CONCURRENT SESSIONS E - G

SESSION E - Fatty acid synthesis & metabolism

(E1) SUCKLING RATS ACTIVELY RECYCLE CARBON FROM a-LINOLENATE (ALA) INTO NEWLY SYNTHESIZED LIPIDS DURING DEFICIENCY OF ω3 POLYUNSATURATES(PUFA)
S.C. Cunnane, M.A. Ryan, Y.H. Lin1, S-Y. Lim1 and N. Salem Jr1. Department of Nutritional Sciences, University of Toronto,Toronto, Canada, and 1Laboratory of Membrane Biochemistry and Biophysics, NIAAA, NIH, Rockville, MD, USA.

Nearly thirty years ago, significant amounts of carbon from ¹⁴C-ALA injected into suckling rats were shown to be incorporated into newly synthesized brain lipids. 'Carbon recycling' of ALA by suckling rats was later reported to exceed its incorporation into docosahexaenoic acid (DHA) by 5-50 fold. The objective of the present study was to establish the extent of ALA carbon recycling in ω3 PUFA deficient suckling rats. At 10 d old, control and second generation ω3 PUFA deficient rats (n=10/group) were gavaged with a single 1 mg dose of ¹³C-ALA in 100 µl olive oil. 24 h later, they were euthanized and liver and brain collected for analysis of 13C enrichment by GC-isotope ratio mass spectrometry. Compared to controls, DHA in ω3 PUFA deficient rats was 82% lower in brain and 97% lower in liver (both p<0.01). In controls, 9 times more ¹³C from ALA appeared in newly synthesized brain lipids (saturates, monounsaturates, and cholesterol) than in brain DHA (p<0.01). In ω3 PUFA deficient rats, the excess of ALA carbon recycled into brain lipids over use in synthesis of brain DHA decreased to 2.9 fold (69% lower; p<0.01). Similar data were obtained for liver and were largely independent of the units of measure. These data confirm that ALA carbon recycling in suckling rats markedly exceeds ALA use to make DHA. Maximal tissue depletion of ω3 PUFA decreases ALA carbon recycling but this pathway still exceeds ALA use to make DHA by nearly 3 fold. The reason for extensive ALA carbon recycling in the suckling period, even in the face of very high demand for DHA, needs investigation.

(E2) LINOLENIC ACID DOES NOT CONTRIBUTE APPRECIABLY TO THE BRAIN SYNTHESIS OF DOCOSAHEXAENOIC ACID IN THE ADULT RAT
James C. DeMar, Kiazong Ma, Lisa Chang, Jane Bell, and Stanley I. Rapoport. Brain Physiology and Metabolism Section,
National Institute on Aging, National Institutes of Health, Bethesda, USA

The extent to which the brain synthesizes docosahexaenoic acid (DHA) in vivo from a-linolenic acid (LNA) is uncertain. Using a kinetic in vivo rat infusion model, we determined the extent to which LNA entering brain from plasma is converted to DHA. Methods: Adult rats (n = 6) were infused i.v. over 5 min with [1-¹⁴C]-LNA, and the brain was removed after determined by scintillation counting. Results: Whole brain fractional-uptake of [¹⁴C]-LNA from plasma was 1.6 + 0.2% x ml plasma x min-1 x g-1 brain. Most of the brain radioactivity, 71 + 6%, was in water-soluble ß-oxidation products, whereas the remainder was in lipids. Of the total lipid radioactivity, 34% was [¹⁴C]-LNA esterified in phospholipids; and no [¹⁴C]-DHA was esterified in phospholipids. Remaining total lipid radioactivity was found as ß-oxidation products, mostly recycled into cholesterol and saturated fatty acids. Some brain DHA synthesis was implied by the presence of [¹⁴C]-LNA-CoA and[14C]-DHA-CoA at a concentration ratio of 8 to 1; as [¹⁴C]-DHA was absent in the plasma. From the unesterified unlabeled LNA concentration in plasma (41 + 13 nmol/ml), we estimate the synthesis rate of DHA in brain is 0.002 nmol x min-1 x g-1 brain, which is ~2% of the published rat brain incorporation rate of DHA from plasma. Conclusion: LNA that enters brain is largely subjected to ß-oxidation, but is not appreciably converted to DHA within phospholipids.

