The Journal of Nutrition Vol. 128 No. 9 September 1998, pp. 1421-1428

Dietary Menhaden and Corn Oils and the Red Blood Cell Membrane Lipid Composition and Fluidity in Hyper- and Normocholesterolemic Miniature Swine¹

Elliott Berlin, Sam J. Bhathena^{*, 2}, Dennis McClure, and Renee C. Peters^*

Metabolism and Nutrient Interactions and ^* Phytonutrients Laboratories, Beltsville Human Nutrition Research Center, Agricultural Research Service, U.S. Department of Agriculture, Beltsville, MD 20705 and  Center for Food Safety and Applied Nutrition, Division of Toxicological Studies, Beltsville Research Facility, Food and Drug Administration, U.S. Department of Health and Human Services, Laurel, MD 20708 

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Fatty acids in the diet are readily incorporated into lipids in various tissues. However, it is not clear whether all tissues have the same level of incorporation. Second, (n-6) unsaturated fatty acids increase the fluidity of membranes, but this has not been shown for (n-3) fatty acids. In this study, we measured the incorporation of (n-6) and (n-3) fatty acids into erythrocyte membrane lipids and studied their effects on the fluidity of erythrocyte membranes. One group of female miniature swine was made hypercholesterolemic by feeding the swine cholesterol and lard for 2 mo; the other group served as controls and was fed a stock diet. Both groups were then fed either corn oil or menhaden oil or a mixture of the two for 23 additional weeks. Blood was collected at 0, 2, 4, 12 and 23 wk after initialization of the experimental diets, and fatty acid composition of phospholipids was assessed. Membrane phospholipids of pigs fed menhaden oil had elevated (n-3) fatty acids (20:5 and 22:6), and lower 18:2 than those fed corn oil. There was no difference in 20:4 content. The fatty acid changes occurred as early as 2 wk after consumption of the corn oil or menhaden oil in pigs previously fed a stock diet, but it took longer in pigs previously fed lard + cholesterol, indicating residual effects of pretreatment. Menhaden oil increased anisotropy (indicating decreased fluidity) more than corn oil for the nonpolar probe diphenylhexatriene (DPH) at earlier time points, but not at 23 wk. Erythrocyte membrane fluidity was significantly related to membrane polyunsaturate content, with (n-6) fatty acids having a greater influence than (n-3) fatty acids. A comparison of the present red blood cell fatty acid compositions with brain synaptosome fatty acid compositions for the same animals showed poor correlations for some of the fatty acids. There was no significant direct relationship between docosahexaenoate (DHA) concentrations in erythrocyte membranes with DHA concentrations in brain synaptosomes from cerebellum, forebrain and caudate nucleus.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

The (n-3) polyunsaturated fatty acids (PUFA)³ have received much attention in recent years because of their effect on the development of the nervous system (Simopoulos 1991), including brain (Bourre et al. 1984), and on the retina (Uauy et al. 1990). Consumption of (n-3) PUFA in fish oil, particularly as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), is recommended for prevention and treatment of hypertension, arthritis, psoriasis and cancer, among other maladies (Simopoulos 1991), and (n-3) and (n-6) types have been recommended for prevention of heart disease. An extensive body of literature describes the incorporation of PUFA into cell membrane phospholipids in humans and experimental animals. Lokesh et al. (1986) reported the effects of menhaden oil feeding on mouse spleen with resulting modulation of prostaglandin production. Hinds and Sanders (1993) demonstrated (n-3) fatty acid incorporation into mouse spleen leukocyte phospholipids. Several reports have described (n-3) replacement of (n-6) fatty acids in rat heart (Charnock et al. 1989), liver (Muriana and Ruiz-Guttierez 1992) and adipose tissue (Leray et al. 1993). Sassen et al. (1993) described (n-3) PUFA incorporation into plasma and aortic plaques in atherosclerotic swine. Because it is impossible to monitor incorporation of dietary fatty acids into tissue phospholipids in humans, Makrides et al. (1995) measured erythrocyte fatty acids to assess (n-3) fatty acids in brain phospholipids in human infants. This approach assumes that the fatty acid composition in erythrocytes reflects that of other tissues. We have reported effects of dietary fats on erythrocytes in humans (Berlin et al. 1989 and 1994a) and in experimental animals (Barnard et al. 1990). In this study, we examined the effects of corn oil (CO) and menhaden oil (MO) on a variety of tissue phospholipids in miniature swine. We have reported the effects of corn and menhaden oil feeding on fatty acid composition in liver (Berlin et al. 1994b), heart membranes and brain synaptosomes (Berlin et al. 1998). Here we report findings with red blood cell membrane phospholipids from the same animals. This permits comparison of PUFA incorporation into heart, liver, and brain phospholipids with PUFA incorporation into red blood cell membranes.

