In Vitro and In Vivo Digestibility of Soybean, Fish, and Microalgal Oils, and Their Influences on Fatty Acid Distribution in Tissue Lipid of Mice

The major FAs in soybean oil are linoleic (C18:2, 54.6%), oleic (C18:1, 21.7%), palmitic (C16:0, 10.9%), linolenic (C18.3, 6.6%), and stearic acid (C18:0, 4.5%), and the main FAs at the sn-2 position are linoleic (46.9%), oleic (22.0%), and linolenic acid (5.8%) (Table 1). The major FAs in fish oil are palmitic (18.5%), EPA (16.2%), palmitoleic (C16:1, 13.1%), DHA (10.8%), myristic (9.2%), and oleic acid (8.8%), showing that this oil is rich in the n-3 FAs of EPA and DHA; the main FAs at the sn-2 position include palmitic (20.0%), EPA (17.3%), palmitoleic (12.3%), DHA (19.0%), and myristic acid (9.8%). Microalgal oil mainly contains DHA (56.8%), DPA (12.3%), and palmitic (22.4%), whereas EPA is present in trace amounts (0.6%); at the sn-2 position, it contains DHA (67.3%), DPA (15.2%), and palmitic acid (11.7%). The DHA and DPA contents were 5.3. and 4.7 times higher in microalgal oil than in fish oil, respectively, whereas EPA was 25.7 times higher in fish oil. The fish oil used in the present study was obtained from menhaden, and the fatty acid composition was similar to commercial Atlantic and Gulf coast menhaden oils [24] (Joseph, 1985). The microalgal oil is a commercial water-extracted Chromista algae oil (Schizochytrium sp.).

Generally, long-chain unsaturated fatty acids (LC-USFAs) tend to be located at the sn-2 position and saturated fatty acids (SFAs) with a relatively small number of carbons are located at the sn-1,3 position of the TAG structure. In soybean oil, C18-USFAs and C16- and C18-SFAs respectively account for 77.4% and 22.5% of the FAs at sn-1,3 and 98.5% and 1.4% at sn-2. Fish oil and microalgal oil are composed of FAs with various carbon numbers and double bonds. At sn-2 and sn-1,3, PUFAs with C20 or higher account for 47.8% and 31.8% of FAs, respectively, in fish oil and 83.5% and 65.8%, respectively, in microalgal oil. Therefore, the constituent FAs in each oil are different with different locations in the glycerol backbone, which affects digestion and absorption upon ingestion.

The degree of oil hydrolysis was expressed in terms of the released free fatty acid (FFA) content (µM) via a step-by-step in vitro simulation of the digestion process with saliva, gastric, duodenal, and bile juices, lasting 30 and 120 min (Figure 1). The released FFA content at 30 min was significantly higher in soybean oil and fish oil than in microalgal oil, and that at 120 min was significantly higher in soybean oil than in fish oil and microalgal oil (p < 0.05).

In the body, the digestibility and absorption rates of oils differ due to variations in FA unsaturation, carbon length, and location in the glycerol backbone (sn-1,3 or sn-2). These variations affect the bioavailability of FAs and lipid metabolism [8]. Oil is emulsified by bile acid after ingestion and absorbed after being hydrolyzed by pancreatic lipase. Most FAs at sn-1,3 of TAG are hydrolyzed during digestion, whereas 22% of FAs at the sn-2 position are hydrolyzed, which is due to the regiospecificity of pancreatic lipase [7]. The degree of USFA hydrolysis by lipase is greater than that of SFA. Moreover, the degree of lipase-mediated short-chain (SC) FA hydrolysis is higher than that of LCFA. Notably, the hydrolysis rate of PUFA is reportedly low when the double bond is closer to the carboxyl group [8].

In the present study, the PUFA content at sn-1,3 was 1.9 times and 1.2 times higher in microalgal oil than in fish oil and soybean oil, respectively; in particular, microalgal oil has 2.0 times more PUFA with carbon length ≥ 20 than fish oil, while soybean oil has none. The double bond present in the FA structure and the carboxyl group were closest in DHA (C22:6, Δ4,7,10,13,16,19) and DPA (C22:5, Δ4,7,10,13,16), followed by EPA (C20:5, Δ5,8,11,14,17), linoleic (C18:2, Δ9,12), and linolenic acid (C18:3, Δ9,12,15). Bottino et al. [25] reported that DHA exhibited greater resistance to pancreatic lipase than EPA, and this phenomenon can be explained by the structural differences between these molecules. In other words, if the terminal methyl and carboxyl groups of PUFA are close, there is a steric hindrance effect on the lipase-mediated hydrolysis. Therefore, the higher polyunsaturation of FAs located in sn-1,3 and the double bond close to the carboxyl group inhibit the TAG hydrolysis activity of lipase, resulting in lower digestion rates of microalgal oil and fish oil compared to that of soybean oil. Aarak et al. [26] analyzed the profile of released FAs after in vitro digestion of salmon oil with human duodenal juice and a commercial enzyme preparation (porcine pancreatin and bile), and reported that the concentration of linoleic acid was the highest, followed by EPA and DHA (p < 0.05).

