Dietary fish oil stimulates hepatic low density lipoprotein transport in the rat

Dietary fish oil stimulates hepatic low density

lipoprotein transport in the rat.

M A Ventura, … , L A Woollett, D K Spady

J Clin Invest.

1989;

84(2)

:528-537.

https://doi.org/10.1172/JCI114195

.

These studies were undertaken to examine the effect of fish oil, safflower oil, and

hydrogenated coconut oil on the major processes that determine the concentration of low

density lipoprotein (LDL) in plasma, i.e., the rate of LDL production and the rates of

receptor-dependent and receptor-inreceptor-dependent LDL uptake in the various organs of the body. When

fed at the 20% level, fish oil reduced plasma LDL-cholesterol levels by 38% primarily by

increasing LDL receptor activity in the liver. Dietary safflower oil also increased hepatic LDL

receptor activity; however, since the rate of LDL production also increased, plasma

LDL-cholesterol levels remained essentially unchanged. Hydrogenated coconut oil had no effect

on LDL receptor activity but increased the rate of LDL-cholesterol production causing

plasma LDL-cholesterol levels to increase 46%. Dietary fish oil had no effect on the

receptor-dependent transport of asialofetuin by the liver, suggesting that the effect of fish oil

on hepatic LDL receptor activity was specific and not due to a generalized alteration in the

physical properties of hepatic membranes. Finally, dietary fish oil increased hepatic

cholesteryl ester levels and suppressed hepatic cholesterol synthesis rates, suggesting that

the up-regulation of hepatic LDL receptor activity in these animals was not simply a

response to diminished cholesterol availability in the liver.

Research Article

Dietary

Fish

Oil Stimulates Hepatic

Low

Density Lipoprotein Transport

in

the

Rat

Mark A.Ventura, Laura A.Woollett,and David K. Spady

DepartmentofInternalMedicine, University ofTexasSouthwesternMedicalCenter, Dallas, Texas 75235-9030

Abstract

Thesestudieswereundertakentoexamine theeffect of fish

oil,

safflower

oil,

and

hydrogenated

coconutoilonthe

major

pro-cessesthatdetermine the concentration oflow

density

lipopro-tein

(LDL)

in

plasma,

i.e.,

therateof LDL

production

and the rates of receptor-dependent and

receptor-independent

LDL uptake in the variousorgansof the

body.

When fedatthe 20% level,fish oil reduced plasma LDL-cholesterol levels

by

38% primarily by

increasing

LDLreceptor

activity

in theliver. Di-etary

safflower

oil also increased

hepatic

LDL receptor activ-ity;however, since therateofLDL

production

also

increased,

plasma LDL-cholesterol levels remained

essentially

un-changed.Hydrogenated coconutoil hadnoeffectonLDL

re-ceptor

activity

butincreasedthe rateofLDL-cholesterol pro-duction

causing plasma

LDL-cholesterol levels to increase 46%.

Dietary fish

oil hadnoeffectonthe

receptor-dependent

transportof asialofetuin

by

the

liver, suggesting

thattheeffect offish oilon

hepatic

LDL receptor

activity

was

specific

andnot

due to a

generalized

alteration in the

physical properties

of hepatic membranes.

Finally, dietary

fish oil increased

hepatic

cholesterylesterlevels and

suppressed

hepatic

cholesterol syn-thesisrates,

suggesting

that the

up-regulation

of

hepatic

LDL receptor

activity

in theseanimalswas not

simply

aresponseto

diminished cholesterol

availability

in theliver.

Introduction

Epidemiologic studies

suggest that the

consumption

ofadiet rich in marine

lipids

is associated withareduction in the inci-dence ofcoronaryheart

disease

(1,

2).

Theactivecomponent

of fish oil

appears to be the

long-chain

w-3

polyunsaturated

fatty acids (eicosapentaenoic

acid and docosahexaenoic

acid),

which

are knowntoaffect both platelet

function

andplasma

lipid

levels (3-5).

Dietary fish oil markedly

reduces plasma

triglyceride

levels in normal and

hyperlipidemic subjects

and is much more

active

than equalamountsof polyunsaturated vegetable oil inthis regard (6-10).

Dietary fish

oil has also been showntolower

plasma

cholesterol levels in normal and

hyper-lipidemic

subjects and, in several reports, was aseffective, if not more

effective,

than equal amounts of polyunsaturated vegetable oil (6, 7, 11-15). Furthermore, the reduction in plasma cholesterol levelsseenwithfish oil feedingis due

pri-marily

to adecrease in cholesterol carried inthe very low and low density lipoprotein (VLDL and LDL) fractions. High

density

lipoprotein

(HDL)cholesterol, incontrast, usually

re-Addressreprintrequeststo Dr.Spady.

Receivedforpublication 4 January 1989 and in revised form 14 March 1989.

mainsunchangedoractuallyincreases(5, 6, 10, 15). Although dietary fish oil usually lowers plasma total and LDL-choles-terollevels,themagnitude of thisresponsevaries widelyfrom studytostudy. Thegreatestreductions in total and LDL-cho-lesterol levels have been achieved when the triglyceride nor-mally present in the diet is largely replaced by fish oil. In contrast, when fish oil is simply added to atypical Western diet, plasma cholesterol levelsmayremain

unchanged

or actu-ally increase (16). Another important variable in these studies is the sourceandcomposition of the marine lipid. The crude fish oil used in manystudies contains substantial amountsof cholesterol and saturated fatty acids

which,

in

susceptible

indi-viduals, may counteract the cholesterol-lowering activity of the w-3polyunsaturated fattyacids.Thus, theresponse to di-etary fish oil in a given individual probably depends on a number of factors including the quantity andsource ofthe marinelipidingested,thedegreetowhichotherdietary fatsare replacedbythe fishoil,andtheunderlying lipid phenotypeof the individual. There is little doubt, however, that specific fatty acids present in the diet can markedly influence the concen-tration of LDL in the plasma. Since an elevated concenconcen-tration of LDL in theplasma is awell-established riskfactor forthe development of coronary heartdisease, it iscritically impor-tant tounderstandhow theselipids regulate circulatingLDL levels.

Theconcentration ofLDLin plasma is determinedby the rateatwhichLDLisintroduced intothe plasma relative to the rate at which LDL is removed from plasma by receptor-de-pendent and receptor-independenttransport processes in the various tissues ofthe body. In animals on a low-cholesterol, low-triglyceride diet, receptor-dependent transport accounts for approximately three-fourths of total LDL turnover (17-19). Receptor-dependent LDLtransport is saturable and canbedescribedin termsofamaximaltransport rate(*Jm)' and theconcentration of LDL inplasmanecessary toachieve one-half of this maximal transport rate(*Km)(20). Theliver is the mostimportantsiteofLDLcatabolismand accountsfor

>80%

of whole-body

receptor-dependent LDL uptake (20). Receptor-dependent LDL transport in the liver is regulable andit isnowapparentthat thetypeandquantity of fat inthe diet profoundly influenceshepatic LDL receptor activity. For example,inthe hamsterdietarycholesterol has been shown to suppressreceptor-dependentLDLuptake in the liver (21, 22). Responsivenesstodietarycholesterol varies widely, however, among different species and even within the same species (21-25). Dietary triglyceridesalso appear to regulate hepatic

1.Abbreviations used in this paper:*Jm,the maximal rate of receptor-dependentLDLuptakein aparticular tissueorthe wholebody;J,,the

totalrateofLDLuptake byaparticulartissueorthewhole body by receptor-dependent and receptor-independent pathways; *Km, the concentration of LDL-cholesterolinplasmanecessarytoachieve half of the maximal uptakerate; *P,theproportionalityconstantfor

re-ceptor-independent LDLuptake ina particular tissueorthewhole

body.

