510 Biochimica et Biophysica Acta 875 (1986) 510-515 Elsevier

BBA 52119

Oxidation of oleic and elaidic acids in rat and human heart homogenates

A.c. Lanser *, E.A. Ernken and J.B. Ohlrogge

Northem Regional Research Center, Agricultural Research Service, Us. Department ofAgriculture n. 1815 North University Street. Peoria. IL 61604 (US.A.)

(Received April 10th. 1985) (Revised manuscript received October 8th. 1985)

Key words: f3-0xidation; Elaidic acid; Oleic acid; Sex difference; (Heart)

Parallel incubations with uniformly 14C-labeled oleic and elaidic acids were conducted to compare oxidation rates in tissue homogenates prepared from rat and human hearts. Radioactivity in 14C02 and 14C-labeled chain-shortened acid-soluble products was used to measure the extent of oxidation. Oxidation rates (pmoljmin per mg heart protein) determined on 14C-labeled acid-soluble products suggest that oleic acid was oxidized 35-40% faster than elaidic acid by both male and female rat heart homogenates, whereas human heart homogenates oxidized these fatty· acids at equal rates. Rates for female heart homogenates were somewhat higher than those for males in rats and humans. Rates of formation of 14C02 were the same for each acid in rat and human heart tissue. Comparative rates of formation of oxidation products expressed as oleic j elaidic ratios from parallel incubations confirm that preferential oxidation of oleic acid occurred with rat heart homogenates, but not with the human heart homogenates. These data suggest that the presence of the trans double bond in elaidic acid does not impair its utilization for energy by human heart muscle.

Introduction Rat studies have shown that some isomeric fatty acids tend to accumulate in lipid classes The content of vegetable fat in the American [4-6]. while others are excluded. suggesting im­ diet has gradually increased over the last 50 years paired metabolism or turnover of certain struc­ to the extent that it now represents almost 50% of tures. Distribution of fatty acids in human heart total fat consumption [1]. Hydrogenation of vege­ tissue indicates that isomeric fatty acids are pre­ table oils to produce consumer products causes sent approximately in the same proportions as positional and geometrical isomerization of the their dietary inclusion [3]. Concern has been ex­ natural cis double bonds in the fatty acids, These pressed that some of these structures may not be isomeric fats. which are 5-8% of the fatty acids in suitable substrates for ,B-oxidation. Because fatty our diet. are readily absorbed and deposited into acids are the main energy source of the heart tissue lipids [2.3]. muscle [7]. it is important to know the capability of the heart to metabolize the various fatty acid structures which are available for oxidation. Dur­ * To whom correspondence should be addressed. ing heart trauma. are all isomeric fatty acids avail­ ** The mention of firm names or trade products does not imply that they are endorsed or recommended by the U,S. able to supply energy or does the heart fail to Department of Agriculture over other firms or similar prod­ oxidize some structures at the same rate and to the ucts not mentioned. same extent as their natural cis counterparts?

0005-2760j86jS03.50 if'; 1986 Elsevier Science Publishers B.V. (Biomedical Division) 511

