Review

Digestion and absorptionof polyunsaturated fatty acids

H Carlier, A Bernard, C Caselli

Département de Nutrition, ENS.BANA, Campus Montmuzard, Université de Bourgogne,F21000, Dijon, France

(Received 20 November 1990; accepted 24 April 1991 )Text presented at the Seminar on Ingestion, Digestion, Absorption and Food Science,

11-13 September 1990, held under the auspices of the Association Française de Nutrition

Summary― Polyunsaturated fatty acids play an important part in the structure and function of cellu-lar membranes and are precursors of lipid mediators which play a key role in cardiovascular and in-flammatory diseases. Dietary sources of essential fatty acids are vegetable oils for either linoleic ora-linolenic acids, and sea fish oils for eicosapentaenoic and docosahexaenoic acids. Because of thespecificity of the pancreatic lipid hydrolases, triglyceride fatty acid distribution is an essential parame-ter in the digestibility of fats. The efficiency of the intestinal uptake depends on the hydrolysis andespecially on their micellarization. n-3 polyunsaturated fatty acid ethyl ester digestion is recognizedto be impaired, but n-3 polyunsaturated fatty acid triglyceride hydrolysis remains a controversialpoint, and to some authors explains differences observed between vegetable and fish oil absorption.So additional studies are required to investigate this intestinal step. In enterocytes, morphologicaland biochemical absorption processes involve reesterification of long-chain fatty acids and lipopro-tein formation. At this level, specific affinity of I- and L- FABPc (cytosolic fatty acid binding proteins)to polyunsaturated fatty acids requires further investigation. A better understanding of the role ofthese FABPc might bring to light the esterification step, particularly the integration of polyunsatura-ted fatty acids into phospholipids. With reference to differences published between fish and vege-table oil absorption, longer-term absorption studies appear essential to some authors. Polyunsatura-ted fatty acid absorption is thought to be not very dissimilar to that of long-chain mono-unsaturatedfatty acid absorption. However, several digestion and absorption specific steps are worth studyingwith reference to the crucial role of polyunsaturated fatty acids in the organism, and for exampleadaptation of possible dietary supplements.

polyunsaturated fatty acid / digestion / absorption pathway / enterocyte esterification / intesti-nal lipoprotein

Résumé ― Digestion et absorption intestinales des acides gras polyinsaturés. Les acidesgras polyinsaturés interviennent dans la structure et la fonction des membranes cellulaires et ils par-ticipent à la synthèse des médiateurs cellulaires lipidiques dont le rôle est primordial dans la préven-tion des maladies cardiovasculaires et la défense de l’organisme. Les sources d’acides gras indis-pensables, linoléique et a-linolénique sont essentiellement d’origine végétale, alors que les acidesgras essentiels, eicosapentaénoïque et docosahexaénoïque, proviennent d’animaux marins. La dis-tribution des acides gras sur les 3 fonctions du glycérol des triglycérides, variable selon leur origine,est un paramètre important de la digestibilité des lipides et ultérieurement de leur absorption et deleur biodisponibilité. Si les éthyl esters des acides eicosapentaénoïque et docosahexaénoi’que sontreconnus pour être difficilement hydrolysés dans la lumière intestinale, des études complémentairess’avèrent cependant nécessaires, pour mieux définir les conditions de l’hydrolyse intestinale des tri-

glycérides d’acides eicosapentaénoi’que et docosahexaénoi’que, compte tenu des résultats contradic-toires obtenus concernant leur absorption. L absorption intestinale des acides gras à longue chainenécessite la mise en place de processus biochimiques et morphologiques pour la réestérification deces acides gras et la formation des lipoprotéines permettant ainsi leur transport dans le milieu inté-rieur. Il semble que les deux FABPc exprimées au niveau intestinal jouent un rôle déterminant dansles processus d’absorption étant donné leur affinité différente pour les différents types d’acides gras.L’étude de la répartition chyloportale des acides gras polyinsaturés atteste de leur absorption préfé-rentielle par la voie lymphatique, avec selon les conditions expérimentales, des caractéristiques spé-cifiques par comparaison avec les modalités d’absorption des acides gras à longue chaine mono-insaturés tel l’acide oléique. Ainsi, une intégration particulière dans les phospholipides des lipopro-téines intestinales est mise en évidence. Cette estérification est influencée par la quantité d’acidesgras polyinsaturés administrée et le degré d’insaturation des lipides d’accompagnement. La distribu-tion entre les VLDL et les chylomicrons est tributaire du degré d insaturation des acides gras adminis-trés. Les différences observées par certains auteurs dans les modalités d’absorption des huiles depoissons par comparaison avec celles d’huiles végétales invite à réaliser des expérimentations delongue durée. Toutefois l’absorption des acides gras polyinsaturés n’apparaît pas fondamentalementdifférente de celle des acides gras à longue chaîne, en particulier mono-insaturés, mais de nombreuxpoints sont à approfondir, étant donné l’importance physiologique des acides gras essentiels. L inté-gration des acides gras polyinsaturés, dès l’entérocyte, dans les phospholipides des lipoprotéines in-testinales, revêt un intérêt nutritionnel particulier dans l’éventualité d’une supplémentation en acidesgras polyinsaturés essentiels motivée par une baisse d’activité des désaturases. Par ailleurs, unemeilleure connaissance des mécanismes de digestibilité, permettra de définir la forme d’administra-tion orale adaptée aux conditions optimales de l’hydrolyse intraluminale.

acide gras polyinsaturé / digestion / voie d absorption / estérification entérocytaire / lipopro-téine intestinale

INTRODUCTION

Polyunsaturated fatty acids play an impor-tant role in the structure of cell membranesand in the synthesis of important media-tors. Cell membranes consist of a lipid bi-layer composed of phospholipids and cho-lesterol, as well as proteins which are

embedded in this lipid bilayer. Differencesexist between various cell types in the in-corporation and metabolization of fatty ac-ids provided with the diet and between in-corporation into specific cellular phospho-lipids (Galloway et al, 1985). Particularlyn-3, polyunsaturated fatty acids have a

very specific role in the membrane of thenervous system (Bourre et al, 1989). Themaintenance of appropriate membrane

physico-chemical properties is essentialfor life processes. Thus, modifications in

dietary fatty acids change membrane

phospholipid fatty acid composition and af-fect functions mediated by membrane pro-teins: receptors, transport pathways andenzymes (Goodnight et al, 1982; Philbricket al, 1987). Polyunsaturated fatty acidsare precursors of lipid mediators which

play a key role in cardiovascular and in-

flammatory diseases. In recent years, evi-dence has been obtained showing that

polyunsaturated fatty acids may be effec-tive in prevention and therapy of cardiovas-cular diseases (Dyerberg et al, 1975,1978; Bang et al, 1976; Kromann andGreen, 1980; Goodnight et al, 1982). Thelow incidence of atherosclerosis andchronic inflammatory diseases among theEskimos of Greenland has been related totheir traditional diet rich in long-chain n-3polyunsaturated fatty acids (Dyerberg et al,1975; Bang et al, 1976).

The 2 real main dietary essential fattyacids are linoleic acid, C18:2, n-6 and a-linolenic acid, C18:3, n-3 because theycannot be synthesized by mammalian or-ganisms. But mammalian organisms aregenerally able to convert linoleic acid

(C18:2, n-6) and a-linolenic acid (C18:3,n-3) into n-6 and n-3 polyunsaturated long-chain fatty acids with the participation ofdesaturases and elongases respectively.On the one hand, arachidonic (C20:4, n-6)and on the other hand, eicosapentaenoic(C20:5, n-3) and docosahexaenoic (C22:6,n-3) acids are synthesized and constitutethe more important physiological polyun-saturated essential fatty acids. Arachidonicand eicosapentenoic acid are precursorsof eicosanoids: prostaglandins, thrombox-anes and leukotrienes, the biosynthesis ofwhich is controlled by cyclooxygenasesand endoperoxidases respectively. Theseeicosanoids have a broad spectrum of bio-logical activity (Goodnight et al, 1982; Sa-muelsson, 1983).