528 M. A. Ventura,L. A. Woollett, and D. KSpady

J.Clin.Invest.

©TheAmericanSocietyfor ClinicalInvestigation,Inc.

0021-9738/89/08/0528/10 $2.00

LDLreceptor

activity

although the mechanism(s) involvedare poorly understood. In hamsters fed a low-cholesterol diet, polyunsaturated vegetableoil modestly increased hepaticLDL receptor activity whereas saturated fatty acids had no effect (22).Whether or notdietarycholesterol or triglyceride affects receptor-dependent LDL transportintheextrahepatic tissues is not known. In contrast to receptor-dependent LDL trans-port,receptor-independentLDL transport, which accounts for about one-fourth of total LDL turnover, is widely distributed throughout all tissues of the body (19, 20). Receptor-indepen-dent transport is nonsaturable andcanbe definedbya

propor-tionality constant (*P). Although little is known about the regulation of receptor-independent transport, several experi-mental manipulations known to alter receptor-dependent LDL uptake in the liver have had no discernible effect on receptor-independent uptake (22).

There is generally little information on the effect of fish oil onthesethreebasic processes that determinecirculatingLDL levels. Thepresentstudiesweretherefore undertakento quan-titate and compare the effects of fish oil (rich in w-3 polyunsat-uratedfatty acids), safflower oil (rich in w-6 polyunsaturated fatty acids), and hydrogenated coconut oil (saturated fatty acidsonly)ontherateof LDLproductionand on the rates of receptor-dependent andreceptor-independentLDLuptake in all of the organsof the rat. Studies werealso carried out to determine the effects of dietary fish oil,safflower oil and co-conutoil on rates of cholesterolsynthesisinthevarioustissues ofthe body. These studiesprovidethe firstquantitativedata onthe mechanismsbywhich these three dietary triglycerides alterplasmaLDL-cholesterol concentrations in the rat.

Methods

Animals and diets. All studies were performedin female

Sprague-Dawleyratsthatwereobtained from SASCO Inc., Omaha, NE in the weight range of 125-150 g. The animalswerehoused inanisolation

roomwithlight cycling(dark from 3a.m. to3 p.m. andlightfrom 3 p.m.to3a.m.) and allowed freeaccess to waterand standard rodent

diet foratleast 2 wkbefore beginningtheexperimental diets. The controldiet used in these studies, ground Wayne Lab Blox (Allied Mills, Chicago, IL), contained only0.025% (wt/wt)cholesterol and

<5%(wt/wt)triglyceride.Thefattyacidcompositionof this diet, as

determined bygas-liquidchromatography of the methyl esters, was

16%ofthe fatty acidsas16:0,4% as18:0, 28%as18:1,44% as18:2, and 6%as18:3. Theexperimentaldietswereprepared bymixingthe

var-ioustriglyceridesinto the controldietusingacommercialfood mixer. The addition of thetriglyceridenecessarilydilutedouttheother

com-ponents of the control diet.Preliminarystudies showed that the addi-tionofupto20%glucose(wt/wt)tothe controldiet hadnoeffecton

receptor-dependent or receptor-independent LDLuptake in any organ. Nevertheless, from a strict nutritional standpoint, the most

propercomparisonsarebetween diets withequalamountsof added

triglyceride.

Thefish oilconcentrate usedin these studies (Sanomega)was pur-chasedfromNipponOil andFatsCompany, Ltd., Tokyo, Japan,and

contained 28.4%eicosapentaenoic acid, 13.7%docosahexaenoicacid,

andvirtuallynocholesterol (<0.05%).Thefish oil dietswerestoredin sealedcontainersundernitrogenat4°C.Alldietswereenriched with vitaminE(I g/kgofdiet). Preliminary studies showed that this amount ofvitaminEhadnoeffectonthe parameters of cholesterol and lipo-protein metabolism thatwere being studied. Animals were fed the

variousexperimental diets ad lib. and weight gain was monitored.All

studieswerecarriedoutduringthe mid-darkphaseofthelight cycle.

Determination oftissue LDL uptake rates in vivo. Plasma was

obtainedfromnormocholesterolemicratorhuman donors. The LDL fractionwasisolated in the density range of 1.020-1.055 g/ml, washed andconcentrated, and labeled with [125I]tyraminecellobiose (26) or

'"'I(27)aspreviouslydescribed (19). The human LDL was also

sub-jectedtoreductivemethylationtocompletely eliminate its recognition

bythe LDL receptor (28). TotalLDLtransport in the organsof the rat

wasmeasuredusingthe homologous LDL preparations while the

re-ceptor-independent component of total LDLtransport was deter-minedusingthe methylated human LDL preparations. In either case,

lipoproteinswere used within 48 hof preparation and were filtered

througha0.45-MuM filter(Millipore/ContinentalWaterSystems,

Bed-ford, MA)immediatelybeforeuse.Aspreviously described, rates of tissueLDLuptakeweredeterminedusingaprimed-continuous infu-sion of

['251]tyramine

cellobiose-labeledLDL(22, 29). The

radioactiv-itypresent in theprimingdoserelativetotheradioactivity infused each hourwasadjustedso as tomaintainaconstantspecificactivity of LDL inplasmaovertheexperimental period. The infusions of

['251I]tyra-mine cellobiose-labeled LDLwerecontinued for6 h at which time eachanimalwasadministeredabolusof

'3"I-labeled

LDLand killed 10 min laterbyexsanguinationthrough the abdominal aorta. Samples of various organswerequicklyrinsed and weighed, and along with ali-quots ofplasmawereassayed forradioactivityin agamma counter

(PackardInstrumentCo., Inc.,DownersGrove, IL).The organs

sam-pledincluded theliver,smallintestine, spleen, kidney,adrenalgland, heart, lung,adiposetissue,andskeletal muscle. The entireremaining

carcass wasthenhomogenizedandaliquotswereassayedfor

radioac-tivity.Sincenotissuewas_discarded,ratesof LDLuptakein individual

organs,aswellasthe wholeanimal,couldbe determined. Theamount

of labeledLDLin each organat10min

(I'II

disintegrationsper minute per gram of tissue dividedbythespecific activityof

'"'I

in theplasma)

andat6h(1251I disintegrationsper minute per gram oftissuedividedby

thespecific activityof1251Iin theplasma)wasthencalculated andhas the units ofmicrogramsof LDL-cholesterol per gram of tissue. The increase in the tissuecontentof LDL-cholesterol with time represents therateofLDLuptakeinmicrogramsof LDL-cholesterol taken up per hour per gram of tissue

(pg/h

perg)or per whole organ

(ag/h

per

organ).Insomestudieshepaticasialofetuin transportwasdetermined

using

'3'I-

and

[1'251]tyramine

cellobiose-labeled asialofetuin.

