Position-Specific Isotope Fractionations Predicted by Molecular Modeling: Amino Acids

Molecular modeling calculations can be used to predict isotopic fractionation in geochemical systems. Recent advances in instrumentation make the prediction of position-specific, equilibrium isotope fractionation and isotopic clumping useful and necessary to contextualize observed isotope effects. We present calculations of equilibrium isotope compositions of amino acids, including position-specific differences, accomplished using density functional theory. For example, differences in site-specific compositions of C atoms within a single amino acid are predicted to reflect atomic oxidation state and bonding partners. At equilibrium, more oxidized atoms, atoms with more bonding partners, and atoms bonded to C, N, and O are predicted to be isotopically heavier, whereas reduced atoms, atoms with less bonding partners, and atoms bonded to H and S are predicted to be lighter. Thus, measurements that find light isotopes prefer specific sites are not by themselves evidence of kinetic fractionation. Organic compounds can seldom be expected to reach full isotopic equilibrium in biogeochemical or industrial systems, but portions of organic reactions are reversible and can be expected to approach isotopic equilibrium. A comparison of equilibrium predictions with measured position-specific effects in methionine reveals the rate-limiting step in industrial synthesis to be associated with lighter C isotopic composition than equilibrium. Position-specific calculations thus have the potential to reveal important details about reactions of organic compounds that could help elucidate the origins and transformations of organics from diverse geochemical sources. However, confounding factors of multiple reaction pathways, multiple reactant sources, and Rayleigh fractionation must also be considered.

Errors arise in these calculations from, among other sources, incompletely considering the importance of conformational variability and solvation. We present calculations which sample the variability in molecular and solvent conformation of aqueous amino acids. The calculations demonstrate that these factors have important secondary effects on the calculated predictions that must be considered when performing molecular modeling.