Subsumed complexity: abiogenesis as a by-product of complex energy transduction

Abstract

The origins of life bring into stark relief the inadequacy of our current synthesis of thermodynamic, chemical, physical and information theory to predict the conditions under which complex, living states of organic matter can arise. Origins research has traditionally proceeded under an array of implicit or explicit guiding principles in lieu of a universal formalism for abiogenesis. Within the framework of a new guiding principle for prebiotic chemistry called subsumed complexity, organic compounds are viewed as by-products of energy transduction phenomena at different scales (subatomic, atomic, molecular and polymeric) that retain energy in the form of bonds that inhibit energy from reaching the ground state. There is evidence for an emergent level of complexity that is overlooked in most conceptualizations of abiogenesis that arises from populations of compounds formed from atomic energy input. We posit that different forms of energy input can exhibit different degrees of dissipation complexity within an identical chemical medium. By extension, the maximum capacity for organic chemical complexification across molecular and macromolecular scales subsumes, rather than emerges from, the underlying complexity of energy transduction processes that drive their production and modification.

This article is part of the themed issue ‘Reconceptualizing the origins of life’.

1. Introduction

A compelling physico-chemical explanation for the origin of life remains elusive [1], with modern workers split between numerous, often mutually exclusive models that hinge on uncertainties between the rates of flux of energy, organic compound productivity and chemical compound interactivity [2–6]. These uncertainties are unlikely to be resolved soon, as the historical data that link the gap between the Last Universal Common Ancestor [7,8] and the as-yet-undetermined chemical phenomena modern scientists might call life or protolife [9] are not directly available for recovery or study. This state of affairs has remained remarkably constant over time, as has the general trend that new models tend to sprout from discoveries in other, more fundamental fields of scientific inquiry [10].

Concurrent with the development of ideas largely driven by progress in biochemistry and molecular biology, there is an increasing awareness of the complexity and inseverability between biology and its environment. This ‘biogeochemical’ vision arguably took root with the works of Vernadsky [11], and has reached its greatest heights with NASA's Astrobiology programme [12]. While the biogeochemical vision has stressed searching for environments where life has managed to survive or flourish, it is not clear that all environments where life can survive are appropriate environments for the origin of life [13], and indeed it may be the case that environments which are particularly hostile to modern organisms are more suited for originating molecular systems with life-like properties.

Nevertheless, the various problems associated with making, concentrating, sorting and organizing chemicals, which are now solved in biology by specialized enzymes and regulatory networks, could very well have been solved by abiotic physico-chemical processes prior to life's emergence. For example, the energy-requiring processes which allow cell division have been suggested to have been mediated by cyclic tidal processes [14], clays have been postulated to have served as replicase-like catalysts for nucleic acid replication [15,16], various metalloenzymes have been proposed to have been co-opted from environmental mineral catalysis [17] and mineral cavities have been proposed as the precursors of cells [18].

The complexity inherent to the production of organic molecules is a logical starting point in the search for prebiologically relevant processes. One should bear in mind, however, that the components of modern biochemistry may in fact be the outcome of a considerable amount of biochemical evolution, and not necessarily crucial for the origins of life [19,20]. Their centrality to origins research may reflect a form of ‘survivorship bias’, and may therefore provide incomplete guidance as a cohesive organizing principle. An environment-centric principle that encompasses different forms of energy and chemical transformations may be a more suitable alternative. The compatibility of stepwise models of prebiotic organic synthesis can then be assessed from the constraints inherent to known hosting environments.

One of the most prominent ideas invoked in discussing the origination of living systems of matter is the concept of ‘emergent complexity’, or the emergence of higher-level patterns of behaviour that transcend the rules followed by the constituent parts of a system at a lower level of interaction with one another and their environment [21–23]. However, ‘complexity’ itself is an ambiguous concept with more than 40 definitions [24]. The underlying challenge is to establish the relationships between various groups of measures [25] and between the corresponding natural phenomena. Regardless of the specific terminology employed, ‘plausibility’ is inherent to this discussion in relating the complexity of a synthetic protocol able to produce a compound of prebiotic interest; the complexity of this protocol is considered a proxy for the complexity of the environment required to carry out this protocol under naturally occurring conditions. If this level of complexity is comparable to processes that occur under naturally occurring conditions (i.e. without human intervention), then the protocol may be considered to be plausible.

