Dietary n-3 LCPUFA from fish oil but not α-linolenic acid-derived LCPUFA confers atheroprotection in mice^[S]

Mice, diets, and study outline

ApoB^100/100LDLr^−/− female mice were created as previously described (15). Mice were of a mixed genetic background (∼75% C57BL/6 and ∼25% 129Sv/Jae). To minimize genetic heterogeneity, sibling controls were used. At 7–8 weeks of age, mice were switched from a diet of regular chow to one of four experimental diets containing 20% calories from fat and 0.02% cholesterol by weight. Two levels of ALA (2% and 4% of energy) from flaxseed oil (Bioriginal Food and Science Corp., Saskatoon, Canada) were used. Isocaloric diets enriched either in cis-monounsaturated fatty acids (mainly oleic acid) (AC HUMKO, Memphis, TN) or fish oil (Omega Protein, Inc., Houston TX) were used as negative and positive controls, respectively. The diet groups consisted of 12 animals, and the diets were administered for 16 weeks. The diets were balanced for all the ingredients; the only variable was the fatty acid composition. See supplementary data for a detailed description of the dietary fatty acid composition (Tables IA and IB) and dietary ingredients (Table II). At baseline and after 2, 4, 8, and 16 weeks of diet, a blood sample was taken by submandibular vein puncture and placed into a tube containing a protease inhibitor cocktail (Sigma) dissolved in 5% EDTA, 5% NaN₃. Plasma was separated after centrifugation at 12,500 g for 15 min at 4°C. After being fed the experimental diets for 16 weeks, mice were fasted for 4 h and then euthanized. At the time of sacrifice, a terminal blood sample was drawn via heart puncture, and the carcass was flushed with saline. Liver samples were collected, weighted, snap-frozen in liquid N₂,and stored in an ultralow freezer at −80°C. The heart with attached aorta (beginning at the aortic sinus and ending at the iliac bifurcation) was removed from the body and stored in 10% neutral buffered formalin. All animals were housed in our American Association for the Accreditation of Laboratory Animal Care (AAALAC)-approved animal facility. All procedures involving mice had prior approval from the Institutional Animal Care and Use Committee of Wake Forest University School of Medicine.

Plasma lipid and lipoprotein measurements

Plasma total cholesterol (TPC) and triglyceride (TG) concentrations were measured using enzymatic assays as previously described (15). Cholesterol concentrations in VLDL, LDL, and HDL were determined after separation of individual plasma samples (n = 12 mice per diet group) by HPLC with a Pharmacia Superose 6 column run at a flow rate of 0.5 ml/min as previously described (15). LDLs were isolated from fresh aliquots of plasma from individual mice and measured for lipid and fatty acid composition. Briefly, lipoproteins were isolated from plasma samples by adjusting the solution density to 1.225 g/ml with KBr. The plasma samples were then centrifuged for 4 h at 100,000 g in a TLA 120.2 rotor in an Optima MAX-E ultracentrifuge (Beckman Instruments). The floated lipoproteins were harvested using a tube slicer. The lipoprotein mixture was injected onto a Superose 6 chromatography column, which was subsequently run at 0.5 ml/min with 0.9% NaCl containing 0.05% EDTA (pH 7.4) and 0.05% NaN₃. Fractions corresponding to LDL were collected, pooled according to elution time, and measured for lipid concentration and fatty acid composition as previously described (15). LDL particle diameter was determined by dynamic light scattering using a zetasizer nano S (Malvern Instruments, United Kingdom).

Liver lipid measurements

Liver lipid concentrations were estimated after performing chloroform-methanol extraction on ∼80 mg of liver tissue as previously described (16). To determine the fatty acid compositions, liver CE, TG, and PL were separated by thin layer chromatography, and the corresponding bands were scraped and methylated. Fatty acids were then quantified by gas-liquid chromatography.

