Effect of Randomly Interesterified Triacylglycerols Containing

Medium- and Long-Chain Fatty Acids on Energy Expenditure

and Hepatic Fatty Acid Metabolism in Rats

Hisami SHINOHARA,y Akiko OGAWA, Michio KASAI, and Toshiaki AOYAMA

Division of Healthcare Science, Research Laboratory, The Nisshin OilliO Group, Ltd.,

1-Banchi, Shinmei-cho, Yokosuka, Kanagawa 239-0832, Japan

Received January 24, 2005; Accepted July 15, 2005

In our previous studies, medium- and long-chain

triacylglycerols (MLCT), randomly interesterified tri-

acylglycerols containing medium-chain and long-chain

fatty acids in the same glycerol molecule, significantly

reduced body fat accumulation in humans and rats. To

clarify mechanism(s) for this effect of MLCT, we

measured energy expenditure and hepatic fatty acid

metabolism in rats by comparison with long-chain

triacylglycerols (LCT) or medium-chain triacylglycerols

(MCT). MLCT, compared with LCT, showed signifi-

cantly lower body fat accumulation, higher 24-h energy

expenditure and acyl-CoA dehydrogenase activity meas-

ured using octanoyl-CoA as a substrate, and similar

lipogenic activity. MCT, compared with LCT, showed

significantly higher energy expenditure, but fat accu-

mulation was comparable. Additionally, MCT exhibited

significantly higher lipogenic activity than the other oils.

These data suggest that enhancement of energy expen-

diture and medium-chain fatty acids (MCFA) oxidation

without activating de novo lipogenesis are responsible at

least for the lower body fat accumulation in rats fed

MLCT. The activation of hepatic lipogenesis by exces-

sive intake of MCFA might counteract their preventive

effects on body fat accumulation.

Key words: medium-chain fatty acid; body fat accumu-

lation; energy expenditure; fatty acid oxi-

dation; lipogenesis

Medium-chain fatty acids (MCFA, composed ofchains of 8 to 10 carbon atoms) are found in edibleoils such as coconut oil. Compared with long-chain fattyacids (LCFA), MCFA have several unique physiologicaland biological characteristics, as reported elsewhereover the past half-century.1–3) Medium-chain triacylgly-cerols (MCT), composed exclusively of MCFA, are

metabolized differently from long-chain triacylglycerols(LCT), composed exclusively of LCFA. MCFA areabsorbed via the portal system and are transported to theliver directly, whereas LCT are absorbed via theintestinal lymphatic ducts and transported as chylomi-crons through the thoracic duct to reach systemiccirculation. MCFA and hence MCT are easily oxidizedbecause their intramitochondrial transport does notrequire carnitine palmitoyltransferase (CPT), a rate-limiting enzyme of mitochondrial �-oxidation.4) MCTare absorbed and metabolized as rapidly as glucose andhave more than twice the caloric density of protein andcarbohydrate. They are utilized to prevent obesity andseveral lifestyle-related diseases.4) However, it is diffi-cult to substitute MCT for LCT in the diet for long-termtherapy, largely because their lower smoke point andhigher tendency to bubble make MCT difficult to use asa cooking oil.Recently, to overcome this disadvantage, we devel-

oped a new cooking oil containing triacylglycerolscomposed of medium- and long-chain fatty acids(MLCT), which are randomly interesterified triacylgly-cerols containing MCFA and LCFA in the same glycerolmolecule.5) MLCT contain about 13% of MCFA, incontrast to MCT, which contain MCFA exclusively.MLCT can replace LCT in the daily diet, and in humanand rat studies this replacement has been shown to causesignificant lower body fat accumulation.6–10) Previousstudies have suggested possible mechanisms for theeffect of MLCT on body fat accumulation, includingenhancement of diet-induced energy expenditure inhumans6) and activation of hepatic fatty acid oxidationin rats.10,11) However, although MLCT contain only one-eighth the amount of MCFA found in MCT, the reasonsMLCT cause significant lower body fat accumulationthan LCT are still unclear. Furthermore, the relationship

y To whom correspondence should be addressed. Tel: +81-46-837-2416; Fax: +81-46-837-2509; E-mail: h-shinohara@nisshin-oillio.com

Abbreviations: ACAD, acyl-CoA dehydrogenase; ACL, ATP-citrate lyase; ACO, acyl-CoA oxidase; CPT, carnitine palmitoyltransferase; FAS,

fatty acid synthase; G6PDH, glucose-6-phosphate dehydrogenase; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; LCAD, long-chain acyl-

CoA dehydrogenase; LCFA, long-chain fatty acids; LCT, long-chain triacylglycerols; MCAD, medium-chain acyl-CoA dehydrogenase; MCFA,

medium-chain fatty acids; MCT, medium-chain triacylglycerols; ME, malic enzyme; MLCT, medium- and long-chain triacylglycerols; PPAR�,peroxisome proliferator activated receptor-alpha; SREBP-1, sterol regulatory element binding protein-1

Biosci. Biotechnol. Biochem., 69 (10), 1811–1818, 2005

between changes in energy expenditure and fatty acidmetabolism with respect to the effect of MLCT on bodyfat accumulation is unknown.In the present study, we examined the relevance of

energy expenditure and fatty acid metabolism to knowmechanism(s) responsible for body fat reduction byMLCT. We measured 24-h energy expenditure inSprague-Dawley rats after 4 weeks of feeding dietscontaining MLCT, LCT, or MCT, and then measuredseveral parameters of fatty acid metabolism in the liver.

