Free radical spin traps as adjuncts for the prevention and treatment of disease.

Free radical spin traps exhibit properties consistent with a role in the prevention or amelioration of diseases mediated by the formation of free radical species in excess of optimum constitutive requirements. At first view, it may be surprising that they have not found a place in clinical medicine. Some studies of diseases in which free radicals and oxidative stress are aberrantly over- or underabundant and which might be ameliorated by inhibiting or augmenting their formation have been reported. A number of pathophysiologic categories in which such agents have been or might be employed are briefly summarized.

In vivo and in vitro biotransformation of the lithium salt of gamma-linolenic acid by three human carcinomas.

Lipid metabolism has been considered recently as a novel target for cancer therapy. In this field, lithium gamma-linolenate (LiGLA) is a promising experimental compound for use in the treatment of human tumours. In vivo and in vitro studies allowed us to assess the metabolism of radiolabelled LiGLA by tumour tissue and different organs of the host. In vitro studies demonstrated that human pancreatic (AsPC-1), prostatic (PC-3) and mammary carcinoma (ZR-75-1) cells were capable of elongating GLA from LiGLA to dihomo-gamma-linolenic acid (DGLA) and further desaturating it to arachidonic acid (AA). AsPC-1 cells showed the lowest delta5-desaturase activity on DGLA. In the in vivo studies, nude mice bearing the human carcinomas were given Li[1-(14)C]GLA (2.5 mg kg(-1)) by intravenous injection for 30 min. Mice were either sacrificed after infusion or left for up to 96 h recovery before sacrifice. In general, the organs showed a maximum uptake of radioactivity 30 min after the infusion started (t = 0). Thereafter, in major organs the percentage of injected radioactivity per g of tissue declined below 1% 96 h after infusion. In kidney, brain, testes/ovaries and all three tumour tissues, labelling remained constant throughout the experiment. The ratio of radioactivity in liver to tumour tissues ranged between 16- and 24-fold at t = 0 and between 3.1- and 3.7-fold at 96 h. All tissues showed a progressive increase in the proportion of radioactivity associated with AA with a concomitant decrease in radiolabelled GLA as the time after infusion increased. DGLA declined rapidly in liver and plasma, but at a much slower rate in brain and malignant tissue. Seventy-two hours after the infusion, GLA was only detected in plasma and tumour tissue. The sum of GLA + DGLA varied among tumour tissues, but it remained 2-4 times higher than in liver and plasma. In brain, DGLA is the major contributor to the sum of these fatty acids. Data showed that cytotoxic GLA and DGLA, the latter provided either by the host or by endogenous synthesis, remained in human tumours for at least 4 days.

Mechanisms of the selective cytotoxic actions of certain essential fatty acids.

Polyunsaturated fatty acids (PUFA) have a selective cytotoxic/cytostatic effect on a number of tumor cell lines in culture. Although this process may be enhanced by the addition of iron there is a minimum level of PUFA necessary for potentiation of cell death. Vitamin E blocks PUFA cytotoxicity when added up to 5 days after fatty acid administration. Levels of thio-barbiturate reactive material (TBARM) in the medium rise in parallel with cell death. However, they are not affected by small alterations in temperature or oxygen tension. Incubating cells with PUFA causes marked alterations in the fatty acid patterns of both neutral and phospholipid fractions. Membrane fluidity is increased and the activity of membrane-bound receptors may be influenced directly or through the actions of eicosanoids derived from the exogenous fatty acid. PUFA may be an effective way of influencing tumor growth and a safe approach for the management of human cancer.

Exogenous gamma-linolenic acid alters hormone stimulated cyclic AMP levels in U937 cells.

Polyunsaturated fatty acids are selectively cytotoxic in culture. Incorporation of these fatty acids leads to profound changes in membrane fatty acid composition which in turn may alter the activity of transmembrane receptor/effector systems. In U937 cells, hormone stimulated production of cyclic AMP can be reduced by 30% following incubation with gamma-linolenic acid (18:3n-6). It is suggested that beta-adrenoreceptor number, subtype and adenylyl cyclase stimulation may be regulated by alterations in membrane fatty acid composition as a result of changes in the levels of polyunsaturated fatty acids and alterations in eicosanoid production.

Vitamin E blocks the cytotoxic effect of gamma-linolenic acid when administered as late as the time of onset of cell death--insight into the mechanism of fatty acid induced cytotoxicity.