Changes induced in cell membrane composition by dietary fat and cholesterol affect membrane physical-chemical properties including fluidity, and could modulate membrane-associated physiologic processes. Effects of dietary PUFA on cell function could relate to membrane fluidity modulation: fluidity is a major determinant of many membrane-associated functions, such as receptor binding, neurotransmitter release and reuptake, ion transport, protein phosphorylation and membrane-bound enzyme activity (Engelhard et al. 1976, Ghosh et al. 1993). There are potential differences in the effects elicited by various fatty acids that result from differences in chain length and in the numbers and positions of the double bonds.

In this study, we examined effects of dietary (n-6) acids (predominantly 18:2), (n-3) acids (predominantly 20:5 and 22:6) and cholesterol on the fluidity of erythrocyte membranes from miniature swine. Membrane fluidity in mid-bilayer and surface domains was assessed by measuring fluorescence anisotropy with the neutral hydrocarbon probe diphenylhexatriene (DPH) and its polar trimethylammonium (TMA-DPH) and propionic acid (DPH-PA) derivatives. Fluorescence anisotropy data derive from the intensities of fluorescence emission having parallel and perpendicular orientation with respect to a polarized excitation beam. The extent of depolarization, i.e., the decrease in emission that is parallel to the plane of the excitation beam, is thought to reflect the relative motional freedom of the membrane lipids surrounding the reporter fluorophore (Lentz 1993).

The three probes used in this study sample different depths of the cell membrane bilayer, thus reflecting upon different membrane domains (Engel and Prendergast 1981, Kitagawa et al. 1991). The polar derivatives, TMA-DPH and DPH-PA, are associated with the phospholipid headgroups through electrostatic interactions; they therefore describe membrane domains on the periphery of the bilayer, including that part of the fatty acid tail that is close to the headgroup. Kitagawa et al. (1991) demonstrated association of the cationic derivative TMA-DPH with the inner membrane leaflet in platelets, whereas the anionic derivative DPH-PA reflected on that region of the bilayer in proximity to neutral phospholipids, i.e., the outer leaflet of the plasma membrane. In contrast, DPH is usually described as colinear with and intercalated between the fatty acyl tails of the membrane phospholipids. In fluid lipid states, however, DPH might assume various other orientations, including one parallel to the membrane surface in the bilayer center (Borenstain and Barenholz 1993, Lentz 1993).

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Animals and treatment.  Thirty female miniature Hormel-Hanford swine (FDA Swine Breeding Facility, Balok Farms, Suffolk, VA), 4-11 y old and weighing 60-110 kg, were divided into five experimental groups and fed diets (150 g fat/kg feed) containing corn oil, menhaden oil or mixtures thereof for 6 mo. The experimental protocol was approved by the U.S. Food and Drug Administration Institutional Animal Care and Use Committee. The research reported here is part of a larger study (Berlin et al. 1994b and 1998) designed to assess fish oil toxicity and biochemical effects of fish oil feeding on blood components and heart, liver, and brain tissue in normo- and hypercholesterolemic pigs. Twelve of the 30 pigs were first made hypercholesterolemic by consumption of 130 g lard/kg diet + 20 g cholesterol/kg diet for 2 mo at 20 g feed/(kg body wt·d); the other pigs continued to receive a basal stock diet, USDA 1160 (Table 1). At the end of 2 mo, the pigs were started on the various corn and menhaden oil experimental diets. The experimental protocol and diets fed are described in Table 2. No litter mates were assigned to the same group. Pigs were fed 13 g feed/(kg body wt·d) for 6 mo, corresponding to 1.95 g fat/(kg body wt·d). Pigs were weighed each week, and the amount of feed for each pig was adjusted each week to maintain intake at 13 g feed/(kg body wt·d). Diets were prepared as indicated in Table 1 to provide test diets containing 150 g CO/kg diet, 142.5 g CO + 7.5 g MO/kg diet, 75 g CO + 75 g MO/kg diet or 150 g MO/kg diet. For each 100 g of feed, the total diet provided 15 g protein, 50 g carbohydrate and 17 g fat. The total fat included 2 g extractable from the corn and soybean meal in the basal diet plus the 15 g fat added (Table 1). The menhaden oil contained 14% eicosapentaenoate (EPA) and 8% docosahexaenoate (DHA). The high and low doses of (n-3) fatty acid (EPA + DHA) for a typical 80-kg minipig are thus 156 and 1.7 g MO/d, respectively. The menhaden oil was provided through the National Institutes of Health Fish Oil Test Material Program (Charleston, SC). The menhaden oil as supplied was fortified with vitamin E and tert-butylhydroquinone (TBHQ) as antioxidants to prevent deterioration. The contents of EPA and DHA in the oils was verified on several samples by gas chromatography. Corn oil was provided by Best Foods (Best Foods, Union, NJ), and appropriate quantities of vitamin E and TBHQ were added to the corn oil to correspond to the amount present in the menhaden oil. Feed stability was verified by monitoring EPA and DHA concentrations and by testing for peroxides during the study.

