Eur J Nutr 42 : 293–296 (2003) DOI 10.1007/s00394-003-0422-6
Received: 2 January 2003 Accepted: 19 March 2003 E. Sarkadi-Nagy (present address) Semmelweis University
Faculty of Medicine Institute of Pathophysiology Budapest, Hungary M.-C. Huang
Kaohsiung Medical University School of Medicine
Dept. of Public Health Kaohsiung, Taiwan G.-Y. Diau
Division of Pediatric Surgery Dept. of Surgery
Tri-Service General Hospital Nai-Whu Taipei, Taiwan
R. Kirwan McGill University Montreal, Quebec, Canada A. Chueh Chao · C. Tschanz · J. T. Brenna ()
Division of Nutritional Sciences Savage Hall, Cornell University Ithaca, NY 14853 USA Tel.: +1-6 07/2 55-91 82 Fax: +1-6 07/2 55-10 33 E-Mail: [email protected]
■ Summary Background Addition
of highly polyunsaturated fatty acids to infant formulas raises the possibility of increased lipid perox-idation. Aim of the study We deter-mined the effects of increasing lev-els of dietary docosahexaenoic acid (DHA) and arachidonic acid (AA) on lipid peroxidation and perox-idative potential in piglet tissues.
Methods Four groups of piglets
(n = 6) were bottle-fed a formula containing one of four treatments: no long chain fatty acid (Diet 0) and three different levels of DHA/AA at 1-fold (0.3 %/0.6 % FA; Diet 1) 2-fold (0.6 %/1.2 % FA; Diet 2) and 5-fold (1.5 %/3 % FA; Diet 5) concentration used in some human infant formulas, and all with equal amount of vitamin E (5.7 IU/ 100 kcal formula) for four weeks.
Results There were no significant
differences between the groups in conjugated diene and glutathione
(GSH) levels in the liver, and thio-barbituric acid-reactive substance (TBARS) in plasma. TBARS levels of the erythrocyte membranes in-creased in a dose-dependent man-ner when in vitro oxidation was induced with 10 mM hydrogen per-oxide (H2O2) for 30 minutes. The TBARS levels of the liver homo-genates of the Diet 5 and Diet 2 groups were significantly different than those of the membranes of the Diet 0 group when the in vitro oxidation was induced with H2O2.
Conclusion The results show that
dietary vitamin E effectively pre-vented lipid peroxidation at the LCP concentrations investigated and suggest that levels presently in infant formulas are sufficient.
■ Key words LCP supplementa-tion – piglets – lipid peroxidasupplementa-tion – vitamin E – red blood cells
SHORT CONTRIBUTION
Eszter Sarkadi-Nagy Meng-Chuan Huang Guan-Yeu Diau Ryan Kirwan Angela Chueh Chao Carolyn Tschanz J. Thomas Brenna
Long chain polyunsaturate
supplementation does not induce
excess lipid peroxidation of piglet tissues
Abbreviations
AA arachidonic acid DHA docosahexaenoic acid EPA eicosapentaenoic acid GSH glutathione
LCP (C ≥ 20) long chain polyunsaturated fatty acids PUFA polyunsaturated fatty acid
RBC red blood cells SCO single-cell oil
TBARS thiobarbituric acid reactive substances
Introduction
Long chain polyunsaturated fatty acids (LCP) are recog-nized as playing a key role in the development of the hu-man central nervous system in the perinatal period [1]. Starting in the mid-1990s, infant formulas with LCP have appeared in countries around the world, and re-cently in the USA and Canada. The goal of supplemen-tation is to increase tissue content of LCP, and it is well known that supplementation is particularly effective at increasing tissue docosahexaenoic acid (DHA) content [2]. Tissues with greater unsaturation levels are at risk
294 European Journal of Nutrition, Vol. 42, Number 5 (2003) © Steinkopff Verlag 2003
for increased oxidative susceptibility. In the absence of adequate endogenous antioxidant, the propagation of free radicals can lead to the reaction of secondary lipid autooxidation products with macromolecules such as membrane constituents, enzymes, and DNA. These con-cerns are relevant at all ages, but particularly for infants who are growing rapidly.An obvious choice for assisting in protection of LCP is vitamin E, which has long been known to function as an antioxidant for protection of polyunsaturated fatty acids in cell membranes through its peroxyl radical trapping activity [3].
