The plant nonheme iron dioxygenase flavonol synthase performs a regioselective desaturation reaction as part of the biosynthesis of the signaling molecule flavonol that triggers the growing of leaves and flowers. These compounds also have health benefits for humans. Desaturation of aliphatic compounds generally proceeds through two consecutive hydrogen atom abstraction steps from two adjacent carbon atoms and in nature often is performed by a high-valent iron(IV)-oxo species. We show that the order of the hydrogen atom abstraction steps; however, are opposite of those expected from the C-H bond strengths in the substrate and determines the product distributions. As such flavonol synthase follows a negative catalysis mechanism. Using density functional theory methods on large active site model complexes we investigated pathways for desaturation and hydroxylation by an iron(IV)-oxo active site model. Against thermochemical predictions, we find that the oxidant abstracts the hydrogen atom from the strong C 2 -H bond rather than the weaker C 3 -H bond of the substrate first. We analyzed the origin of this unexpected selective hydrogen atom abstraction pathway and find that the alternative C 3 -H hydrogen atom abstraction would be followed by a low-energy and competitive substrate hydroxylation mechanism hence, should give considerable amount of by-products. Our computational modelling studies shows that substrate positioning in flavonol synthase is essential as it guides the reactivity to a chemo-and regioselective substrate desaturation from the C 2 -H group leading to desaturation products efficiently.

The nonheme iron enzyme ScoE catalyzes the biosynthesis of an isonitrile substituent in a peptide chain. To understand details of the reaction mechanism we created a large active site cluster model of 212 atoms that contains substrate, the active oxidant and the first- and second-coordination sphere of the protein and solvent. Several possible reaction mechanisms were tested and it is shown that isonitrile can only be formed through two consecutive catalytic cycles that both use one molecule of dioxygen and α-ketoglutarate. In both cycles the active species is an iron(IV)-oxo species that in the first reaction cycle reacts through two consecutive hydrogen atom abstraction steps: first from the N–H group and thereafter from the C–H group to desaturate the NH-CH2 bond. The alternative ordering of hydrogen atom abstraction steps was also tested but found to be higher in energy. Moreover, the electronic configurations along that pathway implicate an initial hydride transfer followed by proton transfer. We highlight an active site Lys residue that is shown to donate charge in the transition states and influences the relative barrier heights and bifurcation pathways. A second catalytic cycle of the reaction of iron(IV)-oxo with desaturated substrate starts with hydrogen atom abstraction followed by decarboxylation to give isonitrile directly. The catalytic cycle is completed with a proton transfer to iron(II)-hydroxo to generate the iron(II)-water resting state. The work is compared with experimental observation and previous computational studies on this system and put in a larger perspective of nonheme iron chemistry.

Ethylene is an important signaling molecule in plants that triggers the growth of leaves, flowers, and fruits. One of the enzymes involved in the biosynthesis of ethylene is the ethylene-forming enzyme (EFE), which is an usual nonheme iron enzyme that biodegrades α-ketoglutarate into three CO 2 molecules and ethylene. As the detailed mechanism of EFE in the biosynthesis of ethylene remains controversial and particularly the function of the co-substrate L-arginine, we decided to pursue a density functional theory study on the possible pathways of the enzyme leading to ethylene biosynthesis and test many possible pathways and mechanisms. A large active site cluster model of 322 atoms was created, which contains all the features of the firstand second-coordination sphere of the active site and substrate (αketoglutarate) binding pockets. The calculations identify a persuccinate intermediate that triggers a bifurcation pathway in the enzyme and either react with a molecule of CO 2 to form a carbonate or forms a high-valent iron(IV)-oxo species through heterolytic dioxygen bond cleavage. Our studies show that both the bifurcation pathways converge to the same intermediate again and can lead to ethylene products, although the two pathways have different kinetics. Interestingly, our studies also show that the iron(IV)-oxo itself can form a carbonate and ethylene but through much higher barriers. As a matter of fact, these barriers are higher in energy than the typical aliphatic hydroxylation barriers and may not be competitive with arginine hydroxylation. Inclusion of the L-arginine co-substrate into the model leads to minor changes in the structure and fold, but its charge and dipole moment does not seem to affect the first stage of the catalytic cycle. Moreover, the key activation barriers seem less affected by the inclusion of L-arginine into the model. We, therefore, believe that the role of L-arginine is to lock α-ketoglutarate and its products into a tight binding pocket to enable its degradation and to prevent early release of CO 2 . Our studies show that due to the distinct differences in α-ketoglutarate positioning between different arginine activating nonheme iron dioxygenases in the co-substrate binding pocket and its tighter binding in EFE, we predict that the release of CO 2 is prevented in the first stage of the oxygen activation mechanism. This enables attack of CO 2 on a persuccinate complex to form carbonate products, leading to ethylene formation. This work gives suggestions on the engineering of EFE into a hydroxylase or improving the ethylene biosynthesis.

Plants produce flavonol compounds for vital functions regarding plant growth, fruit and flower colouring as well as fruit ripening processes. Several of these biosynthesis steps are stereo-and regioselective and are being carried out by nonheme iron enzymes. Using density functional theory calculations on a large active site model complex of flavanone-3hydroxylase (FHT), we established the mechanism for conversion of naringenin to its dihydroflavonol, which is a key step in the mechanism of flavonol biosynthesis. The reaction starts with dioxygen binding to the iron(II) centre and a reaction with -ketoglutarate cosubstrate gives succinate, an iron(IV)-oxo species and CO 2 with large exothermicity and small reaction barriers. The rate-determining reaction step in the mechanism; however, is hydrogen atom abstraction of an aliphatic C-H bond by the iron(IV)-oxo species. We identify a large kinetic isotope effect for the replacement of the transferring hydrogen atom by deuterium. In a final step the OH and substrate radicals combine to form the alcohol product with a barrier of several kcal mol-1. We show that the latter is the result of geometric constraints in the active site pocket. Furthermore, the calculations show that a weak tertiary C-H bond is shielded from the iron(IV)-oxo species in the substrate binding position and therefore the enzyme is able to activate a stronger C-H bond. As such, the flavanone-3hydroxylase enzyme reacts regioselectively with one specific C-H bond of naringenin by avoiding activation of weaker bonds through tight substrate and oxidant positioning.