In an in vitro hydrolysis study by Ikeda et al. [27], wherein tridocosahexaenoyl glycerol (TriDHA) and trieicosapentaenoyl glycerol (TriEPA) were treated with porcine pancreatic lipase, the hydrolysis rates of TriEPA and TriDHA in the initial period were low, but it increased markedly afterwards; in the end, most TriEPA and 80% of TriDHA were hydrolyzed. Similarly, in this study, when the duration of in vitro digestion was increased from 30 to 120 min, the released FFA contents of soybean oil and fish oil increased by 1.41 and 1.34 times, respectively, whereas that of microalgal oil (51.5% DHA at sn-1,3; 67.3% at sn-2) increased by 1.53 times, indicating that longer digestion led to higher hydrolysis of DHA present in TAG at high concentrations.

The body weights of mice increased in all four groups during the 4-week period. The body weight gain (3.28~5.55 g) and feed efficiency ratio (FER, 4.23~6.54%) of mice fed the experimental diet were significantly higher than that of the normal group (0.37 g, 0.48%) (p < 0.05), and there was no significant difference in body weight gain and FER among the SO, FO, and MO groups (p > 0.05) (Table 2). Experimental diet-fed mice had significantly heavier liver and epididymal fat than normal diet-fed mice; the liver weight was the highest in the MO group and lowest in the SO group (p < 0.05). There were no significant differences in brain and heart weights among the diet groups (p > 0.05) (Table 2). The crude lipid contents of the brain and epididymal fat were similar in all four diet groups (p > 0.05), while the lipid content of the liver was significantly lower in the FO compared to those in the SO and MO groups (p < 0.05) (Table 2). According to the study of Yuan et al. [28], the rat supplemented with 10% fish oil in a high-fat high-cholesterol diet showed a significantly reduced hepatic triacylglycerol content than the 10% lard diet, demonstrating that fish oil protects against high-fat high-cholesterol diet-induced non-alcoholic fatty liver disease by improving lipid metabolism and ameliorating hepatic inflammation.

The apparent digestibility values were determined by measuring the coefficients of digestibility for soybean oil, fish oil, and microalgal oil in mice (Table 3). The fecal crude lipid content in each group was in the following order: microalgal oil (15.87%) > fish oil (9.24%), soybean oil 20% (8.89%) > soybean oil 7% (6.87%). The fecal lipid content in the microalgal oil intake group was significantly higher than in the other groups (p < 0.05). The apparent digestibility of the soybean oil 20% (SO) was significantly higher than that of the 7% intake group (NO) (p < 0.05). Among the experimental diet groups, the apparent digestibility of the microalgal oil group (91.49%) was significantly lower than the others (p < 0.05), and there was no significant difference between the soybean oil (96.50%) and fish oil (96.99%) groups (p > 0.05). This suggests that microalgal oil, due to its lower digestibility and absorption rate compared to soybean oil and fish oil, was not be absorbed easily in the body and is secreted as feces, resulting in low apparent digestibility.; This is consistent with the results of the in vitro digestibility model (Figure 1). According to the in vivo study of Tou et al. [21] using rats, krill oil (93.1%) had the highest EPA and DHA contents but lower apparent digestibility compared to salmon oil (98.8%), tuna oil (98.0%), and menhaden oil (97.2%). The EPA and DHA concentrations in krill oil were 19.92~22.83% and 10.60~12.33%, respectively [29], and its DHA concentration was lower than that of microalgal oil in the present study (56.8%).

Table 3 shows the apparent digestibility values of individual FAs in each diet group. In the soybean oil 7% and soybean oil 20% groups, the digestibility coefficient of individual FA was significantly higher for C18:3 (96.2~96.9%), followed by C18:2 (93.6~94.7%), C18:1 (91.5~92.7%), C16:0 (88.2~88.4%), and C18:0 (85.9~86.5%) (p < 0.05). FA digestibility in the fish oil group was significantly higher for C20:5 (95.7%), followed by C22:6 (94.5%), C16:1 (93.8%), C18:1 (92.7%), C16:0 (89.2%), and C18:0 (76.8%) (p < 0.05). FA digestibility in the microalgal oil group was in the following order: C22:5 (99.9%) > C22:6 (93.2%) > C16:0 (80.8%) (p < 0.05). Generally, FA digestibility increased with reduced chain length and further increased with increasing unsaturation level [8]. In all groups, the digestibility coefficients of palmitic acid and stearic acid were significantly lower than those of other unsaturated FAs, because the long-chain SFA positioned at sn-1,3 in each oil has a lower rate of pancreatic lipase-mediated hydrolysis; during digestion, free palmitic and stearic acids form insoluble calcium soap and are excreted in feces [8]. In fish oil, EPA and DHA had higher digestibility than oleic acid and palmitoleic acid, and this was in agreement with the results reported by Ikeda et al. [27], indicating that n-3 FAs are slowly released from the sn-1,3 positions by pancreatic lipase at the initial period of digestion, but as digestion progresses, the hydrolysis rate increases significantly and may be even higher than that of oleic acid. In the microalgal oil group, the digestibility of DPA was significantly higher than that of DHA, since DHA at the sn-1,3 position resists pancreatic lipase hydrolysis, whereas DPA is hydrolyzed more easily [25].