Biological effectsand nutritional essenfialityof polyunsaturated fatty acids

In response to numerous stimuli, phosphol-ipase A2 splits off the arachidonic acid

(C20:4, n-6) which is metabolized by cyclo-oxygenase both to thromboxane A2, a pot-entiator of platelet aggregation and a vaso-constrictor, and to prostaglandin 12. Dietaryeicosapentaenoic acid (C20:5, n-3), thepredominant n-3 fatty acid in the plateletphospholipids of seafood consumers, is re-leased froom phospholipids and can betransformed by cyclooxygenase both into

prostaglandin 13 which is as active as the

vasodilatory and antiaggregatory prosta-glandin IZ derived from arachidonic acid,and into a small quantity of thromboxaneA3 which is almost inactive as platelet ag-gregator and vasoconstrictor (Fischer and

Weber, 1983). Moreover, n-3 and n-6 fattyacids compete for the desaturase and

elongase enzymes. As a consequence,n-3 fatty acids inhibit the synthesis ofarachidonic acid from linoleic acid and adecrease in thromboxane A2 synthesismay result (Holman, 1964). These bio-chemical and physiological aspects explainsome of the proposed benefits of fish oils(Dyerberg et al, 1978; Fischer and Weber,1983) and the intense interest developedin the biological effects and nutritional es-sentiality of n-6 and particularly n-3 longchain polyunsaturated fatty acids.

According to the interaction of the n-6and n-3 polyunsaturated fatty acids, andthe importance of their respective metabo-lism and biological effects, recommenda-tions for dietary fat/fatty acid intake havebeen proposed assuming 2 600 kcal/day.Total polyunsaturated fatty acids havebeen evaluated at 6-7% of calories withn-6/n-3 = 4/1, linoleic acid, a-linolenic acid,and eicosapentaenoic acid and docosa-hexaenoic acid together at 4.8-1.0 and0.27% of calories respectively. Saturatedfatty acids have been evaluated at 6-7%of calories and total mono-unsaturated fat-

ty acids at 12-14% of calories (Lasserre etal, 1985; Simopoulos, 1989). There is a

limiting step in the transformation of linole-ic acid and a-linoelic acid into derivatives.If a loss of desaturase activity preventsendogenous synthesis of arachidonic acid,or of eicosapentaenoic acid, the dietarydoses of C18 essential fatty acids must bemodulated with certain disease states andwith aging (Brenner, 1981) implying a pos-sible supply, for example, in y-linolenicacid, C18:3, n-6.

Source of polyunsaturated fatty acidsand chemical characteristics (table I)

Vegetable oils such as sunflower, soybean, peanut, and corn oils constitute a

substantial source of linoleic acid. More-over, some vegetable oils such as soybean and low erucic acid rapeseed (can-bra oil in table I) oils contain both n-3 andn-6, C18 polyunsaturated fatty acids. Eve-ning primrose oil, blackcurrant oil, and bor-age oil provide significant amounts of

C18:3, n-6 y-linolenic acid and some

C18:4, n-3 stearidonic acid, particularlyblackcurrant oil (Lawson and Hughes,1988c). The derived C20 and C22 essen-tial fatty acids may be found in animals.While land animals provide arachidonicacid (C20:4, n-6), sea animals and espe-cially fish oils constitute some importantnatural sources of n-3 polyunsaturated fat-ty acids: eicosapentaenoic acid (C20:5, n-3) (EPA) and docosahexaenoic acid

(C22:6, n-3) (DHA). However there is awide variation in fatty acid composition ac-cording to the species. For example, her-ring and mackerel depot fats are rich in

gadoleic (C20:1) and erucic (C22:1) acids(17-26% and 11-16% respectively) andpoor in eicosapentaenoic (C20:5, n-3) anddocosahexaenoic (C22:6, n-3) acids (6-7% and 5-8%). In contrast, menhaden oilcontains 16 and 8% respectively of eicosa-pentaenoic and docosahexaenoic acids.

Triglyceride fatty acid distribution is an

important parameter in the digestibility offats (table II). Unsaturated fatty acids oc-cupy the 2-position of glycerol in fats ofmost mammals, except for pigs. Gamma-linolenic acid, C18:3, n-6, is found in 2-and 3-positions of evening primrose,blackcurrant, and borage (Lawson and

Hughes, 1988c). In the same manner, sat-urated fatty acids are in 1- and 3-positionsand linoleic acid in the 2-position of glyce-rol in most plant oils. Several differencesoccur in the marine oil triglyceride chemi-cal structure. For example, eicosapentae-noic acid and docosahexaenoic acid areoften found in the 2-position in the glycerolof mackerel, herring, and cod oils, where-as in fat of seal and polar bear eicosapen-

taenoic and docosahexaenoic acids occu-

py the 1- and 3-positions (Brockerhoff andYurkowski, 1966; Brockerhoff et al, 1966a,1968; Bottino et al, 1967). In MaxEPA, apurified fish oil preparation, eicosapenta-enoic acid occupies both the 2- and 1-,3-positions and docosahexaenoic acid

preferentially the 2-position (Chernenko etal, 1989).

Intestinal availabilityof polyunsaturated fatty acids

The availability of these different poly-unsatured fatty acids is determinant fortheir biological utilization. This availabilityis conditioned by digestion and intestinal

absorption of polyunsaturated fatty acid tri-glycerides, since triglycerides remain themain source of dietary lipids.

The digestion and absorption of saturat-ed and mono-unsaturated long-chain fattyacid triglycerides have been largely re-

viewed (Brindley, 1977; Thomson, 1978;Friedman and Nylund, 1980; Thomson andDietschy, 1981; Tso and Simmonds, 1984;Bernard and Carlier, 1989; Thomson et al,1989). Conversely, long-chain polyunsatu-rated fatty acid triglyceride digestion andabsorption studies began only a few yearsago (Nelson and Ackman, 1988) andsometimes gave conflicting results (Mc-Donald et al, 1980; Chen et al, 1987b; Nils-son et al, 1987; Chernenko et al, 1989; Pa-vero et al, 1989). The assumption of mostinvestigators was that polyunsaturated fat-ty acids are digested and absorbed

through normal processes similar to thoseof mono-unsaturated long-chain fattyacids. However, some researchers sug-gested a possible substantial blood ab-

sorption pathway for polyunsaturated fattyacids with reference to their hydrosolubilityproperties (McDonald et al, 1980, 1987).

The aim of this review is to exposelong-chain fatty acid triglyceride digestionand absorption processes with emphasison polyunsaturated fatty acid absorption.Several aspects are worth studying: i), Isthe gastric and intestinal hydrolysis of

polyunsaturated fatty acid triglycerides, ormore generally of polyunsaturated fattyacid esters, similar to that of saturated andmono-unsaturated fatty acid triglycerides?ii), Does the presence of double bonds in-crease mucosal catabolism of these fattyacids during their absorption? iii), What isthe chyloportal partition of polyunsaturatedfatty acids? iv), If the lymphatic pathwayremains the preferential mode of transportof polyunsaturated fatty acids, what is thechemical form of integration of these fattyacids into intestinal lipoproteins: phospho-lipids or triglycerides? v), What are the in-testinal lipoproteins which transfer them:very low density lipoproteins (VLDL) or

chylomicrons? These last 3 points are de-terminant for their biological utilization ei-ther as an energy supply or as precursorsof lipidic mediators.

Stomach and intestinal stepsof dietary lipid hydrolysis

The digestion of lipids is based on hydroly-sis of the different lipid constituents of thediet. Triglycerides are the main compo-nents (96%) of human dietary lipids corre-sponding to 80-120 g per day; dietaryphospholipids represent 2! g per day, en-riched with 7-20 g of endogenous phos-pholipids from the bile and from the shedintestinal cells, and by sterols, especiallycholesterol, and by fat-soluble vitamins.

In the stomach, the lingual lipase se-creted by lingual glands in murine or gas-tric lipase in rabbit or in man (DeNigris etal, 1988; Moreau et al, 1988) starts hydro-lysis of triglycerides. This hydrolysis takesplace at the low pH of the stomach lumen

in the absence of bile salts. In adults, dueto the preferential cleaves of the 1- and/or1- and 3-positions of triglycerides, lingualor gastric lipase mainly produces partialglycerides and free fatty acids. Such am-phiphilic molecules help disperse the fat inthe stomach and also in the proximal smallintestine. This emulsification of lipids en-hances further hydrolysis by pancreatic lip-id hydrolases in the small intestine, sincethese enzyme activities depend on theemulsification of dietary fat. Thus, Roy etal (1979) demonstrated significantly im-

proved digestion when rats deprived of sal-ivary lipase were fed lipids microdispersedby ultrasonication. Furthermore, Liu et al

(1987) showed that uptake of docosahexa-enoic acid-fish oil was improved by disper-sion in preterm infants known to have lowlevels of pancreatic lipase and bile salts.And Harris and Williams (1989), observedthe very rapid and extensive absorption ofemulsified fish oil in man.