Quantification of atherosclerosis

The extent of atherosclerosis was measured by quantifying the accumulation of CE in the aorta, including the entire vessel from the heart to the iliac bifurcation, according methods previously described (17). Briefly, the aortas were cleaned by removing all adherent adipose and connective tissue, and lipids were extracted in chloroform-methanol using 5α-cholestane as an internal standard. Cholesterol mass was measured by gas-liquid chromatography. Aortic CE (esterified cholesterol × 1.67) was calculated as the difference between free and total cholesterol as measured before and after saponification by extraction of the sterol into hexane. The delipidated tissue protein was digested in 1N NaOH, and total protein content was determined according to the Lowry method (18).

Quantification of mRNA levels

Real-time PCR analyses of liver tissue were performed according to methods previously published (19). Briefly, total mRNA was extracted from ∼80 mg of liver with Trizol (Invitrogen Life Technologies). After resuspension in 300 μl of diethyl pyrocarbonate water, mRNA was reverse transcribed into cDNA using Omniscript reverse transcriptase (Qiagen) under the following conditions: 37°C for 1 h and 93°C for 5 min. RT-PCR was done in triplicate with 5 μl of cDNA, 12.5 μl of SYBR GREEN PCR master mix (Applied Biosystems), 5.5 μl of diethyl pyrocarbonate water, and 1 μl of forward and reverse primer for a final reaction volume of 25 μl. PCR was then run on the Sequence Detection System 7000 under the following conditions: 50°C for 2 min; 94°C for 10 min; and 40 cycles of 94°C for 10s and 60°C for 1 min. The fluorescence measurement used to calculate threshold cycle (Ct) was made at the last time point. A dissociation step was run to ensure a single amplification product. The arbitrary unit value was calculated as follows: 1 × 10⁹ × e(−0.6931 × Ct). All values were normalized to cyclophillin mRNA concentration of the sample to take total mRNA concentration into account. Primer sequences used for qPCR are shown in supplemental Table III.

Statistical analyses

Data are expressed as the mean ± SEM. Results were analyzed using one-way ANOVA followed by Tukey's test for post hoc analyses. Differences were considered significant at P < 0.05. All analyses were performed using GraphPad Prism (version 4) software.

ALA-enriched diets are ineffective in lowering plasma cholesterol levels compared with fish-oil diet

TPC responses to the different fatty acids during the intervention period are illustrated in Fig. 1A. In as early as two weeks of diet intervention, dietary fat-specific differences were found among the diet groups. Both cis-monounsaturated fat and ALA-enriched diet-fed animals showed a significant increase in TPC compared with baseline, while fish oil-fed mice maintained lower TPC concentrations. This trend continued at 4, 8, and 16 weeks, indicating that increased amounts of ALA (high-flax diet) were unable to provide additional benefit compared with cis-mono and moderate flax diets. A similar pattern was observed when the cholesterol concentration in the LDL fraction was determined (Fig. 1B). At two weeks into the study, groups fed cis-mono and ALA-enriched diets exhibited LDL-cholesterol (LDL-c) concentrations of about 400 mg/dl, which were higher than the LDL-c values in the fish-oil group (∼206 mg/dl). At 4, 8, and 16 weeks, cis-mono and ALA-enriched diet-fed animals showed further increases in LDL-c, reaching values near 500 mg/dl at the end of the study, while fish-oil-fed animals did not show increasing LDL-c concentrations and values remained similar throughout the intervention period. When cholesterol concentrations were measured in the VLDL fraction, lower VLDL-c levels (24.3 mg/dl) were found in mice fed the high-flax diet compared with cis-mono and moderate-flax diet-fed animals (61 and 37 mg/dl, respectively), but VLDL-c levels were still higher compared with those measured in fish-oil-fed animals (5.7 mg/dl). Fish-oil diet was effective in reducing plasma triglyceride levels compared with baseline (111.2 mg/dl versus 61.2 mg/dl), and the hypotriglyceridemic effect persisted during diet intervention. No effect on plasma TG was measured in either cis-mono or ALA-enriched diet-fed animals (data not shown).

