comparison of hypocholesterolemic activity of tea seed oil with commonly used vegetable oils in...

jfbc_425 859..876

COMPARISON OF HYPOCHOLESTEROLEMIC ACTIVITY OFTEA SEED OIL WITH COMMONLY USED VEGETABLE OILS

IN HAMSTERS

LEI GUAN1, HAU YIN CHUNG1,2,3 and ZHEN YU CHEN2,3

1Department of Biology

2Food and Nutritional Sciences ProgrammeThe Chinese University of Hong Kong

Shatin, N.T., Hong Kong, China

Accepted for Publication August 15, 2009

ABSTRACT

Interest in tea seed oil (named tea oil) as a cooking oil is increasing.However, its effect on blood cholesterol is not known. This study was thereforeconducted to compare the hypocholesterolemic activity of tea oil with grapeseed, canola and corn oils. Results showed that plasma total cholesterol (TC),non-high-density lipoprotein-cholesterol (non-HDL-C) and triacylglycerols(TG) in hamsters fed with a 0.1% cholesterol diet containing tea, grape,canola or corn oil were significantly reduced compared with those in lard-fedgroup. Tea oil decreased only non-HDL-C and had no or little effect onHDL-C concentration, while grape oil reduced both. Unlike grape oil, tea oilup-regulated sterol regulatory element binding protein (SREBP-2) and low-density lipoprotein receptor. Besides, tea oil-fed hamsters excreted less neutralbut greater acidic sterols compared with other three oils. Differences betweentea oil and the tested vegetable oils could be attributable partially to thegreatest oleic acid (>80%) and least phytosterol content in tea oil.

PRACTICAL APPLICATIONS

Although tea leaves are used worldwide, tea oil is only used in someAsian countries. Extensive research has shown the health benefits of teadrinking. However, benefit associated with the consumption of tea seed oilremains unclear. We have shown that tea seed oils were able to lower plasmacholesterol equally as grape, canola and corn oil in hamsters fed a 0.1%

3 Corresponding author. TEL: +(852) 26096149; Fax: +(852) 26035745; EMAIL: [email protected] or [email protected]

DOI: 10.1111/j.1745-4514.2010.00425.x

Journal of Food Biochemistry 35 (2011) 859–876.© 2011 Wiley Periodicals, Inc. 859

cholesterol diet. The hypocholesterolemic activity of tea oil was characterizedby decreasing only non-high-density lipoprotein-cholesterol (non-HDL-C)and having no effect on HDL-C concentration. Most importantly, tea oil couldup-regulate sterol regulatory element binding protein 2 and low-density lipo-protein receptor. It was further demonstrated that hamsters fed the tea oil dietexcreted lesser neutral but greater acidic sterols compared with other threevegetable oils. These results suggest that the tea seed oil could be an alternativehealthy oil for human consumption.

INTRODUCTION

Tea leaves (Camellia sinensis) are popularly used to brew different bev-erages worldwide, whereas tea seed has been utilized to produce cooking oil insome Asian countries. Extensive research has linked a wide range of healthbenefits to drinking tea including lowering the risk of certain cancers and heartdisease as well as weight loss and protection against Alzheimers (Stangl et al.2006; Chen et al. 2008). However, benefit associated with the tea seed oil(known as tea oil) remains relatively unexplored except for the two recentreports showing that tea oil possessed the antioxidant activity (Fazel et al.2009) and anti-lipogenic activity (Kim et al. 2008). Tea oil has been utilized asa cooking oil for more than 1,000 years in China, particularly in SouthernChina such as Hunan province where more than 50% of the cooking oil is teaoil (Shanan and Ying 1982; Xia et al. 1993). Tea oil is chemically stablebecause it is rich in oleic acid (18:1n-9).

Accumulating evidence has demonstrated that grape seed and winepolyphenols are able to scavenge the free-radicals, lower plasma cholesterollevel, and reduce the risk of cardiovascular disease and cancer (Castilla et al.2006). Grape seed oil (called grape oil) has been long used for salad dressings,marinades, deep frying, flavored oils, baking, massage oil, sunburn repairlotion, and hand creams in all over the world. Chemically, grape oil is knownfor abundance in both linoleic acid (18:2n-6) and oleic acid. However, nutri-tional properties of grape oil remains unknown compared with other vegetableoils.

Blood total cholesterol (TC) and low-density lipoprotein cholesterol(LDL-C) correlate directly with the risk of heart diseases, whereas high-density lipoprotein cholesterol (HDL-C) correlates inversely with risk. Dietaryfatty acids affect plasma cholesterol level largely mediated by their interactionwith two transcriptional factors, sterol regulatory element binding protein-2(SREBP-2) and liver X receptor (LXR-a) (Rotheblat et al. 2002; Eberle et al.2004; Kastelein 2007). SREBP-2 governs the transcription of LDL receptor(LDL-R) and 3-hydroxy-3-methylglutary-CoA (HMG-CoA) reductase

860 L. GUAN ET AL.

(HMGR). LDL-R is responsible for the removal of LDL-C from the circula-tion, whereas HMGR is a key enzyme in cholesterol synthesis in the liver. Onthe other hand, LXR-a regulates the transcription of CYP7A1-encoding cho-lesterol 7a-hydroxylase, which is a rate-limiting enzyme in conversion ofcholesterol to bile acids in the liver, and is responsible for elimination ofexcessive cholesterol in the bile fluid. No study to date has characterized theinteraction of tea and grape oil with these transcriptional factors, receptor andenzymes that are involved in cholesterol metabolism.

Despite extensive research on vegetable oils, little is known on how teaand grape oils affect blood cholesterol and how they interact with these genesand proteins involved in cholesterol metabolism in vivo. The present study wastherefore undertaken to: (1) use hamsters as a model to examine the relativehypocholesterolemic activity of tea oil compared with grape oil, canola oil andcorn oil with lard being a reference animal fat; and (2) characterize theinteraction of dietary tea oil with SREBP-2, LXR, HMGR, LDLR andCYP7A1.