Materials and Methods

Materials. MCT (consisting of octanoate and dec-anolate), MLCT (consisting of MCFA and LCFA), andLCT (rapeseed oil) were from Nisshin OilliO (Tokyo).Their fatty acid compositions as measured by gaschromatography are shown in Table 1. Acetyl-CoA,malonyl-CoA, octanoyl-CoA, and palmitoyl-CoA werepurchased from Sigma-Aldrich Japan (Tokyo).

Animals and diets. All rats were treated in accordancewith the guidelines established by the Japanese Societyof Nutrition and Food Science (Law No. 105 andNotification No. 6 of the Japanese Government). MaleSprague-Dawley rats (6 weeks old) were obtained fromJapan SLC (Hamamatsu, Shizuoka, Japan). The ratswere housed individually in a room with controlledtemperature (20� to 24 �C), humidity (40% to 60%), andlighting (lights on from 0800 to 2000), and given acommercial standard diet (PicoLab Rodent Diet20 5053,PMI Feeds, St. Louis, MO) and tap water ad libitum for7 d. Fifteen rats were randomly divided into threegroups. Each group of rat was allowed free access toan experimental diet and water for 31 d. All diets wereformulated according to the AIN-93G diet,12) which wasmodified to contain sucrose at 40 g/kg of diet andpregelatinized instead of dextrinized cornstarch. Thecomposition of the experimental diets is shown inTable 2. At the end of the 4-week feeding period, therats were measured for 24-h energy expenditure. After3 d, the rats were weighed and killed by decapitation

under anesthesia with diethyl ether without fasting.Blood was collected in centrifuge tubes (VenojectII;Terumo, Tokyo). Serum was separated by centrifugationat 600� g for 15min and stored at �70 �C untilanalyzed. The liver was removed, weighed, and usedfor RNA analysis. The remaining portion of the liverwas frozen using liquid N2 and stored at �70 �C untilassayed. Carcass fat weight was measured using Soxhletextraction apparatus, and expressed as a percentage ofbody weight. Carcass protein was determined by theKjeldahl method.

Energy expenditure. Rats were placed in a plasticchamber and allowed test-diet and water ad libitum.Oxygen consumption and carbon dioxide productionwere measured for 26 h on a metabolism measuringsystem for small animals (MK-5000R; MuromachiKikai, Tokyo). Oxygen consumption and carbon dioxideproduction were measured every 3min, but the datafrom 0 to 2 h after placing the animals in a chamber wereexcluded to avoid variation due to restlessness in thenew environment. Total energy expenditure was calcu-lated according to the method reported previously.13)

Preparation of liver fraction and enzyme activitymeasurements. Fractions measuring fatty acid metabo-lism-related enzyme activities in rat liver were preparedas described previously,11) except that about 6 g of eachliver was homogenized. Carnitine palmitoyltransferase(CPT), acyl-CoA dehydrogenase (ACAD), acyl-CoAoxidase (ACO), fatty acid synthase (FAS), ATP-citratelyase (ACL), and glucose-6-phosphate dehydrogenase(G6PDH) were measured with a spectrophotometer.14–19)

ACAD activity was measured using octanoyl-CoA orpalmitoyl-CoA as a substrate. Activities were expressedas nanomoles per min per milligram protein. Proteinconcentration was determined with a BCA protein assaykit (Piece, France) using bovine serum albumin as astandard.

Table 1. Fatty Acid Compositions of Test Oils1

Fatty acid2 LCT3 MCT MLCT

(g/100 g total fatty acids)

8:0 ND4 74.4 9.4

10:0 ND 25.6 3.2

16:0 4.2 ND 3.7

18:0 2.0 ND 1.7

18:1 59.6 ND 51.5

18:2(n-6) 20.4 ND 18.6

18:3(n-3) 10.2 ND 9.0

Others 3.6 ND 2.9

1 Values determined by gas chromatography.2 Number of carbon atoms: number of double bonds.3 Rapeseed oil.4 Not detected.