In the proximal part of the small intes-tine, ie the end of the duodenum and the

jejunum, the intestinal lumenal phase ofthe digestion of lipids by pancreatic lipidhydrolases takes place. In presence of bi-carbonate-rich pancreatic juice and bile, tri-glycerides are included in the emulsion

phase previously made in the stomach.Pancreatic bicarbonates increase the in-

testinal lumenal pH to a value allowing anoptimal activity of pancreatic enzymes. Si-multaneously under the detergent action ofbile salts, the diameter of the emulsion oildroplets is decreased from 50 000 down to2 000 A. Intestinal lumen lipid emulsifica-tion is thought to be a limiting step for ab-sorption of most long-chain fatty acids

(Hofmann, 1976).Pancreatic cholesterol ester hydrolase

activated by bile salts completely hydrolyz-es cholesteryl esters into free fatty acidsand free cholesterol. Dietary phospholip-ids, less resistant than biliary phospholip-ids, are hydrolyzed by activated pancreatic

phospholipase A2 in the presence of tryp-sin, calcium ions and bile salts. They yield1-lysophospholipids and free fatty acids.Above the critical micellar concentration ofbile salts, pancreatic colipase, activated bytrypsin, strongly binds to the lipase to allowthe specific enzymatic action of the

pancreatic lipase on triglycerides at theoil-water interface. Thus, triglycerides arecleaved into 2-monoglycerides and free

fatty acids (Chapus et al, 1975; Borgstr6m,1977).

Due to the specificity of pancreaticlipid hydrolases, the hydrolysis products in-cluded free fatty acids, free cholesterol,1-lysophospholipids and 2-monoglycerides.Free fatty acids and 2-monoglycerides arethe major constituents of the terminal

phase of dietary lipid intestinal lumen di-

gestion.

n-3, polyunsaturated fatty acidtriglyceride digestion

Brockerhoff et al (1966a), Yurkowski andBrockerhoff (1966) and Bottino et al (1967)reported that marine oil has an unusual fat-ty acid distribution pattern and observedthat certain long-chain polyunsaturated fat-ty acids esterified in triglyceride oils ofwhales were resistant in vitro to porcinepancreatic lipolysis. This resistance, whichseems to be independent of the position ofpolyunsaturated fatty acids on the glycerolmolecule, was attributed to the introductionof a double bond in the 6-2 through 5-5 po-sition of the fatty acid chain. It is a contro-versial point in vitro in the rat, when lipaseand colipase were in sufficient excess overthe dietary load, MaxEPA fish oil triglyce-rides were totally hydrolyzed into monogly-cerides and free fatty acids (Chernenko etal, 1989). However, Harris and Connor

(1980), reported the decrease of postpran-dial hypertriglyceridemia in humans fed a

fatty meal, where the source of fat was sal-

mon oil. Vahouny (1985) and Chen et al(1987a) demonstrated in rats that the totalfatty acids recovered in lymph after respec-tive salmon oil or menhaden oil and fish oilconcentrate feeding were significantly low-er than levels found after corn oil feeding.Lawson and Hughes (1988a) showed thatin man, eicosapentaenoic and docosahex-aenoic acids from fish oil free fatty acidswere found to be completely absorbed, butonly two-thirds of eicosapentaenoic anddocosahexaenoic acids from the fish oil tri-

glycerides were found to be absorbed.

They observed that absorption of docosa-hexaenoic acid but not of eicosapentaeno-ic acid from fish oil triglycerides was signifi-cantly improved, from 68 to 90%, by co-ingestion with a high-fat meal. This im-

provement was attributed to an enhancedpancreatic lipase activity. In vivo, in the rat,Yang et al (1989) observed that hydrolysisof menhaden oil resulted in a preferentialretention of a high proportion of the polyun-saturated long chain fatty acids in the sn-2monoglycerides and in the residual trigly-cerides; in contrast, the digestion of rape-seed oil led to a preferential release of freelong-chain mono-unsaturated fatty acids.

They concluded that the differential lumen-al release of long-chain and polyunsaturat-ed fatty acids from fish and rapeseed oilsis largely due to their characteristic distri-bution between the primary and secondarypositions in the glycerol molecule and to amuch lesser extent to a chain length dis-crimination by pancreatic lipase. Additionalstudies are required to investigate intesti-nal modalities of polyunsaturated fatty acidtriglyceride hydrolysis, especially C20, andC22, n-3 polyunsaturated fatty acid trigly-cerides.

n-3, polyunsaturated fatty acidethyl ester digestion

Despite the fact that ethyl ester forms offatty acids are not natural components of

the human diet, enriched preparations ofeicosapentaenoic and docosahexaenoicacids of fish oils in the form of ethyl estershave been proposed as dietary supple-ments. Opinions differ about fish oil C20and C22, n-3 polyunsaturated fatty acid tri-glyceride hydrolysis compared to vegeta-ble oil, C18:1, n-9 and C18:2, n-6 unsatu-rated fatty acid triglyceride hydrolysis. Thepoor intestinal absorption of ethyl esters ofC20 and C22, n-3 polyunsaturated fattyacids compared to free fatty acid absorp-tion of eicosapentaenoic and docosahex-aenoic acids, is unanimously attributed totheir impaired intestinal hydrolysis (Hama-zaki et al, 1982; El Boustani et al, 1987;Lawson and Hughes, 1988a, b). It is nowdemonstrated that intestinal absorption ofeicosapentaenoic and docosahexaenoicacids administered as ethyl esters takesplace after their intestinal hydrolysis. In-

deed, no eicosapentaenoic or docosahex-aenoic ethyl esters were found either in 4-h lymph samples collected from rats re-

ceiving ethyl ester, or in blood of man orrats after ethyl ester administration (Hama-zaki ef al, 1982, 1987; Terano et al, 1983;Reicks ef al, 1990). In man, Lawson and

Hughes (1988a), noted that the ethyl es-ters of eicosapentaenoic and docosahex-aenoic acids were absorbed in a range of20% as well as the free fatty acids. Fur-

thermore, they showed (Lawson and

Hughes, 1988b), that absorption of both ei-cosapentaenoic acid and docosahexaeno-ic acid, from fish oil ethyl esters was in-creased 3-fold to = 60% by co-ingestionwith a high-fat meal, indicating that ab-

sorption of polyunsaturated fatty acid ethylesters is highly dependent on the amountof co-ingested fat. In vivo in the rat, Yanget al (1989), showed that the methyl andethyl esters are hydrolyzed = 4 times moreslowly than the corresponding triglyce-rides, which is sufficient to maintain a satu-rated micellar solution of fatty acids in theintestinal lumen during digestion.

Diffusion acrossthe unstirred water layerof intestinal hydrolysis products (fig 1)

As intestinal hydrolysis progresses, the hy-drolysis products can move from the oil-water interface into the aqueous phase ofthe lumenal content allowing pancreatic lip-id hydrolases to pursue their action. Partic-ularly ionized free fatty acids which consti-tute more than half of the fatty acids at thepH of the jejunum and 2-monoglyceridesenter into bile micelles, to form with phos-pholipids mixed micelles, multimolecular

aggregates of 30 to 50A. They constitutethe micellar phase of lipid absorption whichenables apolar lipid monomers such assaturated and mono-unsaturated long-chain fatty acids to cross the unstirred wa-ter layer and mucus lining enterocyte mi-crovillus of the intestinal villus. The transferof the lipolysis products from the oil phaseto the micellar phase appears as a criticalstep in the absorption of dietary fatty acids.Micellar solubilization increases uptake oflipid digestion products by increasing theiraqueous concentration gradients across

the unstirred water layer (Wilson et al,1971 The partition of free fatty acids and2-monoglycerides between the aqueousphase and the micellar phase depends onfatty acid polarity, in particular the chainlength and the degree of unsaturation (Hof-mann, 1976; Dietschy, 1978). The concen-tration of long-chain fatty acids in the waterphase is low compared to that of short andmedium-chain fatty acids and rises with

polyunsaturated fatty acids. Mixed mi-

celles, indispensable for the diffusion of

apolar lipids across the unstirred water

layer and mucus, diffuse less rapidly andrelease their solubilized products less

readily than polar fatty acid monomers,such as short and medium-chain fatty ac-ids and perhaps some polyunsaturated fat-ty acids.

Uptake of long-chain fatty acids

On the one hand the thickness of the intes-tinal water layer is the lowest at the distalpart of the villus (Thomson and Dietschy,1981 on the other hand, as the maturityof enterocyte increases as cells migratefrom the villus crypts to the villus tips, theenzyme activity involved in the monoglyce-ride pathway simultaneously increases

(Johnston, 1976). Thus, the site of lipid ab-sorption is the proximal part of the small in-testine, in an area corresponding to the

upper third of the villus. Dissociation ofmixed micelles occurs prior to lipid absorp-tion and only monomers stemmed from thelipid lumenal hydrolysis are able to be tak-en up by enterocytes. This dissociation is

permitted by the low pH microclimatewhich is maintained by the presence ofmucus all along the lumenal side of the en-terocyte microvillous membrane (Shiau etal, 1985; Shiau, 1990).

