Oleic Sunflower Oil Into VLDL Triacylglycerols of Hypertensiv

Home , Trimyristin

Incorporation of dietary triacylglycerols from olive oil and high- oleic sun¯ower oil into VLDL triacylglycerols of hypertensive patients

V Ruiz-Gutierrez1*, JS Perona1, YM Pacheco1, FJG Muriana1 and J Villar2

1Instituto de la Grasa, CSIC; and 2Unidad de HipertensioÂn Arterial y LõÂpidos, Servicio de Medicina Interna, Hospital Universitario Virgen del RocõÂo, Seville, Spain

Objectives: To establish whether the ingestion of diets enriched with olive oil or high-oleic sun¯ower oil may produce changes in the composition of VLDL triacylglycerols from hypertensive patients. It could be relevant for the uptake and metabolism of triacylglycerol-derived metabolites by extrahepatic tissues. Design: Patients were assigned to the diets in a random-order sequence. Subjects: The participants were 24 hypertensive patients recruited from a religious community. Interventions: The study was conducted over two four week periods with a four week washout period between both MUFA diets. Results: Dietary olive oil kept in balance the content of saturated fatty acids and decreased the content of arachidonic acid in VLDL triacylglycerols. HOSO diet reduced the content of palmitic acid and increased the content of linoleic acid. There was also a decrease in trioleate-glycerol and an increase in tripalmitate-glycerol of VLDL after the MUFA diets, but these effects were more pronounced in the HOSO group. Intake of olive oil decreased the content of disaturated triacylglycerols and increased the content of dioleate-containing triacylglycerols. A decrease in palmitate-dioleate-glycerol after dietary HOSO was observed. Olive oil (but not HOSO) promoted the presence of long-chain PUFA of n-3 family at the sn-2 position of VLDL triacylglycerols. Conclusions: Our data indicate that olive oil and HOSO, providing a similar concentration of MUFA (oleic acid), differ in the formation of VLDL triacylglycerols in hypertensive patients. Sponsorship: This study was supported by a grant (ALI96-0456 and OLI96-2126) from the CICYT, Spain. Descriptors: VLDL; oleic acid; triacylglycerols; olive oil; high-oleic sun¯ower oil; hypertension; human

Introduction atherogenic index (total cholesterol=HDL cholesterol) and the LDL : HDL cholesterol ratio in the plasma of normo- Recommendations to intake monounsaturated fatty acids cholesterolaemic and hypercholesterolaemic hypertensive (MUFA)-rich oils were one of the emergent strategies in patients (Ruiz-Gutierrez et al, 1998a). However, olive oil the last years for modulating the plasma lipid pro®le in but not HOSO had in¯uence on the normalising of struc- humans (Heyden, 1994; Katan et al, 1994). Accordingly, tural and functional alterations of erythrocyte cell mem- the so-called Mediterranean diet with a high content of brane in patients affected with essential hypertension MUFA is actually associated with a low incidence of (Muriana et al, 1997a, b). This raises the question of coronary heart disease (CHD) (Keys et al, 1986). whether other lipoproteins rather than LDL and HDL Olive oil contains a high amount of MUFA (mainly cholesterol are important in regulating cellular processes oleic acid (18:1 n-9)) and predominates in Mediterranean when blood pressure is higher than normal. dietary patterns. Diets supplemented with olive oil are Indeed, there are increasing evidences that relate pre- reported to have bene®cial effects on plasma lipids and mature CHD and the metabolism of triacylglycerol-rich lipoprotein pro®le in hyperlipidaemic patients (Nydahl et lipoproteins (Ginsberg et al, 1995; Criqui, 1998). Among al, 1994), but the effects on healthy subjects are still this heterogeneous group of large lipoprotein particles, controversial (Truswell & Choudhury, 1998). In addition, VLDL are secreted mainly by the liver and their chief recent studies have shown that not only olive oil but also function is the transport of endogenously synthesised high-oleic sun¯ower oil (HOSO) may reduce the risk for triacylglycerols into extrahepatic tissues (Thompson, CHD, both oils having a similar effect in diminishing the 1994). Formation of remnant particles and conversion of VLDL to LDL are dependent upon removal of triacyl- *Correspondence: Dr V Ruiz-Gutierrez, Instituto de la Grasa (CSIC), glycerol core molecules by lipolytic pathways, such as Apartado 1078, 41012 Seville, Spain. Contributors: VR-G, JSP, YMP and FJGM conducted most laboratory those mediated by hepatic lipase and lipoprotein lipase, analyses and were blinded to the dietary assignments. JV was in charge of targeting cells with fatty acids for either energy or storage the enrolement of the participants, and was present twice a week in the (Deckelbaum et al, 1992). The fatty acid pattern of kitchen during the preparation of the meals and remained blinded along triacylglycerols in VLDL appears to be in¯uenced by the with the subjects to changes in VLDL triacylglycerol