Closely related colon cancer cell lines display different
sensitivity to polyunsaturated fatty acids, accumulate
different lipid classes and downregulate sterol regulatory
element-binding protein 1
Svanhild A. Schønberg
1
, Anne G. Lundemo
1
, Torill Fladvad
1
, Kristin Holmgren
2
, Hilde Bremseth
2
,
Asbjørn Nilsen
2
, Odrun Gederaas
2
,Ka
˚re E. Tvedt
1
, Kjartan W. Egeberg
2
and Hans E. Krokan
2
1 Department of Laboratory Medicine, Children’s and Women’s Health, Norwegian University of Science and Technology, Trondheim,
Norway
2 Department of Cancer Research and Molecular Medicine, Norwegian University of Science and Technology, Trondheim, Norway
Keywords
cell-cycle arrest; lipid droplets; lipid
peroxidation; polyunsaturated fatty acids;
SREBP1
Correspondence
S. A. Schønberg, Department of Laboratory
Medicine, Children’s and Women’s Health,
Norwegian University of Science and
Technology, Laboratoriesenteret, St. Olav’s
Hospital, N-7006 Trondheim, Norway
Fax: +47 725 76400
Tel: +47 725 73362
E-mail: svanhild.schonberg@ntnu.no
(Received 10 March 2006, accepted 24 April
2006)
doi:10.1111/j.1742-4658.2006.05292.x
N-6 polyunsaturated fatty acids (PUFAs) may be associated with increased
risk of colon cancer, whereas n-3 PUFAs may have a protective effect. We
examined the effects of docosahexaenoic acid (DHA), eicosapentaenoic
acid and arachidonic acid on the colon carcinoma cell lines SW480 derived
from a primary tumour, and SW620 derived from a metastasis of the same
tumour. DHA had the strongest growth-inhibitory effect on both cell lines.
SW620 was relatively more growth-inhibited than SW480, but SW620 also
had the highest growth rate in the absence of PUFAs. Flow cytometry
revealed an increase in the fraction of cells in the G
2
⁄M phase of the cell
cycle, particularly for SW620 cells. Growth inhibition was apparently not
caused by increased lipid peroxidation, reduced glutathione or low activity
of glutathione peroxidase. Transmission electron microscopy revealed for-
mation of cytoplasmic lipid droplets after DHA treatment. In SW620 cells
an eightfold increase in total cholesteryl esters and a 190-fold increase in
DHA-containing cholesteryl esters were observed after DHA treatment. In
contrast, SW480 cells accumulated DHA-enriched triglycerides. Arachidon-
ic acid accumulated in a similar manner, whereas the nontoxic oleic acid
was mainly incorporated in triglycerides in both cell lines. Interestingly,
nuclear sterol regulatory element-binding protein 1 (nSREBP1), recently
associated with cell growth regulation, was downregulated after DHA
treatment in both cell lines. Our results demonstrate cell-specific mecha-
nisms for the processing and storage of cytotoxic PUFAs in closely related
cell lines, and suggest downregulation of nSREBP1 as a possible contribu-
tor to the growth inhibitory effect of DHA.
Abbreviations
AA, arachidonic acid; ACAT, acyl CoA:cholesterol acyltransferase; BHA, butylated hydroxyanisole; BHT, butylated hydroxytoluene; COX,
cyclooxygenase; DGAT, diacylglycerol acyltransferase; DHA, docosahexaenoic acid; EPA, eicosapentaenoic acid; FAMEs, fatty acid methyl
esters; GSH, glutathione; GSH-Px, glutathione peroxidase; h-ALAS, human 5-aminolevulinate synthase; h-PBGD, human porphobilinogen
deaminase; LOX, lipoxygenase; LPPs, lipid peroxidation products; MDA, malondialdehyde; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide; NDGA, nordihydroguaiaretic acid; NS-398, N-[2-cyclohexyloxy)-4-nitrophenyl]-methanesulfonamide; OA, oleic
acid; PI, propidium iodide; PUFAs, polyunsaturated fatty acids; SOD, superoxide dismutase; SREBP, sterol regulatory element-binding
protein; TBA, 2-thiobarbituric acid; TBARS, thiobarbituric acid reactive substances; TUNEL, terminal deoxynucleotidyl transferase (TdT)-
mediated dUTP-biotin nick end labelling.
