Atherosclerosis 197 (2008) 524533

Orally administered eicosapentaenoic acid reduces and stabilizes atherosclerotic lesions in ApoE-decient mice
Miwa Matsumoto a , Masataka Sata a,b, , Daiju Fukuda a , Kimie Tanaka a , Masaaki Soma c , Yasunobu Hirata a , Ryozo Nagai a
a

Department of Cardiovascular Medicine, University of Tokyo Graduate School of Medicine, Tokyo 113-8655, Japan b Department of Advanced Clinical Science and Therapeutics, University of Tokyo Graduate School of Medicine, Tokyo 113-8655, Japan c Mochida Pharmaceutical Co., Ltd., Tokyo, Japan Received 6 March 2007; received in revised form 7 July 2007; accepted 15 July 2007 Available online 4 September 2007

Abstract Accumulating evidence demonstrates that dietary intake of n-3 polyunsaturated fatty acids (PUFAs) is associated with reduced incidence of cardiovascular events. However, the molecular mechanisms by which n-3 PUFAs prevent atherosclerosis are not fully understood. Here, we examined the effect of eicosapentaenoic acid (EPA), a major n-3 PUFA, on the pathogenesis of atherosclerosis in ApoE-decient mice. Five-week-old ApoE-decient male mice were fed on western-type diet supplemented with 5% (w/w) EPA (EPA group, n = 7) or not (control group, n = 5) for 13 weeks. An analysis of the fatty acid composition of liver homogenates revealed a marked increase of the n-3 PUFA content in the EPA group (n-3/n-6 ratio: 0.20 0.01 vs. 2.5 0.2, p < 0.01). En face Sudan IV staining of the aorta and oil red O-staining of the aortic sinus revealed that EPA signicantly suppressed the development of atherosclerotic lesions. We also observed anti-atherosclerotic effects of EPA in LDL-receptor-decient mice. The lesions of the EPA group contained more collagen (19.6 2.4% vs. 32.9 3.9%, p < 0.05) and smooth muscle cells (1.3 0.2% vs. 3.6 0.8%, p < 0.05) and less macrophages (32.7 4.1% vs. 14.7 2.0%, p < 0.05). Pretreatment with EPA attenuated the up-regulation of VCAM-1, ICAM-1 and MCP-1 in HUVECs as well as the expression of MMP-2 and MMP-9 in macrophage-like cells induced by TNF-. The anti-inammatory effects of EPA were abrogated when the expression of peroxisome proliferator-activated receptor alpha (PPAR) was suppressed. EPA may potentially reduce and stabilize atherosclerotic lesions through its anti-inammatory effects. 2007 Elsevier Ireland Ltd. All rights reserved.
Keywords: Inammation; Pathology; Atherosclerosis; Leukocytes; Prostaglandins

1. Introduction Previous epidemiologic studies have demonstrated that higher consumption of sh is associated with reduced risk of cardiovascular events [1]. N-3 polyunsaturated fatty acids (n3 PUFAs) are major active ingredients of sh that contribute to prevention of cardiovascular events. The blood levels of n-3 fatty acids are strongly associated with a reduced risk of sudden death among men without evidence of prior cardiovascular disease [2]. A strong inverse association exists between dietary intake of n-3 fatty acids and risk of definite myocardial infarction and nonfatal coronary events [1]. Moreover, treatment with n-3 PUFAs signicantly low-

Abbreviations: n-3 PUFAs, n-3 polyunsaturated fatty acids; AA, arachidonic acid; EPA, eicosapentaenoic acid; DPA, docosapentaenoic acid; DHA, docosahexaenoic acid; HUVECs, human umbilical vein endothelial cells; MMP, matrix metalloproteinase; siRNA, small interference RNA; PPAR, peroxisome proliferator-activated receptor; VCAM, vascular cellular adhesion molecule; ICAM, intercellular adhesion molecule Corresponding author at: Department of Cardiovascular Medicine, University of Tokyo Graduate School of Medicine, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan. Tel.: +81 3 3815 5411; fax: +81 3 3814 0021. E-mail address: msata-circ@umin.net (M. Sata). 0021-9150/$ see front matter 2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.atherosclerosis.2007.07.023

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

525

ered the risk of death, non-fatal myocardial infarction and stroke, by 10% in patients who had had a myocardial infarction, indicating that dietary supplementation with n-3 PUFAs provided a clinically important and statistically signicant benet regarding prevention of cardiovascular mortality [3]. A recent study showed that the long-term use of highly puried eicosapentaenoic acid (EPA), in addition to HMGCoA reductase inhibitor, prevented cardiovascular events in Japanese patients with hypercholesterolemia [4]. To provide molecular basis of the epidemiological evidence, several animal studies demonstrated antiatherosclerotic effects of n-3 PUFAs [5,6]. Wang et al. demonstrated that sh oil reduces the formation of atherosclerotic lesions in ApoE/ mice by reducing oxidative stress [6]. Renier et al. reported that dietary n-3 PUFAs inhibit the macrophage ability to secrete several effecter molecules that may be involved in the pathogenesis of atherosclerosis [5]. However, the molecular mechanism by which n-3 PUFAs reduce atherosclerosis is not fully understood. Here, we investigated the effects of highly puried eicosapentaenoic acid, a major n-3 PUFA, on the quantity and quality of atherosclerotic lesions in a mouse model of hyperlipidemia. The results suggest that EPA reduces and stabilizes atherosclerotic lesions through its anti-inammatory actions.