Pigs were housed individually (in heated pens in winter) with indoor and outdoor runs and free access to water. Several pigs died before the end of the study primarily because of respiratory problems that were unrelated to the diet treatment. The numbers of pigs receiving each diet and surviving until the study's end are shown as n values in the tables.

Blood sampling and erythrocyte membrane isolation.  Morning blood samples were taken immediately before starting the corn and menhaden oil diets and again at 2, 4, 12 and 23 wk. Blood was drawn into vaccutainer tubes with potassium EDTA as an anticoagulant. After removal from plasma and platelets by differential centrifugation (15 min, 1000 × g), the erythrocytes were dispersed in isotonic phosphate buffer (0.119 mol PO₄/L, pH 7.4) and washed two or three times by centrifugation (20 min, 1000 × g). Erythrocyte membranes were prepared by hypotonic lysis in 7.6 mmol PO₄/L (pH 7.4) according to the procedure of Dodge et al. (1963). Membrane preparations were washed in the 7.6 mmol PO₄/L until the supernatant was clear to remove hemoglobin and other cytoplasmic components. Aliquots were removed for fluidity measurements; the remainder of the membrane preparations were stored at 85°C for later chemical analyses.

Fluidity measurement.  Membrane fluidity was measured by fluorescence polarization of polar and nonpolar hydrocarbon probes in membrane samples. Increasing values for polarization or anisotropy are taken as indicators of decreasing fluidity. In this paper, all data are reported as steady-state anisotropy rather than polarization because a simple mathematical relation exists between these quantities. Steady-state fluorescence anisotropy measurements were made at 37 and 4°C with the nonpolar probe DPH (Shinitzky and Barenholz 1978), the cationic derivative TMA-DPH (Kuhry et al. 1983) and the anionic derivative DPH-PA (Trotter and Storch 1989). Anisotropies are reported as r_DPH, r_TMA-DPH, or r_DPH-PA, with the subscripts corresponding to the fluorescent probe. The three probes (0.002 mol/L in dimethylformamide) were diluted 1000-fold into the aqueous membrane suspensions, which were then incubated with agitation at 35-37°C for 2 h. Steady-state fluorescence polarization was measured with a SLM Model 4800 spectrophotofluorometer (Milton Roy, SLM-Aminco, Rochester, NY) equipped with Glan-Thompson prism polarizers in the T-optical format. Excitation and emission wavelengths were 366 and 430 nm, respectively. Light scattering errors were minimized by ensuring that the anisotropies were of independent concentration.

Chemical analyses.  Membrane phospholipid fatty acyl compositions were determined by capillary gas chromatography of the corresponding methyl esters prepared by transesterification with methanolic hydrochloric acid (Matusik et al. 1984). Membrane protein (Lowry et al. 1951) and lipid phosphorus (Bartlett 1959) concentrations were determined by colorimetry. Membrane cholesterol content was determined enzymatically (Allain et al. 1974) with the use of cholesterol esterase and cholesterol oxidase. Lipids were extracted by an adaptation (Matusik et al. 1984) of the method of Sperry and Brand (1955). Phospholipids were separated from neutral lipids by silicic acid chromatography, using Unisil (Clarkson Chemical, Williamsport, PA). Fatty acid methyl esters were prepared by transesterification with methanolic hydrochloric acid. After clean-up by Alumina chromatography (Fisher Scientific, Fair Lawn, NJ), methyl ester samples were dissolved in isooctane, and chromatography was performed with a Hewlett-Packard (Avondale, PA) model 5890A gas chromatograph. The instrument was equipped with dual-flame ionization detectors, a model 7673A automatic sampler, and a model 3396A integrator. Chromatography was performed with a Supelco (Bellefonte, PA) SP-2560 fused silica 100 m, 0.25-mm i.d. capillary column and 0.20-µm film.

Statistical analysis.  Data were analyzed by two-way ANOVA to measure differences between dietary treatment and time effects, and by linear regression analysis for the various measurements using SAS programs, Version 6.03, (SAS 1988). The model we tested included variation in measurements because of changes in dietary fat. Duncan's multiple range test (SAS 1988) was used to determine differences in model-classified composition data.

	RESULTS

Abstract Introduction Methods Results Discussion References

Table 3 shows the erythrocyte membrane phospholipid fatty acid composition of pigs fed the stock diet or the cholesterol + lard diet for 2 mo. Total (n-3) fatty acids were significantly lower in cholesterol-fed pigs than in stock diet-fed pigs. This was due primarily to lower levels of 22:5. Although fatty acid composition of erythrocyte membrane phospholipids was measured at 2, 4, 12 and 23 wk after feeding the experimental diets (corn oil, menhaden oil or a mixture of the two), only the values for 23 wk are shown in Table 4. Significant dietary fat effects were evident at 2 wk when the 18:2 concentration was significantly lower in pigs fed corn oil. No significant differences were observed in 20:4 in any of the pigs at 2 wk.

