Intestinal metabolism of plasma free fatty acids.Intracellular compartmentation andmechanisms of control.

A Gangl, R K Ockner

J Clin Invest. 1975;55(4):803-813. https://doi.org/10.1172/JCI107991.

Fatty acid metabolism in intestinal mucosa has been examined primarily in regard to lipidabsorption. Since earlier studies suggested intestinal utilization of plasma free fatty acids(FFA), we investigated mucosal metabolism of plasma FFA in rats. Mucosal radioactivity (1per cent of administered) was maximal 2 min after i.v. [14C]palmitate. Of mucosal 14C, 42percent was in water-soluble metabolites, including CO2 and ketoacids, 28 percent inphospholipids, and only 16 per cent in triglycerides. The specific activity of mucosaltriglyceride fatty acids (TGFA) was 11 times that of serum TGFA, confirming in situsynthesis. Double isotope experiments showed marked differences in the metabolism offatty acids entering mucosa simultaneously from lumen and plasma. Whereas luminal fattyacids were chiefly esterified to triglyceride, plasma FFA were preferentially oxidized andincorporated into phospholipids. Crypts did not differ from villi, indicating that intestinalmetabolism of plasma FFA is related to their site of entry into epithelial cells. Mucosalmetabolism of i.v. [14C]palmitate was minimally affected by glucose administration.However, intraduodenal isocaloric ethanol inhibited mucosal oxidation of FFA by 60 percent, and increased incorporation into triglycerides nearly twofold. During lipid absorption,mucosal uptake of plasma FFA doubled and incorporation into intestinal lymph triglycerideswas increased sixfold. These studies demonstrate an intracellular compartmentation of fattyacids in the intestinal epithelium. In contrast to absorbed […]

Research Article

Find the latest version:

http://jci.me/107991-pdf

Intestinal Metabolism of Plasma Free Fatty Acids

INTRACELLULARCOMPARTMENTATIONAND

MIECHANIS'MS OF CONTROL

ALFREDGANGLand ROBERTK. OCKNER

From the Department of Medicine, University of California School of Medicine,San Francisco, California 94143

A B S T RA C T Fatty acid metabolism in intestinal mu-cosa has been examined primarily in regard to lipidabsorption. Since earlier studies suggested intestinalutilization of plasma free fatty acids (FFA), we in-vestigated mucosal metabolism of plasma FFA in rats.Mucosal radioactivity (1% of administered) was maxi-mal 2 min after i.v. ['4C]palmitate. Of mucosal "4C,42% was in water-soluble metabolites, including C02and ketoacids, 28% in phospholipids, and only 16% intriglycerides. The specific activity of mucosal triglyce-ride fatty acids (TGFA) was 11 times that of serumTGFA, confirming ini sitit synthesis.

Double isotope experiments showed marked differ-ences in the metabolism of fatty acids entering mucosasimultaneously from lunmen and plasma. Whereas lumi-nal fatty acids were chiefly esterified to triglyceride,plasnma FFA were preferentially oxidized and incorpo-rated into phospholipids. Crvpts did not differ fromvilli, indicating that intestinal metabolism of plasmaFFA is related to their site of entry into epithelialcells. Mucosal metabolism of i.v. ['4C]palmitate wasminimally affected by glucose administration. However,intraduodenal isocaloric ethanol inhibited mucosal oxi-dation of FFA by 60%, and increased incorporationinto triglycerides nearly twofold. During lipid absorp-tion, mucosal uptake of plasmca FFA doubled and in-corporation into intestinal lynmph triglycerides was in-creased sixfold.

These studies demonstrate an intracellular compart-mentation of fatty acids in the intestinal epithelium. Incontrast to absorbed luminal fatty acids, plasma FFA

Dr. Gangl is the recipient of a Research Fellowship fromthe Max Kade Foundation and Dr. Ockner is the recipientof a Research Career Development Award (AM-36586)from the National Institutes of Health.

Rcceived for publication 12 Aulguist 19741 anid in reviscdforin 9 Dccenbcr 1974.

in the fasting state are both an energy source and asubstrate for the svnthesis of tissue phospholipid. Thefasting contribution of plasnma FFA to mucosal andlymph triglyceride is minimal, but it increases duringethanol administration and fat absorption.

INTRODUCTION

Studies of fatty acid metabolism in intestinal mucosahave dealt almost exclusively with luminal fatty acids,which are predominantly esterified to trigly ceride. Incontrast, plasma free fatty acids (FFA) undergo oxi-dation in many tissues, and the possibility that intes-tinal mucosa also utilizes plasma FFA was suggestedby previous studies (1-3). Although this would implya unique duality of fatty acid sources for the intestine(lumen and plasnma), this interesting possibility hasreceived little attention. Furthermore, to the extentthat plasma FFA might be utilized by the mucosa, itis not known whether esterification or oxidation isthe major pathway. Available information suggests thatincorporation of plasmia FFA into intestinal lymph lipo-protein triglyceride if of only minor significance inthe fasting state, but it is possible that this contributionassumes greater importance under certain conditions.The present studies were designed to investigate themetabolism of plasma FFA in the intestinal mucosa.both in the fasting state and as affected by possibleregulatory factors. They have been reported, in part, inabstract form (4, 5).

METHODSAiinimals and operativc proceditres. Male Sprague-Dawley

rats (300-365 g) were used in all experiments, and werefed standard rat chow (Berkeley Diet Rat and Mouse Food,Feedstuff Processing Co., San Francisco, Calif.) ad lib.until operation. Under ether anesthesia a polyethylene cathe-ter (PE 90, Clay-Adams, Inc., Parsippany, N. J.) was in-

The Journal of Clinical Investigation Volume 55 April 1975-803-813 80o3

serted into the duodenum and secured with a purse-stringsuture. In some rats the mesenteric lymph duct was can-nulated by a modification of the technique of Bollman, Cain,and Grindlay (6) using Silastic medical-grade tubing, 0.020inch ID (Dow Corning Corp., Medical Products Div., Mid-land, Mich.) Operated rats were placed in restraining cages(7) and received intraduodenal 0.85% saline, 3.0 ml/hthroughout the entire experiment, unless otherwise stated.

Preparation of ["C] palmnitic acid-rat serum comnplex.About 20 juCi[U-'4C]palmitic acid (735 mCi/mmol, NewEng-land Nuclear, Boston, Mass.; radiochemically > 98%o pure)in 20 Al of methyl ethyl ketone, was neutralized with a slightexcess of ethanolic KOH, slowly mixed with 2.3 ml ratserum, and diluted 1: 1 with 0.85% saline. This preparationwas stable at 40C for at least 4 days, but in all cases wasused within 2 days. After i.v. injection of this preparation,the half time of the initial disappearance of plasma ["C]-FFA was 1.5 min. Of the injected `C, 20% was expiredin "CO2 by 30 min, and 38% by 3 h (unpublished observa-tions). These findings are similar to those of other workersin regard to the metabolism of plasma FFA (8, 9) andestablish the validity of the preparation employed in ourstudies. Approximately 1.0 ml of this preparation (0.5 mlserum) was injected per rat, representing an acute expan-sion of the plasma FFA pool by less than 5%o.

