Brominating oxidants generated by human eosinophils.

Eosinophils are white blood cells that in humans are found in association with helminthic infections and various inflammatory disease processes. These cells contain a unique lysosomal peroxidase that oxidizes halides to generate highly reactive and toxic hypohalous acids. Although chloride is found in vivo at concentrations at least 1000-fold greater than those of other halides, human eosinophils did not preferentially oxidize chloride under physiologic conditions. Instead, eosinophils used bromide, a halide with a hitherto unknown function in humans, to generate a halogenating oxidant with characteristics similar, if not identical, to those of hypobromous acid. These results indicate that physiological concentrations of bromide arm human eosinophils with the ability to generate and release an unusual oxidant capable of destroying a wide range of prokaryotic and eukaryotic targets.

A rapid turbidimetric assay of phagocytosis and serum opsonizing capacity.

An in vitro assay has been developed to measure the opsonizing capacity of serum and the extent of bacterial uptake by phagocytes. Various micro-organisms were preopsonized for 10 min with a serum concentration previously determined to be optimal for the respective types of micro-organism. Subsequently, neutrophils from a healthy donor were added to the preopsonized bacteria in a cuvette of a spectrophotometer. The decrease in turbidity at 400 nm, resulting from the uptake of the micro-organisms by the neutrophils, was measured for 20-30 min and the area under the curves was taken as a measure of the opsonizing capacity of the serum or the phagocytic capacity of the neutrophils. The results correlated well with standard opsonophagocytic assays. By excluding Ca2+ from the buffer of the assay, phagocytosis was distinguished from the combined response of phagocytosis and aggregation. In the presence of Ca2+ ions, both phagocytosis and aggregation contributed to the decrease in turbidity. In the absence of Ca2+, phagocytosis was normal, but aggregation was completely inhibited. Phagocytosis in the absence of Ca2+ was also observed using microscopic and radiometric methods of evaluation. Neutrophils from a patient with a deficiency of leukocyte adhesion molecules, ingested as many bacteria as did normal neutrophils without Ca2+. Experiments with NaF, to inhibit phagocytosis, indicated that the change in turbidity measured in the absence of Ca2+ was mainly caused by phagocytosis, not by attachment of bacteria to the neutrophils. The opsonizing capacity of sera, as determined in our assay, depended both on antibodies and on an intact complement system and the inter-assay variance was less than 5%. We found a close correlation between turbidity changes measured in the presence or absence of Ca2+, suggesting that both phagocytosis and aggregation are opsonin-dependent. This assay is applicable to a variety of opsonizing fluids and micro-organisms, and can be used for assessing the phagocytic capacity of patients' neutrophils as well as for assessing the opsonizing capacity of patients' sera.

Killing of schistosomula by taurine chloramine and taurine bromamine.

Eosinophils generate hypochlorous acid when stimulated with opsonized particles. The hypochlorous acid can react with beta-amino acids such as taurine to produce chloramines with a long lifetime. In the presence of bromide, eosinophils generate hypobromous acid which can react with taurine to generate taurine bromamine. As taurine is abundantly present in leukocytes, eosinophils have the potential to generate taurine chloramine or taurine bromamine. With regard to the role of eosinophils in killing schistosomula, we have shown that both taurine chloramine and taurine bromamine in physiological concentrations are able to kill the schistosomula of Schistosoma mansoni.

Characterization of the interaction of human eosinophils and neutrophils with opsonized particles.

The interaction of human eosinophils with opsonized particles was compared with that of human neutrophils. When eosinophils are stimulated with serum-opsonized zymosan particles, the lag time in H2O2 production is twice as long as found with neutrophils. Moreover, the concentration of these IgG + C3-coated particles required for optimal stimulation is about four times as high for eosinophils as for neutrophils. Under these conditions, the two cell types generate similar amounts of H2O2. However, eosinophils produce twice as much H2O2 as do neutrophils when stimulated with the soluble agent phorbol myristate acetate. Thus, although the oxidase capacity of eosinophils is larger than that of neutrophils, opsonized zymosan is a weak trigger for this activity in eosinophils. This phenomenon may be due to differences between the two cell types in the plasma membrane receptors or in the receptor oxidase transducing signal. The following are indications for the first possibility. i) IgG interacts poorly with the Fc gamma receptors on the eosinophil surface compared with those on neutrophils. This was shown by the inability of IgG-coated zymosan or IgG-coated latex to trigger any substantial H2O2 production by eosinophils unless brought into close contact with these cells by centrifugation. In contrast, neutrophils are stimulated by these particles both in suspension and in a pellet. The dissimilarity of the Fc gamma receptors on eosinophils and neutrophils was also shown with respect to antigenicity, determined by the monoclonal antibodies 3G8 and CLB-FcR-1. ii) Eosinophils contain about half as many receptors for C3b and C3bi on their surface as do neutrophils, also detected with monoclonal antibodies. The interaction of IgG subclasses with functional Fc gamma receptors on eosinophils and neutrophils showed that eosinophils release twice as much H2O2 as do neutrophils upon interaction with IgG1-, IgG2-, or IgG3-coated Sepharose beads, but this difference becomes fivefold with IgG4-coated Sepharose. This might be of relevance to the situation of chronic antigenic stimulation, e.g., in chronic schistosomiasis, in which eosinophil numbers and IgG4 antibody levels are elevated.

Purification of eosinophils from normal human blood, preparation of eosinoplasts and characterization of their functional response to various stimuli.