Data on oxidation of isomeric fatty acids are Materials and Methods based mainly on animal studies. In vivo experi­ ments show that isomeric fats are metabolized at Chemicals. L-Carnitine, coenzyme A. cyto­ similar rates and to the same extent as the cis fatty chrome c. NAD and fatty acid-free human serum acids [8-10]. For example, in rats and dogs, the albumin were obtained from Sigma Chemical trans isomers of oleic and linoleic acids were ab­ Company (St. Louis, MO): ATP from pol Bio­ sorbed and catabolized equally as well as the chemicals, Inc. (Milwaukee, WI). [U- 14 ClOleic acid corresponding cis acid [8]. In vitro studies, how­ was purchased from New England Nuclear (Bos­ ever, have indicated that cellular- or tissue-specific ton, MA): methyl oleate from Hormel Institute differences toward fatty isomers may exist [11-14]. (Austin, MN): SAS (solubilizer for aqueous sam­ For example, Lawson and Holman [14] used ples) from Research Products International Corp. polarographic procedures to determine that oleic (Mount Prospect. IL): PPO (2,5-diphenyloxazole) acid was oxidized significantly faster than elaidic from Aldrich Chemical Company, Inc., Milwau­ acid in male rat heart mitochondria. In contrast. kee, WI). 14 expired 14 CO2 trapped during incubations with Preparation ojjarry acid substrares. [U- ClOleic heart tissue homogenates from young male and acid (0.025 ,umol) was esterified with di­ female rats maintained on various diets led Menon azomethane and diluted with unlabeled methyl and Dhopeshwarkar [15] to report that carbon-14­ oleate to a specific activity of 1 ,uCij,umo!. Iso­ labeled elaidate and oleate were oxidized equally merization to methyl elaidate was accomplished welL regardless of sex or dietary differences. Even with p-toluenesulfinic acid [18]. The elaidate/ though large variability causes values to overlap. oleate mixture was separated by argentation thin­ the reported averages suggest that elaidate may be layer chromatography (TLC). Bands were scraped oxidized 2-3-times faster than oleate in heart ho­ and the esters eluted from the silica gel with ethyl mogenates from rats fed a stock diet and from ether. [U- 14 ClElaidic acid was prepared by male rats fed a high-trans diet. saponification of the elaidyl ester with ethanolic­ Differences in conclusions in previous studies potassium hydroxide followed by acidification with may be related to variations in methods used to hydrochloric acid. The elaidic acid was purified by estimate oxidation rates. Studies which have mea­ TLC by developing the plates with diethyl ether/ sured 14C02 release may reflect only a minor petroleum ether/acetic acid (1: 1: 0.01). Radio­ portion of the total /3-oxidation activity. This has chromatogram scan indicated better than 99% pur­ been shown to be the case for rat and human heart ity. oxidation of palmitate [16,17] where the majority . [U- 14 ClOleic acid was diluted with unlabeled of oxidation products were recovered as chain­ oleic acid (HormeL 99% + ) to a specific activity of shortened acid-soluble intermediates. Oxygen up­ 1 ,uCij,umo!. take measurements also do not reveal the potential After esterification with diazomethane. the pur­ partial oxidation of acyl chains with subsequent ity of elaidic and oleic acids was determined by release of chain-shortened derivatives. capillary gas chromatography (Packard Model 428: For the above reasons our studies have re-ex­ 25 m capillary column coated with OV-275 at amined the question of /3-oxidation of cis versus 160°C) to be greater than 95 and 99%. respec­ trans unsaturated fatty acids. We have measured tively. oxidation of uniformly labeled oleic and elaidic Each acid was complexed at 37°C to human acids by trapping both CO2 released and acid­ serum albumin in a molar ratio of 5 : 1 in 4.2 ml of soluble intermediates. and we have used tissue 5 mM Tris buffer (pH 7.4) containing 0.12 M homogenates which contain cytosolic enzymes and sucrose and 1 mM EDTA. peroxisomes as well as mitochondria, and thus Preparation oj heart muscle homogenares. Hu­ more closely approximate the intact organism. man heart muscle (right atrial appendage) was obtained from patients (ages 48-74 years) under­ going heart by-pass surgery (0.06-0.61 g wet weight). Rat hearts were obtained from Sprague- 512

