Prebiotic Reaction Networks in Water

Abstract

A prevailing strategy in origins of life studies is to explore how chemistry constrained by hypothetical prebiotic conditions could have led to molecules and system level processes proposed to be important for life's beginnings. This strategy has yielded model prebiotic reaction networks that elucidate pathways by which relevant compounds can be generated, in some cases, autocatalytically. These prebiotic reaction networks provide a rich platform for further understanding and development of emergent "life-like" behaviours. In this review, recent advances in experimental and analytical procedures associated with classical prebiotic reaction networks, like formose and Miller-Urey, as well as more recent ones are highlighted. Instead of polymeric networks, i.e., those based on nucleic acids or peptides, the focus is on small molecules. The future of prebiotic chemistry lies in better understanding the genuine complexity that can result from reaction networks and the construction of a centralised database of reactions useful for predicting potential network evolution is emphasised.

Keywords: RNA; autocatalysis; complexity; protometabolism; systems chemistry.

Figure 1

Proposed mechanisms and data analysis plots for the Miller-Urey experiment. (A) Two pathways proposed for amino acid production, namely Strecker and Bucherer-Bergs syntheses, which rely on the formation of α-aminonitriles. While the Strecker synthesis relies on slow hydrolysis of aminonitriles, Bucherer-Bergs relies on the more efficient hydrolysis of hydantoins formed from reaction between CO₂ and aminonitriles. (B) Van Krevelen diagram of Miller-Urey products categorised by their H/C versus O/C ratios. (C) Kendrick maps displaying homologous series of CH₂, HCN, NH, and CO. Plots in (B,C) reprinted with permission from reference [66].

Figure 2

A proposed autocatalytic loop that revolves around the formamide-catalysed hydration of HCN. Formamide, first produced from HCN hydration, can subsequently serve as a catalyst for its own production [79].

Figure 3

Simplified scheme of borate-mediated formose reaction proposed by Benner and co-workers [50,96,97]. The main reaction sequence produces aldoses and ketoses of increasing complexity following a series of aldol additions with formaldehyde starting from glycolaldehyde. Branched pentoses 12 and 13, which cannot further enolise, undergo retro-aldol fragmentation to produce glycolaldehyde and glyceraldehyde or molybdate-catalysed Bilik conversion (reaction not shown) to linear ketoses 9 and 10, respectively. In presence of molybdate, ribulose equilibrates to ribose. B: borate. Compounds are depicted as d- or l-isomers but exist as racemic mixtures.

Figure 4

Geochemical scenario for mutual synthesis of purine and pyrimidine ribonucleotides proposed by Carell and co-workers [105]. An initial mixture of cyanoacetylene 14, (hydroxyimino)malononitrile 16, hydroxylurea 15, and methylthioamidine 17 goes through a series of wet-dry cycles which relies on the separate delivery of urea and formic acid via rain as well as Zn/Zn²⁺ chemistry to produce pyrimidine and purine nucleobase precursors 20–22. A stream carrying these precursors then merges with another carrying ribose, and a subsequent wet-dry cycle drives the coupling reactions. The figure image was redrawn based on Figure 5B of reference [105].

Figure 5

Activated pyrimidine ribonucleotide synthesis demonstrated by Sutherland and co-workers [108]. The first step involves the reaction between glycolaldehyde 4 and cyanamide 26, which after cyclisation, yields 2-aminooxazole 27. d-glyceraldehyde 5d reacts with 27 to form d-arabinose aminooxazoline 28 among other stereoisomers (not shown). Compound 28 further reacts with cyanoacetylene 14 to yield 2,2′-anhydrocytidine 29. Phosphorylation of 29 yields cytidine-2′,3′-cyclic phosphate 30, which can undergo UV-promoted deamination to form uridine-2′,3′-cyclic phosphate 31.

Figure 6

Cyanosulfidic geochemical scenario proposed by Sutherland and co-workers [26,111,112]. (A) Reaction scheme for the cyanocuprate-photocatalytic synthesis of ribonucleotide and amino acid intermediates. Not shown are pathways for additional amino acid and phospholipid precursors. For more details, see reference [112]. (B) Scheme for arabinose aminooxazoline synthesis 28 via the Powner-Sutherland pathway from potential evaporites that could have yielded the necessary starting materials. Image redrawn based on Figure 2D of reference [112]. (C) A post-meteoritic impact geochemical scenario where streams produced from rainfall carry different starting materials derived from various evaporites as shown in (B). The image in (C) was reprinted with permission from reference [112].

Figure 7

Production of ribonucleotide precursors by γ-radiolysis of briny HCN solutions demonstrated by Fahrenbach and co-workers [51]. (A) Radiolysis of briny water produces the reducing and oxidising species required for the synthesis of glycolaldehyde 4 and cyanamide 26 which come together in the first step of the Powner-Sutherland pathway to produce 2-aminooxazole 27. (B) Scheme for the experimental procedure that starts with gamma radiolysis followed by dry-down and heating. Schemes were redrawn based on Figure 2 and Figure 3 from reference [51].

Figure 8

Iron-promoted reaction network capable of synthesis and breakdown demonstrated by Moran and co-workers [23]. The highly interconnected reaction network starting from glyoxylate 38 and pyruvate 39 produced 9 of 11 intermediates of the TCA cycle and 8 of 9 intermediates of the glyoxylate cycle. The addition of hydroxylamine and Fe⁰ to the reaction network affords reductive amination of glyoxylate, pyruvate, α-ketoglutarate, and oxaloacetate yielding glycine, alanine, glutamic acid, and aspartic acid, respectively. The scheme was redrawn based on Figure 1A of reference [23].

Figure 9

Linked cycles analgous to the TCA which rely on alternating additions of glyoxylate and H₂O₂ demonstrated by Springsteen, Krishnamurthy and co-workers [22]. The reaction network can be initiated by the aldol addition of glyoxylate 38 with either pyruvate 39, oxaloacetate 49 or malonate 52. Aspartic acid can be generated via the reaction of malonate and hydroxyglycine, formed from glyoxylate and ammonia, followed by decarboxylation in the presence of Mg²⁺. Scheme redrawn based on Figure 2 from reference [22].

Figure 10

Iron dependency of nonenzymatic reactions in a chemical network reminiscent of glycolysis and the pentose phosphate pathway. Reactions accelerated by iron are coloured with brown arrows while those slowed down are shown with grey arrows. Iron-independent reactions are shown with white arrows. Scheme redrawn based on Figure 4B from reference [146].

Figure 11

Autocatalytic chemical network consisting of biologically relevant reactions capable of bistability and oscillating behaviours demonstrated by Whitesides and co-workers [147]. (A) Scheme of the organic reactions that comprise the autocatalytic chemical system based on reference [147]. (B) The continuously stirred tank reactor (CSTR) used to study the emergent properties of the chemical system. (C) Hysteresis curve based on experimental steady-state RSH (free thiol) concentrations as a function of space velocities (normalised flow rates). (D) A kinetic model simulating oscillating concentrations of RSH, maleimide, and AlaSEt based on experimental observations. Images in (B–D) were reprinted with permission from reference [147].