MATERIALS AND METHODS

Animals and Diets

Five diets were prepared as previously described by Lam et al. (2008).The basal diet was formulated by mixing the following ingredients per kilo-gram diet (g): corn starch, 500; casein 200; sucrose, 100; AIN-76 mineral mix,40; AIN-76 A vitamin mix, 20; gelatin, 20; DL-methionine, 1; cholesterol, 1.The five experimental diets were prepared by adding 10% lard, 10% tea oil,10% grape oil, 10% canola oil and 10% corn oil, respectively, into the basaldiet (Table 1).

Male Golden Syrian hamsters (n = 50, 5 wk, 100–120 g) were obtainedfrom the Laboratory Animal Services Centre, The Chinese University of HongKong. Experiments were approved and conducted in accordance with theguidelines set by the Animal Experimental Ethical Committee, The ChineseUniversity of Hong Kong. Hamsters were housed (two hamsters per cage) inwire-bottomed cages in an animal room at 25C with 12:12-h light-dark cycles.In brief, hamsters were allowed free access to a standard cereal-based chowdiet (PicoLab Rodent Diet 20-Lab Diet, Australia) for a 2-week acclimationperiod. Afterwards, hamsters were weighed, ear-punched and randomlydivided into five groups (n = 10 each) maintained on one of the five diets fora period of 12 weeks. Body weight was recorded once a week, and total fecaloutput was collected, and food consumption was recorded every 2 days. Bloodwas collected from the retro-orbital sinus into a heparinized capillary tube

861HYPOCHOLESTEROLEMIC ACTIVITY OF TEA SEED OIL

under light anesthetization, using a mixture of ketamine, xylazine and saline(v/v/v, 4:1:5) after a 16-h food deprivation at week 0, 4, 8 and 12. At the endof 12 weeks, all the hamsters were sacrificed after overnight fasting. Bloodwas collected via the abdominal aorta. The liver, heart, kidney, testis, perirenalfat and epididymal fat were also removed, rinsed with ice-cold saline,weighted, flash frozen in liquid nitrogen and stored at -80C until analysis.

Determination of Fatty Acid Composition and Phytosterol Contentin Oils

To determine the fatty acid composition of lard, tea oil, canola oil andgrape oil, acid-catalyzed methylation was used in the present study. In brief,each sample (25 mg) were transesterified in 2 mL of boron trifluoride inmethanol (14%, Sigma, St. Louis, MO) under N2. The methylation tube wasplaced in a heat block at 80C for 1 h and then cooled to room temperature.Hexane (4 mL) and distilled water (1 mL) were then added and mixed thor-oughly. After centrifugation, the top hexane layer containing fatty-acid methylesters (FAME) was saved. The FAME mixtures were analyzed on a flexiblesilica capillary column (HP-INNOVAX, 30 m ¥ 0.32 mm inner diameter [i.d.],Agilent Technologies, Santa Clara, CA) in a SHIMADZU GC-2010 gas-liquidchromatograph (GC) equipped with a flame-ionization detector and an auto-mated injector (Shimadzu, Tokyo, Japan). Column temperature was pro-grammed from 120 to 220C at a rate of 2C/min and then held for 20 min.Injector and detector temperature was set at 250C and 270C, respectively.Helium was used as the carrier gas at a head pressure of 105 kPa.

To determine the phytosterol composition in each oil/fat, 100 mg of thesample was subjected to saponification in 5 mL of NaOH in 95% ethanol. The

TABLE 1.COMPOSITION OF FIVE HIGH FAT DIETS CONTAINING

0.1% CHOLESTEROL

Main nutrients g/kg

Corn starch 500Casein 200Sucrose 100Fat* 100Cholesterol 1Vitamin mixture 20Mineral mixture 40Gelatin 20Methionine 1

*Refer to 100 g of lard or one of four vegetable oils.

862 L. GUAN ET AL.

unsaponifiable matters were extracted into cyclohexane and derivatized intothe trimethylsilyl (TMS)-esters in a commercial TMS-reagent (Sil-A regent,Sigma, St. Louis, MO). The TMS-ether derivatives were dissolved into 400 mLhexane with 1 mL being injected onto a fused silica capillary column(SACTM-5, 30 m ¥ 0.25 mm, i.d.; Supelco, Inc., Bellefonte, PA) in a SHI-MADZU GC-2010 gas-liquid chromatograph equipped with a flame-ionization detector and an automated injector (Shimadzu). Helium was used asa carrier gas at a constant flow of 1.0 mL/min. Column temperature wasprogrammed from 90 to 270C at a rate of 30C/min, and then to 300C at1.5C/min, where it was held for 12 min. Individual phytosterol was quantifiedaccording to the amount of 5a-cholestane added.

Plasma Lipid and Lipoprotein Determinations

Plasma TC and total triacyglycerol (TG) levels were determined usingenzymatic kits from Infinity (Waltham, MA) and Stanbio Laboratories(Boerne, TX), respectively. The concentration of HDL-C was measured afterprecipitation of LDL and very low-density lipoprotein (VLDL) with phospho-tungstic acid and magnesium chloride, using a commercial kit (Stanbio Labo-ratories). Non-HDL-C was calculated from the difference between TC andHDL-C.