How does one measure and compare the ‘complexity’ of a given physico-chemical system? Lloyd & Pagels [26] have outlined a universal definition of ‘complexity as thermodynamic depth’, consisting of the difference between fine- and coarse-grained entropy of a macro-scale physical system. In the specific case of the prebiotic production of reactive macromolecules, macromolecule A is thermodynamically deeper than macromolecule B if more entropy has to be generated in the process of constructing A than B. The additive property of thermodynamic depth implies that complexity increases as new depth-producing physico-chemical reactions come into play using the products of previous reactions. Though conceptually straightforward, the measurement of thermodynamic depth by itself provides no predictive framework for comparatively evaluating the likelihood that a given energy input source is more or less capable of imparting the production of prebiotic macromolecules.

We introduce here the term ‘subsumed complexity’ to describe the co-opting of external, inorganic energy transduction processes by internal, organic energy dissipation processes via intermolecular exchange of bond energy. This concept is compatible with the definition of complexity as thermodynamic depth outlined by Lloyd and Pagels. According to the concept of subsumed complexity, living systems did not arise as a result of the emergence of ‘new’ complexity, but rather from the stepwise instantiation, in the form of interacting organic compounds, of increasingly complex elements of energy transduction processes that already existed in the background environment. A search for conditions that give rise to ‘emergent complexity’ within a prebiotic system of interacting organic molecules may be overlooking a subtle but important thermodynamic implication: that the complexity of a physico-chemical system giving rise to life may have been essentially constant throughout all stages of abiogenesis. In this view, chemical complexity arose from, and is perhaps inexorably constrained by, a pre-existing framework of energetic complexity. However, if one were to track only the organic chemical components of a prebiotic system as they instantiated increasingly diverse environmental mechanisms, they would appear to increase in complexity over time. In this way, complex energy input pathways are said to have been gradually ‘subsumed’ into wholly organic, self-replicative forms or network structures.

Within the definition of complexity as thermodynamic depth, this would imply that different forms of energy input have variable levels of complexity that are discernible as entropic and energy dissipation characteristics. The search for the origins of life, then, may be informed by study of the most complex (i.e. lowest entropy) energy input processes. The elements of complex energy input into the chemical component of the system may then obviate the invocation of complicated biosynthetic pathways at life's origin [27].

2. Entropy and energy dissipation

Different forms of energy dissipate to a given ground state in different ways. For example, a single 1 MeV gamma-ray photon directed into a vial of liquid water will displace electrons and form ions and free radical species via three primary mechanisms of energy transduction. The first is that a gamma photon may be completely absorbed by an atomic electron, and the electron is emitted from the atom, which is termed the photoelectric effect [28]. The photoelectric effect dominates interactions with low photon energies and with target atoms of high atomic number. The second is Compton scattering, wherein a photon interacts with an atomic electron and is scattered with reduced energy, and an electron is ejected from the atom with the remaining amount of energy in the form of momentum [29]. This process dominates interactions between light atoms and typical gamma photon energies. The third process is pair production, which occurs for extremely high energies (i.e. associated with galactic cosmic rays); a photon interacts close to the nucleus of an atom and generates both an electron and a positron, which are ejected from the atom [30]. The first and second forms of energy exchange dominate energy dissipation cascades associated with a 1 MeV photon, and the ejected (secondary) electrons may themselves emit Bremsstrahlung photons (X-rays) of lower energy or induce the emission of Cerenkov (linearly polarized UV) radiation before the system equilibrates with the thermal energy of the ground state (figure 1a) [31]. In the course of this process, some of the energy does not reach the ground state, and instead is retained in the form of residual chemical species produced by free radical recombinations or redox reactions. The most common and stable of these species for a liquid water medium are H₂, O₂, H₂O₂, solvated electrons or OH or H ions [32].

Figure 1.

A schematic comparison of two equivalent amounts of photonic energy input, but with non-equivalent thermodynamic depths, into a target medium of liquid water molecules. (a) Input of 1 MeV gamma-ray photons. (b) Input of 1 million 1 eV near-infrared photons. (Online version in colour.)

In an alternative experiment, the same amount of energy in the form of 1 000 000 near-infrared photons at 1 eV each interact with the same vial of liquid water (figure 1b). These photons are unlikely to trigger the same chemical responses generated in the first example. Infrared photons are most likely to fractionally increase the momentum or excitation state of many of the target molecules, and then the system will re-emit this excess energy as photons as the system thermally equilibrates to the ground state. The production of X-rays, UV photons or even visible photons is highly improbable.

