The Journal of Nutrition Vol. 128 No. 6 June 1998, pp. 1011-1017

Dietary Structured Triacylglycerols Containing Docosahexaenoic Acid Given from Birth Affect Visual and Auditory Performance and Tissue Fatty Acid Profiles of Rats^1,2

Merete M. Christensen, Søren P. Lund^*, Leif Simonsen^*, Ulla Hass^*, Svend E. Simonsen^*, and Carl-Erik Høy³

Center for Food Research, Department of Biochemistry and Nutrition, The Technical University of Denmark, DK-2800 Lyngby, Denmark and ^* National Institute of Occupational Health, Department of Toxicology and Biology, DK-2100 Copenhagen Ø, Denmark

	ABSTRACT

Abstract Introduction Methods Results Discussion References

To examine whether it is possible to enhance the level of 22:6(n-3) in the central nervous system, newborn rats were fed dietary supplements containing oils with either specific or random triacylglycerol structure, but similar concentrations of polyunsaturated fatty acids. In the specific structured oil, 22:6(n-3) was located in the sn-2 position, whereas it was equally distributed among the three positions in the triacylglycerol molecule in the randomized oil. A reference group was fed rat milk before weaning and nonpurified diet after weaning. After 12 wk, the levels of 22:6(n-3) in brain and liver phospholipids were higher in the groups fed the experimental diets than in the reference group. The specific structured oil resulted in the highest level of 22:6(n-3) in the brain, whereas the level of 22:6(n-3) was highest in the liver of the group fed randomized oil, indicating differences in metabolism of fatty acids resulting from their position in the dietary triacylglycerol molecule. The higher levels of 22:6(n-3) were accompanied by significantly lower levels of the long-chain (n-6) polyunsaturated fatty acids compared with the reference group. The fatty acid profiles, including the level of 22:6(n-3), in the retina phospholipids were not affected by the three different diets apart from a lower level of 20:4(n-6) in rats fed the experimental diets, indicating a strong tendency to maintain a high level of 22:6(n-3) in the retina. The changes in the fatty acid profiles did not result in differences in learning ability, but caused changes in visual function, evidenced by higher latency of the b-wave and lower oscillatory potential, and in auditory brainstem response, evidenced by generally greater amplitude of wave Ia in the group fed specific structured oil.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Large amounts of docosahexaenoic acid [22:6(n-3)] are incorporated into the structural lipids of the developing central nervous system. In the phospholipids of the photoreceptor outer segment membrane in the retina and in the synaptic membranes in the brain, 22:6(n-3) may constitute from 35 to 60% of the fatty acids (Neuringer et al. 1988). The largest accretion of the polyunsaturated fatty acids (PUFA)⁴ in the nervous system takes place during the last trimester and in the first 6-10 mo after birth in humans (Clandinin et al. 1980), although it continues up to 2 y of age (Martinez, 1992). In rats, the major deposition takes place during the first 15 d (Dobbing and Sands 1979), which makes the newborn rat a relevant model for the preterm infant with respect to brain development. The fetus receives the (n-3) PUFA from the mother through the placenta, and the newborn mammal obtains them from the maternal milk. The optimal level of 22:6(n-3) in the diet and hence in the brain phospholipids is unknown. However, because reduced amounts of (n-3) PUFA in the central nervous system caused by dietary deprivation of these fatty acids have been correlated with reduced visual ability in rhesus monkeys (Connor et al. 1990) and with effects on learning ability in rats (Bourre et al. 1989), a high level of these fatty acids may be beneficial for the newborn.

