LL were purchased from a local herbal market (Daegu, South Korea) and extracted with hot water by heating for 2 h at 70°C. The extract was then lyophilized and re-dissolved in water to 0.1 mg/ml.

GS (1 kg, moisture content, 93%) was provided by Foodwell Co., Ltd. (Daegu, South Korea), and was extracted with hot water for 2 h, then concentrated using a rotary evaporator (N-1000; Tokyo Rikakikai Co., Ltd., Tokyo, Japan).

Acai (Euterpe oleracea Mart.) puree was obtained from All-In-One International Company (Seoul, South Korea). Acai puree (moisture content 86.5%) was lyophilized, extracted in hot water at 70°C, and centrifuged (2,000 x g) to separate solids.

Total phenolic content of the extracts was determined by using phosphomolybdic-phosphotungstic acid including Folin-Ciocalteu reagent (22). Briefly, each sample (0.1 mg/ml, 50 ml) was filtered with Whatman #1 paper and an aliquot (0.5 ml) of the sample added to 2.5 ml of 0.2 N Folin-Ciocalteu reagent for a 5-min incubation. Na₂CO₃ of 2 ml 75 g/l were then added and the total volume was raised to 25 ml with distilled water; this solution was incubated at room temperature for 2 h. Absorbance was measured at 760 nm using a 1-cm cuvette in a Perkin-Elmer UV-Vis Lambda 25 spectrophotometer. Tannic acid (0–800 mg/l) was used to produce a standard calibration curve. Total phenolic content was expressed in mg of tannic acid equivalents (TAE per gram of extract.

Total flavonoid content was determined by the Dowd method as modified by Meda et al (23). Five milliliters of 2% aluminum trichloride (AlCl₃) in methanol were mixed with the same volume of extract (0.4 mg/ml). Absorption readings were obtained at 415 nm after 10 min using a UV-Vis spectrophotometer (Ultrospec 2100 Pro; Amersham Biosciences, Amersham, UK) against a blank sample consisting of 5 ml extract solution with 5 ml methanol without AlCl₃. Total flavonoid content was determined using a standard curve of catechin dilutions (0–100 mg/l). Total flavonoid content was expressed as mg of catechin equivalents per gram of extract.

Total anthocyanin content was determined spectrophotometrically by the pH differential method (24) and calculated using mg/l equivalents of cyanidin-3-glucoside with an extinction coefficient of 29,600 (25).

Ferric reducing ability (FRA) was determined using a previously described method (26) with slight modifications (27). Briefly, freshly prepared FRA reagent (300 μl) was mixed with each extract or tissue homogenate as an antioxidant source, after which FRA was determined by measuring the absorbance at 593 nm every 20 sec over a 10-min period at 25°C using a DU800 spectrophotometer (Beckman Coulter, Fullerton, CA, USA) equipped with a MultiTemp III thermocirculator (Amersham, Uppsala, Sweden).

Fresh human low-density lipoprotein (LDL; 300 μg of protein) was incubated with each extract (final concentration, 1 μg/ml) in a 500-μl reaction volume in the presence of 10 μM (final concentration) CuSO₄ for up to 2 h. During incubation, conjugated diene formation was monitored at 234 nm at 24.5°C (28) using a DU800 spectrophotometer equipped with a MultiTemp III thermocirculator. In order to verify spectroscopic data, oxidized LDL samples were subjected to electrophoresis on a 0.5% agarose gel for comparison of relative electromobility (29).

Human plasma was obtained from healthy male volunteers at the Yeungnam University Hospital (Daegu, South Korea). ApoA-I was purified from the human plasma high-density lipoprotein (HDL) fraction (1.063<d<1.225) using ultracentrifugation and column chromatography following the method described by Brewer et al (30). Purified apoA-I was lyophilized and stored at −80°C until use.

Determination of apoA-I glycation was carried out according to the method of McPherson et al (31) with slight modifications, including the use of human apoA-I (20,21). The extent of glycation was determined by reading the fluorometric intensity at 370 nm (excitation, Ex) and 440 nm (emission, Em) using a spectrofluorometer (LS55; Perkin-Elmer, Norwalk, CT, USA) with WinLab software package 4.00 (Perkin-Elmer). Each extract (final concentration, 10 μg/ml; 10 μl volume) was treated to test its inhibitory ability in a 100-μl reaction volume. Percentage inhibition was calculated from fluorescence intensity (FI) with the following formula: %   inhibition = 100   x  { 1 - [ sample   FI - background   FI negative   control   FI - background   FI ] }where sample FI = apoA-I (0.1 mg/ml; 82 μl) + fructose (final concentration, 250 mM; 8 μl) + each water extract (final concentration, 10 or 100 μg/ml; 10 μl); background FI = apoA-I; and negative control FI = apoA-I + fructose + water.