Determination of Organ Cholesterol

Cholesterol in organs was determined using a method described as pre-viously described (Chan et al. 1999). The liver (100 mg) and heart (300 mg)were used to determine the cholesterol level. In brief, the tissue sample and1 mg stigmastanol, as an internal standard, were homogenized in 15 mLchloroform-methanol (2:1, v/v) and 3 mL saline. The chloroform-methanolphase was removed and dried down under a nitrogen steam. After 1 h mildhydrolysis with 5 mL of 1 N NaOH in 90% ethanol at 90C, 1 mL of water and6 mL of cyclohexane were added for extraction of cholesterol. The cyclohex-ane phase was evaporated to dryness under nitrogen, and cholesterol wasconverted to its TMS-ether derivative. The TMS-ether derivative was dissolvedin hexane for gas chromatography (GC) analysis.

Determination of Fecal Neutral and Acidic Sterols

Neutral and acidic sterols in the feces of the hamsters were determined asdescribed previously with slight modifications (Chan et al. 1999). 300 mgdried fecal samples were mildly hydrolyzed with 1 N NaOH in 90% ethanol at90C for 1 h. Then, total neutral sterols were extracted with cyclohexane, andconverted into their trimethylsilyl derivatives. The acidic sterol-containing


lower aqueous phase were saponified and converted into their trimethylsilylderivatives. The two trimethylsilyl derivatives were separately analyzed withGC.

Western Blotting Analysis of Liver SREBP-2, LDL-R, HMGR, LXRand CYP7A1

Liver protein was extracted according to the method described previouslyby Vaziri and Liang with some modification (Vaziri and Liang 1996). In brief,the liver was homogenized in a homogenizing buffer containing 20 mM Tris–HCL (pH 7.5), 2 mM MgCL2, 0.2 M sucrose and Complete protease inhibitorcocktail (Roche, Mannheim, Germany). The extract was centrifuged at13,000 ¥ g for 15 min at 4C and the supernatant was collected (total protein).The total protein was centrifuged at 126,000 ¥ g for 60 min at 4C. The pelletwas resuspended in the same homogenizing buffer.

The pellet protein was separated by electrophoresis in a 7% sodiumdodecyl sulfate polyacrylamide gel electrophoresis gel and transferred to poly-vinylidene difluoride membranes (Millipore, Billerica, MA) using a semi-drytransfer system. Membranes were then blocked in 5% nonfat milk Tris-buffered saline with Tween-20 for 1 h and overnight at 4C in the same solutioncontaining 1:600 anti-LDL-R antibody (Santa Cruz Biotechnology, Inc., SantaCruz, CA), 1:500 anti-HMGR (Upstate USA Inc., Lake Placid, NY), 1:200anti-CYP7A1 (Santa Cruz Biotechnology, Inc.), 1:400 anti-LXR antibody oranti-SREBP-2 antibody (Santa Cruz Biotechnology). The membrane was thenincubated for 1 h at 4C in diluted (1:3,000) horseradish peroxidase-linked goatanti-rabbit IgG (Santa Cruz Biotechnology, Inc.), donkey anti-rabbit IgG(Santa Cruz Biotechnology, Inc.) or goat anti-mouse IgG (Calbiochem, EMDChemicals, Inc., San Diego, CA). Then, membranes were developed withenhanced chemiluminescence agent (Santa Cruz Biotechnology, Inc.) andwere subjected to autoradiography on SuperRX medical X-ray film (Fuji,Tokyo, Japan). Densitometry was quantified using the Bio-Rad Quantity onesoftware (Bio-Rad Laboratories, Hercules, CA). Data on abundance of LDLR,HMG-R, CYP7A1, LXR and SREBP-2 were normalized with b-actin (SantaCruz Biotechnology, Inc.).

Statistics

Data are expressed as mean � standard deviation. Treatment effects werestatistically analyzed among groups using one-way analysis of variance andpost hoc Least Significant Difference test on SigmaStat Advisory StatisticalSoftware (SigmaStat version 16.0, SPSS Inc., Chicago, IL). P-value less than0.05 are regarded statistically significant. Finally, regression analysis was

864 L. GUAN ET AL.

performed and Pearson’s correlation coefficients were determined to evaluatethe relationship between LXR protein abundance and plasma and hepatic lipidlevels.

RESULTS

Fatty Acid and Phytosterol Composition

Fatty acid composition is different among the five diets (Table 2). Lardhas the greatest, while canola oil has the least amount of palmitic acid (16:0)among the five diets. In contrast, tea oil has the greatest amount of oleic acid(18:2n-9), while the grape oil has the least. Regarding linoleic acid (18:2n-6),corn oil has the greatest, while tea oil has the least. However, tea oil has theleast a-linolenic acid (18:3n-3), whereas canola oil has the most.

Four vegetable oils contained the varying amount of phytosterols withcorn oil having the most followed by canola oil, grape oil and tea oil in adecreasing order. In contrast, there was no phytosterols detected in lard(Table 2).

Food Intake, Body and Organ Weights

No difference was seen in mean food consumption among the five groups(Table 3). It was found that dietary fat had no effect on the final body weight

TABLE 2.FATTY ACID COMPOSITION (g/100 g OIL) AND PHYTOSTEROL CONTENT (mg/100 g OIL)

IN FIVE OILS OR FAT

Lard Tea oil Grape oil Canola oil Corn oil

14:0 1.95 0 0 0 016:0 25.54 7.60 8.24 4.00 11.6418:0 14.41 2.09 3.64 1.97 1.9818:1n-9 43.16 77.36 28.57 55.60 29.0818:2n-6 10.69 7.51 49.54 20.41 50.2218:3n-3 0.51 0.11 2.50 9.34 0.94M:S 1.08 7.98 2.41 9.32 2.14P:S 0.27 0.79 4.38 4.99 3.76Campesterol 0 22.82 29.66 187.43 120.51Stigmasterol 0 10.31 20.95 ND 43.66b-sitosterol 0 32.37 118.15 254.36 425.57Total 0 65.50 168.76 441.78 589.74

M:S, monounsaturated: saturated fatty acids; ND, no detected; P:S, polyunsaturated: saturated fattyacids.


of the five groups. Liver, kidney, heart and epididymal fat pad weights were notsignificantly different among the five groups except that corn oil-fed hamstershad significant lower relative liver weight than the lard-fed group, and thehearts in canola oil-fed and corn oil-fed groups weighed lesser than the otherthree groups.

