Reactive oxygen species are classically described as occurring as an accidental byproduct of respiration, and are generally thought to be deleterious to biologic systems. The phagocyte nicotinamide adenine dinucleotide phosphate oxidase provides an example of deliberate reactive oxygen species generation, but the function of this enzyme is to oxidatively modify bacteria as part of bactericidal mechanisms. The discovery of a family of nicotinamide adenine dinucleotide (phosphate) oxidases related to the phagocyte oxidase, the Nox/Duox family, provides additional examples of deliberate generation of reactive oxygen species. This article describes this new family of enzymes and considers hypotheses for their function. Potential roles of Nox/Duox in generation of reactive oxygen species that function in cell signaling (related to growth and angiogenesis), immune function, hypoxic response, and oxidative modification of extracellular matrix proteins are discussed.

The flavoprotein NADPH-adrenodoxin reductase and the iron sulfur protein adrenodoxin function as a short electron transport chain which donates electrons one-at-a-time to adrenal cortex mitochondrial cytochromes P-450. The soluble adrenodoxin acts as a mobile one-electron shuttle, forming a complex first with NADPH-reduced adrenodoxin reductase from which it accepts an electron, then dissociating, and finally reassociating with and donating an electron to the membrane-bound cytochrome P-450 (Fig. 9). Dissociation and reassociation with flavoprotein then allows a second cycle of electron transfers. A complex set of factors govern the sequential protein-protein interactions which comprise this adrenodoxin shuttle mechanism; among these factors, reduction of the iron sulfur center by the flavin weakens the adrenodoxin-adrenodoxin reductase interaction, thus promoting dissociation of this complex to yield free reduced adrenodoxin. Substrate (cholesterol) binding to cytochrome P-450scc both promotes the binding of the free adrenodoxin to the cytochrome, and alters the oxidation-reduction potential of the heme so as to favor reduction by adrenodoxin. The cholesterol binding site on cytochrome P-450scc appears to be in direct communication with the hydrophobic phospholipid milieu in which this substrate is dissolved. Specific effects of both phospholipid headgroups and fatty acyl side-chains regulate the interaction of cholesterol with its binding side. Cardiolipin is an extremely potent positive effector for cholesterol binding, and evidence supports the existence of a specific effector lipid binding site on cytochrome P.450scc to which this phospholipid binds.

NADPH-dependent superoxide generation can be reconstituted in a cell-free system using recombinant cytosolic factors (p47-phox, p67-phox, and Rac) plus flavocytochrome b558. Rac1 and Rac2 are closely related small GTPases, differing primarily in the C-terminal 10 residues where Rac1 but not Rac2 contains a polybasic sequence. In their nonisoprenylated forms, Rac1 was highly effective in reconstituting NADPH oxidase activity (low EC50, high Vmax), whereas Rac2 was only minimally effective (high EC50, low Vmax). In contrast, low concentrations of isoprenylated Rac1 and Rac2 both supported high rates of superoxide generation. Like full length Rac2, truncated forms of both Rac1 and Rac2 in which the C-terminal 10 residues were eliminated were poorly activating, pointing to the C terminus of Rac1 as a determinant of activity. Mutation of single positively charged residues in the C terminus of nonisoprenylated Rac1 markedly reduced its ability to support superoxide generation, affecting both its EC50 and the Vmax. In contrast, mutation or truncation of the C terminus failed to affect the activation of PAK, a Rac-regulated protein kinase. The EC50 for Rac1 increased with increasing salt concentrations, whereas that of Rac2 was independent of salt, implicating the involvement of electrostatic forces for the former. Using flavocytochrome b558 reconstituted into phosphatidylcholine vesicles, the EC50 for Rac1 but not Rac2 decreased (increased binding) when an acidic phospholipid (phosphatidylinositol) was present, supporting a role for the Rac1 polybasic C terminus in binding to the membrane. A model in which Rac must associate simultaneously both with p67-phox and with the membrane to activate the NADPH oxidase can account for the above observations.

No abstract

We reported previously that diacylglycerol (diC8) and GTP gamma S synergize with an anionic amphiphile such as sodium dodecyl sulfate (SDS) to produce high rates of superoxide generation in a cell-free system consisting of neutrophil plasma membrane plus cytosol [Burnham, D. N., Uhlinger, D. J., & Lambeth, J. D. (1990) J. Biol. Chem. 265, 17550-17559]. Here we investigate the effects of these activating factors on the plasma membrane association in an in vitro translated radiolabeled recombinant p47-phox protein. Apparent translocation, assayed by cosedimentation with plasma membranes, required the presence of excess cytosol and an anionic amphiphile, was enhanced by both GTP gamma S and diC8, and was inhibited by high salt, correlating qualitatively with activation; up to 70% cosedimentation was observed with the combination of activators (compared with less than 20% in their absence). Similar results were obtained using heat-inactivated cytosol, wherein another oxidase component, p67-phox, has been inactivated. Unexpectedly, from 50 to 80% of the apparent translocation occurred in the absence of membranes, indicating that protein aggregation accounted for a significant part of the observed translocation. Nevertheless, the percent translocation was increased in all cases by the presence of membranes, indicating some degree of protein-membrane interaction. While a control in vitro translated protein failed to translocate, cosedimentation of p47-phox occurred equally well when red blood cell or neutrophil plasma membranes lacking cytochrome b558 were used. Also, the peptide RGVHFIF, which is contained within the C-terminus of the large subunit of cytochrome b558, failed to inhibit translocation/aggregation of p47-phox, despite its ability to inhibit cell-free activation of the oxidase. The data are consistent with the following: (a) SDS, diC8, and GTP gamma S all act on cytosolic components to alter protein-protein and/or protein-membrane associations, and these changes are necessary (but not sufficient) for activation; (b) these altered associations are likely to function by increasing the local concentration of p47-phox and other components at the plasma membrane; (c) a high background of nonspecific associations in the cell-free activation system is likely to obscure any specific, functionally relevant associations (e.g., with cytochrome b558); and (d) the mechanism of translocation in the cell-free system differs from that seen in intact neutrophils.