composition. fatty acid composition of dietary fats, so that the metabolic Guarantor: Dr V Ruiz-Gutierrez. Received 30 October 1998; revised 24 February 1999; accepted 15 fates of VLDL and the clearance of triacylglycerols may March 1999 be markedly altered (Campos et al, 1995; Montalto & Incorporation of dietary triacylglycerols V Ruiz-Gutierrez et al 688 Bensadoun, 1993). However, the effects of dietary tria- cholesterol intake was maintained during the three diet cylglycerols on the composition of VLDL in hypertensive periods at 280 ± 300 mg=d. Dietary instructions were given patients are unknown. by a dietitian before entry into the study. Each participant Interestingly, olive oil and HOSO may contain similar was told what food items should be eaten for breakfast, amounts of MUFA (oleic acid) but the oils differ in their lunch, dinner, and in-between meals. No other food items triacylglycerol composition (Perez-Jimenez et al, 1995; except mineral water, coffee, and tea were allowed to be Carelli & Cert, 1993). More importantly, olive oil but not consumed during the study periods. Three duplicate food HOSO may reduce the systolic and diastolic blood pres- portions corresponding to each weekday were collected and sures in healthy subjects (Ruiz-Gutierrez et al, 1997) and homogenised to be analysed for their fat content and other hypertensive patients (Ruiz-Gutierrez et al, 1996). These nutrient. Different menus were prepared as previously observations suggest a possible link between dietary tria- described (Perez-Jimenez et al, 1995). Virgin olive oil cylglycerols and the mechanisms involved in the regulation (Olea europaea, Extra, Baena, Spain) or HOSO of blood pressure. In this present study, we focus our (Helianthus annuus, Coreysa, Seville, Spain) was used for investigation on determining the effects of olive oil and cooking and salad dressing and was occasionally spread on HOSO intake on triacylglycerol composition of VLDL bread slices. To avoid any isomerisation, only oils obtained from plasma of patients affected with essential hyperten- after the ®rst frying were used. Fatty acid composition, sion. Total fatty acid composition and sn-2 positional triacylglycerols and fatty acids at sn-2 position of the oils distribution of fatty acids in VLDL triacylglycerols were are depicted in Tables 2 and 3. also evaluated. Blood sampling and biochemical determinations Methods During the last days of each period, blood samples were obtained from fasting (12 h) patients by venous puncture Subjects, experimental design and diets into EDTA-containing (1 g=L) tubes. Plasma was separated Twenty-four hypertensive volunteers aged 55.7 Æ 3.6 y by centrifugation at 15006g at 4C for 30 min. Lipoprotein were recruited for this study (Table 1). The criterion for fractions were isolated from fresh plasma samples by hypertension was a systolic blood pressure 140 mmHg centrifugation at 105 0006g at 4C for 20 h with the use and a diastolic blood pressure 90 mmHg recorded on at of sequential ¯otation (Schumaker & Puppione, 1986). least three different occasions after the subject had rested VLDL were harvested at density 1.006 kg=L (Havel et al, supine for 10 min. Before enrolment, all participants had a comprehensive review of their medical history, as well as physical examination and a clinical chemistry analysis to Table 2 Fatty acid and triacylglycerol compositions of the virgin olive exclude the possibility of any secondary cause of hyperten- oil (VOO) and high-oleic sun¯ower oil (HOSO) sion. None of them had diabetes mellitus, hypothyroidism VOO HOSO or received any antihypertensive drug or another treatment affecting lipid metabolism. No history of alcohol abuse or mol=100 mol cigarette smoking was detected. All subjects gave their Fatty acid informed consent before participating in the study. 16:0 11.79 4.30 The study was conducted over two four week periods, 18:0 2.79 4.72 during which each participant ate olive oil or HOSO diets, 18:1 n-9 n-7 79.22 80.18 18:2 n-6 3.45 9.44 including a washout (four week) period between the two 18:3 n-3 0.60 0.06 MUFA diets. Assignment was in a random order sequence Others 2.24 1.30 and in a crossover design. The washout period was long Saturated 15.67 9.96 enough to ensure the re-establishment of basal conditions. Monounsaturated 80.28 80.54 Polyunsaturated 4.05 9.50 The study included an initial 28 d period during which all Triacylglycerola,b participants consumed a similar diet (baseline) that con- POP PLS 3.12 0.23 tained 30% of energy as fat (11% saturated fatty acids PPoO PLP 0.58 NDc (SFA), 16% MUFA and 3% PUFA), 22% as protein and POS 1.42 0.35 48% as carbohydrate. Olive oil and HOSO diets were POO 29.45 9.63 PLO PoOO 4.33 2.05 characterised by a lower amount of SFA (6%) and a SOO 5.27 10.32 higher amount of MUFA (21%). The consumption of OOO 45.82 63.21 olive oil and HOSO was 50 g=d, indicating that both OLO 6.54 6.71 oils contributed around 75% of total dietary fat. The daily OLL 0.42 3.30 energy consumption of participants was 9.1 MJ. Dietary LLL ND 1.42 Others 3.05 2.78