Dawley rats (ages 4-24 months) which had been the septum cap at various times from 0 to 180 min. euthanized with halothane (0.60-1.43 g). Heart Incubations were continued for an additional 30 tissue was maintained on ice (human) or in ice-cold min after termination of oxidation in the last saline (rat) until homogenized. After removal of sample of the series. The acidified incubation mix­ fat and connective tissue, the muscle tissue was ture was centrifuged and 200 ttl of the supernatant weighed, then minced with scissors and homoge­ were assayed for radioactivity by liquid scintilla­ nized in 1.6 ml (human) or 3 ml (rat) of 10 mM tion counting in 10 ml of toluene-SAS cocktail Tris-HCl (pH 7.4) containing 0.25 M sucrose, 2 (2833 ml toluene, 500 ml SAS, 20 g PPO). The mM EDTA and 50 units/ml heparin. The tissue same cocktail was used to count the 14C02 trapped suspension was homogenized twice with a Poly­ in the hyamine hydroxide. Protein was assayed tron tissue processor (Brinkmann Instruments) for according to Read and Northcote [24] with bovine 2.5 s at a rheostat setting of 6.5. This process serum albumin as a standard. releases subsarcolemmal, but not interfibrillar, Rate/ ratio calculations. Rates for 14C02 pro­ mitochondria from heart muscle [19]. After two duction and for appearance of 14C-labeled acid­ centrifugations for 5 min at 500-600 X g, the su­ soluble oxidation products were obtained from pernatant was decanted into 3 ml (human) or 5 ml graphs of the carbon-14 content in these oxidation (rat) of a solution containing 1.0 mM malate, 50 products (disintegrations per min) vs. time and mM sucrose, 60 mM KCI, 150 mM Tris-HCl (pH were determined along the linear portion of the 7.4), 20 mM K 2P04, 10 mM MgCl 2 and 2 mM curve. Absolute rates (pmoljmin per mg super­ EDTA. The supernatant still contained cytosolic natant heart protein) were calculated at 10% enzymes and peroxisomes and closely represents in oxidation of the fatty acid substrate. vivo conditions [20,21]. Rat heart suspensions were Ratios (oleic/elaidic) were calculated for each further diluted with water to a total volume of 9 heart assay from the slopes of the individual rate ml to give a final concentration of 7-16 mg pro­ curves and also from the absolute rates determined tein/ml. Human tissue samples were not further from parallel incubations. diluted and had a concentration range from 2 to 15 mg protein/ml. Each heart tissue homogenate Results became the stock solution for addition to the series of cis and trans fatty acids. Chloroform/methanol (2: 1, vIv) extraction of rat and human heart Because long-chain fatty acids and their CoA homogenates used in these incubations yielded and carnitine esters are not acid soluble, the pro­ similar weights of extractable total lipid per gram duction of 14C02 and 14C-labeled percWoric of wet tissue (2.30 vs. 2.14 mg/g, respectively). acid-soluble products from 14C-labeled fatty acids OXidations. Fatty acid oxidation was de­ reflects the overall rate of fatty acid ,B-oxidation termined radiochemically on a microscale [22] [23,25]. Release of 14C02 during oxidation of oleic according to the general procedure of Van Hins­ and elaidic acids showed an initial lag period bergh et al. [23]. Oleic or elaidic acid (100 ttM) followed by linear production (Fig. 1). Radioactiv­ bound to albumin in as: 1 molar ratio was used ity in the perchloric acid-soluble fraction generally as substrate. Incubations were carried out in a increased at a linear rate during the first 30 min of final volume of 200 ttl. After 10 min of preincuba­ oxidation, often followed by a leveling off of the tion at 37°C of both the heart homogenate and curve, possibly due to loss of enzyme activity in fatty acid-albumin complex, the reaction was the heart cell preparation. started by addition of homogenate (180 ttl) to the Average oxidation rates for oleic and elaidic fatty acid (20 ttl) in 1.5 ml plastic centrifuge vials acids to 14C02 and 14C-labeled acid-soluble prod­ which were held in 20 ml glass scintillation vials ucts by male and female rat and human heart containing 0.05 ml hyamine hydroxide. The vials homogenates are given in Table 1. The rates of were capped with rubber septa and maintained at 14C02 production from oleic and elaidic acids 37°C with shaking. The oxidations were stopped were similar in the male rat and in the female rat by addition of 100 ttl 12% perchloric acid through heart homogenates, although the rates from the 513