Plasma Lipoprotein Concentrations

Plasma TC increased in all five groups compared with the baseline values(Table 4). Compared with that of lard group, plasma TC in the four vegetableoil groups was significantly reduced at the end of week 12. No significantdifference in plasma TC was observed among the four vegetable oil groups atthe end of week 12 (Table 4), though grape oil and canola oil groups appearedto have lower plasma TC than tea oil and corn oil groups. Similarly, Non-HDL-C was markedly reduced in hamsters fed with tea oil, rape seed oil,canola oil and corn oil compared with that of hamsters maintained on the larddiet at the end of the experiment. Plasma HDL-C increased in all five groupscompared with the baseline values throughout the experiment. However, grapeoil and corn oil groups had HDL-C significantly lower than the lard-fedhamsters at the end of the experiment. Interestingly, only tea oil and canola oilcould decrease significantly the ratio of non-HDL-C to HDL-C at the end ofexperiment.

TABLE 3.CHANGE IN FOOD CONSUMPTION, BODY WEIGHT AND RELATIVE ORGAN WEIGHTSIN MALE GOLDEN SYRIAN HAMSTERS FED ONE OF THE FIVE DIETS CONTAINING

0.1% CHOLESTEROL AND 10% LARD, TEA OIL, GRAPE OIL, CANOLA OIL OR CORN OILFOR 12 WEEKS


Food intake 9.5 � 0.9 9.2 � 0.6 9.4 � 0.5 9.1 � 0.5 9.3 � 0.6Body weight (g)

Initial 112.5 � 9.2 112.0 � 6.7 113.0 � 5.4 112.5 � 10.3 113.0 � 8.6Final 130.0 � 15.3 127.5 � 7.2 130.5 � 7.6 132.5 � 14.6 134.0 � 15.6

Relative organ weight (% body weight)Liver 4.01 � 0.25a 3.79 � 0.19a 3.99 � 0.20a 3.84 � 0.24a 3.76 � 0.22b

Kidney 0.82 � 0.04 0.80 � 0.02 0.78 � 0.06 0.79 � 0.04 0.80 � 0.04Heart 0.35 � 0.03a 0.34 � 0.02a 0.35 � 0.02a 0.33 � 0.01b 0.33 � 0.02b

Testis 2.92 � 0.20 2.92 � 0.20 2.94 � 0.12 2.97 � 0.22 2.92 � 0.20Epididymal fat pad 0.95 � 0.13 0.91 � 0.17 0.88 � 0.12 0.93 � 0.27 0.88 � 0.21Perirenal fat pad 1.64 � 0.23 1.61 � 0.21 1.68 � 0.23 1.71 � 0.37 1.69 � 0.26

Data were expressed as mean � standard deviation; n = 10 each group; means at the same row withdifferent superscript (a, b, c) differ significantly at P < 0.05.

866 L. GUAN ET AL.

Plasma TG concentrations in all five groups also increased substantiallycompared with their baseline values throughout the entire experiment.However, the four vegetable oils could equally be effective to suppress theelevation in plasma TG level (Table 4).

Liver cholesterol accumulation varied in the five groups. Liver choles-terol was greatest in tea oil group, followed by canola oil group, grape oilgroup, lard group and corn oil group in a decreasing order (Table 4).

TABLE 4.CHANGE IN PLASMA AND ORGAN LIPID CONCENTRATIONS IN MALE GOLDEN

SYRIAN HAMSTERS FED ONE OF THE FIVE DIETS CONTAINING 0.1% CHOLESTEROLAND 10% OF LARD, TEA OIL, GRAPE OIL, CANOLA OIL OR CORN OIL, RESPECTIVELY,

FOR 12 WEEKS


Total cholesterolInitial 121 � 17 120 � 11 120 � 19 121 � 13 120 � 124th week 170 � 30a 148 � 43ab 146 � 24ab 144 � 30ab 124 � 208th week 217 � 15a 185 � 19bc 196 � 29ab 175 � 33bc 170 � 18c

12th week 234 � 31a 202 � 33b 183 � 29b 184 � 18b 198 � 30b

HDL-CholesterolInitial 60 � 5 57 � 6 58 � 8 59 � 5 59 � 44th week 84 � 10a 79 � 11a 70 � 4b 75 � 13ab 54 � 3c

8th week 82 � 11a 73 � 12ab 72 � 8b 73 � 9ab 66 � 8b

12th week 101 � 9a 99 � 13a 86 � 11b 96 � 13ab 90 � 12b

Non-HDL-CholesterolInitial 61 � 13 61 � 7 62 � 16 61 � 11 61 � 104th week 86 � 28 78 � 32 77 � 22 69 � 19 70 � 208th week 136 � 16a 112 � 17c 124 � 27b 102 � 26c 104 � 18bc

12th week 133 � 29a 103 � 30b 97 � 26b 88 � 22b 108 � 30b

Non-HDL-C/HDL-CInitial 1.01 � 0.18 1.05 � 0.11 1.07 � 0.29 1.04 � 0.19 1.03 � 0.154th week 1.03 � 0.35 1.01 � 0.39 1.10 � 0.29 0.92 � 0.15 1.29 � 0.378th week 1.70 � 0.37 1.57 � 0.40ab 1.74 � 0.42ab 1.39 � 0.25 1.62 � 0.36b