BPX70 capillary column (0.25 mm internal diameter 30 m, SGE International Ltd., Melbourne, Australia). Tricosanoic acid, C23:0 was used as the internal standard. 2.3. Quantication of atherosclerotic lesions Atherosclerotic lesion severity was assessed in the aortas as previously described [11]. In brief, at euthanasia, animals were perfused with PBS. The entire mouse aorta was dissected from the proximal ascending aorta to the bifurcation of the iliac artery by using a dissecting microscope. Adventitial fat was removed, and the aorta was opened longitudinally, pinned at onto a black rubber board, stained with Sudan IV [11], and photographed using a CCD camera (VB6010, Keyence, Osaka). The photographs were digitized, and total aortic areas and lesion areas were calculated by using Adobe Photoshop Version 7.0. The results were reported as a percentage of the thoracic aortic area that contained lesions. As a second assessment of atherosclerosis, lesions of the aortic root (aortic sinus) were analyzed [12]. The crosssectional lesions of the aortic root were analyzed according to the modied method of Paigen et al. [13]. Each heart was cut in the plane between the lower tips of the right and left atria. The upper portion was snap-frozen in OCT compound (TissueTek, Tokyo). Then, the aortic root was sectioned serially (5 m intervals) from the appearance of the aortic valve to the ascending aorta until the valve cusps were no longer visible. Frozen sections were used for characterization of atherosclerotic lesions in the aortic root. Atherosclerotic area in the aortic sinus was measured from the images digitized by a CCD camera (Digital Sight DS-2Mv, Nikon) after staining by oil red O and hematoxylin. 2.4. RNA isolation from atherosclerotic aota Total RNA was isolated from the aorta of the ApoE/ mice fed on western type diet supplemented with EPA (EPA group, n = 4) or not (Control group, n = 4) for 12 weeks starting at 9 weeks old. In brief, at euthanasia, animals were perfused with PBS. The entire mouse aorta was dissected from the proximal ascending aorta to the bifurcation of the iliac artery by using a dissecting microscope. Adventitial fat was removed, and total RNA was prepared with the use of RNazol reagent (TEL-TEST, Inc., Friendswood, TX). 2.5. Histological characterization of atherosclerotic lesions in the aortic root Immunohistochemistry was performed on sections approximately 50 m above the beginning of the aortic sinuses [14]. Frozen sections were incubated with anti-F4/80 monoclonal antibody (Serotec, Oxford, UK) followed by the avidin-biotin complex technique and Vector Red substrate (Vector, Burlingame, CA). Smooth muscle cells were identied by immunostaining with an alkaline-phosphatase-

2. Materials and methods 2.1. Animals ApoE-decient mice (ApoE/ mice, C57BL/6 background) [7] and LDL-receptor-decient mice (LDL-R/ mice, C57BL/6 background) [8] were purchased from The Jackson Laboratory (Bar Harbor, Me). C57BL/6J mice were purchased from SLC Japan (Shizuoka, Japan). Ultra-pure eicosapentaenoic acid ethylester (EPA; >99% purity) was provided by Mochida Pharmaceutical Co., Ltd. (Tokyo). Male ApoE/ mice or LDL-R/ mice were fed on western-type diet (0.15% cholesterol, 15% butter) supplemented with 5% EPA (w/w) (EPA group, 5.35 kcal/g) or not (Control group, 5.00 kcal/g) for the periods indicated. All mice were kept in microisolator cages on a 12-h day/12-h night cycle. All experimental procedures and protocols were approved by the Animal Care and Use Committee of the University of Tokyo and complied with the Guide for the Care and Use of Laboratory Animals [9]. 2.2. Analysis of fatty acids composition The fatty acids composition of serum and homogenized liver tissue (100 mg of tissue/ml of saline) was determined by capillary gas chromatography [10]. Total lipids were extracted by Folchs procedure and then fatty acids were methylated with boron triuoride and methanol; methylated fatty acids were analyzed using a Shimadzu GC-17A gas chromatograph (Shimadzu Corporation, Kyoto, Japan) and a

526

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

conjugated monoclonal antibody against -smooth muscle actin (-SMA, clone 1A4, Sigma). Sections were counterstained with hematoxylin. 2.6. Sirius red polarization method for collagen staining Sirius red polarization microscopy was performed to visualize interstitial collagen as previously described [15]. Fresh-frozen sections (5 m) were rinsed with distilled water and incubated with 0.1% Sirius red (SigmaAldrich, St. Louis, MO) in saturated picric acid for 90 min. Sections were rinsed twice with 0.03N HCl for 1 min each time and then immersed in distilled water. After dehydration with 70% ethanol for 30 s, the sections were coverslipped. The stained sections were examined under a polarization microscope (Eclipse LV100POL, Nikon, Tokyo). Images were digitized by a CCD camera (Digital Sight DS-2Mv, Nikon). As the bers thickness increases, the color changes from green to red. 2.7. Thioglycollate-induced peritonitis The effect of EPA on macrophages inltration in vivo was evaluated in a thioglycollate-induced peritonitis model [16]. Five-week-old ApoE/ mice were fed on western-type diet supplemented with 5% EPA (EPA group, n = 6) or not (control group, n = 5) for 4 weeks. Each mouse was injected intraperitoneally with 0.5 ml of 4% thioglycollate medium (SigmaAldrich). After 4 days, cells in the peritoneal exudate were harvested by washing the peritoneal cavity with 4 ml of PBS. Cells in the lavage were stained with an anti-F4/80 antibody, followed by Cy 5-conjugated anti-rat Ig antibody. F4/80-positive cells were counted by owcytometry (EPICS XL, Beckman Coulter, Fullerton, CA). 2.8. Cell cultures

