Main

The ω-6 and ω-3 fatty acids AA (C20:4ω-6) and DHA (C22:6ω-3) are important components of membrane phospholipids in the brain (1) and can be synthesized in vivo from the dietary precursors, linoleic (C18:2ω-6) and linolenic (C18:3ω-3) acid, respectively (2). AA and DHA are also available in the diet in fats of animal origin, and are provided to the breast-fed infant in human milk in concentrations representing approximately 0.5–0.7% and 0.2–0.4% milk fatty acids, respectively (3, 4). It is well known that AA and DHA are accumulated in large amounts in the brain during perinatal brain growth (5, 6), and many animal studies have shown that diets very low in C18:2ω-6 and C18:3ω-3 reduce brain AA and DHA and learning behaviors (7). Changes related to C18:3ω-3 fatty acid deficiency include reduced DHA in brain and retina, altered electroretinograms, decreased looking and VEP acuity, and changes in behavior (learning) in some, but not all, studies (7–13). Recently, lower auditory evoked brain stem response was reported for offspring of rats fed fish oil (providing DHA) in gestation (14). Despite the well-described changes in membrane fatty acids, however, the mechanism by which dietary ω-6 and ω-3 fatty acids alter learning, visual, and auditory functions has not been established.

It is well known that plasma and red blood cell AA and DHA concentrations in infants fed formulas with no AA or DHA are lower than in breast-fed infants or infants fed formulas supplemented with the latter long-chain ω-6 and ω-3 fatty acids (15–21). This, however, may simply reflect the difference in dietary AA and DHA intake, and is not ipso facto evidence of reduced brain AA and DHA. Studies using stable isotopes have shown preterm and term infants can desaturate C18:2ω-6 and C18:3ω-3 to AA and DHA, respectively (22–24), and brain astrocytes can synthesize AA and DHA (25). However, it is not clear whether the rate of C18:3ω-3 conversion in liver or brain is sufficient to meet the needs of growing tissues for DHA (23). Autopsy data show lower DHA, but not AA, in brain of infants fed formula rather than those who were breast-fed (26, 27), as well as higher VEP and looking acuity. Higher scores in tests of neurodevelopment in infants fed formula with DHA than in infants fed formulas without DHA (16–18, 28) is consistent with a hypothesis that dietary DHA is important for optimal brain DHA assimilation. On the other hand, lower brain DHA may be explained by inadequate dietary C18:3ω-3, or high C18:2ω-6/C18:3ω-3 resulting in inhibition of C18:3ω-3 desaturation (29, 30). Thus, it remains unclear as to whether dietary DHA has a specific role in developing brain, or whether the discrepancies among studies are explained by differences in the formula C18:3ω-3, C18:2ω-6/C18:3ω-3 ratio, or other confounding variables (31).

Recent studies from this laboratory demonstrated reduced dopamine, serotonin, and norepinephrine in frontal cortex of piglets fed formula low in C18:2ω-6 and C18:3ω-3 (32). Furthermore, inclusion of 0.4% AA plus 0.3% DHA in the formula prevented the decrease in monoaminergic neurotransmitters that resulted from feeding with the low C18:2ω-6 and C18:3ω-3 formula for only 18 d from birth (32). Cognitive and behavioral processes involve many brain regions, including striatum, hypothalamus, and hippocampus, in addition to frontal cortex. Superior colliculus is considered essential for visual-motor integration, whereas inferior colliculus is a source of auditory and tactile localization cues and vocalization (33–36). The current article reports the effects of feeding piglets a formula with low C18:2ω-6 and C18:3ω-3 or C18:2ω-6 and C18:3ω-3 at similar concentrations to infant formula, and the effects of AA and DHA supplementation on monoaminergic neurotransmitters in hippocampus, striatum, hypothalamus, and inferior and superior colliculus of formula-fed piglets. Results for frontal cortex are included for comparison and demonstrate the region- and neurotransmitter-specific effects of the C18:2ω-6 plus C18:3ω-3 and AA plus DHA content of formula. AA was included with DHA in the supplemented formula to avoid problems of reduced AA that may result from supplementation with only DHA (15, 37).

METHODS

Animals and diets.