In pigs previously fed the stock diet and then the menhaden oil diet, EPA and DHA were significantly increased at 2 and 4 wk. No such increase was observed in the pigs previously fed the cholesterol + lard diet. At 12 and 23 wk, all of the pigs fed menhaden oil showed higher EPA and DHA levels than those fed corn oil. Thus, the residual effects of cholesterol + lard in the diet on fatty acid composition of erythrocyte membrane phospholipids persisted for at least 4 wk.

Pigs fed the cholesterol + lard diet and those fed the stock diet for 8 wk had blood plasma cholesterol concentrations of 9.04 ± 0.03 and 2.2 ± 0.04 mmol/L, respectively (Berlin et al. 1998). In erythrocyte membranes, the mean cholesterol levels were 0.33 ± 0.14 and 0.14 ± 0.04 mmol/g protein in pigs fed the cholesterol + lard diet and the stock diet, respectively, for 8 wk. No significant differences in either cholesterol or phospholipids were observed at any time up to 23 wk in pigs fed either corn oil or menhaden oil. However, there was a progressive, significant increase in all diet groups in cholesterol and phospholipids during the course of the study.

Fluorescence anisotropies for the various probes in the red cell membranes at each sampling for all five groups of minipigs are given in Table 5. At time 0, erythrocyte membranes from the hypercholesterolemic pigs were significantly (P < 0.0001) less fluid than were membranes from the normocholesterolemic pigs as determined by DPH fluorescence, with values of 0.233 ± 0.001 and 0.227 ± 0.001, respectively. These values, not shown in Table 5, are averages for all hyper- and normocholesterolemic pigs, regardless of their subsequent diet treatment. Similar differences were noted with the polar probes. Significantly (P < 0.02) higher anisotropies, r_TMA-DPH, were shown in TMA-DPH for the hypercholesterolemic pigs, than for the normocholesterolemic pigs, 0.260 ± 0.001 and 0.256 ± 0.001, respectively. For the hypercholesterolemic and normocholesterolemic pigs, DPH-PA anisotropies, r_DPH-PA, of 0.274 ± 0.001 and 0.270 ± 0.001, respectively, were significantly different (P < 0.002).

We compared mean values for the fluorescence anisotropy for the subsequent samplings from all pigs fed the same fat diet regardless of dietary pretreatment before starting the experimental diets. At all intervals up to 12 wk and with all probes, the erythrocyte membranes isolated from the pigs fed menhaden oil were less fluid than were the membranes from the pigs fed corn oil. This difference in fluidity disappeared after 23 wk of consumption of the experimental diets. We compared the erythrocyte fatty acid composition at 23 wk with the brain fatty acid composition when these animals were killed at 6 mo. The brain fatty acid data have already been published (Berlin et al. 1998).

We performed regression analyses to examine the relationships between red blood cell mole fractions of various fatty acids (18:2, 20:4, 20:5, 22:5, 22:6, (n-3) and (n-6)) and the respective synaptosome mole fractions in our previous paper (Berlin et al. 1998). Figures 1 and 2 show a comparison of (n-6) fatty acids in erythrocyte membrane phospholipids and fatty acids from caudate nucleus and cerebellum synaptosomes, respectively. Significant direct relationships existed for several fatty acids, with the following equations relating total erythrocyte membrane (n-6) acids to caudate nucleus (CN) and cerebellum (CB) (n-6) acids: Intercepts, slopes, probabilities (P) and correlation coefficients (r) for each equation obtained for the selected long-chain PUFA are given in Table 6. There were significant direct correlations between red blood cell and synaptosome 18:2, 20:5, 22:5 and (n-6), but there was no significant relationship for arachidonate. There were significant inverse relationships between erythrocyte DHA and synaptosomal DHA in caudate nucleus and cerebellum, but there was no significant correlation for forebrain DHA. Total (n-3) fatty acids showed the same pattern, but this was probably because of DHA.

View larger version (18K): [in this window] [in a new window]   Fig 1. Relationship between (n-6) fatty acid mole proportions of caudate nucleus synaptosomal phospholipids and erythrocyte membrane phospholipids of minipigs. The relationship followed the equation where XCN is the sum of (n-6) fatty acids of caudate nucleus synaptosomes and XRBC is the sum of (n-6) fatty acids of erythrocyte membranes.