In this study, we have investigated the effect of a high intake of 22:6(n-3) in rats starting at birth. The rats received a liquid diet containing either an oil with a specific triacylglycerol structure, with 22:6(n-3) in the sn-2 position, or a randomized oil in which 22:6(n-3) was equally distributed in the triacylglycerol molecule. The location in the sn-2 position may enhance the absorption of the particular fatty acid (Jensen et al. 1994). After weaning, the rats were fed solid diets containing the same oils. A reference group received ordinary rat milk, followed by a nonpurified rat diet after weaning. To investigate possible effects of the dietary contents of 22:6(n-3) in the brain and retina, visual and auditory performance as well as learning ability of the rats were tested after 10 wk of consuming the diets. The rats were then killed and fatty acid profiles of brain, retina, and liver phospholipids and adipose tissue triacylglycerols were examined.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Synthesis of the specific structured and the randomized oil.  The specific structured oil was made from fish oil triacylglycerol, which was reesterified with free fatty acids from butter fat catalyzed by Lipozyme, a 1,3-specific lipase (Novo Nordisk A/S, Bagsværd, Denmark) (Elliott and Parkin 1991, Haraldsson et al. 1989). In this process, a triacylglycerol with long-chain polyunsaturated fatty acids from the fish oil in the sn-2 position, and predominantly short- and medium-chain fatty acids from the butter oil in the sn-1 and sn-3 positions was produced. The randomized oil was produced from the structured oil by sodium methoxide catalyzed interesterification (Zeitoun et al. 1993). The fatty acid profile of triacylglycerols was determined by gas liquid chromatography (GLC) (Table 1). The fatty acid composition in the sn-2 position was determined by Grignard degradation (Jensen et al. 1994), and the fatty acid composition in the sn-1 and sn-3 positions was calculated.

Animals.  The experiments were approved by the Danish Animal Experiments Inspectorate. Pregnant Wistar rats, purchased from Møllegård Breeding Centre A/S, LI. Skensved, Denmark, were housed in plastic cages with free access to tap water. The temperature was maintained at 24°C, the relative humidity was 45%, with a 12-h light:dark cycle. After delivery, the litters were reduced to 6 or 8 pups, and dams and pups were divided into three groups: One group of 4 dams and 24 pups received diet with specific structured oil; a similar group received diet with randomized oil, and a reference group of 3 dams with 8 pups each received nonpurified diet (Altromin no. 1324, Ringsted, Denmark).

After birth the pups stayed with their dams during the first 2 d. At 0800 h on d 3, pups and dams in the experimental groups were separated, and the pups were placed under a heating device maintained at 32°C. At 1000, 1300 and 1600 h, the pups were weighed and then tube fed the liquid diet (see below) until their weight had increased by 0.1 g (0.3 g/d). This weight gain was gradually increased to 0.3 g per feeding during the experimental period (0.9 g/d). Immediately after the last feeding at 1600 h, the pups were returned to their dams where they stayed until the next morning. This feeding regimen continued until the pups were weaned at the age of 14 d when they received a solid diet (see below) containing the same oils as the liquid diets. The reference pups stayed with their dams throughout the feeding period and were transferred to rat nonpurified diet at the time of weaning.

At the age of 7 wk, the rats were transferred to the National Institute of Occupational Health and accustomed to a reverse light:dark cycle suitable for obtaining the electroretinograms (ERG). During wk 9, the rats were tested in a Morris water maze (see below); during wk 10 and 11, ERG and auditory brainstem response (ABR) tests were performed. During wk 12, all rats were decapitated, and brain, liver, adipose tissue located dorsal to the kidneys and retina were immediately excised and frozen in liquid nitrogen.

Liquid diets.  Powder for liquid diet (Code 189 without added fat, Altromin, Lange, Germany), was mixed with water and the manufactured oil in the ratio: 60 g powder:15 g oil:0.15 L water, which resulted in the following composition: carbohydrate, 143 g/kg diet; fiber, 5 g/kg diet; protein, 84 g/kg diet; and fat, 67 g/kg diet.

Solid diets.  After weaning, the rats received 20 g diet/d as described in Table 2.

Analyses of tissue lipids.  Brain, liver, adipose tissue and retina were extracted according to Folch et al. (1957) (retinas from four rats were pooled). Phospholipid classes in brain and liver were separated by TLC and methylated with BF₃ (Morrison and Smith 1964). Triglyceride from adipose tissue was isolated by TLC (Høy et al. 1983) and methylated with KOH (Christopherson and Glass 1969). The fatty acid methyl esters were analyzed by GLC by using a Fisons Instruments HRGC Mega 2 series 8560 gaschromatograph with FID (Fisons Instruments, Milan, Italy) and a BPX-70 fused silica capillary column (25 m × 0.22 mm i.d.) (SGE, Victoria, Australia). On column injection was used. Initial oven temperature was 40°C for 2 min, followed by a temperature gradient of 10°C/min to 150°C, followed by a gradient of 3°C/min to 215°C, followed by a gradient of 5°C/min to a final temperature of 220°C, which was maintained for 5 min. Identification was made by comparison with actual standards (Nu-Chek-Prep, Elysian, MN).

