Effect of docosahexaenoic acid on rate of differentiation of HL-60 human leukemia.

We have utilized an experimental model of cell lipid modification that allows study of the effect of a polyunsaturated fatty acid on the linked processes of cellular differentiation and growth arrest. HL-60 human leukemia cells were grown in media supplemented with 10 microM concentrations of the fatty acid docosahexaenoic acid (22:6) or oleic acid (18:1) or in unsupplemented media. Gas chromatographic analysis of phospholipid extracts from HL-60 cells grown in unmodified or 18:1-supplemented media revealed 39% and 36% 18:1, 13 and 12% polyenoics, and 2 and 3% 22:6, respectively. In contrast, cells from 22:6-supplemented cultures had 22% 18:1, 18% total polyunsaturated fatty acids, and 10% 22:6. Retinoic acid was added to cells grown in the various media, and phorbol ester-induced superoxide generation, nitroblue tetrazolium reduction, and growth arrest were determined as measures of differentiation. Unmodified and 18:1-enriched cells showed inducible oxidative burst activity beginning at 48 h after the addition of retinoic acid and continuing to increase for 5 days. In marked contrast, the 22:6-enriched leukemia cells exhibited an increased oxidative activity as early as 24 h which is equivalent to about one division cycle time. G1/0-specific growth arrest was associated with the oxidative phenotypic differentiation in all three cell types. However, cells enriched with 22:6 demonstrated early growth arrest and differentiation considerably in advance of 18:1-modified or unmodified cells. An effect on the cellular differentiation process could be detected after even a brief 1-h exposure of the cells to 22:6. Therefore, a highly polyunsaturated fatty acid which is actively incorporated into membrane structures appreciably accelerates the differentiation process of this human neoplastic cell.

Subcellular distribution of doxorubicin: comparison of fatty acid-modified and unmodified cells.

We have examined the subcellular localization of doxorubicin and evaluated the effect of fatty acid modification on specific intracellular localization. L1210 leukemia cells enriched with docosahexaenoic acid (22:6) or oleic acid (18:1) were incubated with radiolabeled or unlabeled doxorubicin. After equilibration the cells were ruptured and the subcellular fractions were isolated by differential centrifugation and sucrose gradient separation. The doxorubicin localized primarily in nuclei, as expected, but appreciable amounts were also detected in mitochondria and smaller amounts in plasma membranes, microsomes, and cytoplasm. Subcellular distribution of another anticancer drug which binds to DNA, mitoxantrone, was similar. There were increased amounts of doxorubicin contained in the nuclei and all organelles of the 22:6-enriched cells. Although polyunsaturated fatty acid modification influenced the total amount of doxorubicin in fractions, the relative distribution of drug among the fractions was not different from that of the 18:1-enriched and unmodified cells. We conclude that enrichment with polyunsaturates influences total drug uptake but not proportional distribution of doxorubicin.

Modification of the fatty acid composition of L1210 leukemia subcellular organelles.

We have examined the extent to which it is possible to modify the fatty acid composition of subcellular organelles of L1210 leukemia cells. A polyunsaturated fatty acid, docosahexaenoic acid, or a monounsaturated fatty acid, oleic acid, were added to the culture media. After 48 hr, the cells were ruptured and the subcellular fractions isolated. Fatty acid analysis revealed that nuclei, mitochondria, plasma membranes and microsomes of the cells grown in media supplemented with docosahexaenoic acid contained increased amounts of polyenoic fatty acids compared with cells grown in oleic acid. We conclude that it is possible to experimentally modify the lipids of multiple intracellular structures of L1210 cells by the addition of fatty acids to the growth media.

Membrane lipid alteration: effect on cellular uptake of mitoxantrone.