Table 2. Compositions of Experimental Diets

IngredientsDietary group

LCT1 MCT MLCT

(g/kg)

Cornstarch 457.5 457.5 457.5

Casein 200.0 200.0 200.0

Pregelatinized cornstarch 132.0 132.0 132.0

Sucrose 40.0 40.0 40.0

LCT1 70.0 20.0 0.0

MCT 0.0 50.0 0.0

MLCT 0.0 0.0 70.0

Cellulose 50.0 50.0 50.0

Mineral mix2 35.0 35.0 35.0

Vitamin mix2 10.0 10.0 10.0

L-Cystine 3.0 3.0 3.0

Choline hydrogen tartrate 2.5 2.5 2.5

Tert-butylhydroquinone 0.014 0.014 0.014

1 Rapeseed oil.2 Based on AIN-93G mixture.

1812 H. SHINOHARA et al.

RNA extraction and RT-PCR analysis. Total RNAwas isolated from the liver using an RNeasy Mini Kit(Qiagen, Tokyo) according to the manufacturer’s in-structions. To remove genomic DNA, DNase I digestionwas performed with an RNase-Free DNase Set (Qiagen,Tokyo). Reverse transcriptional reaction was carried outusing a 1st Strand cDNA Synthesis Kit for RT-PCR(AMV) (Roche Diagnostics, Tokyo). Real-time quanti-tative PCR was performed using a fluorescence temper-ature cycler (LightCycler; Roche Diagnostics). Toconfirm the exactitude of RNA extraction and RT-PCR, all samples were subjected to PCR amplification,using a LightCycler-FastStart DNA Master SYBRGreen I kit (Roche Diagnostics). The primers used inthis study were designed on-line with Primer 3 software(http://frodo.wi.mit.edu/cgi-bin/primer3/primer3 www.cgi) and purchased from Sigma Genosys Japan(Hokkaido, Japan). The primer sequences used were asfollows: long-chain acyl-CoA dehydrogenase (LCAD)(sense, 50-CGCCCGATGTTCTCATTCT-30; antisense,50-GGCTTTCTCCCATTCTTCGT-30), medium-chainacyl-CoA dehydrogenase (MCAD) (sense, 50-CGCC-CCAGACTACGATAAAA-30; antisense, 50-CAAGA-CCACCACAACTCTCC-30), malic enzyme (ME)(sense, 50-CGGCACAGAAAATGAGGAGTT-30; anti-sense, 50-CCTTCTTGCTGCTCTTCGGAT-30), perox-isome proliferator activated receptor-alpha (PPAR�)(sense, 50-TGAAAACAAGGAGGCAGAGG-30; anti-sense, 50-AAGGAGGACAGCATCGTGAA-30), sterolregulatory element binding protein-1 (SREBP-1)(sense, 50-CAAAACCAGCCTCCCCAGA-30; anti-sense, 50-AGTCCCCATCCACGAAGAAA-30), andglyceraldehyde 3-phosphate dehydrogenase (GAPDH)(sense, 50-GTCGGTGTGAACGGATTTG-30; anti-sense, 50-GTGGTGAAGACGCCAGTAGA-30). Thesewere based on the mRNA sequences of a database (theaccession nos. are NM 012819 for LCAD, NM 016986for MCAD, NM 012600 for ME, NM 013196 forPPAR�, AF286470 for SREBP-1, and NM 017008for GAPDH). PCR-grade water instead of first strandDNA was used as a negative control. Melting curveanalysis was used to determine the specific PCRproducts. We used an internal standard curve basedon serial dilutions of first-strand DNA from the liver ofrats fed a commercial standard diet. The level oftranscripts for the constitutive housekeeping geneproduct GAPDH was measured in each sample, andthe data were normalized to the expression levels ofthis housekeeping gene. These were expressed asarbitrary units.

Analytical measurement. Serum triacylglycerol, cho-lesterol, and non-esterified free fatty acid were measuredwith a 7170 automated system (Hitachi, Tokyo) by anenzymatic method. Serum acetoacetic and 3-hydroxy-butyric acids were measured with a JCA-BM12 auto-mated system (JEOL, Tokyo). Liver lipids were ex-tracted with an ice-cold mixture of chloroform and

methanol (2:1, v/v), and triacylglycerol was analyzedusing Triglyceride E-test Wako (Wako Pure Chemicals,Osaka, Japan).

Statistical analysis. Results are expressed asmeans� SEM. The significance of the changes wasevaluated by analysis of variance, followed by Tukey’sHSD test for multiple comparisons. Significant dif-ferences between groups were identified using SPSS forWindows (version 10.0J; SPSS Japan, Tokyo). Resultswere considered significant at P < 0:05.

Results

After 4 weeks of consumption of the test diets, nosignificant differences were seen among the groups infinal body weight, body weight gain, or food intake, andhence, food efficacy (Table 3). Daily average intake ofMCFA was calculated to be 4.1 g/d/kg body weight inthe MCT group and 0.7 g/d/kg body weight in theMLCT group. As shown in Table 4, epididymal andmesenteric adipose tissue weights were significantlylower in the MLCT group than in the LCT group. On theother hand, although there was a tendency for adiposetissue weights to be lower in the MCT group than in theLCT group, the difference was not significant. Nochanges were observed in liver weights or carcassprotein contents among the groups. The hepatic triacyl-glycerol contents of MCT-fed rats were significantlyhigher than those of LCT-fed rats. The hepatic triacyl-glycerol contents of MLCT-fed rats showed a tendencyto be higher than those of LCT-fed rats, but thedifference was not significant.As shown in Table 5, MCT-fed rats had significantly

lower serum triacylglycerol concentrations than didLCT-fed rats. The triacylglycerol concentrations ofMLCT-fed rats showed a tendency to be lower thanthose of LCT-fed rats, but the difference was notsignificant. Even though MLCT contain MCFA, totalketone body concentrations in MLCT-fed rats weresignificantly lower than those in MCT-fed rats.To evaluate energy expenditure in rats fed the

experimental diets, we measured energy expenditureover 24 h in individual animals after 4 weeks on each