Eosinophils from the blood of normal individuals were purified by centrifugation over discontinuous Percoll gradients. Eosinophil suspensions were obtained with a mean purity of 96% and a mean recovery of 64% (n = 19). When incubated with phorbol-myristate acetate, eosinophils consumed twice as much oxygen as did neutrophils from the same donors. With serum-treated zymosan, 70% and 100% of the maximal oxidative response (i.e. the response to phorbol-myristate acetate) was obtained with eosinophils and neutrophils, respectively. The calcium ionophore A23187 is a weak stimulus that triggered only 2.5% of the eosinophil and 10% of the neutrophil oxidative capacity. The response of both cell types to formyl-methionyl-leucyl-phenylalanine (fMLP) was rapid, with a maximum after 3 min. The magnitude of this eosinophil reaction was half that of neutrophils. Although the activities of the granule enzymes beta-glucuronidase and arylsulphatase were 2.5 and 6 times higher in eosinophils than in neutrophils, respectively, the exocytosis of these enzymes in response to various stimuli was lower in eosinophils. The high yield of eosinophils from our separation method enabled us to prepare eosinoplasts by centrifugation of eosinophils over discontinuous Ficoll gradients that contained cytochalasin B. Eosinoplasts are plasma membrane vesicles derived from eosinophils, filled with cytoplasm but devoid of granules and nucleus. The eosinoplasts contained 30% of the cytoplasm and plasma membrane present in intact eosinophils. Eosinoplasts still possessed a functionally intact oxidase enzyme that could be stimulated with various stimuli. Therefore, eosinoplasts may provide a valuable tool to study separately the role of the oxidase products and that of the granule contents in eosinophil functions.

Eosinophils do respond to fMLP.

Eosinophils were isolated from normal human blood by separation over Percoll gradients, which resulted in eosinophil suspensions of a purity higher than 95% and recoveries of about 65%. Normal human eosinophils were found to respond to formyl-methionyl-leucyl-phenylalanine (fMLP) at concentrations greater than 10(-7) mol/L with an increase in the concentration of intracellular free calcium, oxygen consumption, nitroblue tetrazolium reduction, and chemiluminescence. The maximal response of eosinophils to fMLP was lower than that of neutrophils isolated from the same blood samples and required at least ten times as much fMLP as was needed for neutrophils. Low fMLP concentrations (approximately 10(-8) mol/L), which in themselves did not stimulate O2 consumption by either eosinophils or neutrophils, primed these cells to respond to a suboptimal concentration of another stimulus. Purification of eosinophils after treatment of whole blood with fMLP showed that these eosinophils had lost their ability to respond to fMLP. We conclude that normal eosinophils do respond to fMLP and that therefore fMLP should not be used to isolate eosinophils.

Bactericidal action of eosinophils from normal human blood.

The ability of normal human eosinophils to ingest and kill Staphylococcus aureus and Escherichia coli was investigated and compared with the reactions shown by neutrophils from the same donors. The rate of phagocytosis of S. aureus by eosinophils was 50% of that shown by neutrophils. Unlike neutrophils, eosinophils were not able to kill ingested S. aureus at low bacterium/phagocyte ratios. The degree of S. aureus killing increased with increasing ratios, being equal to that of neutrophils when bacterium/phagocyte ratios of about 15 were used. This was probably due to a better triggering of the eosinophil oxidase system at high bacterium/phagocyte ratios. The early kinetics of the association of bacteria with eosinophils, the perforation of the bacterial envelope and the inactivation of bacterial proteins, was monitored in the ML-35 mutant strain of E. coli. The association of E. coli with eosinophils was 70% of that with neutrophils. Eosinophils had only 25% of the capacity of neutrophils to perforate the E. coli envelope. E. coli loses its colony-forming ability when the bacterial envelope has been perforated, indicating that eosinophils also kill E. coli more slowly than do neutrophils. This was confirmed with a plating assay for colony formation. The perforation of E. coli is independent of peroxidase-mediated reactions. Hence, the defective bactericidal action of eosinophils is probably not related to the differences between myeloperoxidase and eosinophil peroxidase. On the other hand, the inactivation of bacterial proteins is peroxidase dependent and was also seen to occur to a lesser extent in eosinophils compared with neutrophils. We conclude that eosinophils ingest E. coli but only slowly perforate (kill) these bacteria and barely inactivate the bacterial enzymes. In contrast, neutrophils quickly ingest and perforate (kill) E. coli and quickly inactivate the bacterial enzymes.

Excretion of superoxide by phagocytes measured with cytochrome c entrapped in resealed erythrocyte ghosts.

Resealed erythrocyte membranes (ghosts) filled with (Fe3+)cytochrome c were used as an assay system to measure the release of superoxide (O-2) from human phagocytes into the incubation medium. Neutrophils, activated by either opsonized zymosan particles or the soluble stimulus phorbol myristate acetate, released O-2, which subsequently entered the ghosts and reduced (Fe3+)cytochrome c. This reaction was dependent on the time of incubation, the concentration of neutrophils, the concentration of stimulus, and the concentration of ghosts. The reaction was completely inhibited by superoxide dismutase and by 4,4'-diisothiocyano-2,2'-disulfonic acid, a specific blocker of anion channels in membranes. The reduction of (Fe3+)cytochrome c free in solution was about four times as fast as the reduction of (Fe3+)cytochrome c in the ghosts. Human eosinophils stimulated by phorbol myristate acetate reacted similarly to human neutrophils; the rate of O-2 production/cell was about twice as high for eosinophils as for neutrophils. In contrast, eosinophils stimulated with opsonized zymosan particles only reduced (Fe3+)cytochrome c free in solution, but not (Fe3+)cytochrome c in ghosts. This lack of reaction was not due to production of an inhibitor or below threshold generation of O-2 for the ghost assay. These results indicate: 1) activated human neutrophils and eosinophils can release O-2 or a similar product into the incubation medium; and 2) reduction of (Fe3+)cytochrome c free in solution is no proof for O-2 excretion by phagocytes.