FEBS Journal 273 (2006) 2749–2765 ª2006 The Authors Journal compilation ª2006 FEBS 2749
Epidemiological studies indicate that there is an
inverse association between intake of polyunsaturated
fatty acids (PUFAs) and incidence of breast, colon
and prostate cancers, although these studies are not
uniformly conclusive [1–4]. Also, cancer xenografts in
immunosuppressed mice as well as cell culture studies
demonstrate that PUFAs may slow down cancer cell
growth, induce apoptosis and increase the efficiency of
chemotherapeutic drugs [5–8]. The mechanisms behind
these effects are clearly complex and not well under-
stood. However, factors possibly implicated in the
PUFA-mediated effects include modification of tumour
cell membranes which can affect cell signalling path-
ways [9], lipid peroxidation and oxidative stress [10],
eicosanoid production [11], fatty acid metabolism [12]
and the regulation of gene expression [13]. Also, the
activity of different antioxidant defence enzymes seems
to vary among different cell lines and between normal
and malignant cells, and may be of importance for
cancer cells susceptibility towards n-3 PUFAs [14,15].
One of the major mechanisms of PUFA-mediated tox-
icity on cancer cells is thought to be the lipid peroxida-
tion process, which is initiated by free radical attack
on membrane PUFAs, leading to the formation of a
wide range of very reactive and genotoxic lipid peroxi-
dation products (LPPs) [16]. Altogether, these data
suggest that intake of n-3 PUFAs may modulate cell
behavior and growth by a variety of mechanisms.
Previously we have shown that the growth inhibitory
effect of docosahexaenoic acid (DHA) and eicosapen-
taenoic acid (EPA) on A-427 cells (human lung adeno-
carcinoma cell line) may be reversed by vitamin E,
sodium selenite and ebselen, a synthetic glutathione
peroxidase (GSH-Px) mimic, indicating that lipid per-
oxidation is responsible for the cytotoxic effect [10].
The high PUFA sensitivity of this cell line seems to be
caused by a low level of the antioxidant defence
enzyme GSH-Px. We have also shown that this cell
line is sensitive to the hydroperoxy derivatives of the
PUFAs, indicating that cytotoxicity of n-3 PUFAs
may be mediated via the formation of these primary
products [17].
In this study we wanted to explore possible mecha-
nisms of cytotoxicity induced by PUFAs. For these
studies we used two colon carcinoma cell lines, SW480
and SW620, derived from a primary and a secondary
tumour of the same patient, respectively. We find that
both cell lines are growth-inhibited by PUFAs, although
SW620 cells display a higher degree of PUFA sensitivity
than SW480 cells. Our results indicate that lipid peroxi-
dation or deficient defence against oxidants is not
involved in sensitivity, contrary to some other cell lines.
Rather, our results indicate that differences between cell
lines in regard to processing and storage of potentially
harmful PUFAs and downregulation of nuclear sterol
regulatory element-binding protein 1 (nSREBP1) may
be related to effects on cancer cell growth.
Results
Growth inhibitory effects of n-3 and n-6 PUFAs
The effects of DHA, EPA and arachidonic acid (AA)
on the growth of SW480 and SW620 are shown in
Fig. 1. All three PUFAs inhibited the growth of both
cell lines in a time- and concentration-dependent man-
ner. Cell proliferation of SW480 was reduced by 77%
compared with control cultures after exposure to
70 lmDHA for 144 h, while EPA and AA reduced
cell growth by 50 and 44%, respectively. Cell prolifer-
ation of SW620 was somewhat more affected; DHA
decreased cell proliferation by 95% after 144 h incuba-
tion compared with control cultures, whereas EPA
and AA reduced cell proliferation by 75% each. The
strongest effect on cell growth was seen with DHA for
both cell lines, which is in agreement with previous
results using a different cell line [10]. SW620 was relat-
ively more growth-inhibited than SW480, but SW620
also had the highest growth rate in the absence of
PUFAs (Fig. 1). Oleic acid (OA) had no effect on cell
growth (results not shown).
