Evidence that changes in Se-glutathione peroxidase levels affect the ...

Carcinogenesis vol.18 no.10 pp.1897–1904, 1997

Evidence that changes in Se-glutathione peroxidase levels affect the sensitivity of human tumour cell lines to n-3 fatty acids

Svanhild A.Schønberg1,4, Parveen K.Rudra1,4, Randi Nøding2, Frank Skorpen1, Kristian S.Bjerve2 and Hans E.Krokan1,3 1UNIGEN

Center for Molecular Biology, Norwegian University of Science and Technology, and 2Department of Clinical Chemistry, University Hospital of Trondheim, N-7005 Trondheim, Norway

3To

whom correspondence should be addressed

4Svanhild

A.Schønberg and Parveen K.Rudra contributed equally to this

article

The human lung adenocarcinoma cell line A-427 is significantly more sensitive to cytotoxic lipid peroxidation products of eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) than the human lung adenocarcinoma cell line SKLU-1, and the glioblastoma cell lines A-172 and U-87 MG. The cytotoxic effect as well as lipid peroxidation were abolished by vitamin E. The differential sensitivities of the cell lines were not correlated to the levels of lipid peroxidation products (measured as the end product malondialdehyde), indicating differences in sensitivities to products of lipid peroxidation. The high sensitivity of A427 is apparently due to a low level of selenium-dependent glutathione peroxidase (GSH-Px), because pretreatment with sodium selenite (250 nM) increased the GSH-Px activity 3- to 4-fold and protected the cells almost completely against the growth inhibitory effect of DHA. Furthermore, 2-phenyl-1,2-benzisoselenazol-3(2H)-one (ebselen) a seleno-organic GSH-Px mimic, suppressed the cytotoxic action of DHA to A-427 in a dose dependent manner. Northern analysis demonstrated that pretreatment with sodium selenite (250 nM) was accompanied by an increased level of GSH-Px mRNA (1.8-fold) in A-427 cells, while the level remained unchanged under the same conditions in DHA/EPA-resistant A-172 cells. In addition, the level of selenophosphate synthetase mRNA (SelD), a key intermediate in tRNASec formation, increased 1.2- to 1.7-fold in A-427 and A-172 cells after pretreatment with sodium selenite. These results indicate that upregulation of GSH-Px activity by sodium selenite in the EPA/DHA sensitive cell line A-427 may be due to an increase in mRNAs for GSH-Px and a precursor important for formation of tRNASec which is required for incorporation of selenocysteine in GSH-Px during translation. These results demonstrate an important role for GSH-Px in the cellular defence against cytotoxic lipid peroxidation products. Furthermore, measurement of GSH-Px activities in tumour cells may be one useful biochemical predictor for their sensitivities to polyunsaturated fatty acids.

*Abbreviations: DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; PUFA’s, polyunsaturated fatty acids; MDA, malondialdehyde; MTT, 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide); DMSO, dimethyl sulfoxide; TBA, 2-thiobarbituric acid; GSH, glutathione; GSH-Px, selenium dependent glutathione peroxidase. © Oxford University Press

Introduction Numerous animal studies have shown that the diet may affect tumour growth and invasiveness. Thus, diets enriched in marine polyunsaturated fatty acids (PUFAs*) of the n-3 type reduce colon tumour incidence and multiplicity after exposure to azoxymethane (1), decrease mammary tumourigenesis after exposure to 7,12-dimethylbenz[a]anthracene (2) and decrease preneoplastic lesions in rat pancreas after treatment with azaserine (3). n-3 Fatty acids have also been shown to reduce the growth of human tumour cells in nude mice (4–7). In addition, eicosapentaenoic acid (20:5 n-3, EPA) may reduce in vitro tumour invasiveness and metastasis in mice possibly due to reduced production of collagenase IV (8). A marked suppression by EPA of cachexia accompanying tumour growth has been observed in mice bearing the colon adenocarcinoma cell line MAC16. This effect is apparently specific for EPA (9). The mechanism of inhibition of tumour cell growth is generally not well understood, but a number of factors that are involved in growth control may be altered by n-3 fatty acids. Thus, modification of tumour cell membranes by n-3 fatty acids alters membrane fluidity (10), transport systems for metabolites (11), uptake of cytostatic drugs (10,12,13), as well as receptor binding (14) and eicosanoid production (15– 19). More recently, it has become clear that cytotoxic lipid peroxidation products formed from PUFAs are likely to be involved in inhibition of tumour cell growth (20–23). An inverse relationship between tumour growth and the concentration of lipid peroxidation products was observed in one study (20). Lipid peroxidation is initiated by free radical attack on membrane PUFAs and this gives rise to a wide range of lipid peroxidation products, some of which are very reactive, but the relative cytotoxicity of different molecular species is not well established. A class of reactive aldehydes, 4-hydroxyalkenals, formed in the course of lipid peroxidation of both n-6 and n-3 PUFAs are cytotoxic, and inhibit certain enzymes containing critical SH-groups, such as the DNA repair enzyme O6-methylguanine-DNA methyltransferase (24,25). Such aldehydes are likely candidates to be responsible for some of the cytotoxicity. In addition to hydroperoxides of unsaturated fatty acids, oxidative processes may lead to reactive oxygen metabolites which can interact with nucleic acids and other biomolecules. If not detoxified, hydroperoxides may accumulate and amplify lipid peroxidation deleterious to the cell. There is evidence that experimental tumour cells may frequently have reduced antioxidant defence mechanisms (26,27). Seleniumdependent glutathione peroxidase (GSH-Px), an enzyme important in the cellular antioxidant defence systems, catalyses the reduction of hydrogen peroxide and fatty acid hydroperoxides (28,29). This enzyme belongs to a group of selenoproteins containing the unusual amino acid selenocysteine encoded by a UGA codon (30, a review). Selenophosphate synthetase (encoded by selD) generates selenophosphate, which is essential for formation of tRNASec required for incorporation of selenocysteine into the selenoproteins during translation. The 1897

S.A.Schønberg et al.