Inthesestudies totalLDLuptakewasmeasured inthe experimen-tal animalsusingLDLderived from donorratsfed standard rodent diet.Inpreliminary studies, however,wefound that LDLpreparations

obtained from animals fed fishoil,safflower oil,or coconutoilwere

transportedatsimilarrateswheninfusedinto controlanimals(Spady,

D.K.,unpublished observation).

Determinationofcholesterolsynthesisrates in vivo.Aspreviously

described (30, 31), the animals wereadministered - 50 mCi of

[3H]H20

i.v. The animalswerethenreturnedtoindividualcages under

afume hood. 1 h after the injection of[3H]H20 theanimalswere

anesthetized andexsanguinated throughtheabdominalaorta.Aliquots

ofplasmawere taken for the determination ofbodywaterspecific activity,andsamplesof the various organsweretakenfor the isolation ofdigitonin-precipitablesterols andfattyacids. In thecaseofthe small intestine and theremainingcarcass, the entire tissuewassaponified

andbroughtto anexactvolume. Ratesof sterolsynthesisareexpressed

asthemicromoles of

[3H]H20

incorporatedintodigitonin-precipitable

sterols per hour per gram of tissue(,umol/hperg)orper whole organ

(Amol/hperorgan).

Cholesterol, triglyceride, andfattyacidcontentinplasma, tissues, anddiets.Totalplasmacholesterolandtriglyceridelevelswere

deter-minedusing enzymatickitsobtainedfrom Boehringer Mannheim

Bio-chemicals, Indianapolis,IN.Thedistribution of cholesterol in plasma

wasdetermined by simultaneously centrifuging plasma at densities of

aspreviously described(32).Thefattyacidprofileofplasma,tissues,

and diets was determined afterextracting total lipids intochloroform/ methanol(2:1)(33).Thephospholipids, triglycerides,andcholesteryl

esterswereseparatedbyTLC. Eachof theselipidfractionswasmethyl

esterified and thefatty acid profile determinedby capillarygas

chroma-tography.

Calculations. Plasma LDL-cholesterol levels variedbymorethan

three-foldinanimalsingestingthe variousexperimentaldiets.Thus,to

relate the changes in LDL-cholesterol uptakeinthe organs of these animals tochanges in LDL receptoractivity,itwasnecessaryto

super-impose theexperimentally determined uptake rates on the kinetic curvesdefiningratesof LDLuptakein the various tissues ofcontrol animalsas afunction ofplasmaLDL-cholesterol concentrations(20,

22, 29). Total LDL-cholesterol uptake by aparticulartissueorthe wholebodyis equal to theratesof LDL-cholesteroluptakeby

recep-tor-dependentandreceptor-independent pathways. Therelationship

between total LDL-cholesterol uptake (J,) andplasma LDL-choles-terolconcentrations(C,)in normal animalscanbe describedbythe

equation:

J,

=(*J=C,)/(*Km+C,)+*PC,,where*Km and *Jf equal

the apparent Michaelisconstantand maximal transportvelocity,

re-spectively, for thereceptor-dependentprocess, and*Pequalsthe ap-parent uptake constant forreceptor-independent LDLuptake (20).

Thekinetic curves for normal LDL transport used in these studieswere

constructedusingpreviously publishedvaluesfor*J",*Km, and*P

obtained in controlrats(20).Usingthese curves, theregulationof LDL receptoractivity by the various experimental diets could be deter-minedbycomparingratesofreceptor-dependentLDL transport in the

experimentalanimalsto ratesofreceptor-dependent LDLtransport that would beseenin control animalsatthesameplasma LDL-cho-lesterolconcentration.ValuesforLDL receptoractivityarepresented

aspercentagesoftheappropriatecontrol values.

The dataare presentedas meanvalues±l SEM.In most

experi-ments, three differenttriglycerideswerefedatthreedifferent levels(5,

10, or 20%).Statisticalanalysisof the effects of thesetwofactorswas

carried outusingtwo-way analysis ofvariance. Whena significant triglycerideordoseeffectwasshown,individual contrasts were per-formed using Student's t testsatthe P<0.01 level.

Results

In

preliminary experiments

we noted that

dietary

fish oil markedly lowered plasma cholesterol and

triglyceride

levelsin the rat. As illustrated in Table

I,

total

plasma

cholesterol equaled 62 mg/dl in control animals andwasreduced

by

23,

31,and 29% when fish oilwas

fed

atthe 5,

10,

and20%levels, respectively. Most ofthereduction in total

plasma

cholesterol occurredinthe d<1.095

g/ml density

fractions. For

example,

when fedatthe20% level, fish oil markedly lowered choles-terolinthe d< 1.020

g/ml

(60% reduction),

d= 1.020-1.063 g/ml(36% reduction), and

d

= 1.063-1.095

g/ml (42%

reduc-tion) fractionsbuthadrelativelylittle effectonthed> 1.095 g/ml (18%

reduction)

density fraction.

Intherat,cholesterol in the

d

= 1.063-1.095

g/ml

fraction iscarried

predominantly

in apoprotein

E-containing

HDL,

particles. Thus, dietary

fishoil

primarily

lowered those

lipoproteins which,

by

virtue of

their

apoprotein

B and/or

apoprotein

E content, areclearedfrom plasma by theLDL receptorpathway.

In contrast tothefish

oil, safflower oil

hadrelatively little effecton

plasma

cholesterollevelswhile

hydrogenated

coconut oil raised cholesterol in the d = 1.020-1.063 g/ml and

d

= 1.063-1.095 g/ml fractions and,to alesserextent,in the

d

> 1.095 g/ml fraction.Asalso shownin TableI,dietaryfish oil

TableI. PlasmaCholesterol and TriglycerideConcentrations inAnimalsFed Fish Oil, Safflower Oil, or Hydrogenated Coconut Oil

Plasma cholesterolconcentration Density fraction

Plasma

g/ml triglyceride

Diet Total d<1.020 d= 1.020-1.063 d = 1.063-1.095 d > 1.095 concentration

mg/dl

Control 62±3 5±1 11±1 12±1 34±2 64±3 Fish oil

5% 48±5 3±0.5 8±1 8±2 29±2 39±2 10% 43±4 2±0.3 6±1 7±1 28±2 31±2 20% 44±3 2±0.3 7±1 7±1 28±1 27±3

Saffloweroil

5% 62±4 6±1 11±2 13±1 32±2 60±3 10% 58±3 5±1 10±1 10±2 33±3 61±2 20% 58±5 4±1 10±2 10±1 34±2 56±3 Coconut oil

5% 66±4 5±1 12±1 13±2 36±3 66±6 10% 76±6 5±1 15±1 18±2 38±4 59±3 20% 79±5 4±1 17±2 19±2 39±2 60±4 Groups of animals were fed diets enriched with varying amounts of fish oil,saffloweroil, orhydrogenated coconut oil for 1 mo. Each value represents themean±1 SEM for data obtained in six animals. An analysis of variance showed that plasma cholesterol (total and various density fractions) and triglyceride were significantly affected by the type of triglyceride present in the diet. Individual contrasts showed that total choles-teroland cholesterol carried in thed= 1.020-1.063 andd= 1.063-1.095 g/ml density fractionswere lower with dietary fish oil than with di-etary safflower oil and lower withdietary safflower oil than with dietary coconut oil (P < 0.01). Plasma triglyceride and cholesterol carried in the d<1.020 andd> 1.095g/ml fractionswerelowerin animals fed fish oil than in animals fedsaffloweroil or coconut oil (P < 0.01).