TABLE I I'C0 RATES OF OXIDATION OF OLEIC AND ELAIDIC ACIDS 2 BY RAT AND HUrvlAN HEART HOMOGENATES Oleic Oleic Oxidation rates are expressed as pmoljmin per mg supernatant Elaidic protein; rates were calculated when 10% of the substrate had , been oxidized and are an average of n samples. The means ± '" S.D. are given. Differences between fatty acids. between sexes x a..--=='---'----'_---'-_L----! E 16 "C·acid Solubles and between species were determined to be non-significant at .§- Oleic Oleic 0.05 level. 12 Elaidic

Sex Fatty n 14C02 14C_acid • E!aidic acid solubles Rat male oleic 4 8.0±3.0 102.2 55.0 20 40 60 80 100 120 a 20 40 60 80 100 120 Time, min elaidic 4 7.7±4.0 61.6± 39.0 female oleic 4 12.4±4.6 263.4 ± 11 0.1 Fig. 1. Representative oxidation curves of uniformly 14 C-Iabeled elaidic 4 11.7±7.4 168.9± 43.4 oleic and eIaidic acids by rat (A) and human (B) heart homo­ genates. Oxidations were measured by assay of carbon-14 in Human CO2 and perchloric acid-soluble products during parallel in­ male oleic 5 7.8 ±2.6 116.0 41.2 cubations of albumin-bound oleic (X) and elaidic (e) acids eIaidic 5 7.2 ± 2.4 106.5± 34.9 with heart homogenates. female oleic 6 1O.6±4.8 137.6± 42.4 elaidic 6 11.4±5.8 139.0± 39.5

However, acid-soluble products show that oxida­ female heart homogenates were higher. In con­ tion rates are the same for oleic and elaidic acids trast oleic acid displayed a 35-40% faster rate of in heart homogenates from human males as well as oxidation into acid-soluble products in both the from females. Rates determined with female heart male and female rat heart homogenates. These homogenates are slightly higher than for males as rates were again higher for female than for male noted previously for rats. heart homogenates. The absolute fatty acid oxidation rates of oleate Data for 14C02 production by human heart and elaidate varied widely from subject to subject homogenates is nearly identical to rat 14C02 data. Therefore, in order to compare the oxidation of

TABLE II RATIOS OF OXIDATION PRODUCTS FORMED DURING INCUBATIONS OF UNIFORMLY 14C-LABELED OLEIC AND ELAIDIC ACIDS WITH RAT AND HUMAN HEART HOMOGENATES

Species Sex n 14C02 14C-Iabeled acid soluble n Total 14 C oleic/ eIaidic oleic/elaidic oleic/elaidic

Rat C male 4 1.90 ± 0.43 a 1.88 ± 0.28 4 1.71 ± 0.34 b female 4 1.61 ±0.20 1.71 ± 0.28 4 1.70±0.16 * * * Human C male 5 1.13 ±0.32 1.08 ± 0.16 5 1.11 ±0.28 female 6 0.98 ±0.12 1.08 ± 0.12 6 0.99±0.05

* Significant species differences at 0.01 level. a Average ratios ± S.D. based on rates determined on n samples. Values were determined by computer as the slope of the best-fit line between data points during linear production of oxidation products (dpm/min). b Average ratios±S.D. of paired rates (pmoljmin per mg supernatant protein) calculated from the sum of 14C02 + 14C-labeled acid soluble products at 10% substrate oxidation.