human homologue of bacterial selD has recently been cloned and shown to be functional in mammalian cells as well as in bacteria (31). The activity of GSH-Px has been shown to vary widely among different cell types and between comparable normal and malignant cells (32,33). Protection against oxidative damage mediated through GSH-Px has been widely studied in vivo by manipulating dietary selenium content (34), while fewer studies have been carried out on isolated cells. Ochi et al. (35) have reported that GSH-Px probably plays an important role in 15HPETE-induced cytotoxicity on vascular endothelial cells since both sodium selenite and ebselen (a glutathione peroxidase mimic) prevented cell injury. This report as well as other studies on both normal and tumour cells (36–38) emphasize the importance of the glutathione redox cycle as a potent antioxidant defence mechanism. In the present study, we have investigated the effect of DHA and EPA on the growth of four different tumour cell lines in culture. We find that the growth inhibitory effect of DHA and EPA on A-427, the only highly sensitive cell line, can be reversed by vitamin E, indicating that lipid peroxidation products are responsible for the cytotoxic effect. The high sensitivity of A-427 appears to be due to a low level of GSHPx activity in this cell line. Our results indicate that deficiencies in antioxidant defence systems may be responsible for the sensitivity to PUFAs of at least some tumour cell lines, and that one or more cytotoxic lipid peroxidation products are substrates for GSH-Px. Materials and methods Chemicals Pure (99%) cis-4,7,10,13,16,19-docosahexaenoic acid (DHA), cis5,8,11,14,17-eicosapentaenoic acid (EPA), 3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide (MTT), 2-thiobarbituric acid (TBA), 1,1,3,3tetramethoxypropane, α-tocopherol, β-nicotinamide adenine dinucleotide phosphate (NADPH), reduced form, glutathione (GSH), reduced form, 5,59dithio-bis(2-nitrobenzoic acid) (DTNB), sodium selenite, H2O2 (3%), GSHreductase (type IV) from Bakers yeast and GSH-peroxidase from bovine erythrocytes were obtained from Sigma Chemical Co. (St. Louis, MO, USA). Ebselen [2-phenyl-1,2-benzisoselenazol-3(2H)-one] was purchased from ICN Pharmaceuticals Inc. (Costa Mesa, CA, USA). Growth media and additives were purchased from Gibco BRL, Life Technologies (Inchinnan, Scotland). Cell cultures Human cancer cell lines A-427, SK-LU-1, A-172 and U-87 MG were obtained from The American Type Culture Collection (Rockville, MD). The A427 (lung carcinoma), SK-LU-1 (lung adenocarcinoma) and U-87 MG (glioblastoma/astrocytoma) cells were all grown in Eagle’s minimum essential medium (MEM) supplemented with 10% fetal calf serum (FCS) with Lglutamine (3%, 80 mg/l), gentamicin (45 mg/l), non-essential amino acids (3100, 10 ml/l) and sodium pyruvate (1 mM). A-172 (glioblastoma) cells were maintained in Dulbecco’s Minimum Essential Medium (DMEM) with 4.5 g/l glucose supplemented with 10% FCS, L-glutamine (80 mg/l), gentamycin (50 mg/l) and sodium pyruvate (1 mM). All cell lines were routinely maintained in a humidified atmosphere of 5% CO2:95% air at 37°C. Fatty acid supplementation and MTT-survival Stock solutions of fatty acids (250 mM) were prepared in 96% (v/v) ethanol and stored at 220°C, and further diluted in complete growth medium with 10% FCS before experiments such that the final concentration of ethanol was ,0.5% (v/v). The growth inhibitory effect of EPA and DHA was determined by plating 103 or 3 3 103 cells/well in quadruplicate in 96-well microtitre plates. Cell cultures were incubated at 37°C in a 5% CO2 atmosphere for 4 h before changing to medium containing fatty acids. Cultures were supplemented with 0, 10, 20, 40 or 70 µM fatty acids in growth medium and incubated for 1–3 days. The growth of the cells was assessed by the MTT reduction assay essentially as described (39). After the incubation period, MTT (3 mg/ml in PBS) was added to each well (50 µl per 100 µl medium) and the plates were incubated at 37°C for 3 h in 5% CO2. Subsequently, 100 ml of medium were removed from each well and 100 ml of dimethyl sulfoxide (DMSO)