Morris water maze.  Testing of spatial learning in a Morris water maze was performed as described earlier (Hass et al. 1995). The maze consisted of a black plastic pool, 100 cm in diameter and 40 cm deep. Four directional points on the rim of the pool designated N, E, S and W were used as starting points and for dividing the pool into four arbitrary quadrants. The pool was filled to a depth of 30 cm with tap water at room temperature (20°C). A circular platform (diameter 10 cm) was situated on a solid support 1 cm below the surface of the water.

The latency to reach the platform was measured by stopwatch, and the route was automatically tracked by means of a wide-angle lens video camera connected to a computer with an image-analysis program. The maximum trial latency was 60 s. The rats were tested in blocks of four trials using the four starting points assigned in a pseudorandom sequence. Two cages with four or five rats were placed near the pool. Each rat was started at the first point, tested and then returned to the cage. The same procedure was repeated for the other three starting points. When the rat swam to and climbed onto the platform, the trial was ended. If the animal failed to locate the platform within 60 s, it was gently guided to the platform, left there for 15 s and then returned to the cage.

The following scheme was used: 1) training: with the platform situated in the centre of the SW quadrant, the animals were tested in three blocks of four trials, one block per day on three consecutive days; and 2) new platform location: with the platform situated in the center of the NE quadrant opposite the original location, the animals were tested in one block of four trials.

Electroretinography (ERG).  Animals were anesthetized with Mebumal (7 mg/100 g body weight, intraperitoneal) after dark adaption for a minimum of 6 h. The following procedures were performed under dim red light illumination. A drop of tetracainhydrochloride (10 g/L) in saline was applied to the left eye for local anesthesia. The rat was placed in an isolated box of 50 × 60 × 70 cm and kept on a thermostatically controlled plate; rectal temperature was kept constant at 38.8 ± 0.2°C during the recordings. The luminance of the flashes was measured with a photometer (Mastersix, Gossen A.G., Erlangen, Germany). A photo stimulator (Type 3G22, NEC San-ei Instruments, Tokyo, Japan), providing 100 µs flashes, was placed behind an opaque acrylic plate, 25 cm in front of the rat. The plate covered 60° of the visual field. A contact lens electrode (Kyoto Contact Lens, Kyoto, Japan) was placed on the left eye of the rat and a reference electrode was placed in the mouth with good contact to the mucous membrane. The electrodes were connected to an amplifier (Dantec 15C01 EMG-amplifier, Dantec, Copenhagen, Denmark), and the ERG waveforms were sampled on a 486 computer at a 4-kHz sampling rate by a DT2821 data acquisition board (DATA TRANSLATION, Malborough, MA) using the Asyst 4.0 software package (Keithley, Taunton, MA).

After the mounting of the electrodes, at least 3 min passed before the recordings, which were performed in the dark. The recordings were divided into four parts using luminance of 0.015, 1.5, 55 and again 0.015 cd/m². Series of six recordings were made at each luminance, with intervals of 10, 30, 60 and 10 s, respectively. A 3-min pause was inserted before the last series of recordings. The amplitudes and the peak latencies of the a- and b-waves were measured on the ERG-recordings after filtering with a digital zero-phase lowpass filter with a cut-off frequency of 54 Hz. The oscillatory potentials (OP) were measured at the 55 cd/m² recordings after filtering with a zero-phase highpass filter with a cut-off frequency of 86 Hz.

Auditory brain response (ABR).  The ABR was measured immediately after the ERG in the same anesthesia period. A silver wire placed subcutaneously at the back of the head served as an active electrode with the same reference electrode as in the ERG measurements.

The general procedures applied have been previously described (Simonsen and Lund 1995); however, some modifications were introduced. The signals for the single-frequency tonebursts (4, 8, 16 and 32 kHz from 25 to 95 dB sound pressure level (SPL) in 10-dB steps) were generated by a 16-bit digital signal processor (Ariel DSP16+, Ariel Corporation, Cranbury, NJ) with a 125-kHz sampling frequency, amplified by a computer-controlled audio amplifier (NAD2100) with a 44-kHz second-order lowpass input filter, and emitted from a piezoelectrical horn tweeter (type KSN 1016, Motorola, Huntsville, AL). The sound intensities at the individual frequencies were calibrated using a Brüel and Kjær peak sound level meter (type 2218) equipped with an octave filter (type 1613, Brüel & Kjær, Copenhagen, Denmark). After amplification by an EMG-amplifier (Dantec 15C01 EMG amplifiers), the ABR was sampled on a 486 DX2 computer at a 46.5-kHz sampling rate with a 16-bit data acquisition board (DAS-HRES, Keithley, MA) and the ASYST 4.0 software package. Instead of using the highpass filter of the amplifier, the ABR was digitally filtered by a zero-phase Butterworth filter with a cut-off frequency of 2.5 kHz and a stopband of 4.0 kHz.