C Ratios from rat data are significantly different (P < 0.01) from the theoretical value of 1.0 (equal oxidation of oleic and elaidic acids) whereas ratios from human data are not significant at the 0.01 level. 514 these two substrates. we calculated the ratio of [16]. 14C02 accounted for an increasing percentage oleate and elaidate oxidation rates for parallel of oxidation products during the course of our incubations with the same tissue (Table II). Ratios incubations, reaching a maximum of 24% after 120 greater than unity indicate that there is a prefer­ min. Apparently, the lag time before 14C02 was ential oxidation of the cis fatty acid by rat heart. detected and the linear increase of 14C02 as oxida­ Ratio averages determined for 14C02 and acid­ tion products can be explained by the time re­ soluble products for male and female rat heart quired for the acetate fragments from f3-oxidation homogenates were between 1.61 and 1.90 with to traverse the tricarboxylic acid cycle. ratios for female rats being somewhat lower than Human heart homogenates yielded nearly iden­ for males. In contrast. homogenates prepared from tical rates to those for rat heart homogenates for human hearts show no preferential oxidation of oxidation of these unsaturated fatty acids to 14C02. either fatty acid in men or women. No difference was seen between oleic and elaidic acids with human male and female heart homo­ Discussion genates. Rates for female heart homogenates ap­ pear slightly higher in both 14C02 and acid-solu­ Previous studies have shown that trans un­ ble products, but not significantly. Oxidation rates saturated acids are absorbed and readily incor­ for the two acids measured as acid-soluble prod­ porated into tissue lipids and can account for from ucts were also the same for human male and 1-9% of the total fatty acids in human heart female heart homogenates. [26,27]. More recently, Ohlrogge et al. have de­ Our data on cis/trails oxidation rates from termined that elaidic acid contributed about 20% male rats agree very well with similar ratios (1.68 of the trans-octadecenoate content in human heart and 1.86) determined by Lawson and co-workers [3]. [13,14] using oxygen uptake with preparations of Oxidation of fatty acids not only produces CO2, heart mitochondria. Cis/trans ratios calculated but also tricarboxylic acid cycle intermediates. either from slopes of the best-fit line between data Stanley and Tubbs [28] reported an accumulation points during linear production of oxidation prod­ of these intermediates which were themselves ucts or from absolute rates at 10% oxidation (Ta­ oxidized only after depletion of the original fatty ble II, last column), are similar. The ratios for acid substrate. Radioactive CO2 represented both male and female rat heart homogenates are 2-25% of the oxidation products in incubations greater than one. indicating that there is prefer­ (up to 30 min) with [1- 14 C]palmitate with rat ential oxidation of oleic acid. tissues and only 1-2% in incubations of [U­ Oxidations with human heart homogenates 14C]palmitate with rat heart homogenates [16]. yielded cis/trans ratios near unity for both males With human muscles, 14C02 accounted for 1-3% and females in CO2 and acid-soluble products. of the oxidation products from [1- 14 C]palmitate These data show that. in contrast to rats. the [16]. Furthermore, in contrast to mitochondrial human heart can oxidize oleic and elaidic acids oxidation, peroxisomal oxidation of fatty acids is equally well and that the presence of the trans not complete [21]. Short-chain acid-soluble inter­ double bond in elaidic acid does not impair its mediates from peroxisomal