In order to confirm spectrofluorometry data, glycated apoA-I was electrophoresed on 15% SDS-PAGE gels with or without extract; more-extensively glycated apoA-I showed a tendency to multimerize, as previously reported (32,33).

Recombinant HDL (rHDL) was synthesized in accordance with the method described by Cho (34) with trace amounts of [³H]-cholesteryl oleate (TRK886, 3.5 μCi/mg of apoA-I; GE Healthcare, Uppsala, Sweden) to act as a CE donor containing apoA-I and cholesteryl oleate. Human serum (0.05 ml) and LDL (0.25 mg/ml, 0.05 ml volume) were used as CETP and CE-acceptor sources, respectively. Different concentrations of each extract (in 0.01 ml volume) were added as CETP inhibitors. CETP inhibition was calculated as follows: %   inhibition = 100   ×   { 1 - [ sample   ( cpm ) - blank   ( cpm ) control   ( cpm ) - blank   ( cpm ) ] }where sample is extract treated as a source of inhibitor and the control lacks inhibitor.

LDL (1.019<d<1.063) was purified from healthy human plasma obtained from the blood bank of Yeungnam University Medical Center (Daegu, South Korea) by ultracentrifugation according to standard protocols (35). The density was appropriately adjusted by the addition of NaCl and samples were centrifuged for 22 h at 10°C and 100,000 x g using a Himac CP-90α (Hitachi, Tokyo, Japan) at the Instrumental Analysis Center of Yeungnam University. Oxidized LDL (oxLDL) was produced by incubating the LDL (1.019<d<1.063) fraction with CuSO₄ (final concentration, 10 μM) for 4 h at 37°C. OxLDL was then filtered (0.2 μm) and analyzed using the thiobarbituric acid reacting substances (TBARS) assay to determine the extent of oxidation, as described previously (36).

LDL acetylation (to generate acLDL) was performed using saturated sodium acetate and acetic anhydride, according to a previously described method (37). After acetylation and subsequent dialysis, the acLDL protein content was determined and samples were filtered prior to use.

THP-1 cells, a human monocyte cell line, were obtained from the American Type Culture Collection (ATCC) (#TIB-202; Manassas, VA, USA) and maintained in RPMI-1640 medium (Hyclone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS), until needed for experimentation. Cells that had undergone no more than 20 passages were incubated in medium containing phorbol 12-myristate 13-acetate (PMA; final concentration, 150 nM) in 24-well plates for 48 h at 37°C in a humidified incubator (5% CO₂ and 95% air) to induce differentiation as macrophages.

To compare anti-atherosclerotic activity using a cellular model, each extract (0.1 mg/ml) was assayed for its effect on macrophage oxLDL uptake, as previously reported by our research group (34). Differentiated and adherent macrophages were then rinsed with warm PBS and incubated with 400 μl of fresh RPMI-1640 medium containing 1% FBS, 50 μl of oxLDL (50 μg of protein in PBS) or acLDL (50 μg of protein in PBS), and 50 μl of each extract, for 48 h at 37°C in a humidified incubator. After incubation, the cells were stained with Oil Red O solution (0.67%) to visualize the amount of lipid species in the cell.

Zebrafish maintenance and experimental procedures were approved by the Committee of Animal Care and Use of Yeungnam University (Gyeongsan, South Korea). A high cholesterol (HC) diet containing 4% cholesterol was made by soaking tetrabit [Tetrabit Gmbh D49304; 47.5% crude protein, 6.5% crude fat, 2.0% crude fiber, 10.5% crude ash, containing vitamin A (29,770 IU/kg), vitamin D3 (1,860 IU/kg), vitamin E (200 mg/kg), and vitamin C (137 mg/kg); (Melle, Germany)] in a solution of cholesterol (Sigma #C-3045; Sigma-Aldrich, St. Louis, MO, USA) in diethyl ether. After ether evaporation, HC diet was mixed with lyophilized fruit extract (a final concentration of 10% w/w of powder/tetrabit). Each group (n=70) consumed the designated diet (20 mg/day/fish), as shown in Table II. The zebrafish were maintained at 28±1°C under a 14 h:10 h light:dark cycle.

After feeding for 5 weeks, blood (2 μl) was drawn from the heart of adult fish and combined with 5 μl of PBS-EDTA, then collected into EDTA-treated tubes (final concentration, 1 mM). Serum total cholesterol (TC), HDL-cholesterol, and triglycerides (TG) were determined using a commercial assay kit (cholesterol, T-CHO; triglycerides, Cleantech TS-S; were from Wako Pure Chemical, Osaka, Japan). CETP activity of zebrafish serum samples was compared among groups.

