Combined optical and resonance Raman studies have revealed the formation of an O2-adduct upon exposure of 4e- chemically reduced stearoyl-acyl carrier protein Delta9 desaturase to stearoyl-ACP and 1 atm O2. The observed intermediate has a broad absorption band at 700 nm and is remarkably stable at room temperature (t1/2 approximately 26 min). Resonance Raman studies using 16O2 gas reveal vibrational features of a bound peroxide [Vs(Fe-O2), 442 cm-1; Vas(Fe-O2), 490 cm-1; V(O-O), 898 cm-1] that undergo the expected mass-dependent shifts when prepared in (16)O(18)O or 18(O2). The appearance of two Fe-O2 vibrations, each having a single peak of intermediate frequency with 16(O)18(O), provs that the peroxide is bound symmetrically between the two iron atoms in a mu-1,2 configuration. The same results have been obtained in the accompanying resonance Raman study of ribonucleotide reductase isoform W48F/D84E [P. Moënne-Loccoz, J. Baldwin, B. A. Ley, T. M. Loehr, and J. M. Bollinger, Jr. (1998) Biochemistry 37, 14659-14663], thus making it likely that other members of the class II diiron enzymes form related peroxodiferric intermediates. Study of the reactivity of peroxodiferric Delta9D revealed that this intermediate underwent 2e- reduction leading to an oxidase reaction and recovery of the resting ferric homodimer. In contrast, biological reduction of the same enzyme preparations using ferredoxin reductase and [2Fe-2S] ferredoxin gave catalytic desaturation with a turnover number of 20-30 min-1. The profound difference in catalytic outcome for chemically and enzymatically reduced Delta9D suggests that redox-state dependent conformational changes cause partition of reactivity between desaturase and oxidase chemistries. The Delta9D oxidase reaction represents a new type of reactivity for the acyl-ACP desaturases and provides a two-step catalytic precedent for the "alternative oxidase" activity recently proposed for a membrane diiron enzyme in plants and trypanosomes.

No abstract

No abstract

The stearoyl-ACP delta 9 desaturase from plants is a new example of a growing number of proteins that contain oxo- or hydroxo-bridged diiron clusters. On the basis of differences in primary sequence motifs providing the cluster ligands and upon structural differences elucidated by X-ray crystallography, we now propose that the presently known, soluble diiron-oxo proteins can be grouped into two classes, I and II. Class I contains hemerythrin, myohemerythrin, and, possibly, purple acid phosphatase. Class II contains ribonucleotide reductases, bacterial hydrocarbon hydroxylases (methane monooxygenase, toluene-4-monooxygenase, and phenol hydroxylase), rubrerythrin, and stearoyl-ACP desaturases. Through the use of resonance Raman spectroscopy, we have detected symmetric (vs = 519 cm-1) and asymmetric (vas = 747 cm-1) vibrational modes in the castor stearoyl-ACP delta 9 desaturase, which are typical of oxo-bridged diiron clusters. These frequencies shift by -18 and -34 cm-1, respectively, in H218O, proving that the bridging ligand is readily exchangeable with solvent (t1/2 = 7 min). Calculation of an approximately 123 degrees Fe-O-Fe angle from the position of vs and vas and from the 18O-dependent shift in these frequencies suggests that the diiron-oxo cluster in the desaturase is triply bridged in the diferric state. In the diferrous state, the two iron sites of the cluster are structurally inequivalent, as shown by differential temperature dependence of the Mössbauer quadrupole splittings. For the class II diiron-oxo proteins, primary sequence alignments reveal conserved amino acid residues which act as iron cluster ligands, participate in a hydrogen-bonding network, and are potentially involved in O2 binding and activation. Based on this conservation, a structural model for the stearoyl-ACP delta 9 desaturase active site is proposed that has strong similarity to both ribonucleotide reductase and methane monooxygenase. However, after single turnover of the diferous state with 18O2, 18O is not detected in the oxo bridge of the castor desaturase. This is in contrast to the outcome observed for ribonucleotide reductase, suggesting the desaturase and ribonucleotide reductase differ in certain aspects of their respective O2-activation reactions.

No abstract