Preparation of lipids for intradulodenial infusion. [9,10-3H]palmitic acid (200 mCi/mmol, New England Nuclear)was purified radiochemically by thin-layer chromatography(TLC) to > 96%. Chemical purity (gas-liquid chromatog-raphy) was > 99%. Just before use, a tracer quantity (about1 ,uCi) was dissolved in 1.0 ml of 10 mMsodium taurocholatein 0.85% NaCl, pH 7.4. Mixed micelles consisted of a clearsolution of 10 mMoleic acid (Calbiochem, San Diego, Calif;A grade, > 99% pure) and 5 mM glyceryl monooleate(Calbiochem; 90% pure) dissolved in 10 mMsodium tauro-cholate (Calbiochem; >95%o pure) in 0.85%o NaCl, pH7.0-7.5. Emulsions contained 10 times these amounts ofoleic acid and glyceryl monooleate in 10 miM taurocholate,pH 7.0-7.5, and were prepared by sonication (Sonifier celldisruptor model W 185D, Heat Systems-Ultrasonics, Inc.,Plainview, N. Y.).

Experimiental procedure. The [1'C]palmitic acid-rat serumcomplex was injected acutely into a tail vein of consciousrats. Blood was collected by decapitation at specified timeintervals. The abdomen was opened immediately and thesmall intestine, from ligament of Treitz to cecum, was re-moved and flushed with 40 ml ice-cold saline. The timefrom decapitation to completion of the saline flush was 60-90 s. The small intestine then was divided into proximal(jejunal) and distal (ileal) halves.

Tissue preparation. Whole mucosa was extruded fromthe intact intestinal segment. In some experiments, villi andcrypts were prepared (10). The mucosal scrapings were

homogenized in a Teflon-glass homogenizer in 3 vol of either0.85% NaCl or in 0.28 Mmannitol in 0.01 M Na2HPO4, pH7.4, containing 0.05 M EDTA. In several experiments, theliver was homogenized in 5 vol of 0.85%c NaCl.

Where indicated, mucosal homiogenates were subsequentlysonicated (Sonifier cell disruptor, model W 185D, HeatSystems-Ultrasonics, Inc.) for 625 W*s; a droplet was

examined by light microscopy to determine completeness ofcell disruption.

Subcellular fractions were prepared after homogenizationand sonication of intestinal mucosa by the method of Rod-gers, Riley, Drummey, and Isselbacher (11). The floating

fat layer was separated from the 105,000-g supernate bymeans of a tube slicer.

Analytical mecthods. Lipids were extracted from mucosaland liver homogenates, cell fractions whole serum, and de-fibrinated lymph by the method of Folch, Lees, and SloaneStanley (12). Lipid classes were separated on precoatedTLC plates (silica gel, 0.25 mm, EM Reagents, E. Merck,Darmstadt, West Germany), in petroleum ether: diethylether: acetic acid, 90: 15: 1.5. Plates were "washed" beforeuse by allowing them to develop overnight in chloroform:methanol, 49:1 (13). Lipid zones, demonstrated with iodinevapors or rhodamine 6G and identified by comparison withstandards, were scraped and eluted or transferred directlyinto counting vials. The phospholipid zone at the origin waseluted by extractions with chloroform: methanol: acetic acid:water, 25: 15: 4: 2 (twice), followed by methanol, and me-thanol acetic acid: water, 94: 1: 5 (recovery > 98%). Ma-terial in the combined eluates was then developed in a secondTLC system, chloroform: methanol: acetic acid: water, 25:15: 4: 2 (14), to separate indivdual phospholipid classes.

Triglycerides and fatty acids were eluted from the silicagel by three extractions with chloroform: heptane: methanol,280: 210: 10 (recovery > 98%c). These eluates were used fordetermination of fatty acids (13) and of triglycerides (15).

In several experiments, water-soluble radioactivity wasmeasured directly in the upper (water-methanol) phase ofthe Folch extraction system, and was consistently found tobe less than the value calculated as the difference betweenlipid-soluble and whole homogenate radioactivity, by anamount similar to the "CO2 content of the homogenate. Sincethe Folch extraction involves acidification of the aqueousphase, "CO2 was undoubtedly lost during this procedure.Accordingly, water-soluble radioactivity was routinely takenas the difference between radioactivity in the whole homo-genate and the lipid extract.

"CO2 in tissue was determined by acidifying homogenateswith 66% citric acid in 25-ml flasks fitted with a center wellcontaining 0.3 ml hydroxide of hyamine, and incubated ina metabolic shaker for 20 min, 370C, 100 cpm (16). Centerwells were then removed and transferred to counting vialsfor radioassay.

After evolution of 1"CO2, p-ketoacids were decarboxylatedwith aniline citrate (16, 17) during further incubation (37 C,2 h, 100 cpm); the evolving "CO2 was collected and assayedas described above. Since the [14C]palmitate employed inthese studies was uniformly labeled, and since decarboxyla-tion of acetoacetic acid drived from this palmitate wouldyield only one labeled carbon atom (of the four which po-tentially were labeled), the measured disintegrations perminute were multiplied by four.

Radioassays. Homogenates, cell fractions, whole serum,and defibrinated lymph were assayed for radioactivity inLiquifluor toluene solution (New England Nuclear) con-taining 10% Biosolv (Beckman Instruments, Inc., Fullerton,Calif.) in a Beckman liquid scintillation system (modelLS-250). For lipid-soluble extracts, Biosolv was not added.Quenching was corrected for by an automatic externalstandard, and results were expressed as disintegrations perminute.

Calculation of data. To permit comparison of resultsamong rats injected with differing amounts of isotope, allradioactivity data were normalized to an injection of 1.0,uCi. Groups of three or more rats were compared by un-paired or paired t tests. Statistical significance of differencesamong more than two groups were determined by analysis ofvariance (18).

804 A. Gangl and R. K. Ockner

RESULTS

Uptakc and mnetabolisin of Plasmiia FFA by intestinialsnucosa. At 2 min after i.v. administration of L[CC]-palmitate, 0.96±0.09% of the radioactivity was foundin the niucosa of the jejuno-ileum. Proximal and distallhalves (0.50+0.06 and 0.45±0.03%) did not differ, andafter 2 min, mucosal "C gradually declined. Becausemucosal radioactivity was greatest at 2 min, and toavoid the potentially confusing effects of subsequentsecretion of nmucosal L"C]lipid into intestinal lymph, orthe absorption of labeled bile phospholipid, experi-nments were conducted at this time period, unless other-xvise indicated. (In general, FFA uptake in thesestudies is assumed to equal total tissue radioactivitysince, as noted below, contamination by plasma wasinsignificant.)

Distribution of ["C]palmitic acid among intestinalmucosal metabolites is slhowvn in Table I. In bothproximal and distal mucosa, incorporation into water-soluble metabolites (42.2 and 34.2%, respectively) andinto plhosphlolipids (27.6 and 35.0%) together accountedfor the preponderance (- 70%) of mucosal radioactiv-itx . Incorporation into triglyceride (16%) was minor,and only small amounts of "C were present in diglyce-ride, cholesterol esters, and tissue FFA. Proximal anddistal mucosa were similar except in regard to con-version of ["C]palmitate to water-soluble metabolites,cholesterol esters, and phospholipids, as shown. Thisgeneral pattern was similar at 5 and 15 min.

Approximately 20-25% of wvater-soluble "C were in"CO2 and 8-ketoacids (Table I), the remaining un-identified 75-80% probably representing citric acid-cycle intermediates (19). The possibility that water-soluble mucosal radioactivity was due to contaminationby serum is most unlikely since admixture of 0.6 mlserunm/g mucosa would have been required (the actualadmixture was far less: e.g., mucosal ["C]FFA/g wasonly about 4% of that in serum). Thus, the water-soluble metabolites almost certainly were formed in themucosa itself.