We have studied the effect of membrane structural alteration on the cellular association of the anticancer drug mitoxantrone whose uptake is not carrier-mediated. Membrane fatty acids of L1210 cells were modified by incubating the cells with the highly unsaturated docosahexaenoic acid (22:6), which results in isolated plasma membranes with 37% of the fatty acids as 22:6, or with the monounsaturated oleic acid (18:1), which results in 58% of the fatty acids as 18:1. The rate of uptake by 22:6-enriched cells during the first min was 62% greater than by those enriched with 18:1. The higher rate was recorded at 0.5-16 microM, pH 6.6-7.6 and temperatures 10-40 C. The difference in cell-associated drug apparently was not due simply to a change in mitoxantrone solubility as measured by partitioning of the drug in lipophilic-hydrophilic systems containing lipids from the fatty-acid altered cells. We conclude that the type of fatty acids contained in L1210 cell membranes can affect the cell association of mitoxantrone. This effect could be on transmembrane flux or be due to differences in binding of the drug to intracellular structures.

In vitro and in vivo effects of treatment by platelet-activating factor on N-formyl-met-leu-phe-mediated responses of polymorphonuclear leucocytes.

Two chemoattractants, the peptide N-formyl-met-leu-phe (FMLP), and the ether phospholipid, platelet activating factor (PAF), each stimulate a variety of in vitro responses in polymorphonuclear leucocytes (PMN). Because often more than one inflammatory mediator is active during inflammation, we determined the effect on PMN of sequential stimulation with these two agents. Before FMLP stimulation, human PMN were exposed to PAF, at concentrations which gave little or no response when administered alone. PAF enhanced FMLP-elicited superoxide release in a dose-dependent fashion. Likewise, release of granular lysozyme from the cells was increased in PAF treated cells. Similar treatment with other phospholipids, including the lyso derivation of PAF, failed to produced these effects. Incubation with nordihydroguaiaretic acid, an inhibitor of arachidonic acid metabolism, had little effect on the enhancement of lysozyme release by PAF. To determine if enhancing effects by PAF might occur also in vivo, we studied rabbits receiving PAF and/or FMLP intravenously. When rabbits received 0.01 micrograms PAF (a dose which does not elicit the sustained neutropenia observed with higher doses of PAF) followed by 0.05 micrograms FMLP the absolute granulocyte count (AGC) dropped at 1 min (46 +/- 11% of original value), and continued to fall (24 +/- 12% at 10 min). Controls, treated with the suspending fluid for PAF, and then 0.05 micrograms FMLP, had a similar 1 min AGC value, but at 10 min AGC returned to 65 +/- 6.1% (P less than 0.001 for comparison of 10 min values). Thus PAF pretreatment enhanced FMLP-elicited granulocytopenia in vivo. Study of in vitro human PMN aggregation revealed that, at certain relative concentrations of PAF and FMLP, aggregation was enhanced. These studies show that both in vitro and in vivo responses of FMLP-stimulated PMN may be exaggerated by pre-exposure to PAF.

The effect of vitamin E on rabbit neutrophil activation.

Chemoattractants such as N-formylmethionyl leucyl phenylalanine (FMLP) cause neutropenia in vivo. The sequestered neutrophils may block the microvasculature and contribute to respiratory distress. Neutrophils from humans receiving 1600 units vitamin E per day have reduced oxidative activity. To test whether vitamin E attenuates the responses of neutrophils to FMLP in vivo we gave rabbits four daily intramuscular injections of 100 mg vitamin E. Serum levels of the vitamin were 2.34 +/- 0.15 mg% compared to 0.19 +/- 0.04 mg% in control rabbits receiving placebo injections. On the fifth day testing was done before and after injecting FMLP. Variables monitored were the absolute granulocyte count (AGC), systolic, diastolic and mean blood pressures (MBP), heart rate, PO2, PCO2, pH and respiratory rate. When 0.5 microgram FMLP was injected intravenously the AGC decreased (at 2.5 min the percentage change was -89.7 +/- 8.0 with vitamin E and -97.0 +/- 2.7 without vitamin E; P = 0.2). MBP decreased also (% change, -29.0 +/- 13.0 with vitamin E and -36.3 +/- 16.0, without vitamin E). By 15 min recovery was seen (AGC % change, -26.0 +/- 17 with vitamin E and -78.7 +/- 10.5, without vitamin E; P = 0.01; MBP % change, -9.3 +/- 3.8 with vitamin E and -52.3 +/- 10.1 without vitamin E). Chromatographic analysis of serum extracts revealed increases in 6-keto-PGF1 alpha after stimulation. Studies with [3h]thymidine-labelled neutrophils showed that the sequestered cells return to the circulation. Vitamin E might facilitate this return by altering the adherence of neutrophils to endothelium. This possibility was tested by measuring the adherence to cultivated rabbit aorta endothelial monolayers of FMLP-stimulated neutrophils from vitamin E-treated rabbits. The percentage of neutrophils adhering was 32.5 +/- 3.5 with vitamin E and 60.0 +/- 7.1, without vitamin E. Thus vitamin E promotes the return of neutrophils to the circulation after chemotactic challenge and may do so by reducing their adherence to endothelium.