(E3)THE CONTROL OF SAPIENIC ACID BIOSYNTHESIS
Guillou H, D'Andrea S, Rioux V, Bouriel M, Catheline D, Legrand P,Laboratoire de Biochimie ENSAR-INRA, Agrocampus, Rennes cedex, FRANCE.

We recently reported that a mammalian Δ6-desaturase not only acts on n-6 and n-3 polyunsaturated fatty acids(PUFA) but also on palmitic acid. The Δ6-desaturation of palmitic acid leads to the biosynthesis of sapienic acid (C16:1n-10), a major fatty acid of human sebum. Through the study of recombinant desaturases we further investigated the biosynthesis of sapienic acid. We showed that Δ9-desaturase expression limits C16:1n-10 synthesis in COS-7 cells expressing Δ6-desaturase. This effect depends on the level of Δ9-desaturase expression. Our data provide biochemical evidence that Δ9-desaturase indirectly reduces the biosynthesis of sapienic acid in cells expressing Δ6-desaturase. This biosynthesis may also depend on competition between palmitic acid and other Δ6-desaturase substrates, such as essential 18 carbon PUFA. These data highlight some new aspects of desaturases biochemistry with particular relevance to human sebaceous gland function, sebum lipid synthesis and skin homeostasis.

(E4) ELONGATION AND DESATURATION OF 18-CARBON NUTRIENTS DIFFER FROM 20-CARBON IN ADULT RATS AND HUMAN NEONATES
YH Lin, AR Llanos1, RJ Pawlosky, P Mena1, R Uauy1, N Salem, Jr.,Laboratory of Membrane Biochemistry & Biophysics, NIAAA, NIH, USA. 1Clinical Nutrition, INTA Univ de Chile, Santiago, Chile.

This study was designed to compare the formation of the C22 HUFA from C18 or C20 PUFA for both the n-3 and n-6 fatty acids metabolism pathways using a sophisticated, multiple-isotope procedure with mass spectrometric detection. 20:5n3, ¹³C-U-18:2n6, and ²H₅-20:3n6 with a ratio of 10:1:50:1 (w/w). Venous blood samples were drawn over an concentration-time course curves (AUC) was calculated for each of the isotopomers. For the isotopomers of 22:5n3 for the labelled 22:6n3 in infants. This was similar for 22:5n6. Physiologic compartmental models were constructed to carbon precursors was 6-7 fold greater than that from the 18-carbon precursors in rats. It was about 2-3 fold greater interval of 7 (infants) or 30 (rats) days. Time course data of the labelled fatty acids was plotted and the area under the (also 24:5n3 and 24:6n3 in rats), AUC determinations were similar. However, the AUC for the ¹³C20-22:6n3 was less than that of the ²H₅-22:6n3. When analyzing the data as percent of dose, the amount of labelled 22:6n3 from the 20- compartmental modelling software WinSAAM. A greater rate constant coefficient for the conversion of ²H₅-22:5n3 to seven infants. This resulted in an hourly synthetic rate of 47 nmol (or 16 µg) for the 18:3n3-derived 22:6n3 compared to 17 nmol (or 5.9 µg) for the 20:5n3-derived 22:6n3 (P=0.041). These results also demonstrate that multiple simultaneous experiments can be performed in vivo using this novel approach.

(E5)SREBP-1c: A KEY REGULATOR OF DE NOVO LIPOGENESIS
E.Tartling, A.Salter1 & A.Bennett,School of Biomedical Sciences, Queen's Medical Centre & 1School of Biosciences, Sutton Bonnington Campus,University of Nottingham, UK.