Table II. Total and Receptor-independent LDL-Cholesterol Uptake Rates and Plasma LDL-Cholesterol Concentrations in Animals Fed Fish Oil,

Safflower

Oil,orHydrogenatedCoconutOil

Receptor-independent

LDL-TotalLDL-cholesteroluptake cholesterol uptake G. Plasma

A.Final LDL-cholesterol

Diet body weight B. Liverweight C. Hepatic D.Extrahepatic E.Hepatic F. Extrahepatic concentration g g/100g bodywi ug/hper100 g body wt mg/dl Control 218±13 3.2±0.2 48±4 30±3 4.6±0.5 19.6±2 13±1

Fish oil

5% 220±18 3.4±0.1 56±4 22±2 3.2±0.2 13.6±1 9±2 10% 229±11 3.7±0.3 58±3 20±1 2.9±0.2 12.4±1 8±1

20% 225±14 4.1±0.2 61±4 20±3 2.9±0.2 12.8±2 8±1 Saffloweroil

5% 222±12 3.2±0.3 58±5 28±2 4.2±0.2 18.1±2 12±2

10% 230±16 3.6±0.1 59±4 30±2 4.8±0.4 19.8±1 13±1

20% 234±16 3.7±0.2 61±5 29±3 4.3±0.3 18.4±1 12±1

Coconutoil

5% 218±10 3.3±0.2 55±5 34±3 5.3±0.4 22.7±2 15±2 10% 222±12 3.4±0.2 68±3 41±3 6.4±0.3 27.0±2 18±1

20% 222±18 3.6±0.3 71±6 43±4 6.6±0.3 28.2±1 19±2

Groupsof animals werefed dietsenrichedwithvarying amounts offishoil,saffloweroil, orhydrogenated coconut oil for1mo. Each value represents the mean±1SEM lor data obtained in six animals.

markedlyreduced plasma triglyceride levels whereas

safflower

oil and hydrogenatedcoconutoil had

relatively

little

effect.

Studies

werenextundertakentoexamine the effect offish

oil, safflower oil,

and

hydrogenated

coconutoil onthose pa-rametersthat determine the concentration ofLDL inplasma,

i.e.,

therate

of

LDL

production

and the ratesof

receptor-de-pendent

and

receptor-independent

LDL

uptake

by

thevarious

tissues

of the

body.

The results of these studiesareshownin Table II. In these

experiments,

the three

triglycerides

were

fed

atthe 5,

10,

or20% level

(wt/wt)

for 1 mo.

Weight gain

was

similaramonganimals fed the variousexperimentaldiets

(col-umnA). Liver

size, however,

tendedtoincreasewhen

increas-ing

amounts

of

triglyceride

wereaddedtothediet.

Thus,

when

fed

atthe 20% level, fish

oil, safflower oil,

and

hydrogenated

coconut oil increased liversize

by 28, 16,

and

13%,

respec-tively (column B).

ColumnsC and D show the absoluterates

of total

(receptor-dependent plus receptor-independent)

LDL-cholesterol

uptake

in the liver and

extrahepatic

tissues of ani-mals

fed

the three

triglycerides.

All

uptake

rates are

normal-ized

to a

body weight of

100g.Total LDL-cholesterol

uptake

in the

liver equaled

48±4

,ug/h

per

100-g animal

in control animals and

increased

modestly

when

increasing

amountsof fish

oil, safflower oil,

or coconut oil were added to the diet

(column

C).

Rates of LDL-cholesterol

uptake

in individual

extrahepatic

tissues were

generally

quite

low.

Therefore,

up-takeratesin allof the

extrahepatic

tissueswere

combined

and thesevaluesare

presented

incolumn D. In control

animals,

LDL-cholesterol uptake in the

extrahepatic

tissues

equaled

30±3

_gg/h

per

100-g

animal.

Extrahepatic

LDL-cholesterol uptake decreased in animals fed fish

oil,

remained

unchanged

inanimals fed safflower

oil,

andincreased in animalsfed

co-conutoil.Itshouldbe noted thattotal LDL-cholesterol

uptake

in the liver plus the

extrahepatic

tissues equals whole-body LDL-cholesterol uptake

which,

in asteady state, must

equal

therateof LDL-cholesterol production. The receptor-indepen-dent component

of

total LDL-cholesterol uptake in the

he-patic

and

extrahepatic tissues is

shown in columnsEandF. In control

animals, receptor-independent

LDL-cholesterol up-takeequaled

4.6±0.5

and 19.6±2

,ug/h

per100-ganimal in the

liver

and

extrahepatic tissues, respectively.

Inboth the

liver

and

extrahepatic tissues, receptor-independent

LDL-choles-terol

uptake

decreased in

animals

fed fish oil,

remained un-changed in

animals fed safflower

oil,

andincreased in

animals

fed

coconut

oil.

Asshown in column

G,

the

plasma

LDL-cho-lesterol

concentration

equaled 13 mg/dl in control

animals.

In animals

fed

20%

triglyceride, plasma

LDL-cholesterol levels decreased 38% with fish

oil,

remained

virtually unchanged

withsafflower oil, and increased 46% withcoconutoil. From Table

II,

rates

of

receptor-dependent LDL

uptake

can be

obtained by subtracting

the rateof

receptor-indepen-dent LDL

uptake

from the rateof total

(receptor-dependent

plus

receptor-independent)

LDL uptake. However,

since

re-ceptor-dependent

LDL transport

is

saturable and

since plasma

LDLlevels varied

widely

in

animals fed

thethree

triglycerides,

changes

in therateofreceptor

dependent

LDL

uptake

can not

be

equated

directly

with changes

inLDL receptor

activity

(20).

Inordertorelate the

changes

in absoluterates

of

LDL-choles-terol

uptake

to

changes

inreceptor

activity,

each of the values shown in TableII was

superimposed

onthe

kinetic

curvesthat define the

relationship

between LDL-cholesterol

uptake

and

repre-250

Lb

1

0<

_E

-4

CM 2

luw

o E

~I8

0

5 10 15 20 25 30

PLASMA LDL-CHOLESTEROL (mg/di)

Figure 1. LDL-cholesteroluptakeinthe whole liver(top)and allof

the extrahepatictissuescombined(bottom)in animals fed 20% fish

oil,saffloweroil,orhydrogenatedcoconutoil for1mo.Theshaded

areasrepresentthe kineticcurvesfor total(stippled)and receptor-in-dependent (hatched)LDL-cholesteroluptakein control animals fed

standard rodent diet. Thesecurves werecalculatedasdescribed in

Methods. Theindividualpoints superimposedonthese kinetic

curvesshow theratesoftotal(opensymbols)and

receptor-indepen-dent(solid symbols)LDL-cholesterol uptakein theexperimental

ani-mals plottedas afunction of theplasmaLDL-cholesterol levelsin

thesameanimals. Eachpointrepresentsthemean±1 SEM for data

obtained in six animals.

sentsthe receptor-dependentcomponent of total LDLuptake. Superimposedonthesenormal kineticcurves aretheobserved ratesof total (open

symbols)

and

receptor-independent (closed

symbols) LDL uptakein animals fed the 20%triglyceridediets.