1898

(alternatively 2-propanol with HCl) was added to solubilize the MTT formazan. The plate was placed on a mechanical shaker for 20–60 min at room temperature for complete solubilization. The optical density of each well was read on a Titertek Multiscan Plus Reader using a 588-nm wavelength filter and the cell number was determined from a standard curve of absorbance against cell numbers. Lipid peroxidation assay Lipid peroxidation was measured in cell cultures by the TBA assay (40). This assay was slightly modified to achieve reproducible results. 105 cells were seeded in flasks (25 cm2), and after 24 h treated with DHA (40 µM), EPA (40 µM), vitamin E (30 µM), DHA plus vitamin E and EPA plus vitamin E in the growth medium, and incubated at 37°C. After every 24 h, 2 ml of trichloroacetic acid (20% w/v) and 4 ml of TBA (0.67% w/v) were added to the culture flask. All flasks were kept in a water bath for 20 min at 97°C. The content of the flask was then centrifuged at 3500 g for 20 min at 4°C and the absorbance of the supernatant was measured using a Hitachi U-200 spectrophotometer reading at 532 nm. The TBA-absorbance from the media was subtracted from the TBA-absorbance of the culture. Absorbance was converted to nmol malondialdehyde (MDA) from standard curves generated with 1,1,3,3-tetramethoxypropane. All solutions were freshly prepared on the day of assaying. Enzyme assays Measurement of GSH-Px activity in different cell lines were performed spectrophotometrically. 3 3106 cells were seeded in 175 cm2 tissue culture flasks. After a 4-h period, medium was replaced with fresh medium supplemented with Na2SeO3 (250 nM) for 20 h. Media were then replaced with media containing DHA (35 µM) or control medium, and the cells were further incubated for 48 h. Cells were washed twice in cold phosphate buffered saline (PBS), harvested by scraping and then sonicated. The resulting homogenates were used as source for enzyme measurements. GSH-Px activity was determined according to the method of Wendel (41). The method is based on monitoring the oxidation of NADPH at 340 nm (Hitachi U-2000 spectrophotometer) and enzyme activity was calculated using a molar extinction coefficient of 6.223103 M–1/cm–1 for NADPH. One unit (U) was defined as the amount of enzyme catalysing the oxidation by H2O2 of 1 µmol of reduced GSH to oxidized GSH per min at pH 7.0 at 25°C. GSH assay Cells were seeded at a concentration of 1.43106 cells in tissue culture flasks (75 cm2). After 4 h, medium was replaced with complete control medium or complete medium with DHA (35 µM). At the end of the incubation period (48 h), the cell layer was rinsed twice with cold PBS, harvested by scraping and then sonicated. This resulting homogenate was used as source for GSH measurements. Total GSH content was assayed according to the glutathione reductase assay (42). The GSH content was quantitated by comparison with a standard curve generated with known amounts of GSH (reduced form) and expressed as nmol per mg protein. Protein concentrations were determined by the BioRad-assay using bovine serum albumin as standard. Northern blot analysis Total RNA was extracted from cells treated with Na2SeO3 and/or DHA as described above for the measurements of GSH-Px activities, using the MicroScale Total RNA Separator Kit protocol (Clontech Lab, Inc., Palo Alto, CA, USA). Total RNA (20 µg) was electrophoresed on 1% agarose gels containing formaldehyde, stained with ethidium bromide, blotted onto Hybond N1 membrane (Amersham, UK) by vacuum, and crosslinked to the membrane by baking (20 min at 120°C). Hybridization was carried out overnight in ExpressHybTM hybridization solution (Clonetech Lab. Inc., CA, USA) to either a 856 bp cDNA encoding the human selenium-dependent GSH-Px (GPX1) (43), a 1.35 kb XhoI fragment of SelD4 cDNA (31) encoding the entire human selenophosphate synthetase, or to a glyceraldehyde 3-phosphate dehydrogenase (GAPDH) cDNA. All probes were radioactively labelled in random primer extension reactions (RediprimeTM labelling kit, Amersham, UK) with [α-32P]dCTP (Amersham). The membrane was washed 5 times for 20 min in 23SSC/1% SDS at 65°C and twice for 20 min in 0.13SSC/0.5% SDS at 68°C, enclosed in plastic wrap, and exposed to Kodak X-Omat AR film and/or analysed on a Molecular Dynamics PhosphorImager SF. Quantitation of band densities was performed using the Molecular Dynamics ImageQuant Software version 3.3. Prior to rehybridization, the membrane was stripped by boiling in 0.5% SDS for 10 min followed by an additional 10 min in the hot solution after removing it from heat.

Results Lipid peroxidation and growth inhibitory effect of n-3 fatty acids on human tumour cells The effect of DHA and EPA on the growth of one lung carcinoma, one lung adenocarcinoma and two glioblastoma

Sensitivity of human tumour cell lines to n-3 fatty acids

Table I. The effect of n-3 fatty acids DHA and EPA on lipid peroxidation in different human tumour cell linesa Cell line/Treatment

Fig. 1. Effect of increasing concentrations of DHA on the growth of the two human lung carcinoma cell lines A-427 and SK-LU-1 and the human glioblastoma cell lines A-172 and U-87 MG. j, Control, u, 10 µM DHA, d, 20 µM DHA, s, 40 µM DHA. Cell numbers were assessed using the MTT-assay as described in Materials and methods. The values represent the mean of four parallels 6 SD from one experiment, and are verified through at least two experiments.

SK-LU 1 Control DHA (40 µM) EPA (40 µM) Vit.E (30 µM) DHA 1 vit. E EPA 1 vit. E A 427 Control DHA (40µ M) EPA (40µ M) Vit.E (30 µM) DHA 1 vit. E EPA 1 vit. E A 172 Control DHA (40 µM) EPA (40 µM) Vit. E (30 µM) DHA 1 vit. E EPA 1 vit. E U87-MG Control DHA (40 µM) EPA (40 µM) Vit.E (30 µM) DHA 1 vit. E EPA 1 vit. E

24 h MDA (nmol/106cells)



1.75 4.54 6.44 0.97 2.19 3.42

6 6 6 6 6 6

1.43 1.73b 1.14b 0.93 0.97 0.66

1.83 5.20 5.35 1.24 2.38 3.18

6 6 6 6 6 6

0.08 1.23b 0.82b 0.43b 0.73 0.83b

1.16 6 0.46 3.82 6 0.41b 4.89 6 0.41b 0.82 6 0.20 0.70 6 0.07 1.11 6 0.45

1.12 3.53 3.38 0.95 1.08 2.30

6 6 6 6 6 6

0.42 1.84b 0.88b 0.51 0.46 0.30b

0.46 2.15 2.50 0.58 0.56 0.65

6 6 6 6 6 6

0.16 1.02b 1.42b 0.14 0.35 0.26

0.35 2.24 2.01 0.40 0.27 0.49

6 6 6 6 6 6

0.27 1.49 2.17 ND 0.70 ND

6 0.38 6 1.09 6 0.39b

0.44 6 0.09 4.60 6 0.75b 3.70 6 2.01b ND 0.63 6 0.32 0.5060.31

0.46 2.80 1.89 ND 0.43 ND

6 0.18 6 0.90b 6 0.81b

6 0.05 6 0.63b 6 0.49b

0.13 0.87 0.83 0.07 0.06 0.12

6 6 6 6 6 6

0.35 2.76 1.50 0.35 0.33 0.41

6 0.13 6 6 6 6 6 6

0.12 1.28b 0.52b 0.08 0.09 0.23

0.14 1.66 1.59 ND 0.15 0.21

6 0.09 6 0.06

0.11 0.35b 0.59b 0.12 0.15 0.18

6 0.46

0.06 0.34 0.27b 0.01 0.02 0.09

aFor

each cell line two independent experiments were carried out. MDA values are mean 6 SD from two independent experiments, each in duplicate. ND , 0.25 nmol MDA. bSignificantly different at P , 0.05.