Table 3. Initial and Final Body Weights, Body Weight Gain, Food

Intake, and Food Efficiency of Rats Fed LCT, MCT, or MLCT for 4

Weeks1

Dietary group

LCT MCT MLCT

Initial body weight (g) 233� 3 230� 3 230� 3

Final body weight (g) 412� 9 403� 12 401� 9

Body weight gain (g) 180� 6 173� 12 172� 10

Food intake (g) 718� 12 725� 38 696� 17

Food efficiency (%)2 25:0� 0:7 23:8� 0:7 24:6� 0:9

1 All values are expressed as means� SEM.2 Body weight gain (g)/food intake (g) � 100 (%).

Effect of Structured Triacylglycerols in Rats 1813

diet, and calculated the area under the curve of energyexpenditure. As shown in Fig. 1, both MLCT- andMCT-fed rats showed significantly higher expenditurevalues than LCT-fed rats.

We used spectrophotometric methods to evaluateenzyme activity related to hepatic fatty acid metabolism.With regard to �-oxidation activity, the activity ofACAD measured using octanoyl-CoA as a substrate was

Table 4. Weights of Adipose Tissues and Liver, and Content of Liver Triacylglycerol and Carcass Protein of Rats Fed LCT, MCT, or MLCT for

4 Weeks1

Dietary group

LCT MCT MLCT

Epididymal adipose tissue (g/100 g BW)2 2:27� 0:15a 2:15� 0:11ab 1:64� 0:19b

Mesenteric adipose tissue (g/100 g BW) 2:34� 0:09a 2:16� 0:13ab 1:89� 0:04b

Perirenal adipose tissue (g/100 g BW) 3:18� 0:23 3:27� 0:28 2:65� 0:19

Carcass fat (g/100 g BW) 8:12� 0:36 7:27� 0:74 5:76� 0:73Liver weight (g/100 g BW) 4:50� 0:15 4:53� 0:14 4:47� 0:08

Liver triacylglycerol (mg/g liver) 20:3� 2:3a 29:8� 2:8b 24:6� 1:4ab

Carcass protein (g/100 g BW) 13:8� 0:6 14:0� 0:3 14:8� 0:4

1 All values are expressed as means� SEM. Data were analyzed by ANOVA, followed by Tukey’s HSD test for multiple comparisons. Values with different

superscript letters indicate significant difference, P < 0:05.2 All values are expressed as the percentage of final body weight.

Table 5. Concentrations of Serum Triacylglycerol, Non-Esterified Fatty Acid, Total Cholesterol, and Ketone Bodies of Rats Fed LCT, MCT, or

MLCT for 4 Weeks1

Dietary group

LCT MCT MLCT

Triacylglycerol (mg/dl) 322� 48a 193� 24b 238� 18ab

Non-esterified fatty acid (mEq/l) 0:424� 0:035 0:371� 0:031 0:400� 0:029Total cholesterol (mg/dl) 87:1� 2:7 88:3� 7:8 84:0� 5:9

Total ketone bodies (mmol/l) 115� 5ab 139� 12b 99:6� 4:6a

Acetoacetic acid (mmol/l) 47:2� 1:6a 60:0� 5:1b 40:2� 2:2a

3-Hydroxybutyric acid (mmol/l) 67:4� 4:0 78:8� 7:5 59:4� 2:5

1 All values are expressed as means� SEM. Data were analyzed by ANOVA, followed by Tukey’s HSD test for multiple comparisons. Values with different

superscript letters indicate significant difference, P < 0:05.

Fig. 1. Energy Expenditure in Rats Fed LCT, MCT, or MLCT for 4 Weeks.

Rats were fed 7% LCT as a control, 5% MCT plus 2% canola oil, or 7% MLCT for 4 weeks, and each animal was placed in a metabolic

chamber for 24-h respiratory gas analysis. Oxygen consumption and carbon dioxide production were measured for each session, and the energy

expenditure was calculated from these values. All values are expressed as means� SEM, n ¼ 5 rats. Data were analyzed by ANOVA, followed

by Tukey’s HSD test for multiple comparisons. LCT, open box; MCT, black box; MLCT, gray box. Bars with different letters indicate significant

difference, P < 0:05.

1814 H. SHINOHARA et al.

significantly higher in the MLCT group than in the LCTgroup. In contrast, the activity was comparable whenpalmitoyl-CoA was used as a substrate (Table 6). Nodifferences were observed in the activities of otherenzymes in the �-oxidation pathway. On the other hand,the activities of the lipogenic enzymes measured wereall markedly higher in the MCT group than in the others.