a Table 1 Characteristics of the subjects enrolled for the study Nomenclature of fatty acids: P, palmitic acid, hexadecanoic acid, 16:0; O, oleic acid, cis-9 octadecenoic acid, 18:1 n-9; L, linoleic acid, cis-9,12 Value octadecadienoic acid, 18:2 n-6; S, stearic acid, octadecanoic acid, 18:0; Po, palmitoleic acid, cis-9 hexadecenoic acid, 16:1 n-9. Age (y) 55.7 Æ 3.6 bNomenclature of triacylglycerols: POP, palmitate-oleate-palmitate- Body mass index (kg=m2) 25.1 Æ 3.1 glycerol; PLS, palmitate-linoleate-stearate-glycerol; PPoO, palmitate- Diastolic blood pressure (mm Hg) 93.5 Æ 4.1 palmitoleate-oleate-glycerol; PLP, palmitate-linoleate-palmitate-glycerol; Systolic blood pressure (mm Hg) 162.6 Æ 11.9 POS, palmitate-oleate-stearate-glycerol; POO, palmitate-oleate-oleate- Total cholesterol (mmol=L) 5.8 Æ 0.3 glycerol; PLO, palmitate-linoleate-oleate-glycerol; PoOO, palmitoleate- LDL-cholesterol (mmol=L) 4.0 Æ 0.2 oleate-oleate-glycerol; SOO, stearate-oleate-oleate-glycerol; OOO, oleate- HDL-cholesterol (mmol=L) 1.2 Æ 0.2 oleate-oleate-glycerol; OLO, oleate-linoleate-oleate-glycerol; OLL, oleate- linoleate-linoleate-glycerol; LLL, linoleate-linoleate-linoleate-glycerol. Values are means Æ s.d. n 24. cND, not detected. Incorporation of dietary triacylglycerols V Ruiz-Gutierrez et al 689 Table 3 Fatty acid sn-2 positional distribution of the virgin olive oil (Chrompack, Middleburg, The Netherlands) ®tted with a (VOO) and high-oleic sun¯ower oil (HOSO) ¯ame ionisation detector and a split injection system VOO HOSO (splitting ratio 1:30) (Carelli & Cert, 1993; Ruiz-Gutierrez & Barron, 1995). Separation was conducted on a high Fatty acid mol=100 mol temperature aluminium-clad fused-silica capillary column (25 m60.25 mm i.d.) coated with methyl 65% phenylsili- 12:0 0.18 0.03 14:0 0.03 0.04 cone (0.1 mm ®lm). The initial column temperature was 16:0 0.63 0.11 350 C, which was held for 1 min, then programmed to 16:1 n-7 0.40 NDa change from 350 ± 360C at 0.5C=min and remaining at 18:0 0.12 0.14 360C for 8 min. The injection and detector temperatures 18:1 n-9 95.16 93.23 were 360 and 365C, respectively. Helium was used as the 18:2 n-6 2.63 6.45 18:3 n-3 0.85 ND carrier gas at a column head pressure of 130 kPa. Trilino- Saturated 0.96 0.32 lein, trimyristin, trinonadecanoin, triolein, palmitate- Monounsaturated 95.56 93.23 dioleate-sn-glycerol, tripalmitin, tripalmitolein and tris- Polyunsaturated 3.48 6.45 tearin were used as reference standards (Nu-Chek-Prep aND, not detected. Inc., Elysian, Minnesota, USA). Tripentadecanoin was used as internal standard (Sigma-Aldrich Quimica SA, Madrid, Spain). Triacylglycerols were then identi®ed by 1955). Compliance with the diets was determined by their elution times, their expected elution orders and by evaluation of daily food questionnaires and by analysis of comparison with previous studies (Ruiz-Gutierrez et al, fatty acid composition of the cholesterol ester fraction in 1992a, b), because retention is affected not only by the LDL (Sarkkinen et al, 1994). number of carbon atoms but also by the number of double bonds. Separation of triacylglycerols from VLDL Lipids from VLDL were extracted with chloro- Fatty acids in the sn-2 position of triacylglycerols form=methanol (2:1, v=v) and separated by TLC on silica Triacylglycerols were partially hydrolysed by pancreatic gel 60 plates using a solvent system of hexane= lipase from pig (EC 3.1.1.3; Sigma-Aldrich Quimica SA) diethylether=acetic acid (80:20:1, v=v=v). The band corre- and then separated by TLC on silica gel 60 plates by using sponding to triacylglycerols was scraped off and eluted a solvent system of hexane=diethylether=acetic acid with hexane=chloroform (9:1, v=v). (60:40:2, v=v=v) (Ruiz-Gutierrez & Mazuelos 1985). The band corresponding to monoacylglycerols was scraped off, Analysis of fatty acid methyl esters eluted with hexane and treated as above for analysis of fatty Triacylglycerols were saponi®ed by heating for 10 min with acid methyl esters. 0.2 mol=L sodium methylate at 80C. Fatty acid methyl esters were formed by heating again for 5 min with 60 g=L Blood pressure measurements H2SO4 in anhydrous methanol. After extraction with Blood pressure measurements were performed at the right hexane, fatty acid methyl esters were analysed in a Hew- brachial artery of participants using a mercury-gauge lett ± Packard 5890 series II gas chromatograph (Hewlett ± sphygmomanometer. At each visit three blood pressure Packard Co., Avondale, Pennsylvania, USA) equipped with readings were recorded and the average was used to a ¯ame ionisation detector and using an Omegawax 320 determine eligibility. In addition, blood pressure was fused-silica capillary column (30 m60.32 mm i.d., recorded at the beginning and end of each period of 0.25 mm ®lm). The initial column temperature was 200C, MUFA diet. which was held for 10 min, then programmed