Fig. 2. Effect of increasing concentrations of EPA on the growth of the two human lung carcinoma cell lines A-427 and SK-LU-1, and the human glioblastoma cell lines A-172 and U-87 MG. j, Control, u, 10 µM EPA, d, 20 µM EPA, s, 40 µM EPA. Cell numbers were assessed using the MTT-assay as described in Materials and methods. The values represent the mean of four parallels 6 SD from one experiment, and are verified through at least two experiments.

cell lines is shown in Figures 1 and 2. Both fatty acids inhibited the growth of cell line A-427 in a time- and dose-dependent manner. Cell proliferation of A-427 was reduced by ~50% compared with control cultures after exposure to 20 µM

DHA, while EPA at a concentration of 40 µM reduced cell proliferation by ~40%. SK-LU-1, A-172 and U-87 MG were much more resistant to both DHA and EPA at all concentrations tested (Figures 1 and 2). We have also tested the effect of a number of other fatty acids, both n-3 and n-6 fatty acids on the growth of A-427, SK-LU-1 and A-172. In general, A-172 is not sensitive to any of the fatty acids tested, SK-LU-1 shows a slight growth-inhibitory response to some fatty acids, while the growth of A-427 is significantly reduced both by n-3 and n-6 fatty acids (data not shown). The strongest effects were, however, seen with DHA. In order to determine whether the sensitivity of different cell lines to n-3 fatty acids is related to differences in DHA/ EPA-induced lipid peroxidation, we measured the production of MDA in the lung carcinoma, lung adenocarcinoma and glioblastoma cell lines after exposure to 40 µM DHA or EPA for 24, 48 or 72 h. A several-fold increase in MDA production was observed in all cell lines upon exposure to DHA or EPA (Table I) after a 24-h period. The lung carcinoma cell line A427 and the lung adenocarcinoma cell line SK-LU-1 have a higher level of MDA formation in the absence of fatty acids. However, lipid peroxidation appears to be more rapidly induced in these cell lines and the absolute levels are also higher after 24 h exposure than in the glioblastoma cell lines. On the other hand, after 48 h the least sensitive cell line (A-172) has produced higher MDA levels than the most sensitive cell line (A-427). Also if these data are expressed as nmol MDA/cell 1899


Table II. GSH level in different cell lines and the effect of DHA administration on the GSH-levela Cell line

nmol GSH/mg protein 6 SD

A-427, control A-427, DHA 35 µM SK-LU-1, control SK-LU-1, DHA 35 µM A-172, control A-172, DHA 35 µM U-87 MG, control U-87 MG, DHA 35 µM

35 6 3 29 6 2 150 6 5 148 6 1 112 6 2 124 6 5 42 6 1 39 6 1

aThe values represent the mean 6 SD of triplicate measurements from one experiment. The values are verified through two separate experiments.

culture rather than nmol/106 cells, the least sensitive cell line has produced more MDA than the most sensitive cell line. Vitamin E alone had no effect or suppressed MDA formation to values below those in control cultures, but had no significant effect on cell growth. In the presence of EPA or DHA, vitamin E reduced MDA levels to those observed in control cultures and also restored cell growth (Table I, Figure 4). Effects of DHA-treatment on GSH levels The total concentrations of GSH in the different cell lines are shown in Table II. The level of GSH varies between the cell lines, and the cell line most sensitive to DHA- and EPAtreatment, A-427, had 3- to 5-fold lower GSH-level than A172 and SK-LU-1. However, the sensitivity of the cell lines to n-3 fatty acids does not seem to correlate with GSH-level, since the DHA-resistant cell line U-87 MG had almost as low a level of GSH as A-427. Exposure of the cells to DHA (35 µM) for 48 h did not significantly change cellular GSH contents. GSH-Px in tumour cell lines and effects of sodium selenite and DHA on GSH-Px activities and cell growth To examine whether differences in activity of enzymes involved in antioxidant defence could explain the variation in sensitivity to lipid peroxidation products, we measured GSH-Px activity in different cell lines treated or not treated with DHA (Figure 3). Compared with GSH-Px levels in rat liver extracts (data not shown) the activity of this enzyme was relatively low in SK-LU-1, A-172 and U-87 MG, but even 3-fold lower in A427 in which the activity was close to the detection limit of the present assay. When the different cell lines were exposed to DHA for a period of 48 h there was no significant change in the GSH-Px activity (Figure 3), although a slight increase in cultures of A-427 supplemented with DHA was observed in one experiment (Figure 5). Pretreatment of the most sensitive cell line A-427 with sodium selenite (250 nM) for 20 h before DHA-administration, resulted in an almost complete recovery of cell growth (Figure 4) and reduction of MDA from 2.5 6 0.8 nmol/106cells (DHA alone) to 0.70 6 0.2 nmol/106cells (sodium selenite and DHA). Furthermore, pretreatment of A427 with sodium selenite (250 nM) for 20 h followed by DHA administration or control medium, increased the activity of GSH-Px to a level comparable to those observed in the other cell lines (Figure 5). To obtain further evidence for the significance of GSH-Px in the resistance to PUFAs, we also examined the effect of ebselen, a GSH-Px mimic, on DHAinduced cellular injury towards A-427. As shown in Figure 6, pretreating A-427 cells with ebselen for 20 h before DHA administration, resulted in a dose dependent increase in cell 1900

Fig. 3. Effect of DHA (35 µM) on the GSH-Px activity of the human lung carcinoma cell lines A-427 and SK-LU-1, and the glioblastoma cell lines A172 and U-87 MG. Cells were treated with DHA for 48 h, and after incubation period, scraped and sonicated as described in Materials and methods. Results are expressed in nmol NADPH oxidized/min/mg protein. Values represent the mean 6 SD from one experiment, each determination was done in quadruplicate, and verified through at least two experiments. j, Control, u, DHA (35 µM).