The thermodynamic difference between these two energy dissipation processes is quantifiable only in the most general of terms but derives from the underlying entropy of the input energy. A statistical physics treatment can be used to demonstrate that the ratio of entropy production can be approximated by the ratio of the number of input to output particles carrying the energy dissipated within the system [33], a numerical value that reflects the thermodynamic depth of each energy input. There is a directionality associated with this difference: a gamma-ray photon can, with a high probability, produce a physico-chemical effect similar to that produced by multiple infrared photons of an equivalent amount of energy (especially if sufficient time is allowed to pass to allow the set-up to fully equilibrate with the ground state), but the reverse cannot be said to be true under most naturally occurring circumstances. It is in this way that the dissipation of a gamma-ray photon follows a more complex set of rules than does the dissipation of an equivalent amount of energy of near-infrared photons: gamma-ray photons of a given amount of energy can exhibit more complex physico-chemical transduction processes than can an equivalent amount of energy of infrared photons.

The reasons for these differences are themselves complex. Gamma-ray photons interact with a target medium at the subatomic and atomic levels, and these energy dissipation processes are not particularly sensitive to the atomic composition of the target medium—similar effects would be expected for targets composed of N, S, C, P, etc., maintaining the overall characteristic of ‘directionality’ of energy dissipation from a fine scale to a coarse scale, until photons and the chemical bonds of the system are able to interact less irreversibly with one another (figure 2). However, the molecular composition of the medium can have ancillary effects on the attenuation of energy at the subatomic and atomic levels. The displacement of electrons can produce free radicals, which can produce oxidized molecular species upon recombination, which in turn can quench free radicals more readily than can reduced species. Alternatively, the joining of atoms into molecules with shared electron orbitals (i.e. aromatic compounds) can also make the overall molecule more or less susceptible to interaction with free radical species, altering the coarse-scale chemical output to the energy input initiated at the fine scale (figure 1a, dashed line). As a result, the ‘evolution’ of chemical species at the molecular level alters the energy attenuation at lower, atomic levels. The overall energy dissipation topology of the system can develop in complex ways at future time steps. This is a product of the simultaneous dissipation and channelization of energy through the system as more varied chemical species that come to compose the molecular population of the system are coupled to one another, or when atoms capable of forming versatile bonds between many atoms (i.e. carbon) can effectively carry forward traces of past chemical transformations from previous time steps.

Figure 2.

Fundamental forces and associated energy transfer particles (right), energy retention mechanisms (left) and associated energy retention structures (centre). Energy dissipation generally proceeds upwards through the chart. (Online version in colour.)

The nonlinear relationship between energy channelization and dissipation bears many of the characteristics of ‘emergent’ systems, whereby larger, coherent patterns of entities emerge from the interaction of smaller entities. Such systems are characterized by feedback loops, hierarchically nested groups of objects at different scales of observation, downward causation and robustness of order to external perturbations [21,34,35]. In effect, by introducing energy characterized by interactions below the level of molecules, molecules can be resolved as a coarse-grained level of organization [36]. It is important to note that a boundary between ‘fine’ and ‘coarse’ levels of observation can be drawn at multiple different points on a plot such as figure 2.

3. Chemical networks and complexity

Life exhibits emergent properties at macromolecular and integrated systems (i.e. species) levels and above, with substrate molecules effectively serving as a finest grain level of order and energy input (figure 3). Research into prebiotic chemical reaction networks focuses on the dynamical properties of systems of reactions and interactions, i.e. time evolution of concentrations according to mass-action kinetics or stochastic kinetics, or on the reconstruction of chemical networks from kinetic behaviour. These are part of broader efforts to replicate or engineer emergent chemical systems between molecular and higher levels by creating autocatalytic cycles [37,38] or informational polymers [20].

Figure 3.

Simplified depiction of major energy flow pathways characteristic of living organisms. Energy dissipation processes proceed upwards. Dark blue boxes indicate proven physico-chemical retention structures. Polymeric information describing and controlling configurations of the system constitutes feedback with the potential to maintain or alter future configurations of the system. (Online version in colour.)