1 M of sodium AA, sodium EPA, or sodium bezabrate (Kissei Pharmaceutical Co., Ltd., Matsumoto, Japan) for 48 h. THP-1 cells were stimulated with TNF- (10 ng/ml) for 24 h. Total RNA was isolated with RNazol reagent as already described [17]. 2.9. Reverse transcription-polymerase chain reaction Reverse transcription was performed with 1 g of total RNA, random hexamer primers and MMLV reverse transcriptase (ReverTraAce-, TOYOBO, Osaka) [17]. The polymerase chain reaction (PCR) primers were as follows: VCAM-1, 5 -GATACAACCGTCTTGGTCAGCCC-3 and 5 -CGCATCCTTCAACTGGCCTT-3 ; ICAM-1, 5 -CAGTGACCATCTACAGCTTTCCGG-3 and 5 -GCTGCTACCACAGTGATGATGACAA-3 ; human MMP-2, 5 -CCTGATGGCACCCATTTACAC-3 and 5 -GAGCTCCTGAATGCCCTTGA-3 ; human MMP-9, 5 -GCTCACCTTCACTCGCGTGTA-3 and 5 -TCCGTGCTCCGCGACA-3 ; murine MMP-2, 5 -CCCCATGAAGCCTTGTTTACC-3 and 5 -TTGTAGGAGGTGCCCTGGAA-3 ; murine MMP-9, 5 -AGACCAAGGGTACAGCCTGTTC-3 and 5 -GGCACGCTGGAATGATCTAAG-3 ; human ABCA-1, 5 -GGAGGCAATGGCACTGAGGAA-3 and 5 -CCTGCCTTGTGGCTGGAGTGT-3; murine ABCA-1, 5 -CCTCAGCCATGACCTGCCTTGTAG-3 and 5 -CCGAGGAAGACGTGGACACCTTC3; and glyceraldehyde 3-phosphate dehydrogenase (GAPDH), 5 -GGGAAGCCCATCACCATCTTC-3 and 5 -GTTCTGGGCAGCCCCACGGCC-3 . Specic mRNAs were quantied by SYBR Green real-time PCR Master Mix (TOYOBO) in an ABI PRISM 7000 thermocycler (Applied Biosystems, Foster City, CA) under standard manufacturers conditions. Data are expressed in arbitrary units that were normalized by correction for the signal obtained in the same cDNA preparation for GAPDH mRNA. 2.10. ELISA

Human umbilical vein endothelial cells (HUVECs) were purchased from Cambrex Bio Science Walkersville Inc. (Walkersville, MD). HUVECs were cultured in EGM2 medium (Clonetics, Walkersville) containing either arachidonic acid (1 M, AA, Na salt, Sigma), or EPA [1 M, cis-5,8,11,14,17-eicosapentaenoic acid (EPA, Na salt, Sigma)] for 48 h. HUVECs were stimulated with TNF- (10 ng/ml, R&D Systems Inc., Minneapolis, MN) for 24 h. Total RNA was obtained with RNazol reagent (Tel-Test, Friendswood, TX) as described elsewhere [17]. THP-1 cells (human monocytic cell line) and RAW264.7 cells (murine macrophage cell line) were purchased from the American Type Culture Collection (Manassas, VA). THP-1 cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37 C in a humidied atmosphere of 5% CO2 . Differentiation of THP1 cells (5 105 to 106 per ml) into macrophage-like cells was induced by treatment with 200 nM 12-O-tetradecanoylphorbol-13-acetate (TPA, Sigma) in the presence of either

HUVECs were pre-incubated in EGM-2 medium containing either sodium AA (1 M) or sodium EPA (1 M) for 48 h, followed by stimulation with TNF (10 ng/ml) for 24 h. MCP-1 in the medium was measured by enzyme-linked immunosorbent assay (ELISA) according to the manufacturers instructions (Amersham Biosciences, Tokyo). 2.11. Transfection of siRNA A duplex 21-nucleotide small interfering RNA (siRNA) against human peroxisome proliferator-activated receptor alpha (PPAR) and human peroxisome proliferator-activated receptor gamma (PPAR) were designed by Ambion (Austin, TX, No. 5348 and 5731, respectively) according to the current guidelines for effective knockdown by this method. The two double-stranded PPAR-targeting siRNAs were 5 -GGAUAGUUCUGGAAGCUUUTT-3 (sense) and 5 -AAAGCUUCCAGAACUAUCCTC-3 (antisense).

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

527

The two double-stranded PPAR-targeting siRNAs were 5 -GGGCGAUCUUGACAGGAAATT-3 (sense) and 5 UUUCCUGUCAAGAUCGCCTC-3 (antisense). A nontargeting scrambled RNA duplex siRNA containing 21 nucleotides with the same GC content sequences was used as a control (Ambion). Differentiation of THP-1 cells was induced by treatment with TPA (200 nM) for 48 h. Approximately 3 105 cells/well were grown in a six-well microplate. Differentiated THP-1 cells were transfected with siRNA using a commercially available kit (EXPRESS-si Delivery Kit, Genospectra, Fremont, CA) according to the manufacturers instructions. Briey, siRNA was mixed with the transfection reagent (200 l/well) and added dropwise onto the cells. Cells were incubated with siRNA (20 nM) for 4 h and then complete growth media was added. After 48 h, gene silencing was monitored by RT-PCR. Cells were incubated with 1 M of either sodium AA or sodium EPA for 2 days. Then, cells were stimulated with TNF (10 ng/ml) for 24 h. Total RNA and nuclear extracts were collected. 2.12. Trans-AM NF-B transcriptional factor assay NF-B activation was evaluated using a commercially available kit (Trans-AM NF-B Kit, Active Motif, Inc., Carlsbad, CA). Differentiated THP-1 cells were pretreated with sodium AA or sodium EPA, followed by stimulation with TNF (10 ng/ml) for 24 h. Nuclear extracts were isolated. The Trans-AM NF-B kit contains a 96-well plate on which oligonucleotide containing the NF-B consensus site (5 GGGACTTTCC-3 ) has been immobilized. The active form of NF-B specically binds to this oligonucleotide. The primary antibody used to detect NF-B recognizes an epitope on p50 that is accessible only when NF-B is activated and bound to its target DNA. A secondary HRP-conjugated antibody provides a sensitive colorimetric readout easily quantied by spectrophotometry. 2.13. Statistical analysis All measurements were conducted by two observers blinded to the treatment of the mice. All values are presented as the mean S.E.M. Means were compared by unpaired Students t test. A p-value of <0.05 was considered statistically signicant.