View larger version (18K): [in this window] [in a new window]   Fig 2. Relationship between (n-6) fatty acid mole proportions of cerebellum synaptosomal phospholipids and erythrocyte membrane phospholipids of minipigs. The relationship followed the equation where XCB is the sum of (n-6) fatty acids of cerebellum synaptosomes and XRBC is the sum of (n-6) fatty acids of erythrocyte membranes.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

In this study, we observed residual effects of cholesterol + lard feeding on the fatty acid composition of erythrocyte membrane phospholipids up to 4 wk when pigs were fed either corn oil or menhaden oil; these effects disappeared after 12 wk. This was probably the result of turnover of erythrocytes since the time of initiation of the fat diets; erythrocytes usually survive for 3-7 mo (Vacha 1979), and red blood cells are considered "old" (Garby and Hjelm 1963) after 80 d. In another study (Barnard et al. 1990), monkey erythrocyte fatty acid composition stabilized after 15 wk of consumption of trans fatty acids.

We have previously shown that the cholesterol + lard diet fed for 4 wk significantly increases plasma cholesterol levels in minipigs (Berlin et al. 1991). We observed similar increases in plasma cholesterol in this study (Berlin et al. 1998) after 8 wk of feeding the lard + cholesterol diet. However, after 2 wk of feeding either corn oil or menhaden oil, no differences were observed in either cholesterol or phospholipid levels of erythrocyte membranes from the animals fed the experimental diets. There was, however, a progressive significant increase in cholesterol and phospholipid during the course of the study. These increases with consumption of the high fat diets, together with the differences observed at 12 wk between the pigs fed lard + cholesterol and those fed the stock diet, demonstrate a significant effect of elevated dietary fat on membrane lipid composition regardless of dietary fatty acid composition.

We have observed effects of high fat diets, regardless of unsaturation, in previous in vivo studies with human females (Berlin et al. 1989) and in vitro studies with Chinese hamster ovary cells (Hannah et al. 1995). The women were fed diets containing 20 or 40% of energy as fat with polyunsaturated to saturated fatty acid (P/S) ratios of 0.3 or 1.0. Consumption of the higher PUFA and total fat diet resulted in drastic increases in erythrocyte 20:5 concentrations and smaller increases in 22:5 and 22:6 concentrations. Thus the combined modification of PUFA level and total fat consumption had specific effects on certain fatty acids. In this study, the experimental fat diets represent large increases in total fat compared with the stock diet normally eaten by the minipigs. Some of the unusual changes presently observed in the membrane concentrations of the long-chain (n-3) PUFA may describe specific effects of this large increase in total fat consumption.

The observed increase in membrane cholesterol may have occurred as part of a homeostatic mechanism (Schouten et al. 1984 and 1985) to maintain membrane integrity and overcome the excessive fluidity that could follow from the incorporation of a high concentration of unsaturated fatty acids into membranes. Compensatory incorporation of cholesterol and unsaturated fatty acids has been reported for erythrocytes in rabbits (Schouten et al. 1984) and in rats (Schouten et al. 1985).

We have examined the present data for influences of PUFA composition on erythrocyte membrane cholesterol by performing regression analyses of relationships between the ratio of molar cholesterol to phospholipid (C:P) and various parameters that describe unsaturation, i.e., mole percentages of (n-3) acids and (n-6) acids, total polyunsaturates, total monounsaturates and the unsaturation index, which is calculated as a function of the sum of the mole percentages of the unsaturated fatty acids multiplied by the number of olefinic double bonds.

Data from the samples taken at 23 wk were used because they should represent an equilibrium situation. The only relationship that approached significance (P < 0.06; r = 0.31) was an inverse one with the monounsaturate mole percentage according to the equation: C/P = 5.14  0.15 × (MUFA). When the analysis was performed by expressing the mass of cholesterol per gram of protein, the relation was significant (P < 0.03 and r = 0.41); however, the equation again indicated an inverse relationship between membrane cholesterol and unsaturation.

The factors that determine lipid domain fluidity (Smith 1987) include the molar ratio of cholesterol to phospholipid, the degree of unsaturation of phospholipid acyl chains and the ratio of phosphatidylcholine to sphingomyelin. In this study, we did not measure phosphatidylcholine or sphingomyelin. Regression analyses showed no significant relationship among the anisotropies for the three probes and the unsaturation index. No significant correlations were observed in heart plasma membranes from the same pigs for DPH-PA and TMA-DPH (data not shown). A similar lack of correlation between anisotropy and the unsaturation index was observed with liver plasma membranes (Banks et al. 1996) in stock-fed miniature swine in an aging study. Significant inverse relationships were observed in this study between DPH anisotropy and the polyunsaturate content, with more contribution from (n-6) fatty acids. Equations relating the steady-state anisotropy to the mole fractions, X, are as follows: The correlation between r_DPH and X_n-3 was not significant, but r_DPH followed a significant, direct relation to the mole fraction of monounsaturated fatty acid (X_MUFA), which is difficult to rationalize. Significant inverse relationships obtained for the polar probes were as follows: The DPH-PA anisotropy was directly related to the monounsaturated mole percentage; the TMA-DPH anisotropy was directly related to the (n-3) mole percentage.