In a recent study [4], we reported that increasing amounts of LCP fed to piglets in the first month of life did not alter serum chemistry or other measures of well-being. Here, we report the effects of increasing doses of dietary DHA and arachidonic acid (AA) and fixed amount of vitamin E supplementation on lipid peroxi-dation and in vitro susceptibility to oxiperoxi-dation in piglet tissues from that study.
Materials and methods
■ Animals and diets
All procedures involving animals were approved by the Cornell Institutional Animal Care and Use Committee (IACUC). Six sows from the Cornell University swine fa-cility were delivered spontaneously at term and four healthy male piglets from each litter were randomly as-signed to one of the treatment groups. Diet 0 served as an LCP-free control; Diet 1, Diet 2, and Diet 5 contained 34/17, 68/34 and 170/85 AA/DHA in units of mg fatty
acid per 100 kcal formula. All diets contained 5.77 IU
vi-tamin E per 100 kcal formula. Seventy-five percent of the vitamin E content was alpha-tocopherol with the re-minder being of unspecified stereochemistry. The diet composition and preparation procedure is detailed else-where [4].
■ Analyses
Red blood cell (RBC) FA composition was determined by gas chromatography as previously described [4]. A representative coefficient of variation (CV) for compo-nents of minor abundance, e. g., LCP, is about 10 %. De-termination of glutathione (GSH) in the liver was car-ried out by modification of the method of Ellman [5]. Briefly, liver homogenates (10 % w/v) were prepared and were combined with sodium azide (2 mM) and sulfosal-icylic acid (4 % w/v). Following centrifugation, super-natants were analyzed for acid-soluble free sulfhydryl group content using 5,5’-dithiobis(2-nitrobenzoic acid), with representative CV of about 11 %. Conjugated di-enes were assayed in an isooctane extract of liver
ho-mogenate by absorption at 234 nm and calibrated using a molar absorption coefficient of 29,500 M–1cm1 (CV~16 %). Plasma thiobarbituric acid reactive sub-stances (TBARS) was measured by the method of Placer et al. [6]. Quantification of TBARS was performed by us-ing 1,1,3,3 tetraethoxypropane standard (CV~8 %). Ox-idative susceptibility was assessed by determination of TBARS concentration after exposure of RBCs to 10 mM hydrogen peroxide (H2O2) [7] (CV~10 %), and liver ho-mogenates to a range of hydrogen peroxide concentra-tions (0, 1, 5, 10 mM) for 30 minutes (CV~14 %) [8].
■ Statistics
Kruskal-Wallis one-way ANOVA by ranks was applied to the data. Post hoc comparisons between pairs of means are made by using Wilcoxon rank sum test, with down-ward adjustment of the alpha level to compensate for multiple comparisons; significance was set at P < 0.083 (0.05/6 comparisons = 0.083) and performed by SPSS re-lease 10.0 for Windows. For liver TBARS the effect of the diet was determined by the nonparametric Friedman two-way ANOVA procedure followed by a post-hoc test performed by SAS 8.1 (Cary, NC).
Results
The major results are presented in Table 1. Erythrocyte RBC membrane total lipid DHA and AA increased with dietary DHA/AA supplementation. These modifications resulted in a concomitant dose-dependent increase in the membrane unsaturation index. In vivo lipid peroxi-dation was assessed by measuring conjugated diene and glutathione levels in the liver, and by plasma TBARS. There were no significant differences between the groups in any of these parameters.
Oxidative susceptibility to 10 mM hydrogen peroxide was determined in RBC membranes. The TBARS level of Diet 1 was not different from the control Diet 0, or from Diet 2 and Diet 5. Diet 2 and Diet 5 had significantly greater TBARS than Diet 0, indicating greater suscepti-bility to this extreme oxidative insult, and consistent with great tissue unsaturation.
Fig. 1 shows graphically the liver homogenate oxida-tive susceptibility after incubation with increasing con-centrations of hydrogen peroxide and measured by TBARS. With the Friedman procedure keeping the H2O2 effect constant, we found the effect of diet significant (P < 0.05). The significant differences were detected be-tween Diet 0 vs. Diet 2 and Diet 0 vs. Diet 5 when the ob-servations were pooled for all four H2O2levels.