12th week 1.32 � 0.30a 1.01 � 0.30b 1.13 � 0.31ab 0.94 � 0.32b 1.20 � 0.33ab

Triglyceride (mg/dL)Initial 96 � 32 90 � 29 114 � 43 106 � 33 109 � 304th week 201 � 46 196 � 78 190 � 69 159 � 99 167 � 848th week 241 � 79 156 � 56b 130 � 34b 141 � 51b 134 � 56b

12th week 269 � 94a 185 � 47b 185 � 38b 170 � 45b 175 � 69b

Organ Cholesterol (mg/g)Liver 72 � 11bc 90 � 12a 74 � 20bc 81 � 11ab 67 � 13c

Heart 3.3 � 0.2bc 3.4 � 0.1ab 3.5 � 0.2a 3.2 � 0.1c 3.3 � 0.2abc

Non-HDL-C = [TC] - [HDL-C].Data were expressed as mean � SD; n = 10 each group; means at the same row with differentsuperscript (a, b, c) differ significantly at P < 0.05.HDL-C, high-density lipoprotein-cholesterol; TC, total cholesterol.


Fecal Output of Neutral and Acidic Sterols

Concentration of neutral sterols in the feces of the hamsters is shown inTable 5. Total fecal neutral sterol content in hamsters fed canola oil and cornoil was significantly greater than that in hamsters fed the lard, tea and grape oil.Compared with the lard group, tea oil increased the excretion of total fecal bileacids although the difference was statistically significant. Hamsters fed thelard and tea oil, however, had greater excretion of total fecal bile acids than theother three groups.

Cholesterol Balance. Total intake of cholesterol was compared with theexcretion of neutral and acidic sterols (Table 5). Cholesterol retention wascalculated according to the difference between the intake and excretion of bothneutral and acidic sterols. It was found that cholesterol retention was lesser inthe canola and corn oil groups than that in the other three groups. When thecholesterol retention was expressed as a percentage of cholesterol intake perhamster, lard group had the greatest apparent cholesterol absorption followedby grape oil group, tea oil group, corn oil group and canola oil group in andecreasing order (Table 5).

Western Blot of CYP7A1, HMGR, LDLR, LXR, and SREBP-2

No difference in hepatic CYP7A1 and HMGR was seen among the fivegroups (Fig. 1). However, LDL-R appeared to be greater in tea oil group thanthat in the lard and corn oil groups. Except for the tea oil group, all thevegetable oil groups had LXR significantly greater than the lard group. Like-wise, all the four vegetable oil diets could up-regulate SREBP-2 protein,however, only tea and canola oil groups statistically had greater SREBP-2compared with control group (Fig. 1).

DISCUSSION

The present report was the first of its kind to examine the relative hypo-cholesterolemic activity of tea oil compared with that of grape oil, canola oiland corn oil with lard being as a control. The results clearly demonstrated thattea oil and grape oil were equally effective as canola oil and corn oil inlowering blood cholesterol in hamsters. However, tea oil and grape oil affecteddifferently the lipoproteins with the former decreasing only non-HDL-C andhaving no or little effect on HDL-C concentration, while the latter decreasingboth non-HDL-C and LDL-C concentrations. In this regard, grape oil behavedlike corn oil, while tea oil acted like canola oil. The varying effect of tea oil andgrape oil on the ratio of non-HDL-C to HDL-C can be explained by their

868 L. GUAN ET AL.

TAB

LE

5.C

HO

LE

STE

RO

LB

AL

AN

CE

INH

AM

STE

RS

FED

TH

ED

IET

SC

ON

TAIN

ING

0.1%

CH

OL

EST

ER

OL

AN

D10

%L

AR

D,T

EA

OIL

,GR

APE

OIL

,C

AN

OL

AO

ILA

ND

CO

RM

OIL

,RE

SPE

CT

IVE

LY,F

OR

12W

EE

KS*

Lar

dTe

aoi

lG

rape

oil

Can

ola

oil

Cor

noi

l

Cho

lest

erol

inta

ke(m

g/da

y)10

.02

�1.

199.

23�

0.47

9.76

�0.

379.

39�

0.50

9.63

�0.

23Fe

cal

tota

lne

utra

lst

erol

outp

ut(m

g/da

y)2.

22�

0.67

b1.

78�

0.60

b3.

00�

0.49

b6.

00�

2.00

a6.

13�

1.86

a

Feca

lto

tal

acid

icst

erol

outp

ut(m

g/da

y)3.

56�

1.12

a4.

90�

1.55

a1.

79�

0.53

b1.

84�

0.71

b1.

31�

0.68

b

Net

Cho

lest

erol

equi

vale

ntre

tain

ed(m

g/da

y)5.

45�

1.46

a3.

99�

1.60

a4.

87�

1.11

a2.

15�

0.48

b2.

64�

0.90

b

App

aren

tch

oles

tero

lab

sorp

tion

(%in

take

)53

.39

�2.

13a

41.1

4�

1.39

b48

.57

�2.

82ab

22.0

9�

1.22

d27

.61

�1.

31c

*Dat

aw

ere

expr

esse

das

mea

n�

stan

dard

devi

atio

n;n

=10

each

grou

p;A

ppar

ent

chol

este

rol

abso

rptio

n=

[(ch

oles

tero

lin

take

-ex

cret

ion

ofne

utra

lan

dac

idic

ster

ols)

/cho

lest

erol

inta

ke];

Mea

nsat

the

sam

ero

ww

ithdi

ffer

ent

supe

rscr

ipt

(a,b

,c)

diff

ersi

gnifi

cant

lyat

P<

0.05

.