to change from 200 ± 230C at 2C=min. The injection and detector Statistical analyses temperatures were 250 and 260C, respectively. The ¯ow The signi®cance of the differences between the groups was rate of helium was 2 mL=min. Peak areas were calculated assessed by one-way (repeated measures) ANOVA with by a Hewlett ± Packard 3390A recording integrator. Indivi- Tukey's post-hoc comparison of the means (Neter et al, dual fatty acid methyl esters were identi®ed by comparison 1985). Data were transformed reciprocally before statistical of their retention times with those of reference standards. analysis. All comparisons were considered statistically Fatty acid methyl esters for which no standard was avail- signi®cant at P < 0.05. The analyses were done with the able were quanti®ed using calibration tables of relative GraphPAD InStat (GraphPAD Software, San Diego, Cali- response ratios constructed according to carbon number fornia, USA) and CoStat (CoHort Software, Berkeley, with the use of gas chromatography-mass spectrometry, California, USA) statistical packages. performed on a Konik KNK-2000 chromatograph (Konik Co., Barcelona, Spain) interfaced directly to an AEJ Results MS30=70 VG mass spectrometer (VG Analytical, Manche- ster, UK), using the electron impact ionisation mode. The Diets ion source temperature was maintained at 200C, the Compliance with the diets was estimated to be higher than multiplier voltage was 4.0 kV, the emission current was 90% from the evaluation of daily food questionnaires and 100 mA and the electron energy was 70 eV. The data were by analysis of the fatty acid composition of the plasma processed with a VG 11=250 data system. cholesterol ester fraction of LDL during ingestion of olive oil and HOSO diets (Sarkkinen et al, 1994), indicating a Analysis of triacylglycerols good adherence to the diets. Fractions of triacylglycerols were redisolved in hexane and The fatty acid composition of olive oil and HOSO was analysed in a Chrompack CP9000 gas chromatograph characterised by a high content of MUFA (mainly oleic Incorporation of dietary triacylglycerols V Ruiz-Gutierrez et al 690 acid (18:1 n-9) (80% of total fats)) (Table 2). Palmitic acid PUFA of n-6 family, but kept in balance the content of SFA ((16:0) 75% of total SFA) and a-linolenic acid ((18:3 n-3) and PUFA of n-3 family in VLDL triacylglycerols of 15% of total PUFA) were more abundant in olive oil than in patients with essential hypertension. HOSO, whereas the content of linoleic acid (18:2 n-6) was HOSO diet had similar effects as the olive oil diet when greater in HOSO (9% of total fats) than in olive oil (3% of compared with baseline, except that linoleic acid (18:2 n-6) total fats). Trioleate-sn-glycerol (OOO) was the major was signi®cantly increased (21%, P < 0.001), palmitic acid triacylglycerol found in HOSO (63% of total triacylglycer- was more markedly decreased (58%, P < 0.001) and ara- ols), whereas OOO (46% of total triacylglycerols) and chidonic acid was unchanged (Table 4). These ®ndings palmitate-dioleate-glycerol ((POO) 30% of total triacylgly- were consistent with an increase in MUFA and PUFA of n- cerols) were signi®ed as abundant in olive oil (Table 2). 6 family, and a decrease in SFA. Oleic acid was the predominant fatty acid located at the sn- 2 position of triacylglycerols in both olive oil and HOSO (Table 3). Triacylglycerol composition of VLDL As in a previous study (Ruiz-Gutierrez et al, 1996), there Olive oil and HOSO diets signi®cantly decreased the was a signi®cant reduction of systolic and diastolic blood content of trioleate-glycerol (OOO, 64% for olive oil and pressures in hypertensive patients after olive oil ingestion 71% for HOSO, P < 0.001) and increased the content of (data not shown). tripalmitate-glycerol (PPP, 73% for olive oil and > 100% for HOSO, P < 0.001) in VLDL of hypertensive patients Fatty acid composition of VLDL triacylglycerols (Table 5). Intake of olive oil, in particular, signi®cantly With regard to baseline, olive oil diet resulted in signi®cant decreased the content of disaturated triacylglycerols increases in lauric acid ((12:0), > 100%, P < 0.001), myr- (linoleate-dimirystate-glycerol (MLM, 39%, P < 0.001) istic acid ((14:0), > 100%, P < 0.001), palmitoleic acid and oleate-dipalmitate-glycerol (POP) plus palmitate- ((16:1 n-9), 65%, P < 0.001), stearic acid ((18:0), 21%, linoleate-stearate-glycerol (PLS) (29%, P < 0.001)) and P < 0.01), oleic acid (15%, P < 0.001), octadecatetraenoic increased the content of dioleate-containing triacylglycer- acid ((18:4 n-3), > 100%, P < 0.001), arachidic acid ols (palmitate-dioleate-glycerol (POO, 16%, P < 0.001), ((20:0), > 100%, P < 0.001) and eicosenoic acid ((20:1 n- linoleate-dioleate-glycerol (OLO, 25%, P < 0.05)). 9), > 100%, P < 0.001) of VLDL triacylglycerols (Table 4), and in signi®cant decreases in palmitic acid (18%, P < 0.001), hexadecenoic acid ((16:1 n-7), 41%, Table 5 Triacylglycerol composition of plasma VLDL at baseline and P < 0.001), a-linolenic acid (24%, P < 0.001), eicosatri- after consumption of the virgin olive oil (VOO) and high-oleic sun¯ower enoic acid ((20:3 n-6), 22%, P < 0.001) and arachidonic oil (HOSO) diets for four weeks1, 2 acid ((20:4 n-6), 40%, P < 0.001). Therefore, the ingestion Baseline VOO HOSO of olive oil tended to increase MUFA and to decrease Triacylglycerol3,4 mol=100 mol