E. Sarkadi-Nagy et al. 295 Vitamin E protects against peroxidation with LCP supplementation
Discussion
Dietary long chain fatty acids are readily incorporated into cell membranes and the alteration in fatty acid com-position is expected to make tissue and plasma more susceptible to free radical attack and lipid peroxidation. Increasing membrane unsaturation increases oxidative susceptibility and in principle increases the need for an-tioxidant protection. Therefore, the recommendation for vitamin E is higher when large amounts of LCP are consumed.
A human infant formula that has proven beneficial in a recent clinical trial provides 21.3 IU vitamin E per gram of LCP [9]. The piglet formulas Diet 1, Diet 2, and
Diet 5 contain 113 IU, 57 IU, and 23 IU vitamin E, re-spectively, per gram of AA and DHA. Comparing the amount of vitamin E contained in the piglet formulas to the human infant formula, Diet 1 contained 5.3-fold, Diet 2 2.7-fold, and Diet 5 1.1-fold vitamin E per gram DHA and AA. Diet 5 therefore contained an equivalent amount of vitamin E/gram LCP as the human LCP-con-taining formula. Because greater relative amounts of vitamin E did not yield lower values for piglet tissue ox-idation, e. g., in Diet 1 or Diet 2, we conclude that in-creased amounts of vitamin E would not improve oxida-tive status of tissues for human infants on that formula. A dose response was observed for increasing oxida-tive susceptibility with increasing LCP. The hydrogen peroxide concentrations used in these experiments pro-vided a much greater oxidative challenge than would be seen under physiological conditions, and were chosen to produce an effect. The diet effect was significant for liver TBARS and the RBC membrane responded to the 10 mM hydrogen peroxide concentration. For both the RBC and liver Diet 2 and Diet 5 produced significantly more TBARS than the controls. It is not clear if there are im-portant physiological consequences to increasing oxida-tive susceptibility, and therefore these data alone do not support increasing vitamin E in formula.
The TBARS test is sensitive to many secondary prod-ucts formed during lipid peroxidation, and it is accepted as an overall estimate of lipid peroxidation in tissues and plasma. This assay is thought to be preferable to other methods when the content of dietary fat is varied [10, 11]. The other indices of peroxidation that we mea-sured, conjugated dienes and glutathione, are consistent with the TBARS data. Conjugated dienes are produced during the first phase of lipid peroxidation and glu-tathione is essential for maintenance of protein thiols and has other antioxidant functions.
Diet 0 Diet 1 Diet 2 Diet 5 Red blood cell DHA2 0.9±0.2a 2.8±0.5b 3.6±0.4b 4.0±0.5b Red blood cell AA2 4.0±0.2a 5.4±0.9b 6.5±0.7b, c 8.2±0.4c Red blood cell UI3 106 110 119 119 Liver glutathione
(µmol/g wet tissue) 5.2±0.6 5.8±0.8 5.1±0.8 5.9±0.3 Liver conjugated dienes 35.5±19.4 27.3±9.6 30.0±13.1 36.2±7.8 (nmol/g wet tissue)
Plasma TBARS (nmol/ml plasma) 10.0±1.9 9.4±0.8 10.0±2.1 9.2±1.4 RBC oxidative susceptibility4 71.6±4.5a 78.8±8.5a, b 81.8±2.7b 92.6±10.8b (nmol/g hemoglobin)
1Values are the means ± SD, n = 6. Different superscripts in a row indicate significantly different values (P < 0.05) by one-way ANOVA. The Diet 0 served as an LCP-free control, Diet 1 contains LCP in some infant for-mulas in the USA, Diet 2 and Diet 5 have approximately 2- and 5-fold greater concentration of LCP than Diet 1. 2Expressed as wt% of total fatty acids
3UI Unsaturation index =∑
i(Fxx dbx), where F is percentage of fatty acid x, and db is number of double bonds in fatty acid x.
4TBARS measured after incubating RBC in 10 mM hydrogen peroxide for 30 minutes Table 1 Red blood cell fatty acid composition and
antioxidant status of 4-wk-old piglets fed formulas with increasing amount of PUFA1
Fig. 1 Susceptibility of liver homogenate to hydrogen peroxide-induced lipid per-oxidation. TBARS levels were measured in the liver homogenates after exposure of increasing concentration of hydrogen peroxide. Values are means ± SD (n = 6). The effect of the diet was significant (P < 0.05) keeping the H2O2effect constant. The significant differences were detected between Diet 0 vs. Diet 2 and Diet 0 vs. Diet 5 when the observations were pooled for all four H2O2levels
296 European Journal of Nutrition, Vol. 42, Number 5 (2003) © Steinkopff Verlag 2003
There are many studies of human adults on fish oil supplementation, used as a source of DHA and eicosa-pentaenoic acid (EPA, 20:5n-3). Results conflict, showing either increased or decreased oxidative sensitivity [12, 13]. In preterm human studies, where formula was sup-plemented with 0.35 % DHA and 0.65 % EPA, RBC do not have elevated TBARS levels [14]. Supplementation of preterms with 0.4 % AA and 0.25 % DHA does not result
in increased lipid peroxidation measured by the AA metabolite F2 isoprostanes [15].