oxidation may accu­ utilization for energy by human heart muscle. This mulate if further oxidation by mitochondria does would suggest that human heart tissue inherently not occur. contains enzymes or cellular mechanisms not The assay procedure employed here provides available in rat heart tissue or that years of con­ more accurate and sensitive determinations of fatty sumption of a diet high in trans fatty acids has acid oxidation than measuring expired CO2 or 02 allowed the human heart enzyme system to adjust consumption because both CO2 and acid-soluble to the presence of these structures. Such an 'ad­ products are monitored. The acid-soluble inter­ justment' mechanism was seen after 6 weeks in an mediates represented the predominant products experiment using heart mitochondria from rats fed during the course of these incubations which is in diets containing either hydrogenated fat or hydro­ agreement with earlier palmitate oxidation studies genated fat plus corn oil [11]. Ratios of oxidation 515 rates decreased toward unity by 7 and 20%, re­ 11 Hsu. C.M.L. and Kummerow. F.A. (1977) Lipids 12. spectively, from values determined after 1 week on 486-494 these diets. . 12 Lawson. L.D. and Kummerow. F.A. (1979) Lipids 14. 501-503 13 Lawson. L.D. and Kummerow. F.A. (1979) Biochim. Bio­ Acknowledgments phys. Acta 573. 245-254 14 Lawson. L.D. and Holman. R.T. (1981) Biochim. Biophys. The authors thank Dr. D.M. Geiss and H.J. Acta 665. 60-65 Gasdorf for assistance in obtaining tissue samples. 15 Menon, N.K. and Dhopeshwarkar. G.A. (1983) Biochim. Biophys. Acta 751. 14-20 This work was supported in part by a grant from 16 Veerkamp. J.H.. Van Moerkerk. H.T.B.. Glatz. J.F.C. and the Institute of Shortening and Edible Oils. Van Hinsbergh. V.W.M. (1983) Biochim. Biophys. Acta 753.399-410 References 17 Van Hinsbergh. V.W.M.. Veerkamp. J.H. and Bookelman. H. (1979) J. ~lol. Cell. Cardiol. 11. 1245-1252 18 Gibson. T.W. and Strassburger. P. (1976) J. Org. Chern. 41. Rizek. R.L., Welsh, S.O.. Marston. R.1vL and Jackson. E.M. 791-793 (1983) in Dietary Fats and Health (Perkins. E.G. and Visek. 19 Palmer. J.W., Tandler. B. and Hoppel. c.L. (1977) J. BioI. W.J.. ed.). Ch. 2. pp. 13-43, American Oil Chemists' Society. Chem. 252. 8731-8739 Champaign . 20 Mathias. M.M .. Dupont. J. and Hwang. D.H. (1973) Life 2 Wood, R. (1979) in Geometrical and Positional Fatty ACid Sci. 13. 257-267 Isomers (Emken. E.A. and Dutton. H.J.. ed.). Ch. 9. pp. 21 LazarO\\'. P.B. (1978) J. BioI. Chern. 253. 1522-1528 213-281, American Oil Chemists' Society. Champaign 22 Beaven. M.A.. Wilcox. G. and Terpstra. G.K. (1978) Anal. 3 Ohlrogge, J.B., Emken. E.A. and Gulley. R.M. (1981) J. Biochem. 84. 638-641 Lipid Res. 22. 955-960 23 Van Hinsbergh. V.W.M.. Veerkamp. J.H. and Van Moer­ 4 Reichwald-Hacker, I.. Grosse-Oetringhans. S.. Kiewett. I. kerk. H.T.B. (1978) Biochem. Med. 20. 256-266 and Mukherjee. K.D. (1979) Biochim. Biophys. Acta 575. 24 Read. S.M. and Nonhcote. D.H. (1981) Anal. Biochem. 327-334 116. 53-64 5 Hov. C.E. and Holmer, G. (1979) Lipids 14, 727-733 25 Van Hinsbergh. V.W.M., Veerkamp. J.H. and Van Moer­ 6 W;od, R. (1979) Lipids 14, 975-982 kerk. H.T.. (1-978) Arch. Biochem. Biophys. 190. 762-771 7 Neely, J.R. and Morgan, H.E. (1974) Annu. Rev. Physiol. 26 Johnston. P.V., Johnson. P.c. and Kummerow. F.A. (1957) 36.413-459 Science 126. 698-699 8 Ono. K. and Fredrickson, D.S. (1964) J. BioI. Chern. 239. 27 Kaufmann, H.P. and Mankel. G. (1964) Fette Siefen 2482-2488 Anstrichm. 66. 6-13 9 Anderson. R.L. (1967) Biochim. Biophys. Acta 144. 18-24 28 Stanley. K.S. and Tubbs. P.H. (1975) Biochem. J. 150. 10 Coots. R.H. (1964) J. Lipid Res. 5. 468-472 77-88