To examine the underlying mechanism(s) responsiblefor changes in fatty acid metabolism after feedingMLCT, we measured mRNA expression levels by real-time PCR (LightCycler). The data for each sample werenormalized to the expression level of the housekeepinggene GAPDH and expressed as arbitrary units. Theexpression of this housekeeping gene showed nosignificant differences in the respective parametersamong the groups (data not shown). As shown in

Fig. 2, among fatty acid oxidation genes, the mRNAlevel of LCAD was significantly higher in the MCT andMLCT groups than in the LCT group. No differenceswere observed in the mRNA levels of MCAD or PPAR�among the groups. The mRNA levels for the hepatic denovo lipogenic enzymes ACC and ME showed nodifferences between the MLCT and LCT groups. On theother hand, in the MCT group the hepatic de novolipogenic enzymes measured were significantly up-regulated compared with the other groups. SREBP-1,the nuclear transcription factor for lipogenesis, was alsoup-regulated.

Discussion

The purpose of this study was to examine therelevance of energy expenditure and hepatic fatty acidmetabolism to the body fat-lowering effect of MLCT.We observed a significantly lower body fat accumula-tion in MLCT than in LCT, while MCT showed anintermediate effect. Also, we observed a significantenhancement of energy expenditure in MLCT and MCTover that in LCT. With respect to fatty acid metabolism,MLCT but not MCT significantly activated ACAD asmeasured with an octanoyl-CoA substrate comparedwith LCT. In contrast, the activities of all the lipogenicenzymes measured were significantly enhanced in MCTcompared with the other oils. Although no enhancedactivities of fatty acid oxidation-related enzymes wereobserved in MCT-fed rats compared with LCT-fed rats,a significant elevation of serum ketone bodies was seen(Tables 5 and 6). These results suggest an enhancementin MCFA oxidation in the livers of MCT-fed rats. On theother hand, ACAD activity measured with an octanoyl-CoA substrate was significantly enhanced in MLCT-fed

Table 6. Activities of Hepatic Key Enzymes in Fatty Acid Metabolic

Pathways in Rats Fed LCT, MCT, or MLCT for 4 Weeks1

EnzymesDietary group

LCT MCT MLCT

(nmol/min/mg protein)

ACAD2

C8:0 14:3� 1:9a 19:7� 1:4ab 21:6� 1:7b

C16:0 15:8� 0:8 15:1� 0:7 14:6� 0:7

CPT 0:49� 0:11 0:32� 0:05 0:46� 0:11

ACO 0:90� 0:09 0:63� 0:04 0:74� 0:08FAS 6:72� 0:87a 9:74� 0:64b 7:41� 0:23ab

ACL 14:1� 1:7a 23:2� 0:9b 16:7� 1:1a

G6PDH 26:0� 3:6a 48:2� 2:4b 26:3� 2:8a

1 All values are expressed as means� SEM. Data were analyzed by

ANOVA, followed by Tukey’s HSD test for multiple comparisons. Values

with different superscript letters indicate significant difference, P < 0:05.2 ACAD activity was measured using octanoyl-CoA (C8:0) or palmitoyl-

CoA (C16:0) as a substrate.

Fig. 2. Messenger RNA Expression Levels in Rats Fed LCT, MCT, or MLCT for 4 Weeks.

Rats were fed 7% LCT as a control, 5% MCT plus 2% canola oil, or 7% MLCT for 4 weeks. At the end of 4 weeks of feeding, mRNA

expression levels of enzymes related to fatty acid metabolism were measured by real-time PCR (LightCycler). Values from rats fed LCT (open

boxes), MCT (black boxes), or MLCT (gray boxes) were normalized to the expression levels of housekeeping gene GAPDH, and expressed as

arbitrary units. All values are expressed as means� SEM, n ¼ 5 rats. Data were analyzed by ANOVA, followed by Tukey’s HSD test for

multiple comparisons. Bars with different letters indicate significant difference, P < 0:05.

Effect of Structured Triacylglycerols in Rats 1815

rats compared with LCT-fed rats (Table 6). The resultsthus suggest that MCFA are subjected to fatty acidoxidation more than LCFA in the livers of MLCT-fedrats. The mRNA expression levels of MCAD andPPAR� showed no differences among the groups(Fig. 2). LCFA, especially polyunsaturated fatty acids,are known to serve as ligands for PPAR�, whereasMCFA are poor activators,20) and PPAR� is known toregulate the mRNA expression level of MCAD.21) Theseobservations suggest that after intake of MCT or MLCT,the unresponsiveness of the mRNA expression level ofMCAD might be the result of unresponsiveness of thePPAR� gene. LCAD is the first-step enzyme inmitochondrial �-oxidation of C6–C18 fatty acids,22) andits mRNA expression level is known to be under theregulation of PPAR�.21) In the present study, the mRNAexpression level of LCAD was significantly higher bothin MLCT- and MCT-fed rats than in LCT-fed rats,although ACAD activity measured using palmitoyl-CoAas a substrate showed no difference among the groups(Fig. 2, Table 6). The reason for this discrepancy ispresently unclear.MLCT is not a physically mixed oil made of MCFA

and LCFA but rather randomly interesterified triacyl-glycerols in the same glycerol molecule. We cannotcompletely rule out the possibility that the body fat-lowering effect of MLCT reflects their structuralcharacteristics. Some studies have indicated that intes-tinal absorption after intake of the test oil variesdepending on whether the oil was structured.23,24)