Sterol regulatory element-binding protein 1c (SREBP-1c) is a member of the basic helix-loop-helix family of transcription factors that stimulate sterol and fatty acid biosynthesis. SREBP-1c specifically regulates the expression of genes involved in the production of fatty acids and is expressed in liver, adipose, intestine and skeletal muscle. Sequence alignments show that the human and mouse SREBP-1c promoters share only 42.0% similarity suggesting they may be regulated by different mechanisms and pathways. We have cloned the human SREBP-1c promoter and compared the sequence to that of the mouse to identify putative transcription binding sites. While a number of sites unique to the human promoter have been identified, including HNF-4 (hepatocyte nuclear factor 4), and PDX-1 (pancreatic duodenal homeobox 1), LXREs (liver X receptor response element), thought to be responsible for regulation of expression by polyunsaturated fatty acids (PUFAs) appear to be conserved across the species. Preliminary experiments transfecting human HepG2 hepatocarcinoma cells and rat McA-RH7777 hepatoma cells with human SREBP1c promoter/reporter gene constructs suggest its expression is positively regulated by oxysterols and inhibited by PUFAs potentially through conserved LXRE sites.

(E6)ROLE FOR ELOVL3 AND FATTY ACID CHAIN LENGTH IN VIVO
Rolf Westerberg, Jan-Erik Månsson, Lars Norlén, Andreas Jakobsson, Walter H. Holleran, Rune Toftgård, Mario R. Capecchi and Anders Jacobsson The Wenner-Gren Institute, The Arrhenius Laboratories F3, Stockholm University, Sweden.

Very little is known about the in vivo regulation of mammalian fatty acid chain elongation enzymes as well as the role of specific fatty acid chain length in cellular responses and developmental processes. By disruption of the Elovl3 gene by homologous recombination in mouse, we show that ELOVL3, which belongs to a highly conserved family of microsomal enzymes involved in the formation of very long chain fatty acids (VLCFA), participates in the formation of specific neutral lipids which are necessary for the function of the skin. This was most prominent within the triglyceride fraction where longer fatty acids than 20 carbon atoms were almost undetectable. A functional consequence of this is that Elovl3-ablated mice exhibited a severe defect in water repulsion and increased trans-epidermal water loss. In addition, despite the lack of Elovl3, these mice show no obvious dysfunction of brown adipose tissue were Elovl3 expression is normally elevated more than two hundred-fold in cold-stressed wild-type mice. However, although cold acclimated Elovl3-ablated mice experienced an increased heat loss, due to impaired skin barrier, they were unable to hyper-recruit their brown adipose tissue. Instead they used muscle shivering in order to maintain body temperature.

Session F - Cell signaling, transcription factors & gene expression

(F1) DOCOSAHEXAENOIC ACID: A POSITIVE MODULATOR OF AKT-SIGNALING IN NEURONAL SURVIVAL
Hee-Yong Kim, Mohammed Akbar, Frances Calderon and Zhiming Wen,Section of Mass Spectrometry, Laboratory of Membrane Biochemistry and Biophysics, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, MD, USA.

PI3 kinase/Akt signaling, regulated by extracellular stimuli, is a critical pathway in neuronal survival that entails protein-membrane interaction. Previously, we have reported that docosahexaenoic acid (DHA, 22:6n-3), which is highly concentrated in neuronal membranes, increases membrane PS and prevents neuronal apoptosis. Here we demonstrate that membrane PS accumulation by DHA promotes survival signaling through facilitated membrane substantially faster in DHA-enriched cells in comparison to non-enriched control. This effect was dependent on the capacity of DHA to accumulate PS in neuronal membranes. DHA-enrichment enhanced Akt-phosphorylation as well as its activity, however, did not alter PI3K activity, indicating that Akt translocation is the target for the observed effect of DHA. Docosapentaenoic acid (DPA) which replaces DHA during n-3 fatty acid deficiency was less effective in accumulating PS, and accordingly less efficient in facilitating Akt translocation and preventing apoptosis. Consistently, in vivo reduction of hippocampal PS by dietary depletion of n-3 fatty acids increased the susceptibility of hippocampal neurons to apoptosis in cultures. These results demonstrate a novel mechanism that membrane PS concentration by DHA has a significant impact on neuronal survival by affecting membrane-Akt interaction.