ALA incorporation in plasma LDL lowers cholesteryl oleate percentage and enriches LDL-CE with EPA

Studies in nonhuman primates and transgenic mice have reported alterations in LDL particle size and composition of lipid and fatty acid induced by dietary fat type (15, 16, 20, 21). The core lipid of LDL particles was somewhat different in composition in the fish-oil group, with the triacylglycerol percentage being the lowest compared with cis-mono and ALA-enriched diets (1.7% versus 3.67% and 2.5%, respectively) while a slightly lower cholesteryl ester percentage (41% versus 44.6%) was also found. Particle diameters, assessed by light scattering means, were not different among dietary fat groups, reflecting the modest differences in lipid composition (data not shown).

LDL cholesteryl ester fatty acid (CEFA) composition has been reported to be predictive of atherosclerosis (20–22). Fig. 2 illustrates LDL CEFA composition measured after 16 weeks of diet intervention. As shown in Fig. 2A, the predominant fatty acids in CE were monounsaturated fatty acids (mostly oleic acid), which made up close to 50% of the total. Compared with the cis-mono diet, the ALA-enriched diet resulted in a lower percentage of cholesteryl oleate (44% versus 55.6%), while the lowest percentage was found in mice fed the fish-oil diet (32.4%). Similar saturated and n-6 polyunsaturated CE percentages were found among diet groups. In both ALA- and fish-oil-diet groups, a significant enrichment in n-3 polyunsaturated CE was found. To identify which n-3 fatty acid contributed the most, the percentages of ALA, EPA, DPA, and DHA were determined as shown in Fig. 2B. In both ALA and fish-oil groups, EPA was the predominant n-3 fatty acid, even though the extent of n-3 enrichment of LDL-CE was quite different. In the moderate- and high-flaxseed-oil groups, EPA made up to 5.7% and 9.4%, respectively, of LDL-CE fatty acids, while in fish-oil group, 20% of LDL-CE fatty acid was represented by EPA. In mice fed ALA-enriched diets, a dose-dependent increase in ALA and EPA percentages was observed without a change in DHA percentages. Fish-oil-fed animals exhibited the greatest enrichment in both EPA and DHA, which together made up close to 28% of the total fatty acids in CE.

ALA lowers hepatic lipids and is incorporated into lipids as EPA

The primary source of cholesteryl oleate in plasma has been identified to be the hepatic enzyme acyl-CoA:cholesterol acyl-transferase 2 (ACAT2) (23–25). ACAT2 esterifies cholesterol using fatty acyl-CoA, mainly oleoyl-CoA. ACAT2 is the isoform of ACAT responsible for the synthesis of CEs incorporated into apoB-containing lipoproteins secreted by the liver and intestine (26, 27). Studies in hypercholesterolemic transgenic mice have elegantly demonstrated that alterations of dietary fatty acids shift the hepatic composition of fatty acyl-CoAs toward the diet composition so that diets enriched in PUFA provide less oleoyl-CoA available for esterification by ACAT2 (16). This shift would presumably result in a lower hepatic CE deposition in ALA- and fish-oil-fed animals. Fig. 3 shows the effect of dietary ALA and n-3 PUFA from fish oil on hepatic neutral lipid (CE and TG) concentrations. The fish-oil group had the lowest CE concentration while the cis-mono group had the highest hepatic CE accumulation (Fig. 3A). Although both ALA groups had lower hepatic CE concentrations compared with the cis-mono group (46 and 35.5 mg/g protein liver, respectively, versus 84 mg/g protein liver), the CE concentrations were higher than that seen in livers from the fish-oil group. A similar pattern was shown for hepatic TG concentrations (Fig. 3B), although in this case, the high-flax-diet group was as effective as fish oil in maintaining lower hepatic TG levels. We have previously observed that moderate amounts of ALA enriched hepatic lipids differently, with ALA being the main n-3 fatty acid in neutral lipids (CE and TG) and EPA the predominant fatty acid in phospholipids (PL) (unpublished observations). We analyzed the hepatic lipid fatty acid composition to determine whether higher amounts of ALA would provide a greater n-3 PUFA enrichment of hepatic lipids (Fig. 4). In agreement with our previous results, the extent of n-3 PUFA enrichment was different among hepatic lipids, with PL being the most enriched and CE the least enriched. In both ALA-diet groups, ALA was the predominant n-3 PUFA in neutral lipids, but it was absent from PL. ALA was incorporated in a dose-dependent manner in both CE and TG, and small percentages of EPA and DHA were found in these fractions as well. The n-3 PUFA enrichment of CE and TG was greater in the fish-oil group with DHA being the main n-3 PUFA. Phospholipid fatty acid composition differed from that of neutral lipids, and PL of both ALA groups had similar EPA and DHA percentages that made up close to 25% of the total fatty acids in PL; this pattern of n-3 fatty acid enrichment was not significantly different from that obtained by the fish-oil diet. In summary, compared to moderate, higher amount of ALA modestly increased the ALA and EPA percentages in CE and TG, but a higher amount of ALA provided an n-3 PUFA enrichment of hepatic PL as EPA that was comparable to that measured in the fish-oil group.