on HDL level (14.3 1.7 mg/dl vs. 15.0 7.0 mg/dl, NS). The plasma triglyceride level was elevated in the EPA group at 13 weeks after the treatment (41 18 mg/dl vs. 312 37 mg/dl, p < 0.01). An analysis of the fatty acids composition of sera revealed that EPA supplementation signicantly increased EPA concentration (13.7 0.5 g/ml vs. 3542.5 474.0 g/ml, p < 0.01) (0.3 0.1 mol% vs. 42.6 0.3mol%, p < 0.01) and decreased AA concentration (265.4 13.2 g/ml vs. 51.7 4.7 g/ml, p < 0.01) (5.2 0.1 mol% vs. 0.6 0.1mol%, p < 0.01) in the peripheral blood (on line supplementary Table 2). These changes were associated with alternation of fatty acids composition of the liver homogenate. EPA markedly increased the proportion of EPA (0.3 0.1 mol% vs. 18.8 0.5 mol%, p < 0.01) and decreased the content of AA (10.6 1.3 mol% vs. 1.7 0.1 mol%, p < 0.01) in the liver. The content of docosapentaenoic acid (DPA) (0.5 0.1 mol% vs. 5.8 0.2 mol%, p < 0.01) and docosahexaenoic acid (DHA) (5.4 0.5 mol% vs. 7.6 0.3 mol%, p < 0.05) also signicantly increased by EPA. When EPA was administered to 5-week-old LDL-R/ mice, EPA markedly decreased serum level of total cholesterol (658 98 mg/dl vs. 273 62 mg/dl, p < 0.01) and triglyceride (69 8 mg/dl vs. 24 4 mg/dl, p < 0.01). 3.2. EPA reduces atherosclerotic lesions and changes plaque composition EPA signicantly suppressed the development of atherosclerotic lesions (18.7 3.1% vs. 5.1 1.2, p < 0.01) (Fig. 1A). The effect of EPA on atherosclerosis development was also assessed in cross-sectional lesions of the aortic root. Heart aortic sinus atherosclerosis was measured in the ApoE/ mice treated with EPA (EPA group, n = 6) or not (Control group, n = 8) for 12 weeks starting at 9 weeks old. EPA signicantly suppressed the development of atherosclerotic lesions in aortic sinus (429 31 103 m2 vs. 311 25 103 m2 , p < 0.01) (Fig. 1B). The anti-atherosclerotic effect of EPA was also tested in LDL-receptor/ mice. Male LDL-receptor/ mice were treated with EPA (EPA group, n = 7) or not (Control group, n = 6) for 12 weeks starting at 4 weeks old. EPA signicantly suppressed the development of atherosclerotic lesions (8.9 3.0% vs. 0.6 0.3%, p < 0.05) as determined by en face Sudan IV staining of the aorta (Fig. 1C). Cross-sectional analysis of the aortic root was performed to characterize the constituents of atherosclerotic plaques. The content of interstitial collagen in atheromas was increased by EPA treatment (19.6 2.4% vs. 32.9 3.9%, p < 0.05), as determined by picrosirius red staining analyzed by polarization (Fig. 2A). Immunohistochemistry against -smooth muscle actin (-SMA) revealed that EPA induced a signicant increase of smooth muscle cells in atherosclerotic plaques (1.3 0.2% vs. 3.6 0.8%, p < 0.05) (Fig. 2B). Macrophages inltration of atherosclerotic lesions

3. Results 3.1. Effect of EPA on serum lipid level and fatty acids composition of cell membrane Orally administered EPA tended to decrease total cholesterol levels without statistical signicance at 13 weeks after the treatment (955 110 mg/dl vs. 714 117 mg/dl, NS) (on line supplementary Table 1). EPA had no effect

528

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

Fig. 1. Measurement of atherosclerosis in hypercholesterolemic mice in the EPA and control groups. (A) Five-week-old male ApoE/ mice were fed on western-type diet (0.15% cholesterol, 15% butter) without supplemented with 5% EPA (8 g/kg/day; EPA group, n = 7) or not (control group, n = 5) for 13 weeks. Atheromatous plaques in the aortas were visualized by en face Sudan IV staining. The extent of Sudan IV-positive areas was measured and expressed as percentage. * p < 0.01. (B) Nine-week-old ApoE/ were fed on western-type diet supplemented with 5% EPA (EPA group, n = 6) or not (control group, n = 8) for 12 weeks. Heart aortic sinus atherosclerosis was measured after staining by oil red O and hematoxylin. * p < 0.05. (C) Four-week-old LDL-receptor-decient mice were fed on western-type diet supplemented with 5% EPA (EPA group, n = 6) or not (control group, n = 8) for 12 weeks. Atheromatous plaques in the aortas were visualized by en face Sudan IV staining. The extent of Sudan IV-positive areas was measured and expressed as percentage. * p < 0.01.

was markedly attenuated by EPA as determined by anti-F4/80 immunostaining (32.7 4.1% vs. 14.7 2.0%, p < 0.05) (Fig. 2C). Oil red O staining revealed that EPA tended to prevent lipid deposition (20.4 2.2% vs. 10.9 5.3%, p < 0.05). Taken together, these results suggest that EPA substantially stabilized atherosclerotic lesions. Oil red O positive area was also signicantly decreased in EPA group (n = 3) compared with the control group (n = 3) (27.6 1.9% vs. 23.1 1.6%, p < 0.05) in ApoE/ mice treated for 13 weeks starting at 5 weeks old. Oil red O positive area was also signicantly decreased in EPA group (n = 6) compared with the control group (n = 8) (27.6 3.0% vs. 16.9 12.8%, p < 0.05) in ApoE/ mice treated for 12 weeks starting at 9 weeks old. 3.3. EPA inhibits inltration of inammatory cells To investigate the molecular mechanism by which EPA effectively prevented and stabilized atherosclerotic lesions without improving the serum lipid prole, we investigated the effect of EPA on inammatory responses. First, the expression of adhesion molecules by endothelial cells was evaluated. HUVECs were pre-incubated with either 1 M sodium AA or 1 M sodium EPA for 48 h and then stimulated with 10 ng/ml TNF for 24 h. RT-PCR revealed that TNF markedly up-regulated the expression of VCAM-1 and ICAM-1 mRNA in HUVECs. Pre-treatment of HUVECs