Relation of respiratory burst and arachidonate metabolism during phagocytosis by guinea pig alveolar macrophages.

We have examined the respiratory burst and arachidonic acid oxygenation that accompany phagocytosis in macrophages. Guinea pig alveolar macrophages were stimulated with opsonized zymosan in the presence of inhibitors of arachidonic acid metabolism: ASA, indomethacin, and ETYA, ASA, at concentrations as high as 60 micrograms/ml, had no effect on either oxygen consumption or superoxide ion formation. Indomethacin (4 x 10(-4) M) and ETYA (2 x 10(-5) M) did inhibit oxygen utilization and superoxide production. However, no indomethacin or ETYA inhibition of oxygen utilization was detected in the presence of 1 mM KCN, suggesting that the inhibitable portion of the respiratory burst observed with indomethacin or ETYA was dependent on mitochondrial respiration. Further study with ETYA showed that the inhibitor at 2 x 10(-5) M had little effect on uptake of 125I-labeled zymosan but did abolish the conversion of 14C-arachidonic acid to a compound that co-migrated with authentic 12-HETE on silica gel plates. Lower concentrations of ETYA (5 x 10(-6) M), which had no effect on the respiratory burst of phagocytosing alveolar macrophages, also inhibited arachidonic acid metabolism. We conclude therefore that the inhibition of oxygen consumption and superoxide production by ETYA at 2 x 10(-5) M is unrelated to inhibition of arachidonic acid metabolism. Furthermore, the oxygenation of arachidonic acid requires little of the oxygen consumed by phagocytosing alveolar macrophages.

Metabolic, membrane, and functional responses of human polymorphonuclear leukocytes to platelet-activating factor.

The phospholipid mediator of anaphylaxis, platelet-activating factor (PAF) is chemotactic for polymorphonuclear leukocytes (PMN). We have examined this agent's effects on several other PMN functions. Human PMN were prepared from heparinized venous blood by Ficoll gradient. Metabolic burst was examined by measurement of O2 use and O2.- production in the presence or absence of PAF (10(-6)--10(-9) M). Unless cells were treated with cytochalasin-B (5 micrograms/ml), no significant respiratory burst was demonstrated. However, pretreatment with PAF (10(-7) M) enhanced approximately threefold the O2 utilization found when cells were subsequently stimulated with 10(-7) M FMLP. PAF also stimulated arachidonic acid metabolism in 14C-arachidonic acid-labeled PMN. Thin-layer chromatography analysis of chloroform-methanol extracts showed substances that comigrated with authentic 5-hydroxyeicosatetraenoic acid had a marked increase in radioactivity following PAF stimulation at 10(-7) M. PAF failed to stimulate release of granule enzymes, B-glucuronidase, lysozyme, or myeloperoxidase unless cytochalasin-B were added. PAF from 10(-6) M to 10(-10) M affected PMN surface responses. PMN labeled with the fluorescent dye, chlorotetracycline, showed decreased fluorescence upon addition of PAF, suggesting translocation of membrane-bound cations. Further, the rate of migration of PMN in an electric field was decreased following PAF exposure, a change consistent with reduced cell surface charge. PMN self-aggregation and adherence to endothelial cells were both influenced by PAF (10(-6) M--10(-9) M). Aggregation was markedly stimulated by the compound, and the percent PMN adhering to endothelial cell monolayers increased almost twofold in the presence of 10(-8) M PAF. Thus, PAF promotes a variety of PMN responses: enhances respiratory burst, stimulates arachidonic acid turnover, alters cell membrane cation content and surface charge, and promotes PMN self-aggregation as well as adherence to endothelial cells.