Uptake of long-chain fatty acids is be-lieved to be an energy-independent pro-cess. This passive permeation increaseswith the lipophilic properties of fatty acids;consequently it increases with the chain

length and decreases with the degree ofunsaturation. However, at low concentra-

tion, polyunsaturated fatty acids such aslinoleic and arachidonic acids may be ab-sorbed through a facilitated diffusion mech-anism without energy requirement (Chowand Hollander, 1978a,b). The role of a spe-cific long-chain fatty acid binding protein(FABPpm) acting as a membrane receptoris not excluded (Stremmel et al, 1985).Fatty acid absorption would represent adual, concentration-dependent uptakemechanism consisting of a passive diffu-sional transport process and an active car-rier-mediated translocation mechanismwhich would be predominant at low sub-strate concentration as demonstrated byrecent works with some long-chain fattyacids, among them linoleic and linolenicacids (Stremmel, 1988). At the cytosolicside of the enterocyte microvillous mem-brane, cytosolic FABPs may be responsi-ble for the removal of long-chain fatty acidsfrom their strong binding to the lumenal en-terocyte membrane.

Mucosal phase of fatty acid absorption

The chain length of fatty acids appears tobe determinant for absorption processesand particularly the absorption pathway

(fig 2). Short-chain fatty acids are trans-ported by the portal blood as free fattyacids bound to albumin, while it has been

recognized for a long time that long-chainfatty acids are mainly esterified into trigly-cerides and delivered to the lymph includ-ed in lipoproteins: VLDL and chylomicrons(Bloom et al, 1951; Blomstrand, 1955;Borgstr6m, 1955; Clement et al, 1963;Greenberger et al, 1966; Hyun et al, 1967;Carlier, 1971; Carlier and Bezard, 1975;Vallot et al, 1985; Bugaut and Carlier,1987). Intestinal absorption of saturatedand monounsaturated long-chain fattyacids involves correlated complex bio-chemical and morphological enterocyte

events (fig 3). Intestinal lipoprotein synthe-sis by the masking of the hydrophobicgroups of apolar long-chain fatty acids per-mits their transfer in the aqueous media ofthe lymph and then of the blood vascularcompartments.

Cytosolic fatty acid binding proteins(FABPc) bind to the apolar long-chain fattyacids and transfer them through the aque-ous compartment of the enterocyte cytosolfrom the microvillous membrane to the en-

doplasmic reticulum vesicles. In 1976,Ockner and Manning attributed to the fattyacid binding protein Z isolated from the in-testine a higher affinity towards polyunsat-urated long-chain fatty acids comparedwith saturated long-chain fatty acids. In

fact, a few years later, 2 cytosolic fatty acidbinding proteins were identified in the en-

terocyte: the I-FABPC (15124 Da), specificof the intestine and the L-FABPc (14273Da) expressed in the enterocyte but ex-pressed alone in the liver (Bass, 1985).These FABPc seem to have no affinity forshort- and medium-chain fatty acids. Ac-cording to the work of Lowe et al (1987)I-FABPC may have an equal affinity for pal-mitic and arachidonic acids, whereas L-

FABPc may have a higher affinity only forthe arachidonic acid.This point needs fur-ther investigation because of the interest inthe fate of polyunsaturated fatty acids inthe organism. These cytosolic fatty acid

binding proteins transfer long-chain fattyacids towards their site of activation, thenthe activated long-chain fatty acids to-

wards their sites of esterification in the en-

terocyte, that is the smooth and rough en-doplasmic reticulum. The triglycerides areresynthesized mainly through the 2-

monoglyceride pathway (Clark and Hubs-cher, 1960), rather than the glycerol-3-phosphate pathway (Kern and Borgstr6m,1965). Simultaneously with the triglyceridesynthesis within the smooth endoplasmicreticulum, phospholipid and cholesteryl es-ter synthesis takes place within the roughendoplasmic reticulum in which an active

protein synthesis is also observed, particu-larly the synthesis of apoprotein B48. Thetriglyceride synthesis through the 2-

monoglyceride pathway requires the pres-ence of 2-monoglycerides as acceptors for

activated fatty acids. Lumenal hydrolysis ofdietary triglycerides provides enterocytesin exogenous 2-monoglycerides. They areacylated by 2 acyltransferases (a mono-acylglycerolacyltransferase and a diacyl-glycerolacyltransferase) to 1,2-diglyceridesand triglycerides respectively. These ester-ification processes might partly explain theimpairment of the absorption of ethyl es-ters of eicosapentaenoic and docosahexa-enoic acids administered alone, that is,without either triglycerides or monoglyce-rides, since an improvement in their ab-

sorption is observed when co-ingestedwith a high- fat meal in man (Lawson andHughes, 1988a,b). However, in rats,Reicks et al (1990), showed that additionof olive oil to the eicosapentaenoic anddocosahexaenoic ethyl ester lipid emul-sion did not significantly influence lympheicosapentaenoic and docosahexaenoicacid recoveries.

The vesiculation observed in thesmooth endoplasmic reticulum and after-wards in the rough endoplasmic reticulumsaccules of the supranuclear area of theenterocytes takes place simultaneouslywith the resynthesis of triglycerides, cho-lesteryl esters and phospholipids (Cardellet al, 1967; Carlier, 1971). During theirtransfer and progression within the roughendoplasmic reticulum, triglycerides, cho-lesteryl esters and phospholipids are inte-grated in VLDL and chylomicrons by addi-tion of apoproteins (Bisgaier and Glikman,1983). These newly synthesized lipopro-teins are transferred towards the Golgicomplex. Structural changes in the ente-

rocytes during lipid absorption affect the

Golgi apparatus, which exhibits a majorenlargement of its vesicles filled with thenascent lipoproteins (Cardell et al, 1967;Carlier, 1971). This organelle appears tobe an essential step for the final strutura-tion of the intestinal lipoprotein particles,particularly their glycosylation. This pack-aging is necessary in order to allow the

transport of lipoproteins within secretoryvesicles towards the lateral plasma mem-brane and to their exocytosis from the en-terocyte into the intercellular spaces, in-

volving the participation of themicrotubules for this migration (Sabesin,1976; Bernard et al, 1979a). Intestinal li-

poproteins of the intercellular spaces trav-el towards the basement membranewhere opening gaps allow a direct com-munication with the lamina propria. Theirtransfer into the lacteals takes place main-ly through gaps between adjacent endo-thelial cells (Cardell et al, 1967; Carlier,1971; Sabesin, 1976; Bernard et al,1979b). Electron microscope micrographsof rat enterocytes after administration ofcanbra oil (57% oleic acid, 22% linolenicacid, 9% linolenic acid) (Bernard et al,1979a,b) and electron microscope micro-graphs of rat intestinal lymph particles af-ter administration of either corn oil (30%oleic acid, 52% linoleic acid) or menhadenoil (9% oleic acid, 16% eicosapentaenoicacid, 7% docosahexaenoic acid) or codliver oil (21% oleic acid, 10% eicosapen-taenoic acid, 9% docosahexaenoic acid)(Caselli et al, 1979; Rayo et al, 1990)(fig 4), seem to indicate that absorptionprocesses of polyunsaturated fatty acidsare similar to those of saturated andmono-unsaturated fatty acids.

To clarify such an assumption, we willexpose the actual state of knowledgeabout polyunsaturated fatty acid absorp-tion with reference to both medium-chain

fatty acid and to saturated and mono-unsaturated long-chain fatty acid absorp-tion.