with EPA signicantly attenuated the expression of VCAM1 (21.7 0.7 vs. 13.8 2.0 arbitrary units, p < 0.05) and ICAM-1 (8.6 0.6 vs. 3.2 0.1 arbitrary units, p < 0.01) as determined by real-time quantitative PCR (Fig. 3A and B). MCP-1 secretion from HUVECs was measured by ELISA. Pretreatment of HUVECs with EPA blunted the up-regulation of MCP-1 secretion after stimulation with TNF (2.5 0.2 ng/ml vs. 2.0 0.1 ng/ml, p < 0.05) (Fig. 3C). Furthermore, real time RT-PCR analysis of the aortic homogenate revealed that EPA down-regulated MCP1 (7.2 4.7% vs. 3.7 2.8%, NS), VCAM-1 (6.6 6.2 vs. 2.5 1.1 arbitrary unit, NS) and ICAM-1 expression (2.0 1.1 vs. 0.8 0.4 arbitrary unit, NS), although the difference did not reach statistical difference. These results suggest that EPA inhibited the interaction of inammatory cells with endothelial cells. Consistent with this notion, orally administered EPA signicantly decreased the number of macrophages inltrating the peritoneal cavity in ApoEdecient mice upon injection of thioglycollate (7.1 1.5 vs. 3.7 0.7 105 /mouse, p < 0.05) (Fig. 3D). 3.4. Effects of EPA on MMP expression by macrophages TNF up-regulated the expression of MMP-2 and MMP9 in RAW264.7 cells (a murine macrophage cell line). Pre-treatment of RAW264.7 cells with EPA signicantly attenuated the expression of MMP-2 (0.3 0.1 vs. 0.1 0.1

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

529

Fig. 2. Histological characterization of atherosclerotic lesions in the aortic root. Histological analyses of the atherosclerotic plaques were performed on cross sections of the aorta obtained from approximately 50 m above the beginning of aortic sinuses. (A) Sirius red polarization method for collagen staining. Frozen sections (5 m) were stained with 0.1% Sirius red in saturated picric acid. After dehydration in 70% ethanol for 30 s, the sections were coverslipped and observed under a polarization microscope. Images were digitized without (Top) or with (Bottom) polarization. The Sirius red positive area in atheromas was quantied in each section. A, atheroma; M, media. Scale Bar, 100 m. * p < 0.05. (B) Smooth muscle cells were identied by immunostaining with an alkaline-phosphatase-conjugated monoclonal antibody against -smooth muscle actin (-SMA, clone 1A4, Sigma) and Vector Red substrate. The sections were counterstained with haematoxylin. The -SMA-positive area in atheromas was quantied in each section. A, atheroma; M, media. Scale bar, 50 m. * p < 0.05. (C) Macrophages were identied with anti-F4/80 monoclonal antibody followed by the avidinbiotin complex technique and Vector Red substrate. The F4/80-positive area in atheromas was quantied in each section. A, atheroma; M, media. Scale bar, 50 m. * p < 0.05.

arbitrary units, p < 0.05, n = 3 for each group) and MMP-9 (49.5 11.0 vs. 13.7 4.4 arbitrary units, p < 0.05, n = 3 for each group) as determined by real time PCR (Fig. 4A). These results indicate that pre-treatment with EPA attenuates MMP expression by macrophages in response to cytokine stimulation. Consistent with the in vitro experiment, down regulation of MMP-2 (1.6 0.2 vs. 0.6 0.2 arbitrary unit, p < 0.05, n = 4 for each group) and MMP-9 (4.5 0.6 vs. 2.9 0.8 arbitrary unit, NS, n = 4 for each group) was documented in the aorta in the ApoE treated with EPA for 4 weeks starting at 5 weeks old (Fig. 4B).

3.5. PPAR-dependent inhibition of NF-B activation by EPA Treatment of differentiated THP-1 cells with EPA upregulated the expression of PPAR in a dose-dependent manner (Fig. 5A). Similarly, in the atherosclerotic aorta of ApoE-decient mice, EPA signicantly increased the expression of PPAR and ABCA-1, which are PPAR target genes (1.91 0.26 vs. 3.30 0.18, arbitrary unit, p < 0.01, 0.82 0.08 vs. 1.49 0.22, arbitrary unit, p < 0.05, respectively). The expression of ABCA-1 in differentiated

530

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

Fig. 3. Inhibitory effects of EPA on endothelial cells and macrophages interaction. (A and B) Human umbilical vein endothelial cells (HUVECs) were treated with either arachidonic acid (AA, 1 M) or EPA (1 M) for 48 h. After stimulation with TNF- (10 ng/ml) for 24 h, total RNA was isolated. Expression of VCAM-1 (A) and ICAM-1 (B) was quantied by real-time PCR. (*) and (**) indicate p < 0.05 and p < 0.01, respectively. (C) HUVECs were pre-incubated with either sodium arachidonic acid (AA, 1 M) or sodium EPA (1 M) for 48 h. After stimulation with TNF- (10 ng/ml) for 24 h, the concentration of MCP-1 in the medium was measured by ELISA. (D) Five-week-old ApoE/ mice were fed on western-type diet upplemented with 5% (w/w) EPA (EPA group, n = 6) or not (control group, n = 6) for 4 weeks. Each mouse was injected intraperitoneally with 0.5 ml of 4% thioglycollate medium. After 4 days, cells in the peritoneal exudate were harvested by washing the peritoneal cavity with 4 ml of PBS. Cells in the lavage were stained with an anti-F4/80 antibody, followed by Cy 5-conjugated anti-rat Ig antibody. F4/80-positive cells were counted by owcytometry. * p < 0.05.