The inverse relations between anisotropy and unsaturation represent a direct correlation between fluidity and unsaturation. Membrane fluidity results from local disordering of the bilayer because of the kinks induced by cis double bonds in the phospholipid fatty acid tails. The presence of additional multiple double bonds in the fatty acyl components of a membrane bilayer does not necessarily result in increased fluidity. The multiple double bonds in the (n-3) acids, EPA and DHA, could induce more ordering in the membrane, with their presence actually decreasing fluidity. Niebylski and Salem (1994) have reported additional ordering in phospholipids that contain DHA, as determined calorimetrically.

The higher anisotropy of the polar probes is related to the more restricted motion (Stubbs et al. 1984) experienced by these molecules. The TMA-DPH molecule is tethered at the interface, and it has a limited capacity to enter the hydrocarbon bilayer, thus describing a true conical volume of displacement during its rotation (Engel and Prendergast 1981). This contrasts with DPH, which has greater freedom of rotation. The correlations between TMA-DPH and DPH-PA anisotropies and fatty acid composition indicate that they reflect on fluidity associated primarily with EPA and DHA, which have double bonds in positions 5, 8, 11, 14 and 17 and in positions 4, 7, 11, 14, 17 and 20, respectively. Thus, it is possible that the polar probes could describe the effects of the double bonds in positions 4, 5, 7 and 8, whereas oleate and linoleate influence bilayer fluidity in the vicinity of carbons 9-12.

Cholesterol usually is thought of as reducing fluidity; however, regression analyses yielded the following equations: It is entirely reasonable that the cholesterol fluidizes the bilayer region made rigid by the (n-3) acids in proximity to the headgroups. Cholesterol modulates fluidity in either direction as appropriate for the fatty acid complement of the membrane bilayer.

Tissue fatty acids often are modified by dietary fats, as demonstrated in this study with respect to EPA and DHA incorporation into erythrocytes. Incorporation has also been demonstrated in liver and heart tissue (Berlin et al. 1994b) and in brain fractions (Berlin et al. 1998). Jones et al. (1995) described differences in fatty acid accretion in rats as a function of dietary fat and energy status. They observed that adipose tissue was more responsive to diet than was liver tissue, which in turn was more responsive than heart tissue. Erythrocyte DHA concentrations in suckling rats were enriched by DHA present in maternal diets (Carlson et al. 1986, Yeh et al. 1990). Carlson et al. (1986) found that erythrocyte and neural tissue DHA were decreased in neonatal rats fed a diet deficient in linolenic acid [18:3(n3)], and they suggested that red blood cell membrane DHA might be used to interpret the relative DHA status of human infants. Similarly, Uauy et al. (1990) and Liu et al. (1987) used erythrocyte DHA as an index of brain DHA. Innis (1992), however, stressed that phospholipid fatty acids in the circulation might not reflect the status of (n-6) and (n-3) PUFA in developing organs. Winters et al. (1994) cautioned against the use of erythrocyte DHA as an index of DHA status in tissues capable of in situ synthesis.

Our results underscore the need for caution in assuming that erythrocyte phospholipid fatty acids are indicators of phospholipid fatty acids in other organs and tissues, especially in different regions of the brain. This is especially important when animals have passed the point in development at which brain lipids are deposited. Second, unsaturated (n-3) fatty acids are less effective in increasing the fluidity of erythrocyte membranes than are (n-6) fatty acids.