Polyunsaturated fatty acid absorption

In rats under vascular perfusion, 24-35%of the infused !4C decanoic acid radioac-tivity was recovered in the mesenteric

portal venous blood during the hourwhich followed the intraduodenal infusionof either 90 pmol of an equimolar mixtureof !4C decanoic acid, oleic acid and

monopalmitin or of !4C decanoic acidalone (Vallot et al, 1985). In contrast, inthe same experimental conditions, only1.8-2.2% of infused 14C linoleic acid and2.6 to 3% of infused 14C arachidonic acidwere recovered in the mesenteric portalvenous blood (Bernard and Carlier, 1991 ).As suggested by McDonald et al (1980),we observed with 14C linoleic acid ad-ministered alone, for infused loads equalor inferior to 30 pmol, a significant in-crease in blood absorption as linoleic aciddoses decreased, but all the values re-mained lower than 5% of the infused radi-

oactivity (Bernard et al, 1991) (fig 5). Inrats with fistulated main mesenteric lym-phatic duct using the same labelled lipidemulsions (fig 6), ie either 90 pmol of !4Cfatty acid alone or 90 pmol of an equimo-lar mixture of !4C fatty acid, oleic acidand monopalmitin, only 0.36-3% of theinfused 14C decanoic acid was recoveredin the lymph instead of 32-48% of the in-fused !4C linoleic acid (Vallot et al, 1985;Bernard et al, 1991) and of 37-43% ofthe infused 14C arachidonic acid duringthe 6 h following duodenal lipid infusion(Pavero ef al, 1989). In the lymph, we ob-served a slightly higher recovery of !4Clinoleic acid radioactivity rather than that

of !4C arachidonic acid, as observed byNilsson et al (1987).

Twenty-four h after gastric administra-tion of 300 pmol of either oleic acid orarachidonic acid or eicosapentaenoic acid,in rats with fistulated left thoracic lymphaticchannel Chen et al (1985) demonstrated

that the overall appearance of arachidonicand eicosapentaenoic acids in the lymphwas quantitatively equivalent to that of

oleate, although there were apparent dif-ferences in the rates of lymphatic absorp-tion of these fatty acids. In these experi-ments the lymph recovery was 76.2-

85.6% of the quantify infused. Studies car-ried out on rats with fistulated main mesen-teric lymphatic duct, with 90 ymol of equi-molar mixtures of oleic acid, monopalmitinand either !4C arachidonic acids or !4C ei-

cosapentaenoic acid, gave similar recover-ies of labelled lipids in lymph (45.1 and

43.7% respectively of the radioactivity in-

fused intraduodenally 6 h before) (fig 6).The same similitude appeared in the re-

sults when the infusates were composedof 5 ymol of either !4C arachidonic acid di-luted with 25 umol of linoleic acid or !4C ei-cosapentaenoic acid diluted with 25 ymolof arachidonic acid in the presence of 30

ymol of oleic acid and 30 umol of monopal-mitin (Pavero et al, 1989). Whatever theexperimental conditions and the dose ad-ministered, unesterified polyunsaturatedfatty acids are efficiently absorbed throughthe lymphatic pathway, even if some differ-ences occur in the absorption profiles, asnoted by Chen et al (1985). Thus in our

data, compared to oleic acid absorption,the peak of absorption was delayed by 30min for arachidonic acid and that of linoleicacid by 60 min, and the lymph recoverywas significantly lower (Pavero et al,1989). Such differences might be explainedby the integration of these polyunsaturatedfatty acids into pools of mucosal phospho-lipids before their effective integration inVLDL and chylomicrons.

Polyunsaturated fatty acid integrationin lymph phospholipids

Most investigators noted a significant in-

crease of labelled lymph phospholipids af-ter labelled polyunsaturated fatty acid ad-ministration by comparison for examplewith a mono-unsaturated fatty acid such asoleic acid. Despite the fact that oleic,arachidonic and eicosapentaenoic acidswere largely recovered in lymphatic trigly-cerides, particularly when administeredalone (Chen et al, 1985; Nilsson et al,

1987; Pavero et al, 1989), Chen et al

(1985) showed a greater incorporation ofarachidonic and eicosapentaenoic acidsinto lymphatic phospholipids (respectively4.5 and 4.3% of the radioactive lipids) thanoleic acid (2.1 %). Can such an esterifica-tion pathway be explained, as in the liver,by a preference of 1-lysophosphatidyl-acylCoA transferase for unsaturated fattyacids (Nisson et al, 1987)? In vitro, withmouse jejunal explants incubated for 15

min at 37 °C in an oxygenated culture me-dium enriched with lipids (1.2 mM: equimo-lar in monopalmitin, oleic acid and either14C oleic or !4C linoleic acid) !4C oleic acidwas integrated mainly into triglycerides:82.5 nmol per mg of explant proteins ver-sus 37 nmol/mg of explant proteins into

phospholipids, whereas 14C linoleic acidwas mainly integrated into phospholipids:74.5 nmol per mg of explant proteins ver-sus 49 nmol/mg of explant proteins into tri-glycerides (Bernard and Carlier, 1991; in

press). This integration in lymph phospho-lipids is particularly enhanced in our experi-mental conditions in vivo in the rat, when30 pmol of monopalmitin of !4C fatty acidswere administered with 30 pmol of oleicacid and 30 pmol of monopalmitin. Thus,at the maximum of the radioactive lymphrecovery (from the 30th to the 120th minafter the intraduodenal infusion), 1.1 to

3.1% of the lymph lipid radioactivity wasrecovered in phospholipids, after !4C oleicacid infusion, 7.3 to 19% of the lymph lipidradioactivity was found on phospholipidsafter 14C arachidonic acid administration

(Pavero et al, 1989) (fig 7).The integration of polyunsaturated fatty

acids into lipoprotein phospholipids ap-

pears to be influenced by the quantity oflipids infused and the degree of unsatura-tion of the other lipids of the lipid emul-sions. Thus, in the rat, Nilsson et al (1987),observed that the incorporation of both lin-oleic and arachidonic acids into different

lipid classes varied with the proportion ofunsaturated fat in the meal: the proportion

of the 2 polyunsaturated fatty acids in

phospholipids was higher when fed in

highly saturated fat ie cream instead of in-tralipid. Some of our results corroboratedthis observation: at the maximum of the ra-dioactive lymph recovery, the integrationof !4C arachidonic acid into phospholipidswas significantly decreased when this fattyacid was administered in the presence oflinoleic acid and mono-olein rather thanwhen 14C arachidonic acid was adminis-tered with oleic acid and monopalmitin(2.8-6.5% and 7.3-19% respectively ofthe lymph lipid radioactivity was recoveredin the phospholipids) (Pavero et al, 1989).

Marine oil absorption

The above discussion concerns absorptionstudies of unesterified polyunsaturated fat-ty acids administered alone or with otherlipids. What happens with vegetable oilsrich in C18 polyunsaturated fatty acids,preferentially of the n-6 family, and withmarine oils rich in C20 and C22 polyunsat-urated fatty acids of the n-3 family? Harrisand Connor (1980) observed that the in-

gestion of salmon oil did not give a typicalfat tolerance curve in human subjects,compared to those given a control mealcontaining animal and vegetable fats.These results were in agreement with Va-houny’s data in the rat (Vahouny, 1985).Then Chen et al (1987b) found that corn oilwas better absorbed than menhaden oil orfish oil concentrate: 24 h after duodenal in-fusion of 170 mg of each emulsified oil re-

spectively 219.7-164.4 and 152.2 mg offatty acids were recovered in rat thoracic

lymph. In contrast, Chernenko et al (1989)did not observe any significant differencein the lymphatic absorption of 0.5 ml ofMaxEPA or olive oil given intraduodenallyin a bolus to Sprague-Dawley rats, over 6and 24 h, over which times = 40 and 70%of the administered dose was recovered in

lymph triglycerides. In the same manner

Rayo et al (in press) did not observe anysignificant difference in lymph triglycerideoutput of rats fed with 0.5 ml of either cornoil, or menhaden oil or cod liver oil, but theprofiles of lymph absorption were different(fig 8). Menhaden oil was more rapidly ab-sorbed than cod liver oil; these differencesmight be explained by the presence of

gadoleic and erucic acids in cod liver oil.

Undoubtedly, differences between authorsare to be attributed to differences in experi-mental conditions; the role of lingual or

gastric lipase must be also taken into ac-count, as well as the fractionation deliveryof the chyme from the stomach. All these

studies dealt with short-term absorption;longer-term animal studies appeared es-sential for some authors to conclude about

the bioavailability of polyunsaturated fattyacids (Hamazaki et al, 1987).