THP-1 cells also signicantly increased in the EPA group (0.11 0.006 vs. 0.163 0.01, arbitrary unit, p < 0.05). To investigate the potential involvement of PPAR in the antiinammatory effect of EPA, the expression of PPAR was suppressed using small interference RNA (siRNA). Real time-PCR revealed that the PPAR level was effectively down-regulated by siRNA against PPAR (77.2 2.24%), whereas transfection of scramble siRNA did not affect the expression of PPAR (Fig. 5B). When HUVECs were transfected with scramble siRNA, pre-treatment with EPA signicantly prevented NFB activation upon TNF stimulation (Fig. 5C). On the other hand, when PPAR expression was inhibited by siRNA, EPA did not inhibit NFB activation (Fig. 5C). Similarly, when PPAR expression was inhibited by siRNA in differentiated THP-1 cells, pre-treatment with EPA failed to suppress MMP expression upon TNF

stimulation (Fig. 5D). Similarly, when PPAR expression was inhibited by anti-PPAR siRNA in differentiated THP-1 cells, pre-treatment with EPA also failed to suppress MMP expression upon TNF stimulation (Fig. 5E).

4. Discussion In this study, we investigated the effect of EPA on the progression and instability of atherosclerotic lesions in ApoE / and LDL-receptor/ mice. Orally administered EPA markedly increased the cellular content of n-3 PUFA without a signicant effect on serum levels of total cholesterol and HDL in ApoE/ mice. EPA signicantly suppressed the development of atherosclerotic lesions in ApoE/ and LDL-receptor/ mice. The atherosclerotic plaques of EPA-

Fig. 4. Inhibitory effect of EPA on MMP expression in vitro and in vivo. (A) RAW264.7 cells were pre-incubated with arachidonic acid (AA, 3 M) or EPA (3 M) for 48 h. Then, RAW264.7 cells were stimulated with TNF- (10 ng/ml) for 24 h. Total RNA was isolated. Expression of MMP-2 and MMP-9 was quantied by quantitative real-time PCR. N = 3 for each group. * p < 0.05. (B) Total RNA was isolated from the aorta of the ApoE/ mice fed on western type diet supplemented with EPA (EPA group, n = 4) or not (Control group, n = 4) for 12 weeks starting at 9 weeks old. Expression of MMP-2 and MMP-9 was quantied by quantitative real-time PCR. * p < 0.05.

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

531

Fig. 5. PPAR-dependent inhibition of NFB activation by EPA. (A) Up-regulation of PPAR mRNA by EPA. THP-1 cells were stimulated to differentiate into macrophage-like cells with 200 nM TPA for 48 h in the presence of either sodium EPA (0, 100 or 300 M) or sodium bezabrate (30 or 100 M) for 48 h. Total RNA was harvested. PPAR expression was quantied by real-time PCR. Beza, sodium bezabrate. * p < 0.05 vs. control (0 M EPA). ** p < 0.01 vs. control. (B) Suppression of PPAR expression by siRNA against PPAR. A duplex 21-nucleotide small interfering RNA (siRNA) or a non-targeting scrambled RNA against human PPAR was transfected to differentiated THP-1 cells for 48 h. Then, cells were incubated with arachidonic acid (AA) or EPA for 48 h. Gene silencing was monitored at the mRNA levels by real time PCR. (C) Differentiated THP-1 cells were transfected with siRNA against PPAR (20 nM) or a scrambled RNA. Cells were incubated with 10 M of either sodium AA or sodium EPA for next 2 days. Then, cells were stimulated with TNF (10 ng/ml) for 24 h. Total RNA and nuclear extracts were collected. NF-B p50 activity was measured using a commercially available kit (Trans-AM NF-kB p50 kit). * p < 0.05 vs. AA. (D and E) No effect of EPA on MMP expression in the differentiated THP-1 cells transfected with PPAR (D) or PPAR (E)-targeting siRNAs. Differentiated THP-1 cells were transfected with PPAR or PPAR-targeting siRNA. After 48 h, the cells were incubated with 1 M of either sodium AA or sodium EPA for 48 h. Then, the cells were stimulated with TNF (10 ng/ml) for 24 h. Total RNA was isolated. Expression of MMP-2 and MMP-9 was evaluated by quantitative real time PCR.

treated mice displayed a stabilized morphology that was characterized by less lipid deposition, decreased accumulation of macrophages, more smooth muscle cells, and increased collagen content. EPA suppressed the expression of adhesion molecules and MCP-1 by endothelial cells and the production of MMPs by macrophages in a PPAR-dependent manner. Although accumulating evidence suggests that a higher consumption of sh or n-3 PUFAs is associated with a reduced risk of cardiovascular events [1], relatively little is known about the effects of EPA on the pathogenesis of atherosclerosis. Previous studies reported that n-3 PUFAs have various benecial effects on eicosanoid metabolism,

inammation, beta oxidation, and endothelial dysfunction [18]. However, few studies have directly demonstrated that EPA has benecial effects on atherosclerotic lesions in vivo [19]. Our ndings provide evidence that ultra-pure EPA can effectively inhibit atherosclerosis progression in a mouse model of atherosclerosis. It is a widely accepted view that acute coronary syndrome results from the rupture of vulnerable plaques in the majority of cases [20]. Vulnerable plaques are characterized by thinning of the brous cap and a decrease in the content of collagen as well as in the number of smooth muscle cells [20]. Disrupted brous caps are usually heavily inltrated by

