In vivo reduction of cyclopropane by Azotobacter vinelandii nitrogenase

Abstract

NITROGENASE from N₂-fixing aerobes such as Azotobacter vinelandii is stable to air in crude extracts but on fractionation into its two protein components, Fe–Mo protein and Fe protein it manifests the O₂ sensitivity characteristic of other nitrogenases¹. This difference between the air-stable and air-labile forms of the enzyme has been ascribed to various causes^2–4. More generally, the incompletely defined correlations between in vivo and in vitro nitrogenase function with respect to optimal component ratio⁵, local pH⁶, ATP stoichiometry⁷, reductant⁸, catalytic behaviour^4,9 and other important enzymatic parameters are a source of ambiguity in biochemical studies on the cell-free enzyme. In principle, effects from changes in many of these parameters can be studied by parallel in vivo and in vitro substrate reduction and inhibition studies. In practice¹⁰, restrictive limitations are imposed by the paucity of nitrogenase substrates that are at once (1) well bound, (2) capable of diagnostically complex chemical interactions, and (3) compatible with function of the enzyme in living microorganisms. For example, among known substrates other than N₂ itself, C₂H₂ is deficient in reduction complexity¹⁰, HCN is excluded by its cytotoxicity and acrylonitrile, which displays interesting product patterns¹¹, is a poor substrate¹². Recently, cyclopropene was shown to undergo ATP-dependent nitrogenase-catalysed reduction to propene and cyclopropane in vitro¹³. The product ratio of ∼2:1 alkene:cycloalkane was found to be insensitive to purification over a specific activity range of 40–1,200 in conventional assay conditions. We report here results of further nitrogenase-cyclopropene investigations which establish that: (1) nitrogenase in living A. vinelandii catalyses cyclopropene reduction to the products observed in vitro in the same ∼2:1 ratio; the K_m for formation of either product is very similar if not identical to that obtained with cell-free enzyme; (2) with purified components (specific activity up to 2,000) at optimal stoichiometry, the K_m for formation of the reduction products (propene and cyclopropane) is 0.01 atm, corresponding to a molar K_m as low as that of N₂ (and C₂H₂). These findings strengthen the case for catalytic similarity between in vivo and in vitro nitrogenase. Conversely, they substantiate the importance of cyclopropene product ratio as a chemical probe for the enzyme and as an important new criterion for nitrogenase biomimetic chemistry.