Of the ["C]lipid recovered in phospholipids, morethan 50% exhibited mobility similar to that of phospha-tidvl choline in a second TLC system ( Methods).About 80%c was recovered in this and in phosphatidvlethanolamine (12.6±1.2%), phosphatidvl serine andphosphatidvl inositol (8.0±0.3%), and sphingomivelin(5.5±-0.3%). The remaining 20% (similar in mo-bilitv to phosplhatidvl glycerol, phosphatidic acid, andcardiolipin) was not further identified.

Hepatic uptake of plasma FFA accounted for 17.6±+1.6% of the injected radioactivity at 2 min, corn-parable to earlier studies (20). In contrast to intes-tine (Fig. 1), "4C was mainly incorporated into tri-glvcerides, with only a small fraction in water-soluble

TABLE I

AMetabolism of ix. ['4C]Palmitic Acid by Intestinal Alucosa

Proximal Distal(5,677 +554 (5,831 -+419

Mucosal radioactivity. dp.m, g mucosa) dpm, g mucosa)

H2O-soluble metaholites,%C of musocal radioactivity

Total 42.2 +2.2 34.2 +4-(30*CO2 7.9 +41.0 6.2 ± ()0.8*Ketoacidcs 2.6 +0.8 0.4 i)0.41$

Lipids, %of mucosal r}adioactitvityTotal 57.8 +2.2 65.8 +-3. 0*

Phospholipids 27.6 +2.1 35.0 +2. 1§Digly-cerides 6.0( 40.4 6.8 4+0.4Fatty acids 6.1 +0.7 6.6+(0.5Triglycerides 16.3 -+0.9 16.2 A 1.0Cholesterol esters 1.8 +0.2 1.2+0.11

['4C]Palmitate was injected i.v. into fasting rats. At 2 min, rats were killedby decapitation; intestinal mucosa was prepared and analyzed for 14C inwhole homogenate, water-soluble metaholites, and lipids (Methods).Mean+ISE, ai = 10. Significance of differences, proximal vs. distal(paired t):* P < 0.005.

P <0.05.§ P < 0.0005.

metabolites. Percent incorporation into phospholipids(30.5-4.6% ) was similar to intestinal nmucosa.

Distribution. of plasma [lC]FFA in sibccllulfar frac-tions, crypts, and ztilli of intestinal mtnucosa. About half(48%) of the "C in proximal mucosa was recovered inthe plasma menmbrane-nuclei fraction. Of the ["C]lipidin this fraction, more than three-quarters was in esters(predominantly phospholipids), indicating that con-

tamination by plasma FFA was not significant. Super-nate accounted for a lesser amount (28%), and stillsmaller quantities of "C were found in the microsomal(14%c) and mitochondrial (8%) fractions. Floating fat

50r

u

-v)

0

40

30

20

10

0

O WATERSOLUBLE

0 TRIGLYCERIDES

O PHOSPHOLIPIDS* FFA, DIGLYCERIDES, CHOLESTERYLESTERS

T

INTESTINE LIVER

FIGURE 1 Utilization of plasma FFA by intestinal mucosaand liver of fasting rats. After an overnight fast, rats wereinjected i.v. with ["C]palmitate. Radioactivity in water-soluble metabolites and lipids at 2 min was determined inhomogenates of liver and mucosa of proximal small intestine(Methods). Results are expressed as percentage of totaltissue radioactivity. Mean± SE, ii. = 4.

Intestinal Metabolism of Plasma FFA 805

TABLE I IUptake and Metabolism of ix. [P4C]Palmitic Acid by

Intestinal Villi and Crypts

Villi Crypts(3,655 4±362 (5,2 12 :±5 70

Mucosal radioactivity. dpm/g mucosa) dpm/g mucosa)*

H20-soluble metaholites,%0 of mucosal radioactivity

Total 47.0±45.5 49.2 ± 1.7C02 1 1.2 ±)0.8 3.8 ±)0.5*Ketoacids 1.9±+-0.4 0.8 ±4-0.0

Lipids, %of mucosal radioactivityTotal 53.0±)-5.5 5(0.8 i 1. 7

Phospholipids 23.1+±2.0 26.9± 1.0Diglycerides 3.4±+0.4 4.4 ±0.2Fatty acids 3.1±i0.2 9.6±4-1.41Triglycerides 20.9 ±3.4 8.8 ± 1.9*Cholesterol esters 2.5 ±t0.4 1.1 ±-0. 1

At 2 min after i.v. injection of [14C]palmitate into fasting rats, preparationsof villi and crypts from proximal intestine were analyzed for 14C in water-soluble metabolites and in lipids. Mean i±SE, n = 4. Significance of dif-ferences (paired I test):* P < 0.005.$P < 0.05.

contained only 2% of mucosal radioactivity, and wasnot further characterized. Proximal and distal intestinewere generally similar.

Water-soluble compounds accounted for 92% of super-natant "4C, whereas only 18% of microsomal '4C waswater soluble, consistent with the major role of the en-doplasmic reticulum in fatty acid esterification. Dis-tribution of label in lipid classes (phospholipids > tri-glyceride > FFA), expressed as a percentage of thetotal radioactivity in each cell fraction, did not differgreatly among particulate cell fractions and wholesonicate.

Total esterified '4C in all cell fractions was about10% less than in whole sonicate; conversely, ['4C]FFA

600 -

E

E

U

UJ1,<.V.

500-_

400_-

31

2C

1I

°-0 MUCOSALTGFA

0---* SERUM TGFA

-SA -- - - -- - - - - - - --.

2 5

I

I,

DO ~~~//l A .- -4_

/

15TIME (min)

FIGURE 2 Specific activity of triglyceride-fatty acids(TGFA) in intestinal mucosa and serum after i.v. [14C]-palmitic acid. After an overnight fast, rats received i.v. ["4C1-palmitate, and specific activities of mucosal and serum TGFAwere determined at the indicated times. Mean±-SE: 2 min,it = 10 (P < 0.0002, paired t); 15 min, it = 3; 5 and 30 min,it = 2.

in the combined cell fractions exceeded the sonicateby a comparable amount. These discrepancies mostlikely reflect the action of mucosal lipases (21) duringhomogenization and centrifugation. Although its effecton overall lipid recoveries was not of major signifi-cance, this phenomenon precludes meaningful interpre-tation of mucosal FFA specific activities.

M\Ietabolism of plasma FFA also was compared incrypts and villi. Uptake per gram tissue was some-what greater in cryts (Table II). Percent conversionto C02, triglycerides, and cholesterol esters was greaterin villi, while [14C]FFA was greater in crypts. Possiblythe crypt preparation was contaminated by plasma ["4C]-palmitate, but the relatively small absolute difference in["4C] FFA between the fractions was small comparedto the difference in uptake.

Of the total mucosal radioactivity, 21% or less wasrecovered in triglycerides (Tables I and II), raisingthe possibility that this small amount reflected con-tamination by plasma (hepatogenous) triglyceridesrather than in situ synthesis. As shown in Fig. 2, how-ever, the specific activity of mucosa triglyceride fattyacids (TGFA) 1 at 2 and 5 min was much greater thanthat of serum TGFA, thereby excluding any significantcontribution by plasma ['4C]triglyceride, and estab-lishing that plasma FFA are incorporated into tri-glycerides by intestinal mucosa.

