The Effects of Iron on In Silico Simulated Abiotic Reaction Networks

Abstract

Iron is one of the most abundant elements in the Universe and Earth's surfaces, and undergoes a redox change of approximately 0.77 mV in changing between its +2 and +3 states. Many contemporary terrestrial organisms are deeply connected to inorganic geochemistry via exploitation of this redox change, and iron redox reactions and catalysis are known to cause significant changes in the course of complex abiotic reactions. These observations point to the question of whether iron may have steered prebiotic chemistry during the emergence of life. Using kinetically naive in silico reaction modeling we explored the potential effects of iron ions on complex reaction networks of prebiotic interest, namely the formose reaction, the complexifying degradation reaction of pyruvic acid in water, glucose degradation, and the Maillard reaction. We find that iron ions produce significant changes in the connectivity of various known diversity-generating reaction networks of proposed prebiotic significance, generally significantly diversifying novel molecular products by ~20%, but also adding the potential for kinetic effects that could allow iron to steer prebiotic chemistry in marked ways.

Keywords: Maillard reaction; chemical reaction networks; combinatorial chemistry; formose reaction; glucose; iron chemistry; iron-sulfur world; origins of life; prebiotic chemistry; pyruvic acid.

Figure 1

Overview of the workflow explored here. Generic organic transformations mediated by iron were collected from the experimental chemical literature and converted to machine readable form. These were then used to construct CRNR using the MØD software package. These modeled CRNRs were then analyzed with regard to their products and network topology using computational tools.

Figure 2

(A). proposed mechanism for the observed conversion of glyceraldehyde to lactic acid in the presence of mineral surface-bound iron (III) (mineral surface indicated by the orange box) adapted from [11]. In this reaction iron plays the role of catalyst, as it is not consumed or altered during the reaction; (B). A generic representation of the reaction mechanism presented in (A) in the MØD-based workflow. The MØD algorithm looks for compounds with substructures matching L and transforms the bonds between them as per the rule to produce the substructure in R. The context K is analogous but not identical to the “intermediate” of a conventional reaction mechanism. Here, bonds colored in blue represent bonds broken in the course or the reaction and those in green are bonds created by the reaction. Atom coloring is merely a guide for the eye. Asterisks represent “wild card” atoms which may be of any specified type, allowing such mechanisms to be generally applied to any molecule containing the specified substructure. In this reaction mechanism Fe³⁺ is a catalytic species that must be present for the Fe³⁺-catalyzed reaction to proceed, but which is not altered or consumed in the reaction. An equivalent “uncatalyzed” reaction mechanism can be written lacking Fe³⁺.

Figure 3

Examples of iron-independent (A) and iron-dependent (B) reactions as encoded in MØD. (A). Aldol condensation reaction; (B). A two-electron oxidation of an aldehyde accompanied by the reduction of two iron (III) atoms to two iron (II) atoms.

Figure 4

Plots showing the measured and projected increase of products in four model reactions in the absence (blue) or presence (red) of iron species as a function of reaction generation. The hatched bars in plots C and D show the predicted number of compounds for generations beyond those computed.

Figure 5

Log number of products vs. reaction generation for each reaction for iron-containing (red) or iron-free (blue) CRNRs. Data points are computed values, dashed lines are best fits to the computed data points.

Figure 6

Modeled product counts for 15 generations of the iron-catalyzed formose reaction CRNR with an imposed upper mass limit of 100 AMU. After generation 13, no new compounds are discoverable in this network.

Figure 7

Reaction edges vs. reaction generation in the four model reactions in the absence (blue) or presence (red) of iron species. In plots C and D, hatched bars represent the predicted number of reactions in generations four and five.

Figure 8

Total (in- plus out-) degree comparisons the studied reaction networks. Blue bars: iron-free reactions, red bars: iron-containing reactions. With few exceptions, the order of the degree of comparison does not change significantly in the networks. The magnitude of the degree for the common compounds of the networks always increases in the iron-containing network. Note that the comparisons are drawn only for the ten highest-degree compounds of the iron-containing network. Refer to SI (Spreadsheet S2) for the degree of all products produced in all the networks.

Figure 9

Gephi visualization of modeled reaction networks. The left plots are iron-free networks and the right plots are the iron-containing networks for each reaction. Iron-utilizing reaction edges are colored in red.