Cell-cycle analysis and apoptosis
Total cellular DNA content and DNA fragmentation
were analysed by flow cytometry. Treatment of the cells
for 96 h with DHA (70 lm) resulted in a significant
increase in the population of cells in the G
2
⁄M phase
compared with control cells, 1.7- and 2.5-fold for
SW480 and SW620, respectively (Fig. 2). However, the
strong inhibitory effect on growth indicates that other
phases must also be affected. After 120 h, the popula-
tion of cells arrested in the G
2
⁄M phase was diminished
to 1.3- and 1.7-fold increase relative to controls for
SW480 and SW620, respectively (data not shown). The
terminal deoxynucleotidyl transferase (TdT)-mediated
dUTP-biotin nick end labelling (TUNEL) assay
revealed no evidence for DNA-fragmentation and
therefore no evidence for apoptosis (data not shown).
Effect of vitamin E, acyl CoA:cholesterol
acyltransferase inhibitor and inhibitors of eicosa-
noid synthesis on DHA-induced growth inhibition
In order to determine whether vitamin E could coun-
teract the DHA-induced growth inhibition of the two
DHA-induced toxicity in colon cancer S. A. Schønberg et al.
2750 FEBS Journal 273 (2006) 2749–2765 ª2006 The Authors Journal compilation ª2006 FEBS
cell lines, SW480 and SW620 were treated with DHA
(70 lm) and vitamin E (10 or 50 lm) simultaneously.
Vitamin E was not able to prevent growth inhibition
in either cell line and even enhanced the antiprolifera-
tive effect of DHA in SW480 cells (Fig. 3). Vitamin E
alone had no effect on cell survival (data not shown).
We also examined whether the lipoxygenase inhibitor
nordihydroguaiaretic acid (NDGA) and the cyclooxyg-
enase inhibitor N-[2-cyclohexyloxy)-4-nitrophenyl]-
methanesulfonamide (NS-398) could reverse the
growth inhibition induced by DHA in these two cell
lines. When the cells were coincubated with DHA
(70 lm) and NDGA (0.01–1 lm) or NS-398 (0.01–
10 lm), no effect was seen on cell survival (data not
shown). NDGA and NS-398 had no effect on cell
survival alone at the concentrations indicated. Also,
coincubation of cells with the acyl CoA:cholesterol
acyltransferase (ACAT) inhibitor Sandoz 58-035 (2.1–
8.5 lm) and DHA, had no effect on cell survival (data
not shown).
Lipid peroxidation
Because generation of free radicals and subsequent
lipid peroxidation are important for PUFA-induced
toxicity in several tumour cell lines, we measured the
level of malondialdehyde (MDA), a major secondary
lipid peroxidation product, after DHA treatment alone
or in combination with vitamin E, both in cells and
culture medium. A several fold increase in the MDA
level was observed in both SW480 and SW620 cells
after DHA treatment (70 lm) for 72 h when compared
Fig. 1. Effect of docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA) and arachidonic acid (AA) on the growth of the human colon car-
cinoma cell lines SW480 and SW620. Cell survival was assessed using the MTT assay as described in Experimental procedures. The values
represent the mean of at least four parallels ± SD from one experiment, and are verified through at least two experiments. D, attenuance.
S. A. Schønberg et al. DHA-induced toxicity in colon cancer
FEBS Journal 273 (2006) 2749–2765 ª2006 The Authors Journal compilation ª2006 FEBS 2751
with control (Table 1). A combinatory treatment with
DHA (70 lm) and vitamin E (50 lm) reduced the
MDA level below the level in control cultures. Only
minor amounts of MDA were found in the culture
medium and the level did not exceed that in control
cultures of the cell lines.
Effect of DHA on glutathione levels and GSH-Px
activities
The total level of glutathione (GSH) in SW480 and
SW620 cells is shown in Fig. 4. The level of GSH is
approximately equal in the two cell lines and
comparable with the GSH level in other PUFA-sensi-
tive cell lines [10]. An approximately twofold increase
in the total amount of glutathione was observed in
both cell lines after DHA treatment for 48 h. This
increase may represent a rebound effect, which was
similar in the two cell lines. However, the sensitivity of
the cell lines to PUFAs does not seem to correlate with
the GSH level.