0.0

0.4

0.8

1.2

1.6

0.0

0.4

0.8

1.2

1.6

0.0

0.4

0.8

1.2

1.6

0.0

0.4

0.8

1.2

1.6

0.0

0.4

0.8

1.2

1.6

CYP7A1

β -Actin

Lardoil

Grapeoil

Canolaoil

Cornoil

HMGR

β -Actin

LDLR

β -Actin

LXR

β -Actin

SREBP-2

β -Actin

HM

GR

LX

RL

DL

-R

bc ab ab a

ba aab

ab

a ab abb

b

c

CY

PA

1SR

EB

P-2

Tea

Lard oil Grape oil Canola oil Corn oilTea oil

FIG. 1 EFFECT OF TEA OIL, GRAPE OIL, CANOLA OIL AND CORN OIL COMPARED WITHLARD ON THE RELATIVE IMMUNOREACTIVE MASS OF HEPATIC CYP7A1 ENCODING

CHOLESTEROL 7A-HYDROXYLASE, 3-HYDROXY-3-METHYLGLUTARY-COAREDUCTASE (HMGR), LDL RECEPTOR (LDL-R), LIVER X RECEPTOR (LXR) AND STEROL

REGULATORY ELEMENT BINDING PROTEIN (SREBP-2). DATA WERE NORMALIZEDWITH B-ACTIN SO THAT VALUE OF THE CONTROL (LARD) GROUP WAS REGARDED AS

1.0. VALUES WERE EXPRESSED AS MEAN � STANDARD DEVIATION, n = 10

870 L. GUAN ET AL.

respective fatty acid composition. It is known that oleic acid has a lesserLDL-C lowering effect but possesses a greater HDL-raising effect comparedwith linoleic acid (Katan et al. 1994; Kris-Etherton and Yu 1997; Trautweinet al. 1999). In fact, tea oil and canola oil are rich in oleic acid, while grape oiland corn oil are rich in linoleic acid. Therefore, a smaller ratio of non-HDL-C/HDL-C in tea oil group and canola oil group was expected compared withthat in grape oil and corn oil-fed hamsters (Table 2).

Polyunsaturated-to-saturated fatty acids (P : S) ratio in diet is a key deter-minant of plasma cholesterol levels. It is proved that the higher the ratio is, thelower the plasma TC is (Scott and Margo 1990; McNamara 1992). To beexpected, plasma TC levels among the five groups were inversely correlatedwith their corresponding P : S-values (Table 4). Dietary P : S ratio may be alsoassociated with plasma HDL-C concentration. Ehnholm et al. (1982) reportedthat the plasma HDL-C concentration decreased as the ratio of the P : Sincreased in the diet. In addition, in contrast to the n-6 polyunsaturated fattyacid-enriched vegetable oils that lower HDL-C concentration, n-3 polyunsatu-rated fatty acid-rich oils did not or mildly decrease HDL-C concentration(Conner 2000). The present study found that TC and HDL-C concentrationswere affected by not only P : S but also n-3: n-6 fatty acids. This was reflectedby the observation that tea oil with the lowest P : S ratio and the lowestpercentage of n-6 fatty acid had the highest HDL-C concentration among thefour unsaturated cooking oils, although the difference was not statisticallywarranted.

Four vegetable oils were equally effective in reducing plasma TG level(Table 4). Elevated plasma TG has emerged as a significant and independentrisk factor for major coronary events (Paul 2000). In the present study, exces-sive cholesterol and lard intake induced a marked increase in plasma TGconcentration; however, feeding four vegetable oils could significantly attenu-ate the hypertriacylglycerolemic response induced by cholesterol and lardfeeding. It is known that dietary polyunsaturated fatty acids suppress transcrip-tion of a group of hepatic genes encoding glycolytic and lipogenic enzymes,thus leading to reduction in tissue and plasma TG (Teran-Garcia et al. 2007).Therefore, reduced plasma TG levels were expected in the four vegetable oilgroups because these vegetable oils are rich in monounsaturated or polyun-saturated fatty acids. In addition, LXR has emerged as a key regulator ofcholesterol and lipid metabolism (Tontonoz and Mangelsdorf 2003; Barish andEvans 2004; Li and Glass 2004). The present study found a striking positivecorrelation between protein expression of LXR and plasma TG levels (Fig. 2).It was further affirmed that hamsters fed the four unsaturated vegetablecooking oil diets showed a marked up-regulation of LXR protein abundancecompared with the lard-fed group, although the difference between the tea oiland lard groups was not statistically significant.


Tea oil could up-regulate, while grape oil did not significantly increaseexpression of SREBP-2 and LDL-R (Fig. 1). In both humans and hamsters,most non-HDL-C is taken up by the LDL-Rr-mediated pathway (Chen et al.1996). LDL-R expression is a key determinant of plasma cholesterol levels. Itis evident that saturated fatty acids reduce clearance of LDL-C by suppressionof LDL-R, whereas unsaturated fatty acids relatively activate the activity ofLDL-R and decrease the LDL-C concentration (Spady and Dietschy 1989;Elke et al. 1997). In response to changes in plasma cholesterol levels, the livermaintains cholesterol homeostasis by regulating internalization of LDL par-ticles and de novo cholesterol synthesis (Cheema et al. 1997). Under normalconditions, as much as 80% of the circulating LDL particles are cleared by theliver mainly through an LDL-R-mediated process (Dietschy 1995). At anylevel of dietary cholesterol, receptor-dependant LDL uptake by the liver isalways higher in animals fed unsaturated fatty acids than in animals fedsaturated fatty acids (Horton et al. 1993). In the present study, the hepaticcholesterol accumulation was significantly elevated in the liver of tea oil groupand, to a lesser degree, in canola oil group. These changes are in accordancewith previous studies indicating that monounsaturated fatty acid-rich diet mayincrease hepatic cholesteryl esters (Spady et al. 1993; Elke et al. 1997). Teaoil, rich in monounsaturated fatty acid, decreased LDL-C concentrationaccompanied with a higher level of LDL-R compared with the other groups,suggesting higher rates of LDL clearance, thereby resulting in a higher level ofhepatic cholesterol concentration in the present study (Table 4).