(F2) REPLACEMENT OF DHA BY DPA IN N-3 FATTY ACID DEFICIENCY DOWN REGULATES G PROTEIN-COUPLED SIGNALING IN RAT RETINAL ROD OUTER SEGMENTS
Shui-Lin Niu, Drake C. Mitchell, Sun-Young Lim, Zhi-Ming Wen, Hee-Yong Kim, Norman Salem, Jr., and Burton J. Litman,Laboratory of Membrane Biochemistry & Biophysics, NIAAA, National Institutes of Health, Rockville, Maryland, USA.

N-3 FA deficiency is associated with sub-optimal function in learning, memory, olfactory-based discrimination, spatial learning, and visual acuity. Here, the effect of n-3 FA deficiency on GPCR signalling in retinal rod outer segment (ROS) membranes from rats raised on n-3 adequate or deficient diets was studied. ROS membranes of second generation, n-3 FA deficient rats had lost ~ 80% DHA (22:6n3) relative to n-3 adequate rats. DHA was replaced by DPA (22:n6). Reduced DHA in n-3 deficient ROS correlated with lower rhodopsin activation, metarhodopsin (MII)- transducin (Gt) coupling, and phosphodiesterase (PDE) activity and slower formation kinetics for MII and the MII-Gt complex. N-3 deficient ROS also had a higher degree of acyl order than ROS from n-3 adequate rats. The only difference between n-3 deficient and adequate ROS samples was the relative DHA and DPA content. Therefore, we conclude that differences in signalling efficiency are related to changes in membrane properties resulting from the replacement of DHA acyl chains by DPA. The findings reported here provide an explanation of the changes in the retinal ERG a-wave observed in n-3 FA deficient rodents and non-human primates. Since members of the GPCR family are widespread in signalling pathways in the nervous system, the effect of reduced GPCR signalling due to the loss of membrane DHA may serve as an explanation for other neuronal related deficiencies observed in n-3 FA deficiency.

(F3)INHIBITION OF PHENYLEPHRINE-INDUCED CARDIAC HYPERTROPHY BY DOCOSAHEXAENOIC ACID
Rafat A. Siddiqui1,2,3, Saame Raza Shaikh3, Richard Kovacs2, William Stillwell3, and Gary Zaloga1,2. 1Cellular Biochemistry Laboratory,Methodist Research Institute; 2Department of Medicine, and 3Department of Biology, Indiana University, Indianapolis, IN, USA .

Many of the cardiovascular benefits of fish oil result from the actions of the n-3 polyunsaturated lipids docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA). The beneficial effects of DHA/EPA in modulation of the myocardial hypertrophic response was investigated during present study. Hypertrophy was assessed in neonatal cardiomyocytes exposed to phenylephrine (PE) by measuring cell surface area, total protein synthesis (¹⁴C leucine incorporation), and the organization of sarcomeric a-actinin and by monitoring expression of atrial natriuretic factor (ANF). We report that PE induced a twofold increase in cell surface area and protein synthesis in cardiomyocytes. Treatment of cardiomyocytes with 5 µM DHA effectively prevented PE-induced hypertrophy as shown by inhibition of surface area expansion and protein synthesis, inhibition of ANF expression, and prevention of a-actinin organization into z-bands. DHA treatment prevented PE-induced-accumulation of RAS in detergent resistant fractions of cell membranes (lipid rafts) and consequently inhibited RAS and Raf-1 kinase activation. The upstream inhibition of RAS→Raf-1 effectively prevented translocation and nuclear localization of phosphorylated Erk1/2. These effects consequently led to inhibition of nuclear translocation, and hence, activation of the downstream signaling enzyme p90 ribosomal S6 kinase (p90^rsk). These results indicate that PE-induced cardiac hypertrophy can be minimized by DHA. Our results suggest that inhibition of RAS→Raf-1→Erk1/2→p90^rsk→hypertrophy is one possible pathway by which DHA can inhibit cardiac hypertrophy. In vivo studies are needed to confirm these in vitro effects of DHA.

