1978b), the fidelity of reproduction of biochemical information plays a critical role in the emergence and evolution of complex life. The final key breakthrough to life was the discovery of macromolecules which were capable of carrying information but which had regular physical properties that are independent of precise composition and sequence. Benner (2004) has pointed out that it is the polyanionic nature of nucleic acids, together with the similar, but not identical nature of the four basic building blocks that underpins their development as genetic molecules. So how did these molecules become integrated with the evolving autocatalytic networks?

Once sugar phosphate derivatives became available as components of soluble proto-metabolism they, like other molecules, could undergo further chemistry. Phosphate chemistry is useful in channelling sugar chemistry into interesting metabolites (Muller et al., 1990) opening the way to nucleotides (Powner, Gerland and Sutherland, 2009) These derivatives, like other proto-metabolites could undergo dehydrative couple to generate oligomers and functional oligomers could be selected and evolve. Initially the selection would be based on the catalytic capabilities of these molecules (Gesteland, Cech and Atkins, 2006; White, 1976) but then a second property emerged, namely the templated synthesis (Sievers, and Kiedrowski, 1994) that underpins genetic replication mechanisms. For the first time oligomers could be made with reliable sequences. A modified version of the RNA world (Koonin and Martin, 2005) was at hand.

When amino acids linked to these simple oligonucleotides as ester derivatives, they could undergo a new type of templated peptide synthesis which could piggy back on the increased level of sequence control inherent in RNA formation. Thus began the journey to the ribosome (Hsiao et al., 2009). The succeeding journey further illustrates the importance of reproduction fidelity as a key selection pressure in the generation of entities with the complexity of microbial life. The conversion of RNA to DNA to provide more chemically stable oligomers, capable of carrying greater amounts of genetic information, and the development of modified bases in DNA to increase the effectiveness of DNA repair processes all pay testament to Eigen’s key insights into error thresholds and the selection pressure of information.

14. Concluding Remarks: The First Life and Experimental Testing

Finally, with the combination of an underlying metabolism supporting the production of both proteins and nucleic acids of controlled composition and reproducing within robust compartments, an entity has emerged that is subject to true Darwinian evolution and can be considered to have crossed the threshold to life. That the emergence of an apparently irreducibly complex system through a succession of proto-metabolic innovations, both explains how the complexity of a cell can arise and also provides insights into the testing of the theory. If life emerges from a process based on combinatorial chemistry, which proceeds through increasingly complex autocatalytic networks, then it should be possible to test this hypothesis using methodology, such as microfluidic technology (Kreutz et al., 2010), surveying the autocatalytic potential of different combinations of putative protometabolites and catalytic species.