Newborn male piglets weighing >1 kg at birth and <12 h old were assigned to be fed one of four formula diets (n = 6/group) which differed only in the composition of the fat (Table 1). Two of the formulas were low in C18:2ω-6 and C18:3ω-3, providing about 0.8% energy C18:2ω-6 and 0.05% energy C18:3ω-3 (low C18:2ω-6 + C18:3ω-3); the other two formulas had 8.3% energy C18:2ω-6 and 0.8% energy C18:3ω-3 (C18:2ω-6 + C18:3ω-3 adequate). One of each of the low and adequate C18:2ω-6 + C18:3ω-3 formulas had 0.2% energy (0.4% total fatty acids) AA and 0.16% energy (0.3% fatty acids) DHA. The AA and DHA were from single cell oils and were included in the formula during preparation by Nestle Research Centre, Lausanne, Switzerland. Other details concerning the formulas and feeding have been reported (32). Littermates were not assigned to the same diet. The procedures involving the piglets were approved by the Animal Care Committee of the University of British Columbia and conformed to the guidelines of the Canadian Council on Animal Care.

Table 1 Major fatty acid components of formulas varying in linoleic and linolenic acid without or with arachidonic and docosahexanoic acid from single cell (SC) oil Results are given in grams per 100 g.
Full size table

Sample collection.

Piglets were anesthetized at 18 d of age (ketamin hydrochloride, 37.5 mg/kg, MTC Pharmaceuticals, Cambridge, Canada; and xylazine hydrochloride, 3.75 mg/kg, Bayvet Division, Chenango, Etobicoke, Canada), then killed by intracardiac injection of 200 mg/kg pentobarbital. The brain was rapidly removed and weighed, and the frontal cortex, hippocampus, striatum, superior colliculus, inferior colliculus, and hypothalamus were dissected and frozen in liquid nitrogen. All of the samples were stored at −80°C until analysis.

Biochemical analysis.

Frontal cortex (100 mg), hippocampus (40 mg), striatum (40 mg), hypothalamus (50 mg), superior colliculus (40 mg), and inferior colliculus (30 mg) were homogenized (Sonics Materials Inc, Danbury, CT, U.S.A.) in 760 μL of perchloric acid (0.1 mol/L) with 40 μL of 3,4-dihydroxybenzylamine (3 μg/mL) as an internal standard, then centrifuged at 172,381 ×g for 1 h at 4°C. The supernatant was then used for analysis of monoamine concentrations by high-pressure liquid chromatography, as described in detail (32).

Frontal cortex, hippocampus, and striatum lipids were extracted (38), then PC, PE, PI, and PS were separated by thin-layer chromatography, as recently described (31). The separated phospholipids were recovered, and their fatty acid components were separated, identified, and quantified by gas-liquid chromatography (39). Inadequate sample was available from sufficient individual piglets for phospholipid fatty acid analysis of hypothalamus and superior and inferior colliculus.

Statistical analyses.

Results were compared between the groups using two-way ANOVA with the level of C18:2ω-6 + C18:3ω-3 (low or adequate) and no addition or addition of AA and DHA as the main effects. Statistically significant differences among the four groups were then determined using Fisher's least significant difference only with ANOVA results with p < 0.05.

RESULTS

The effects of the amount of C18:2ω-6 + C18:3ω-3, and the inclusion of single cell oils with AA and DHA in the formulas on dopamine, serotonin, and norepinephrine, and their metabolites, in different regions of the piglet brain are shown in Table 2. The levels of neurotransmitter pathways are heterogeneous in different areas, and as demonstrated by the results, affected differently by dietary ω-6 and ω-3 fatty acids. The concentrations of dopamine and its metabolites DOPAC and HVA and of serotonin and its metabolite 5-HIAA were all significantly lower in frontal cortex of piglets fed the low C18:2ω-6 + C18:3ω-3 diet than in piglets fed the other three diets. In contrast, despite the much higher concentrations of dopamine, DOPAC, and HVA in striatum than frontal cortex, the amount of ω-6 and ω-3 fatty acids in the formula had no statistically significant effect on striatum dopamine or its metabolites. An interaction was found between the amount of C18:2ω-6 + C18:3ω-3 and the addition of AA + DHA on the concentrations of serotonin in the striatum. The individual group comparisons indicated that feeding AA + DHA decreased serotonin in the striatum of piglets fed the C18:2ω-6 + C18:3ω-3 adequate formula, but not in those fed the low C18:2ω-6 + C18:3ω-3 formula (Table 2). Concentrations of dopamine and 5-HIAA in both the inferior and superior colliculus were significantly higher in piglets fed the formulas with AA and DHA than in piglets fed the formulas without AA + DHA. The individual group comparisons showed that the concentrations of dopamine and 5-HIAA in the superior colliculus and 5-HIAA in the inferior colliculus were significantly increased when AA and DHA were added to either the formula with low C18:2ω-6 + C18:3ω-3 or with adequate C18:2ω-6 + C18:3ω-3. The concentrations of dopamine in the inferior colliculus was statistically significantly increased when AA + DHA was included in the formula with adequate C18:2ω-6 + C18:3ω-3, but the higher dopamine in the inferior colliculus of piglets fed the low C18:2ω-6 + C18:3ω-3 formula with versus without AA + DHA was not of statistical significance.