(F4) TETRADECYLTHIOACETIC ACID (TTA) INDUCES LIGAND DEPENDENT COACTIVATION OF PPARs BY SRC COACTIVATORS AND PGC-1
Røst, TH.1, Mellgren, G.2, Berge RK.1,From the 1Lipid Research Group and 2Hormone Laboratory, Institute of Medicine, University of Bergen.

Peroxisome proliferator-activated receptors (PPAR) are believed to serve as mediators of proliferation inhibition in cancer cells. Interaction with members of the steroid receptor coactivator (SRC) family, and other coactivators, may increase the transcriptional activity of the PPARs. Tetradecylthioacetic acid (TTA) is potent in modulating critical steps in lipid metabolism, and TTA has shown anti cancer effects in vitro and in vivo, the effects being regulated by both PPAR-dependent and -independent mechanisms. Here we show the potential of TTA as a PPAR activator in human breast cancer cells. In transiently transfected cells TTA stimulated PPAR-dependent expression in the following ranking order: PPARδ>PPARγ>PPARα. In cotransfection studies including the SRCs and PPARγ coactivator-1 (PGC-1), PGC-1 gave most pronounced coactivation on all PPARs, following the same ranking order as described for TTA. PPARδ activity was further enhanced by addition of specific agonist or TTA. Compared to the absence of PGC-1, TTA increased the activity more than the specific PPARδ and PPARγ agonists. Transcription intermediary factor 2 (TIF2), one of the SRCs, effectively coactivated PPARγ and PPARδ. SRC-1 induced a less prominent PPAR-dependent transcription than PGC-1 and TIF2. However, SRC-1 and TIF2 increased PPARδactivity synergistically, whereas other combinations of the coactivators showed competing effects on PPAR-dependent transcription. These data demonstrate TTA as a potent activator of PPARs, and enhancer of coactivation by various coactivators, PGC-1 in special.

(F5) FATTY ACID DERIVATIVES AS PPAR-LIGANDS
Laila N. Larsen, Linda Granlund*, Anne Kristin Holmeide**, Hilde I. Nebb*, Jan I. Pedersen*, Jon Bremer,Department of Biochemistry, *Department of Nutrition, **Department of Chemistry, University of Oslo, Norway .

Saturated fatty acids with a sulfur atom in ß-position are known to have hypolipemic effects when fed to rats. They also induce peroxisomal proliferation and ß-oxidation, but are not ß-oxidized themselves. Tetradecyl-thio-acetic acid (TTA) is a derivative that has these effects, and it is a known PPARa-ligand. In order to study the effects of derivatives with varying chainlength, saturated 3-thia fatty acids with chainlength from 11 to 20 carbon atoms have been tested as ligands for mouse PPAR-a, ß and γ, by transient transfections in COS-1 cells. Fatty acids with chainlength from 15 to 18 (including the sulfur atom) were preferred as ligands for PPARa and ß whereas longer fatty acid derivatives were stronger ligands for PPARγ. Saturated sulfur-substituted fatty acid derivatives, normal fatty acids, and polyunsaturated fatty acids were modified into a-methyl fatty acid derivatives. Adding a methyl-group to the a-position in the fatty acid increased their binding to PPAR-ligand binding domain significantly. To verify our results in the transfection study, the inductive effects of the derivatives on transcription of PPARα target genes were studied by northern hybridization. Also the enzyme-activity of the PPARα-target gene, acyl-CoA oxidase, was measured after treatment of 7800 C1 Morris hepatoma cells with a-methyl fatty acids.

(F6)THE EFFECT OF N-3 PUFA-ENRICHED OXIDISED-LDL ON THE EXPRESSION OF CARDIOVASCULAR-RELATED GENES IN A MONOCYTIC CELL LINE
Ian C. Kavanagh1, Carole E. Symes1, David S. Leake2 and Parveen Yaqoob1,1Hugh Sinclair Unit of Human Nutrition, School of Food Biosciences, 2Animal and Microbial Sciences, The University of Reading, Reading, UK.