Microalgal oil extracted from Schizochytrium sp. was purchased from Source-Omega LLC. (Chapel Hill, NC, USA). Fish oil extracted from menhaden was purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA), and soybean oil was purchased from Ottogi (Pyeongtaek, Gyeonggi-do, Korea). Bile salts, α-amylase, lipase from porcine pancreas, bovine serum albumin, pancreatin from porcine pancreas, mucin from porcine stomach, sodium molybdate, chromic oxide, and potassium dichromate were purchased from Sigma-Aldrich Chemical Co. Triundecanoin, a standard for fatty acid analysis, was purchased from NU-CHEK PREP, Inc. (Elysian, MN, USA) and Supelco 37 Component FAME Mix was purchased from Supelco Inc. (Bellefonte, PA, USA).

Samples (50 mg), 1 mL triundecanoin in isooctane (5 mg/mL), and 1.5 mL of 0.5 N methanolic NaOH were added to a test tube, followed by saponification at 85 °C for 10 min. After cooling, 2 mL of BF₃-methanol were added, then methylated at 85 °C for 10 min and cooled. Then, 2 mL of isooctane were added and vortexed for 1 min, and 1 mL of saturated NaCl solution was added. After centrifugation (1224× g, 5 min), the supernatant was passed through a sodium sulfate column to remove moisture. The FA composition was analyzed using gas chromatography equipped with a flame ionization detector (GC-2010 Plus, Shimadzu Corp., Kyoto, Japan) and SP^TM-2560 column (100 m × 0.25 mm × 0.2 µm, Supelco Inc.). The oven temperature was maintained at 100 °C for 5 min, increased to 240 °C at 4 °C/min, and maintained for 40 min. The column flow (carrier gas) was 1.00 mL/min (N₂), the injector and detector temperatures were 250 and 260 °C, respectively, the split ratio was 100:1, and the injection volume was 1 µL. Each FA was identified based on the retention time of the standard chromatogram, and the area of each peak was expressed as % of total FAs.

Oil (25 mg), tris-HCl buffer (pH 7.6, 25 mL), 0.05% bile salt (6.25 mL), 2.2% CaCl₂ (2.5 mL), and pancreatic lipase (25 mg) were mixed in a test tube, and allowed to react for 3 min at 37 °C, with 30-s vortexing repeated 3 times. Then, 6 mL of diethyl ether were added and vortexed for 1 min. After centrifugation (1224× g, 5 min), the supernatant was passed through an anhydrous sodium sulfate column, concentrated with N₂, and loaded onto a thin-layer chromatography (TLC) F₂₅₄ silica plate (20 × 20 cm, Merck, Kenilworth, NJ, USA). After separation with a developing solvent (n-hexane: diethyl ether: acetic acid = 50:50:1, v/v/v), the 2-MAG band in the TLC plate was taken and analyzed with GC, and positional FAs were calculated using the following equation:

Fatty acid composition at sn-1,3 position (%) = [3 × total fatty acid composition (%)—fatty acid composition at sn-2 (%)]/2.

The in vitro digestion rate of the oil was measured by modifying the method described by Versantvoort et al. [37] and Chang et al. [20]. All digestive juices were prepared and used on the day of the experiment. Oil (100 mg) and saliva juice (1.2 mL) were mixed in an Erlenmeyer flask at 37 °C and 80 rpm for 5 min, and gastric juice (2.4 mL) was added and reacted for 2 h. NaHCO₃ solution (0.4 mL), bile juice (1.2 mL), and duodenal juice (2.4 mL) were added, followed by hydrolysis reaction for 30 and 120 min. For control samples (0 min), duodenal juice without pancreatin and lipase were added. After the reaction, the lipase inhibitor 4-bromophenylboronic acid (100 μL) was added to stop digestion. For lipid extraction of the digested solution, n-hexane was added and centrifuged at 1763× g for 5 min, and the supernatant was passed through an anhydrous sodium sulfate column (repeated 3 times). Next, 1N HCl (0.5 mL) was mixed with the remaining lower part for 1 min, and 10 mL of n-hexane were mixed and centrifuged, and the supernatant was removed, and the moisture removed (repeated 3 times). The supernatants were combined, and after removing n-hexane with N₂, 10 mL of ethanol:n-hexane (1:1, v/v), and 1 mL of 1% phenolphthalein were added, and titrated with 50 mM KOH solution. The in vitro digestion rate of each oil was expressed as released free fatty acids (µM) using the following equation [38]:(1)ReleasedfreefattyacidsμMoftheoils=volumeofKOHmL×normalityofKOHmN×100weightoftheoilmg