Table 2 Monoaminergic neurotransmitter concentrations in frontal cortex, striatum, hippocampus, hypothalamus, and superior and inferior colliculus in 18-d-old piglets fed formula varying in linoleic and linolenic acid without or with arachidonic and docosahexanoic acid from single cell (SC) oil Values are reported in micromoles per gram of tissue, and represent mean ± SEM, n = 6; concentrations are reported only for monoamines at >0.05 μmol/g. 1 main effect of amount of C18:2ω-6 + C18:3ω-3 2 main effect of inclusion of SC oil with AA C20:4ω-6 + DHA C22:6ω-3 3 significant interactions between the amount of C18:2ω-6 + C18:3ω-3 and SC oils; values within a row with a different letter superscriptabc are significantly different, p < 0.05.
Full size table

The major focus of this study was analysis of changes in monoaminergic neurotransmitters and their metabolites caused by including AA + DHA in formula with different amounts of C18:2ω-6 and C18:3ω-3. Sufficient tissue was available to separate and analyze PC, PE, PI, and PS only for frontal cortex, hippocampus, and striatum, for which the concentrations of AA and DHA are shown in Figure 1. These analyses show both phospholipid class and region-specific effects of the formula C18:2ω-6 + C18:3ω-3 and AA (C20:4ω-6) + DHA (C22:6ω-3) content. Inclusion of 0.4% AA and 0.3% DHA in the low C18:2ω-6 + C18:3ω-3 formula resulted in significantly higher frontal cortex AA in PC and PI and DHA in PC, PE, PS, and PI; higher hippocampus DHA in PC, PE, and PS; and higher striatum AA in PS. Inclusion of AA and DHA in the C18:2ω-6 and C18:3ω-3 adequate formula, on the other hand, had no significant effect on the frontal cortex AA, but resulted in lower AA in PC, PE, and PI in hippocampus, and higher AA in PS in striatum. Inclusion of AA + DHA in the formula with adequate C18:2ω-6 and C18:3ω-3 also increased DHA in PE and PI in frontal cortex, and increased AA in PS in striatum.

Figure 1
figure 1

Concentration of AA (C20:4ω-6) and DHA (C22:6ω-3) in frontal cortex, hippocampus, and striatum of piglets fed formula low or adequate in C18:2ω-6 + C18:3ω-3 without (−) or with (+) AA + DHA from single cell oils (SC). Results are presented as mean + SEM and were analyzed by ANOVA; numbers at top of each graph indicate significant main effects of (2) amount of C18:2ω-6 + C18:3ω-3, (3) amount of AA + DHA, and (4) the interaction. Groups with a different letter (a, b, c, d) are significantly different at p < 0.05. FA, fatty acid.

Full size image

DISCUSSION

The results of this study establish that the dietary intake of ω-6 and ω-3 fatty acids during the first 18 d after birth can have marked and diverse effects on monoaminergic neurotransmitters in brain regions involved in learning, behavior, and the integration and transmission of visual and auditory information. The decreased concentrations of dopamine, serotonin, and norepinephrine in the frontal cortex of piglets given approximately 1% dietary energy C18:2ω-6 and 0.05% C18:3ω-3, and the prevention of this decrease by feeding 0.25% energy AA + 0.15% DHA, have recently been described (32).

Of particular significance, the studies in this report show that supplementation of the diet fed to piglets with AA and DHA at levels similar to those in human milk (3, 4) resulted in increased dopamine and 5-HIAA, the metabolite of serotonin, in the superior and inferior colliculus in the formula-fed piglet. In contrast to the other brain regions analyzed, supplementation of the formula with AA + DHA increased dopamine in both the superior and inferior colliculus of piglets fed a formula with C18:2ω-6 + C18:3ω-3 concentrations similar to those in current infant formulas in North America. Linolenic acid (C18:3ω-3) represented 1.5% of the formula fatty acids, about 0.8% energy, with a C18:2ω-6/C18:3ω-3 ratio of 10:1 in the formula with adequate C18:2ω-6 and C18:3ω-3. Thus, it is possible that higher intakes of C18:3ω-3, or a lower C18:2ω-6/C18:3ω-3, would have facilitated higher dopamine and 5-HIAA concentrations with no further effect of supplementation with AA and DHA. Insufficient superior and inferior colliculus was available for separation and analysis of phospholipid fatty acids. However, supplementation of the C18:2ω-6 + C18:3ω-3 adequate formula with AA + DHA either increased or had no effect on AA and DHA in striatum and frontal cortex phospholipids, but decreased AA in PC, PE, and PI and decreased DHA in PE and PS in hippocampus. These results show that potential diet-induced changes in membrane fatty acids in superior and inferior colliculus cannot be estimated from the diet-induced changes in other brain regions, but require specific analysis. The demonstration that the effects of dietary ω-6 and ω-3 fatty acids is heterogeneous among different brain regions has been noted by others (40). Possible explanations include the stage of development in different regions when the dietary treatment was imposed, the relative proportions of myelin and nonmyelin membranes in the different regions, or potentially regional differences in AA and DHA synthesis, or membrane phospholipid synthesis and turnover. However, it needs to be considered that this study did involve analysis of neurotransmitters and fatty acids from multiple regions of the brain, raising the possibility of a type 1 error.