Oxidised low-density lipoprotein (ox-LDL) exhibits many atherogenic effects, including the promotion of monocyte recruitment to the arterial endothelium and the induction of scavenger receptor expression. Increased consumption of formation of foam cells has not previously been studied. LDL and n-3 PUFA-enriched LDL were oxidised to a mild, moderate and high degree and cultured with THP-1 cells for 24 hours. Gene array analysis of 588 cardiovascular related genes demonstrated a striking up-regulation of a number of potentially pro-atherogenic genes by moderately ox-LDL, but not by mildly and highly ox-LDL. A number of cell survival genes were subject to the same pattern of regulation. A smaller subset of genes was upregulated by all species of ox-LDL. The predominant effect of fish oil was that enrichment of LDL with n-3 PUFA tended to reverse its effects on gene expression when moderately oxidised. In conclusion, gene array analysis demonstrated moderately ox-LDL to be more atherogenic than mildly or highly ox-LDL, suggesting that the lipid hydroperoxides (which peak in moderately ox-LDL) upregulate a number of pro-atherogenic genes. Enriching the LDL particle with n-3 PUFA appears to prevent this upregulation to some degree.

SESSION G - Maternal and Infant Nutrition

(G1) DIFFERENT N-6 POLYUNSATURATED FATTY ACID COMPOSITION OF MILK FROM ALLERGIC AND NON-ALLERGIC MOTHERS
C. Beermann, A. Möller, J. Jelinek, G. Boehm. Numico Research, Friedrichsdorf, Germany.

Background: Polyunsaturated fatty acids (PUFA) serve as precursors for prostaglandines and connecting fat metabolism and immune system. Prostaglandines are responsible for balancing cytolytic (Th1) and humoral (Th2) immune response. According to the immune type paradigm Th2 is correlated to allergy. The arachidonic acid (20:4n6) dependent PGE2 suppress Th1, the dihomo-gamma-linolenic acid (20:3n6) dependent PGE1 suppress immunoglobuline E mediated reactions. Neonates are dominated by Th2 during the prenatal phase. The developing immune competence of infants after delivery might be influenced by PUFA. Methods: The fatty acid composition of breast milk obtained at week 1-8 from type I allergic (n=9) and non- allergic mothers (n=10) was analysed relatively (r) and absolutely (a). Results: Breast milk of allergic mothers showed less linoleic acid (18:2n6) (a p=0.0152; r p=0.0582), gamma-linolenic (18:3n-6) (a p=0.0604; r p=0.0308), and dihomo-gamma-linolenic acid (a p=0.0488) compared to milk of non-allergic mothers but did not differed in arachidonic acid content. The correlation of linoleic to gamma-linolenic (a p=0.0019)and dihomo-gamma-linolenic acid (a p=0.0007) indicates a substrate to product dependency. The fat content from milk of allergic mothers is 21% (although not significant p=0.152) less and correlates to the gamma-linolenic acid content (a p=0.0004; r p=0.0082). Conclusion: The reduced gamma-linolenic and dihomo-gamma-linolenic acid content of breast milk from allergic mothers reflects the linoleic acid availability depending on the fat content. Whether the fat content is associated to allergy needs further investigations.

(G2)PERINATAL DHA STATUS, EARLY FEEDING PRACTICE, AND PROBLEM BEHAVIOUR OF 7-YEAR-OLD CHILDREN
L. Krabbendam (1), E. Bakker (2), G. Hornstra (2), J. van Os (1), Departments of Psychiatry and Neuropsychology (1) and
Human Biology (2), Maastricht University, PO Box 616, 6200 MD Maastricht, The Netherlands.