Hepatic tissues were collected from the same lobes of at least 15 zebrafish, flash-frozen in liquid nitrogen, and stored at −70°C. Frozen liver sections were sliced to a thickness of 7 μm using a microtome (CM1510S; Leica, Nussloch, Germany), and stained with Oil Red O and hematoxylin in order to evaluate any light-microscopic features or fatty liver changes. Morphological changes were compared with a Nikon Eclipse TE2000 microscope (Tokyo, Japan).

An aliquot of hepatic tissue (50 mg of liver in 0.5 ml of PBS) from each group was homogenized for 3 min (150 x g) in an ice bath using a tissue homogenizer (Euro-ST; Eurostar, IKA-Werke, Staufen, Germany). After brief centrifugation (10,000 x g, 3 min), protein concentration in the supernatants was determined and supernatants were resuspended at an equal concentration.

Malondialdehyde (MDA) levels in each homogenate were determined by adding 10 mM 1-methyl-2-phenylindole to each sample for color development, as described previously (38). Reactions were initiated by adding HCl (final concentration, 5%) and incubated for 1 h at 45°C. After the reaction, mixtures were centrifuged and absorbance readings of supernatants were taken at 586 nm. The MDA content was calculated from a calibration curve using an MDA standard (1,1,3,3-tetramethoxy propane, Sigma #T9889; Sigma-Aldrich).

All data are expressed as the mean ± SD from at least three independent experiments with duplicate samples. Comparisons between results were made by analysis of variance (ANOVA) and differences were assessed by the Student’s t-test using the SPSS program (version 14.0; SPSS, Inc., Chicago, IL, USA) for in vitro experiments. In Table II, for in vivo tests, data were evaluated via one-way ANOVA using SPSS and the differences between the means were assessed using the Duncan’s multiple-range test. Statistical significance was defined as P<0.05.

We performed hot-water extractions on fruit samples to isolate antioxidant compounds. As shown in Table I, phenolics and flavonoids were enriched in GS extract, which contained 2.5- and 1.9-fold more total phenols than LL and acai, respectively. In the same quantity of extract, LL possessed 5.5-fold higher flavonoid content than acai; however, the acai extract exhibited 1.3-fold more phenol content than the LL fraction. Anthocyanin content was similar in GS and acai extracts although they were not identified in LL.

As shown in Fig. 1A, acai, LL, and GS extract demonstrated the strongest ferric ion reductive ability (FRA), with 77, 65 and 63% increases from initial levels, respectively. However, the GS extract showed the highest initial absorbance at 593 nm (data not shown), indicating that its FRA was most pronounced.

In a DPPH radical-scavenging assay (Fig. 1B), the LL fraction showed the strongest scavenging activity, removing approximately 92% of radicals. LL and acai fractions showed 47 and 38% radical removal, respectively, during the same length of incubation (Fig. 1B).

Cupric-ion mediated LDL oxidation was inhibited by all extracts (Fig. 2A). LDL exposed to each extract treated showed no increase in A234 except in samples treated with LL; these demonstrated an absorbance increase of approximately 23% (Fig. 2A). Agarose gel electrophoresis confirmed that electromobility of extract-treated LDL was reduced relative to oxidized LDL, which demonstrated a smear with faster migration due to apo-B fragmentation and increased negative charge. As a positive control, vitamin C treated LDL (final concentration, 0.5 mM) showed no increase of A₂₃₄ and the slowest electromobility.

As shown in Fig. 3, acai extract exhibited the strongest anti-glycation activity of those assayed, at 89% relative to the control at a final concentration of 0.1 mg/ml, although LL and GS extracts could also reduce glycation concentration-dependently (68 and 58% inhibition, respectively). Electrophoresis on 15% SDS-PAGE gels identified a higher degree of multimerization in glycated apoA-I samples (lane 7) relative to control apoA-I (Fig. 3B; lane 8). However, apoA-I exposed to high dosage of extracts (final concentration, 0.1 mg/ml) showed almost no multimerization (Fig. 3B; lanes 1–3).

Oxidized LDL was phagocytosed strongly by macrophages compared with native LDL (Fig. 4A). Among the extracts, treatment with acai evoked the highest LDL uptake (Fig. 4D), leading to an approximately 47% reduction in Oil Red O staining relative to PBS treatment in the presence of oxLDL. However, treatment with the LL (Fig. 4C) and GS (Fig. 4E) fractions elicited weaker inhibitory activity, with approximately 11% and 29% less staining than PBS treatment, respectively.