1.2

1.4

1.6

1.8

2

50 100 150 200 250 300 350 400 450 500

Serum Triacylglycerols (mg/dl)

LX

R a

bund

ance

r = -0.426P < 0.01

FIG. 2 CORRELATION SCATTER PLOT OF HEPATIC LIVER X RECEPTOR (LXR) PROTEINAND SERUM TRIACYLGLYCEROLS. DATA WERE OBTAINED FROM ALL INDIVIDUAL

ANIMALS (N = 50)The pearson’s correlation coefficients (r-values) and level of significance are noted..

872 L. GUAN ET AL.

Tea oil affected differently the excretion of neutral and acidic sterolscompared with other three vegetable oils (Table 5). The major finding was thattea oil had greater excretion of bile acids (acidic sterols) and lesser excretionof neutral sterols compared with the other three vegetable oils. Firstly, thiscould be attributed to the lowest phytosterol in tea oil compared with the otherthree oils (Table 1). It is known that phytosterols could compete with choles-terol for absorption and thus increase the excretion of neutral sterols. Secondly,lesser excretion of neutral sterols and greater excretion of acidic sterols asso-ciated with tea oil could be attributable to it high oleic acid content. It has beenshown that monounsaturated fatty acid (olive oil) and linoleic acid (corn oil)affected the bile composition with the former having relative lesser neutralsterols and greater acidic sterols in the bile compared with the latter in rats(Bravo et al. 1998). When total intake of cholesterol was compared with theexcretion of neutral and acidic sterols, cholesterol retention was calculatedaccording to the difference between the intake and excretion of both neutraland acidic sterols (Table 5). It was found that cholesterol retention was less infour groups fed with the vegetable oils compared with that in the lard-fedhamsters. The present result was in agreement with that of Bravo et al. (1998),who found that mono-or n-6 polyunsaturated fatty acids reduced the absorp-tion of cholesterol compared with dietary saturated fatty acids. The presentstudy suggests that cholesterol-lowering activity associated with thesefour vegetable oils be mediated by their inhibitory effect on cholesterolabsorptions.

In summary, we found that like canola and corn oil, tea and grape seedoils were able to lower plasma cholesterol in hamsters fed with a 0.1%cholesterol diet. However, tea oil decreased only non-HDL-C and had no orlittle effect on HDL-C concentration, while grape oil reduced both non-HDL-C and LDL-C concentrations. Unlike grape oil, tea oil could up-regulateSREBP-2 and LDL-R. Hamsters fed the tea oil diet excreted less neutral butgreater acidic sterols compared with the other three vegetable oils. The presentstudy suggests that cholesterol-lowering activity associated with these fourvegetable oils be mediated by their inhibitory effect on cholesterol absorp-tions. This could be explained, at least in part, why tea oil is rich in oleic acid,a monounsaturated fatty acid.

ACKNOWLEDGMENTS

The author would like to thank Miss R. Jiao, Miss K. Y. Ma and Miss S.N. Lim for their assistance in animal handling, GLC analysis and Westernblotting analyses.


REFERENCES

BARISH, G.D. and EVANS, R.M. 2004. PPARs and LXRs: Atherosclerosisgoes nuclear. Trends Endocrinol. Metab. 15, 158–165.

BRAVO, E., FLORA, L., CANTAFORA, A., DE LUCA, V., TRIPODI, M.,AVELLA, M. and BOTHAM, K.M. 1998. The influence of dietary satu-rated and unsaturated fat on hepatic cholesterol metabolism and thebiliary excretion of chylomicron cholesterol in the rat. Biochim. Biophys.Acta 390, 34–48.

CASTILLA, P., ECHARRI, R., DÁVALOS, A., CERRATO, F., ORTEGA, H.,TERUEL, J.L., LUCAS, M.F., GÓMEZ-CORONADO, D., ORTUÑO, J.and LASUNCIÓN, M.A. 2006. Concentrated red grape juice exerts anti-oxidant, hypolipidemic, and antiinflammatory effects in both hemodialy-sis patients and healthy subjects. Am. J. Clin. Nutr. 84, 252–262.

CHAN, P.T., FONG, W.P., CHEUNG, Y.L., HUANG, Y., HO, W.K.K. andCHEN, Z.-Y. 1999. Jasmine green tea epicatechins are hypolipidemic inhamsters (Mesocricetus auratus) fed a high fat diet. J. Nutr. 129, 1094–1101.

CHEEMA, S.K., CIKALUK, D. and AGELLON, L.B. 1997. Dietary fatsmodulate the regulatory potential of dietary cholesterol on cholesterol7a-hydroxylase gene expression, J. Lipid Res. 38, 315–323.

CHEN, J., SONG, W. and REDINGER, R.N. 1996. Effects of dietary choles-terol on hepatic production of lipid and lipoproteins in isolated hamsterliver. Hepatology 24, 424–434.

CHEN, Z.Y., JIAO, R. and MA, K.Y. 2008. Cholesterol-lowering nutraceuti-cals and functional foods. J. Agric. Food Chem. 56, 8761–8773.

CONNER, W.E. 2000. Importance of n-3 fatty acids in health and disease.Am. J. Clin. Nutr. 71, 171S–175S.

DIETSCHY, J.M. 1995. Experimental mechanism: Regulation of plasma LDLcholesterol. Am. J. Clin. Nutr. 62, 679S–688S.

EBERLE, D., HEGARTY, B., BOSSARD, P., FERRE, P. and FOUFELLE, F.2004. SREBP transcription factors: Master regulators of lipid homeosta-sis. Biochimie 86, 839–848.