As shown in thetoppanel, the control animals usedinthese studies manifestedratesof total (0)andreceptor-independent (@) LDL-cholesterol uptakein the liver of 48±4 and 4.6±0.5

pg/hper l00-g animal, respectively,ata meanplasma

LDL-cholesterol concentration of 13 mg/dl. Since these values fell

onthenormal kineticcurvesfor LDLtransport inthe liver, hepatic LDLreceptor activity in the control animalswas

as-signedavalue of 100%.In contrast, totalhepatic

LDL-choles-terol uptakein animalsfed20% fish oil(A) equaled61±4 pg/h

ataplasma LDL-cholesterolconcentration of 8mg/dlwhereas

normal animals would be expected to transport only - 30

Ag/h

atthisLDL-cholesterolconcentration. Since

receptor-in-dependent uptake (-) was normal in these animals, the

in-creasein totalLDL-cholesterol uptakecould be attributed

en-tirelytoanincreaseinreceptor-dependentLDL transport.

In-deed, it could be calculated that hepatic receptor-dependent LDL-cholesterol uptake in animals fed 20% fish oilwas209%

ofthatinnormal animalsatthesameplasma LDL-cholesterol concentration.Similarly,therateof total LDL-cholesterol

up-Z->

2

()8

a-Jo

-i.)p

200

150

100

8

W

wr jI-cc)

150

100

0 5 10 15

DIETARY TRIGLYCERIDE(%) 20

Figure 2.LDL-receptoractivity in the whole liver(top)and all of the extrahepatic tissues combined (bottom)in animals fedvarying

amountsof fish oil, safflower oil,orhydrogenatedcoconutoilfor I

mo.The values shownrepresenttheratesofreceptor-dependent LDL-cholesterol uptake in experimentalanimalsaspercentagesof theratesof receptor-dependent LDL-cholesterol uptake that would

occurincontrol animalsatthesameLDL-cholesterol concentration (as illustrated in Fig. 1). Each pointrepresentsthemean±1SEM for

data obtained in six animals.Two-way analysisofvariance showed

that thetypeoftriglycerideandtheamountof triglyceride fed af-fected hepatic LDLreceptoractivity. Individualcontrastsofthe tri-glyceride effect showed that hepatic LDLreceptoractivitywashigher

inanimalsfed fishoil thaninanimals fed safflower oil and higher in

animals fed safflower oil than in animals fedcoconutoil(P<0.01).

take in the liverwassignificantlyincreased (relativetonormal animalsatthesameLDL-cholesterollevel) in animals fed 20% safflower oil. Again, receptor-independent LDL-cholesterol uptakewas normal and it could be calculated that

receptor-dependent LDL-cholesterol transport was 138% of that in

control animalsatthesameplasmaLDL-cholesterol

concen-tration. In animals fed hydrogenated coconut oil, total and receptor-independent uptake rateswere notdisplaced signifi-cantly from the normal kineticcurves,indicatingthat neither

transportprocess wasaffected by this dietary manipulation. As shown in the bottom panel,ratesoftotal and receptor-inde-pendent LDL-cholesterol uptake in the extrahepatic tissues

were not displaced significantly from the normal kinetic

curvesbyanyof the experimental diets.

From thetypeof analysis illustrated in Fig. 1, the changes in absolute rates of LDL uptake shown in Table II could be convertedtochanges in LDLreceptoractivity (definedasthe

rateof receptor-dependent LDL uptake in experimental ani-mals relativetotherateof receptor-dependent uptake in

con-trol animalsatthesame_{LDL-cholesterol concentration) and}

these valuesarepresentedinFig. 2. Asseenin thetoppanel, dietary fish oil stimulated hepatic LDLreceptoractivity by 75, 95, and 105% when fed atthe 5, 10, and 20% level,

respec-532 M. A. Ventura,L. A. Woollett,and D. K._Spady M

O *

Conrol

i0 A otas Upta

13 Se;lSbwoi

10

20

~Receptor-lndependervt

Uptdwe

M~~~~~~~~~~~~~~uIIIIIlIhlIIu

.,lIIIlIIIIII,I

I Fsh Oil * _SafflowerOI

r

CooonutOil

a a

I:-.

W₂

Z c

w8 125

00

0..

<

100

UJ CC<4

,, X 75

wr _I

z

-1OZ 125

IZUJ 100

<0L

VDt

75

LUa

z z20

0

0-0. 150

0.0

t~w

I.w

o 100

0 0

-j I

0 5 10 15 20

DIETARY TRIGLYCERIDE(%)

Figure 3.Receptor-independent LDL-cholesterol uptakeinthe whole liver (top) and all of the extrahepatic tissues combined (bot-tom)inanimals fedvaryingamountsof fishoil,saffloweroil,or hy-drogenatedcoconutoilfor 1mo.The values shownrepresentthe

ratesof receptor-independent LDL-cholesterol uptakein

experimen-tal animalsaspercentagesof theratesofreceptorindependent

up-take that wouldoccurin control animalsatthesame LDL-choles-terol concentration(as illustratedinFig. 1).Eachpointrepresents

themean±I SEM for data obtainedinsixanimals.Two-way analysis

of variance showed that neither thetypeoftriglyceridenorthe

amountoftriglyceridefedsignificantlyalteredreceptor-independent

LDLuptakeintheliverorextrahepatictissues.

tively. Safflower oil also stimulated hepatic LDL receptor

ac-tivity but to a much smaller degree whilehydrogenated

co-conut oil had no significant effect on hepatic LDL receptor

activity. As shown in the bottom panel, none ofthe experi-mental diets had a major effect on receptor activity in the extrahepatic tissues. Fig.3shows the effectofthe threedietary triglycerides on receptor-independent LDL transport in the

liver and extrahepatic tissues. The values showninthisfigure

were determined by superimposing the absolute rates of

re-ceptor-independent LDL-cholesterol uptakeshown in Table II

onthe normal kineticcurvesforreceptor-independent trans-port(as illustratedinFig. 1). Receptor-independenttransport in the experimental animalswasthenexpressed as a

percent-age of the rate ofreceptor independent transportthatwould

occurin controlanimalsatthesameLDL-cholesterol

concen-tration. As isapparent, the three dietary lipids tested had no

significant effect on receptor-independent LDL transport in

the liverorin the extrahepatic tissues.