The latency and the amplitude of peak Ia, the latency of peak III as well as the latency between peak Ia and peak III were measured at the frequencies tested. In addition, the auditory thresholds, defined as the lowest sound pressure level at which component I of the recorded ABR could be identified by visual inspection, were determined for each rat.

Statistics.  Fatty acid determinations and ERG recordings were analyzed by one-way ANOVA, and means were compared by Newman-Keuls Multiple Range Test with a 5% level of significance (Montgomery 1991). ABR recordings were treated as repeated measurements within each frequency and were tested by one-way ANOVA of repeated measurements using the general linear hypothesis. Water maze data were tested similarly to the ABR recordings. Values in the text are means ± SD.

	RESULTS

Abstract Introduction Methods Results Discussion References

Body weights.  The weight gains did not differ in the experimental groups and were less than in the reference group, resulting in higher body weight of the reference group from weaning and throughout the study period. Final body weights were: reference group, 254 g ± 71 g; randomized oil group, 193 g ± 42 g; and specific structured oil group, 209 g ± 56 g.

Brain phospholipids.  In the experimental groups the levels of 22:6(n-3) and other long-chain (n-3) fatty acids in brain phosphatidyl ethanolamine (PE) and phosphatidyl serine (PS) were higher (P < 0.05) than in the reference group. This was accompanied by significantly lower levels of long-chain (n-6) fatty acids (Table 3). One exception was 20:3(n-6), which was higher in the experimental groups. A higher level of 22:6(n-3) was found in brain PE in the group of rats fed structured oil compared with randomized oil, and no other differences were found between the two experimental groups. In brain phosphatidyl choline (PC) and phosphatidy inositol (PI), similar differences among the dietary groups were found (data not shown). The levels of the saturated and monounsaturated fatty acids did not differ among the groups.

Liver phospholipids.  The major differences were in the polyunsaturated fatty acids. The levels of the long-chain (n-3) fatty acids were higher (P < 0.05) in the experimental groups compared with the reference group. This was accompanied by reduced levels (P < 0.05) of the (n-6) fatty acids (Table 4). The level of 20:4(n-6) in particular was greatly reduced in the experimental groups, from 23% and 22% to 6.5% and 3.5% in PE and PS, respectively. Similar differences were observed for liver PC (data not shown).

Retina phospholipids.  No significant difference was seen in the levels of 22:6(n-3) in the retina phospholipids (Table 5), whereas small differences were found between the reference and the experimental groups in the levels of (n-3) and (n-6) fatty acids, as also observed in the brain phospholipids.

Adipose tissue.  No differences between the experimental groups were observed. The fatty acid composition of the dietary fats influenced the adipose tissue triacylglycerols (Table 5). Long-chain (n-3) PUFA were significantly higher in the experimental groups, as were the levels of 20:1 and all saturated and monounsaturated fatty acids with chain length  18. The levels of 18:2(n-6), 18:3(n-3) and long-chain (n-6) PUFA were higher (P < 0.05) in the reference group compared with the experimental groups.

Morris water maze.  No significant differences were found among the groups in the spatial learning test (data not shown).

ERG.  An example of the electroretinograms obtained is shown in Figure 1, with the a- and b-wave and the oscillatory potential (OP) indicated. The peak latency of wave b was significantly higher at a luminance of 1.5 cd/m² (P < 0.05) in the rats fed structured oil compared with both the reference group and the rats fed randomized oil (Table 6). The amplitude of the oscillatory potentials was significantly smaller (P < 0.05) in both experimental groups compared with the reference group (Table 6). The amplitude and peak latency of wave a, the amplitude of wave b and the peak latency of the oscillatory potential were not significantly different between the groups (data not shown).

View larger version (9K): [in this window] [in a new window]   Fig 1. (A) Single unfiltered electroretinogram (ERG) recorded after a flash stimulation of 55 cd/m2 given at time F. The amplitudes and latency of the a- and b-wave of the ERG are indicated. (B) The same ERG as A, but filtered with a zero-phase highpass filter with a cut-off frequency of 86 Hz leaving the oscillatory potentials (OP).