MPP 1.67 Æ 0.54b 2.20 Æ 0.78b 3.18 Æ 0.94a Table 4 Fatty acid composition of plasma VLDL triacylglycerols at MLM 5.66 Æ 1.14a 3.45 Æ 0.85b 5.14 Æ 1.80a baseline and after consumption of the virgin olive oil (VOO) and high- PPP 5.85 Æ 1.79c 10.10 Æ 4.07b 15.69 Æ 3.70a oleic sun¯ower oil (HOSO) diets for four weeks1, 2 MOP 0.60 Æ 0.13 0.40 Æ 0.29 0.70 Æ 0.61 POP PLS 6.82 Æ 1.21a 4.87 Æ 0.65b 6.10 Æ 1.68a Baseline VOO HOSO MOO ND5 ND 0.89 Æ 0.65 PPoO PLP 6.21 Æ 0.99 6.46 Æ 0.80 6.25 Æ 2.58 Fatty acid mol=100 mol PLPo MLO 2.15 Æ 0.76a 1.59 Æ 0.29b 1.95 Æ 1.07a MLL 0.94 Æ 0.34 1.04 Æ 0.74 0.65 Æ 0.46 12:0 0.59 Æ 0.22b 1.54 Æ 0.66a 1.73 Æ 0.23a POS ND ND 0.66 Æ 0.62 14:0 0.23 Æ 0.06c 0.57 Æ 0.13a 0.41 Æ 0.06b POO 24.75 Æ 3.15b 28.76 Æ 6.67a 20.97 Æ 3.57c 16:0 25.16 Æ 0.14a 20.51 Æ 4.77b 10.49 Æ 0.42c PLO PoOO 18.87 Æ 1.78b 22.67 Æ 2.74a 19.03 Æ 0.84b 16:1 n-9 0.78 Æ 0.10b 1.29 Æ 0.57a 1.08 Æ 0.26a PLL 5.39 Æ 1.02 5.23 Æ 0.73 5.04 Æ 1.01 16:1 n-7 3.35 Æ 0.15a 1.96 Æ 0.57c 2.45 Æ 0.41b SOO 1.99 Æ 0.22a 0.26 Æ 0.19c 1.18 Æ 0.16b 18:0 5.74 Æ 0.22b 6.94 Æ 2.38a 7.18 Æ 0.93a OOO 13.94 Æ 3.60a 5.04 Æ 0.81b 4.01 Æ 1.18c 18:1 n-9 32.35 Æ 0.40b 37.22 Æ 4.32a 39.51 Æ 6.32a SOL ND 1.53 Æ 1.07 1.75 Æ 1.33 18:1 n-7 1.79 Æ 0.01b 1.97 Æ 0.18b 2.34 Æ 0.41a OLO 4.46 Æ 1.05b 5.57 Æ 1.45a 5.25 Æ 1.55ab 18:1 t 0.41 Æ 0.12b 0.97 Æ 0.53a 0.89 Æ 0.43a OLL 0.70 Æ 0.24b 0.83 Æ 0.78b 1.56 Æ 0.72a 18:2 n-6 20.49 Æ 2.44b 19.23 Æ 4.68b 24.81 Æ 6.69a 18:3 n-3 0.74 Æ 0.18a 0.56 Æ 0.11b 0.80 Æ 0.22a 1Baseline values were obtained after hypertensive patients consumed a 18:4 n-3 0.34 Æ 0.09b 0.72 Æ 0.09a 0.40 Æ 0.07b low-monounsaturated diet for 28 d period. 20:0 0.22 Æ 0.10b 0.69 Æ 0.62a 0.33 Æ 0.07b 2Values are means Æ s.d., n 24. Means with different superscripts are 20:1 n-9 0.23 Æ 0.05b 0.57 Æ 0.21a 0.20 Æ 0.04b signi®cantly different by post-hoc Tukey's test (P < 0.05). Data were 20:3 n-6 1.16 Æ 0.12a 0.90 Æ 0.25b 1.23 Æ 0.18a transformed reciprocally before statistical analysis. 20:4 n-6 4.48 Æ 0.38a 2.71 Æ 0.75b 4.26 Æ 1.20a 3Nomenclature of fatty acids: M, myristic acid, tetradecanoic acid, 14:0; P, 22:4 n-6 0.25 Æ 0.08a 0.22 Æ 0.09a,b 0.17 Æ 0.06b palmitic acid, hexadecanoic acid, 16:0; O, oleic acid, cis-9 octadecenoic 22:5 n-3 0.39 Æ 0.07a 0.48 Æ 0.29a 0.20 Æ 0.11b acid, 18:1 n-9; L, linoleic acid, cis-9, 12 octadecadienoic acid, 18:2 n-6; S, 22:6 n-3 1.30 Æ 0.94 0.95 Æ 0.35 1.52 Æ 0.63 stearic acid, octadecanoic acid, 18:0; Po, palmitoleic acid, cis-9 Saturated 31.94 Æ 0.57a 30.25 Æ 4.43a 20.14 Æ 1.08b hexadecenoic acid, 16:1 n-9. Monounsaturated 38.91 Æ 0.61b 43.98 Æ 4.56a 46.47 Æ 6.75a 4Nomenclature of triacylglycerols: MPP, myristate-palmitate-palmitate- Polyunsaturated 29.15 Æ 2.80a,b 25.77 Æ 4.93b 33.39 Æ 6.84a glycerol; MLM, myristate-linoleate-myristate-glycerol; PPP, palmitate- palmitate-palmitate-glycerol; MOP, myristate-oleate-palmitate-glycerol; 1Baseline values were obtained after hypertensive patients consumed a MOO, myristate-oleate-oleate-glycerol; PLPo, palmitate-linoleate- low-monounsaturated diet for 28 d period. palmitoleate-glycerol; MLO, myristate-linoleate-oleate-glycerol; MLL, 2Values are means Æ s.d., n 24. Means with different superscripts are myristate-linoleate-linoleate-glycerol; PLL, palmitate-linoleate-linoleate- signi®cantly different by post-hoc Tukey's test (P < 0.05). Data were glycerol; SOL, stearate-oleate-linoleate-glycerol. transformed reciprocally before statistical analysis. 5ND, not detected. Incorporation of dietary triacylglycerols V Ruiz-Gutierrez et al 691 Table 6 Fatty acid sn-2 positional distribution of plasma VLDL properties implicated in the pathogenesis of essential triacylglycerols at baseline and after consumption of the virgin olive oil hypertension (Muriana et al, 1997a, b). These observations 1, 2 (VOO) and high-oleic sun¯ower oil (HOSO) diets for four weeks suggest that dietary triacylglycerols may be of relevance in Baseline VOO HOSO the metabolism of vascular cells from hypertensive patients. We have recently reported the in¯uence of olive Fatty acid mol=100 mol oil and HOSO diets on the composition of VLDL triacyl- 12:0 2.34 Æ 0.22b 3.01 Æ 0.31a 2.75 Æ 0.53a glycerols in healthy subjects (Ruiz-Gutierrez et al, 1998b). 14:0 0.88 Æ 0.24b 1.24 Æ 0.35a 0.81 Æ 0.32b Consequently, the present study was designed to determine 16:0 18.44 Æ 5.21b 20.62 Æ 2.60a,b 22.65 Æ 5.94a how both MUFA diets may also in¯uence the composition 16:1 n-9 0.61 Æ 0.22c 3.05 Æ 0.92a 1.49 Æ 1.35b of total fatty acids, triacylglycerol molecular species and a b c 16:1 n-7 2.91 Æ 0.88 2.02 Æ 0.76 1.51 Æ 0.24 fatty acids at sn-2 position of VLDL triacylglycerols in 18:0 18.32 Æ 3.77a 12.91 Æ 2.46b 14.64 Æ 5.41b 18:1 n-9 29.81 Æ 2.20b 35.25 Æ 5.32a 36.55 Æ 7.99a patients affected with essential hypertension (Table 1). 