Visual function has been assessed in infants in relation to the amounts of DHA in milk and formula diets using behavioral measures of looking acuity and VEP, usually in response to counter-phase checkerboard or grating stimulation. Several studies have found lower VEP and looking acuity in preterm and term gestation infants fed standard formula than in infants fed formulas supplemented with DHA (16–18, 20) in similar amounts to that used in the studies here with piglets. Another study, however, has not found differences (15). The VEP represents part of the activity of the visual cortex in response to information passed by the retina and the geniculostriate path (33, 41), with the cortical visual threshold estimated by extrapolation of the measures to derive the acuity limit (41). Behavioral measures of looking acuity, on the other hand, involve transmission of the visual information, integration, and a motor response of the infant in orienting toward the grating stimuli (42). More recently, studies with young rats found that maternal dietary fish oil reduced auditory evoked brain stem responses in offspring (14), but 2.5% dietary DHA in the milk fed after birth increased auditory brain responses (S. de la Presa Owens, S.M. Innis, unpublished data). The superior colliculus, which was analyzed here, is a laminated mass of gray fibers in the mesencephalon that receives and is involved in the conveyance of visual information and is essential in visual motor integration, such as orienting behavior, and avoidance reaction, and in facilitating shifts in gaze (33, 34, 36). The inferior colliculus, on the other hand, is involved in processing of auditory information and visual and tactile localization cues (33, 35). Furthermore, monoamines are known to be involved in both visual and auditory pathways (43–45). The demonstration in our study that supplementing formula with AA + DHA increased superior and inferior colliculus dopamine and 5-HIAA suggests differences in visual and auditory responses, and behavioral tests of learning that involve integration of visual cues in infants and animals fed diets varying in ω-3 fatty acids (9–12, 16–18, 20, 28, 46) may involve changes in dopamine or serotonin metabolism.

Inasmuch as our studies involved supplementation with both AA + DHA, it is not possible to conclude a specific role of DHA in monoamine metabolism. However, lower dopamine and serotonin receptor binding, and lower dopamine concentrations in frontal cortex (47, 48) and reduced dopamine in synaptic vesicles after a learning task (49, 50) have been reported for second-generation rats fed a C18:3ω-3 fatty acid–deficient diet. Thus, information is available to suggest dietary ω-3 fatty acids may alter brain monoamine metabolism at multiple levels, possibly involving synthesis, receptor-mediated neurotransmitter release and uptake, and vesicle storage.

In summary, this study has shown that the amount of C18:2ω-6 + C18:3ω-3, and inclusion of small amounts of AA and DHA, in formula fed to neonatal piglets for as little as 18 d from birth had diverse and region-specific effects on monoaminergic neurotransmitters and phospholipid fatty acids. It seems possible that the functional changes associated with dietary ω-6 or ω-3 fatty acid deprivation or supplementation are likely to be the consequence of altered neurotransmitter metabolism. The marked decrease in dopamine, serotonin, and norepinephrine after dietary C18:2ω-6 and ω-3 fatty acid restriction strongly argues against early nutrition without adequate provision of these fatty acids in the parenteral or enteral diet. The increase in dopamine and 5-HIAA in superior and inferior colliculus of piglets fed formula with low C18:2ω-6 + C18:3ω-3 or with C18:2ω-6 + C18:3ω-3 concentrations typical of infant formula suggests dietary AA or DHA may have a specific role in dopamine and serotonin metabolism in areas involved in transmission of visual and auditory information. Whether changes in dopamine or serotonin metabolism are involved in the changes in visual or auditory function, or performance in developmental tests found in some studies with infants and animals fed ω-3 fatty acid–modified diets, remains to be determined.