To examine whether deficiencies in the antioxidant
defence mechanisms could be used as a determinant
for PUFA sensitivity, we also measured the GSH-Px
activities in the two cell lines (Fig. 5). Compared with
the PUFA-senstitive cell line A-427, in which the
540
450
360
270
180
90
00 64 128 192 256 320 384 448 512
540
450
360
270
180
90
00 64 128 192 256 320 384 448 512
540
450
360
270
180
90
00 64 128 192 256 320 384 448 51
2
540
450
360
270
180
90
00 64 128 192 256 320 384 448 512
AC
BD
Fig. 2. Docosahexaenoic acid (DHA)-induced G
2
⁄M arrest in SW480 and SW620 cells. Cells were treated with DHA (70 lM) for the indicated
times, stained with propidium iodide (PI) and examined by flow cytometry as described in Experimental procedures. (A) Untreated SW480
cells after 96 h; (B) SW480 cells treated with DHA (70 lM) for 96 h; (C) untreated SW620 cells after 96 h; (D) SW620 cells treated with
DHA (70 lM) for 96 h.
DHA-induced toxicity in colon cancer S. A. Schønberg et al.
2752 FEBS Journal 273 (2006) 2749–2765 ª2006 The Authors Journal compilation ª2006 FEBS
activity level was very low [10], the activity level in
SW480 and SW620 was significantly higher, about
two- and sixfold, respectively. However, the fact that
the most PUFA-sensitive cell line, SW620, had a 2.5-
fold higher level of activity compared with SW480,
indicates that there is no relationship between the
activity level of GSH-Px and PUFA sensitivity in this
case. DHA treatment (70 lm) alone and pretreatment
with sodium selenite (250 nm) for 20 h before DHA
administration did not lead to any significant change
in the GSH-Px activity level in the two cell lines. In
accordance with this, pretreating SW480 and SW620
with sodium selenite (250 nm) did not affect cell
growth after DHA treatment (data not shown). This
Fig. 3. Lack of effect of vitamin E (10 and 50 lM) on DHA-induced
growth inhibition in SW480 and SW620 cells at different time
points. Cell survival was assessed using the MTT assay as des-
cribed in Experimental procedures. Control, n; Control ⁄EtOH, h;
DHA (70 lM), ; DHA(70 lM)⁄Vit. E (10 lM), ; DHA (70 lM)⁄Vit.E
(50 lM), . The values represent the mean of at least four parallels
± SD from one experiment, and are verified through at least two
experiments.
Table 1. Effect of DHA alone or in combination with vitamin E on
lipid peroxidation measured as MDA
a
in SW480 and SW620 cells.
Cell line ⁄treatment
MDA (nmolÆmg protein
)1
)±SD
Cell extract Released into media
SW480
Control 3.20 ± 0.21 0.64 ± 0.15
DHA (70 lM) 44.82 ± 0.71 3.00 ± 0.40
DHA (70 lM)⁄vit.E (50 lM) 2.46 ± 0.29 2.17 ± 0.24
SW620
Control 1.13 ± 0.09 1.06 ± 0.08
DHA (70 lM) 12.62 ± 0.60 0.88 ± 0.07
DHA (70 lM)⁄vit.E (50 lM) 0.48 ± 0.05 0.79 ± 0.07
a
The value is the mean calculated from measurements performed
in triplicate ± SD in one of two representative experiments.
Fig. 4. Effect of 48 h treatment with DHA (70 lM) on the GSH level
in the human colon carcinoma cell lines SW480 and SW620. Values
represent the mean ± SD of triplicate measurements from three
separate experiments. *Results analysed by Student’s t-test and
considered significant different from control (P<0.05).
Fig. 5. Effect of 48 h treatment with DHA (70 lM) on the GSH-Px
activity in the human colon carcinoma cell lines SW480 and
SW620. GSH-Px activity was measured with or without pretreat-
ment with sodim selenite (250 nM) for 20 h before DHA supple-
mentation for 48 h in both cell lines. Results are expressed in
nmol NADPH oxidizedÆmin
)1
Æmg
)1
protein. Values represent the
mean ± SD from one experiment. Each determination was per-
formed in triplicate, and verified through at least two experiments.
S. A. Schønberg et al. DHA-induced toxicity in colon cancer
FEBS Journal 273 (2006) 2749–2765 ª2006 The Authors Journal compilation ª2006 FEBS 2753