Fig. 4. Effect of pretreatment with sodium selenite (250 nM) and αtocopherol (50 µM) for 20 h before DHA-administration (70 µM) for 48 h on the growth of A-427 cells. Cell numbers were assessed using the MTTassay as described in Materials and methods, and are expressed as percent of control. Each value represents the mean of four parallels 6 SD from one experiment, results are reproduced in three experiments.

survival. These results strongly indicate that GSH-Px protects against lipid peroxidation products produced from PUFAs. Effects of Na-selenite and DHA on the levels of GSH-Px- and SelD mRNAs In order to examine whether the increase in GSH-Px activity observed after selenium-supplementation in A-427 cells was due to an increase in the levels of GSH-Px or SelD transcription, northern blot analysis was performed using total RNA isolated from A-427 cells (DHA-sensitive) and A-172 cells (DHAresistant) treated with selenium and/or DHA as described for the measurements of GSH-Px activities. Northern analysis demonstrated moderately increased levels of GSH-Px mRNA


Fig. 5. Effect of pretreatment with sodium selenite (250 nM) for 20 h before DHA-supplementation (35 µM) for 48 h on the GSH-Px activity of A-427 cells. After incubation period, cells were scraped and sonicated, as described in Materials and methods. Results are expressed in nmol NADPH oxidized/min/mg protein, and are the mean of two separate experiments 6 SD with each determination performed in at least triplicate.

Fig. 7. Northern blot analysis of RNA isolated from A-427 and A-172 after selenium pretreatment and/or DHA supplementation, probed against cDNA encoding the human GSH-Px (GPX1) and selD. Gel-picture of RNAloading is shown below.

Table III. Levels of GSH-Px and SelD mRNAs in A-427 and A-172 cells after pretreatment with Na-selenite and/or DHA-supplementationa

A-427

Se/DHA

–/1

1/2

1/1

GSH-Px

0.7 (0.8) 0.6 (0.6) 0.9 (1.0) 1.3 (1.5)

1.7 (1.3) 1.2 (0.9) 1.0 (1.2) 1.7 (2.1)

1.8 (1.3) 1.4 (1.0) 1.1 (1.8) 1.6 (2.7)

SelD A-172

GSH-Px SelD

aThe

values are expressed relative to untreated cells. The values shown in brackets are normalized to the amount of GAPDH mRNA.

Fig. 6. Effect of pretreatment with ebselen (1, 10 and 20 µM) for 20 h before DHA-administration (35 µM) on the growth of A-427 cells. Cell numbers were assessed using the MTT-assay as described in Materials and methods, and are expressed as percentage of control values. Each value represent the mean of two separate experiments 6 SD, each performed with eight parallels.

(1.7- to 1.8-fold) and SelD mRNA (1.2- to 1.4-fold) in A-427 cells pretreated with selenium and then treated with DHA or mock treated (Figure 7, Table III). DHA treatment alone was accompanied by a slight decrease in both GSH-Px and SelD mRNAs. In A-172 cells the level of GSH-Px mRNA was virtually unchanged by selenium and/or DHA-supplementation, while there was a slight increase in the level of SelD mRNA (Figure 7, Table III). Somewhat surprisingly, the levels of GSH-Px mRNA as well as SelD mRNA were essentially similar in A-427 cells and A-172 cells in spite of the lower activity of GSH-Px in A-427 cells as compared with A-172 cells. mRNA for the house-keeping protein GAPDH, which is frequently used to normalize results, was also induced in A427 cells pretreated with selenium and this can apparently not

be ascribed to differences in amounts of RNA loaded on the gel since the ethidium bromide staining indicated even loading (Figure 7, lower panel). The increase in GAPDH mRNA in A-427 cells that grow better after treatment with sodium selenite is consistent with the previously reported upregulation of GAPDH mRNA and protein in the S-phase of cultured cells and possibly precludes the use of GAPDH for normalizing mRNA levels (44). Because of this, we prefer to compare mRNA for GSH-Px and SelD to untreated cells rather than normalizing relative to GAPDH mRNA. The increases in GSH-Px and SelD mRNAs are both likely to contribute to the increase in GSH-Px activity observed after selenium pretreatment of A-427 cells. Discussion In the present study we show that n-3 fatty acids inhibit the growth of the human lung carcinoma cell line A-427 in vitro, whereas the lung adenocarcinoma cell line SK-LU-1 and glioblastoma cell lines A-172 and U-87 MG are relatively resistant. The sensitive cell line examined in the present study (A-427) appears to have a deficient expression of GSH-Px as a basis for its sensitivity. Presently, we do not know how common such a cause of sensitivity to n-3 fatty acids is. In fact, we do not know how common sensitivity to n-3 fatty acids among tumour cells is, since only one cell line was 1901


examined in most studies. However, a few reports, as well as unpublished results from our laboratory, indicate that there are significant differences in sensitivities of different cell lines to n-3 and n-6 PUFAs (Nøding et al, unpublished results). In addition, it has been observed that the concentration by which PUFAs inhibit cell growth varies with cell densities as well as with the microenvironment of the cell culture (45). We have made similar observations. Thus, we have repeatedly observed that SK-LU-1 responds towards DHA and EPA in a dose- and time-dependent manner when grown in tissue culture bottles, but we have not been able to reproduce these results in 96well microplates. Rose and Connolly (46) observed that DHA and EPA inhibited growth of the prostate cancer cell line PC3 in a dose-dependent manner, whereas the effect was less pronounced in the prostate cancer cell line DU 145. The effect of PUFAs on the growth of various breast cancer cell lines was examined by Chajes et al. (47). They found EPA and DHA to be effective in arresting cell growth in all cell lines but one. α- and γ-linolenic acid exerted variable effects on cell proliferation depending on the cell line investigated. A recent study indicated that there is a correlation between sensitivities of cell lines in vitro and in vivo after implantation in nude mice (7). These results indicate that differential sensitivities of different cell lines to fatty acids are not exclusively an in vitro phenomenon and also indicate that some cells are sensitive enough to make intake of a concentrated preparation of n-3 fatty acids a possible therapeutic supplement. In agreement with previous studies (46,47), our present results strongly support the view that lipid peroxidation products are major cytotoxic products in tumour cells exposed to n-3 fatty acids. These results are best explained by differential sensitivity of different tumour cell lines to metabolites formed during lipid peroxidation, since some of the resistant cell lines produced higher levels of malondialdehyde than the most sensitive cell line. Our data demonstrate that the basis for this differential sensitivity is related to the level of cellular expression of GSH-Px, a central enzyme in the defence against oxidative stress. Since the lung carcinoma cell line A-427 (highly sensitive) and the lung adenocarcinoma cell line SKLU-1 (resistant), as well as the glioblastoma cell line U-87 MG (resistant) were all cultured in an identical culture medium supplemented with FCS from the same batch selected for its ability to support growth rates for all the cell lines used, the differences in sensitivity cannot be explained by different culture conditions. This is more than a trivial fact because some commercial media and sera do not supply cultured cells with sufficient amounts of selenium and this may lead to low expression of glutathione peroxidases (48). Our data also demonstrate that different cell lines vary in their abilities to utilize the trace amounts of selenium present in the medium. Furthermore, the present data suggest that the increase in GSH-Px activity observed in A-427 cells after pretreatment with selenium is at least in part due to increased levels of GSH-Px and selenophosphate synthetase (selD) mRNAs. However, the pathway for generation of tRNASec in human cells has not been completely established and the rate limiting factor(s) in biosynthesis of GSH-Px is not known; consequently, these results must be interpreted with caution. Furthermore, we cannot exclude the possibility that biosynthesis of other selenoproteins may also be induced by addition of selenite and that this may contribute to protection against toxic effects of PUFAs. Our results are consistent with those of Ochi et al. (35) who 1902