Since triglyceride synthesis takes place in the endo-plasmic reticulum (22), it was expected that the specificactivity of microsomal TGFA would exceed that of thewhole homogenates. However, specific activities of mi-crosomal and sonicate TGFA (263±61 and 299±50dpm/iimol, respectively) did not differ. This findingsuggests that there is rapid equilibration of newlysynthesized triglyceride within the cell, and/or thatmicrosomal TGFA specific activity is diminished bythe admixture of unlabeled fatty acid absorbed fromthe intestinal lumen.

Compartmentation of fatty acids in intestinal mucosa.Fatty acids which enter intestinal mucosa from thelumen are mainly esterified to triglyceride, only a smallfraction being oxidized. In contrast, plasma FFA areincorporated into mucosal triglyceride only to a minorextent (Table I). This finding suggested compart-mentation of mucosal fatty acids.

To investigate this question further, tracer quantitiesof [3H]palmitate in 1.0 ml 10 mMtaurocholate wereadministered intraduodenally to fasting rats, simultane-ously with i.v. administration of ['4C]palmitate. In-corporation of each label into mucosal metabolites was

measured at 2 min, by which time 41% of luminal[3H]palmitate had been absorbed.

1 Abbreviations used in this paper: FABP, fatty acid-bind-ing protein; TGFA, triglyceride fatty acids.

806) A. Gangl and R. K. Ockner

A striking difference in the mucosal metabolism ofthe palmitate administered by these two routes wasobserved (Fig. 3). As expected, more than 60% of theabsorbed luminal (8H) fatty acid had been incorporatedinto triglyceride, and only 17% was recovered in water-soluble metabolites. By contrast, triglyceride accountedfor only 21% of the i.v. ["C]palmitic acid radioactivity,whereas 44% were found in water-soluble metabolites.The marked differences in metabolism of fatty acidssimultaneously administered by different routes stronglysupports the concept that there is compartmentation offatty acids in the intestinal mucosa. That this com-partnmentation does not simply reflect the anatomicaland functional heterogeneity of the mucosa (e.g., thedifferences between villi and crypts) was established byexperiments in which double isotope administration[3H] palmitate intraduodenally; [1C] palmitate i.v.) wascombined with differential mucosal scraping, permittingcomparison of villi and crypts. As shown in Fig. 4,mucosal radioactivity derived from luminal palmitatewas greater in villi than in crypts (consistent with therole of villi in fat absorption), and that from i.v. palmi-tate was slightly greater in cryptes (as in Table II).Despite this, in both villi and crypts the major fateof Plasma FFA was oxidation to water-soluble metabo-lites, whereas this fate accounted for only a minor frac-tion of lunlinal fatty acids (P < 0.005).

These findings strongly suggest that the observedcompartmentation of fatty acids is intracellular. Inother words, the metabolism of fatty acid in the epi-thelial cell appears to be determined by its site of entryinto the cell. The basis for this compartmentation re-mains to be elucidated, but may be related in somemanner to the structural polarity of the cell (see Dis-cussion).

Incorporation of plasmza FFA into initestinial lyinphlipids. To determine whether mucosal lipids synthe-sized from plasma FFA were incorporated into intes-tinal lipoproteins and secreted into lymph, rats withintestinal lyvmph fistulas were studied. After an over-night fast, ["C] palmitate was administered i.v., andintestinal lymph and blood were collected at intervalsup to 2 h; in some experiments, proximal mucosa wasstudied at 10, 20, and 30 min. Lymph flow averaged2.7±0.2 ml/h, and lymph TGFA concentration was6.28±0.68 /Amol/ml. Results are shown in Fig. 5.

Consistent with the findings at 2 and 5 mm in ratswithout lymph fistulas (Fig. 2), specific activity ofmucosal TGFA at 10 min in fistula rats was signifi-cantly higher than in serum (Fig. 5). Labeled tri-glyceride, which first appeared in intestinal lynmphsafter 10 miii, increased in specific activity between 20and 30 min. Although the data do not conclusivelyestablish a precursor-product relationship between mu-

70 r.sE

v

0

0

0a

0

u

60k

50 e-

40

30 e-

20 H

10 e

0

Q TG

Q H20 SOLUBLE

LUMINAL H)14

i.v. C)

FIGURE 3 Intestinal metabolism of palmitic acid: effect ofroute of administration. After an overnight fast, rats re-ceived, simultaneously, ["C]palmitate i.v. and ['Hipalmitateintraduodenally. Radioactivity of triglycerides (TG) andwater-soluble metabolites at 2 min was determined andexpressed as a percentage of the total mucosal radio-activity for each isotope. Mlean-+SE, t = 4. Significance ofdifferences (paired t) in triglycerides and water-solublemetabolites, respectively (i.v. vs. intraluminal), were P <0.001 and < 0.05.

cosal and lymph triglyceride, available evidence indi-cates that plasma very low density lipoproteins (theo-retical alternative "precursors") do not enter intestinallymph (23, 24).

Of the lipid-extractable radioactivity in lymph duringthe 20-30 min interval, only 10.0±2.8% was present inphospholipids, whereas triglyceride accounted for 68.5±5.2%, despite the fact that in the mucosa incorpora-

LU

-

0

co

L)co

0CM

I

50

E40

0 _.0

0° > 30E *j

° w 200

_ m00

I-

10

LUMINAL (3H)El i.v. (4c)

OSF]

+_

[2'

±

0 L I I I I i IVILLI CRYPTS

TOTAL MUCOSAL 8.2±0.3 6.2±0.8 3.4+0.2 7.2+0.7RADIOACTIVITY x105 X 13 X105 X 13(dpm/gl

FIGuRE 4 Intestinal metabolism of luminial and i.v. palmiticacid: villi vs. crypts. After an overnight fast, rats receivedsimultaneous i.v. and intraduodenal ["C] - and ['H 1palmitatc(as in Fig. 3). Radioactivity in homogenates and water-soluble metabolites of villi and crypts at 2 mmil was deter-mined and expressed as a percentage of the total mucosalradioactivity for each isotope. Mean± SE, nt = 4.

Intestinal Metabolism of Plasma FFA 807

F

0E

700-a 700.. AMUCOSA

600 ~~~~~~0LYMPH°- 600 30 60 SERUM

500-

U~ 00

~'300-U3

200-

100 1

15 30 45 60 75 90 105 120

TIME (min)

FIGURE 5 Specific activity of triglyceride fatty acids(TGFA) in intestinal mucosa, lymph, and serum after i.v.["4C]palmitate. After an overnight fast, lymph fistula ratsreceived i.v. ["C] palmitate. Intestinal lymph was collectedfor 10-immin periods until 30 min, and for 30-min periods to2 h. In experiments ended at 10, 20, and 30 min, mucosalhomogenates were prepared and blood was collected afterdecapitation; in more prolonged experiments, blood wascollected from the tail vein at the midpoint of the indicated30-min lymph collection periods. TGFA specific activity wasdetermined as in Methods. Mean+SE, it . 3. Values forlymph are plotted at the midpoint of the collection period.

tion of plasnma FFA into plhospholipid greatly exceededthat into triglyceride (Table I).