The influence of vitamin E on human polymorphonuclear cell metabolism and function.

These studies on the effect of administration of 1,600 units of vitamin E to humans indicated the following responses to the PMNs (TABLE 6). Functional alterations occur with an increased ability to ingest particles but a mild decrease in bactericidal potency of the PMN. Although the respiratory burst is slightly enhanced as is superoxide anion release, H2O2 release from the PMN is markedly impaired. The hexose monophosphate shunt activity, which is dependent on intracellular H2O2 is decreased during phagocytosis. Membrane responses such as changes in order parameter during phagocytosis as reported by the stearic acid analogue probe 5DS are similar to those of normal PMNs. The release of arachidonic acid from membranes of vitamin E PMNs during phagocytosis of opsonized zymosan is slightly enhanced, indicating normal phospholipase A2 activation. NADH oxidase-derived H2O2 is not impaired within phagocytic generated by NADPH oxidase in phagocytic vesicles, accounting for impairment in HMPS activity and bactericidal activity in these cells.

Membrane fluidity changes accompanying phagocytosis in normal and in chronic granulomatous disease polymorphonuclear leukocytes.

We have studied membrane fluidity changes in polymorphonuclear leukocytes (PMN) during phagocytosis. Membrane fluidity was assessed by electron spin resonance (ESR) using a nitroxide-substituted stearic acid analog (5DS) as a spin probe. PMN from normal subjects and from 3 CGD patients (2 males, 1 female) were incubated in Kreb's Ringers phosphate with or without opsonized zymosan. ESR spectra were obtained and the order parameter (S), which is inversely related to membrane fluidity, was calculated. Without zymosan addition, S for normal (0.638) and for CGD (0.635) were not significantly different (p less than 0.35). The S values indicate that under resting conditions the molecular environment of the CGD membrane is similar to that of normal PMN membranes. However, with addition of opsonized zymosan, the normal, but not the CGD, PMN showed a significant increase (CGD, S = 0.638; normal, S = 0.647; p less than 0.001). This change in S for the normals is consistent with a more restricted movement of 5DS. Treatment of normal PMN with a mixture of scavengers specific for H2O2 (catalase, 1600 U/ml), O2-.(superoxide dismutase, 100 micrograms/ml), and for HO., (sodium benzoate, 1mM) during zymosan stimulation gave S values similar to those of resting cells. Catalase alone also lowered S value, suggesting that H2O2 was instrumental in causing the initial S value increase. This idea was supported by studies in which CGD cells were incubated with zymosan in the presence of glucose oxidase, an enzyme that catalyzes glucose oxidation resulting in the direct reduction of molecular oxygen to H2O2. Our results indicate that reduced O2 by-products, particularly H2O2, can cause altered biophysical properties of PMN membrane during phagocytosis.

Amphotericin-B promotes leukocyte aggregation of nylon-wool-fiber-treated polymorphonuclear leukocytes.

Severe pulmonary reactions have been reported in patients receiving leukocyte transfusion and amphotericin-B. To study the interaction of amphotericin-B with polymorphonuclear leukocytes (PMN), purified human PMN were incubated with 200 mg of nylon wool fiber for 60 min either in the absence or presence of 2 mM EDTA. PMN were recovered in acid citrate dextrose solution and were suspended in balanced salt solution for determination of their aggregation properties. The cells exposed to nylon wool fibers without EDTA aggregated in response to concentration as low as 1.25 micrograms/ml of amphotericin-B. Cells initially treated with EDTA, however, failed to aggregate. Serum from a patient treated with amphotericin-B aggregated PMN exposed to nylon wool fiber but not control cells, whereas serum taken before amphotericin was given without effect on the PMN treated with nylon wool fiber. Amphotericin-B at 5 micrograms/ml failed to potentiate the release of beta-glucocuronidase or lactic dehydrogenase by PMN treated by nylon wool beyond that seen with exposure to the fibers alone. Rabbit peripheral blood was similarly incubated with nylon wool fibers and the recovered PMN were infused into recipient rabbits that had received 1 mg/kg of amphotericin-B intravenously 1 hr prior to the infusion of the leukocytes. Rabbits were sacrificed 30 min after transfusion of PMN, and their lungs were excised for histologic sectioning. Those rabbits receiving a combination of amphotericin-B and 4 x 10(7) nylon-wool-fiber-treated PMN had evidence of pulmonary hemorrhage and accumulation of leukocytes in the pulmonary vasculature whereas those animals who received such cells alone had normal appearing lung tissue. In summary, amphotericin-B at concentrations achievable in vivo enhanced the aggregation of PMN damaged by incubation with nylon fiber with subsequent accumulation of the phagocytes in pulmonary tissue.