ABR.  The hearing thresholds were not significantly different among the groups, but the amplitude of wave Ia at 8 and 16 kHz was significantly higher (P < 0.05) in the rats fed structured oil than in the other groups at sound intensities of 95, 85 and 75 dB, and at 8 kHz also at 65 dB (Table 7). This difference was also found at 4 and 32 kHz at the higher sound pressure levels (data not shown). No other significant differences were found among the three groups.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

In most mammalian species, the level of long-chain polyunsaturated fatty acids from the (n-3) family, primarily 22:6(n-3), is very high (from 35 to 60% in PE and PS) in the central nervous system, especially in the synaptic membranes in the brain and the photoreceptor membrane in the retina (Neuringer and Connor 1986). These fatty acids accumulate during the embryonic period and in the first months of life in human infants (Clandinin et al. 1980), but the optimal level is unknown.

In this experiment, we have tried to increase the accumulation of 22:6(n-3) during development in a rat model by dietary intervention from birth, and subsequently investigated the effect by learning ability tests and electrophysiologic measurements of nerve function.

In both brain and liver, long-chain (n-3) PUFA replaced (n-6) PUFA in the experimental groups compared with the reference group. A small but significantly higher level of 20:3(n-6) in the brain of the experimental groups compared with the reference group after 12 wk of consuming the diets indicated a possible inhibition of the conversion of 20:3(n-6) to 20:4(n-6) by the  5 desaturase, suggested also by the significantly lower level of 20:4(n-6) in these groups, caused by the high level of (n-3) fatty acids in the diet (Rioux and Innis 1992, Yonekubo et al. 1993).

The difference in the structure of the dietary triacylglycerols seemed to influence the brain phospholipids after 12 wk of consuming the diets. In all brain phospholipids, the level of 22:6(n-3) was higher in the group fed structured oil (significant only in PE), suggesting a higher accessibility of the fatty acids in the sn-2 position of the triacylglycerol. In the liver, however, a higher level of 22:6(n-3) was found in the group fed randomized oil, suggesting differences in the metabolism of fatty acids related to their positions in the dietary triacylglycerols. We have recently demonstrated that the distribution of PUFA in dietary triacylglycerols is reflected in the chylomicron triacylglycerols following intestinal resynthesis (Christensen and Høy 1996). After intake of randomized triacylglycerols, the chylomicron triacylglycerols may thus be poor substrates for lipoprotein lipase, due to the presence of 22:6(n-3) in the 1- and 3-positions, allowing 22:6(n-3) to accumulate in the chylomicron remnant and return to the liver.

The level of 22:6(n-3) in the retina phospholipids was not influenced by the diet, and only the level of 22:5(n-3) was greater in the experimental groups after 12 wk. Significantly lower levels of all of the long-chain (n-6) fatty acids were found in these groups compared with the reference group, which was compensated by higher levels of 16:0 and 18:1. These results suggest that the retina maintained a certain level of 22:6(n-3), which was not the case for the long-chain fatty acids of the (n-6) family. Levels of 22:6(n-3) in the retina as low as 7% in PE compared with 36% in control groups were found by Neuringer et al. (1986) in extreme deficiency; this is important when investigating the correlation between fatty acid composition and visual performance, for example, but of no relevance in the determination of optimal dietary (n-3) intake. At high dietary levels of 22:6(n-3) as in this experiment, a further increase in the intake does not affect the incorporation of (n-3) PUFA into the retina. Under these circumstances, an optimal level in the retina may have been reached.

The differences found in the adipose tissue fatty acid composition were largely reflections of the differences in the dietary fatty acid composition as also has been found by others (Sheppard and Herzberg 1992), but the significantly higher levels of the long-chain (n-3) fatty acids after the experimental diets suggest that a surplus of 22:6(n-3) was present.