The effect of the three dietary triglycerideson

LDL-choles-terolproductionis shown in Fig.4.Rates of LDL-cholesterol

production increased in a dose-dependent fashionwhen

co-conutoilwasaddedtothe diet.Althoughnotstatistically sig-nificantatthe P<0.01 level, dietary safflower oilalso tended

to increase the rate ofLDL-cholesterol production. In

con-trast, theaddition of fishoiltothe diet hadnoeffecton

LDL-cholesterol production. It should benoted, however, that di-etary fish oilactuallyreduced therateofLDL-cholesterol

pro-0 5 10 15 20 DIETARY TRIGLYCERIDE (%)

Figure 4. Ratesof LDL-cholesterol production in animals fed vary-ing amounts of fish oil, safflower oil, or hydrogenated coconut oil for

I mo. Eachvalue represents the mean±I SEM for data obtained in sixanimals. Two-way analysis of variance showed that both the type and the amountoftriglyceridefed significantly affected the rate of LDL-cholesterol production. Individual contrasts of the triglyceride effect showed that the rate of LDL-cholesterol production was signif-icantlyhigher in animals fed coconut oil than in animals fed fish or safflower oil (P<0.01).

duction when compared with equal amounts of the other dietary triglycerides.

To investigate the mechanism bywhich fish oil regulates hepatic LDL receptoractivity,theeffect of fish oil on hepatic asialofetuin transport wasmeasured.Asialofetuin isaplasma glycoprotein rapidly cleared by the liver via a receptor pathway similar to the LDL receptor pathway (34, 35). As shown in Fig. 5, hepatic asialofetuin transport was notsignificantly altered byany of the threedietarylipidsused inthese studies.

Thefinalgroupof experimentswasundertakento seeifthe changesinhepaticLDL receptoractivitydescribedabovewere accompanied by alterations inotheraspectsof cholesterol me-tabolisminthe liver. Fig. 6 shows the effect of the three dietary triglycerides on free and esterified cholesterol levels and de novocholesterol synthesis rates inthe liver. As shownin the toppanel, none of the threedietarytriglycerides significantly altered the free cholesterol content of the liver. In contrast, hepatic cholesteryl esters,which equaled 0.2 mg/g in control animals, increasedby 2.5-, 6.5-, and2.5-fold in animalsfed fish oil, safflower oil, or coconut oil, respectively (middle panel). As showninthe bottompanel,ratesofhepatic choles-terol synthesis, when expressed per gram ofliver, were

sup-,, 150 _{Figure5.Whole-liver}

RFHOR asialofetuinuptakein D_125aO*Saffow

_CofooOil

cml

_animals

_fed

_varying

amountsof fish

oil,

saf-T

X00 .izzzi

flower

oil,

or

hydroge-y

,^ I natedcoconutoilfor 1

- mo.Each value

repre-!Z 75 _{.sentsthemean±1 SEM}

,

_,.1

: for dataobtained in six

0 5 10 15 20 animals. Two-way

anal-ysisof variance revealed

no

significant

effectof

thetype oftriglycerideortheamountoftriglyceridefedonhepatic

asialofetuin uptake.

OControl

* Fish Oil

* Safflower011

O Coconut Oil

f

. _.- a

,-.~S2~tLI

I

r.--I

*FishOil

MSaffowerOi

OCooonutO

t

?

L

.

l].'11"

0

3

9 2 U-IE

IICo

-F-30

E

-J

0

I-cc

S

CL

0

X

0

-i

C,)

05

0.5 1.0 1.5 2.0 2

WHOLE-BODYLDL RECEPTOR ACTIVITY (multiples of normal)

77

§O-

OM-rbO

SOMw

Figure6.Hepatic unesterifiedandesterified cholesterol levels and

hepatic cholesterol synthesisratesinanimals fed20% fishoil, saf-floweroil,orhydrogenatedcoconutoilfor 1 mo.All valuesare

ex-pressedper gramofwetweightliverandrepresentthemean±1SEM

for data obtainedin sixanimals.

pressed by 28%inanimalsfedfishoil andby46% in animals

fed saffloweroil butwereunaffectedbycoconutoil.Although the dataarenotshown,cholesterolsynthesisratesand freeand

esterified cholesterol levelsin theextrahepatic tissueswerenot

affectedby the three dietary lipids.

Discussion

These studiesprovide the first quantitative description of the mechanismsby which dietary polyunsaturated (w-3 and w-6) and saturated fatty acids regulate plasma LDL levels in the whole animal. The concentration ofLDLinplasma is deter-mined by therate atwhich LDL is produced relativetotherate atwhich LDLisremoved fromplasma by receptor-dependent and receptor-independent transport processes located in the

various tissues of the body. Inthepresentinvestigations,none

of thetriglycerides studied significantly altered receptor-inde-pendent LDL transport in any tissue. Thus, the changesin

plasma LDLlevels induced by triglyceride feeding weredue

entirelytochangesintherateofLDLproductionorchangesin

whole-body LDLreceptoractivity. These relationshipsare

il-lustrated in Fig. 7, where plasma LDL-cholesterol levelsare

plottedasafunction of whole-bodyreceptoractivity and LDL productionrates. Thesolidcurvesshowthe steady-state

LDL-Figure 7. Plasma LDL-cholesterol concentrationsas afunctionof

whole-bodyLDLreceptoractivityand LDL-cholesterolproduction

ratesinanimals fed 20% fishoil,saffloweroil,orhydrogenated

co-conutoil for 1mo.The solidcurvesshow theplasma

LDL-choles-terol concentrations thatwilloccurincontrol animalsfed standard

rodent diet under circumstances whereLDLreceptoractivityis

var-ied from half normaltotwoand one-half timesnormal at

LDL-cho-lesterolproductionratesofhalfnormal, normal,ortwicenormal.

Theindividual datapointsrepresentthemean±1 SEM fordata

ob-tainedinsixanimals.

cholesterol concentrations thatwilloccurunder circumstances where whole-body LDL receptor activity is varied from half normalto twoandone-half times normalatLDLproduction

ratesof halfnormal, normal, ortwice normal. These kinetic curvesfor normal LDLtransport intheratwereconstructed usingpreviously published values for thetransport parameters

thatdescribereceptor-dependent (*Jm, *Km) and receptor-in-dependent(*P)LDL-cholesteroluptake in control animals fed

a standard low-cholesterol, low-triglyceride rodent diet. Su-perimposed on thesecurves are the data foranimals fed the

three triglycerides at the 20% level. The control animals in these studies (0), witha normal productionrateand normal

receptoractivityhadaplasma LDL-cholesterol concentration

of 13mg/dl. When fish oilwasaddedtothe diet, the produc-tion rate remained essentially unchanged while whole-body LDLreceptoractivity increased approximately twofold

caus-ing LDL-cholesterol levelstofall 38%to8 mg/dl. Whole-body LDLreceptoractivity also increased (- 40%) in animals fed saffloweroil. Intheseanimals, however, the enhancedreceptor

activity waslargely offset byan increase intherate of LDL-cholesterol productionsothat plasma LDL-cholesterol levels

remained essentially unchanged. Finally, hydrogenated

co-conut oil significantly increased the rate ofLDL-cholesterol production while having no effect on LDL receptor activity andasaresultplasma LDL-cholesterol levels rose - 50% to

19 mg/dl. Thus, when compared with equal amountsof saf-floweroilorcoconutoil, dietary fish oil reduced plasma

LDL-cholesterol levelsbothby reducing therateof LDL-cholesterol production and by increasing the rateofreceptor dependent

LDLtransport.