ALA does not protect mice from aortic CE deposition

Previous studies have shown that fatty acid compositional changes can translate into effects on atherosclerosis (15, 20). In this study we measured the accumulation of cholesteryl ester in the entire aorta as a chemical endpoint to quantitatively measure the extent of atherosclerosis after 16 weeks of diet intervention. Fig. 5 illustrates the atherogenic response to dietary fatty acids in the aorta. In agreement with previous reports (16, 21), cis-monounsaturated fat-rich diet promoted the highest aortic CE concentration at about 30 mg/g. Plasma and hepatic n-3 PUFA enrichment observed in both ALA groups did not result in a significant benefit to atherosclerosis, as indicated by aortic CE levels that were similar to those measured in the cis-mono group. As anticipated from earlier studies, the fish-oil group had a significantly lower aortic CE concentration.

Fish-oil-fed animals had significantly lower hepatic CE levels (Fig. 3) compared with both cis-mono- and ALA-fed animals, and this pronounced hypocholesterolemic effect, along with the lowest LDL-c concentration (Fig. 1B) and LDL cholesteryl oleate percentage (Fig. 2A), apparently contributed to the atheroprotection (Fig. 5). The cholesterol-lowering effect of fish oil can likely be attributed to one or more mechanisms, including reduced cholesterol synthesis and esterification, lowered fatty acyl-CoA synthesis, and increased cholesterol catabolism by conversion into bile acids. Several hepatic genes influence these processes and were analyzed by RT-PCR. No major differences were found among livers of the four diet groups in cholesterol synthesis or bile acid metabolism-related genes, with the exception of cholesterol-7-α-hydroxylase (CYP7A1) (Fig. 6A, B). CYP7A1 transcript was found lower (up to 70%) in fish-oil-fed mice compared with the other diet groups, which is opposite from the directional change anticipated. Both dietary and endogenously synthesized fatty acids are required to be converted into fatty acyl-CoAs by long chain acyl-CoA synthetases (ACSL) before being incorporated into CE, TG, and PL. Four isoforms (ACSL1, ACSL3, ACSL4, and ACSL5) have been found in rodent liver, and few reports have indicated that they might exert independent functions and exhibit differential affinity toward fatty acids (28). As indicated in Fig. 6C, only ACSL1 of the four ACSL isoforms was different among diet groups. As previously reported, no significant differences were found in ACAT2 mRNA levels among dietary fat groups (16, 29) while stearoyl-CoA-desaturase 1 (SCD1) transcript was found lower when polyunsaturated fat-enriched diet was fed compared with the cis-mono diet (Fig. 6D).

To date, a shorter dietary study (4 weeks) carried out under the same experimental conditions produced almost identical outcomes with regard to plasma and hepatic lipid concentrations and n-3 PUFA enrichment, suggesting that diet-induced alterations take place rapidly and contribute to the subsequent aortic response (data not shown).

Collectively these data indicate that in a mouse model of LDL-driven atherogenesis, robust reductions in TPC, LDL-c concentrations, and LDL cholesteryl oleate percentage are required for atheroprotection to be achieved. Fish-oil-fed animals experiencing all of these benefits were the only group protected from atherosclerosis.