The differences in brain levels of 22:6(n-3) and 20:4(n-6) did not result in any differences in the learning ability in the Morris water maze. Larger differences in fatty acid composition such as those found in (n-3) deficiency, would therefore be necessary to introduce differences in learning ability (Lamptey and Walker 1976, Yamamoto et al. 1987). Yonekubo et al. (1993 and 1994), however, showed that higher levels of 22:6(n-3) in brain PE and PS during gestation, caused by giving diet supplemented with fish oil to the dam, resulted in a better performance of the young rats (6 wk old) in a swimming test. The increased levels of 22:6(n-3) in the fetus brains, did not, however, persist in the brains of the neonates; at wk 7 postnatally, no differences in the brain levels of 22:6(n-3) were found between the fish oil group and the control group. The control diet contained 2.0% 18:3(n-3) and had a ratio of 18:3(n-3)/18:2(n-6) of 0.1, which may be low for optimal incorporation of long-chain (n-3) fatty acids when no preformed long-chain (n-3) PUFA are present in the diet (Arbuckle et al. 1992, Clark et al. 1991, ISSFAL 1994).

There were no differences in the a-wave [the hyperpolarization of the photoreceptors after a flash of light (Ikeda 1987)] among the groups, indicating that the photoreceptor processes were unaffected. The peak latency of the b-wave [the depolarization of the on-bipolar cells and the Müller cells (Ikeda 1987, Xu and Karwoski 1995)] in the group fed structured oil was altered, and the size of the OP (elicited by the bipolar and amacrine cells (Garner and Lee 1994) in the experimental groups were significantly less than those of the controls. In the analysis of both the b-wave and the OP, the differences in body weight had very little influence on the differences between groups. The observed changes in the b-wave and the OP, when the a-wave is unchanged, are caused by changes in the inner nuclear layer of the retina (Folk 1991) and reflect changes in the signal transduction between the neurons as well as interactions between neurons and glial cells. The retina is part of the central nervous system, and the changes in the nuclear and internal plexiform layer of the retina may reflect changes found throughout the central nervous system. The changes in the ERG of the rats in the experimental groups demonstrated that the high level of polyunsaturated fatty acids may have an effect on the physiologic processes in the brain.

Wave I of the rat ABR is usually split in two separate waves, Ia and Ib, (Chen and Chen 1991), with wave Ia representing the first wave of the compound action potential in the acoustic nerve (Chen and Chen 1991). The group fed specific structured oil had a higher amplitude of wave Ia than the other groups. The higher amplitude of wave Ia may result from either enhanced excitability in the hair cells of the cochlea or in the first neuron of the auditory pathway, i.e., the bipolar neurons of the spinal ganglion. Enhanced exitability has been demonstrated in adult rats exposed to 800 ppm dearomatized white spirit for 6 mo (Lund et al. 1996). The physiologic importance of the effect in this study is unclear; it may, however, turn out to be an untoward effect. An increase in auditory brainstem conduction time has thus been reported in rat pups after 29 d intake of milk from dams fed a diet containing 22% of energy as fat with a content of 22:6(n-3) corresponding to 6% of the total dietary fat (Stockard et al. 1997).

The differences in the electrophysiologic tests between the group fed structured oil and the other two groups are not related to differences in body weight because the two experimental groups had similar weights. The major differences in fatty acid composition were found between the experimental groups and the reference group, and only the higher level of 22:6(n-3) in brain PE differentiated the group fed structured oil from the group fed randomized oil. It is not clear if the level of 22:6(n-3) has any influence on the electrophysiologic variables, but it is important to determine whether the major replacement of (n-6) PUFA by (n-3) PUFA in the central nervous system has any undesirable effects. It has been demonstrated previously, for example, that a reduced level of 20:4(n-6) in plasma in humans is accompanied by a reduced growth rate (Carlson et al. 1991 and 1992, Koletzko and Braun 1991), although the mechanism remains unclear.

	ACKNOWLEDGMENTS

Kirsten Kjær from Thyborøn Andelsfiskeindustri provided the fish oil used in the experiment. We thank Trine Brix Jensen for excellent assistance with the lipid analysis and Jørgen Holm-Jensen and Maria Hammer for excellent assistance with the ERG/ABR measurements and the water maze tests.