Forty male ICR mice, aged 6 weeks (initial body weight of approximately 30 g), were obtained from Samtako BioKorea (Osan, Korea) and housed in a temperature-controlled environment at 22 ± 3 °C and 50 ± 10% relative humidity. The mice were maintained on a 12-h light/dark cycle and provided with a designated diet (3.5 g) and ad libitum access to water. After one week of acclimation with a standard diet (Samtako BioKorea), mice were randomly assigned to one of four dietary treatment groups, and two mice were housed in one cage for 4 weeks. The NO group was fed a normal diet with 7% soybean oil, and SO, FO, and MO groups received experimental diets containing 20% soybean oil, fish oil, and microalgal oil, respectively. Body weight and food intake were measured twice a week by each cage, in which food intake was calculated by subtracting the amount of remaining food from the amount of provided food by each cage. The composition of the diets, based on the AIN-93G DIET conditions, is presented in Table 6.

At week 4 of the diet intervention, a diet containing 0.5% chromic oxide was provided for 5 consecutive days, and green feces were collected. The mice were fasted for 24 h before sacrifice, and dissection was performed between 10:00 am and 12:00 pm. The liver, brain, and epididymal fat of the mice were removed, rinsed, weighed, and stored in a deep freezer (Nihon Freezer Co., Ltd., Saitama, Japan). This study was conducted in accordance with the Institutional Animal Care and Use Committee of the Southeast Medi-Chem Institute (IRB No: SEMI-16-10).

Freeze-dried feces (0.5 g), 2 mL of pyrogallol in ethanol (50 mg/mL), and 10 mL of 8.3 M HCl were placed in a vial and vortexed for 30 s. The lipid of feces was extracted in a shaking water bath at 80 °C and 200 rpm for 1 h; the vial was taken out every 20 min and vortexed for 30 s. After extraction, 15 mL diethyl ether was added, mixed for 1 min, and centrifuged (1763× g, 3 min). The upper layer was mixed with 15 mL of petroleum ether and centrifuged. The supernatant was passed through a sodium sulfate column to remove moisture, and after removing the solvent with N₂, the crude lipid content was calculated.

For lipid extraction from tissues, the liver, brain, and epididymal fat were cut into pieces and placed in a test tube, mixed with 2 mL of 0.9% saline, and homogenized using an ultrasonic processor (Sonics & Materials, Inc., Newtown, CT, USA) with 30% power for 1 min twice. The crude lipid was extracted with 12 mL of chloroform: methanol (2:1, v/v) for 30 min in a shaking water bath at 50 °C and 190 rpm, then centrifuged (441× g) for 5 min. The bottom layer was concentrated with N₂. The fatty acid compositions of fecal and tissue lipids were analyzed by GC.

Green feces obtained by feeding a chromic-oxide-supplemented diet were collected, lyophilized, and stored at −80 °C. A digestion solution was prepared with 450 mL of nitric acid, 150 mL of perchloric acid, and sodium molybdate (0.6 g). Crushed feces (250 mg) was mixed with the digestion solution (10 mL) in a 100-mL Kjeldahl flask and heated at 300 °C. The heating was continued until brownish smoke appeared from the digestion solution; at this point, the heating was stopped when the color changed to yellowish or orange. After cooling to room temperature, the digested solution was filtered, placed in a 50-mL volumetric flask, filled with distilled water, and absorbance was measured at 440 nm using a UV spectrophotometer (Optizen 2120UV, Mecasys Co., Ltd., Daejeon, Korea). After obtaining a standard curve with potassium dichromate solution, the chromic oxide content in the feces was calculated. The apparent digestibility of oil (or FA) was determined by the coefficient of digestibility, which is a measure of bioavailability expressed as the percentage of ingested oil (or FA) that was not excreted in the feces [15]. The coefficient of digestibility of oil and individual FAs was measured using the following equation:(2)Coefficientofdigestion%=100−crudelipidfattyacidcontentinfeces%oilfattyacidcontentindiet%×chromicoxidecontentofdiet%chromicoxidecontentoffeces%×100

Analysis of variance (ANOVA) was performed with Statistical Analysis System 9.2 (SAS Institute Inc., Cary, NC, USA), and the statistical significance of the means was determined by Duncan’s multiple range test at p < 0.05.