Regulation of Rat Brain Polyunsaturated Fatty Acid (PUFA) Metabolism during Graded Dietary n-3 PUFA Deprivation

Abstract

Knowing threshold changes in brain lipids and lipid enzymes during dietary n-3 polyunsaturated fatty acid deprivation may elucidate dietary regulation of brain lipid metabolism. To determine thresholds, rats were fed for 15 weeks DHA-free diets having graded reductions of α-linolenic acid (α-LNA). Compared with control diet (4.6% α-LNA), plasma DHA fell significantly at 1.7% dietary α-LNA while brain DHA remained unchanged down to 0.8% α-LNA, when plasma and brain docosapentaenoic acid (DPAn-6) were increased and DHA-selective iPLA₂ and COX-1 activities were downregulated. Brain AA was unchanged by deprivation, but AA selective-cPLA₂, sPLA₂ and COX-2 activities were increased at or below 0.8% dietary α-LNA, possibly in response to elevated brain DPAn-6. In summary, homeostatic mechanisms appear to maintain a control brain DHA concentration down to 0.8% dietary DHA despite reduced plasma DHA, when DPAn-6 replaces DHA. At extreme deprivation, decreased brain iPLA₂ and COX-1 activities may reduce brain DHA loss.

INTRODUCTION

The brain is enriched in docosahexaenoic acid (DHA, 22:6n-3) and arachidonic acid (AA, 20:4n-6), which are critical for its normal structure and function [1–3]. In vertebrates, these polyunsaturated fatty acids (PUFAs) cannot be synthesized de novo from 2-carbon fragments, but can be elongated in liver (minimally in brain or heart) from their respective shorter-chain PUFA precursors, α-linolenic acid (α-LNA, 18:3n-3) and linoleic acid (LA, 18:2n-6) [4–7]. In humans, a low dietary n-3 PUFA intake or a low plasma DHA concentration has been correlated with increased risk for neuropsychiatric and/or neurodegenerative diseases [8, 9]. Dietary n-3 PUFA supplementation may be beneficial in these conditions [8, 10].

Multiple animal studies have been conducted to understand how dietary-derived n-3 PUFAs influence body integrity and metabolism. For example, in rats fed a DHA-free diet containing α-LNA at 4.6% total fatty acid, brain, heart and liver DHA concentrations are sufficient to maintain organ function, so this diet is considered n-3 PUFA "adequate" [2, 11]. In contrast, in rats fed a DHA-free diet containing 0.2% α-LNA, brain DHA concentrations are reduced, behavior is disturbed and brain derived neurotrophic factor (BDNF) is reduced compared with the 4.6% α-LNA diet, so this diet is considered n-3 PUFA “inadequate" or "deficient" [2, 12, 13]. Brain changes in rats fed this deficient diet include a prolonged DHA half-life; an increased concentration of docosapentaenoic acid (DPAn-6, 22:5n-6), an AA elongation product; reduced expression of enzymes that regulate DHA metabolism, Ca²⁺-independent phospholipase A₂ (iPLA₂ Type VI, iPLA₂β) [14–17] and cyclooxygenase (COX)-1 [18, 19]; and increased expression of enzymes that regulate AA metabolism, cytosolic cPLA₂ Type IV, secretory sPLA₂ Type II and COX-2 [14, 20].

The brain lipid and enzyme changes in animals exposed to dietary n-3 PUFA deprivation, noted above and reported elsewhere [3, 21, 22], may not be clinically relevant because deprivation was too severe and prolonged, sometimes spanning several generations. This severity also limits the ability to identify causes and effects. To overcome these limitation, in the present study we exposed rats after weaning to 15 weeks of graded reductions in dietary n-3 PUFA content below the 4.6% α-LNA "adequate" level, and estimated when statistically significant changes in different lipid parameters first appeared (thresholds) in plasma, brain and liver.