18:1 n-7 2.15 Æ 0.61a 1.76 Æ 0.12b 1.53 Æ 0.21b Olive oil and HOSO had a similar amount of MUFA, 18:1t 0.69 Æ 0.29 0.53 Æ 0.38 0.60 Æ 0.28 however both oils differed in their concentrations of 18:2 n-6 17.99 Æ 1.80a 11.16 Æ 1.46b 12.08 Æ 2.36b minor fatty acids and triacylglycerol composition (Table 18:3 n-3 0.12 Æ 0.09c 2.28 Æ 1.15a 1.50 Æ 0.92b 18:4 n-3 0.18 Æ 0.10c 0.76 Æ 0.11a 0.56 Æ 0.12b 2), and sn-2 positional distribution of fatty acids (Table 3). 20:0 1.46 Æ 0.32a 1.00 Æ 0.37b 0.93 Æ 0.32b In addition to the low content of 20:4 n-6 in membranes 20:1 n-9 0.14 Æ 0.09b 0.28 Æ 0.08a 0.32 Æ 0.11a of hypertensive patients when compared to that in mem- 20:3 n-6 0.33 Æ 0.10b 0.73 Æ 0.19a 0.28 Æ 0.18b branes of healthy subjects (Ruiz-Gutierrez et al, 1996; a b b 20:4 n-6 2.17 Æ 1.73 0.88 Æ 0.36 0.81 Æ 0.21 Ruiz-Gutierrez et al, 1997; Villar et al, 1996), we found 22:5 n-3 0.42 Æ 0.08b 0.85 Æ 0.22a ND3 22:6 n-3 1.04 Æ 0.23b 1.67 Æ 0.58a 0.99 Æ 0.14b that patients with untreated essential hypertension could Saturated 41.44 Æ 5.78 38.78 Æ 3.03 41.78 Æ 5.52 transport more 20:4 n-6 in VLDL (Table 4) than normal Monounsaturated 36.31 Æ 2.46b 42.89 Æ 5.56a 42.00 Æ 8.05a (Ruiz-Gutierrez et al, 1998b). Inclusion of olive oil in the Polyunsaturated 22.25 Æ 2.01a 18.33 Æ 2.93b 16.22 Æ 2.84c diet of hypertensive patients resulted in accumulation of 1Baseline values were obtained after hypertensive patients consumed a 20:4 n-6 in membranes (Ruiz-Gutierrez et al, 1996), here low-monounsaturated diet for 28 d period. we have observed a depletion of 20:4 n-6 in VLDL (Table 2Values are means Æ s.d., n 24. Means with different superscripts are 4). A similar reduction in its precursor (18:2 n-6) was also signi®cantly different by post-hoc Tukey's test (P < 0.05). Data were found in VLDL of healthy subjects (Ruiz-Gutierrez et al, transformed reciprocally before statistical analysis. 3ND, not detected. 1998b). This interesting reverse effect of dietary olive oil in membranes of patients affected with essential hypertension was concomitant with a decrease in 18:2 n-6 and an Contrary to olive oil diet, the HOSO diet produced a increase in 22:5 n-6, which rather well correlates with an signi®cant decrease (15%, P < 0.05) in palmitate-dioleate- enhanced metabolic conversion through the elongation and glycerol. desaturation of 18:2 n-6 to 20:3 n-6 and 20:4 n-6, and of 20:4 n-6 to 22:4 n-6 and 22:5 n-6. However, the content of 18:2 n-6 remained unchanged in VLDL of hypertensive Fatty acids at the sn-2 position of VLDL triacylglycerols patients during the period of eating the olive oil diet. It can Both oils contributed to VLDL enriched in palmitoleic acid be currently explained by some metabolic defect(s) asso- (16:1 n-9, > 100%, P < 0.001) and oleic acid (18:1 n-9, ciated to a poor formation or fast clearance of VLDL 20%, P < 0.01) but depleted in stearic acid (18:0, 20 ± 30%, triacylglycerols containing long-chain PUFA of n-6 P < 0.05) and PUFA of n-6 family (linoleic acid (18:2 n-6, family. After HOSO diet, VLDL of hypertensive patients 35%, P < 0.001) and arachidonic acid (20:4 n-6, 60%, were enriched in 18:2 n-6, without any change in the P < 0.001)) at the sn-2 position of triacylglycerols (Table content of 20:4 n-6 (Table 4). This effect may be related 6). Olive oil (but not HOSO) further promoted the presence to the content of 18:2 n-6 in the oils, which was three times of long-chain PUFA of n-3 family (docosapentaenoic acid greater in HOSO (9% of total fats) than in olive oil. (22:5 n-3, > 100%, P < 0.001) and docosahexaenoic acid Fatty acids are released from circulating VLDL triacyl- (22:6 n-3, 60%, P < 0.001)) in the sn-2 position of VLDL glycerols by the action of the heparin-binding enzyme, triacylglycerols from hypertensive patients. lipoprotein lipase (EC 3.1.1.34) (Wang et al, 1992). The enzyme is tightly regulated by the composition of dietary fats (Montalto & Bensadoun, 1993), can differentiate Discussion between substrates (Calder et al, 1994) and exhibits speci- Growing evidence indicates that plasma levels of triacyl- ®city with respect to the position of fatty acid chains in the glycerol-rich lipoproteins including VLDL, play an impor- glycerol backbone (Wang et al, 1982). Accordingly, the tant role in the development of premature CHD (Ginsberg composition of VLDL triacylglycerols appears to be a et al, 1995; Criqui, 1998). The availability of triacylglycer- determinant for the conversion of VLDL into other lipo- ols is the major driving force in the formation of VLDL. proteins and the uptake and metabolism of triacylglycerol Therefore, dietary factors in¯uencing the supply of triacyl- metabolites by cells. It was surprising to ®nd a signi®cant glycerols may regulate their incorporation into VLDL and decrease of OOO in