However, in our previous study, a physical mixture ofLCT and MCT, which contained the same amount ofMCFA found in MLCT, showed a significantly lowerbody fat accumulation than LCT, and that the adiposetissue weight of rats fed MLCT and the physical mixtureof LCT and MCT were not significantly different.8) Wesuppose that the effect of MLCT on body fat accumu-lation might not be determined by the structure ofMCFA and LCFA within triacylglycerol but rather bythe amount of MCFA ingested.Many studies have reported effects of MCT on body

fat accumulation,25–28) suggesting that daily intake ofMCT can reduce body fat accumulation and that it mightbe effective for preventing obesity and lifestyle-relateddiseases. However, in the present study, although MCT-fed rats exhibited a tendency for adipose tissue weight tobe lower than that of LCT-fed rats, the difference wasnot significant (Table 3). The body fat accumulation ofMLCT-fed rats was less than that of LCT-fed rats, eventhough the amount of MCFA consumed in the MLCT-fed rats was less than that in the MCT-fed rats (Table 3).The experimental diets used in previous studies weregenerally composed of high MCT (45 to 60% of totalenergy) and low carbohydrate (10 to 40% of totalenergy). On the other hand, the diet used in the presentstudy was based on AIN-93G, which is composed of lowMCT (11% of total energy) and high carbohydrate (64%of total energy) in the MCT group. The vitamins and

minerals were also different from our AIN-93G standarddiets. Dietary composition has been shown to influencebody weight, body fat, and whole-body energetics.29) Itis thus likely that one cause of the discrepancies is thedifference in the compositions of the experimental diets.

Metabolic processes such as lipogenesis, gluconeo-genesis, and ketogenesis are known to influence thecalculation of rates of carbohydrate and fat oxidationfrom measurements of oxygen consumption and carbondioxide production.30) Hill et al.31) observed in humansan increase in energy expenditure and fasting totalserum triglycerides and a decrease in total lipidoxidation on day 6 as compared with day 1 of anMCT overfeeding week. They reported that the meas-ured energy expenditure on day 1 might be due toincreased formation and oxidation of ketone bodies, andthat the much larger energy expenditure on day 6 thanon day 1 most likely reflects increased de novo fatty acidsynthesis. Hence they concluded that enhanced energyexpenditure after intake of MCT was probably due toenhanced lipogenesis in the liver. Accordingly, theenhancement of energy expenditure observed in MCT-fed rats in the present study might be due to enhanced denovo lipogenesis and ketogenesis in the liver. On theother hand, unlike MCT, MLCT showed significantenhancement of energy expenditure without enhancingde novo lipogenesis and ketogenesis, suggesting thepossibility that enhancement of fatty acid oxidation inother tissues besides the liver might well be involved.Furthermore, no significant difference in body fataccumulation was observed in MCT-fed rats, in spiteof apparent enhancement of energy expenditure ascompared with LCT-fed rats (Table 3, Fig. 1). St-Ongeet al.32) reported that long-term consumption of MCTsignificantly enhanced energy expenditure and fatoxidation as compared with LCT, but resulted in onlya small, insignificant difference in body composition inoverweight women. The body fat-lowering effect ofMCFA is thus difficult to explain by an enhancement ofenergy expenditure alone.

Daval et al.,33) in a study using chickens, clarified thecontribution of enhanced de novo lipogenesis to body fataccumulation. Hwang et al.34) reported that fat deposi-tion after MCT feeding is accomplished not by incor-poration of dietary fatty acids but by de novo fatty acidsynthesis. Other reports have indicated the ability ofMCT to enhance de novo lipogenesis in the liver.35–38) Itis thought that the magnitude of the enhancement inlipogenesis by MCFA might be one reason for theinconsistency in demonstrating a body fat-loweringeffect of MCT. Its effect appears to depend onexperimental conditions and the characteristics of sub-jects. In the present study, liver triacylglycerol contentswere significantly higher in MCT-fed rats than in ratsfed other oils (Table 4). Excess acetyl-CoA produced by�-oxidation of MCFA was used in the synthesis ofLCFA without being oxidized in liver mitochondria.39)

MCT-fed rats showed significantly higher hepatic de

1816 H. SHINOHARA et al.

novo lipogenesis than did LCT-fed rats (Table 6),indicating that excessive intake of MCFA tends toinduce fatty liver.

The correlation between MCFA intake and otherfactors such as the effect on body fat accumulation,energy expenditure, fatty acid oxidation, fatty acidsynthesis, and energy intake requires further study,40)

but we propose that high intake of MCFA does notnecessarily produce a beneficial effect on body fat mass.The results of our previous studies in humans supportthis.9,41) Further studies are in progress in our laboratory.

In summary, this study indicates that the ability ofMLCT to induce lower body fat accumulation relative toLCT depends not only on enhancement of total energyexpenditure, but also on a failure to enhance de novolipogenic capacity. Oil containing MCFA is more easilyoxidized than oil containing LCFA, and enhancesenergy expenditure, but excessive intake of MCFA alsoappears to activate de novo lipogenesis, erasing theirfavorable body fat-lowering effect.