The sequence of events covered by the term ‘abiogenesis’ (i.e. the progression from non-living to living states of matter) can be represented as a coevolution of energy input and chemical networks of reactions up to the point where a regulatory function associated with advanced catalysts and/or informational polymers separates from the metabolic function so that both specialized subnetworks head towards Darwinian evolvability (system chemistry). Recasting abiogenesis as a problem of chemical network evolution and specialization displaces the evolvability of chemical networks into environmental contextual constraints. Network evolvability can be understood as a modification of the network structure via addition/removal of reactions and interactions rather than the time evolution of concentrations. Such changes are ultimately driven by environments hosting these networks. Environmental factors such as available substrates, reagents, energy sources and spatial connectivities parametrize reaction networks in the same manner as connecting glassware with input lines and setting up thermostats parametrize laboratory experiments. The evolution of an environment amounts to a drift in the network parameter space. A lack of evolvability [39] was one of the points of criticism of the ‘metabolism-first’ scenario proposed by Morowitz [40]. This scenario placed a very particular cycle, a non-enzymatic version of the reverse tricarboxylic acid cycle, into a non-equilibrium environment that lacked any specific features [41]. This is opposite to the approach advocated in the present contribution, which aims to first specify energy input characteristics and then to evaluate their ability to drive reaction networks.

An environment-centric approach facilitates assessment of the complexity of prebiological chemistry using existing synthetic organic studies as a proxy for a natural process of debatable, but fundamentally irreducible, plausibility. Much of this work has focused on uncovering one-pot synthetic routes to prebiological molecules [42,43] that incorporate sequential network-like schemes [44]. The latter network implements a form of Kiliani–Fischer homologation and relies on the externalized environment to combine three geochemical zones, rich in CN-, H₂S and NH₃, respectively, and UV exposure that effectively introduces two additional reducing environments. The development of the chemical network is contingent on the specific connectivity of the geochemical zones and precise exposure of these zones to UV radiation. This scheme represents a single realization of environmental connectivity and UV exposure that is favourable for the synthesis of biogenic amino acid precursors. In this scenario, the thermodynamic depth of the energy input environment itself (i.e. the stepwise progression of heat input and UV irradiation and the methodological addition of the reactants) represents the full physico-chemical complexity that leads to a highly ordered final set of products. It is in this sense that the statistically improbable distribution of the final products can be said to subsume the complexity of the entire energy input and physical mixing protocols.

4. Evaluation of two different energy input scenarios

Subsumed complexity, via comparison through thermodynamic depth, provides a conceptual framework for evaluating the likelihood that a given form of energy input can lead to complex chemical systems. This is achieved by comparing well-described energy dissipation phenomena, the relative entropy of energy inputs, and the transduction of energy through different forms of chemical bond formation and degradation as a subsystem of a larger, conservative system, and observing how much excess energy is dissipated to arrive at the final configuration of product molecules. Chemical bonds in this dissipative dynamical system are regarded as energy retention mechanisms, or mechanisms that inhibit some fraction of the input energy from reaching the ground state [36]. The retention of energy at dissimilar scales opens the possibility for chemical emergence.

Two generalized energy input scenarios can be visualized for comparative study. The first consists of a combination of thermal energy and highly reactive small molecules, approximating environments such as hydrothermal vents, geothermal pools, geysers or near-surface shallow puddles. The second consists of energy input via atomic or subatomic low-entropy radiation, approximating environments such as the interiors of asteroids, comets or dust grains [45–48], gaseous nebulae [49] or radioactive mineral seams [50–52].

Energy dissipation associated with heated reactive molecules is relatively simple. Heat is dissipated through convective and conductive intermolecular interactions [53]. At least some of this energy can be channelled into abiotic synthesis of small organic compounds [54]. Most of the input chemical energy (in the form of simple compounds such as H₂, H₂S and FeS) creates other relatively simple reactive compounds that interact with one another within a molecular interaction network that produces heat and a small number of higher molecular weight organic compounds [2]. It has not yet been shown that organic compounds of high molecular weight can be formed in this class of environment under these energy input conditions [55,56]. It is also not clear, on the basis of the structure of energy input and conversion, whether there is sufficient thermodynamic depth to differentiate between coarse and fine levels of molecules, though a hypothetical difference is often implied between large molecules or molecular networks and polymeric macromolecules (figure 4, dashed line).

Figure 4.

Simplified depiction of major energy flow pathways associated with a hydrothermal vent. Energy dissipation proceeds upwards. Dark blue boxes approximate physico-chemical retention structures. Black dashed line denotes possible boundary between fine- and coarse-grained levels of organization. (Online version in colour.)