We studied the relationship between perinatal DHA status (DHA concentration in cord plasma phospholipids) and early feeding practice (breast- or bottle feeding) on later behaviour (parent-completed Child Behaviour Checklist) of about 390 seven-year-old children. All associations were adjusted for plasma DHA status at age 7, the child's sex and birth weight, maternal smoking and drinking during pregnancy, parental education (3 levels) and parenting skills (based on a self-reporting questionnaire, yielding two factors: parental warmth and parental restrictive control). Linear regression analysis demonstrated a significant, association between DHA status and internalising problem behaviour, but not with externalising problem behaviour, with lower levels of DHA being associated with higher levels of internalising problem behaviour (B (SE) - 1.11 (0.52), beta -0.15, p = 0.035). Stratified analysis revealed that the association with DHA was markedly present in the infants fed with artificial formula (n=170, B (SE) -2.62 (0.70), beta - 0.32, p < 0.0001), but absent in the infants fed with human milk (n=215, B (SE) 0.74 (0.75), beta 0.11, p=0.325). Associations with arachidonic acid were not significant. Current DHA status was unrelated to problem behaviour. Since internalising problem behaviour in childhood has been suggested to increase the risk for depression in adults, our finding may point to programming of later depression by early DHA availability. Our results further suggest that breastfeeding may protect against the increased risk of problem behaviour in children with poor DHA status at birth.

(G3) ALTERATIONS IN THE FATTY ACID COMPOSITIONS OF VARIOUS TISSUES AFTER DIETARY SUPPLEMENTATION WITH DPAn-6 OR DHA
N. Salem, Jr., S-Y Lim, J. Loewke, K.D. Stark,Laboratory of Membrane Biochemistry & Biophysics, NIAAA, NIH, Rockville, MD, 20852, USA

The n-3 fatty acid literature indicates that an increase in brain or retinal DPAn-6 coupled to a decrease in DHA results in suboptimal function. However, it is not clear which of these variables is responsible. Moreover, sources of DPAn-6 have been introduced into the human diet with little understanding of how this might impact organ composition or function. In this experiment, rats were fed artificial milk from the d2 of life to adulthood containing LA, L+DPAn-6, LA+DHA or LA+DHA+DPAn-6 as the sources of polyunsaturates and compared with dam reared rats (an LA+LNA diet). At 21 wks of age, the animals were sacrificed, various tissues collected and lipid extracted and the fatty acid profiles measured by gas chromatography. Examination of the fatty acid profiles of the brain, liver, heart, kidney, testes and plasma indicated that in no case did the addition of dietary DPAn-6 lead to a decline in DHA content. However, there was generally a decline in AA in most tissues when DPAn-6 was fed in relation to the LA only diet. For animals in the DPAn-6 supplemented diet, the plasma and especially the heart showed significant elevations in DPAn-6 content relative to the LA diet. The brain exhibited no change in composition as a result of the addition of DPAn-6 to the LA only diet. Adding DHA together with DPAn-6 lead to a decrease in the DPAn-6 content relative to the LA only diet and good support for the DHA content in most tissues including the brain. However, although the testes contained high levels of DPAn-6 in all diets, DPAn-6 supplementation did not lead to an increase over the level in the LA only diet. These data provide an initial framework for understanding the effects of dietary DPAn-6.

(G4) FISH OIL SUPPLEMENTATION OF LACTATING MOTHERS AFFECTS CYTOKINE PRODUCTION IN 21/2-YEAR-OLD CHILDREN
L.Lauritzen1, T.Kjær2, H.Frøkiær2, and K.F.Michaelsen1,1Dept. Human Nutr., The Royal Vet. & Agricult. Univ., Frederiksberg, and 2BioCentrum, The Techn.Univ., Lyngby, Denmark .