IEffect of glucose and ethaniol ont the intestinal me-tabolisin of plasina FFA. Rats received a constant

intraduodenal infusion of glucose (20% wt/vol in iso-tonic saline) over 15 h (40 kcal/rat), resulting in a25% fall in serum FFA concentration (Table III).Since tissue uptake of plasma FFA is proportional toFFA concentration (25, 26), this suggests that mucosaluptake of plasma FFA was diminished during glucoseadministration, possibly contributing to the observeddecrease in mucosal FFA concentration (Table III).(Mucosal FFA concentrations may be spuriously in-creased by about 10% due to action of mucosal lipasesduring tissue preparation, but it is unlikely that thesignificant differences among the three groups can beattributed to this.) Although a somewhat lower per-centage of plasma FFA entering the mucosa underthese conditions was oxidized, metabolism was generallysimilar to that in fasting controls (Table IV).

In another group of animals, ethanol (0.33 g/kg/h)was substituted isocalorically for a portion (< 1/3)of the glucose, and the two substances were infusedsimultaneously over 15 h. The rate at which ethanolwas infused in these experiments was similar to itsmaximal rate of metabolism (27). These animals ex-hibited no evidence of intoxication, and could not bedistinguished on inspection from glucose-infused con-trols.

Relative to fasting (saline-infused) controls, therewas a 40% fall in serum FFA concentration during theethanol-glucose infusion (Table III), suggesting thatnet mucosal uptake of plasma FFA was diminishedcorrespondingly. Despite the low infusion rate of etha-nol, it resulted in a significantly increased triglycerideconcentration in serum and, to a lesser extent, in mu-cosa (Table III). The most striking effect of theethanol-glucose infusion, however, was a 60-65% de-crease in the conversion of i.v. ['4C]palmitate to water-soluble (oxidative) metabolites, and a near doubling ofits incorporation into triglyceride, relative to both sa-

TABLE IIIEffect of Glucose and Ethanol on FFA and Triglycerides in Serum and Intestinal Mucosa

FFA Triglyceride

Serum Mucosa Serum Mucosa

pmol/ml Amol/g JAmolml ,Amol/gSaline 0.51 +0.02 1.57+0.21 0.30+0.02 0.84+0.07Glucose 0.38+0.03 0.87+0.10 0.38+0.04 1.054+0.13Ethanol-glucose 0.30+0.06 1.08+0.19 0.63 +0.08 1.26+0.11

Analysis of variance F = 10.63 F = 4.38 F = 18.54 F = 4.55P < 0.01 P < 0.05 P < 0.01 P < 0.05

Rats received intraduodenal infusions, over 15 h, or 0.85% saline, glucose (2 g/kg/h), orethanol (0.33 g/kg/h) combined with glucose (1.42 g/kg/h). Serum and mucosal concentra-tions of FFA and triglycerides were determined. Mean+SE, na 7, compared by one wayanalysis of variance for unequal sample sizes.

808 A. Gangl and R. K. Ockner

TABLE IREffect of Glucose and Ethanol on Uptake and Metabolism of iv. [14C]Palmitic Acid by Intestinal MVucosa

Ethanol and Analysis ofSaline Glucose glucose variance

F P

MIuicosal radioactivity, dpm/g mucosa 5,6774+554 4,516+447 6,220±)408 2.13 NS

H11O-soluble metabolites,%of mucosal radioactivity 42.2±2.2 35.943.3 14.745.4 15.87 <(.(1

Lipids, %of mucosal radioacttivitPhospholipids 27.6±2.1 35.7+3.1 40.3+ 7.1 3.56 <0.05Diglycerides 6.040.4 7.8±40.8 10.9+1.0 11.19 <0.01Fatty acids 6.1+(0.7 4.4±i0.4 3.7 +0.6 3.30 NSTriglycerides 16.3±0.9 14.5+1.2 27.0±44.9 9.31 <0.01Cholesterol esters 1.8±(0.2 1.7±40.2 3.4±(0.9 4.88 <0.05

After overnight infusions of saline (n = 10), glucose (n = 7), or ethanol-glucose (n = 4) as in Table V, rats wereinjected i.v. with ['4C]palmitate; 2 min later radioactivity in mLucosal metabolites was determined (Methods).Mleani±SE, compared by one way analysis of variance for unequal sample sizes.

line and glucose-infused rats (Table IV). A slightincrease in the incorporation of plasma FFA into tri-glyceride by distal mucosa (not shown) was statisti-cally insignificant, suggesting that the effect of ethanolwas localized to its site of absorption. Incorporation ofisotope into other esterified lipids was also increasedsomewhat. Differences in total mucosal radioactivityamong the three groups (Table IV) may reflectchanges in mucosal blood flow, with correspondingeffects on FFA uptake.

Glucose and ethanol-glucose infusions did not affecthepatic FFA uptake. Similar to intestinal mucosa,however, ethanol-glucose infusion was associated witha decrease in the conversion of plasma FFA to water-soluble metabolites (15.1±3.1 and 13.8±2.2% of totalhepatic radioactivity in fasting and glucose-infused,respectively; 3.9±11.7% in ethanol infused), with cor-responding increases in esterification.

Effect of fat absorption oil intestinal metabolism ofplasma FFA and their incorporation. into lymph lipids.

TABLE VEffect of Intraduodenal Taurocholate and Lipid on Uptake and Metabolism of i..

[E4C]Palmitic Acid by Intestinal Mucosa

Lipid in 10 mMtaurocholate

Micelles EmulsionSaline Taurocholate (8.2 mg total (82.0 mg total Anal-sis of0.85% 10 mM lipid) lipid) variance

F P

Mucosal radioactivity, dpm/g mucosa 4,671±368 5,208±687 6,188±810 10,398±623 19.13 <0.01

H20-soluble metabolites,%of mucosal radioactivity 51.2±2.3 57.2±2.3 46.3±3.8 42.6±3.4 4.23 <0.05

Lipids, %of mucosal radioactiv.itPhospholipids 22.9i0.4 19.1±0.6 24.3±1.2 24.0±2.1 3.31 NSDiglycerides 3.5±0.2 3.3±0.3 3.8±0.3 4.4±0.3 3.82 <0.05Fatty acids 5.0±1.(0 6.0±0.6 6.3±1.4 4.8±0.5 0.60 NSTriglycerides 15.7+1.6 13.1±2.3 18.6±i3.8 23.9±2.6 2.79 NSCholesterol esters 1.7±0.2 1.4±0.1 0.7±:0. 0 0.3±0. 0 36.25 <(1.01

Serum FFA, MUmol/lmil 0.49±0.07 0.52±0.07 0.66±4-0.05 0.47±0.07 1.90 NS

After an overnight fast, rats received 30 min intraduodenal infusions of saline (n = 3), 10 mMI taUrocholate (n = 4), mixedmicelles in 10 mMtaurocholate (n = 4), or lipid emulsion in 10 mMl tatirocholate (n = 4), followed by i.v. [E4C]palmitate.Serum FFA concentration and radioactivity in mucosal metabolites were measured at 2 mill (Methods). Meani±SE, coIm-pared by one way analyNsis of variance for unequal sample sizes.