532

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533

macrophages, which are associated with high expression of MMPs that potentially weaken the brous cap, predisposing it to rupture [21]. Our results demonstrated that EPA signicantly prevented macrophage inltration, and this was associated with an increase in collagen content and of the number of smooth muscle cells. Pretreatment with EPA signicantly attenuated the expression of adhesion molecules and MCP-1 by endothelial cells stimulated with TNF. In a thioglycollate-induced peritonitis model, macrophages extravasation was signicantly attenuated when mice were treated with EPA. Moreover, EPA effectively inhibited MMP expression by macrophage-like cells. Thus, it is likely that EPA attenuates inltration and activation of inammatory cells in atherosclerotic lesions, preventing thereby plaque destabilization. Consistently, it was reported that oral intake of sh oil supplements increased the content of n-3 PUFAs in atherosclerotic plaques, inducing histological changes that enhanced the stability of atherosclerotic plaques with reduction in macrophage inltration in patients awaiting carotid endarterectomy [22]. There are several biological actions of EPA that may mediate its anti-inammatory effects [19,23]. PPAR is known to regulate inammatory responses by interacting with the inammatory signaling pathways at several levels [24]. It was reported that EPA up-regulated PPAR expression and/or activity in endothelial cells [25] and macrophages [26]. Mishra et al. reported that EPA inhibited NF-B activation in endothelial cells derived from wild-type mice, but not in those from PPAR-decient mice [27]. Consistently, EPA potently inhibited leukocyte-endothelial cells interaction in wild-type mice, but not in PPAR-decient mice [28]. We found that pretreatment with EPA attenuated TNF-induced production of MMPs by macrophage-like cells. When PPAR activity was suppressed by siRNA, EPA failed to inhibit NF-B activation and induction of MMPs. These results suggest that PPAR-dependent inhibition of NFB might be involved, at least in part, in the anti-inammatory effects of EPA. Consistent with this notion, others reported that EPA modulates PPAR activity by regulating its expression and serving as a ligand for PPAR [2931]. In addition, when PPAR activity was suppressed by siRNA, EPA also failed to inhibit induction of MMPs. Others reported that sh oil could be ligands for various other nuclear receptors including PPAR [32,33]. It is plausible that multiple nuclear receptors potentially mediate the anti-atherogenic actions of EPA. EPA and DHA are both components of sh and sh oil. N-3 PUFAs can be supplemented as sh oil, puried EPA, DHA, or a combination of EPA and DHA. It still remains to be determined which of these fatty acids is most effective in preventing cardiovascular diseases. De Caterina et al. reported that DHA, but not EPA, decreased in a dose- and timedependent fashion the expression of VCAM-1 and ICAM-1. DHA treatment also reduced the adhesion of human monocytes and of monocytic U937 cells to cytokine-stimulated endothelial cells, suggesting that these properties of DHA may contribute to antiatherogenic and anti-inammatory

effects of N-3 PUFAs [34]. They also reported that compared with DHA, EPA was a weaker inhibitor of the expression of these molecules and monocyte adhesion [23]. Here, we observed that EPA markedly suppressed the development and instability of atherosclerotic lesions in ApoE-decient mice and LDL-receptor-decient mice. In contrast, other authors reported that sh oil or puried DHA had no signicant effect on lesion size in ApoE-decient mice [35]. Our results, however, suggest that administration of n-3 PUFA as ultrapure EPA is sufcient to prevent atherosclerosis, at least in ApoE-decient mice and LDL-receptor-decient mice. Usually, n-3 PUFAs intake lowers the serum level of triglycerides in humans and animals [36]. We also conrmed that EPA effectively decreased the level of triglycerides in wild-type mice (data not shown) and in LDL-receptordecient mice. In contrast, EPA markedly elevated the level of serum triglycerides in ApoE-decient mice. Other authors have also observed that sh oil has a triglyceride-raising effect in apoE-decient [37] or ApoE*3-Leiden transgenic mice [38]. Functional ApoE appears to be essential for n3 PUFAs to lower the plasma concentration of triglycerides in mice with a C57BL/6 background, although the genetic background may inuence the effect of n-3 PUFAs on lipid metabolism in ApoE-decient mice [39]. It was reported that lipoprotein clearance was impaired due to the absence of the functional systems of triglyceride-rich lipoproteins anchoring to endothelial cells and to the lack of remnant uptake by the liver, even though sh oil reduced the rates of triglycerides synthesis [37]. In summary, our results demonstrate that ultra-pure EPA prevents the progression and destabilization of atherosclerotic lesions. Anti-inammatory effects other than triglyceride-lowering effects likely account for the atheroprotective effects of EPA. Our ndings may provide the basis for the use of ultra-pure EPA as a prophylactic treatment of patients with atherosclerosis.

Acknowledgements This study was supported in part by grants from the Ministry of Education, Culture, Sports, Science and Technology and the Ministry of Health, Labor and Welfare of Japan.

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.atherosclerosis. 2007.07.023.

References
[1] Iso H, Kobayashi M, Ishihara J, et al. Intake of sh and n3 fatty acids and risk of coronary heart disease among Japanese: the Japan Public Health