MATERIALS AND METHODS

Materials

1-Palmitoyl-2-[1-¹⁴C] arachidonoyl-sn-glycerol-3-phosphorylcholine was purchased from PerkinElmer (Boston, MA, USA) and had a specific activity of 60 mCi/mmol. 1-Palmitoyl-2-[1-¹⁴C] palmitoyl-sn-glycerol-3-phosphorylcholine was purchased from GE Healthcare (Buckinghamshire, UK) and had specific activity of 53 mCi/mmol. The purity of each was > 95%, as determined by TLC, scintillation counting and GC. 1-Palmitoyl-2-arachidonoyl-sn-glycerol-3-phosphorylcholine, 1-palmitoyl-2-[1-¹⁴C] palmitoyl-sn-glycerol-3-phosphorylcholine and phosphatidylinositol 4, 5-bisphosphate were obtained from Avanti (Alabaster, AL, USA), protease inhibitor cocktail from Roche (Indianapolis, IN, USA). A high capacity cDNA reverse transcription kit, Taqman® gene expression master mix, and specific primers for real time RT-PCR were purchased from Applied Biosystems (Foster City, CA, USA).

Animals

Fischer-344 (CDF) male rat pups (19 days old) and their surrogate mothers, purchased from Charles River Laboratories (Portage, MI, USA), were housed in an animal facility with regulated temperature, humidity, and a 12 h light/12 h dark cycle. Lactating rats had free access to water and rodent chow formulation NIH-31 18-4 (Zeigler Bros, Gardners, PA, USA), which contained 4% (wt/wt) crude fat and (as percent total fatty acid) α-LNA (5.1%), eicosapentaenoic acid (20:5n-3) (2.0%), DHA (2.3%), LA (47.9%) and AA (0.02%) [23].

After nursing for 21 days, the pups were divided randomly into six groups and placed on a predetermined diet. They had free access to food and water, and their food was replaced every 2 or 3 days. After 15 weeks on a given diet, they were asphyxiated by CO₂ inhalation and decapitated. The brains and liver were excised rapidly and frozen in 2-methylbutane cooled by dry ice to −50 °C, then stored at −80 °C until used. Blood was collected from the abdominal aorta with EDTA, and centrifuged at 1,500 rpm for 5 min. Plasma was removed and stored at 80 °C until assayed. The protocol was approved by the Animal Care and Use Committee of the Eunice Kennedy Schriver National Institute of Child Health and Human Development, and followed the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publication No. 80-23).

Graded n-3 PUFA diets

The different n-3 PUFA diets, prepared by Dyets Inc. (Bethlehem, PA, USA), were based on the AIN-93G formulation [24, 25]. Each diet contained 10% crude fat, but a different amount of flaxseed oil. Fatty acid composition of each diet (µmol/g food, percent total fatty acid, or percent energy) is shown in Table 1. The n-3 PUFA “adequate” diet contained 7.8 µmol/g α-LNA (4.6 % of total fatty acid) [2]. The extreme “deficient” diet contained 0.25 µmol/g α-LNA (0.2% total fatty acid). The less deficient diets contained α-LNA at 3.8, 2.6, 1.7, or 0.8 % of total fatty acid. Other n-3 PUFAs were absent from all diets. Each diet contained 40 µmol/g LA (23–24% total fatty acid).

Table 1.

Composition of graded n-3 PUFA diets.

Ingredient		4.6 % ALA diet	3.8 % ALA diet	2.6 % ALA diet	1.7% ALA diet	0.8 % ALA diet	0.2 % ALA diet
				g/100 g food
Carbohydrate		60	60	60	60	60	60
Protein		20	20	20	20	20	20
Fat		10	10	10	10	10	10
Hydrogenated coconut oil		6.00	6.12	6.25	6.37	6.50	6.62
Safflower oil		3.23	3.26	3.29	3.32	3.35	33.8
Flaxseed oil		0.77	0.62	0.46	0.31	0.15	0
Other ingredient		10	10	10	10	10	10
Fatty acid concentration				% of total fatty acid concentration
12:0		28.8	30.3	30.7	31.6	32.4	34.0
14:0		12.5	12.9	13.1	13.3	14.0	14.6
14:1n-5		0.04	0.04	0.03	0.04	0.05	0.05
16:0		9.5	9.6	9.6	9.5	9.8	9.9
16:1n-7		0.04	0.04	0.04	0.06	0.06	0.13
18:0		8.0	7.9	8.1	8.3	9.3	8.0
18:1n-9		8.4	7.7	8.0	7.6	7.1	5.9
18:2n-6		27.9	27.7	27.9	27.9	27.5	27.2
18:3n-3		4.6	3.8	2.6	1.7	0.8	0.2
Total saturated		58.8	60.7	61.5	62.7	64.5	66.5
Total Monounsaturated		8.5	7.8	8.0	7.7	7.2	6.1
n-6 PUFA		27.9	27.7	27.9	27.9	27.5	27.2
n-3 PUFA		4.6	3.8	2.6	1.7	0.8	0.2
n-6/n-3		4.2	5.1	7.0	11.8	23.4	132.9
Fatty acid content1				µmol/g diet
18:2n-6		40.5 ± 3.6	40.6 ± 1.8	44.1 ± 3.0	38.1 ± 0.8	41.1 ± 5.0	35.8 ± 2.7
18:3n-3		6.7 ± 0.3	5.5 ± 0.6	4.0 ± 0.6	2.4 ± 0.1	1.2 ± 0.2	0.21 ± 0.03

