1
From Chemical Invariance to Genetic Variability

Günter Wächtershäuser

The archaeologist of nature is at liberty to go back to the traces that remain of nature's earliest revolutions, and, appealing to all he knows or can conjecture about its mechanism, to trace the origin of that great family of creatures…down even to mosses and lichens, and finally down to the lowest perceivable stage of nature, to crude matter. From this and from the forces within, by mechanical laws, like those that are at work in the formation of crystals, seems to be derived the whole technique of nature.

Immanuel Kant [1]

1.1 Heuristic of Biochemical Retrodiction

Darwin (1863) wrote in a letter to Hooker [2]: “It is mere rubbish, thinking at present of the origin of life; one might as well think of the origin of matter.” Studies of nucleosynthesis are now quite advanced, but research into the origin of life is still an immature science. The problem of early evolution of life is unique and requires its own heuristic. A commonly used heuristic consists of one-to-one back-extrapolations of individual biochemical features (Figure 1.1a), for which Lipmann [3] coined the term backward projection. More and more backward projections add evermore ingredients to the recipe. Inevitably, this way of thinking leads to the notion of a “primordial broth.” No one has ever spelled out all that what would or would not have been in the broth and how precisely the organization of life could have come about within such a chaotic situation.

Figure 1.1 Heuristics of (a) parallel backward projection and (b) convergent biochemical retrodiction (F1, F2, etc., extant features; P, P1, P2,etc., precursor features).

(From Ref [47] © (2010), Springer.)

This conceptual hodgepodge is overcome by a heuristic of convergent back-extrapolation, termed biochemical retrodiction (Figure 1.1b) [4]. Some extant features are still projected all the way back to the origin. Typically, however, several extant biochemical features are related to one simpler common functional precursor feature and several precursor features are related to a still deeper common precursor. This pattern is applied over and over again, drawing in more and more extant features, progressing to ever deeper, fewer, and simpler precursor features, and generating an overall pattern of backward convergence. Ultimately, the heuristic of biochemical retrodiction aims at a restricted set of chemical compounds and processes, which cooperate to form a distinct chemical entity with the ability to reproduce and evolve: the “pioneer organism.” Its chemistry is intrinsically synthetic, thereby imposing from the start directionality from simple to complex on the overall process of evolution. In this sense, the pioneer organism paves the way for all future evolution, hence its name.

Specifically, we include in our platform for retrodiction extant biochemical features, which combine aspects of evolutionary change with aspects that have been largely invariant over time by the universal laws of chemistry. It is precisely the aspect of chemical invariance within evolved biochemical features that provides directionality and allows us to unravel evolutionary history backward to the very pioneer organism of life. For the biochemical retrodiction of multistep biosynthetic pathways we employ in addition the Florkin–Granick rule [5, 6] that earlier steps in a pathway have greater evolutionary antiquity than later steps. This rule is based on the assumption that biosynthetic pathways evolve by terminal extensions. We should apply caution, however, because pathway evolution comprises also lateral branchings, recruitments, reversals, and eliminations [7].

The results of biochemical retrodiction are evaluated empirically by chemical experiments and theoretically by quantum-chemical calculations [8], with the perceived geochemical scenario determining the parameters. Biochemical retrodiction suggests chemical experiments and experimental results inform revised retrodictions, and such iterative procedure promises a progressive exploration of the pioneer organism. Finally, when our procedure leads to competing hypotheses, we prefer the one with the greater explanatory power, that is, the ability to explain a greater number of extant biochemical facts with fewer evolutionary assumptions [9]. We shall now use this methodology for a step-by-step reconstruction of the pioneer organism, beginning at the simplest level: the elements of life.