VLDL of hypertensive patients after ultimately their cellular metabolic utilisation. both MUFA diets (Table 5), despite OOO being the main According to previous studies, the ingestion of triacyl- triacylglycerol found in olive oil (46% of the total) and glycerols from either olive oil or HOSO could similarly HOSO (63% of the total). In addition, olive oil and HOSO modify the plasma lipid and lipoprotein pro®les in hyper- promoted the formation of VLDL enriched in unsaturated tensive patients (Ruiz-Gutierrez et al, 1998a), but olive oil and saturated triacylglycerols, respectively. Our data sup- (and not HOSO) particularly reduced blood pressure (Ruiz- port the concept that absorption, metabolism and=or clear- Gutierrez et al, 1996) and normalised cell membrane ance of dietary triacylglycerols may constitute nutritional Incorporation of dietary triacylglycerols V Ruiz-Gutierrez et al 692 factors which are modi®ed in the development of human Ginsberg HN, Jones J, Blaner WS, Thomas A, Karmally W, Fields L, essential hypertension. In healthy subjects, we postulated Blood D & Begg MD (1995): Association of postprandial triglyceride and retinyl palmitate responses with newly diagnosed exercise-induced the preference for long-chain PUFA in the hepatocyte myocardial ischemia in middle-aged men and women. Arterioscler. triacylglycerol-synthesising pathways after olive oil Thromb. Vasc. Biol. 15, 1829 ± 1838. intake (Ruiz-Gutierrez et al, 1998b) due to the content Havel RJ, Eder HA & Bradon JH (1955): The determination and chemical and positional distribution of minor fatty acids (16:0, 18:0 composition of ultracentrifugally separated lipoproteins in human and 18:3 n-3) rather than of 18:1 n-9. We now propose that serum. J. Clin. Invest. 34, 1345 ± 1353. Heyden S (1994): Polyunsaturated and monounsaturated fatty acids in the in hypertensive patients a diet rich in triacylglycerols from diet to prevent coronary heart disease via cholesterol reduction. Ann. olive oil (but not from HOSO) enhances the ability of Nutr. Metab. 38, 117 ± 122. VLDL to supply cells with MUFA and PUFA. Among James WPT (1996): Nutritional disorders affecting the heart. In: Diseases different pathways for the synthesis of triacylglycerols, the of the Heart, eds. DG Julian, AJ Gamm, KM Fox, RJC Hall, PA Poole- Wilsor, 1442 ± 1458. London: Saunders WB. diacylglycerol transacylase pathway could be probably Katan MB, Zock PL & Mensink RP (1994): Effects of fats and fatty involved in the formation of dioleoyl-containing triacylgly- acids on blood lipids in humans: An overview. Am. J. Clin. Nutr. 60, cerols after dietary olive oil, although this remains to be 1017S ± 1022S. elucidated. Keys A, Menotti A, Karvonen MJ, Aravanis C, Blackburn H, Buzina R, Interestingly, the ingestion of olive oil (but not HOSO) Djordjevic BS, Dontas AS, Fidanza S, Keys MH, Fromhout D, Nedeljkovic S, Punsar S, Seccareccia F & Toshima H (1986): The promoted the presence of long-chain PUFA of n-3 family diet and 15-year death rate in the Seven Countries Study. Am. J. (22:5 n-3 and 22:6 n-3) in the sn-2 position of VLDL Epidemiol. 124, 903 ± 915. triacylglycerols from patients with untreated essential Kinsella JE (1994): Alpha-linolenic acid: functions and effects on linoleic hypertension (Table 6). The same was true for healthy acid metabolism and eicosanoid-mediated reactions. Adv. Food Nutr. Res. 35, 2 ± 160. subjects (Ruiz-Gutierrez et al, 1998b). These results sug- Montalto MB & Bensadoun A (1993): Lipoprotein lipase synthesis and gest that dietary olive oil may increase the availability of n- secretion: effects of concentration and type of fatty acids in adipocyte 3 fatty acids to be incorporated as monoacylglycerols for cell culture. J. Lipid Res. 34, 397 ± 407. the resynthesis of VLDL triacylglycerols in the liver. This Muriana FJG, Ruiz-Gutierrez V, Guerrero A, Montilla C, Leon-Camacho is of major metabolic importance, given the recent evidence M & Villar J (1997a): Olive oil normalizes the altered distribution of membrane cholesterol and sodium-lithium countertransport activity that plasma exchange of n-3 for n-6 PUFA occurs with a in erythrocyte of hypertensive patients. J. Nutr. Biochem. 8, reduction in blood pressure of hypertensive patients 205 ± 210. (Yosefy et al, 1996). In addition, n-3 PUFA competitively Muriana FJG, Villar J & Ruiz-Gutierrez V (1997b): Intake of olive oil can inhibit the utilisation of 20:4 n-6 for cyclooxygenase path- modulate the transbilayer movement of human erythrocyte membrane cholesterol. Cell. Mol. Life Sci. 53, 496 ± 500. way and output of eicosanoids (James, 1996; Kinsella, Neter J, Wasserman W & Kutner MH (1985): Applied Linear Statistical 1994). Models. Homewood: Richard D Irvin. Nydahl MC, Gustafsson IB & Vessby B (1994): Lipid-lowering diets enriched with monounsaturated or polyunsaturated fatty acids but low in saturated fatty acids have similar effects on serum lipid concentrations in hyperlipidaemic patients. Am. J. Clin. Nutr. 