Data represent averages or means ± SD of 3 analyses

Lipid extraction and methylation

Methods of lipid extraction and methylation have been described [6, 19]. Total lipids from brain, liver and plasma were extracted by the Folch procedure [26], and fatty acids were transmethylated with 0.1% H₂SO₄-methanol for 3 h at 70 °C. Appropriate quantities of di-17:0 PC for total fatty acid analysis and of unesterified 17:0 for unesterified fatty acids were added as internal standards before transmethylation to fatty acid methyl esters.

Gas chromatography

Fatty acid methyl esters from brain and liver (nmol/g wet wt) and from plasma (nmol/ml plasma) were quantified with a gas chromatograph (6890N, Agilent Technologies, Palo Alto, CA, USA) equipped with an SP™-2330 fused silica capillary column (30 m × 0.25 mm i.d., 0.25 µm film thickness) (Supelco, Bellefonte, PA, USA) and a flame ionization detector [27]. Fatty acid concentrations were calculated by proportional comparison of peak areas to the area of the 17:0 internal standard.

Total RNA isolation and real time RT-PCR

Total RNA was isolated from brain using a commercial kit (RNeasy Lipid Tissue Kit; Qiagen, Valencia, CA, USA). cDNA was prepared from total RNA using a high-capacity cDNA Archive Kit (Applied Biosystems). mRNA levels of cPLA₂ IV (Rn 00591916_m1), sPLA₂ II (Rn 00580999_m1), iPLA₂ VI (Rn 01504424_m1), COX-1 (Rn 00566881_m1), COX-2 (Rn 00568225_m1), 5-LOX (Rn 00563172_m1), 12-LOX (Rn 01461082_m1), 15-LOX (Rn 00696151_m1) and BDNF (Rn 01484928_m1) were measured by real time quantitative RT-PCR, using the ABI PRISM 7000 sequence detection system (Applied Biosystems). The fold change in gene expression was determined by the ΔΔC_T method [28]. Data are expressed as relative level of the target gene in the n-3 PUFA deficient diet group normalized to the endogenous control (β-globulin, Rn_00560865_m1) and relative to the level with the n-3 PUFA “adequate” diet (calibrator). Each experiment was performed in triplicate with 6 independent samples per diet group.

Phospholipase A₂ activities

A radioactivity method designed by the Dennis Group [29–31] was used to determine brain cPLA₂ and iPLA₂ activities. A commercial kit (Cayman, Ann Arbor, MI, USA) was used to measure sPLA₂ activity.

Sample preparation

Brain tissue was homogenized with 3 vol homogenization buffer (10 mM HEPES, pH 7.5 containing 1 mM EDTA, 0.34 M sucrose and protease inhibitor cocktail (Roche)) using a glass homogenizer. The homogenized sample was centrifuged at 100,000 g for 1 h at 4 °C, and the supernatant was used for all PLA₂ enzyme activity analyses. Supernatants were kept at −80 °C until use. Protein concentration was determined by the Bradford assay [32] (Bio-Rad, Hercules, CA, USA).

