We have previously reported inhibition of cell-free activation of the neutrophil superoxidegenerating NADPH oxidase by a soluble cationic protein of neutrophil granules and by low concentrations of human defensin. Subcellular fractionation carried out in the current study indicated that the inhibitory substance was derived from azurophilic granules, was released into the medium on cell stimulation, and was resistant to phenylmethylsulfonyl fluoride (PMSF). Phorbol ester was the most effective stimulus for the release of the blocking activity. The possibility was raised that granule protein(s) act in vivo as negative modulators of superoxide production. Gel filtration of granule extract revealed a markedly retarded protein peak exhibiting oxidase-blocking activity and containing lysozyme as the main protein. Because lysozyme did not exert inhibitory effects on oxidase activation, association of the inhibitory protein with lysozyme was assumed. Indeed a column of immobilized lysozyme retained a fraction of the granule extract's oxidase-blocking activity. Elution with a low-pH buffer recovered a component capable of inhibition of the NADPH oxidase in stimulated neutrophils and in the cell-free system. The main 29-kDa protein band in the eluted fraction was identified as proteinase 3, a serine protease of azurophilic granules. Enzymatically active as well as PMSF-blocked conventionally purified proteinase 3 interfered with phorbol myristate acetateinduced superoxide release. These findings support the hypothesis that exocytosed granule constituents may prevent excessive activation of the NADPH oxidase.

Chromatography of human myeloperoxidase (MPO) on a heparin-agarose column demonstrated a tight association of the protein with the resin. The electrophoretic mobility of mixtures of MPO and heparin in polyacrylamide gels under nondenaturing conditions was consistent with a strong interaction of the cationic enzyme with the polyanionic polysaccharide. Purified MPO prebound to bovine aorta endothelial cells (BAEC) and supplemented with hydrogen peroxide dose- and time-dependently abrogated the interaction of coagulation factor IX (FIX) with factor IX-binding protein (FIXBP) on the surface of BAEC reflecting oxidative modification of the binding protein. This inactivation of FIXBP required the presence of chloride implicating hypochlorite in the reaction. Hypochlorite and activated neutrophils exerted a similar effect. The oxidative modification of FIXBP was only partially dependent on the addition of hydrogen peroxide and was abolished by exogenous heparin which displaced MPO from the cell surface, emphasizing the functional differences between cell-bound and free enzyme.

The effects of neomycin, fluoride and the non-hydrolysable guanine nucleotide analogue GTP gamma S on the kinetics of cell-free activation of NADPH oxidase in membranes of resting human neutrophils were investigated. Arachidonate-mediated activation of the oxidase followed a first-order reaction course (kobs. = 0.39 min-1 at 26 degrees C). In the presence of NaF during the activation process, activity was enhanced while the activation rate was slightly reduced (kobs. = 0.25 min-1 at 26 degrees C). Neomycin blocked activation (half-maximal effect at 25 microM) without affecting rates of superoxide release by preactivated enzyme in vitro or in vivo. In spite of reduced specific activity neither the first-order rate constant of the activation nor the Km of the oxidase were altered by neomycin. Oxidase activated in the presence of GTP gamma S exhibited increased specific activity and unchanged Km; the course of the reaction deviated from first-order kinetics. Kinetic evidence is presented for two separate activation reactions: a GTP gamma S-independent, basal, first-order process and a GTP gamma S-dependent sigmoid activation process. The results are compatible with the existence in neutrophil membranes of two separate pools of dormant oxidase. An alternative scheme of the formation of two active forms of NADPH oxidase is also presented.

Farnesylthiosalicylic acid (FTS), a synthetic analog of the terminal prenylcysteine present in signaling proteins induces generation of superoxide ions, phospholipase C-driven hydrolysis of inositol lipids and calcium elevation in human neutrophils and DMSO-differentiated HL60 cells. These effects were ascribed to an interaction of the analog with elements responsible for recognition of specific prenylated proteins. The present study demonstrated that in addition to the release of intracellular calcium stores, FTS enhanced entry of Ca(2+) and Mn(2+) from the medium. The biphasic dependence of the influx on the concentration of FTS, as well as its insensitivity to inhibition by PMA and La(3+) suggest that the influx pathway activated by FTS is distinct from the previously described store-operated calcium channels of neutrophils. Consistent with the participation of a cellular membrane component in the interaction, FTS enhanced (45)Ca uptake in neutrophils and neutrophil cell membranes, but not in multilamellar vesicles. To establish specificity of the farnesyl moiety of FTS (C(15)), effects of three other analogs, geranylthiosalicylate, GTS (C(10)), geranylgeranylthiosalicylate, GGTS (C(20)), as well as the carboxymethyl ester FTS-Me on calcium homeostasis and superoxide production were investigated. GGTS dose-dependently elevated [Ca(2+)](i), induced quenching of the 360 nm Fura-2-calcium fluorescence by Mn(2+) and stimulated superoxide release, while GTS and FTS-Me were inactive. These results defined specific structural requirements for the functional interaction of prenylcysteine analogs with myeloid cells.

The reduction with dithionite of neutrophil cytochrome b-558, implicated in superoxide generation by activated neutrophils, was investigated by a stopped-flow technique in non-ionic-detergent extracts of the membranes and in crude membrane particles. The dependence of the pseudo-first-order rate constants on the concentration of dithionite was consistent with a mechanism of reduction that involves the dithionite anion monomer SO2.- as the reactive species. The estimated second-order rate constant was 7.8 X 10(6) M-1 X S-1 for Lubrol PX-solubilized cytochrome b-558 and 5.1 X 10(6) M-1 X S-1 for the membrane-bound protein. The similarity of the kinetic constants suggests that solubilization did not introduce gross changes in the reactive site. Imidazole and p-chloromercuribenzoate, known as inhibitors of NADPH oxidase, did not affect significantly cytochrome b-558 reduction rates. The reaction rate of cytochrome b-558 with dithionite exhibited a near-zero activation energy. The first-order rate constant for reduction decreased with increasing ionic strength, indicating a positive effective charge on the reacting protein.