Acknowledgments

We thank Kaori Yabuki and Naohisa Nosaka of theHealthcare Science Research Laboratory of our com-pany for their technical assistance. We further thankDr. Michihiro Sugano, Professor Emeritus of KyushuUniversity, for his valuable advice in revision of themanuscript.

References

1) Kaunitz, H., Slanetz, C. A., Johnson, R. E., Babayan, V.K., and Barsky, G., Nutritional properties of thetriglycerides of saturated fatty acids of medium chain-length. J. Am. Oil Chemists’ Soc., 35, 10–13 (1958).

2) Kaunitz, H., Slanetz, C. A., Johnson, R. E., Babayan, V.K., and Barsky, G., Relation of saturated, medium- andlong-chain triglycerides to growth, appetite, thirst andweight maintenance requirements. J. Nutr., 64, 513–524(1958).

3) Freund, G., and Weinsier, R. L., Standardized ketosis inman following medium chain triglyceride ingestion.Metabolism, 15, 980–991 (1966).

4) Bach, A. C., and Babayan, V. K., Medium-chaintriglycerides: an update. Am. J. Clin. Nutr., 36, 950–962 (1982).

5) Negishi, S., Shirasawa, S., Arai, Y., Suzuki, J., andMukataka, S., Activation of powdered lipase by clusterwater and the use of lipase powders for commercialesterification of food oils. Enzyme Microb. Technol., 32,66–70 (2003).

6) Matsuo, T., Matsuo, M., Kasai, M., and Takeuchi, H.,Effects of a liquid diet supplement containing structuredmedium- and long-chain triacylglycerols on body fataccumulation in healthy young subjects. Asia Pacific J.Clin. Nutr., 10, 46–50 (2001).

7) Takeuchi, H., Kubota, F., Itakura, M., and Taguchi, N.,Effect of triacylglycerols containing medium- and long-chain fatty acids on body fat accumulation in rats. J.

Nutr. Sci. Vitaminol., 47, 267–269 (2001).8) Noguchi, O., Shimada, H., Kubota, F., Tsuji, H., and

Aoyama, T., Nutritional effects of randomly interesteri-fied and physically mixed oils containing medium-chainfatty acids on rats. J. Oleo Sci., 51, 699–703 (2002).

9) Kasai, M., Nosaka, N., Maki, H., Negishi, S., Aoyama,T., Nakamura, M., Suzuki, Y., Tsuji, H., Uto, H.,Okazaki, M., and Kondo, K., Effect of dietary medium-and long-chain triacylglycerols (MLCT) on accumula-tion of body fat in healthy humans. Asia Pacific J. Clin.Nutr., 12, 151–160 (2003).

10) Matsuo, T., and Takeuchi, H., Effects of structuredmedium- and long-chain triacylglycerols in diets withvarious levels of fat on body fat accumulation in rats. Br.J. Nutr., 91, 219–225 (2004).

11) Shinohara, H., Shimada, H., Noguchi, O., Kubota, F.,and Aoyama, T., Effect of medium-chain fatty acids-containing dietary oil on hepatic fatty acid oxidationenzyme activity in rats. J. Oleo Sci., 51, 621–626 (2002).

12) Reeves, P. G., Nielsen, F. H., and Fahey, G. C., AIN-93purified diets for laboratory rodents: final report of theAmerican Institute of Nutrition ad hoc writing commit-tee on the reformulation of the AIN-76A rodent diet.J. Nutr., 123, 1939–1951 (1993).

13) Weir, J. B. V., New methods for calculating metabolicrate with special reference to protein metabolism.J. Physiol., 109, 1–9 (1949).

14) Markwell, M. A. K., McGroarty, E. J., Bieber, L. L., andTolbert, N. E., The subcellular distribution of carnitineacyltransferases in mammalian liver and kidney: a newperoxisomal enzyme. J. Biol. Chem., 248, 3426–3432(1973).

15) Hashimoto, T., Miyazawa, S., Gunarso, D., and Furuta,S., alpha-Amanitin inhibits the oxidation of long chainfatty acids in mouse liver. J. Biochem., 90, 415–421(1981).

16) Ide, T., Murata, M., and Sugano, M., Stimulation of theactivities of hepatic fatty acid oxidation enzymes bydietary fat rich in �-linolenic acid in rats. J. Lipid Res.,37, 448–463 (1996).

17) Takeda, Y., Suzuki, F., and Inoue, H., ATP citrate lyase(citrate-cleavage enzyme). Methods Enzymol., 13, 153–160 (1969).

18) Kelley, D. S., Nelson, G. J., and Hunt, J. E., Effect ofprior nutritional status on the activity of lipogenicenzymes in primary monolayer cultures of rat hepato-cytes. Biochem. J., 235, 87–90 (1986).

19) Kelley, D. S., and Kletzien, R. F., Ethanol modulation ofthe hormonal and nutritional regulation of glucose 6-phosphate dehydrogenase activity in primary cultures ofrat hepatocytes. Biochem. J., 217, 543–549 (1984).