	FOOTNOTES

Manuscript received 14 November 1997. Initial reviews completed 25 November 1997. Revision accepted 3 February 1998.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Aaes-Jørgensen E., Hølmer G. Essential fatty acid-deficient rats: I. Growth and testes development. Lipids 1969; 4:501-506
• Arbuckle L. D., Rioux F. M., MacKinnon M. J., Innis S. M. Formula alfa-linolenic (18:3 (n-3)) and linoleic (18:2 (n-6)) acid influence neonatal piglet liver and brain saturated fatty acids, as well as docosahexaenoic acid (22:6 (n-3)). Biochim. Biophys. Acta 1992; 1125:262-267[Medline]
• Bourre J.-M., Francois M., Youyou A., Dumont O., Piciotti M., Pascal G., Durand G. The effects of dietary -linolenic acid on the composition of nerve membranes, enzymatic activity, amplitude of electrophysiological parameters, resistance to poisons and performance of learning tasks in rats. J. Nutr. 1989; 119:1880-1892
• Carlson S. E., Cooke R. J., Rhodes P. G., Peeples J. M., Werkman S. H. Effect of vegetable and marine oils in preterm infant formulas on blood arachidonic and docosahexaenoic acids. J. Pediatr. 1991; 120:S159-S167
• Carlson S. E., Cooke R. J., Werkman S. H., Tolley E. A. First year growth of preterm infants fed standard compared to marine oil n-3 supplemented formula. Lipids 1992; 27:901-907[Medline]
• Chen T. J., Chen S. S. Generator study of brainstem auditory evoked potentials by a radiofrequency lesion method in rats. Exp. Brain Res. 1991; 85:537-542[Medline]
• Christensen M. S., Høy C.-E. Effects of dietary triacylglycerol structure on triacylglycerols of resultant chylomicrons from fish oil- and seal oil-fed rats. Lipids 1996; 31:341-344[Medline]
• Christopherson S. W., Glass R. L. Preparation of milk fat methyl esters by alcoholysis in an essentially nonalcoholic solution. J. Dairy Sci. 1969; 52:1289-1290[Abstract/Free Full Text]
• Clandinin M. T., Chappell J. E., Leong S., Heim T., Swyer P. R. Intrauterine fatty acid accretion rates in human brain: implications for fatty acid requirements. Early Hum. Dev. 1980; 4:121-129[Medline]
• Clark K. J., Makrides M., Neumann M. A., Gibson R. A. Determination of the optimal ratio of linoleic acid to alfa-linolenic acid in infant formulas. J. Pediatr. 1991; 120:S151-S158
• Connor W. E., Neuringer M., Lin D. S. Dietary effects on brain fatty acid composition: the reversibility of n-3 fatty acid deficiency and turnover of docosahexaenoic acid in the brain, erythrocytes, and plasma of rhesus monkeys. J. Lipid Res. 1990; 31:237-247[Abstract]
• Dobbing J., Sands J. Comparative aspects of the brain growth spurt. Early Hum. Dev. 1979; 3:79-83[Medline]
• Elliott J. M., Parkin K. L. Lipase-mediated acyl-exchange reactions with butteroil in anhydrous media. J. Am. Oil Chem. Soc. 1991; 68:171-175
• Folch J., Lees M., Sloan-Stanley G. H. A simple method for the isolation and purification of total lipides from animal tissues. J. Biol. Chem. 1957; 226:497-509[Free Full Text]
• Folk, G. (1991) Retinal physiology. In: Principles and Practice of Clinical Electrophysiology of Vision (Heckenlively, J. R. & Arden, G. B., eds.), p. 69. Mosley Yearbook, St. Louis, MO.
• Garner C. D., Lee E. W. Evaluation of methanol-induced retinotoxicity using oscillatory potential analysis. Toxicology 1994; 93:113-124[Medline]
• Haraldsson G. G., Höskuldsson P. A., Sigurdsson S. T., Thorsteinsson F., Gudbjarnason S. The preparation of triglycerides highly enriched with n-3 polyunsaturated fatty acids via lipase catalysed interesterification. Tetrahedron Lett. 1989; 30:1671-1674
• Hass U., Lund S. P., Simonsen L., Fries A. S. Effects of prenatal exposure to xylene on postnatal development and behavior in rats. Neurotoxicol. Teratol. 1995; 17:341-349[Medline]
• Høy C.-E., Hølmer G., Byrjalsen I., Kirstein D. Acyl group distributions in tissue lipids of rats fed evening primrose oil (gamma-linolenic plus linoleic acid) or soybean oil (alpha-linolenic plus linoleic acid). Lipids 1983; 18:760-771[Medline]