Enzyme assay for radioisotope method

Concentrations were composed for a final incubation volume of 0.5 ml. For cPLA₂ activity analysis, the cytosolic fraction (0.3 mg protein in one assay) was mixed with 100 mM HEPES, pH 7.5 containing 80 µM Ca²⁺, 2 mM dithiothreitol, 0.1 mg/ml fatty acid-free bovine serum albumin in 450 µl, then incubated with 50 µl substrate solution of 100 µM 1-palmitoyl-2-arachidonoyl-sn-glycerol-3-phosphorylcholine and phosphatidylinositol 4,5-bisphosphate (97:3) (containing approximately 100,000 dpm of 1-palmitoyl-2-[1-¹⁴C] arachidonoyl-sn-glycerol-3-phosphorylcholine in one assay) in 400 µM triton X-100 to start the enzyme reaction. For iPLA₂ activity analysis, the cytosolic fraction (0.3 mg protein) was mixed with 100 mM HEPES, pH 7.5, 5 mM EDTA, 2 mM dithiothreitol, and 1 mM ATP in 450 µl, and 50 µl substrate mixture of 100 µM 1-palmitoyl-2-palmitoyl-sn-glycerol-3-phosphorylcholine (containing approximately 100,000 dpm of 1-palmitoyl-2-[1-¹⁴C] palmitoyl-sn-glycerol-3-phosphorylcholine) in 400 µM Triton X-100 was added to start the enzyme reaction.

Substrate preparation for radioisotope method

Substrates for iPLA₂ and cPLA₂ activity analyses were prepared daily. Appropriate amounts of cold and radiolabeled phospholipids were added to an appropriate amount of Triton X-100, and then the mixture was dried under nitrogen gas. Water was added to the residues to give a 10× lipid solution (1 mM phospholipid, 1,000,000 dpm, and 4 mM Triton X-100), which was mixed vigorously.

Enzyme assay

The cytosolic fraction (0.3 mg per assay) was mixed in a 450 µl assay mixture, and 50 µl substrate mixture was added to start the enzyme reaction. The reaction mixture was incubated for 30 min at 40 °C, and then 2.5 ml of Dole reagent (2-propanol, heptane: 0.5 M H₂SO₄, 400:100:20, vol/vol/vol) was added to stop the reaction. 1.5 ml of heptane and 1.5 ml H₂O were added to the mixture, followed by vortexing and centrifugation. The upper phase (about 2 ml) was transferred to a tube, and 200 mg silicic acid (200–400 mesh) was added, followed by vortexing and centrifugation. The supernatant (1.5 ml) was transferred to a scintillation vial to which scintillation cocktail (Ready Safe™, Beckman Coulter, Fullerton, CA) plus 1% glacial acetic acid was added. Radioactivity of released unesterified fatty acid from substrate was counted as described above. Activities of iPLA₂ and cPLA₂ were expressed as the release rate of fatty acid from phospholipid. sPLA₂ activity was measured using an assay kit (Cayman), according to the manufacturer's instructions.

Statistical analysis

Data are presented as mean ± SD (n = 6 for each group). Statistical significance was determined using a one-way ANOVA with Tukey’s multiple comparison. p < 0.05 was used as the cut off for statistical significance.

RESULTS

Effects of graded dietary α-LNA reductions on body weight

Body weight gain did not differ significantly over the 15-week feeding period between rats on any α-LNA deficient diet compared with the control 4.6% α-LNA diet. Average body weight (n = 6 per group) at 21 days equaled 25 ± 2 g (4.6% α-LNA), 25 ± 1 g (3.8%), 26 ± 2 g (2.6%), 26 ± 2 g (1.7%), 26 ± 2 g (0.8%), and 24 ± 3 g (0.2%), respectively (p > 0.05). Final body weight (after 15 weeks on a diet) was 338 ± 25 (4.6% α-LNA), 320 ± 11 g (3.8%), 338 ± 25 g (2.6%), 327 ± 19 g (1.7%), 346 ± 16 g (0.8%), and 325 ± 25 g (0.2%), respectively.