Hypothesis: n-3PUFA may influence atopic development possibly via effects on T-helper cell polarization.Aim: to determine if maternal fish oil (FO)-supplementation during lactation could modify later immune responses in the children.Design: 100 Danish mothers with a low fish intake were randomized to FO (4g/d) or olive oil (OO) 1wk-4mo after delivery. The fatty acid composition of erythrocyte (RBC) in the children was measured at 4mo and 21/2y plus at 21/2y plasma IgE levels and cytokine production in lipopolysaccharide (LPS) stimulated whole blood cultures. Results: 57 children were followed-up with all immune variables. RBC n-3PUFA of infants was higher in the FOgroup (p<0.001), but not different at 21/2y. Plasma IgE was unaffected by FO-supplementation. No unstimulated cytokine production was detectable, except for INF-γ in 1and IL-10 in 7 samples. INF-γ production induced by LPS was 2.2-fold above the limit of detection in the FO-group (median, inter-quartile (IQ) range 1-5), whereas no stimulation was detected in the OO-group (IQ-range 1-2, p=0.03). LPS-stimulated IL-10 production was 18-fold above detection in both groups, whereas IL-12 production was only detectable in 3 children, all from the FO-group. The INF-γ to IL-10 ratio was higher in the FO-group (0.46 vs. 0.18 in the OO-group, p=0.02) and was negatively correlated with EPA/AA in infant RBC (p=0.05).Conclusion: Maternal FO-supplementation may result in a long-term Th1-polarization in breast-fed children, indicating that early n-3PUFA-intake may decrease risk of allergic diseases.

(G5) FA PROFILES IN A BLOOD DROP COLLECTED IN A POPULATION OF 3-DAY-OLD BABIES : COMPARISONS WITH ADULT S AND CORRELATIONS WITH PHYSIOLOGIC,METABOLIC AND LIFE STYLE PARAMETERS
C. Galli1, C. Agostoni2, L. Colombo2 , C. Colombo1and F. Marangoni1,1. Department of Pharmacological Sciences, University of Milan, Italy 2. Department of Paediatrics, S. Paolo Hospital, Milan, Italy

LC-PUFA (LCP) status in the newborn affects post natal growth and development, but data in newborns are limited due to difficulties in sample collection. A new method for FA analysis in a drop of blood (Marangoni et al. Anal.. 72 h after delivery (37-41 wks). Comparisons with an adult population (100 subjects), analyzed with the same technique, showed lower LA and ALA and higher LCP (AA and DHA) levels, higher proportions of LC (24 C) S and MFA in the newborns, revealing major differences in FA intake, metabolism and incorporation in lipid pools between the two groups. Differences in FA profiles occurred also within the newborns, in relation with anthropometric (higher LA in females, higher LCP in lower weight & length babies), metabolic (lower n-3 LCP with high bilirubin), and life style (higher 22:5 n-6/22:6 n-3 when mothers were smoking). The new method provides valuable information on the FA status and metabolic features related to FA at very early stages of post natal development, an age that has not been adequately investigated so far.

(G6) ALPHA-LINOLENIC ACID CONVERSION TO DOCOSA-HEXAENOIC ACID IN RATS IS REGULATED BY THE POLYUNSATURATED FAT CONTENT OF THE DIET
Robert A Gibson 1, Mark A Neumann 1, Eric L Lien 2,1 Child Health Research Institute, Flinders Medical Centre, Adelaide, 2 Wyeth Nutrition, Philadelphia, Pennsylvania USA

Tissue levels of docosahexaenoic acid (DHA, 22:6n-3) in animals and humans are minimally influenced by increasing the level of its precursor alpha linolenic acid (ALA, 18:3n-3) in the diet. We hypothesized that this could be due to acid (LA, 18:2n-6) since all three fatty acids (ALA, LA and 24:5n-3) are substrates for the D6-desaturase. We tested our hypothesis in groups of weaning rats fed a spectrum of 54 separate diets for three weeks. The diets varied in fat content (11.8, 22.2 and 39.4 percent of energy, en%), in the levels of LA (0.07 - 17.1 en% ), ALA (0.02 - 12.1 en%) and in the LA:ALA ratio (0.5:1 to 10:1). The concentrations of DHA in plasma total phospholipids of some dietary groups reached 9% of total fatty acids but the peak of DHA accumulation was seen within a narrow dietary range of 1-3 en% ALA and 1-2 en% LA (when ALA present at 1en%). Outside of this range, DHA levels are uniformly low at around 2% total fatty acids. We conclude that the apparent conversion of ALA to DHA is dependent on the level of PUFA in the diet.