Polyunsaturated fatty acidsand lymph lipoprotein formation

Few works have dealt with the morpholo-gic aspect of lymph lipoprotein particles.Chen et al (1985) subjected lymph sam-ples to ultra-centrifugal separation and not-ed that the distribution of labelled oleic,arachidonic and eicosapentaenoic acids

among major lymph lipoproteins were simi-lar. The same researchers (1987a), dem-onstrated that the overall clearances of theEPA-enriched and oleate-enriched chylo-

microns from the circulation are essentiallythe same. In agreement with biochemicalresults of Chen et al (1985), electron mi-crographs of lymph lipoprotein particles re-veal no significant differences after admin-istration of corn, or menhaden or cod liveroil (fig 4). In fact, percentage and size ofchylomicrons appear correlated with the

degree of unsaturation of the lipids admin-istered. For example, when arachidonicacid was infused with oleic acid and mono-

palmitin (90 J.1mol: 30/30/30 mol/mol/mol)at peak absorption, 6.3% of the lymph lipo-protein particles were chylomicrons, whilewhen arachidonic acid was infused with lin-oleic acid and mono-olein (90 J.1mol: 30/30/30 mol/mol/mol) at peak absorption, 11.2were chylomicrons. When arachidonic acid(5 J.1mol) was diluted with linoleic acid (25J.1mol) and infused with oleic acid and

monopalmitin (30/30 mol/mol) at peak ab-sorption, 7.3% of the lymph particles werechylomicrons, while, as in previous experi-ments, when 5 J.1mol of arachidonic acidand 25 J.1mol of linoleic acid were infusedwith linoleic acid and mono-olein (30/30mol/mol), at peak absorption, 11.3% of thelymph particles were chylomicrons. In the2 cases, the substitution of oleic acid and

monopalmitin by linoleic acid and mono-

olein was followed by an increase in boththe proportion of chylomicrons and in theirsize (fig 9). These observations on the in-fluence of the unsaturation degree of fattyacids corroborate previous results ob-tained with vegetable oils (Caselli et al,1979).

Enterocyte metabolismof polyunsaturatedlong-chain fatty acids

Chernenko et al (1989) indicated that fishoil is absorbed from the rat intestine with-out any major alteration in the acyl chain ofthe triglycerides. The results of the diges-

tion of the lymph samples with lipase indi-cated that the positional distribution ofcharacteristic fish oil fatty acids is similar inboth the MaxEPA and the intestinal lymph.Information concerning eventual oxidationand remodeling of fatty acids in the entero-cyte during their absorption is limited.Some authors have shown that endoge-nous plasma fatty acids were more oxi-dized in the enterocyte than exogenous lu-menal fatty acids (Gangl and Ockner,1975). Although the intestine was an ac-tively metabolizing tissue, it oxidized very

little of the exogenous fatty acids. the

study of catabolic products recovered inmesenteric portal venous blood of rats

(Bernard and Carlier, 1991) shows that themucosal catabolism of polyunsaturatedfatty acids, even higher than the mucosalcatabolism of saturated and mono-

unsaturated long-chain fatty acids, remainslow compared to their absorption. In fact,this oxidation appeared in relation with thechain length, decanoic acid is significantlymore oxidized than long-chain fatty acids,and with the degree of unsaturation, pal-

mitic acid and erucic acids were signifi-cantly less oxidized than oleic, linoleic andparticularly arachidonic acids (Greenber-ger et al, 1965; Vallot et al, 1985; Bernardand Carlier, 1991 ).

Christiansen et al (1986) revealed thatthe microsomal fraction from rat small in-

testine contains a fatty acid elongationactivity; they showed that the activity to-

wards saturated and mono-unsaturated

fatty acids may seem similar in liver andsmall intestine, and that highly polyunsatu-rated fatty acids are markedly poorer sub-strates for the intestinal system. Further-

more, Garg et al (1988) showed the

presence of some desaturase enzymes in

the rat enterocyte. Particularly, they dem-onstrated that rat small intestine possess-es desaturase activity to convert palmiticacid into palmitoleic acid and linoleic acidinto linolenic acid. They suggested that asignificant amount of arachidonic acid mayoriginate from de novo synthesis within theenterocyte via desaturation and chain elon-gation of linoleic acid. In agreement withthis hypothesis, labelled arachidonic acid

appeared in the intestinal lymph of rats

during labelled linoleic absorption, the besttransformation efficiency occurring at peakabsorption (Bernard et al, 1991 ).

A chain shortening, more or less asso-ciated to chain lengthening, of erucic acid,was described when erucic acid was ad-ministered in free form either with triglyce-rides (Thomassen et al, 1985), or with

monoglycerides (Pavero et al, 1990).It may be concluded that the enterocyte

remains essentially an absorptive cell.

CONCLUSION

In conclusion, polyunsaturated long-chainfatty acid absorption does not appear to bevery dissimilar to that of long-chain fattyacid absorption, particularly mono-

unsaturated long-chain fatty acid absorp-tion. Indeed, it has been accepted that sat-urated long-chain fatty acids are poorly ab-sorbed (Bloom et al, 1951; Bernard et al,1987). However, polyunsaturated long-chain fatty acid absorption processes re-

quire further investigation to elucidatesome particularities of the lumenal phasewith reference to the polarity of polyunsatu-rated fatty acids and some specific absorp-tion processes of the mucosal phase, in

particular esterification modalities. Betterinformation on the emulsification, hydroly-sis and micellarization of polyunsaturatedlong-chain fatty acids is essential for the

adaptation of possible dietary supple-ments. Furthermore, a better understand-ing of the role of the enterocyte fatty acidbinding proteins towards polyunsaturatedfatty acids might bring to light the esterifi-cation step crucial for the fate of the poly-unsaturated fatty acids in the organism. In-deed, the energetic or regulator furtherutilization of these essential fatty acids

might depend on their integration into

phospholipids or triglycerides as soon asthey are absorbed.

ACKNOWLEDGMENTS

The authors thank MF Girardier and MC Monnotfor their skillful technical assistance.

REFERENCES

Bass NM (1985) Function and regulation of he-patic and intestinal fatty acid binding pro-teins. Chem Phys Lipids 38, 95-114 4

Bang HO, Dyerberg J, Hjorne N (1976) Thecomposition of food consumed by GreenlandEskimos. Acta Med Scand 200, 59-73

Bernard A, Carlier H, Caselli C (1979a) Étudeau microscope 6]ectronique de la secretiondes chylomicrons chez le rat par 1’ent6rocyte

dans la lumibre des capillaires lymphatiques.CR Acad Sci Paris 289, 1171-1174

Bernard A, Caselli C, Carlier H, Bezard J

(1979b) Electron microscopic investigation ofthe intestinal epithelial cells and the lipopro-tein particles of lymph in rats after ingestionof peanut, canbra or rapeseed oil in the diet.J Physiol (Paris) 75, 559-570

Bernard A, Echinard B, Carlier H (1987) Differ-ential intestinal absorption of two fatty acidinsomers elaidic and oleic acids. Am J Physi-ol 253 (Gastrointest Liver Physiol 16) G751-G759

Bernard A, Carlier H (1989) Absorption and me-tabolism of lipids. In: Intestinal Metabolism ofXenobiotics (Koster AS, Richter E, Lauter-bach F, Hartmann F, eds) Progr PharmacolClin Pharmacol, Gustav Fischer, Stuttgart,217-230

Bernard A, Carlier H (1991) Absorption and in-testinal catabolism of fatty acids in the rat: ef-fect of chain length and unsaturation. ExpPhysiol 76, 445-455

Bernard A, Carlier H (1991) Esterification desacides gras insatur6s au cours de

I’absorption : étude en culture organotypiqued’intestin de souris en fonction de [’Age. Re-prod Nut Dev vol 31, No 3 3

Bernard A, Caselli C, Carlier H (1991) Linoleicacid and chyloportal partition and metabo-lism during its intestinal absorption. Ann NutrMetab (in press)

Bisgaier Cl, Glickman RM (1983) Intestinal syn-thesis, secretion, and transport of lipopro-teins. Ann Rev Physio/45, 625-636

Blomstrand R (1955) Transport form of decano-ic acid 1-14C in the lymph during intestinalabsorption in the rat. Acta Physiol Scand 34,67-704

Bloom B, Chaikoff IL, Reinhardt WO (1951) In-testinal lymph as pathway for transport of ab-sorbed fatty acids of different chain lengths.Am J Physiol 166, 451-455

Borgstr6m B (1955) Transport form of 14C-decanoic acid in portal and inferior venacava blood during absorption in the rat. ActaPhysiol Scand 34, 71-74

Borgstr6m B (1977) The action of bile salts andother detergents on pancreatic lipase andthe interaction with colipase. Biochim Bio-phys Acta 488, 381-391

Bottino NR, Vandenberg GA, Reiser R (1967)Resistance of certain long-chain polyunsatu-rated fatty acids of marine oils to pancreaticlipase hydrolysis. Lipids 2, 489-493

Bourre JM, Dumont 0, Piciotti M, Pascal G, Du-rand G (1989) Polyunsaturated fatty acids ofthe n-3 series and nervous system develop-ment. In: Dietary ur3 and co-6 Fatty Acids(Galli C, Simopoulos AP, eds) NATO ASI Se-ries, Plenum Press, NY, vol 171, 159-175