59, Conclusions 115 ± 122. Our data are consistent with the notion that triacylglycerols Perez-Jimenez F, Espino A, Lopez-Segura F, Blanco J, Ruiz-Gutierrez V, Prada JL, Lopez-Miranda J, Jimenez-Pereperez J & Ordovas JM (1995): from olive oil appear to be metabolically more useful than Lipoprotein concentrations in normolipidemic males consuming oleic those from HOSO for the formation of endogenous tria- acid-rich diets from two different sources: olive oil and oleic acid-rich cylglycerols that resembles a less atherogenic lipid pattern. sun¯ower oil. Am. J. Clin. Nutr. 62, 769 ± 775. We have demonstrated that fatty acids from dietary olive Ruiz-Gutierrez V & Mazuelos F (1985): Efecto de una dieta con aceites oil or HOSO are selectively incorporated within the tria- calentados sobre la lipasa pancreatica de rata. Grasas Aceites 36, 105 ± 108. cylglycerol molecules of VLDL in patients with untreated Ruiz-Gutierrez V, Cert A & Rios JJ (1992a): Determination of phospho- essential hypertension. However, it is still unknown lipid fatty acid and triacylglycerol composition of rat caecal mucosa. whether VLDL triacylglycerols produced after long-term J. Chromatogr. 575, 1 ± 6. intake of olive oil directly contribute to reduce the cardio- Ruiz-Gutierrez V, Montero E & Villar J (1992b): Determination of fatty acids and triacylglycerol composition of human adipose tissue. vascular risk and to restore the abnormalities of membrane J. Chromatogr. 581, 171 ± 178. lipid bilayer in vascular cells of hypertensive patients. Ruiz-Gutierrez V & Barron LJR (1995): Methods for the analysis of triacylglycerols. J. Chromatogr. 671, 133 ± 168. Ruiz-Gutierrez V, Muriana FJG, Guerrero A, Cert AM & Villar J (1996): Plasma lipids, erythrocyte membrane lipids and blood pressure of hypertensive women after ingestion of dietary oleic acid from two different sources. J. Hypertens. 14, 1483 ± 1490. References Ruiz-Gutierrez V, Muriana FJG, Guerrero A, Cert AM & Villar J (1997): Role of dietary oleic acid from two different sources on fatty acid Calder PC, Yaqoob P & Newsholme EA (1994): Triacylglycerol metabo- composition of erythrocyte membrane and blood pressure in healthy lism by lymphocytes and the effect of triacylglycerols on lymphocyte subjects. J. Nutr. Biochem. 8, 689 ± 695. proliferation. Biochem. J. 298, 605 ± 611. Ruiz-Gutierrez V, Muriana FJG & Villar J (1998a): El aceite de oliva Campos H, Dreon DM & Krauss RM (1995): Associations of hepatic and virgen y las enfermedades cardiovasculares. Per®l lipõÂdico en plasma y lipoprotein lipase activities with changes in dietary composition and composicioÂn lipõÂdica de la membrana de eritrocito humano. Grasas low-density lipoprotein subclasses. J. Lipid Res. 36, 462 ± 472. Aceites 49, 9 ± 29. Carelli AA & Cert A (1993): Comparative study of the determination of Ruiz-Gutierrez V, Morgado N, Prada JL, Perez-Jimenez F & Muriana FJG triacylglycerol in vegetable oils using chromatographic techniques. (1998b): Composition of human VLDL triacylglycerols after ingestion J. Chromatogr. 63, 213 ± 222. of olive oil and high-oleic sun¯ower oil. J. Nutr. 128, 570 ± 576. Criqui MH (1998): Triglycerides and cardiovascular disease. A focus on Sarkkinen ES, AÊ gren JJ, Ahola I, Ovaskainen ML & Uusitupa MIJ (1994): clinical trials. Eur. Heart J. 19, A36 ± A39. Fatty acid composition of serum cholesterol esters, and erythrocyte and Deckelbaum RJ, Ramakrishnan R, Eisenberg S, Olivecrona T & Bengts- platelet membranes as indicators of long-term adherence to fat-modi®ed son-Olivecrona G (1992): Triacylglycerol and phospholipid hydrolysis diets. Am. J. Clin. Nutr. 59, 364 ± 370. in human plasma lipoproteins: role of lipoprotein and hepatic lipase. Biochemistry 31, 8544 ± 8551. Incorporation of dietary triacylglycerols V Ruiz-Gutierrez et al 693 Schumaker VN & Puppione DL (1986): Sequential ¯otation ultracentri- Wang CH, Kuksis A & Manganaro F (1982): Studies of the substrate fugation. Methods Enzymol. 128, 155 ± 170. speci-®city of puri®ed human milk lipoprotein lipase. Lipids 17, Thompson GR (1994): A Handbook of Hyperlipidaemia. London: Current 278 ± 284. Science Ltd. Wang CH, Hartsuck J & McConathy WJ (1992): Structure and functional Truswell AS & Choudhury N (1998): Monounsaturated oils do not all have properties of lipoprotein lipase. Biochim. Biophys. Acta 1123, 1 ± 17. the same effect on plasma cholesterol. Eur. J. Clin. Nutr. 52, 312 ± 315 Yosefy C, Viskoper JR, Varon D, Ilan Z, Pilpel D, Lugassy G, Schneider Villar J, Montilla C, MunÄiz-Grijalvo O, Muriana FJG, Stiefel P, Ruiz- R, Savyon N, Adan Y & Raz A (1996): Repeated fasting and refeeding Gutierrez V & Carneado J (1996): Erythrocyte sodium-lithium counter- with 20:5, n-3 eicosapentaenoic acid (EPA): a novel approach for rapid transport in essential hypertension: correlation with membrane lipids. fatty acid exchange and its effect on blood pressure, plasma lipids and J. Hypertens. 14, 969 ± 973. hemostasis. J. Hum. Hypertens. 10, S135 ± S139.