Certain polyunsaturated fatty acids can selectively kill tumor cell lines while causing little to no harm to normal cell lines. However, the mechanism of this cytotoxicity is only partially understood. Antioxidants such as vitamin E have been shown to be capable of completely blocking the cytotoxic response when administered concomitantly with the fatty acid. We report here that when vitamin E was added as late as 6 days following fatty acid treatment, at a time point when the process of cell death was well underway, any further development of cell death was blocked. This implies that the mechanism of fatty acid induced cytotoxicity does not involve a gradual compromising of the cell over the 5-7 day time course of cell death. Instead, the event triggering cell death is an oxidative phenomenon occurring over a short time span of minutes or hours, not days, and is completely blocked by vitamin E.

Metabolism of n-6 fatty acids by NIH-3T3 cells transfected with the ras oncogene.

N-6 fatty acid metabolism was compared in NIH-3T3 cells and DT cells, which differ only in the presence of the v-Ki-ras oncogene. Non-dividing cells were incubated with [1-14C]-labelled fatty acids (18:2n-6, 18:3n-6, 20:3n-6 and 20:4n-6) at different time intervals (2-24 h) and concentration (0-120 microM). In both cells lines, the uptake of different fatty acids from the medium was similar and reached a maximum at 6-8 h. All fatty acids reached the same maximum level in DT cells, whereas, the relative uptake of added fatty acids by NIH-3T3 cells was different: 20:4n-6 > 20:3n-6 > 18:2n-6 = 18:3n-6. Throughout the incubation (2-24 h), desaturation and elongation of n-6 fatty acids was more active in DT cells than in NIH-3T3 cells. However, in both cell lines, incubated with different n-6 fatty acid precursors, the levels of radiolabelled 20:4n-6 were relatively constant. In DT cells, phosphatidylcholine was found to be the major fraction labelled with n-6 fatty acids precursors and those of endogenous synthesis, whereas, in NIH-3T3 cells the neutral lipid fraction, particularly triglycerides, was also strongly labelled. In concentration dependent studies, phospholipid labelling by fatty acids was saturable. At lower concentrations, especially in DT cells, phospholipids were labelled predominantly. As the concentration increased there was an overflow into the triglyceride fraction. Since the differences in fatty acid metabolism between the two cell lines cannot be related to the growth rate, it is suggested that they were a consequence of the expression of the v-Ki-ras oncogene.

Variations in temperature and oxygen content do not alter levels of thiobarbiturate reactive material in human breast tumour cells (ZR-75-1) incubated with gamma-linolenic acid.

The mechanism by which tumour cells may be killed in vitro by exogenous polyunsaturated fatty acids may involve lipid peroxidation. Gamma-linolenic acid caused a dose and time-dependent reduction in ZR-75-1 cell growth. However, altering either the incubator temperature (35, 37 and 39 degrees C) or the oxygen content (16, 21 and 26%) had little effect on either the growth of cells in the presence of gamma-linolenic acid or on thiobarbiturate reactive material levels over a 7 day period. Thus, small changes in cell culture conditions do not affect 18:3n-6 cytotoxicity or markers of lipid peroxidation.

Concentration-dependent effect of iron on gamma-linolenic acid toxicity in ZR-75-1 human breast tumor cells in culture.

Polyunsaturated fatty acids are cytotoxic to ZR-75-1 human breast tumor cells in culture. This effect may be potentiated by the simultaneous addition of iron. When cytotoxicity was measured in the presence of different concentrations of both gamma-linolenic acid and ferrous chloride there was an increase in cell death above concentrations of 9 microM and 0.05 microM, respectively. The potentiation of the effects of 18:3n-6 at low concentrations by the simultaneous addition of Fe(II) ions supports the contention that an alteration in the intracellular Fe(II)/Fe(III) ratio is necessary to promote autocatalytic lipid peroxidation.

Metabolism of radiolabelled 18:2n-6 and 18:3n-6 by NIH-3T3 cells and the DT subclone.

The incorporation and metabolism of delta-6-desaturase substrate and product, [1-14C]-linoleic (18:2n-6) and [1-14C]-gamma-linolenic acid (18:3n-6), was examined in NIH-3T3 cells and the DT subclone which differs only in the presence of the v-Ki-ras oncogene. Similar amounts of post delta-6 and delta-5 desaturase metabolites were found in both cell lines indicating that the activity of these important enzymes of fatty acid metabolism was not affected by the expression of the oncogene. However, measurable quantities of the direct elongation product of 18:2n-6, 20:2n-6, were only found in DT cells. Radiolabel was recovered predominantly from the phospholipid fraction at low fatty acid concentrations, whereas neutral lipid labelling occurred when higher concentrations of exogenous fatty acid were present. This effect was most pronounced in DT cells and may result from the presence of the activated ras oncogene.