M. Matsumoto et al. / Atherosclerosis 197 (2008) 524533 Center-Based (JPHC) Study Cohort I. Circulation 2006;113:195 202. Albert CM, Campos H, Stampfer MJ, et al. Blood levels of longchain n-3 fatty acids and the risk of sudden death. N Engl J Med 2002;346:11138. GISSI-PrevenzioneTrial. Dietary supplementation with n-3 polyunsaturated fatty acids and vitamin E after myocardial infarction: results of the GISSI-Prevenzione trial. Gruppo Italiano per lo Studio della Sopravvivenza nellInfarto miocardico. Lancet 1999;354:44755. Yokoyama M. Effects of eicosapentaenoic acid (EPA) on major cardiovascular events in hypercholesterolemic patients: the Japan EPA Lipid InterventionStudy (JELIS). Circulation 2005;112:33623. Renier G, Skamene E, DeSanctis J, et al. Dietary n-3 polyunsaturated fatty acids prevent the development of atherosclerotic lesions in mice. Modulation of macrophage secretory activities. Arterioscler Thromb 1993;13:151524. Wang HH, Hung TM, Wei J, et al. Fish oil increases antioxidant enzyme activities in macrophages and reduces atherosclerotic lesions in apoEknockout mice. Cardiovasc Res 2004;61:16976. Zhang SH, Reddick RL, Piedrahita JA, et al. Spontaneous hypercholesterolemia and arterial lesions in mice lacking apolipoprotein E. Science 1992;258:46871. Ishibashi S, Brown MS, Goldstein JL, et al. Hypercholesterolemia in low density lipoprotein receptor knockout mice and its reversal by adenovirus-mediated gene delivery. J Clin Invest 1993;92:88393. The National Institutes of Health. Guide for the Care and Use of Laboratory Animals. NIH Publication 1985; 8623. Ozawa A, Takayanagi K, Fujita T, et al. Determination of long chain fatty acids in human total plasma lipids using gas chromatography. Bunseki Kagaku 1982;31:8791. Yagyu H, Ishibashi S, Chen Z, et al. Overexpressed lipoprotein lipase protects against atherosclerosis in apolipoprotein E knockout mice. J Lipid Res 1999;40:167785. Mullick AE, Tobias PS, Curtiss LK. Modulation of atherosclerosis in mice by Toll-like receptor 2. J Clin Invest 2005;115:314956. Paigen B, Morrow A, Holmes PA, et al. Quantitative assessment of atherosclerotic lesions in mice. Atherosclerosis 1987;68:23140. Tanaka K, Sata M, Fukuda D, et al. Age-associated aortic stenosis in apolipoprotein E-decient mice. J Am Coll Cardiol 2005;46:134 41. Fukumoto Y, Libby P, Rabkin E, et al. Statins alter smooth muscle cell accumulation and collagen content in established atheroma of watanabe heritable hyperlipidemic rabbits. Circulation 2001;103:9939. Wang HW, Babic AM, Mitchell HA, et al. Elevated soluble ICAM-1 levels induce immune deciency and increase adiposity in mice. Faseb J 2005;19:101820. Shoji M, Sata M, Fukuda D, et al. Temporal and spatial characterization of cellular constituents during neointimal hyperplasia after vascular injury: Potential contribution of bone-marrow-derived progenitors to arterial remodeling. Cardiovasc Pathol 2004;13:30612. Siess W, Roth P, Scherer B, et al. Platelet-membrane fatty acids, platelet aggregation, and thromboxane formation during a mackerel diet. Lancet 1980;1:4414. Das UN. Benecial effect(s) of n-3 fatty acids in cardiovascular diseases: but, why and how? Prostaglandins Leukot Essent Fatty Acids 2000;63:35162. Falk E, Shah PK, Fuster V. Coronary plaque disruption. Circulation 1995;92:65771.

533

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9] [10]

[11]

[12] [13] [14]

[15]

[16]

[17]

[18]

[19]

[20]

[21] Libby P. Current concepts of the pathogenesis of the acute coronary syndromes. Circulation 2001;104:36572. [22] Thies F, Garry JM, Yaqoob P, et al. Association of n-3 polyunsaturated fatty acids with stability of atherosclerotic plaques: a randomised controlled trial. Lancet 2003;361:47785. [23] De Caterina R, Madonna R, Massaro M. Effects of omega-3 fatty acids on cytokines and adhesion molecules. Curr Atheroscler Rep 2004;6:48591. [24] Zambon A, Gervois P, Pauletto P, et al. Modulation of hepatic inammatory risk markers of cardiovascular diseases by PPAR-alpha activators: clinical and experimental evidence. Arterioscler Thromb Vasc Biol 2006;26:97786. [25] Inoue I, Shino K, Noji S, et al. Expression of peroxisome proliferatoractivated receptor alpha (PPAR alpha) in primary cultures of human vascular endothelial cells. Biochem Biophys Res Commun 1998;246:3704. [26] Michaud SE, Renier G. Direct regulatory effect of fatty acids on macrophage lipoprotein lipase: potential role of PPARs. Diabetes 2001;50:6606. [27] Mishra A, Chaudhary A, Sethi S. Oxidized omega-3 fatty acids inhibit NF-kappaB activation via a PPARalpha-dependent pathway. Arterioscler Thromb Vasc Biol 2004;24:16217. [28] Sethi S, Ziouzenkova O, Ni H, et al. Oxidized omega-3 fatty acids in sh oil inhibit leukocyte-endothelial interactions through activation of PPAR alpha. Blood 2002;100:13406. [29] Marx N, Duez H, Fruchart JC, et al. Peroxisome proliferator-activated receptors and atherogenesis: regulators of gene expression in vascular cells. Circ Res 2004;94:116878. [30] Xu HE, Lambert MH, Montana VG, et al. Molecular recognition of fatty acids by peroxisome proliferator-activated receptors. Mol Cell 1999;3:397403. [31] Larsen LN, Granlund L, Holmeide AK, et al. Sulfur-substituted and alpha-methylated fatty acids as peroxisome proliferator-activated receptor activators. Lipids 2005;40:4957. [32] Jump DB. The biochemistry of n-3 polyunsaturated fatty acids. J Biol Chem 2002;277:87558. [33] Chambrier C, Bastard JP, Rieusset J, et al. Eicosapentaenoic acid induces mRNA expression of peroxisome proliferator-activated receptor gamma. Obes Res 2002;10:51825. [34] De Caterina R, Cybulsky MI, Clinton SK, et al. The omega-3 fatty acid docosahexaenoate reduces cytokine-induced expression of proatherogenic and proinammatory proteins in human endothelial cells. Arterioscler Thromb 1994;14:182936. [35] Zampolli A, Bysted A, Leth T, et al. Contrasting effect of sh oil supplementation on the development of atherosclerosis in murine models. Atherosclerosis 2006;184:7885. [36] Wijendran V, Hayes KC. Dietary n-6 and n-3 fatty acid balance and cardiovascular health. Annu Rev Nutr 2004;24:597615. [37] Asset G, Bauge E, Fruchart JC, et al. Lack of triglyceride-lowering properties of sh oil in apolipoprotein e-decient mice. Arterioscler Thromb Vasc Biol 2001;21:4016. [38] van Vlijmen BJ, Mensink RP, vant Hof HB, et al. Effects of dietary sh oil on serum lipids and VLDL kinetics in hyperlipidemic apolipoprotein E*3-Leiden transgenic mice. J Lipid Res 1998;39:11818. [39] Ide T, Takahashi Y, Kushiro M, et al. Effect of n-3 fatty acids on serum lipid levels and hepatic fatty acid metabolism in BALB/c.KORApoeshl mice decient in apolipoprotein E expression. J Nutr Biochem 2004;15:16978.