Manfred L. Karnovsky

Deficiency of Reduced Nicotinamide-Adenine Dinucleotide Oxidase in Chronic Granulomatous Disease

Reduced nicotinamide-adenine dinucleotide oxidase of normal human polymorphonuclear leukocytes has properties that would qualify it as the enzyme responsible for the respiratory burst during phagocytosis. The enzyme was deficient in leukocytes of five patients with chronic granulomatous disease. This lack of adequate reduced nicotinamide-adenine dinucleotide oxidase could be the basis for the metabolic abnormalities characteristic of these leukocytes and for their diminished bactericidal activity.

Degranulation of leukocytes in chronic granulomatous disease

Quantitative chemical analyses of the subcellular distribution patterns for acid and alkaline phosphatase, beta glucuronidase and peroxidase were obtained for human peripheral blood leukocytes of four patients with chronic granulomatous disease (CGD). Five young adults with acute infections served as controls. The observations were made on fractions obtained by homogenization and centrifugation of leukocytes previously incubated with or without particles for ingestion. Distributions in resting CGD and normal cells were very similar for acid and alkaline phosphatase and peroxidase, but the proportion of beta glucuronidase in the granule fraction of CGD cells was depressed, with an increased proportion in the soluble fraction. Release of granule-bound enzymes during phagocytosis of a variety of particles was the same for CGD and control cells, except that release of beta glucuronidase was less marked in CGD cells. Total enzymatic activity of CGD cells for the hydrolases studied was normal. The data indicated that granular enzymes are released in a normal fashion in phagocytizing CGD cells. Supportive evidence of release of enzymes into the phagocytic vacuole of CGD cells was obtained by an electron microscopic study of myeloperoxidase.

Respiration and glucose oxidation in human and guinea pig leukocytes: comparative studies

A comparison has been made of the metabolic shifts in human and guinea pig leukocytes when they phagocytize. Respiration of guinea pig polymorphonuclear leukocytes (PMN) and the increment during phagocytosis were each about 2(1/2)-fold that of human PMN. This was also true of the direct oxidation of glucose-6-P (hexose monophosphate shunt). Enzymes potentially responsible for these phenomena have been compared in each species. Cyanide-insensitive NADH oxidase and NADPH oxidase were measured and only the formed exhibited adequate activity to account for the respiratory stimulus durintg phagocytosis. The hydrogen peroxide formed by this enzyme stimulates the hexose monophosphate shunt by oxidizing glutathione which upon reduction by an NADPH-linked glutathione reductase provides NADP to drive the hexose monophosphate shunt. Other linkages between respiratory stimulation and that of the hexose monophosphate shunt also pertain in the guinea pig.

Fluoride-mediated activation of the respiratory burst in human neutrophils. A reversible process.

Fluoride ion (F-) is an effective activator of the respiratory burst in neutrophils, as indicated by its ability to induce O2- production by these cells. Other halide ions did not activate the burst, Cl-, in particular appeared to antagonize the effect of F- on O2- production. F- stimulated O2- production showed a requirement for Ca++, but was independent of other exogenous cations. Neither phagocytosis nor degranulation were necessary for respiratory burst activation by F-. The effect of F- on the respiratory burst was reversible. Washing the cells after treatment with F-, while they were still producing large amounts of O2-, returned them to the resting state. They could then be stimulated again to produce O2- in amounts equivalent to those originally produced. Our experiments indicated that restimulation did not represent the activation of a population of cells that had not been activated during the initial exposure to F-, nor did it represent serial activation of different subpopulation of the O2- forming enzyme molecules present in a given cell. Rather, our data suggest that the entire population of O2- forming enzyme molecules was activated in a reversible fashion by F-.

Superoxide release by neutrophils: Synergistic effects of a phorbol ester and a calcium ionophore*

Exposure to combined suboptimal concentrations of 4 beta-phorbol 12-myristate 13-acetate and the calcium ionophore A23187 stimulates superoxide release from guinea pig neutrophils to rates ca. 3.5-fold greater than the sum of the rates elicited by each of the agents added separately. This effect was largely dependent upon the presence of calcium in the extracellular medium. The data are discussed in relation to recent reports concerning the interactions of phorbol-esters with cells and the mechanism of activation of superoxide release by neutrophils.

Superoxide production by polymorphonuclear leukocytes

SummaryPhagocytosis by polymorphonuclear leukocytes triggers a burst of oxidative metabolism resulting in hydrogen peroxide and superoxide production, and these active oxygen species function in the killing of microorganisms. A new cytochemical technique, based on a manganese dependent diaminobenzidine oxidation, has been developed to detect superoxide in these cells. It has been shown that superoxide generation is associated with the plasma membrane in cells activated by particulate (zymosan) and nonparticulate (phorbol myristate acetate) stimuli. This membraned activity is maintained during invagination such that reduced oxygen is generated within the endocytic vacuoles. Reaction product is absent from unstimulated cells; additionally, formation of precipitate is blocked by omission of Mn++, low temperature, glutaraldehyde prefixation, and the presence of superoxide dismutase in the incubation medium.

Correction of metabolic deficiencies in the leukocytes of patients with chronic granulomatous disease

Polymorphonuclear leukocytes from patients with chronic granulomatous disease (CGD) exhibit metabolic and bactericidal deficiencies that may be the result of inadequate production of H(2)O(2). A hydrogen peroxide-generating system was, therefore, inserted into CGD leukocytes. This was accomplished by allowing the cells to phagocytize latex spherules coated with glucose oxidase. This produced an amelioration in the known metabolic deficiencies of these cells during phagocytosis: (a) intracellular (catalatic) formate oxidation dependent upon hydrogen peroxide production was enhanced fourfold; and (b) hexose monophosphate shunt activity, which other workers have shown to be at least partially dependent upon the availability of H(2)O(2), was markedly stimulated. These data strengthen the evidence that the fundamental metabolic lesion in CGD cells during phagocytosis is indeed deficient production of hydrogen peroxide, probably, as previously shown, due to diminished oxidase for reduced nicotinamide adenine dinucleotide.