Brenner RR (1981) Nutritional and hormonal fac-tors influencing desaturation of essential fattyacids. Prog Lipid Res 20, 41-47

Brindley DN (1974) The intracellular phase of fatabsorption. In: Biomenbranes (Smyth DH,ed) Plenum Press, London, vol 4B, 621-672

Brockerhoff H, Yurkowski M (1966) Stereospe-cific analyses of several vegetable fats. J Lip-id Res 7, 62-64

Brockerhoff H, Hoyle RJ, Huang PC (1966a) Po-sitional distribution of fatty acids in the fats ofa polar bear and a seal. Can J Biochem 44,1519-1525

Brockerhoff H, Hoyle RJ, Wolmark N (1966b)Positional distribution of fatty acids in triglyce-rides of animal depot fats. Biochim BiophysActa 125, 55-59

Brockerhoff H, Hoyle RJ, Huang PC, Litchfield C(1968) Positional distribution of fatty acids indepot triglycerides of aquatic animals. Lipids3, 24-29

Bugaut M, Carlier H (1987) Role of intestinal hy-drolases, endogenous subtrates and chylo-portal partition in fat absorption. In: Fat Ab-

sorption (Kuksis A, ed) CRC Press Inc, BocaRaton, FL, vol 1, 197-231

Cardell RR Jr, Badenhausen S, Porter KR

(1967) Intestinal triglyceride absorption in therat. An electron microscopical study. J CellBiol34, 123-155

Carlier H (1971) Absorption en fonction du

temps par un segment d’intestin de rat &dquo;insitu&dquo; de monopalmitine, d’acide palmitique etd’acide ol6ique triti6. J Microsc 12, 193-204

Carlier H, Bezard J (1975) Electron microscopicradioautographic study of intestinal absorp-tion of decanoic and octanoic acids in the rat.J Cell Biol 65, 383-397

Caselli C, Carlier H, Bezard J (1979) Size oflipoprotein particles in the intestinal lymph of

rats fed on corn oil, peanut oil, rapeseed oilor canbra oil. Nutr Metab 23, 73-87

Chapus C, Sari H, Semeriva M, Desnuelle P(1975) Role of colipase in the interfacial ab-sorption of pancreatic lipase at hydrophobicinterfaces. FEBS Lett 58, 155-158

Chen IS, Subramaniam S, Cassidy MM, Shep-pard AJ, Vahouny GV (1985) Intestinal ab-

sorption and lipoprotein transport of (w-3) eic-osapentaneonic acid. J Nutr 115, 219-225

Chen IS, Le T, Subramaniam S, Cassidy MM,Sheppard AJ, Vahouny GV (1987a) Compari-son of the clearances of serum chylomicrontriglycerides enriched with eicosapentaenoicacid or oleic acid. Lipids 22, 318-321

Chen IS, Hotta SS, Ikeda I, Cassidy MM, Shep-pard AJ, Vahouny GV (1987b) Digestion, ab-sorption and effects on cholesterol absorp-tion of menhaden oil, fish oil concentrate andcorn oil by rats. J Nutr 117, 1676-1680

Chernenko GA, Barrowman JA, Kean KT, Herz-berg GR, Keough KMW (1989) Intestinal ab-sorption and lymphatic transport of fish oil

(max EPA) in the rat. Biochim Biophys Acta1004, 95-102

Chow SL, Hollander D (1978a) Arachidonic acidintestinal absorption: mechanism of transportand influence of luminal factors on absorptionin vitro. Lipids 13, 768-776

Chow SL, Hollander D (1978b) Linoleic acid ab-sorption in the unanesthetized rat: mecha-nism of transport and influence of luminal fac-tors on absorption. Lipids 14, 378-385

Christiansen EN, Rostveit T, Norum KR, Tho-massen MS (1986) Fatty acid chain elonga-tion in rat small intestine. Biochem J 273,293-295

Clark B, Hubscher G (1960) Biosynthesis of

glycerides in the mucosa of small intestine.Nature 185, 35-37

Clement G, Clement J, Courel E, Klepping J,Briet S (1963) Absorption des acides gras àchaines moyennes et courtes. In: Biochemi-cal Problems of Lipids (Frazer AC, ed) Else-vier, Scientific Publishing Company, Amster-dam, vol 1, 172-179

DeNigris SJ, Hamosh M, Kasbekar DK, Lee TC,Hamosh P (1988) Lingual and gastric li-

pases: species differences in the origin of

prepancreatic digestive lipases and in the lo-calization of gastric lipase. Biochim BiophysActa 959, 38-45

Dietschy JM (1978) General principles govern-ing movement of lipids across biologicalmembranes. In: Disturbances in Lipid andLipoprotein Metabolism (Dietschy JM, GottoAM Jr, Ontko JA, eds) Am Physiol Soc Be-thesda, MD 1-28

Dyerberg J, Bang HO, Hjorne N (1975) Fattyacid compositon of the plasma lipids inGreenland Eskimos. Am J Clin Nutr 28, 958-966

Dyerberg J, Bang HO, Stofferson E, MoncadaS, Vane JR (1978) Eicosapentaenoic acidand prevention of thrombosis and athero-sclerosis. Lancet i, 117-119 9

El Boustani S, Colette C, Monnier L, DescompsB, Crastes de Paulet A, Mendy F (1987) En-teral absorption in man of eicosapentaenoicacid in different chemical forms. Lipids 22,711-714

Fischer S, Weber PC (1983) Thromboxane A3(TXA3) is formed in human platelets after die-tary eicosapentaenoic acid (C20:5 w3). Bio-chem Biophys Res Commun 116, 1091-1099

Friedman HI, Nylund B (1980) Intestinal fat di-

gestion, absorption and transport. Am J ClinNuti 33, 11 OS-1139

Galloway JH, Cartwright IJ, Woodcock BE,Greaves M, Russel RGG, Preston FE (1985)Effects of dietary fish oil supplementation onthe fatty acid composition of the human plate-let membrane: demonstration of selectivity inthe incorporation of eicosapentaenoic acidinto membrane phospholipid pools. Clin Sci68, 449-454

Gangi A, Ockner RK (1975) Intestinal metabo-lism of plasma free fatty acids. Intracellularcompartmentation and mechanisms of con-trol. J Clin Invest 55, 803-813 3

Garg ML, Keelan M, Thomson ABR, ClandininMT (1988) Fatty acid desaturation in the in-testinal mucosa. Biochim Biophys Acta 958,139-141

Goodnight SH, Harris WS, Connor WE, Illing-worth DR (1982) Polyunsaturated fatty acids,hyperlipidemia and thrombosis. Atheriosde!osis 2, 87-113 3

Greenberger NJ, Franks JJ, Isselbacher KJ

(1965) Metabolism of 1-!4C octanoic and 1-!4C palmitic acids by rat intestinal slices.Proc Soc Exp Biol Med 120, 468-472

Greenberger NJ, Rodgers JB, Isselbacher KJ

(1966) Absorption of medium and long chain

triglycerides. Factors influencing their hydro-lysis and transport. J Clin Invest 45, 217-227

Hamazaki T, Hirai A, Terano T, Sajiki J, KondoS, Fujita T, Tamura Y, Kumagai A (1982) Ef-fects of orally administered ethyl ester of ei-cosapentaenoic acid (EPA C20:5 n-3) on

PGI-like substance production by rat aorta.Prostaglandins 23, 557-567

Hamazaki T, Urakaze M, Makuta M, Ozawa A,Soda Y, Tatsumi H, Yano S, Kumagai A(1987) Intake of different eicosapentaenoicacid containing lipids and fatty acid pattern ofplasma lipids in the rat. Lipids 22, 994-998

Harris WS, Connor WE (1980) The effects ofsalmon oil upon plasma lipids, lipoproteins,and triglyceride clearance. Trans Assoc AmPhysicians 43, 148-155

Harris WS, Williams GG (1989) Emulsificationenhances the absorption of fish oil in man.In: Health Effects of Fish and Fish Oils

(Chandra RK, ed) ARTS Biomedical Publish-ers and Distributors, St John’s, Newfound-land, 211-218 8

Hofmann AF (1976) Fat digestion: the interac-tion of lipid digestion products with micellarbile acid solutions. In: Lipid Absorption: Bio-chemical and Clinical Aspects (Rommel K,Goebell H, Bohmer R, eds) MTP Press Ltd,Lancaster, UK, 3-18 8

Holman RJ (1964) Nutritional and metabolic in-terelationships between fatty acids. Fed Proc23, 1062-1067

Hyun SA, Vahouny GV, Treadwell CR (1967)Portal absorption of fatty acids in lymph andportal vein cannulated rats. Biochim BiophysActa 137, 296-305