• Ikeda, H. (1987) Retinal mechanisms and the clinical electroretinogram (1987). In: A Textbook of Clinical Neurophysiology (Halliday, A. M., Butler, S. R. & Paul, R., eds.), p. 569. John Wiley & Sons, New York, NY.
• ISSFAL The ISSFAL Board Statement "Recommendations for the essential fatty acid requirements for infant formulas." International Society for the Study of Fatty Acids and Lipids. ISSFAL NEWSLETTER 1994; 1:4-5
• Jensen M. M., Christensen M. S., Høy C.-E. Intestinal absorption of octanoic, decanoic, and linoleic acids: effect of triglyceride structure. Ann. Nutr. & Metab. 1994; 38:104-116[Medline]
• Koletzko B., Braun M. Arachidonic acid and early human growth: is there a relation? Ann. Nutr. & Metab. 1991; 35:128-131[Medline]
• Lamptey M. S., Walker B. L. A possible essential role for dietary linolenic acid in the development of the young rat. J. Nutr. 1976; 106:86-93
• Lund S. P., Simonsen L., Hass U., Ladefoged O., Lam H. R., Østergaard G. Dearomatized white spirit inhalation exposure causes long-lasting neurophysiological changes in rats. Neurotoxicol. Teratol. 1996; 18:67-76[Medline]
• Martinez M. Tissue levels of polyunsaturated fatty acids during early human development. J. Pediatr. 1992; 120:S129-S138[Medline]
• Montgomery, D. C. (1991) Design and Analysis of Experiments, 3rd ed. John Wiley & Sons, New York, NY.
• Morrison W. R., Smith L. M. Preparation of fatty acid methyl esters and dimethylacetals from lipids with boron fluoride-methanol. J. Lipid Res. 1964; 5:600-608[Abstract]
• Neuringer M., Anderson G. J., Connor W. E. The essentiality of n-3 fatty acids for the development and function of the retina and brain. Annu. Rev. Nutr. 1988; 8:517-541[Medline]
• Neuringer M., Connor W. E. n-3 Fatty acids in the brain and retina: evidence for their essentiality. Nutr. Rev. 1986; 44:285-294[Medline]
• Neuringer M., Connor W. E., Lin D. S., Barstad L., Luck S. Biochemical and functional effects of prenatal and postnatal omega3 fatty acid deficiency on retina and brain in rhesus monkeys. Proc. Natl. Acad. Sci. U.S.A. 1986; 83:4021-4025[Abstract/Free Full Text]
• Rioux F. M., Innis S. M. Arachidonic acid concentrations in plasma and liver phospholipid and cholesterol esters of piglets raised on formulas with different linoleic and linolenic acid contents. Am. J. Clin. Nutr. 1992; 56:106-112[Abstract/Free Full Text]
• Sheppard K., Herzberg G. R. Triacylglycerol composition of adipose tissue, muscle and liver of rats fed diets containing fish oil or corn oil. Nutr. Res. 1992; 12:1405-1418
• Simonsen L., Lund S. P. Four weeks inhalation exposure to n-heptane causes loss of auditory sensitivity in rats. Pharmacol. Toxicol. 1995; 76:41-46[Medline]
• Stockard, J., Carver, J. D., Benford, V. J., Nelson, R. M. & Barness, L. (1997) The effect of varying levels of dietary docosahexaenoic acid (DHA) upon rat pup auditory brainstem conduction times (ABCTs). Pediatr. Res. 41: 241A (abs).
• Xu X., Karwoski C. J. Current source density analysis of retinal field potentials II. Pharmacological analysis of the b-wave and M-wave. J. Neurophys. 1995; 72:96-105[Abstract/Free Full Text]
• Yamamoto N., Saitoh M., Moriuchi A., Nomura M., Okuyama H. Effect of dietary -linolenate/linoleate balance on brain lipid compositions and learning ability of rats. J. Lipid Res. 1987; 28:144-151[Abstract]
• Yonekubo A., Honda S., Okano M., Takahashi K., Yamamoto Y. Dietary fish oil alters rat milk composition and liver and brain fatty acid composition of fetal and neonatal rats. J. Nutr. 1993; 123:1703-1708
• Yonekubo A., Honda S., Okano M., Takahashi K., Yamamoto Y. Effects of dietary fish oil during the fetal and postnatal periods on the learning ability of postnatal rats. Biosci. Biotechnol. Biochem. 1994; 58:799-801
• Zeitoun M.A.M., Neff W. E., List G. R., Mounts T. L. Physical properties of interesterified fat blends. J. Am. Oil Chem. Soc. 1993; 70:467-471

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