20) Forman, B. M., Chen, J., and Evans, R. M., Hypolipi-demic drugs, polyunsaturated fatty acids, and eicosa-noids are ligands for peroxisome proliferator-activatedreceptors � and �. Proc. Natl. Acad. Sci. U.S.A., 94,4312–4317 (1997).

21) Mandard, S., Muller, M., and Kersten, S., Peroxisomeproliferator-activated receptor � target genes. Cell. Mol.Life Sci., 61, 393–416 (2004).

22) Ghisla, S., and Thorpe, C., Acyl-CoA dehydrogenases: amechanistic overview. Eur. J. Biochem., 271, 494–508(2004).

23) Tso, P., Lee, T., and Demichele, S. J., Lymphatic

Effect of Structured Triacylglycerols in Rats 1817

absorption of structured triglycerides vs. physical mix ina rat model of fat malabsorption. Am. J. Physiol., 277(Gastrointest. Liver Physiol., 40), G333–G340 (1999).

24) Mu, H., and Hoy, C.-E., Effects of different medium-chain fatty acids on intestinal absorption of structuredtriacylglycerols. Lipids, 35, 83–89 (2000).

25) Bray, G. A., Lee, M., and Bray, T. L., Weight gain ofrats fed medium-chain triglycerides is less than rats fedlong-chain triglycerides. Int. J. Obes., 4, 27–32 (1980).

26) Geliebter, A., Torbay, N., Bracco, E. F., Hashim, S. A.,and Van Itallie, T. B., Overfeeding with medium-chaintriglyceride diet results in diminished deposition of fat.Am. J. Clin. Nutr., 37, 1–4 (1983).

27) Lavau, M., Fornari, V., and Hashim, S. A., Ketonemetabolism in brain slices from rats with diet inducedhyperketonemia. J. Nutr., 108, 621–639 (1978).

28) Hashim, S. A., and Tantibhedyangkul, P., Medium chaintriglyceride in early life: effects on growth of adiposetissue. Lipids, 22, 429–434 (1987).

29) Dulloo, A. G., and Girardier, L., Influence of dietarycomposition on energy expenditure during recovery ofbody weight in the rat: implications for catch-up growthand obesity relapse. Metabolism, 41, 1336–1342 (1992).

30) Frayn, K. N., Calculation of substrate oxidation rates invivo from gaseous exchange. J. Appl. Physiol., 55, 628–634 (1983).

31) Hill, J. O., Peters, J. C., Yang, D., Sharp, T., Kaler, M.,Abumrad, N. N., and Greene, H. L., Thermogenesis inhumans during overfeeding with medium-chain trigly-cerides. Metabolism, 38, 641–648 (1989).

32) St-Onge, M.-P., Bourque, C., Jones, P. J. H., Ross, R.,and Parsons, W. E., Medium- versus long-chain trigly-cerides for 27 days increases fat oxidation and energyexpenditure without resulting in changes in bodycomposition in overweight women. Int. J. Obes., 27,95–102 (2003).

33) Daval, S., Lagarrigue, S., and Douaire, M., MessengerRNA levels and transcription rates of hepatic lipogenesisgenes in genetically lean and fat chickens. Genet. Sel.Evol., 32, 521–531 (2000).

34) Hwang, S. G., Yano, H., and Kawashima, R., Theinfluence of dietary medium and long chain triglycerideson growth performances and fat deposition in growingrats. J. Nutr. Sci. Vitaminol., 38, 127–139 (1992).

35) Takase, S., and Hosoya, N., Effect of dietary mediumchain triglyceride on lipogenic enzyme activity in ratliver. J. Nutr. Sci. Vitaminol., 32, 219–227 (1986).

36) Takase, S., and Hosoya, N., Possible role of insulinstatus in the increased lipogenic enzyme activity bydietary medium-chain triglyceride in rat liver. J. Nutr.Sci. Vitaminol., 33, 177–184 (1987).

37) Lavau, M. M., and Hashim, S. A., Effect of mediumchain triglyceride on lipogenesis and body fat in the rat.J. Nutr., 108, 613–620 (1978).

38) Hwang, S. G., Yano, H., and Kawashima, R., Influenceof dietary medium- and long-chain triglycerides on fatdeposition and lipogenic enzyme activities in rats. J. Am.Coll. Nutr., 12, 643–650 (1993).

39) Papamandjaris, A. A., MacDougall, D. E., and Jones, P.J. H., Medium chain fatty acid metabolism and energyexpenditure: obesity treatment implications. Life Sci., 62,1203–1215 (1998).

40) Bach, A. C., Ingenbleek, Y., and Frey, A., The useful-ness of dietary medium-chain triglycerides in bodyweight control: fact or fancy? J. Lipid Res., 37, 708–726(1996).

41) Tsuji, H., Kasai, M., Takeuchi, H., Nakamura, M.,Okazaki, M., and Kondo, K., Dietary medium-chaintriacylglycerols suppress accumulation of body fat in adouble-blind, controlled trial in healthy men and women.J. Nutr., 131, 2853–2859 (2001).

1818 H. SHINOHARA et al.