Intestinal Metabolism of Plasma FFA 809

1,600 1e-0 INTRADUODENALLIPID

1,400 .0-O INTRADUODENALSALINE

E1,200

X 1p00 _U/

V 800 _

f 600

200 I

5 15 25TIME (min)

FIGURE 6 Effect of fat absorption on incorporation ofplasma FFA into intestinal lymph triglyceride. After anovernight fast, lymph fistula rats received 30-min intra-duodenal infusions of saline (n = 4) or of an emulsion con-taining 82 mg lipid in 10 mMtaurocholate (t = 3) (Meth-ods). ['4C] Palmitate was then administered i.v., and lymphwas collected in 10-min periods up to 30 min. Lymph trigly-cerides were isolated and assayed for radioactivity. Mean±SE of cumulative 14C in triglycerides are plotted at themidpoint of collection periods. The difference during thesecond collection period (10-20 min) falls just short ofstatistical significance; for the 20-30 min period, P < 0.002.

After an overnight fast, rats received intraduodenalinfusions of lipid in 10 mMtaurocholate during the 30min period preceding, i.v. ['4C]palmitate. As shown inTable V, administration of 10 mMtaurocholate alone(or, not shown, 30 mMtaurocholate) caused insignifi-cant changes in mucosal uptake and metabolism ofplasma FFA, compared to saline controls. Infusion of10 mMoleic acid and 5 mMglyceryl monooleate (8.2mg total lipid) in 10 mMtaurocholate caused only aslight increase in uptake and percent esterification,compared to taurocholate alone. However, an emulsioncontaining 10 times this amount of oleic acid andglyceryl monooleate (82 mg total lipid) in 10 mMtaurocholate approximately doubled the total mucosaluptake of i.v. ['4C]palmitate, and caused a relative de-crease in oxidation and increase in incorporation intotriglyceride. (The increase in mucosal FFA uptake inthe absence of a rise in plasma FFA concentration mostlikely is due to the increased intestinal blood flow asso-

ciated with fat absorption (28, 29)). Thus, incorpora-

tion of i.v. ['4C]palmnitate into mucosal triglyceride wasincreased nearly fourfold during fat absorption, comn-pared with the taurocholate-infused rats. Mucosal oxi-dation of plasma FFA (despite its relative decrease),increased absolutely, as did incorporation into phos-pholipids and diglycerides. The decreased percent in-corporation of plasma ['4C] palmitate into cholesterolesters may reflect preferential esterification of choles-terol with absorbed oleic acid (30). Serum triglycerideconcentrations (0.23-0.37 mMi) did not differ sig-nificantly among the groups, indicating that the ab-sorbed fat had not entered plasma to any importantextent.

Intestinal metabolism of plasma FFA also was studiedun(ler conditions designed to diminish uptake of lunmi-nal lipid. In these experiments, cholestyramine (0.45g/kg; Questran Mead Johnson Labs., Evansville, Ind.)was infused intraduodenally during the 30 min preced-ing i.v. ["4C]palmitate administration. No effect on up-take or metabolism of plasma FFA by intestinal mucosawas observed (unpublished observations).

The enhanced incorporation of plasma FFA intomucosal triglyceride during the absorption of intra-luminal fat suggested the possibility that plasma FFAmight contribute to lymph triglycerides to a greaterextent under these conditions than in the fasting state.To explore this possibility, rats with intestinal lymphfistulas received an intraduodenal infusion of a lipidemulsion (82 mg) in 10 mt\I taurocholate over 30 minas in Table V; controls received saline. Immediatelyat the completion of this infusion, [3H]palmitate wasadministered intraduodenally, simultaneously with thei.v. administration of ["4C] palmitate. Intestinal lymphwas then collected for three 10-min periods.

At 30 min after the administration of isotope, mu-cosal "4C (from plasma FFA) was significantly higherin lipid-infused animals than in controls (7,168±998vs. 4,631±328 dpm/g mucosa, P < 0.05). Furthermore,mucosal compartmentation of luminal and plasma fattyacids persisted at this later time period. Thus, of theradioactivity remaining in the mucosa at 30 min, the4C (i.v.) was predominantly in phospholipids (40.3%)and water-soluble metabolites (37.8%), whereas mostof the 3H (luminal) was in triglyceride (54.7%).

The effect of fat absorption on the lymphatic outputof i.v. ['4C]palmitate as triglyceride is shown in Fig. 6.The results show that fat absorption is associated not

only with increased mucosal uptake of plasma FFA andincorporation into triglyceride (Table V), but alsowith a marked increase in the mucosal secretion oflymph triglyceride derived from plasma FFA. A lesssignificant increase in lymphatic output of [14C]phos-pholipid also occurred (unpublished observations).

810 A. Gangl and R. K. Ockner

DISCUSSION

The present studies show that, in addition to absorbingluminal fatty acid, the intestinal mucosa also utilizesplasma FFA. This function is not unexpected, since itis shared by many tissues in the body, but the findingthat fatty acids derived from the two sources (lumenand plasma) are metabolized so differently by the samecell population is somewhat surprising and, to ourknowledge, unique. Thus, although there appear to betwo FFA pools within the adipocyte (31), this amountsto a separation of fatty acids entering this cell typefrom those which are exiting, presumably via a com-mon portal. Teleologically, it is understandable thatthese two distinct adipocyte functions (fatty acid uptakeand release) should be separated, since they are re-sponsive to divergent regulatory factors (plasma FFAconcentration and adipocyte hormone-sensitive I ipase,respectively), but the mechanism for this separationis not known.

In the intestine, our studies show that fatty acidswhich enter mucosa from plasma are destined, for themost part, to serve the structural and caloric require-ments of the epithelium itself, whereas those absorbedfrom the lumen serve the organism as a whole. As withthe adipocvte, compartmentation can be understood onteleological grounds but its mechanism is obscure. Pos-sibly, structural factors play a role in this regard. Inthe intestinal epithelium, mitochondria are dispersedthroughout the absorptive cell, in contrast to the endo-plasmic reticulum which is relatively concentrated inthe apical and supranuclear portions (32). Accordingly,at the basal pole of the cell, i.e., the region at whichFFA must certainly enter from plasma, there is a rela-tive preponderance of mitochondria. This structuralpolarization of the epithelial cell could account for anincreased probability that fatty acids entering the basalaspect of the cell would be oxidized. Similarly, theapical preponderance of endoplasmic reticulum couldexplain the predominant esterification of fatty acidswhich enter from the lumen. This tentative scheme isbased on the assumption that the metabolic fate of agiven fatty acid molecule is determined by the enzymaticproperties of the particular cell organelle with whichit associates. Thus, fatty acid activation to its acyl-CoAthioester in the endoplasmic reticulum (22. 33) wouldbe followed by transacylation and glycerolipid synthesis,whereas activation at the mitochondrial surface (34)would favor acylcarnitine synthesis and intramitochon-drial oxidation. This assumption requires only thatlong-chain fatty acyl thioesters, once formed in endo-plasmic reticulum or mitochondria, not equilibrate in acommon intracellular pool. In fact, compartmentationof long-chain acvl-CoA is implicit in the observationthat luminal and plasma fatty acids undergo differing

metabolic fates in the absorptive cell, despite the factthat they are converted to this common intermediate.

An alternative explanation for the observed com-partmentation is that the greater mass of luminal fattyacid in some manner "preempts' 'the triglyceride path-ways. This explanation is unlikely, however, since ourstudies show that neither increased nor decreased lumi-nal fatty acid influx significantly affected the proba-bility that plasma FFA would undergo oxidation inthe mucosa.

An additional possibility is that compartmentation isrelated to cytoplasmic fatty acid-binding protein (FA-BP). Available evidence suggests that in the intestine,FABP participates in the intracellular transport andmetabolism of absorbed fattv acid (35-37). Althoughit is likely that FABP also participates in the utilizationof plasma FFA, its possible role in fatty acid com-partmentation requires further study.