found that sodium selenite and ebselen, a GSH-Px mimic, prevented 15-HPETE-induced injury of endothelial cells in a concentration dependent manner. The fact that ebselen was also able to suppress the inhibitory effect of DHA on A-427, supports our view that GSH-Px plays an important role in protection against DHA-induced cytotoxicity. Hydroperoxides of DHA, being substrates for GSH-Px, are among the likely candidates for mediating the cytotoxic effect. GSH is known to scavenge free radicals in combination with GSH-Px or by direct interaction. Our results show that the cellular GSH level varies considerably between different cell lines, but there seems to be no clear correlation between the cellular GSHlevel and sensitivity of different cells to n-3 fatty acids, since the resistant cell line U-87 MG and the sensitive cell line A427 had an essentially similar level of GSH. Apparently, most normal cells have an excess of GSH, whereas the GSH levels of malignant cells are frequently close to the level required for cell survival (27,49–51). Several studies demonstrate that depletion of intracellular GSH level sensitizes cells to oxidative damage (37,52). It is reasonable to believe that low GSH level, low GSH synthesizing capacity and low GSH-Px activity would be a determinant for the sensitivity of different cell lines to oxidative damage. O2 enhances the toxic effects of n3 fatty acids in cell culture (53). It is conceivable that administration of O2 could enhance the effect of PUFAs in sensitive tumours, alone, or in combination with other forms of treatment, such as oxidative stress induced by γ-irradiation. Although dietary fat, or addition of certain fatty acids to cultured cells, undoubtedly affects several parameters, including growth and invasiveness, the significance of dietary fat in the development of human cancer is still a matter of debate. Thus, an association between high intake of saturated fat and human cancer, especially breast and colorectal cancer, has been reported in a number of epidemiological case-control studies. However, a recent review of seven prospective studies demonstrated no significantly increased risk for breast cancer due to intake of fat, including saturated fat (54 and references therein). Similarly, a prospective study on diet and colon cancer found no evidence for an association with dietary fat (55). It should be pointed out, however, that in studies with experimental animals, the dietary changes are usually extensive compared with variations in diet between different human individuals. Possibly, epidemiology as a tool may not have the sensitivity required for detection of weak associations that might actually exist. Furthermore, the possible lack of association between dietary fat and human cancer does not rule out the possible use of special dietary regimes, such as high concentrates of n-3 fatty acids, as an adjuvant in cancer treatment. Such a treatment would be more likely to be successful if applied to cases where tumours were known to be deficient in GSH-Px, which could be determined from biopsies obtained during surgery. In conclusion, lipid peroxidation products are responsible for the cytotoxic effects of n-3 fatty acids in at least some tumour cell lines, since vitamin E restores cell growth and abolishes lipid peroxidation. The lack of correlation between the amount of lipid peroxidation products formed and inhibition of cell growth between different cell lines indicates a difference in sensitivity of different cell lines to lipid peroxidation products. Our data demonstrate that these differences most probably are due to variation in the GSH-Px activity in at least some cell lines.


Acknowledgements This work was supported by the Norwegian Research Council and the Norwegian Cancer Society, Norsk Hydro Research Center, Porsgrunn, Norway, and Pronova Biocare AS, Oslo, Norway. We also would like to thank the laboratories of Dr C.W.Welsch and Dr C.P.Burns for technical advice related to methods for reproducible measurements of lipid peroxidation. We also thank Dr S.Qichang for generously providing us with the human GPX1 cDNA clone and Dr S.C.Low and Dr M.Berry for providing the selD cDNA clone.