Comparison of the metabolism of alpha-linolenic acid and its delta 6 desaturation product, stearidonic acid, in cultured NIH-3T3 cells.

The incorporation and metabolism of alpha-linolenic acid (18:3n-3) and its delta 6 desaturase product, stearidonic acid (18:4n-3), were compared by NIH-3T3 cells. In the presence of fetal calf serum, cells accumulated exogenously added 18:3n-3 and 18:4n-3 apparently at the expense of oleic acid (18:1n-9). Both 18:3n-3 and 18:4n-3 were elongated and desaturated to eicosatetraenoic acid (20:4n-3), eicosapentaenoic acid (20:5n-3) and docosapentaenoic acid (22:5n-3), but not to docosahexaenoic acid (22:6n-3), and were incorporated into phospholipids and triacylglycerols. Over a 4-d period, the growth of NIH-3T3 cells was slightly stimulated in the presence of 18:3n-3 (20 micrograms/mL) but was strongly inhibited in the presence of 18:4n-3 at the same concentration. This inhibition may be caused by enhanced lipid peroxidation as a result of the high levels of 18:4n-3 present.

The effect of n-6 fatty acids on normal and v-Ki ras transformed NIH-3T3 cells.

The effect of exogenous gamma-linolenic acid (18:3n-6) was examined on NIH-3T3 and a subclone expressing the v-Ki-ras oncogene (DT). 18:3n-6 inhibited DT cell growth more readily than NIH-3T3 cell growth. In comparison, linoleic acid (18:2n-6) had no effect on the growth of either cell line. DT cells elongated and desaturated both 18:2n-6 and 18:3n-6 to dihomo-gamma-linolenic acid (20:3n-6) and arachidonic acid (20:4n-6) to a much greater extent than NIH-3T3 cells and had a much higher membrane fluidity. The presence of the ras gene or its product appears to increase the metabolism of polyunsaturated fatty acids and potentiate the cytostatic actions of 18:3n-6.

Fatty acid changes during the induction of differentiation of human promyelocytic leukemia (HL-60) cells by phorbolmyristate acetate.

We studied fatty acid changes that are likely to occur during phorbolmyristate acetate (PMA)-induced differentiation of HL-60 cells. It was observed that PMA-induced differentiation is associated with increased uptake, but not synthesis, of fatty acids. Fatty acid analysis revealed that arachidonic acid (AA), 20:5 n-3 and 22:6 n-3 levels are reduced with a concomitant increase in 22:5 n-6 in the phospholipid fraction. In the FFA fraction there are increases in free AA, free 20:5 n-3, 22:5 n-3 and 22:6 n-3, and a fall in free 22:5 n-6 in PMA-treated cells. PMA-induced differentiation and nitroblue tetrazolium reduction by PMA-treated cells was only partially inhibited (about 20-30%) by indomethacin and nordihydroguiaretic acid (cyclooxygenase and lipoxygenase inhibitors respectively), but not by superoxide dismutase, catalase or mannitol. These results indicate that PMA-induced differentiation of HL-60 cells is accompanied by specific changes in the fatty acid composition of the cells.

Lipid peroxidation in human breast cancer cells in response to gamma-linolenic acid and iron.

Lipid peroxidation in human breast cancer (ZR-75-1) cells and cancer cell killing were confirmed by using ultraviolet (UV)-spectrophotometry and mass spectrometry (MS). ZR-75-1 cells and human normal fibroblast CCD-41-SK (41Sk) cells were cultured with gamma-linolenic acid (GLA) and ferrous iron Fe (II) combinations. Formation of lipid peroxide and cytotoxic effect were highest in ZR-75-1 cells treated with GLA + Fe (II), though 41Sk cells showed little evidence of either lipid peroxidation or cytotoxity. These results indicate a cancer-cell-specific lipid peroxidation mechanism in association with the selective cancer cell killing effect in response to GLA.

The effect of chemical hepatocarcinogenesis on liver phospholipid composition in rats fed N-6 and N-3 fatty acid-supplemented diets.