PUFA concentrations

Plasma

Compared with control values at 4.6% dietary α-LNA, statistically significant reductions in unesterified plasma α-LNA and DHA concentrations occurred at ≤ 2.6% dietary α-LNA. Total plasma α-LNA and DHA concentrations were reduced, and total plasma DPAn-6 was elevated at ≤ 1.7% dietary α-LNA (Figure 1). At ≤ 0.8% dietary α-LNA, unesterified plasma DPAn-6 was elevated. Neither total nor unesterified plasma AA was changed significantly compared with control by any deficient diet.

Figure 1.

Plasma unesterified and total fatty acid concentrations in rats fed different α-LNA containing diets for 15 weeks. Values are mean ± SD (n = 6 per group). Superscripts show significant differences at p < 0.05 from mean at 4.6% dietary α-LNA.

Liver

Compared with control, total liver DPAn-6 concentration was increased at ≤ 2.6% dietary α-LNA (Figure 2). Total liver α-LNA and DHA concentrations were reduced at ≤ 1.7% dietary α-LNA, but the total liver AA concentration was unchanged at all deprivations.

Figure 2.

Total fatty acid concentrations in liver and brain of rats fed different α-LNA containing diets for 15 weeks. Values are mean ± SD (n = 6 per group). Different superscripts show significant differences at p < 0.05. Superscripts show significant differences at p < 0.05 from mean at 4.6% dietary α-LNA.

Brain

Like plasma and liver AA concentrations, the brain AA concentration did not differ significantly from control at any level of dietary α-LNA deprivation (Figure 2). At 1.7% dietary α-LNA, when brain DHA did not differ from control, unesterified plasma DHA was 36% (4.11 nmol/ml vs. 6.45 nmol/ml) below control, and total plasma (98.3% esterified) DHA was 27% (269 nmol/ml vs. 367 nmol/ml) below control, demonstrating homeostatic regulation. Only at ≤ 0.8% dietary α-LNA was brain DHA reduced compared with control, while brain DPAn-6 was increased.

Brain enzyme mRNA and activity levels

Brain iPLA₂ activity was decreased and cPLA₂ and sPLA₂ activities were increased at ≤ 1.7% dietary α-LNA compared to control values (Figure 3). sPLA₂ mRNA was increased at ≤ 0.8% α-LNA. Compared with control values, at 0.2% dietary α-LNA brain cPLA₂, COX-2 and 12-LOX mRNA levels were increased and iPLA₂ and COX-1 mRNA levels were decreased (Figure 4). No significant difference was found in 5-LOX or 15-LOX mRNA throughout deprivation (Figure 4).

Figure 3.

Brain mRNA and activity levels of iPLA₂, cPLA₂ and sPLA₂ in rats fed different α-LNA containing diets for 15 weeks. Values are mean ± SD (n = 6 per group). Superscripts show significant differences at p < 0.05 from mean at 4.6% dietary α-LNA.

Figure 4.

Brain mRNA levels of COX and LOX enzymes sPLA₂ mRNA in rats fed different α-LNA containing diets for 15 weeks. Values are mean ± SD (n = 6 per group). Superscripts show significant differences at p < 0.05 from mean at 4.6% dietary α-LNA.

Brain BDNF mRNA and protein levels in brain

Compared with control, brain mRNA and protein levels of BDNF were decreased significantly at ≤ 0.8% and at 0.2% dietary α-LNA, respectively (Figure 5).

Figure 5.

Brain mRNA and protein of BDNF, in brains of rats fed different α-LNA containing diets for 15 weeks. Values are mean ± SD (n = 6 per group). Superscripts show significant differences at p < 0.05 from mean at 4.6% dietary α-LNA.