EHNHOLM, C., HUTTUNEN, J.K., PIETINEN, P., LEINO, U., MUTANEN,M., KOSTIAINEN, E., PIKKARAINEN, J., DOUGHERTY, R.,IACONO, J. and PUSKA, P. 1982. Effect of diet on serum lipoproteins ina population with a high risk of coronary heart disease. N. Engl. J. Med.307, 850–855.

ELKE, A.T., ANGELIKA, K.R., JULIANE, D., STEPHAN, D. andHELMUT, F.E. 1997. Effect of dietary fats rich in lauric, myristic, palm-itic, oleic or linoleic acid on plama, hepatic and biliary lipids incholesterol-fed hamsters. Br. J. Nutr. 77, 605–620.

874 L. GUAN ET AL.

FAZEL, M., SAHARI, M.A. and BARZEGAR, M. 2009. Comparison of teaand sesame seed oils as two natural antioxidants in a fish oil modelsystem by radical scavenging activity. Int. J. Food Sci. Nutr. 31, 1–10.

HORTON, J.D., CUTHBERT, J.A. and SPADY, D.K. 1993. Dietary fattyacids regulate hepatic low density lipoprotein (LDL) transport by alteringLDL receptor protein and mRNA levels. J. Clin. Invest. 92, 743–749.

KASTELEIN, J.J. 2007. Refocusing on use of cholesteryl ester transferprotein inhibitors. Am. J. Cardiol. 100, S47–S52.

KATAN, M.B., ZOCK, P.L. and MENSINK, R.P. 1994. Effects of fats andfatty acids on blood lipids in humans: An overview. Am. J. Clin. Nutr. 60,1017S–1022S.

KIM, N.H., CHOI, S.K., KIM, S.J., MOON, P.D., LIM, H.S., CHOI, I.Y., NA,H.J., AN, H.J., MYUNG, N.Y., JEONG, H.J. ET AL. 2008. Green teaseed oil reduces weight gain in C57BL/6J mice and influences adipocytedifferentiation by suppressing peroxisome proliferator-activated receptor-gamma. Pflugers Arch. 457, 293–302.

KRIS-ETHERTON, P.M. and YU, S. 1997. Individual fatty acid effects onplasma lipids and lipoproteins: Human studies. Am. J. Clin. Nutr. 65,1628S–1644S.

LAM, C.K., ZHANG, Z.S, YU, H., TSANG, S.Y., HUANG, Y. and CHEN,Z.Y. 2008. Apple polyphenols inhibit plasma CETP activity and reducethe ratio of non-HDL to HDL. Mol. Nutr. Food Res. 52, 950–958.

LI, A.C. and GLASS, C.K. 2004. PPAR- and LXR-dependent pathways con-trolling lipid metabolism and the development of atherosclerosis. J. LipidRes. 45, 2161–2173.

MCNAMARA, D.J. 1992. Dietary fatty acids, lipoproteins, and cardiovasculardisease. Adv. Food Nutr. Res. 36, 253–351.

PAUL, C. 2000. Evidence that triglycerides are an independent coronary heartdisease risk factor. Am. J. Cardiol. 86, 943–949.

ROTHEBLAT, G.H., DELA LLERA-MOYA, M., FAVARI, F., YANCEY,P.G. and KELLNER-WEBEL, G. 2002. Cellular cholesterol flux studies:Methodological considerations. Atherosclerosis 163, 1–8.

SCOTT, M. and MARGO, A.D. 1990. Dietary influences on serum lipids andlipoproteins. J. Lipid Res. 31, 1149–1172.

SHANAN, H. and YING, G. 1982. The comprehensive utilization of camelliafruits. Am. Camellia Yearbk. 37, 104–107.

SPADY, D.K. and DIETSCHY, J.M. 1989. Interactions of aging and dietaryfat in the regulation of low density lipoprotein transport in the hamster. J.Lipid Res. 30, 559–569.

SPADY, D.K., WOOLLETT, L.A. and DIETSCHY, J.M. 1993. Regulation ofplasma LDL-cholesterol levels by dietary cholesterol and fatty acids.Annu. Rev. Nutr. 13, 355–381.


STANGL, V., LORENZ, M. and STANGL, K. 2006. The role of tea and teaflavonoids in cardiovascular health. Mol. Nutr. Food Res. 50, 218–228.

TERAN-GARCIA, M., ADAMSON, A.W., YU, G., RUFO, C.,SUCHANKOVA, G., DREESEN, T.D., TEKLE, M., CLARKE, S.D. andGETTYS, T.W. 2007. Polyunsaturated fatty acid suppression of fatty acidsynthase (FASN): Evidence for dietary modulation of NF-Y binding tothe Fasn promoter by SREBP-1c. Biochem. J. 402, 591–600.

TONTONOZ, P. and MANGELSDORF, D.J. 2003. Liver X receptor signalingpathways in cardiovascular disease. Mol. Endocrinol. 17, 985–993.

TRAUTWEIN, E.A., RIECKHOFF, D., KUNATH-RAU, A. and ERBERS-DOBLER, H.F. 1999. Replacing saturated fat with PUFA-rich (sunfloweroil) or MUFA-rich (rapeseed, olive and high-oleic sunflower oil) fatsresulted in comparable hypocholesterolemic effects in cholesterol-fedhamsters. Ann. Nutr. Metab. 43, 159–172.

VAZIRI, N.D. and LIANG, K.H. 1996. Down-regulation of hepatic LDLreceptor expression in experimental nephrosis. Kidney Int. 50, 887–893.

XIA, L., ZHANG, A. and XIAO, T. 1993. An introduction to the utilization ofcamellia oil in China. Am. Camellia Yearbk. 48, 12–15.

876 L. GUAN ET AL.

comparison of hypocholesterolemic activity of tea seed oil with commonly used vegetable oils in...

Documents