Energy input by subatomic irradiation results in a fundamentally different configuration of energy transduction pathways. Even for a generic environment of target molecules and radiation particles, there is a remarkable network of physico-chemically coupled energy dissipation pathways and retention structures. Energy input in the form of nuclear fission products can produce photons and particles with energies in the range of MeV which are quickly attenuated by elastic and inelastic collisions with neighbouring subatomic particles. These energies are so high that they can directly induce atomic and molecular reconfigurations in solids such as host minerals [57]. Simultaneously, induced and latent decay of unstable isotopes can alter the electronic and atomic configuration of surrounding atoms in a complex cascade of emission, attenuation and re-emission of many different particle types and photon energies. This can include the photoelectric effect, Compton scattering, electronic displacement (i.e. ionization) and subsequent recombination of ions and free radicals. Simultaneously, the emission of secondary photons from electronic displacement and dispersion can emit X-ray photons known as Bremsstrahlung radiation, and the movement of charged particles through the surrounding medium emits Cerenkov radiation ranging approximately 250–500 nm [31]. Photons of these energies also have enough energy to impart molecular reconfigurations such as ionization and free radical formation. As discussed previously, a fine/coarse-grained distinction can be drawn between atomic and molecular levels of energy retention (figure 5, lower dashed line) in addition to a hypothetical distinction between molecular and macromolecular levels (upper dashed line).

Figure 5.

Simplified depiction of major energy flow pathways associated with near-surface radioactive mineral beds. Energy dissipation processes proceed upwards. Dark blue boxes indicate proven or plausible physico-chemical retention structures. Black dashed lines denote possible boundaries between fine- and coarse-grained levels of organization. (Online version in colour.)

Subatomic energy is channelized into the creation of a relatively small number of chemical species. At the coarse (molecular) level, whether a compound is synthesized or degraded by free radical reactions depends on the reactivity of the product molecule, its susceptibility to attack by the most abundant radical species in the environment, its ability to attenuate ambient background energetic photons without being degraded and the probability of its recombination with other radicals before recombining with its own radical conjugate fragments. In this sense, the attenuation of input energy by a finite number of atomic objects following a probabilistic array of atomic configuration rules effectively channelizes the input energy to a relatively small number of energy retention states, with some level of recognizable feedback between molecular and atomic levels opening the possibility for emergent behaviour.

5. Synthesis and conclusion

It is the relationship of energy to matter and time that fundamentally defines the capacity for relationships among different classes of chemical compounds. Some forms of energy can be channelized into creating chemical products in more specific ways than others. By outlining energy dissipation pathways typical of subatomic processes, a level of atomic-to-molecular emergent behaviour can be resolved that is commonly overlooked in prebiotic synthesis conceptualizations. There is no a priori basis to conclude that emergence at this level is a prerequisite to creating living chemical systems characterized by a higher (i.e. molecular-to-macromolecular) level, but it opens the possibility that there was a tractable source of complexity that existed distinct from, and perhaps as a predecessor to, the process of emergence that characterized abiogenesis. The complexity inherent to subatomic energy dissipation and channelization has the potential to structure an organic chemical environment in ways that reduce reliance upon invoking improbable sequential processes of a naturally occurring environmental setting.

The term ‘subsumed complexity’ describes the possibility that low-entropy sources of energy input are more likely to result in the production of complex chemical compound interaction networks. As increasingly complex chemical energy transduction pathways form, the susceptibility of the chemical system to some energy inputs decreased until they were altogether diminished. At this point, the characteristics of energy input and transduction pathways were ‘subsumed’ into a freestanding organic network. The result was an aggregate chemical system that complexified against a backdrop of energy transduction with emergent characteristics at several energy-retaining object levels. It is possible that this background complexity of energy transduction was a prerequisite to generating life-like chemical systems, and is therefore a more suitable template for generating life-like behaviour than any specific or sequential configuration and alteration of input chemical species. Through the idea of subsumed complexity, the search for life's origins may be reconceptualized as identifying physico-chemical settings that manifest energy transduction attributes commonly associated with living systems such as hierarchical nestedness and downward causation at multiple levels. The reconstruction of abiogenesis as a historical process may then focus on identifying, within the scope of these identified settings, which array of chemical inputs and environmental parameters produce organic compound outputs that most closely resemble the composition and interactive relationships of living systems.

Data accessibility

This article has no supporting data.

Authors' contributions

All authors contributed critical ideas and intellectual content to the manuscript and participated in its writing.

Competing interests

We declare we have no competing interests.

Funding

This work was supported by JSPS KAKENHI Grant-in-Aid for Scientific Research on Innovative Areas ‘Hadean Bioscience’, grant number JP26106003. H.J.C. would like to thank the ELSI Origins Network (EON), which is supported by a grant from the John Templeton Foundation. The opinions expressed in this publication are those of the author(s) and do not necessarily reflect the views of the John Templeton Foundation. Z.R.A. was supported by a Geobiology Postdoctoral Fellowship from the Agouron Institute.