DISCUSSION

Figure 6 (top) relates the thresholds shown in Figures 1–5 to the degree of dietary α-LNA reduction, while Figure 6 (bottom) suggests mechanisms underlying the threshold relations (see below). Statistically significant reductions appeared first for unesterified plasma α-LNA and DHA, accompanied by increased total liver DPAn-6, at 2.6% dietary α-LNA. At ≤ 1.7% dietary α-LNA, total plasma and liver α-LNA and DHA were reduced, total plasma and liver DPAn-6 were elevated, brain iPLA₂ activity was elevated, and brain cPLA₂ and sPLA₂ activities were reduced. At ≤ 0.8% dietary α-LNA, unesterified plasma and total brain DPAn-6 were elevated, total brain DHA was reduced, brain sPLA₂ mRNA was elevated and BDNF protein was reduced. Finally, at 0.2% dietary α-LNA, brain iPLA₂, COX-1 and BDNF mRNA were reduced, while brain cPLA₂, COX-2 and 12-LOX mRNA were increased. Brain, liver and plasma AA concentrations and brain 5-LOX and 15-LOX mRNA levels were not changed significantly by any deprivation regimen, although changes in AA concentration in specific brain phospholipids may have occurred [12].

Figure 6.

Summary (top) and interpretation (bottom) of correlated threshold changes in plasma, brain and liver lipid measurements in rats fed diets containing declining concentrations of α-LNA for 15 weeks, in comparison to values with dietary n-3 PUFA adequate 4.6% α-LNA (see text).

A major finding of this study is that the brain DHA concentration was maintained at its control level at 4.6% dietary α-LNA down to 1.7% dietary α-LNA, despite reductions in unesterified plasma DHA of 36% (4.11 vs. 6.45 nmol/ml) and of total plasma DHA (98.3% esterified) of 27% (269 vs. 367 nmol/ml) below their respective control levels. Even when brain DHA was reduced significantly at 0.8% dietary α-LNA, the reduction was only 11% (15.7 vs. 17.6 µmol/g wet wt), much less than reductions of 64% (2.42 vs. 6.45 nmol/ml) in unesterified plasma DHA and of 54% (168 vs. 367 nmol/ml) in total plasma DHA below control levels. Because unesterified plasma DHA is the replacement source of brain DHA lost by metabolism [33, 34], there must be homeostatic brain mechanisms that resist a decline in brain DHA concentration despite markedly reduced plasma unesterified DHA.

Brain expression of DHA-selective iPLA₂ IV [15, 16, 35]and of COX-1 (which can be functionally coupled to iPLA₂ [18, 36]) did not differ from control at 2.6% dietary α-LNA. To maintain the brain DHA concentration with this diet, other brain DHA-releasing enzymes may have been downregulated, such as plasmalogen-selective PLA₂, cPLA₂γ, sPLA₂, phospholipase C, and iPLA₂γ [14, 37, 38], but this remains to be tested. Reduced β-oxidation of DHA within mitochondria DHA also could have limited DHA loss from brain [39, 40]. Such effects would prolong brain DHA half-life, which was demonstrated in rats fed the 0.2% α-LNA diet [6]. At 1.7% dietary α-LNA, the reduced activity of DHA-selective iPLA₂ IV [15–17] would have helped to limit DHA loss from brain.

Dietary studies by Stark et al. suggest that DPAn-6 does not compete with DHA at for esterification into brain phospholipid, at a normal DHA concentration [41, 42]. At 0.8% dietary α-LNA, however, their data indicate that replacement would occur, because of the reduced plasma unesterified DHA and increased plasma unesterified DPAn-6 available for brain incorporation [40, 41, 43]. The invariance of the total brain AA concentration with all diets, and the absence of a change in AA turnover with 3-generational n-3 PUFA deprivation in rats [44], support this interpretation. As the threshold elevation in total plasma DPAn-6 at 1.7% dietary α-LNA followed the threshold elevation in liver DPAn-6 at 2.6% dietary α-LNA, it is likely that plasma DPAn-6 was increased by its increased synthesis and secretion by the liver, secondary to upregulation of elongases 2 and 5 and Δ6 desaturase. Activities of these enzymes in liver are much higher than in brain and, unlike brain activities, can be upregulated by dietary n-3 PUFA deprivation [19, 45–47].