Uptake of acLDL was also inhibited by treatment with all extracts (Fig. 5). In particular, treatment with acai, LL, and GS extracts caused 76, 70 and 48% inhibition, respectively, of uptake in the presence of acLDL compared with PBS treatment. In cell-based assays, acai extract showed the strongest inhibition of uptake of LDL modified by either oxidation or acetylation.

As shown in Fig. 6, all fractions showed moderate in vitro CETP inhibitory activity of 5–22% with treatment of either 0.1 or 0.01 mg/ml extract (final concentration). Acai manifested the most pronounced CETP inhibitory capacity at 0.01 mg/ml, but no increase was detected with higher concentration (0.1 mg/ml). However, the GS extract did not show notable inhibition activity against human CETP.

Currently, many pharmaceutical companies are developing CETP inhibitors, as CETP is established to be atherogenic (39). CETP promotes the redistribution and equilibration of CE and TG between HDL and apo-B harboring proteins (LDL). It has been proposed that inhibition of CETP might augment levels of HDL-C (40).

After 4 weeks feeding on an HCD, serum TC and TG levels of group 2 (HCD group) were increased around 2.5-fold above levels in group 1 (control) (Table II). Zebrafish groups fed LL (group 3) or acai (group 4) showed 5 and 9% decreases in serum TC, respectively, while the grove consuming GS showed an 8% increase in serum TC compared with the HC control (group 2). LL- and GS-consuming groups showed 20 and 6% decreases, respectively, in serum TG compared with HC control. However, the acai-consuming group demonstrated a 14% increase in serum TG.

Serum glucose levels were 3-fold increased in group 2 compared with controls (group 1), indicating that increased cholesterol uptake might be connected to an increase in plasma blood glucose to elevate serum glucose levels, in good agreement with our previous report (20). LL- and GS-consuming groups demonstrated remarkable hypoglycemia, with blood glucose reduced below that of the ND control (group 1). Although the acai-fed group (group 4) showed lower serum glucose than the HCD group (group 2), its capacity to reduce blood glucose was weakest among extracts.

Interestingly, the acai palm’s primary fatty acid is monounsaturated oleic acid (7). Monoacylglycerol (MAG) containing oleic acid has been reported to exert potent antioxidant effects (41), suggesting that oleic acid might impart some degree of acai’s antioxidant efficacy.

CETP activity was nearly 1.4-fold higher in the serum of hypercholesterolemic zebrafish (group 2) than the normolipidemic (group 1; 0.02 ml serum tested for each group), indicating that HCD consumption elevates CETP activity as well as blood cholesterol. Among the groups fed HCD, those treated with acai exhibited the lowest CETP activity (33% lower than group 2), while the GS-fed group (group 5) showed slightly higher serum CETP activity than group 2. This in vivo observation is in good agreement with results from in vitro assays (Fig. 6), in which the extract of acai showed the strongest inhibitory ability for human CETP, while the extract of GS did not.

Histological analyses of hepatic tissue showed that zebrafish fed a normal diet exhibited normal, non-pathological histochemical features (Fig. 7). However, severe manifestations of fatty liver were observed in HCD-consuming fish; neutral lipid content, visualized as red color, was augmented, and severe infiltration of inflammatory cells was illuminated by the hematoxylin counterstaining. Among the extract-fed groups, those fed LL or acai demonstrated decreases in inflammatory cells and reduced intensity of Oil Red O staining, suggesting a reduction in hepatic fat storage and less inflammation. However, the GS-fed group showed a mild increase in fatty liver with darker-red staining intensity and a visible increase in inflammatory cell numbers.

Malondialdehyde (MDA) levels in hepatic tissue were 4-fold increased in the HCD group compared to those fed a normal diet, suggesting that high cholesterol consumption induced production of oxidized species (Fig. 7B). However, LL- and acai-fed groups had diminished MDA levels, in agreement with histological analyses demonstrating reduced infiltration of inflammatory cells (Fig. 7A). MDA levels and Oil Red O staining intensity were directly correlated.

FRA assays with hepatic homogenate showed the highest antioxidant potential in the ND-fed group, with an increase of 53% from initial levels; the HCD-fed group showed a 9% increase in antioxidant capacity. Among the experimental groups, the group fed GS demonstrated the strongest FRA, at approximately 32% above initial levels, while acai- and LL-fed groups showed 16 and 23% increases, respectively.

In conclusion, LL, GS, and acai extracts shared antioxidant, anti-inflammatory, and anti-atherosclerotic abilities in a cellular model and in hypercholesterolemic zebrafish, although there is incomplete overlap of beneficial activities. These results can be applied to the development of functional foods incorporating these ingredients that can alleviate symptoms and pathology of chronic metabolic diseases such as cardiovascular disease or diabetes.