The effect of dietary fats on essential fatty acid metabolism in rats subjected to chemically induced hepatocarcinogenesis was studied. Sixty male rats were fed a diet supplemented with one of the following three oil compositions: 10% hydrogenated coconut oil (HCO); 5% hydrogenated coconut oil and 5% gamma-linolenic acid (18:3n-6)-rich evening primrose oil (EPO); or 5% hydrogenated coconut oil and 5% marine oil (FO). Half of the animals in each dietary regimen were subjected to hepatocarcinogenesis induction using diethylnitrosamine and 2-acetylaminofluorene (2-AAF) followed by partial hepatectomy, whereas the other half underwent hepatectomy without receiving diethylnitrosamine and 2-acetylaminofluorene. Liver phospholipid composition was analyzed. In comparison to the HCO group, the EPO group showed raised levels of arachidonic acid (20:4n-6) and suppressed n-3 fatty acids. The FO group, on the other hand, showed suppressed levels of n-6 and increased n-3 fatty acids. Hepatocarcinogenesis suppressed the level of 20:4n-6 and this effect was greater in the FO rats. The levels of dihomo-gamma-linolenic acid (20:3n-6) were increased by the hepatocarcinogenic treatment, and this effect was further accentuated in the EPO rats. These results suggest that hepatocarcinogenesis may suppress the activity of delta-5-desaturase, which may be one of the reasons why tumor cell membranes have low levels of long chain fatty acids, especially 20:4n-6 cells, and have an impaired capacity to undergo lipid peroxidation.

Levels of thiobarbituric acid reactive substances and the cytocidal potential of gammalinolenic and docosahexaenoic acids on ZR-75-1 and CV-1 cells.

To clarify the mechanism by which gammalinolenic acid (GLA) is more tumoricidal than docosahexaenoic acid (DHA), we have compared the incorporation of the respective exogenously added ethyl esters GLAe and DHAe into the phospholipids of tumorigenic ZR-75-1 and non-tumorigenic CV-1 cells relative to the ability of the cells to survive and to accumulate thiobarbituric acid reactive substances (TBARS). GLA and DHA were incorporated in the phospholipids to the same extent, but GLA disappeared more rapidly than DHA in both cell lines. GLAe induced about twice as much intracellular TBARS as DHAe in both cell lines, but killed ZR-75-1 cells four times more effectively than DHAe. DHAe induced 11-15 fmoles malondialdehyde-equivalents (MDA-eq)/cell in both ZR-75-1 and CV-1 cells, whereas GLAe induced 5-6 times more TBARS in ZR-75-1 cells (26-30 fmoles MDA-eq/cell) than in CV-1 cells (5-6 fmoles MDA-eq/cell). The results show that there is no difference in GLA and DHA incorporation into phospholipids, but that their metabolism differs in the two cell types. The data also suggest that the cytocidal potential is related to TBARS levels in a nonlinear fashion. The relationship between excess prostaglandin production and excessive cell death due to GLA is discussed.

Intracellular free Fatty-Acid release and lipid-peroxidation in cultured human breast-cancer cells in response to gamma-linolenic Acid with iron (gla + fe).

Intracellular free fatty acid (FFA) release and peroxidation of polyunsaturated fatty acid (PUFA) were studied in cultured human breast cancer cells (ZR-75-1) exposed to gamma-linolenic acid with iron (GLA + Fe). This treatment results in cell death. Increased intracellular FFA were observed in association with both the accelerated peroxidation of PUFA and the killing effect. Vitamin E reduced all three effects. The FFA were methyl esterified and analyzed by gas chromatography-mass spectrometry. The identified FFAs were 16: 0, 18: 3, 20: 3 and 20: 4 (numbers of carbons and double bonds indicated). These results suggest an association of intracellular FFA release with the peroxidation of PUFA and the cancer cell-killing by GLA in the presence of iron.

Effects of eicosapentaenoic acid and arachidonic acid on incorporation and metabolism of radioactive linoleic acid in cultured human fibroblasts.

Effects of exogenous eicosapentaenoic acid, arachidonic acid and oleic acid on incorporation and metabolism of [14C] linoleic acid were examined in cultured human fibroblasts obtained from three donors of different ages. Eicosapentaenoic acid treatment (40 microM) inhibited incorporation of radioactive linoleic acid and actively reduced radioactivity of desaturation-elongation metabolites in phospholipids, predominantly in the phosphatidylethanolamine fraction. In contrast, radioactivities of the metabolites in triacylglycerols were significantly increased with arachidonic acid treatment (40 microM): eicosapentaenoic acid had a smaller effect or none. Oleic acid had virtually no effect. These effects were consistent in the three cell lines, but responses to treatment with the acids differed considerably among individual cells. The pool of linoleic acid metabolites in triacylglycerols may not be negligible. The exogenous fatty acids may influence both the transfer of lipids between the major lipid pools as well as the activities of the desaturation-elongation system.