In summary, our data collectively support the levels of vitamin E now used in US formulas. The data show that no additional protection against oxidation is neces-sary in supplemented piglet tissues at vitamin E to LCP levels five-fold greater than those used in commercial infant formulas in the USA.
References
1. Uauy R, Hoffman DR, Peirano P, Birch DG, Birch EE (2001) Essential fatty acids in visual and brain development. Lipids 369:885–895
2. Abedin L, Lien EL, Vingrys AJ, Sinclair AJ (1999) The effects of dietary alpha-linolenic acid compared with docosa-hexaenoic acid on brain, retina, liver, and heart in the guinea pig. Lipids 345:475–482
3. Packer L (1991) Protective role of vita-min E in biological systems. Am J Clin Nutr 534 (Suppl):1050S–1055S 4. Huang MC, Chao A, Kirwan R, Tschanz
C, Peralta JM, Diersen-Schade DA, Cha S, Brenna JT (2002) Negligible changes in piglet serum clinical indicators or or-gan weights due to dietary single-cell long-chain polyunsaturated oils. Food Chem Toxicol 404:453–460
5. Ellman GL (1959) Tissue sulfhydryl groups. Arch Biochem Biophys 82:70–77
6. Placer ZA, Cushman LL, Johnson BC (1966) Estimation of product of lipid peroxidation (malonyl dialdehyde) in biochemical systems. Anal Biochem 162:359–364
7. Stocks J, Offerman EL, Modell CB, Dor-mandy TL (1972) The susceptibility to autoxidation of human red cell lipids in health and disease. Br J Haematol 236:713–724
8. Godin DV, Garnett ME (1992) Species-related variations in tissue antioxidant status–II. Differences in susceptibility to oxidative challenge. Comp Biochem Physiol B 1033:743–748
9. Birch EE, Hoffman DR, Castaneda YS, Fawcett SL, Birch DG, Uauy RD (2002) A randomized controlled trial of long-chain polyunsaturated fatty acid sup-plementation of formula in term in-fants after weaning at 6 wk of age. Am J Clin Nutr 753:570–580
10. L’Abbe MR, Trick KD, Beare-Rogers JL (1991) Dietary (n-3) fatty acids affect rat heart, liver and aorta protective en-zyme activities and lipid peroxidation. J Nutr 1219:1331–1340
11. Wander RC, Du SH, Ketchum SO, Rowe KE (1996) alpha-tocopherol influences in vivo indices of lipid peroxidation in postmenopausal women given fish oil. J Nutr 1263:643–652
12. Palozza P, Sgarlata E, Luberto C, Pic-cioni E, Anti M, Marra G, Armelao F, Franceschelli P, Bartoli GM (1996) n-3 fatty acids induce oxidative modifica-tions in human erythrocytes depend-ing on dose and duration of dietary supplementation. Am J Clin Nutr 643:297–304
13. Higdon JV, Liu J, Du SH, Morrow JD, Ames BN, Wander RC (2000) Supple-mentation of postmenopausal women with fish oil rich in eicosapentaenoic acid and docosahexaenoic acid is not associated with greater in vivo lipid peroxidation compared with oils rich in oleate and linoleate as assessed by plasma malondialdehyde and F(2)-iso-prostanes. Am J Clin Nutr 723:714–722 14. Hoffman DR, Uauy R (1992)
Essential-ity of dietary omega 3 fatty acids for premature infants: plasma and red blood cell fatty acid composition. Lipids 2711:886–895
15. Stier C, Schweer H, Jelinek J, Watzer B, Seyberth HW, Leonhardt A (2001) Ef-fect of preterm formula with and with-out long-chain polyunsaturated fatty acids on the urinary excretion of F2-isoprostanes and 8-epi-prostaglandin F2alpha. J Pediatr Gastroenterol Nutr 322:137–141