	FOOTNOTES

Manuscript received 18 August 1997. Initial reviews completed 23 October 1997. Revision accepted 15 May 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Allain C. C., Poon L. S., Chan C.S.G., Richmond W., Fu P. C. Enzymatic determination of total serum cholesterol. Clin. Chem. 1974; 20:470-475[Abstract]
• Banks, M. A., Berlin, E., Johnson, W. A. & Peters, R. C. (1996) Vitamin E levels and susceptibility to lipid peroxidation increase with age in heart plasma membrane from miniature swine. J. Gerontol. Biol. Sci. 51A: B409-B416.
• Barnard D. E., Sampugna J., Berlin E., Bhathena S. J., Knapka J. J. Dietary trans fatty acids modulate erythrocyte membrane fatty acyl composition and insulin binding in monkeys. J. Nutr. Biochem. 1990; 1:190-195[Medline]
• Bartlett G. R. Phosphorus assay in column chromatography. J. Biol. Chem. 1959; 234:466-468[Free Full Text]
• Berlin E., Bhathena S. J., Judd J. T., Clevidence B., Peters R. C. Human erythrocyte membrane fluidity and insulin binding are independent of dietary trans fatty acid content. J. Nutr. Biochem. 1994a; 5:591-598
• Berlin E., Bhathena S. J., Judd J. T., Nair P. P., Jones D. Y., Taylor P. R. Dietary fat and hormonal effects on erythrocyte membrane fluidity and lipid composition in adult women. Metabolism 1989; 38:790-796[Medline]
• Berlin, E., Khan, M. A., Henderson, G. R. & Kliman, P. G. (1991) Dietary fat and cholesterol induced modification of minipig lipoprotein fluidity and composition. Comp. Biochem. Physiol. 98A: 151-157.
• Berlin E., Kim C. S., McClure D., Banks M. A., Peters R. C. Brain and heart membrane fatty acid composition in miniature swine fed diets containing corn and menhaden oils. Nutr. Res. 1998; 18:751-765
• Berlin, E., McClure, D., Banks, M. A. & Peters, R. C. (1994b) Heart and liver fatty acid composition and vitamin E content in miniature swine fed diets containing corn and menhaden oils. Comp. Biochem. Physiol. 109A: 53-61.
• Borenstain V., Barenholz Y. Characterization of liposomes and other lipid assemblies by multiprobe fluorescence polarization. Chem. Phys. Lipids 1993; 64:117-127[Medline]
• Bourre J. M., Pascal G., Durand G., Masson M., Dumont O., Piciotti M. Alterations in the fatty acid composition of rat brain cells (neurons, astrocytes, and oligodendrocytes) and of subcellular fractions (myelin and synaptosomes) induced by a diet devoid of n-3 fatty acids. J. Neurochem. 1984; 43:342-348[Medline]
• Carlson S. E., Carver J. D., House S. G. High fat diets varying in ratios of polyunsaturated to saturated fatty acid and linoleic to linolenic acid: a comparison of rat neural and red cell membrane phospholipids. J. Nutr. 1986; 116:718-725
• Charnock J. S., Abeywardena M. Y., McLennan P. L. Dietary lipids, membrane composition and cardiac function. Colloq. INSERM 1989; 195:135-144
• Dodge J. T., Mitchell C., Hanahan D. J. The preparation and chemical characteristics of hemoglobin-free ghosts of human erythrocytes. Arch. Biochem. Biophys. 1963; 100:119-130[Medline]
• Engel L. W., Prendergast F. G. Values for and significance of order parameters and cone angles of fluorophore rotation in lipid bilayers. Biochemistry 1981; 20:7338-7345[Medline]
• Engelhard V. H., Esko J. D., Storm D. R., Glaser M. Modification of adenylate cyclase activity in LM cells by manipulation of the membrane phospholipid composition in vivo. Proc. Natl. Acad. Sci. U.S.A. 1976; 73:4482-4486[Abstract/Free Full Text]
• Garby L., Hjelm M. Ultracentrifugal fractionation of human erythrocytes with respect to age. Blut 1963; 9:284
• Ghosh C., Dick R. M., Ali S. F. Iron/ascorbate-induced lipid peroxidation changes membrane fluidity and muscarinic cholinergic receptor binding in rat frontal cortex. Neurochem. Int. 1993; 23:479-484[Medline]
• Hannah J. S., Yamane K., Berlin E., Howard B. V. In vitro regulation of low-density lipoprotein receptor interaction by fatty acids. Metabolism 1995; 44:1428-1434[Medline]
• Hinds A., Sanders T.A.B. The effect of increasing levels of dietary fish oil rich in eicosapentaenoic and docosahexaenoic acids on lymphocyte phospholipid fatty acid composition and cell-mediated immunity in the mouse. Br. J. Nutr. 1993; 69:423-429[Medline]
• Innis S. M. Plasma and red blood cell fatty acid values as indexes of essential fatty acids in the developing organs of infants fed with milk or formula. J. Pediatr. 1992; 120:S78-S86[Medline]
• Jones P.J.H., Toy B. R., Cha M. C. Differential fatty acid accretion in heart, liver and adipose tissues of rats fed beef tallow, fish oil, olive oil and safflower oil at three different levels of energy intake. J. Nutr. 1995; 125:1175-1182