Reduced brain activities of iPLA₂ IV and COX-1 at ≤ 1.7% dietary α-LNA suggest reduced DHA release from phospholipid and downstream metabolism, tending to preserve DHA [3, 6, 18], and agree with the reported COX-1 protein increase at 0.2% dietary α-LNA [18]. These changes were accompanied by elevated activities of enzymes of the AA cascade, cPLA₂, sPLA₂, COX-2, and 12-LOX, which have been shown to be functionally coupled in various studies [18, 36, 48–50]. However, since the brain AA concentration was unchanged throughout α-LNA deprivation, we propose that the elevated expression of these enzymes might be related to the elevated brain esterified DPAn-6. Supporting this interpretation, DPAn-6 can be metabolized through the LOX pathway [51], and 12-LOX can convert DPAn-6 to anti-inflammatory oxylipins [52]. Further, the elevated brain DPAn-6 concentration in rats fed the 0.2% α-LNA diet is accompanied by increased DPAn-6 turnover in brain phospholipids (Igarashi et al., unpublished observations).

Some of the brain mRNA changes in this study may have reflected the reduced brain DHA concentration [18]. For example, COX-2 and cPLA₂ gene transcription are regulated by nuclear factor (NF)-κB [53, 54], and DHA can inhibit NF-κB activity via a peroxisome proliferator activated receptor-dependent mechanism [55]. DHA and its 15-LOX product, 10,17S-docosatriene, decreased hippocampal NF-κB and COX-2 gene expression in other models [56].

The reduced brain BDNF protein and mRNA levels between 0.8% and 0.2% dietary α-LNA agree with evidence that DHA can be neuroprotective via a BDNF mechanism [13, 21]. At 0.2% dietary α-LNA, BDNF downregulation is accompanied by reduced cAMP response element binding protein transcription factor activity and reduced p38 mitogen-activated protein kinase activity [13]. Reduced BDNF has been reported in human brain disease [57].

Our results with the extreme 0.2% dietary α-LNA deprivation are consistent with other rodent studies on effects of n-3 PUFA deprivation [2, 11, 12, 22]. Such studies could not identify threshold changes because they were too severe, limiting their clinical extrapolation. The present study involving graded reductions in dietary α-LNA in the absence of dietary DHA, with a constant duration (15 weeks) of dietary exposure, leads us to propose the cause and effect relations discussed above and outlined in Figure 6 (bottom).

Our results may clarify whether dietary n-3 PUFA supplementation is necessary in human subjects [58]. That the brain DHA concentration did not fall below control level down to 1.7% dietary α-LNA, despite the 36% reduction in unesterified plasma DHA, suggests that comparable reductions in humans would not necessarily have pathophysiological consequences. Supporting this suggestion, a 33% lesser blood DHA concentration in vegetarians than in omnivores [59] was not associated with a significant difference in mood, or in general mortality or mortality from any cause [60, 61]. Additionally, 4-month dietary DHA supplementation in preschool children, which increased blood DHA from 1.0% to 3.2% of total fatty acid, did not enhance scores on any of four cognitive tests [62].

In summary, by quantifying effects of 15-week-long graded dietary α-LNA reductions in rats, we have shown that the brain DHA concentration, which is considered critical for brain function and metabolism, was maintained at its control level despite a profound reduction in plasma DHA. Brain DHA declined after plasma DPAn-6, likely derived by liver synthesis, was elevated in the presence of reduced plasma unesterified DHA, and this change might be related to increased expression of cPLA₂, sPLA₂ and COX-2. Because human studies involving n-3 PUFA deprivation have not reported increased plasma DPAn-6 levels, the results imply that brain DHA in such studies was not reduced to an extent that caused pathophysiological changes. This conclusion agrees with the absence of reported neuropathological effects in vegetarians compared with omnivores. An adequate brain DHA content with initially falling dietary α-LNA must be maintained by homeostatic mechanisms that remain to be elucidated. With extreme experimental deprivation, additional mechanisms include reduced expression of DHA-metabolizing iPLA₂ and COX-1, and changes are accompanied by reduced brain BDNF expression.

Abbreviations

AA

arachidonic acid

BDNF

brain derived neurotrophic factor

COX

cyclooxygenase

DHA

docosahexaenoic acid

DPAn-6

docosapentaenoic acid

LA

linoleic acid

α-LNA

α-linolenic acid

LOX

lipoxygenase

cPLA₂

cytosolic phospholipase A₂

sPLA₂

secretory PLA₂

iPLA₂

Ca²⁺-independent PLA₂

PUFA

polyunsaturated fatty acid

sn

stereospecifically numbered