Fluidity properties and liquid composition of erythrocyte membranes in Chediak-Higashi syndrome.

We have earlier shown through electron spin resonance (ESR) studies of leukocytes that membranes of cells from both Chediak-Higashi syndrome (CHS) mice and humans have abnormally high fluidity. We have extended our studied to erythrocytes. Erythrocytes were labeled with the nitroxide-substituted analogue of stearic acid, 2-(3-carboxypropyl)-4,4-dimethyl-2-tridecyl-3-oxazolidinyloxyl, and ESR spectra were obtained. Order parameter, S, at 23 degrees C, was 0.661 and 0.653 for erythrocytes of normal and CHS mice (P less than 0.001). S was 0.684 for normal human erythrocytes and 0.675 (P less than 0.001) for CHS erythrocytes at 25 degrees C. Because S varies inversely to fluidity, these results indicate that CHS erythrocytes tend to have higher fluidity than normal. In vitro treatment of both mice and human CHS erythrocytes with 10 mM ascorbate returned their membrane fluidity to normal. We prepared erythrocyte ghosts and extracted them with CHCl3:CH3OH (2:1). Gas-liquid chromatography analysis showed a greater number of unsaturated fatty acids for CHS. The average number of double bonds detected in fatty acids for mice on a standard diet was 1.77 for normal and 2.02 for CHS (P less than 0.04); comparison of human erythrocytes from one normal control and one CHS patient showed a similar trend. Our results suggest that an increased proportion of unsaturated fatty acids may contribute to increased fluidity of CHS erythrocytes. Our observation that both leukocytes and erythrocytes of CHS have abnormal fluidity indicates that CHS pathophysiology may relate to a general membrane disorder.

Membrane fluidity in human and mouse Chediak-Higashi leukocytes.

Polymorphonuclear leukocytes from humans and mice with the Chediak-Higashi syndrome were characterized by spin label electron spin resonance spectrometry. Our results suggest that cells from afflicted mice and humans have membranes more fluid than controls. Order parameters for a spin label that probes near the membrane surface were 0.652 for normals and 0.645 for two Chediak-Higashi patients. Cells from Chediak-Higashi mice showed similar differences, as did isolated plasma membrane fractions. An increased membrane fluidity was also detected with a spin label that probes deeper in the bilayer. In vitro treatment of Chediak-Higashi mouse cells with 0.01 M ascorbate increased the order parameter to normal levels. In vitro incubation of mouse Chediak-Higashi cells with glucose oxidase increased the order parameter, similar to the effect of ascorbate. This increase was abolished when catalase was added to the incubation medium. In vitro incubation with dibutyryl cyclic guanosine monophosphate (1 muM to 0.1 mM) did not normalize order parameters. These results indicate that fluidity of Chediak-Higashi cell membranes was affected by treatments expected to alter the oxidation: reduction potential of the environment but was not affected by treatments expected to alter the ratio of intracellular cyclic nucleotides. The latter treatment would affect microtubule assembly. Therefore, it appears that the membrane fluidity abnormalities as demonstrated by electron spin resonance and the earlier demonstrated microtubule dysfunctions characteristic of Chediak-Higashi cells are coexisting defects and are probably not directly related.