The present studies show that plasma FFA are anenergy source for the intestinal mucosa. However, al-though i.v. palmitate was oxidized to a relatively greaterextent than luminal palmitate, this does not necessarilyindicate that plasma FFA are quantitatively more im-portant in this regard. Thus, calculating from the dataof Baker and Schotz (9), net plasma FFA flux in thefasting rat is 432 Mmol/h/300 g. Our findings showthat about 1 % of this amount, or 4.3 IAmol/h, enters themucosa, of which about 45%, or 1.9 u'mol/h, are oxi-dized. (Since nmucosal uptake was determined at 2 min,or slightly more than one half-life of plasma FFA, itis likely that steady-state uptake is somewhat greaterthan the calculated values.)

For endogenous luminal fatty acids, on the otherhand, flux through the mucosa equals or exceeds fattyacid output in lymph lipids, since the fasting contribu-tion from other sources may be regarded as negligible(2, 23, 38). This output is approximately 17.0 /Amol/ h/300 g (see above, and 2, 23, 39), or about five times themucosal uptake of plasma FFA. Since about 16% ofabsorbed endogenous fatty acids are oxidized (Fig. 3),or 2.7 jtmol/h, endogenous luminal fatty acids provideabout 58% of the fatty acid which is oxidized bywhole mucosa. Presumably, oxidation of luminal fattvacid is more important in the villi, whereas crypt cellsare more dependent on plasma FFA.

Glucose had little effect on intestinal fatty acid me-tabolismi. The finding that plasma FFA are convertedto ketone bodies in intestinal mucosa suggests that theintestine, along with other extrahepatic tissues, includ-ing kidney (40, 41) and forearm muscle (42), may con-tribute to circulating ketones, but this contribution isprobably minor.

Plasma FFA were incorporated into mucosal phos-pholipids to a significant extent. Of these, a relatively

Intcstinal Metabolism of Plasma FFA 811

small quantity was secreted into intestinal lymph, sug-gesting that they served a structural role primarily.The a-glycerophosphate pathway (43) almost certainlyis involved, since Johnston, Paultauf, Schiller, andSchultz (44) showed that, in intestinal microsomes, onlydiglycerides formed via the a-glycerophosphate pathwaycould be incorporated into phospholipid.

In general, these studies confirm earlier evidence thatplasma FFA are incorporated into lymph lipoproteins(1), but the quantitative significance of this contribu-tion appears to be small. Although incorporation ofplasma FFA into mucosal triglycerides was not af-fected by glucose, it was significantly increased in twoother experimental circumstances, i.e., ethanol adminis-tration and absorption of exogenous lipid.

In these studies, the concentration of ethanol in theintraduodenal infusion was only 3.4% wt/vol, andhemorrhagic mucosal erosions seen with higher con-centrations (45) did not occur. The effect of ethanolon intestinal metabolism of plasma FFA resembled thatin liver, in that esterification was increased at theexpense of oxidation. Since mucosal oxidation of etha-nol proceeds via the alcohol dehydrogenase pathway(46, 47), the mechanism for this effect presumably issimilar to that in the hepatocyte, and reflects changesin mucosal NADH/NADratio. In contrast to previousstudies (3), mucosal triglyceride concentrations wereonly nminimally elevated after ethanol, probably reflect-ing different schedules of administration. Our presentfindings support the concept that ethanol inhibition ofmucosal oxidation of FFA (luminal and plasma) con-tributed to the observed increase in mucosal and lymphtriglycerides (3).

The intraduodenal infusion of 82 mg of lipid (ap-proximately 15-20% of average daily dietary intake)essentially doubled the mucosal uptake of plasma FFAand its relative incorporation into triglycerides, andcaused a slight relative decrease in fatty acid oxidation.Since plasma FFA levels remained unchanged, the in-creased uptake probably reflected increased mucosalblood flow (28). As a result, absolute mucosal metabo-lism of plasma FFA increased significantly during fatabsorption and its output in lymph triglyceride was in-creased sixfold. Intraduodenal taurocholate alone hadno significant effect. We explored the possibility thatincreased mucosal FFA uptake during fat absorptionmight be caused by cholecystokinin (1.5 Ivy U/kg rat,i.v.), a hormone which is released during fat absorp-tion and which causes increased nmesenteric blood flow(29), or glucagon (3 or 30 Mg/kg), which increases in

pllasma during fat absorption (48). However, the resultsof these experiments were inconclusive (unpublishedobservations).

ACKNOWLEDGMENTSWe gratefully acknowledge the expert technical help pro-vided by Miss Winnie Leong.

This work was supported in part by the Max KadeFoundation, Research grants AM-13328, AM-14792, andResearch Career Development Award AM-36586 from theNational Institutes cf Health, and the Walter C. Pew Fundfor Gastrointestinal Research.

REFERENCES1. Havel, R. J., and A. Goldfien. 1961. The role of the

liver and of extrahepatic tissues in the transport andmetabolism of fatty acids and triglycerides in the dog.J. Lipid Res. 2: 389-395.

2. Baxter, J. H. 1966. Origin and characteristics of en-dogenous lipid in thoracic duct lymph in rat. J. LipidRes. 7: 158-166.

3. Mistilis, S. P., and R. K. Ockner. 1972. Effects of etha-nol on endogenous lipid and lipoprotein metabolism insmall intestine. J. Lab. C/in. Med. 80: 34-46.

4. Gangl, A., and R. Ockner. 1974. Intestinal metabolismof plasma free fatty acids (FFA). Clin. Res. 22: 358A.

5. Gangl, A., and R. Ockner. 1974. Intestinal metabolismof plasma free fatty acids (FFA). Gastroen~terology.66: 187A.

6. Bollman, J. L., J. C. Cain, and J. H. Grindlay. 1948.Techniques for the collection of lymph from the liver,small intestine, or thoracic duct of the rat. J. Lab.Clin. Med. 33: 1349-1352.

7. Bollman, J. L. 1948. A cage which limits the activityof rats. J. Lab. Cliii. Mcd. 33: 1348.

8. Fredrickson, D. S., and R. S. Gordon, Jr. 1958. Trans-port of fatty acids. Physiol. Rev. 38: 585-630.

9. Baker, N., and M. C. Schotz. 1967. Quanitative aspectsof free fatty acid metabolism in the fasted rat. J. LipidRes. 8: 646-660.

10. Dietschy, J. M., and M. D. Siperstein. 1965. Cholesterolsynthesis by the gastrointestinal tract: localization andmechanisms of control. J. C/in. Invest. 44: 1311-1327.

11. Rodgers, J. B., E. M. Riley, G. D. Drummey, and K.J. Isselbacher. 1967. Lipid absorption in adrenalecto-mized rats: the role of altered enzyme activity in theintestinal mucosa. Gastroecnterology. 53: 547-556.

12. Folch, J., M. Lees, and G. H. Sloane Stanley. 1957.A simple method for the isolation and purification oftotal lipids from animal tissues. J. Biol. C/win. 226: 497-509.

13. Regouw, B. J. M., P. J. H. C. Cornelissen, R. A. P.Helder, J. B. F. Spijkers, and Y. M. M. Weeber. 1971.Specific determination of free fatty acid in plasma.Clini. Chimn. Acta. 31: 187-195.