In animals fed fish oil or safflower oil, the increase in

whole-bodyLDLreceptoractivitywasdue almostentirelyto

anincrease inreceptor activity in the liver.Incontrast,these dietary lipids hadno major effecton receptoractivity in the extrahepatic tissues, which included the small intestine,

534 M. A. Ventura, L. A. Woollett,and D. K.Spady

3

2

a

UJ

-cin

ECE 0_0-I

CLI

I

spleen, kidney, adrenal gland, heart, lung, adipose tissue, and skeletal muscleaswellas the entire residual carcass. Whyfish oil and

safflower

oil alteredreceptoractivity only in the liver is notclear. Itshould be noted,however, that the liver accounts for - 80% of total-bodyLDLreceptoractivityin the rat, as it

does in all other species that have been examined(20,36). Although dietary fish oil clearly enhanced hepatic LDL receptor activity in these studies, the precise mechanism by which it didso is not known. In cultured cells aswell as in whole organs in vivo, denovocholesterolsynthesisand recep-tor-dependentLDLuptakearegenerally regulated by changes incholesterol

homeostasis

within the cellor organ(22, 23, 29,

37, 38).

Thus, when faced with an alteration in cholesterol balance, the liver

responds

by appropriately adjusting therate

of

denovocholesterol

synthesis

and, ifnecessary,by also

regu-lating

therate

of

receptor-dependent LDL uptake. These com-pensatory

mechanisms

allow the livertomaintain cholesterol balance

despite

the presence of significant and variable amountsof cholesterol in the diet. Itis somewhat surprising thatdietary

fatty acids

alsoregulate receptor-dependentLDL transportin theliver. This effect of

fatty

acids, however,may

still

bemediated

by changes

incholesterol homeostasis within the

liver.

Thus, dietary fatty acidsmayregulate the LDL re-ceptor pathway

indirectly

by altering cholesterol balance acrossthe liverorby

altering

the

compartmentation

of

choles-terol

within the liver. Along these lines, dietary fish oil has been

reported

to decrease cholesterol

absorption

from the small

intestine

(5,

39)

andtoenhance the conversion of cho-lesterol tobile salts in the

liver

(5). Both

of

these

effects,

by

decreasing cholesterol availability

in the

liver,

would be

ex-pected

to

trigger

acompensatory rise in cholesterol

synthesis

and,perhaps,anincrease inreceptor-dependent LDL-choles-terol uptake. Inthepresent

studies, however, hepatic

choles-terol

synthesis

wasactually

suppressed

and

hepatic cholesteryl

esters

increased

in animals

fed

fish oilorsafflower

oil,

suggest-ing that the

availability of

cholesterol inthe liverwasincreased rather than decreased.In

addition,

therathasan

exceptionally

high

basalrate

of cholesterol

synthesis

in the

liver which

canbe

regulated

over an

extremely wide

rangeinresponse to

changes

in cholesterol

influx

or

efflux

(23, 30).

Asaconsequence, the

liver

can

usually

compensate

entirely for changes

in choles-terol balance

by

adjusting

therateof denovocholesterol syn-thesis. For example,

experimental manipulations

that decrease cholesterol

availability

in the liver suchas

feeding

cholestyr-amine,

which increases bile acid

synthesis,

or

feeding

sur-former,

which

specifically

blocks cholesterol absorption, are

fully compensated

for

by

enhancedcholesterol

synthesis

while LDL receptor

activity

remains unchanged (23). Thus, it is dif-ficultto

attribute

theincrease in

hepatic

LDL receptor

activity

seenin animals

fed fish

oilorsafflower oiltoagrosschange in

cholesterol

balance acrossthe

liver.

Itmaybe,

however,

that fishoil alters thedistribution of cholesterol within the hepato-cyte so as to shift cholesterol out ofa critical pool(s) that regulates theLDL receptorpathway.

Alternatively,

fish oil could affectthe LDL receptor

path-waythroughamechanismcompletelyunrelated tocholesterol. Forexample,

hepatic

membranes becomemarkedly enriched in w-3 polyunsaturated

fatty

acids

during

fish oil

feeding.

In-deed,inanimals fed fishoilatthe20%

level, nearly

25%ofthe fatty acidspresentin liver

phospholipids

werew-3

polyunsatu-rated fatty acids (Spady, D. K., unpublished

observation).

Thus, fish oil could affect the LDL receptor pathway by alter-ing the composition and, perhaps, the physical properties of hepatic membranes. Dietary fish oil had little effect, however, on the hepatic transport of asialofetuin, a glycoprotein taken up and degraded inthe liver through a receptor pathway simi-lar to that of the LDL receptor pathway (34, 35). Furthermore, fish oil feeding had no detectable effect on receptor-indepen-dent LDL transport as might be anticipated if the effect of the w-3 polyunsaturated fatty acids was simply to induce a nonspe-cific change in the physical properties of hepatic membranes.

How dietary triglycerides regulate the rate of LDL-choles-terol production is also unknown. LDL are produced in the plasma during the catabolism of VLDL. VLDL, in turn, are triglyceride-richparticlessynthesized and secreted by the liver. Several lines ofevidence now suggest that, relative to other fatty acids, w-3 polyunsaturated fattyacids reduce VLDL se-cretion by the liver (40-43). Since LDL are derived from VLDL, such a decrease in the rate of VLDLproductioncould explain the decrease in LDL production seen in the present studies when fish oil was substituted for

safflower

or coconut oil in the diet.

These studies are the first to demonstrate an effect of fish oil on receptor-dependentLDL transport in the liver. In con-trast, previous studies dealing with the mechanism by which fish oil alters plasma LDL levels have generally suggested that fish oil primarily affects the production of LDL rather than LDL receptor activity. For example, Illingworth et al. (15) reported that the 20% decrease in plasma LDL-cholesterol levels seen in seven normal subjects when they were switched from a control diet toone enriched with w-3 polyunsaturated fatty acids was due to areductionin the LDLproduction rate while the fractional catabolic rate for LDL was unchanged. In these studies, however, thetriglyceride content of the control dietequaled 30-40% of total calories and was composed pre-dominantly of saturated andmonounsaturatedfatty acids. In thepresent studies, the addition to the control diet of hydroge-nated coconut oil or, to a lesser extent, safflower oil increased the rate ofLDL-cholesterol production while fish oil had no effect.Thus,relative to the othertriglyceride-richdiets, fish oil lowered LDL levels both by increasing hepatic LDL receptor activity and by reducing the rate of LDL production. How-ever, the decrease in LDL production rates when safflower or coconut oil was replaced by fish oil was due not to the addition ofthe fish oilbut rather to the removal of the vegetable oil. More recently, Roach et al. (44) reported that plasma LDL-cholesterol levels fell by 34% in the rat when standard rodent chow wassupplementedwith fish oil at the 8% level. However, this decrease in plasma LDL levels was associated- with a seeminglyparadoxical37%reduction in the binding of LDL to solubilized hepaticmembranes.It is difficult to reconcile these findings with the present studies. It is possible, however, that changes in LDLbindingtoliver membranes in vitro may not always reflect changes in receptor-dependent LDL uptake in

vivo.

neutralize thepotentiallipid lowering effects of the w-3

poly-unsaturated

fatty

acids.Infuture studies it

will

be

important

to

compare the

cholesterol-lowering

activities of

highly

purified

preparations of

eicosapentaenoic

acid and docosahexaenoic

acid

andto determine how these

fatty

acids interact with di-etary cholesterol and saturated

fatty

acids in

regulating

LDL

production

and

receptor-dependent

LDLtransport.