Modulation of CD4 expression on lymphoma cells transplanted to mice fed (n - 3) polyunsaturated fatty acids.

Groups of adult AKR mice were fed well defined fats controlled diet regimens. These consisted of either saturated (beef tallow: 'BT') or (n - 3) polyunsaturated (fish oil: 'FO') fatty acids supplementation to basal mix mouse food. In other groups, the basal mix was given without any fat supplement ('NF'). Six weeks or more after the initiation of these diet regimens, mice received intraperitoneal injection of histocompatible RDM-4 lymphoma cells. Ascites RDM-4 tumors were harvested approximately two weeks later, and some of their physicochemical properties were studied. It was repeatedly found that: (1) the tumor grew considerably faster in the FO-fed donor than in the BT- or NF-fed donors; (2) cell membrane fluidity, content of C20(n - 3) and of C22(n - 3) fatty acids were significantly higher in the FO groups than in both BT and NF groups, while the content of C20(n - 6) and 22:4(n - 6) fatty acids was concomitantly decreased; (3) expression of the CD4 cell surface marker was always significantly diminished in the FO groups, whereas other markers such as CD8, H2K, Thy-1 and LFA-1 were not affected. Similar results were obtained, whether fats constituted from 1% to 16% by weight of the food intake. Use of a recently selected line of the RDM-4 lymphoma, exhibiting higher CD4 marker expression, resulted in similar observations. On the other hand, CD4 expression on cells from lymphoid organs of healthy adult AKR mice was not detectably modulated by the dietary fats.

Fatty acid and enzymatic compositional changes in the pancreas of rats fed dietary n-3 and n-6 polyunsaturated fatty acids.

The influence of n-3 and n-6 PUFA on the fatty acid composition and the enzyme content of zymogen granules of the normal exocrine pancreas was tested on rats. The animals were fed on different diets comprising 5% fish oil (FO), safflower oil (SFO), and evening primrose oil (EPO) used singly or in combination as dietary fats. The results were compared with those from animals fed 5% hydrogenated beef tallow (HBT). The fatty acid composition and digestive enzyme content were analyzed after a 6-wk feeding period. Differences in the pancreatic fatty acid profiles were related to the fatty acid composition of the ingested fats. Equivalent levels of n-3 fatty acids and 20:3n-6 were obtained with either EPO or FO fed singly or in combination. Similar results were observed with SFO/FO. Higher C20:3n-6/C20:4n-6 ratios were obtained with the oil mixtures. An increase in amylase levels, but a decrease in serine protease (Band 21 kdalton) levels, was associated with EPO. An elevation in procarboxypeptidase levels paralleled an increase in 18:0 levels, whereas the proportion of lipase (Band 49 kdalton) varied inversely with the proportion of C20:3n-6. The SFO/FO mixture elevated the proportions of protease II and proelastase. These results suggest that specific fatty acids influence the proportion of specific digestive enzymes in the zymogen granules.

Susceptibility of RDM4 lymphoma cells to LAK-mediated lysis is decreased in tumor bearers fed fish oil high fat regimen.

RDM4 lymphoma cells were grown intraperitoneally in genetically compatible AKR mice fed either regular mouse chow, or diet supplemented with either saturated fat (hydrogenated beef tallow = HBT) or unsaturated fat (fish oil = FO). It was observed that the lymphoma cells number was significantly greater in FO-fed hosts and lower in HBT-fed hosts, than in the mice fed regular chow. The tumor bearers diet did not dramatically influence the rate of DNA synthesis of RDM4 cells, as measured by [3H]thymidine uptake in culture, a few hours after harvesting from the peritoneal cavity. It was repeatedly found that FO feeding of the tumor bearers elicited an increased resistance of RDM4 cells to lysis by LAK effectors, as appraised in vitro by 51Cr release test and in vivo by the "Winn assay". Different FO percentage of the diet (16%, 8%, 4%) resulted in comparable reduction of susceptibility of RDM4 cells to lysis by LAK effectors. Lipid analysis showed that RDM4 cells grown in mice fed FO diet or HBT diet differed markedly in their fatty acid composition and that their resistance to lysis by LAK cells correlated with the quantity of oxidizable fatty acids especially of the n-6 type.