14. Skipsky, V. P., R. F. Peterson, and M. Barclay. 1964.Quantitative analysis of phospholipids by thin-layerchromatography. Biochemt. J. 90: 374-378.

15. Eggstein, M., and F. H. Kreutz. 1966. Eine neue Bes-timmung der Neutralfette im Blutserum und Gewebe.I. Mitteilung. Prinzip, Durchfuhrung und Besprechungder Method. Klin. Wochenschr. 44: 262-267.

16. Ontko, J. A., and D. Jackson. 1964. Factors affectingthe rate of oxidation of fatty acids in animal tissues.Effect of substrate concentration, pH, and coenzyme Ain rat liver preparations. J. Biol. Chem. 239: 3674-3682.

17. Edson, N. L. 1935. Ketogenesis-antiketogenesis. I. Theinfluence of ammonium chloride on ketone-body forma-tion in liver. Biochcem. J. 29: 2082-2094.

812 A. Gangl and R. K. Ockner

18. Snedecor, G. W., and W. G. Cochran. 1967. StatisticalMethods. Iowa State University Press, Ames, Iowa.6th edition.

19. Sauer, F., S. Mahadevan, and J. D. Erfle. 1971. Theaccumulation of citrate cycle intermediates in rat livercells oxidizing palmitate. Biochiimi. Biophzys. A eta. 239:26-32.

20. Bragdon, J. H., and R. S. Gordon, Jr. 1958. Tissuedistribution of C4 after the intravenous injection oflabeled chylomicrons and unesterified fatty acids in therat. J. Clin. Invest. 37: 574-578.

21. DiNella, R. R., H. C. Meng, and C. R. Park. 1960.Properties of intestinal lipase. J. Biol. Chew. 235: 3076-3081.

22. Johnston, J. M. 1968. Mechanism of fat absorption.Hanidb. PlhVsiol. 3(Sect. 6): 1353-1375.

23. Ockner, R. K., F. B. Hughes, and K. J. Isselbacher.1969. Very low density lipopropteins in intestinal lymph:origin, composition, and role in lipid transport in thefasting state. J. Clin. Invest. 48: 2079-2088.

24. Ockner, R. K., and A. L. Jones. 1970. An electronmicroscopic and functional study of very low densitylipoproteins in intestinal lymph. J. Lipid Rcs. 11: 284-292.

25. Armstrong, D. T., R. Steele, N. Altszuler, A. Dunn,J. S. Bishop, and R. C. De Bodo. 1961. Regulation offree fatty acid turnover. Ami. J. Physiol. 201: 9-15.

26. Spector, A. A., D. Steinberg, and A. Tanaka. 1965.Uptake of free fatty acids by Ehrlich ascites tumorcells. J. Biol. Chewn. 240: 1032-1041.

27. Le Breton, E. 1936. Evolution de la vitesse d'oxydationde l'alcohol 6thylique chez homeothermie au cours dela croissance. C. R. Soc. Biol. 122: 561-564.

28. Jacobson, E. D. 1967. The circulation of the gastroin-testinal tract. Gastroeniterology. 52: 98-112.

29. Fara, J. W., E. H. Rubinstein, and R. R. Sonnenschein.1972. Intestinal hormones in mesenteric vasodilationafter intraduodenal agents. Ams. J. Plysiol. 223: 1058-1067.

30. Karmen, A., M. Whyte, and DeW. S. Goodman. 1963.Fatty acid esterification and chylomicron formationduring fat absorption. 1. Triglycerides and cholesterolesters. J. Lipid Res. 4: 312-321.

31. Vaughan, M., D. Steinberg, and R. Pittman. 1964. Onthe interpretation of studies measuring uptake and es-terification of [J-"C]palmitic acid by rat adipose tissuein vitro. Biochinm. Biophlys. Acta. 84: 154-166.

32. Trier, J. S. 1968. Morphology of the epithelium of thesmall intestine. Handb. Physiol. 3(Sect. 6): 1125-1175.

33. Senior, J. R., and K. J. Isselbacher. 1960. Activation oflong-chain fatty acids by rat-gut mucosa. Biochiimt. Bio-phys. Acta. 44: 399-400.

34. Clark, B., and G. Hiibscher. 1961. Biosynthesis of glyce-

rides in subcellular fractions of intestinal miucosa. Bio-chini. Biophys. Acta. 46: 479-494.

35. Ockner, R. K., J. A. Manning, R. B. Poppenhausen, andW. K. L. Ho. 1972. A binding protein for fatty acidsin cytosol of intestinal mucosa, liver, myocardium, andother tissues. Sciceicc (Wash. D. C.). 177: 56-58.

36. Ockner, R. K., and J. A. Manning. 1974. Fatty acid-binding protein in small intestine. Identification, isolation,and evidence for its role in cellular fatty acid transport.J. Cliii. Invest. 54: 326-338.

37. Ockner, R. K., and J. A. Manning. 1974. Intestinalfatty acid binding protein (FABP) : studies of physio-logical function. J. Clin. Invest. 53: 57a. (Abstr.)

38. Jones, A. L., and R. K. Ockner. 1971. An electron mi-croscopic study of endogenous very low density lipo-protein production in the intestine of rat and man. J.Lipid Res. 12: 580-589.

39. Shrivastava, B. K., T. G. Redgrave, and W. J. Sim-monds. 1967. The source of endogenous lipid in thethoracic duct lymph of fasting rats. Q. J. Exp. Physiol.Cogyn. Med. Sci. 52: 305-312.

40. Owen, 0. E., P. Felig, A. P. Morgan, J. Wahren, andG. F. Cahill, Jr. 1969. Liver and kidney metabolismduring prolonged starvation. J. Cliii. Inrvest. 48: 574-583.

41. Weidmann, M. J., and H. A. Krebs. 1969. The fuel ofrespiration of rat kidney cortex. Biochlewi. J. 112: 149-166.

42. Hagenfeldt, L., and J. Wahren. 1968. Human forearmmuscle metabolism during exercise. III. Uptake, re-lease and oxidation of f3-hydroxybutyrate and observa-tions on the j3-hydroxybutyrate/acetoacetate ratio. Scand.J. Cliji. Lab. Iiiz'est. 21: 314-320.

43. Johnston, J. M. 1959. The absorption of fatty acids bythe isolated intestine. J. Biol. Cheni. 234: 1065-1067.

44. Johnston, J. M., F. Paultauf, C. M. Schiller, and L. D.Schultz. 1970. The utilization of the a-glycerophosphateand monoglyceride pathways for phosphatidyl cholinebiosynthesis in the intestine. Biochiin. Bioph vs. A eta.218: 124-133.

45. Baraona, E., R. C. Pirola, and C. S. Lieber. 1974. Smallintestinal damage and changes in cell population pro-duced by ethanol ingestion in the rat. Gastroctecrology.66: 226-234.

46. Spencer, R. P., K. R. Brody, and B. M. Lutters. 1964.Some effects of ethanol on the gastrointestinal tract.Ami. J. Dig. Dis. 9: 599-604.

47. Mistilis, S. P., and A. Birchall. 1969. Induction ofalcohol dehydrogenase in the rat. Natnre (Lond.). 223:199-200.

48. Bdttger, I., R. Dobbs, G. R. Faloona, and R. H. Unger.1973. The effects of triglyceride absorption upon glu-cagon, insulin, and gut glucagon-like immunoreactivity.J. Cliii. Iiivest. 52: 2532-2541.

Intestinal Metabolism of Plasma FFA S183