References 1. Reddy,B.S., Burill,C. and Rigotty,J. (1991) Effect of diets high in ω-3 and ω6 fatty acids on initiation and postinitiation stages of colon carcinogenesis. Cancer Res., 51, 487–491. 2. Karmali,R.A., Donner,A., Gobel,S. and Shimamura,T. (1989) Effect of n3 and n-6 fatty acids on 7,12 dimethylbenz(a)anthracene-induced mammary tumourigenesis. Anticancer Res., 9, 1161–1168. 3. O‘Conner,T.P., Roebuck,B.D., Peterson,F.J., Lokesh,B., Kinsella,J.E. and Campbell,T.C. (1989) Effect of dietary omega-3 and omega-6 fatty acids on development of azaserine-induced preneoplastic lesions in rat pancreas. J. Natl Cancer Inst., 81, 858–863. 4. Karmali,R.A., Reichel,P., Cohen,L.A., Terano,T., Hirai,A., Tamura,Y. and Yoshida,S. (1987) The effects of dietary ω-3 fatty acids on the DU-145 transplantable human prostatic tumour. Anticancer Res., 7, 1173–1180. 5. Karmali,R.A, Marsh,J. and Fuchs,C. (1984) Effect of omega-3 fatty acids on growth of a rat mammary tumour. J. Natl Cancer Inst., 73, 457–461. 6. Gabor,H. and Abraham,S. (1986) Effect of dietary menhaden oil on tumour cell loss and the accumulation of mass of a transplantable mammary adenocarcinoma in BALB/c mice. J. Natl Cancer Inst., 76, 1223–1229. 7. Mæhle,L., Eilertsen,E., Mollerup,S., Schønberg,S., Krokan,H.E. and Haugen,Aa. (1995) Effects of n-3 fatty acids during neoplastic progression and comparison of in vitro and in vivo sensitivity of two human tumour cell lines. Br. J. Cancer, 71, 691–696. 8. Reich,R., Royce,L. and Martin,G.R. (1989) Eicosapentaenoic acid reduces the invasive and metastatic activities of malignant tumour cells. Biochem. Biophys. Res. Commun., 160, 559–564. 9. Beck,S.A., Smith,K.L. and Tisdale,M.J. (1991) Anticachectic and antitumour effect of eicosapentaenoic acid and its effect on protein turnover. Cancer Res., 51, 6089–6093. 10. Burns,C.P., Luttenegger,D.G., Dudley,D.T., Buettner,G.R. and Spector,A.A. (1979) Effect of modification of plasma membrane fatty acid composition on fluidity and methotrexate transport in L1210 murine leukemia cells. Cancer Res., 39, 1726–1732. 11. Yorek,M.A., Strom,D.K. and Spector,A.A. (1984) Effect of membrane polyunsaturation on carrier-mediated transport in cultured retinoblastoma cells: alteration in taurine uptake. J. Neurochem., 42, 254–261. 12. Burns,C.P. and Spector,A.A. (1987) Membrane fatty acid modification in tumour cells: a potential therapeutic adjunct. Lipids, 22, 178–184. 13. Zulstra,J.G., De Vries,E.G.E., Muskiet,F.A.J., Martini,I.A., TimmerBosscha,H. and Mulder,N.H. (1987) Influence of docosahexaenoic acid in vitro on intracellular adriamycin concentration in lymphocytes and human adriamycin-sensitive and -resistant small-cell lung cancer cell lines, and on cytotoxicity in the tumour cell lines. Int. J. Cancer, 40, 850–856. 14. Gould,R.J., Ginsberg,B.H. and Spector,A.A. (1982) Lipid effects on the binding properties of a reconstituted insulin receptor. J. Biol. Chem., 257, 477–484. 15. Matsumoto,K., Morita,I., Hibino,H. and Murota,S. (1993) Inhibitory effect of docosahexaenoic acid-containing phosholipids on 5-lipoxygenase in rat basophilic leukemia cells. Prostaglandins Leukotrienes Essential Fatty Acids, 49, 861–866. 16. Awad,A.B., Ferger,S.L. and Fink,C.S. (1990) Effect of dietary fat on the lipid composition and utilization of short-chain fatty acids by rat colonocytes. Lipids, 25, 316–320. 17. Fischer,S., Vischer,A., Preac-Mursic,V. and Weber,P.C. (1987) Dietary docosahexaenoic acid is retroconverted in man to eicosapentaenoic acid, which can be quickly transformed to prostaglandin I3. Prostaglandins, 34, 367–375. 18. Karmali,R.A. (1987) Omega-3 Fatty Acids and Cancer: A Review. In Lands,W.E.M. (ed.) Polyunsaturated Fatty Acids and Eicosanoids. The American Oil Chemists’ Society, Champaign, pp. 222–232. 19. Corey,E.J., Shih,C. and Cashman,J.R. (1983) Docosahexaenoic acid is a strong inhibitor of prostaglandin but not lekotriene biosynthesis. Proc. Natl Acad. Sci., 80, 3581–3584. 20. Gonzalez,M.J. and Welsch,C.W. (1993) Inhibition of mammary gland tumourigenesis by dietary fish oil: a phenomenon associated with increased levels of tumour lipid peroxidation products? In Drevon,C.A., Baksaas,I.