1.2 Retrodicting the Elements of Life

The elements of central biochemistry [10] fall into two distinct subsets. (i) The main group nonmetal bioelements (H, C, N, O, P, S, Se) make up the bulk of the biomass with mostly structural roles. They originate deep in the mantle of the Earth and form volcanic gases (H2, N2, CO2, CO, CH4, NH3, H2O, SO2, H2S, H2Se, COS, HCN, CH3SH, P4O10). (ii) The transition metal bioelements (Fe, Co, Ni, V, Mo, W, Mn, Cu, Zn) occur in organisms only in trace amounts, with mostly catalytic functions. Together with the main group biometals Mg, Ca they form essentially stationary crustal minerals. These two classes of bioelements come into close encounter at volcanic-hydrothermal flow sites in the presence of liquid water [11], cf. [12].

Four main group nonmetal bioelements (H, C, N, O) form the structural basis for biochemistry. They are involved in all biochemical reactions, each one with its unique roles, indispensable for life, down to the pioneer organism. Sulfur and its companion selenium have diverse biocatalytic and bioenergetic functions. They are projected into the pioneer organism. Phosphorus is indispensible in genetics and bioenergetics, but limited to phosphate group chemistry. It must have been acquired after the origin of life. Therefore, the pioneer organism is defined prima facie by main group system H–C–O–N–S–Se.

Among the transition metal bioelements we find some of the most crucial, indispensible catalysts of central anaerobic biochemistry. Iron in the form of moderately soft ferrous ions has diverse biochemical functions. It is the most abundant transition metal in aqueous, anaerobic, volcanic-hydrothermal settings. In the same settings, hydrogen sulfide (H2S), the source for soft thio ligands is a ubiquitous volcanic gas. These two locally coinciding components have a high bonding affinity for each other, as evident from the abundance of iron–sulfur clusters in extant metalloenzymes [13] and of iron–sulfur minerals in extant volcanic-hydrothermal flow sites. Therefore, the world of the pioneer organism has been dubbed “iron–sulfur world” [7]. The Fe–S bonding strength under anaerobic conditions serves as a gauge for the suitability of other transition metal bioelements, notably Co and Ni. These three iron group metals form the catalytic core of the pioneer organism.

Cu and Mo were unavailable for the anaerobic pioneer organism, because they form extremely insoluble sulfides (Cu2S and MoS2). They could have entered the biosphere only after oxygenation of the oceans. Zn is also discounted. Under sulfidic conditions, it forms highly insoluble ZnS and it is not redox-active. Mn (as Mn2+) is too hard for the iron–sulfur world. Cr, the group companion of Mo, exists under volcanic-hydrothermal conditions as hard cation (Cr3+) and had no chance to enter early biochemistry. Compared to its extremely low overall abundance in the Solar System, W is highly enriched in the walls of hydrothermal flow ducts and it does not form an extremely insoluble sulfide [14]. It would have been available for the pioneer organism. Vanadium has a remarkable chemical similarity to its diagonal neighbor Mo [15] and a high crustal abundance without being trapped as an insoluble sulfide. We conclude that the pioneer organism was catalytically defined prima facie by the transition metal system Fe–Co–Ni–W–(V).

1.3 Retrodicting Pioneer Catalysis

Extant biocatalysis is dominated by enzymes. These classify into metalloenzymes and nonmetalloenzymes. Nonmetalloenzymes require a large number of weak group interactions to fold and to stabilize transition states. This means high sequence accuracy, that is, late evolutionary arrival. The accuracy need of metalloenzymes is more relaxed. They typically exhibit a few strong coordination bonds to transition metals, reacting molecules, or transition states in addition to weak group interactions. Effectiveness of such coordination bonds is relatively insensitive to sequence variations. Therefore, a few coordination bonds may have been sufficient for protein folding and enzyme catalysis at the beginning of translation.

Next we note that evolutionary variability is not uniform throughout the structure of a metalloenzyme. Variability increases and invariance decreases from the (innermost) transition metal through the (inner) ligators to the (outer) ligand moieties. Hence, in the course of evolution central transition metals and ligators, once established remained invariant, with rare replacements (Fe → Mo, V in nitrogenases; W → Mo in tungsto/molybdopterins; Fe → Co, Ni in tetrapyrroles; S → Se or S → O in Fe–S clusters). The outer protein ligands evolved to modulate the catalytic properties of otherwise invariant transition metals and ligators.

We now apply the heuristic of biochemical...