• Kitagawa S., Matsubayashi M., Kotani K., Usui K., Kametani F. Asymmetry of membrane fluidity in the lipid bilayer of blood platelets: fluorescence study with diphenylhexatriene and analogs. J. Membr. Biol. 1991; 119:221-227[Medline]
• Kuhry J.-G., Fonteneau P., Duportail G., Maechling C., Laustriat G. TMA-DPH: a suitable fluorescence polarization probe for specific plasma membrane fluidity studies in intact living cells. Cell Biophys. 1983; 5:129-140[Medline]
• Lentz B. R. Use of fluorescent probes to monitor molecular order and motions within liposome bilayers. Chem. Phys. Lipids 1993; 64:99-116[Medline]
• Leray C., Raclot T., Groscolas R. Positional distribution of n-3 fatty acids in triacylglycerols from rat adipose tissue during fish oil feeding. Lipids 1993; 28:279-284[Medline]
• Liu C.-C.F., Carlson S. E., Rhodes P. G., Rao V. S., Meydreck E. F. Increase in plasma phospholipid docohexaenoic acid and eicosapentaenoic acid as a reflection of their intake and mode of administration. Pediatr. Res. 1987; 22:292-296[Medline]
• Lokesh B. R., Hsieh H. L., Kinsella J. E. Alteration in the lipids and prostaglandins in mouse spleen following the ingestion of menhaden oil. Ann. Nutr. Metab. 1986; 30:357-364[Medline]
• Lowry O. H., Rosebrough N. J., Farr A. L., Randall R. J. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 1951; 193:265-275[Free Full Text]
• Makrides M., Neumann M. A., Simmer K., Gibson R. A. Erythrocyte fatty acids of term infants fed either breast milk, standard formula, or formula supplemented with long-chain polyunsaturates. Lipids 1995; 30:941-948[Medline]
• Matusik E. J., Reeves V. B., Flanagan V. P. Determination of fatty acid methyl esters. Elimination of tissue-derived contamination and artefacts and preparation of a support-coated acid modified polyester liquid phase. Anal. Chim. Acta 1984; 166:179-188
• Muriana F.J.G., Ruiz-Guttierez V. Effect of n-6 and n-3 polyunsaturated fatty acids ingestion on rat liver membrane-associated enzymes and fluidity. J. Nutr. Biochem. 1992; 3:659-663
• Niebylski C. D., Salem N. A calorimetric investigation of a series of mixed-chain polyunsaturated phosphatidylcholines: effect of sn-2 chain length and degree of unsaturation. Biophys. J. 1994; 67:2387-2393[Medline]
• SAS Institute Inc. (1988) SAS/STAT, Version 6.03, SAS Institute, Cary NC.
• Sassen L.M.A., Lamers J.M.J., Sluiter W., Hartog J. M., Dekkers D.H.W., Hogendoorn A., Verdouw P. D. Development and regression of atherosclerosis in pigs. Effects of n-3 fatty acids, their incorporation into plasma and aortic plaque lipids, and granulocyte function. Arterioscler. Thromb. 1993; 13:651-660[Abstract/Free Full Text]
• Schouten J. A., Beynen A. C., Popp-Snijders C., Mulder C. Effect of a hypocholesterolemic diet on the fatty acid composition of rat erythrocyte phospholipids. Nutr. Rep. Int. 1985; 31:229-236
• Schouten J. A., Beynen A. C., Popp-Snijders C., Van der Veen E. A. Lipid composition of erythrocytes isolated from cholesterol-fed rabbits. Nutr. Rep. Int. 1984; 29:1223-1232
• Shinitzky M., Barenholz Y. Fluidity parameters of lipid regions determined by fluorescence. Biochim. Biophys. Acta 1978; 515:367-394[Medline]
• Simopoulos A. P. Omega-3 fatty acids in health and disease and in growth and development. Am. J. Clin. Nutr. 1991; 54:438-463[Abstract/Free Full Text]
• Smith J. E. Erythrocyte membrane: structure, function, and pathophysiology. Vet. Pathol. 1987; 24:471-476[Medline]
• Sperry W. M., Brand F. C. The determination of total lipides in blood serum. J. Biol. Chem. 1955; 213:69-76[Free Full Text]
• Stubbs C. D., Kinosita K. Jr., Munkonge F., Quinn P. J., Ikegami A. The dynamics of lipid motion in sarcoplasmic reticulum membranes determined by steady-state and time resolved fluorescence measurements on 1,6-diphenyl-1,3,5-hexatriene and related molecules. Biochim. Biophys. Acta 1984; 775:374-380[Medline]
• Trotter P. J., Storch J. 3-[p-(6-Phenyl)-1,3,5-hexatrienyl]-phenylpropionic acid (PA-DPH): characterization as a fluorescent membrane probe and binding to fatty acid binding proteins. Biochim. Biophys. Acta 1989; 982:131-139[Medline]
• Uauy K. J., Birch D. G., Birch E. E., Tyson J. E., Hoffman D. R. Effect of dietary omega-3 fatty acids on retinal function of very-low-birth weight neonates. Pediatr. Res. 1990; 28:485-492[Medline]
• Vacha J. Critical comparative review of the life span of red cells in mammals. Acta Sci. Nat. Acad. Sci. Bohemoslov. Brno 1979; 13:1-47
• Winters B. L., Yeh S. M., Yeh Y. Y. Linolenic acid provides a source of docosahexaenoic acid for artificially reared rat pups. J. Nutr. 1994; 124:1654-1659
• Yeh Y. Y., Winters B. L., Yeh S. M. Enrichment of (n-3) fatty acids of suckling rats by maternal dietary menhaden oil. J. Nutr. 1990; 120:436-443

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences