Main

Arachidonic (20:4n-6) and docosahexaenoic (22:6n-3) acid are among the most abundant fatty acids in the CNS phospholipids. These fatty acids are found in high concentrations in brain synaptic plasma membrane and in photoreceptor cells(1–3). 20:4n-6 plays an important role as a precursor of biologically active molecules like prostanoids, leukotrienes, and other lipoxygenase products(4). 22:6n-3 is involved in providing a specific structural environment within the phospholipid bilayer that influences important membrane functions, such as ion or solute transport, receptor activity, and adenylate cyclase activity(5,6).

Research in infant nutrition has demonstrated that during the last trimester of gestation the fetal brain accrues fatty acids of the n-6 and n-3 types(7). These fatty acids may be derived from the placenta in utero with the formation of major neural tissue requiring approximately 43 mg of n-6 and 22 mg of n-3 fatty acids per week(7–9). The accretion of essential fatty acids in neural tissues is predominantly 20:4n-6 and 22:6n-3(7). It has also been estimated that requirements for n-6 and n-3 fatty acids in neuronal tissue synthesis can only be supplied from labile, hepatic fatty acid reserves for 9 and 2.3 d, respectively(9). The essential fatty acid reserves in the adipose tissue develop during the last trimester of fetal growth(9). Thus, if fetal development is interrupted by premature birth early in the third trimester, the hepatic and adipose reserves cannot meet whole body needs for essential fatty acids and total fat.

Quantitative analysis of the composition of human milk from mothers giving birth to preterm infants(9) indicates that mother's milk provides levels of 20:4n-6 and 22:6n-3 essential fatty acids approximating the predicted requirements at d 16 of life at oral intake levels of approximately 120 kcal/kg of body weight(9). Long-chain essential fatty acids are synthesized from 18:2n-6 or 18:3n-3. However, the amounts produced in vivo may be inadequate to support the accretion rates attained in breast-fed infants(10). Thus, it seems prudent to feed the preterm infant human milk or formulas with a fatty acid balance similar to human milk containing long-chain polyenoic homologs of 18:2n-6 and 18:3n-3(11).

Currently, infant formulas marketed in North America contain linoleic (18:2n-6) and linolenic (18:3n-3) acid and are devoid of 20:4n-6 and 22:6n-3(12,13). Therefore, infants who are fed these formulas must rely on in vivo elongation and desaturation of 18:2n-6 and 18:3n-3 to support the similar rate of accretion of 20:4n-6 and 22:6n-3 attained in breast-fed infants(9,10). It has been proposed(9,11) and recommended(14–17) that formulas fed to preterm infants be designed with a fatty acid balance similar to human milk containing 20:4n-6 and 22:6n-3. In the United Kingdom, Europe, South America, and Australia, 20:4n-6 and 22:6n-3 have been added to preterm infant formulas using single-cell oils and in Europe using phospholipids.

For infant formulas, the question persists if performed 22:6n-3 is needed or if providing more 18:3n-3 can be synthesized into 22:6n-3. In weanling rats, increasing dietary 18:3n-3, by decreasing 18:2n-6 to 18:3n-3 ratio from 7.3:1 to 4:1, increased the 22:6n-3 content in neuronal cell PE but not in other phospholipids from the cerebellum(18). Results from other studies using whole brain(19) or subcellular fractions(20) have also shown increases in 22:6n-3 content with increasing dietary 18:3n-3. However, none of these studies have investigated the effects of increasing 18:3n-3 on individual cell types from whole brain. Therefore, the present study used neonatal rat brain at 2 weeks of age, before the consumption of solid food, to test the hypothesis that increasing maternal dietary 18:3n-3 content from 1.6% (18:2n-6 to 18:3n-3 = 21.6:1) to 17.5% (18:2n-6 to 18:3n-3 = 1:1) of the total fatty acids will increase the 22:6n-3 content of neuronal cell phospholipids of rat pups. The results from this present study show that increasing maternal dietary 18:3n-3, by decreasing 18:2n-6 to 18:3n-3, ratio does not significantly increase the 22:6n-3 content in PC, PE, and PS of neuronal cell phospholipids of rat pups at 2 weeks of age.

METHODS

Animal care. The University of Alberta Animal Ethics Committee approved all animal procedures. Sprague-Dawley rats were obtained from the University of Alberta vivarium. During breeding, three females and one male were housed together for a 2-week mating period. Females were then moved to individual cages in a room maintained at 21 °C with a 12 h light and 12 h dark cycle. Water and food were supplied ad libitum. Laboratory rodent diet, 5001 (PMI Feeds, Inc., St. Louis, MO), was fed to the rats when not receiving experimental diets. Rats were switched to experimental diet on the day of parturition. All litters were culled to 12 pups within 24 h of parturition. Pups received only maternal milk. Pups were killed at 2 weeks of age.

One entire litter of rat pups fed the same diet was sexed and weighed before decapitation. Excised brains were placed in ice-cold 0.32 mol/L sucrose. Six brains from the same sex were pooled per sample. Stomach contents of three rats from each litter were also removed and analyzed for fatty acid composition to reflect the composition of maternal milk. Three litters per diet treatment were used.

Diets. The basal diet fed the rat pups meets all essential nutrient requirements and contained 20% (wt/wt) fat of varying 18:2n-6 and 18:3n-3 fatty acid composition(21). Diet fats were formulated to approximate the fatty acid composition of an existing infant formula providing an 18:2n-6 to 18:3n-3 ratio of 7.3:1. This fat blend served as the control fat treatment. Three experimental diets were formulated by addition of various triglycerides to alter the fatty acid composition of this control fat formulation (Table 1). An 18:2n-6 to 18:3n-3 fatty acid ratio of 21.6:1 was obtained by addition of corn oil to the diet fat blend. The 18:2n-6 to 18:3n-3 fatty acid ratio of 4.4:1 and 1:1 was obtained by the addition of flaxseed oil. These diets were nutritionally adequate, providing for all known essential nutrient requirements [ingredient and concentration (g/kg diet), respectively]: fat, 200; starch, 200; casein, 270; glucose, 207.65; nonnutritive fiber, 50; vitamin mix, 10; mineral mix, 50.85; L-methionine, 2.5; choline 2.75; and inositol, 6.25. The A.O.A.C. vitamin mix (Teklad Test Diets, Madison, WI) provided the following per kg of complete diet: 20 000 IU of vitamin A; 2000 IU of vitamin D; 100 mg of vitamin E; 5 mg of menadione; 5 mg of thiamine-HCl; 8 mg of riboflavin; 40 mg of pyridoxine-HCl; 40 mg of niacin; 40 mg of pantothenic acid; 2000 mg of choline, 100 mg of myoinositol; 100 mg of p-aminobenzoic acid; 0.4 mg of biotin; 2 mg of folic acid, and 30 mg of vitamin B12. Bernhart Tomarelli mineral mix (General Biochemicals, Chargin Falls, OH) was modified to provide 77.5 mg of Mn2+ and 0.06 mg Se2+ per kilogram of complete diet. To minimize any changes in sample composition due to fatty acid oxidation, the diets were sealed under nitrogen and stored in a freezer at -30°C in darkness. Every day the required amount of diet was taken out, mixed, and placed in individual feed cups.

Table 1 Fatty acid composition of experimental diets fed to lactating dams at 2 weeks of age
Full size table

Isolation of neuronal cells. Neuronal cells were isolated according to the method described by Sellinger and Azcurra(22). Briefly, pooled brains were placed in beakers containing 7.5% (wt/vol) polyvinylpyrrolidone and 10 mmol CaCl2/L at pH 4.7 and 25°C. Brain tissue was minced and poured into a 20 mL plastic syringe fitted with a reusable filter unit (Millipore, Swinnex disc holder, 25 mm). The sample was pressed three times through a series of combined nylon mesh filters. The final filtrate volume was adjusted, then layered on a 2-step sucrose gradient of 1.0 mol/L and 1.75 mol/L. Gradients were centrifuged in a Beckman SW-28 rotor at 41 000 × g for 30 min at 4°C.

Neuronal cell bodies were recovered in the pellet. Aliquots of cells were stained with methylene blue and examined for purity under a light microscope (Zeiss, 1600X). Gel electrophoresis and immunoblotting was performed to ensure purity of cell fractions prepared by these procedures(18). Proteins isolated from neuronal cells were compared by gel electrophoresis and immunoblotting to neurofilament and glial fibrillary acid protein standards. Neuronal cells isolated should only contain neurofilament proteins and not glial fibrillary acid proteins.

Lipid analysis. The neuronal cell lipid was extracted by a modified Folch method(23). Separation of individual phospholipids was completed on silica gel, thin-layer chromatography (TLC) H-plates (20×20 cm, Analtech, Newark, DE). The plates were developed in a solvent system containing chloroform:methanol:triethylamine:1-propanol:0.25% wt/vol KCl (30:9:18:25:6, by volume) for approximately 90 min(24). TLC plates were air-dried for 5 min and visualized with 0.1% (wt/vol) aniline naphthalene sulfonic acid in water.

Phospholipid fractions on the plate corresponding to standards were scraped into culture tubes. Fatty acid methyl esters were prepared with 14% wt/wt boron trifluoride in methanol following the method of Morrison and Smith(25).

Fatty acid analysis. Fatty acid methyl esters were separated by automated gas-liquid chromatography (Varian model 6000 GLC equipped with a Vista 654 data system and a Vista 8000 autosampler; Varian Instruments, Georgetown, ON), using a bonded fused silica BP20 capillary column (25 mm × 0.25 mm inside diameter) and quantitated using a flame ionization detector(26). These conditions are capable of separating methyl esters of saturated, cis-monounsaturated, and cis-polyunsaturated fatty acid from 14 to 24 carbons in chain length. Quantitation and identification of peaks was based on relative retention times compared with known standards (PUFA 1 and 2, bacterial methyl ester mix-14; Supelco Canada, Mississauga, Ontario, Canada)(26).

Statistical analysis

The effect of diet treatment and sex of rat pups on the fatty acid composition of neuronal cell phospholipid fractions was assessed by one-way analysis of variance (ANOVA) procedures using the SAS package, version 6.11(27). Significant differences between diet treatments and sex were determined by a Duncan's multiple range test at a significance level of p < 0.05 after a significant ANOVA(28). Values are expressed as mean ± SEM for n = 6.

RESULTS

Growth characteristics. No significant differences were observed between males and females for body weight, total brain weight, or the fatty acid composition of the individual phospholipid fractions. Thus, statistical analyses to test subsequent effects of diet treatments were combined for both sexes. Body and brain weight did not differ significantly between rat pups fed the four experimental diets. Final body weights were (mean ± SEM): 35.8 ± 0.9 g, 35.9 ± 1.0 g, 35.3 ± 0.8 g, and 35.6 ± 1.3 g for 21.6:1, 7.8:1, 4.4:1, and 1:1 diet treatments, respectively. Final brain weights were (mean ± SEM): 1.2 ± 0.1 g, 1.2 ± 0.1 g, 1.2 ± 0.1 g, and 1.1 ± 0.1 g for 21.6:1, 7.8:1, 4.4:1, and 1:1 diet treatments, respectively.

Purity of neuronal cell preparations. The neuronal cell preparations contained only minor cross-contamination (≈5%) from cell membrane fragments and microvessels as determined by microscopic examination. The presence of neurofilament in neuronal samples was previously verified by gel electrophoresis and immunoblotting(18). These results indicate that the cell preparation is primarily neuronal cell bodies with attached extensions.

Fatty acid composition of stomach contents. The fatty acid composition of stomach contents of rat pups was analyzed. These analyses reflected dams' milk composition(18,29,30). The increase in dietary 18:2n-6 or 18:3n-3 fed to the dams altered the stomach contents of the rat pups (Table 2), indicating that the range of dietary fat composition fed in the present experiment produced similar changes in the fat composition of dams' milk.

Table 2 The content of fatty acids in the stomach of rat pups at 2 weeks of age*
Full size table

Neuronal cells phospholipid fatty acid composition. In brain, PC and PE are quantitatively the two most abundant phospholipids and constitute approximately 90% of total brain phospholipid(31). The major fatty acids observed in PC were 16:0, 18:0, and 18:1 (47-52%, 13-14%, 15-16% of total fatty acids, respectively). Feeding a maternal diet providing a ratio of 18:2n-6 to 18:3n-3 of 7.8:1 produced the highest content of 20:4n-6 and 22:6n-3 in neuronal cells (Fig. 1). In PE, 18:0, 20:4n-6, 22:6n-3, and 16:0 (23-28%, 16-20%, 16-22%, and 13-17% of total fatty acids, respectively) were the predominant fatty acids. Increasing 18:3n-3 in the maternal diet did not significantly alter the 20:4n-6 and 22:6n-3 content in neuronal PE of the rat pups (p > 0.05) (Fig. 1).

Figure 1
figure 1

Effect of feeding rats diets varying in 18:2n-6 to 18:3n-3 on the fatty acid content on 20:4n-6 and 22:6n-3 in PC and PE of neuronal cells. Values for a fatty acid without a common letter differ significantly (p < 0.05). Values are mean ± SEM (%wt/wt) with n = 6 for each experimental diet.

Full size image

Analysis of the fatty acid profile in neuronal PS demonstrated that 18:0 and 22:6n-3 (38-40% and 14-28% of the total fatty acid, respectively) were the major fatty acids. The large content of 22:6n-3 in PS of the rat pups was significantly decreased by increasing the maternal dietary levels of 18:3n-3 (p < 0.0001). The maternal diet, providing a ratio of 18:2n-6 to 18:3n-3 ratio of 7.8:1, resulted in the highest level of 22:6n-3 in the rat pups (Table 3). The 16:0 content of PS increased from 6.9-20.1% of the total fatty acids when the 18:2n-6 to 18:3n-3 ratio was lowered from 7.8:1 to 1:1 (Table 3).

Table 3 Effect of varying 18:2n-6 to 18:3n-3 ratio on the fatty acid composition of neuronal phosphatidylserine*
Full size table

In neuronal PI, the major fatty acids were 18:0 and 20:4n-6 (28-37% and 19-34% of the total fatty acids, respectively). When increasing dietary levels of 18:3n-3 were provided in the maternal diet there was a significant increase in 22:6n-3 content of ≈11% from a dietary 18:2n-6:18:3n-3 ratio of 7.8:1 to 4.4:1 (p < 0.05) with a concomitant decrease in 20:4n-6 of ≈13.5% in PI in the rat pups (Table 4).

Table 4 Effect of varying 18:2n-6 to 18:3n-3 ratio on the fatty acid composition of neuronal phosphatidylinositol*
Full size table

DISCUSSION

The present study was initiated to determine the effects of increasing maternal dietary 18:3n-3 content by decreasing the 18:2n-6 to 18:3n-3 ratio from 21.6:1 to 1:1 in the 22:6n-3 content in neuronal cells of rat pups at 2 weeks of age. The results demonstrate that increasing maternal dietary 18:3n-3 content does not significantly increase the 22:6n-3 content of neuronal cell PC, PE, and PS of rat pups at a stage of brain development when 22:6n-3 is needed for rapid neural plasma membrane synthesis.

The reason for the similar 20:4n-6 and 22:6n-3 content in PC and PE of neuronal cells between the four experimental diets may be due to the fact that desaturase activity is age-related(32–35) and that, at 2 weeks of age, activity may be limited. Bourre et al.(35) demonstrated in rats that Δ-6 desaturase activity, a rate limiting step in 20:4n-6 and 22:6n-3 synthesis(36–38), varies during the first 21 d following gestation. Therefore, if Δ-6 desaturase activity is low at 2 weeks of age, the amount of 18:2n-6 and 18:3n-3 added in the diet would not have any significant effect on increasing the 20:4n-6 and 22:6n-3 content of membrane phospholipids.

The significant decrease in 22:6n-3 content in PS with increasing dietary 18:3n-3 may be attributed to different PS molecular species being produced by deacylation and reacylation processes(39).

PI represents approximately 4% of the total brain phospholipids(31). The deacylation of 20:4n-6 from PI followed by reacylation of PI with 22:6n-3 from other phospholipids such as PS could account for the decrease in 20:4n-6 and increase in 22:6n-3 content in PI when the 18:2n-6 to 18:3n-3 ratio was decreased from 7.8:1 to 4:4:1 (Table 4).

PS and PI, although minor phospholipids in brain membranes, are of special interest because both are involved in cellular functions(40). PS is responsible for the activation of several protein kinase C isoforms(41) while PI plays a key role in signal transduction(42) and production of eicosanoids(43). Moreover, the fatty acyl composition of PS and PI has been demonstrated to be one of the regulatory functions in enzyme activation(44). Therefore, although small on the basis of total brain content, the changes observed in PS in 22:6n-3 content and PI in 20:4n-6 and 22:6n-3 content could have functional implications.

In vivo studies have suggested that it is more effective to supply a dietary source of preformed 22:6n-3 to maintain the 22:6n-3 level in membrane phospholipids, rather than increasing the dietary content of 18:3n-3(45,46). The results from the present study appear to support these findings since increasing the dietary 18:3n-3 content by ≈11 fold (Table 1) did not significantly increase the 22:6n-3 content in neuronal cell PC, PE, and PS (Fig. 1 and Table 3).

The findings of this study may have important implications for neonatal feeding. If the present findings in neonatal rats are extrapolated to infants, it appears that increasing the 18:3n-3 content by decreasing the 18:2n-6 to 18:3n-3 ratio in preterm infant formulas will not stimulate an increase in levels of 22:6n-3 in the early neonatal period.