Life and the chemical environment are united in an inescapable feedback cycle. The periodic table of the elements essential for life has transformed over Earth’s history, but, as today, evolved in tune with the elements available in abundance in the environment. The most revolutionary time in life’s history was the advent and proliferation of oxygenic photosynthesis which forced the environment towards a greater degree of oxidation. Consideration of three inorganic chemical equilibria throughout this gradual oxygenation prescribes a phased release of trace metals to the environment, which appear to have coevolved with employment of these new chemicals by life. Evolution towards complexity was chemically constrained, and changes in availability of notably Fe, Zn and Cu paced the systematic development of complex organisms. Evolving life repeatedly catalysed its own chemical challenges via the unwitting release of new and initially toxic chemicals. Ultimately, the harnessing of these allowed life to advance to greater complexity, though the mechanism responsible for translating novel chemistry to heritable use remains elusive. Whether a chemical acts as a poison or a nutrient lies both in the dose and in its environmental history.
1. The periodic table of life
Goldilocks happens upon the house of the three bears in the woods, and discovers among the too hot or too cold porridge, a bowl of porridge that is just right; among the three chairs that are too big or too small, one that is just right; and among the beds that are too hard or too soft, one that is just right. Like this tale, after 4.5 Byr of Earth’s history, to a first order, the elements that are required for life appear to have evolved to be in tune with the elements that are available in abundance to life within the ocean (figure 1). But how has this general matching of abundance and need arisen? Why, of the smorgasbord of elements within the periodic table, have only 25 of the 117 been demonstrated to be essential to all life with a further seven being useful to only some species (figure 1)? At least two factors are at play. First is the unique chemical nature of different elements that perform key parts of biological reactions, which has enabled the evolution of complex and sentient multicellular life. Second, and apparent from the broad correlation between concentration in the environment and biological requirement (figure 1), is the role of evolving availability, dictated by three inorganic equilibria: solubility, complexation and redox chemistry [1,2,3]. Metal availability for life is denoted as the concentration or activity of the free ion, acknowledging that complexation by key ligands can enhance metal availability in some forms for life today. With this perspective and delving into deep time, the key conclusion that can be drawn is that the elements required for life must have evolved over time in a movable feast. A periodic table of elements essential for life 2 billion years ago (Ba) would have looked very different to that of the modern day.
2. Geological controls of the chemistry of the ocean where life evolved
The abundance of elements within bulk Earth was determined by the Big Bang and subsequent supernovae processes. These elements were the starting ingredients, but Earth’s surface is chemically distinct from bulk Earth and the interior, owing to melting and differentiation. This process segregated Earth into the continental and oceanic crust, mantle and core, and partitioned the more labile elements according to their siderophile or lithophile nature.
Water is the only chemical reagent that redistributes elements between the rock reservoir and Earth surface water via solubility. At low temperatures, rivers leaching rocks and soils (accelerated by plants since their Carboniferous land invasion) form the primary conduit for transport of elements from the continental surface to the ocean after chemical weathering. At high temperatures, seawater circulates through newly forming and young ocean crust exchanging elements via leaching and clay formation. Additional loss of elements from the ocean occurs through chemically and biologically mediated precipitates to form ocean sediments that ultimately become subducted or lifted back onto continents. The relative balance of the inputs and outputs of elements to the ocean through time sets the evolving chemical nature of seawater.
The major ions in seawater (Na, Mg, Ca, K, Cl, SO4 and HCO3−) are those that are most soluble from rocks but, importantly, do not have a major sink or output. For example, Na, Cl and K are only removed via the infrequent occurrence of intense evaporite mineral formation that lends these elements a long ‘residence time’ of tens of millions of years and makes them relatively invariant for the majority of Earth’s history.
3. Changing major and trace element concentrations
It is the remarkable similarity between the abundance of the major elements in life and those in the sea that points to a coevolution between these two chemistries (figure 1). Na and Cl are by far the most abundant, followed by Mg, Ca, K and HCO3−. With such long residence times in the ocean, and concentrations being dominantly controlled by solubility and buffering, these major ions have maintained a near-constant relatively high concentration on million year time scales and for much of Earth’s history. Although there is an intimate relationship between the degree of employment for life and complexity of life, whereby the number of proteins using for example Ca increases exponentially as life advances from single bacterial cells to compartmentalized eukaryotic cells, and ultimately to multicellular forms , this is unlikely driven by changes in the environment. In fact, as we shall explore below, the increased use of Ca occurs contemporaneously with the increased use of both Cu and Zn and, we shall argue, is enabled as a result of the advent of those trace metals into the environment triggering enhanced complexity of life.
By contrast to the major ions that, today and always, have the highest concentration in the oceans and life, the transition (and trace) metals have the potential for an extraordinarily flexible redox state, owing to their partially filled d-orbitals in the ground state. This flexibility has two important implications for life. First, the transition metals are unique in providing life with unparalleled opportunities for performing electron transfer, redox and acid–base chemistry. Second, their availability is highly sensitive to the average redox state of Earth’s surface.
The major oxidation/reduction chemical changes of the surface of Earth arose from the uptake of carbon dioxide and nitrogen in combination with hydrogen either as H2 from hydrogen sulfide, or water, into organic compounds, i.e. reduction (in cells), and release of sulfur or oxygen. O2 release (as waste) led to the oxidation of surface minerals and ions and molecules in solution. All this activity was and is driven by light aided by essential catalysts largely based on bound inorganic ions. These changes, small at first, gained systematically as the rate of absorption of light by chemicals, the pool of reduced forms of carbon as a product of life and the inherent release of oxygen all increased.
4. Three chemical equilibria controlling metal availability in the ocean
The increasing redox potential of the environment affects elemental availability via the simple chemistry of three equilibria [3,5]. The general restrictions on the availability of elements as free cations, M+ and M2+ in the sea are insolubility and complex ion formation with redox chemistry controlling the redox state of the ions.
(a) Equilibrium 1: solubility
The solubility at equilibrium is described by equilibrium solubility products , where A is an anion. The order of solubility products of salts of the abundant metal ions, M2+, is that Ca2+>Mg2+ for carbonates and phosphates but Mg2+>Ca2+ for silicates while the order is Ba2+>Sr2+>Ca2+>Mg2+ for sulfate insolubility. These metals do not form sulfides. Such orders and the abundances of the elements have meant that free Mg2+ ions were more concentrated in the sea than Ca. The general order of insolubility of salts in the series of divalent d-block ions, M2+, is This order holds for sulfides and oxides. A major mineral in Earth’s mantle is olivine, Fe2+Mg2+SiO4, which formed through the abundance of Mg and Fe. It is thought that the early sea had high Mg2+ and Fe2+ concentrations, as olivine is relatively soluble. The high Fe2+ reduced the amount of sulfide somewhat as the iron precipitated it as pyrite though it is also not extremely insoluble. Even so the residual sulfide greatly restricted the free ion concentrations of Cu (probably Cu+ which has a very insoluble sulfide), Zn2+ and some heavier metal ions on early Earth.
(b) Equilibrium 2: complexation
The hydrated free ion concentrations, availability in solution, are related to their combinations with ligands in the sea according to the equilibrium K=[Mn+][An−]/[MA]. In the earliest times, the restriction of [M] in the sea would have been primarily due to the presence of hydroxide, carbonate, silicate and sulfide but later by oxyanions of stronger acids, for example, sulfate. The affinity of complex formation for metals closely parallels the above trends in solubility of metal salts. Hydroxide and oxide greatly reduce the free ion concentrations of cations with a charge of more than two. Thus at pH=7, M3+ such as Al3+ could be held by OH− in complexes or precipitates. In the presence of silica Al3+ also forms large soluble aluminosilicates. Decreasing pH increases free Al3+ concentration. The only divalent ion that may have been restricted by complex formation with aluminosilicate is nickel. The later metal ions of the series Mn2+, Fe2+, Co2+, Ni2+, Cu2+, Zn2+ formed sulfide complexes of increasing strength, 1/K, in this order but with Cu2+>Zn2+, the Irving–Williams series . The ions Mg2+ and Ca2+ form few complexes in the sea, while Na+ and K+ form hardly any complexes. Anions can also bind to one another, but rarely, or to organic surfaces.
(c) Equilibrium 3: redox
The metal ion oxidation changes are relatively fast and systematic due to the equilibrium thermodynamics of oxidation/reduction even if full equilibrium is not quite achieved. Redox potential, E, for the equilibrium between two redox states say M+ and M2+ is written as where T is the temperature, F is the Faraday constant and [ ] indicates concentration. Eo, the standard redox potential for [H+]/[H2] at pH=0 is 0.0 V and is −0.42 V at pH=7.0. The range of possible oxidation/reduction potentials is constrained throughout the evolving slow switch from H2 to O2 limiting conditions, where the O2/H2O standard potential at pH=7.0 is +0.8 V.
These three equilibria then prescribe the inevitable direction of change of element availability that occurs in a given order in the environment, according to the reconstructed degree of oxidation. It is proposed that this provides a control of cell chemistry too in evolution. Some elements were directly affected by change of oxidation conditions of the sea as described while other elements were not affected by the presence of hydrogen or oxygen, for example, Na+, K+, F−, Cl−.
5. Probing the past oxidation state of the planet
Although the redox history of the ocean and atmosphere may be more dynamic than originally thought , a consensus is emerging of the timing of the major oxidative switches (figure 2). The most powerful evidence for the great oxidation event (GOE), the first major transition from initially reducing to oxidized conditions at the planet’s surface, is the cessation of the mass-independent fractionation of δ33S, δ34S and δ36S between sulfides and sulfates, dated to between 2090 and 2450 Ma  (figure 2). Gas-phase chemical reactions prior to 2450 Ma are likely to lend the mass-independence to the sulfur isotopic fractionation. This mass-independent signature was diluted or eliminated subsequently in the presence of more abundant oxygen. Possible mechanisms to homogenize the mass-independent isotope signatures of the reduced and oxidized forms include oxidative weathering, or suppression of the atmospheric reactions of sulfur under oxidative conditions.
Molybdenum has become an element of increased interest for providing a window into the past redox conditions of the oceans. Under oxic conditions, this relatively abundant element exists dominantly as the molybdate ion (MoO42−)  but under euxinic conditions with free hydrogen sulfide, Mo is almost entirely scavenged from the water column as the particle-reactive oxythiomolybdate ion (MoOxS4−x2−)  which then captures whole ocean Mo isotopic compositions, sensitive to the proportion of oxic and reducing sediments. The Mo content, reactive iron content and the Mo isotopic signatures of shales all point towards at least one or two further steps in oxygenation significantly at the Precambrian/Cambrian boundary (543 Ma), coincident with the rise of animal life [9,12] (figure 2) and potentially entering into the Carboniferous period some 300 Ma . Crucially despite the GOE approximately 2.2 Ba, only the surface ocean is thought to have become oxic at that time, and the subsequent approximately 2 Byr until 543 Ma are thought to have maintained a euxinic and sulfidic deep ocean, limiting the trace metal solubility .
From this broad oxidation history of the Earth’s surface (figure 2), it becomes apparent that the oxidation state and complexity of life have advanced hand in hand. The major jump in oxygen approximately 543 Ma accompanied the ‘Cambrian explosion’ of life. This explosion in diversity in the skeletal fossil record represents the transition from single celled eukaryotes to differentiated cells, complex multicellularity and to biomineralization.
6. Evidence for evolving transition metal availability
Within the framework of the advancing oxidation potential of the Earth’s surface, the three chemical equilibria prescribe the evolving transition metal availability: declining Fe, Mn and Co and increasing Cu, Zn and Cd across each step of oxygen rise. The geological record provides one approach to testing this chemical framework. Chemical sediments from Proterozoic time may capture the chemistry of palaeoseawater. Potential archives include the banded iron formations (e.g. ), pyrite (in framboidal form)  and shales [9,17]. Banded iron formations consist of alternating magnetite (haematite) and chert. Framboidal pyrite (FeS2), an early diagenetic precipitate, is thought to form quickly after deposition and is therefore best poised to capture seawater as opposed to porewater chemistry . Shales can be thought of as fine-grained detritus, which may adsorb bulk seawater chemical properties as the particles fall through the water column. For each archive however, the chemical contents of the deposit itself cannot be translated directly to seawater concentrations. For the deposit to act as a faithful recorder of evolving metal availability, the metal must partition into the sediment independently of the oxidation state of the ocean, and remain unchanged by post-depositional diagenesis. To date, the picture of evolving Zn, for example, availability from various archives is anything but clear [16,17,18] although there are hints of an increase in seawater Zn concentration towards the modern day . Emerging work suggests that a more promising archive may be within the chemistry of greenalite (an abundant authigenic Fe2+-silicate in Precambrian iron formation, ironstones and shales). Long recognized as an abundant component of Precambrian lithologies but regarded as enigmatic in origin, greenalite occurs throughout the depositional spectrum and is one of the few phases preserved in Precambrian sediments that directly archives multiple trace metal inventories .
Intriguingly, the chemistry and genes of biota seem to provide better recorders than the geological record for the chemical framework of the altering trace metal availability as oxygen evolved and for the order of availability. The chemical make-up of extant phytoplankton which combine the trace metal requirement of the host phagoautotroph with the endosymbiont hints at the later increased availability of Cu, Zn and Cd [20,21]. The more recently evolved red algal superfamily has lower requirements for Fe and Mn, with higher requirements for Cd, Co and Mo than the green algal superfamily. But more than the metallome, an intimate look at comparative genomics and protein structures demonstrates that the prevalence of fold superfamilies specific metals within protein motifs  reiterates a phased employment of transition metals within organisms. Fe, Mn and Mo are preferentially selected by early life before Cu and Zn (and Ca, see above) increased their utilization only after the GOE. Furthermore, bioinformatic approaches to the characterization of the zinc and copper proteomes from all organisms with a full genetic sequence show distinct correlations between the number of zinc- and copper-containing proteins encoded within the genomes, and the complexity of the organism. Of note is the step change in use of zinc-finger containing proteins and zinc hydrolytic enzymes, the copper chaperones, homeostatic proteins and redox enzymes within eukaryotic and multicellular organisms [22,23]. There appears a clear trend of novel and multiple copies of the genetic code for these metal proteins as the metal availability rises in the environment due to oxidation. It is thought that the availability of these key d-block metals acted as a handbrake on the evolution of life (e.g. [1,5,14]), only permitting the functions necessary for eukaryotic cells and multicellularity to arise as and when the metals such as Zn and Cu, best suited to the metallo-centres of enzymes responsible for, e.g. cellular communication, splitting and cohesion between cells, were sufficiently available.
7. The new chemistry of the elements for life
The early establishment and evolution of life on the Earth, then, revolves around a cycle of (1) invention, (2) unwitting generation of initially toxic waste (e.g. O2, Cu and Zn), but which via the (3) evolution of management systems of those toxins towards (4) biological use, ultimately forces or yields invention of novel life chemistry and so on. Even Cd, a chemical latecomer to the ocean in the redox sequence, and commonly thought of as toxic to life due to its striking chemical similarity to Ca and ability to bind irreversibly to Ca enzymes, may have become of use to life. Vertical profiles of dissolved Cd show a near linear relationship with dissolved phosphate (e.g. ), an essential and limiting nutrient in the ocean, with the implication that requirement for life must also extend to Cd. The first enzymatic use of Cd has recently been proposed within a novel form of carbonic anhydrase (CDCA). Despite the inference that the earliest enzymes were likely cambialistic and evolved towards specificity , the relative scarcity of CDCA, so far found only with the diatom class of algae , speaks to the geologically recent novelty in the use of trace metals for life. Whether the uptake of Cd into CDCA accounts quantitatively for the global correlation between Cd and phosphate remains contentious [27,28]. A question for future research will be to disentangle from the widely explored biological quota of metals for phytoplankton , and inference of nutrient-like metals from seawater profiles, the entwined absolute requirement of metals for life from the biological sequestration of mistakenly imported metals.
8. The evolving toxic threshold and dose–response curves
From consideration of the evolution of the new chemistry of the elements, it transpires that it is not only the concentration of an element which dictates its degree of toxicity (Paracelsus ‘[a]ll things are poison and nothing is without poison, only the dose permits something not to be poisonous’) but also its availability within evolutionary history. The evolutionary cycle of invention and unwitting poisoning of life predicts a systematic temporal evolution to the dose–response curve of increasingly complex life as the concentration of a new chemical increased in the environment, e.g. O2, Cu and Zn (figure 3). Across each step in elevated concentration of newly introduced elements in the environment, increasingly complex life would have gained an increase in the toxic threshold, as the initial metal toxicity management systems (recognition, binding and accumulation and/or expulsion) transitioned towards biological use. Furthermore, once biological use of a novel metal is established then any response curve must incur some degree of limitation if that metal confers an inducible physiological advantage, or limitation if that metal becomes employed constitutively for a core physiological use. In this case, invention of a high-affinity uptake system may be additionally triggered to fulfil this biological requirement at low metal concentrations (figure 3).
Since the bioinformatic approaches, detailed above, yield only patterns and phasing of trace metal evolution through time, interrogation of both ends of the dose–response curves of extant organisms with diverse emergences may yield additional insight into absolute concentrations of metals at different times. If these response curves of organisms are reflective of the environment experienced since emergence, as suggested for coccolithophores in response to carbon , then probing a variety of extant organisms of diverse complexity and times of emergence for their toxic thresholds to, for example, Cu or Zn chemistries may reflect the highest concentrations of metals attained in the environment. Similarly, the specificity of the limiting arm of the response curve for different organisms might also approximate the environmental concentrations at the time the metal transitioned from being managed to becoming essential to life, since cambialism characterizes earliest metabolic processes . Both low-affinity and inducible high-affinity Cu and Zn transporters have been demonstrated (with the switch to high affinity for Zn2+ occurring between 10−10.5 and 10−9.5 M, and for Cu+ between 500 and 2000 nM) in marine phytoplankton [31,32], with induction linked to lower availability but essential need of those metals.
Lastly, such dose–response curves which display some degree of limitation may allow exploration of the farther reaches of the periodic table for additional essential elements for life and provide the first indicator that life could be finding novel use for ‘pollutants’ introduced by human modification of environmental chemistry during this anthropogenic era.
9. Concluding remarks
Subsequent to the introduction of oxygen as a waste product into the environment as an inevitable consequence of the emergence of photosynthesis, the periodic table of the elements essential to life evolved systematically according to element availability in the environment, directly controlled by the three inorganic equilibria. These equilibria enabled the advancement of life towards complexity via phased release of key elements into the environment, initially toxic to life but ultimately employed for use. The mechanism by which the systematic inorganic equilibria chemical changes of the environment interact with the chance alterations of the genomes of organisms to result in heritable novel life chemistry and the Goldilocks-like tuning of life’s chemistry to that of the environment remains elusive.
I would like to thank the Philip Leverhulme Award and the ERC Starting Grant ERC SP2-GA-2008-200915 for financial support.
I would like to extend enormous thanks to Bob Williams for his endless inspiration, mind-expanding discussions and undinted support and guidance; and Dave Sansom for invaluable help in preparing the figures.
One contribution of 18 to a discussion meeting issue ‘The new chemistry of the elements’.
- © 2015 The Author(s) Published by the Royal Society. All rights reserved.