BIOCHEMICAL CHARACTERISTICS OF ACTIVATED MACROPHAGES

In the inflammatory processes that occur in rheumatoid arthritis, it is clear that granulocytes play a large role. Macrophages are also present, and the importance of their contribution will presumably be fully comprehended only when their physiologic characteristics in the particular environment are defined. They may be activated to varying degrees. Specific activation is due to phenomena with an immunologic basis; in addition, macrophages may be stimulated by nonspecific means. The definition of “activated” macrophages is a functional one, which indicates that such cells are more efficient with respect to their overall phagocytic and/or antimicrobial activities than are control macrophages. “Activated” macrophages were described mainly by Mackaness in the 1960~. ’ -~ The activation observed has, as mentioned, a specific immunologic basis and is mediated by lymphocytes, but the activated cells manifest their enhanced microbicidal action nonspecifically : that is, they exhibit an augmented capacity to deal not only with the original “activating” organism but also with other organisms. One organism often used for activation of macrophages is Listeria monocytogenes. About 1 week after infection of mice with this bacterium, the peritoneal macrophages may be harvested in markedly greater quantities than is normally the case. They spread out on glass more than do normal macrophages, and they are more phag~cyt ic .~ It is not yet clear whether their enhanced antimicrobial action is due simply to the greater phagocytic ability or whether they have an armamentarium of antimicrobial agents or systems that is qualitatively or quantitatively different from that of control cells. In the context of this Conference, manifestation of antimicrobial potentiat is of minor importance-the capacity to ingest (phagocytize or pinocytize) and destroy tissue components is the important issue. Of concern is the possible activation of macrophages by lymphocytes sensitized to tissue components or their degradation products. Further, macrophages may also be directly (nonspecifically) activated by substances in the synovium, for example. The particular burden of this paper is to indicate some of the biochemical properties of specifically and nonspecifically activated macrophages as compared to control ce1ls.t In addition, one may also compare “resting” and

Manganese-dependent NADPH oxidation by granulocyte particles. The role of superoxide and the nonphysiological nature of the manganese requirement.

Recent work has indicated that superoxide is involved in the manganese-stimulated oxidation of NADPH by crude granule preparations of guinea pig neutrophils. The characteristics of a model manganese-requiring NADPH-oxidizing system that employs a defined O2-generator have now been compared to the original neutrophil-granule system. With respect to pH dependence, cyanide sensitivity, and reduced pyridine nucleotide specificity, the properties of the two systems are very similar. Additional information has been obtained concerning cation specificity and the kinetics of the metal-catalyzed NADPH oxidation. From the similarities between the properties of the model and neutrophil particle systems, we postulate that the manganese-dependent NADPH oxidation observed in the presence of neutrophil granules represents in large part of nonenzymatic free radical chain involving the oxidation of NADPH to NADP, with O2- as both the chain initiator and one of the propagating species. In this reaction, the neutrophil particles serve only as a source of O2-. Further, the same changes in kinetics (decrease in apparent Km for NADPH) observed previously when granules from phagocytizing rather than resting cells were employed could be mimicked by varying the rate of O2-generation by the model system. We conclude from these results that it is unnecessary to invoke a manganese-requiring enzyme as a component of the phagocytically stimulated respiratory system of the neutrophil.

Hydrogen Peroxide Production in Chronic Granulomatous Disease: A CYTOCHEMICAL STUDY OF REDUCED PYRIDINE NUCLEOTIDE OXIDASES

The ability of polymorphonuclear leukocytes (PMN) to produce H(2)O(2) in response to phagocytic stimulation was examined cytochemically using leukocytes from normal individuals and patients with chronic granulomatous disease (CGD). Normal PMN oxidized diaminobenzidine within the phagocytic vacuole by a reaction dependent upon endogenous H(2)O(2) and myeloperoxidase. CGD PMN failed to oxidize diaminobenzidine, which is consistent with the biochemical data showing a lack of H(2)O(2)-generating capacity. A plasma membrane enzyme (oxidase) activated by phagocytosis is capable of H(2)O(2) production in PMN. The localization of this oxidase activity was explored in CGD PMN using a cytochemical technique specific for H(2)O(2). The enzyme activity is stimulated by exogenous NADH, but not NADPH. Reaction product formation, indicative of activity of the oxidase, is dependent upon precipitation of cerium ions by the enzymatically generated H(2)O(2). The advantage of this approach is that enzyme activity of individual cells can be assessed, allowing determination of numbers of reactive cells in the population and their relative degrees of reactivity. NADH oxidase was found to be active both on the plasma membrane and within the phagocytic vacuoles of control PMN, whereas those cells from three CGD patients showed greatly reduced activity in both these sites. Assessment of the reactivity of individual cells showed the number of cells with oxidase activity in CGD to be significantly reduced when compared to control values. Additionally, of those cells that do react, a higher percentage of them are only weakly reactive. Omission of NADH from the incubation medium reduced the percentage of control cells showing enzyme activity but had no effect on CGD PMN, implying that the enzyme is not saturated with substrate in control cells, but in CGD the diminished enzyme is fully saturated. The defect may lie in the fact that in CGD patients there are fewer cells capable of peroxide generation, and a majority of these reactive cells produce only reduced amounts of this bactericidal agent.