and Krokan,H.E. (ed.) Omega-3 Fatty Acids: Metabolism and Biological Effects. Birkha¨user Verlag, Basel/Switzerland, pp. 315–326. 21. Cheeseman,K.H., Emery,S., Maddix,S.P., Slater,T.F., Burton,G.W. and Ingold,K.U. (1988) Studies on lipid peroxidation in normal and tumour tissues. Biochem. J., 250, 247–252. 22. Sagar,P.S., Das,U.N., Koratkar,R., Ramesh,G., Padma,M. and Kumar,G.S. (1992) Cytotoxic action of cis-unsaturated fatty acids on human cervical carcinoma (HeLa) cells: relationship to free radicals and lipid peroxidation and its modulation by calmodulin antagonists. Cancer Lett., 63, 189–198. 23. Burns,C.P. and Wagner,B.A. (1991) Heightened susceptibility of fish oil polyunsaturate-enriched neoplastic cells to ethane generation during lipid peroxidation. J. Lipid Res., 32, 79–87. 24. Esterbauer,H., Zollner,H. and Schaur,R.J. (1988) Hydroxyalkenals: cytotoxic products of lipid peroxidation. ISI Atlas of Science, 0894-3753, 311–317. 25. Krokan,H.E., Grafstrom,R.C., Sundqvist,K., Esterbauer,H. and Harris,C.C. (1985) Cytotoxicity, thiol depletion and inhibition of O6-methylguanineDNA methyltransferase by various aldehydes in cultured human bronchial fibroblasts. Carcinogenesis, 6, 1755–1759. 26. Kumar,K., Thangaraju,M. and Sachdanandam,P. (1991) Changes observed in antioxidant system in the blood of postmenopausal women with breast cancer. Biochem. Int., 25, 371–380. 27. Corrocher,R., Casaril,M., Bellisola,G., Gabrielli,G.B., Nicoli,N., Guidi,G.C. and De Sandre,G. (1986) Severe impairment of antioxidant system in human hepatoma. Cancer, 58, 1658–1662. 28. Michiels,C., Raes,M., Toussaint,O. and Remacle,J. (1994) Importance of Se-glutathione peroxidase, catalase, and Cu/Zn-SOD for cell survival against oxidative stress. Free Radicals Biol. Med., 17, 235–248. 29. Flohe,L. (1985) The glutathione peroxidase reaction: molecular basis of the antioxidant function of selenium in mammals. Curr. Top. Cell. Regulat., 27, 473–478. 30. Bo¨ck,A., Forchhammer,K., Heider,J., Leinfelder,W., Sawers,G., Veprek,B. and Zinoni, F. (1991). Selenocysteine: the 21st amino acid, microReview. Mol. Microbiol., 5, 515–520. 31. Low,S.C., Harney,J.W. and Berry,M.J. (1995). Cloning and functional characterization of human selenophosphate synthase, an essential component of selenoprotein synthesis. J. Biol. Chem., 270, 21659–21664. 32. Chu,F., Esworthy,R.S., Doroshow,J.H., Doan,K. and Liu,X. (1992) Expression of plasma glutathione peroxidase in human liver in addition to kidney, heart, lung and breast in humans and rodents. Blood, 79, 3233–3238. 33. Carmichael,J., Forrester,L.M., Lewis,A.D., Hayes,J.D. Hayes,P.C. and Wolf,C.R. (1988) Glutathione S-transferase isoenzymes and glutathione peroxidase activity in normal and tumour samples from human lung. Carcinogenesis, 9, 1617–1621. 34. Flohe,L. (1982) Glutathione peroxidase brought into focus. In Pryor, W.A. (ed.) Free Radicals in Biology. Academic Press, New York, vol. 5, pp. 223–254. 35. Ochi,H., Morita,I. and Murota,S. (1992) Roles of glutathione and glutathione peroxidase in the protection against endothelial cell injury induced by 15-hydroperoxyeicosatetraenoic acid. Archives biochem. biophys., 294, 407–411. 36. Harlan,J.M., Levine,J.D., Callahan,K.S. and Schwartz,B.R. (1984) Glutathione redox cycle protects cultured endothelial cells against lysis by extracellularly generated hydrogen peroxide. J. Clin. Invest., 73, 706–713. 37. Arrick,B.A., Nathan,C.F., Griffith,O.W. and Cohn,Z.A. (1982) Glutathione depletion sensitizes tumour cells to oxidative cytolysis. J. Biol. Chem., 257, 1231–1237. 38. Geiger,P.G., Lin,F. and Girotti,A.W. (1993) Selenoperoxidase-mediated cytoprotection against the damaging effects of tert-butyl hydroperoxide on leukemia cells. Free Radicals Biol. Med., 14, 251–266. 39. Mosmann,T. (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J. Immunol. Meth., 65, 55–63. 40. Gavino,V.C., Miller,J.S., Ikharebha,S.O., Milo,G.E. and Cornwell,D.G. (1981) Effect of polyunsaturated fatty acids and antioxidants on lipid peroxidation in tissue cultures. J. Lipid Res., 22, 763–769. 41. Wendel,A. (1980) Glutathione peroxidase. In Jakoby,W.B. (ed.) Enzymatic Basis of Detoxification. Academic Press, New York, vol. 1, pp. 333–353. 42. Griffith,O.W. (1980) Determination of glutathione and glutathione disulfide using glutathione reductase and 2-vinylpyridine. Analyt. Biochem., 106, 207–212. 43. Chada,S., LeBeau,M.M., Casey,L. and Newberger,P.E. (1990). Isolation and chromosomal localization of the human glutathione peroxidase gene. Genomics, 6, 268–271.

1903

S.A.Schønberg et al. 44. Mansur,N.R., Meyer-Siegler,K., Wurzer,J.C. and Sirover,M.A. (1993). Cell cycle regulation of the glyceraldehyde-3-phosphate dehydrogenase/uracil DNA glycosylase gene in normal human cells. Nucl. Acids Res., 21, 993–998. 45. Falconer,J.S., Ross,J.A., Fearon,K.C.H., Hawkins,R.A., O’Riordain,M.G. and Carter,D.C. (1994) Effect of eicosapentaenoic acid and other fatty acids on the growth in vitro of human pancreatic cancer cell lines. Br. J. Cancer, 69, 826–832. 46. Rose,D.P. and Connolly,J.M. (1991) Effects of fatty acids and eicosanoid synthesis inhibitors on the growth of two human prostate cancer cell lines. Prostate, 18, 243–254. 47. Chajes,V., Sattler,W., Stranzl,A. and Kostner,G.M. (1995) Influence of n3 fatty acids on the growth of human breast cancer cells in vitro: relationship to peroxides and vitamin-E. Breast Cancer Res. Treat., 34, 199–212. 48. Leist,P., Raab,B., Maurer,S., RosickU. and Brigeliusflohe,R. (1996) Conventional cell-culture media do not adequately supply cells with antioxidants and thus facilitate peroxide-induced genotoxicity. Free Radicals Biol. Med., 21, 297–306. 49. Meister,A. and Anderson,M.E. (1983) Glutathione. Ann. Rev. Biochem., 52, 711–760. 50. Arrick,B.A. and Nathan,C.F. (1984) Glutathione metabolism as a determinant of therapeutic efficacy: a review. Cancer Res., 44, 4224–4232. 51. Kosower,N.S. (1978) The glutathione status of cells. Int. Rev. Cyt., 54, 109–160. 52. Toborek,M. and Hennig,B. (1994) Fatty acid-mediated effects on the glutathione redox cycle in cultured endothelial cells. Am. J. Clin. Nutr., 59, 60–65. 53. Spitz,D.R., Kinter,M.T., Kehrer,J.P. and Roberts,R.J. (1992) The effect of monounsaturated and polyunsaturated fatty acids on oxygen toxicity in cultured cells. Pediatric Research, 32, 366–372. 54. Hunter,D.J., Spiegelman,D., Adami,H-O. et al. (1996). Cohort studies of fat intake and the risk of breast cancer — a pooled analysis. N. Engl. J. Med., 334, 356–361. 55. Giovannucci,E., Rimm,E.B., Stampfer,M.J., Colditz,G.A., Ascherio,A. and Willett,W.C. (1994). Intake of fat, meat, and fiber in relation to risk of colon cancer in men. Cancer Res., 54, 2390–2397. Received on December 12, 1996; revised on July 7, 1997; accepted on July 8, 1997

1904