Johnston JA (1976) Triglyceride biosynthesis inthe intestinal mucosa. In: lipid Absorption:Biochemical and Clinical Aspects (RommelK, Goebell H, Bohmer R, eds) MTP PressLtd, Lancaster, UK, 85-94

Kern F Jr, Borgstr6m B (1965) Quantitative

study of the pathways of triglyceride synthe-sis by hamster intestinal mucosa. BiochimBiophys Acta 98, 520-531

Kromann N, Green A (1980) Epidemiologicalstudies in the Upernavik district, Greenland:incidence of some chronic diseases. ActaMed Scand 208, 401-406

Lassere M, Mendy F, Spielman D, Jacotot B(1985) Effects of different dietary intakes of

essential fatty acids on C20:3 m6 and C20:4w6 serum levels in human adults. Lipids 20,4, 227-233

Lawson LD, Hughes BG (1988a) Human ab-sorption of fish oil fatty acids as triacylglyce-rols, free acids, or ethyl esters. Biochem Bio-phys Res Commun 152, 328-335

Lawson LD, Hughes BG (1988b) Absorption ofeicosapentaenoic acid and docosahexaenoicacid from fish oil triacyglycerols or fish oil

ethyl esters co-ingested with a high-fat meal.Biochem Biophys Res Commun 156, 960-963

Lawson LD, Hughes BG (1988c) Triacylglycerolstructure of plant and fungal oils contain-

ing y- linolenic acid. Lipids 23, 313-317 7Liu CCF, Carlson SE, Rhodes PG, Rao VS,

Meydreck EF (1987) Increase in plasmaphospholipid docosahexaenoic and eicosa-pentaenoic acids as a reflection of their in-take and mode of administration. Pediatr Res22, 292-296

Lowe JB, Sacchettini JC, Laposata M, McQuil-lan JJ, Gordon JI (1987) Expression of rat in-testinal fatty acid binding protein in Escheri-chia coli. J Biol Chem 262, 5931-5937

McDonald GB, Saunders DR, Weidman M, Fis-cher L (1980) Portal veinous transport of longchain fatty acids absorbed from rat intestine.Am J Physiol239, G141-G150

McDonald GB, Weidman M (1987) Partitioningof polar fatty acids into lymph and portal veinafter intestinal absorption in the rat. QuartJ Exp Physiol72, 153-159

Moreau H, Gargouri Y, Bernadal A, Pieroni G,Verger R (1988) Étude biochimique et

physiologique des lipases prdduod6nalesd’origines animale et humaine. Rev Fr CorpsGras 4, 169-176

Nelson GJ, Ackman RG (1988) Absorption andtransport of fat in mammals with emphasis onn-3 polyunsaturated fatty acids. Lipids 23,1005-1014 4

Nilsson A, Landin B, Jensen E, Akesson B(1987) Absorption and lymphatic transport ofexogenous and endogenous arachidonic andlinoleic acid in the rat. Am J Physiol 252(Gastrointest Liver Physiol 15) G817-G824

Ockner RK, Manning JA (1976) Fatty acid bind-ing protein: role in esterification of absorbed

long chain fatty acid in rat intestine. J Clin In-vest 58, 632-641

Pavero C, Grappin S, Caselli C, Bernard A, Car-lier H (1989) Lymphatic transport of arachi-donic and eicosapentaenoic acids in the rat.1st Congress Eurolipid ETIG Paris, vol II,821-828

Pavero C, Caselli C, Bernard A, Carlier H (1991)Absorption and metabolism of erucic acid inthe rat. Rev Fr Corps Gras 9/10, 301-307

Philbrick DJ, Mahadevappa VG, Ackman RG,Holub BJ (1987) Ingestion of fish oil or a de-rived n-3 fatty acid concentrate containing ei-cosapentaenoic acid (EPA) affects fatty acidcomposition of individual phospholipids of ratbrain, sciatic nerve and retina. J Nutr 117,1663-1670

Rayo JM, Cour 0, Caselli C, Bernard A, CarlierH (1991) Absorption intestinale chez le ratd’huile de foie de morue, d’huile de menha-

den, d’ethyl esters d’acides docosahexa6-

noique et 6icosapenta6noique et d’huile demais. Reprod Nutr Dev vol 31, n°3 (in press)

Reicks M, Hoadley J, Subramaniam S, More-house KM (1990) Recovery of fish oil-derivedfatty acids in lymph of thoracic duct-cannulated Wistar rats. Lipids 25, 6-10 0

Roy CC, Roulet M, Lefebvre D, Chartrand L,Lepage G, Fournier LA (1979) The role ofgastric lipolysis on fat absorption and bileacid metabolism in the rat. Lipids 14, 811-815 5

Sabesin SM (1976) Ultrastructural aspects ofthe intracellular assembly, transport and exo-cytosis of chylomicrons by rat intestinal ab-sorptive cells. In: Lipid Absorption: Biochemi-cal and Clinical Aspects (Rommel K, GoebellH, Bohmer R, eds) MTP Press Ltd, LancasterUK, 113-145

Samuelsson B (1983) Leukotrienes: mediatorsof immediate hypersensitivity reactions andinflammation. Science 220, 568-575

Shiau YF (1990) Mechanism of intestinal fattyacid uptake in the rat: the role of an acidic mi-croclimate. J Physiol 421, 463-474

Shiau YF, Fernandez P, Jackson MJ, MacMona-gle S (1985) Mechanisms maintaining a low <microclimate in the intestine. Am J Physiol248, G608-G617 7

Simopoulos AT (1989) General recommenda-tions on dietary fats for human consumption.

In: Dietary w-3 and cu-6 Fatty Acids (Galli C,Simopoulos AP, eds) NATO ASI Series, Ple-num Press, N Y, vol 171, 403-404

Stremmel W (1988) Uptake of fatty acids by jeju-nal mucosal cells is mediated by a fatty acidbinding membrane protein. J Clin Invest 82,2001-2010 0

Stremmel W, Lotz G, Strohmeyer G, Berk PD(1985) Identification, isolation and partialcharacterization of a fatty acid binding proteinfrom rat jejunal microvillous membranes. JClin Invest 75, 1068-1076

Terano T, Hirai A, Hamazaki A, Kobayashi S,Fujita T, Tamura Y, Kumagai A (1983) Effectof oral administration of highly purified eico-sapentaenoic acid on platelet function, bloodviscosity and red cell deformability in healthyhuman subjects. Atherosclerosis 46, 321-331

Thomassen MS, Helgerud P, Norum KR (1985)Chain-shortening of erucic acid and micro-peroxisomal (3-oxidation in rat small intestine.Biochem J 255, 301-306

Thomson ABR (1978) Intestinal absorption of

lipids: influence of the unstirred water layerand bile micelle. In: Disturbances in Lipid andLipoprotein Metabolism (Dietschy JM, GottoAM Jr, Ontko JA, eds) Am Physiol Soc (Be-thesda, MD) 29-56

Thomson ABR, Dietschy JM (1981) Intestinal lip-id absorption: major extracellular and intra-cellular events. In: Physiology of IntestinalTract (Johnson LR, ed) Raven Press, NY,1147-1220

Thomson ABR, Keelan M, Garg ML, ClandininMT (1989) Intestinal aspects of lipid absorp-tion: in review. Can J Physiol Pharmacol 67,179-191

Tso P, Simmonds WG (1984) The absorption oflipid and lipoprotein synthesis. In: Lipid Re-search Methodology (Story JA, ed) Alan RLiss Inc, NY, 191-216 6

Vahouny GV (1985) Inhibition of cholesterol ab-sorption by natural products. In: Drugs Affect-ing Lipid Metabolism VIII (Kritchevsky D,Holmes WL, Paoletti R, eds) Plenum, Press,NY, 265-279

’ Vallot A, Bernard A, Carlier H (1985) Influenceof the diet on the portal and lymph transportof decanoic acid in rats. Simultaneous studyof its mucosal catabolism. Comp BiochemPhysio182A, 693-699

Yang LY, Kuksis A, Myher JJ (1989) Lumenalhydrolysis of menhaden and rapeseed oilsand their fatty acid methyl and ethyl esters inthe rat. Biochem Cell Biol67, 192-204

Wilson FA, Sallee VL, Dietschy JM (1971) Un-stirred water layers in intestine: rate determi-

nant of fatty acid from micellar solutions. Sci-ence174,1031-1033

Yurkowski M, Brockerhoff H (1966) Fatty aciddistribution of triglycerides determined by de-acylation with methyl magnesium bromide.Biochim Biophys Acta 125, 55-59