Acknowledoments

Theauthors wishtothank Janet Hainline and Russell Nelson for their excellent technical assistance andImogeneRobison for her assistance inpreparingthemanuscript.

Thisworkwassupported byU.S. Public Health Service research grants HL-38049 and AM-01221 andagrant-in-aid(870653)from the American Heart Association.Duringthe time that these studieswere

carried out Laura A. Woollett was also supported by U. S. Public Health Service traininggrant AM-07 100 and Mark A. Ventura

re-ceived support from an American GastroenterologicalAssociation Student ResearchFellowship.

References

1.Kromann, N., andA. Green. 1980.Epidemiologicalstudies in theUpernavikDistrict,Greenland. Acta Med.Scand.208:401-406.

2. Kromhout, D., E. B.Bosschieter,and C. de Lezenne Coulander. 1985. The inverse relation between fish consumption and 20-year mortality from coronary heart disease. N. Engl.J. Med.

312:1205-1209.

3.Dyerberg, J., and M.0. Bang. 1979.Haemostatic function and plateletpolyunsaturatedfattyacids in Eskimos. Lancet. 2:433-435.

4.Bang, H. O., and J.Dyerberg. 1972. Plasmalipidsand

lipopro-teins inGreenlandicWest Coast Eskimos.ActaMed.Scand. 192:85-94.

5. Goodnight, S. H.Jr., W. S. Harris, W. E.Connor,and D. R.

Illingworth. 1982. Polyunsaturated fatty acids, hyperlipidemia, and

thrombosis. Arteriosclerosis.2:87-113.

6. Harris, W.S.,W. E.Connor, and M. P.McMurry. 1983. The

comparative reductions oftheplasmalipidsandlipoproteinsby di-etary polyunsaturated fats: Salmon oilversus vegetableoils. Metab. Clin.Exp. 32:179-184.

7.Phillipson,B.E., D. W.Rothrock,W.E.Connor,W.S.Harris,

and D. R.Illingworth. 1985.Reductionofplasmalipids, lipoproteins,

andapoproteins by dietaryfishoils inpatientswith

hypertriglyceride-mia. N.Engl.J.Med. 312:1210-1216.

8.Sanders,T.A.B.,D. R.Sullivan,J.Reeve,and G. R.Thompson.

1985. Triglyceride-lowering effect ofmarine polyunsaturates in pa-tientswithhypertriglyceridemia. Arteriosclerosis.5:459-465.

9.Harris,W.S., W.E.Connor,S. B.Inkeles,and D. R.Illingworth.

1984.Dietary omega-3fattyacidspreventcarbohydrate-induced hy-pertriglyceridemia.Metab. Clin. Exp. 33:1016-1019.

10. Saynor, R., D. Verel, and T. Gillott.1984.Thelong-term effect ofdietary supplementationwithfishlipidconcentrateon serumlipids, bleeding time, plateletsandangina.Atherosclerosis. 50:3-10.

11.Kingsbury,K.J.,D.M.Morgan, C. Aylott, andR.Emmerson. 1961.Effects of ethyl arachidonate, cod-liver oil, andcornoilon the

plasma-cholesterollevel. Lancet. 1:739-744.

12.Ahrens,E. H. Jr.,W. Insull,Jr., J.Hirsch,W. Stoffel,M.L. Peterson,J. W.Farquhar,T.Miller, andH. J.Thomasson. 1959. The effectonhumanserum-lipidsofadietary fat, highlyunsaturated, but poor in essentialfattyacids. Lancet.1:115-119.

13. Keys, A., J. T.Anderson, andF. Grande. 1957. "Essential"

fatty acids, degreeofunsaturation,andeffect ofcorn(maize)oilonthe serum-cholesterol level inman.Lancet.2:66-68.

14.Kinsell,L.W., G.D.Michaels,G.Walker,andR. E.Visintine.

1961. The effect ofafish-oil fractiononplasmalipids. Diabetes. 10:316-319.

15. Illingworth, D. R., W. S. Harris, and W. E. Connor. 1984. Inhibition of lowdensity lipoproteinsynthesis bydietaryomega-3fatty

acids in humans. Arteriosclerosis. 4:270-275.

16.Demke,D.M.,G. R.Peters,0. I.Linet,C. M.Metzler,and K. A. Klott. 1988. Effects ofa fish oilconcentrate inpatientswith

hypercholesterolemia. Atherosclerosis.70:73-80.

17.Steinbrecher,U.P., J. L.Witztum,Y.A.Kesaniemi,and R. L. Elam. 1983.Comparison of glucosylatedlowdensity lipoproteinwith

methylatedorcyclohexanedione-treatedlowdensity lipoproteinin the

measurementofreceptor-independentlowdensity lipoprotein catabo-lism. J.Clin.Invest.71:960-964.

18.Bilheimer,D.W.,Y.Watanabe,and T.Kita. 1982.Impaired receptor-mediatedcatabolismoflowdensity lipoproteinin the WHHL

rabbit,ananimal model of familialhypercholesterolemia.Proc.Natl. Acad. Sci. USA. 79:3305-3309.

19.Spady,D.K.,S. D.Turley,and J. M. Dietschy. 1985. Recep-tor-independentlowdensity lipoproteintransport in the ratinvivo:

quantitation, characterization,and metabolic consequences. J. Clin.

Invest. 76:1113-1 122.

20.Spady,D.K.,J. B.Meddings,and J. M.Dietschy.1986.Kinetic

constantsforreceptor-dependentandreceptor-independentlow

den-sitylipoproteintransport in thetissues oftheratand hamster. J.Clin.

Invest. 77:1474-1481.

21. Spady, D. K., and J. M. Dietschy. 1985. Dietary saturated

triacylglycerolssuppresshepaticlowdensity lipoproteinreceptor

activ-ityin the hamster. Proc.Natl. Acad.Sci. USA. 82:4526-4530. 22.Spady,D. K.,and J. M.Dietschy. 1988. Interaction ofdietary

cholesterol andtriglyceridesin the regulationof

hepatic

lowdensity lipoproteintransport in the hamster. J.Clin.Invest.81:300-309.

23.Spady,D.K.,S. D.Turley,and J. M.Dietschy. 1985. Ratesof

lowdensitylipoprotein uptakeandcholesterolsynthesisareregulated independentlyin theliver.J. LipidRes.26:465-472.

24.Weight,M.J.,A.J. S.Benade,-C. J.Lombard,J. E.Fincham,

M. Marais, B. Dando, J. V. Seier, and D. Kritchevsky. 1988. Low

densitylipoprotein kinetics inAfricanGreenmonkeysshowing vari-ablecholesterolaemicresponsestodietsrealistic for westernised

peo-ple. Atherosclerosis. 73: 1-11.

25. Kuskwaha, R. S., G. M. Barnwell, K. D. Carey, and H. C.

McGill,Jr. 1986. Metabolism ofapoproteinBinselectivelybred ba-boonswith low and highlevelsof lowdensity lipoproteins.J. Lipid

Res.27:497-507.

26.Glass, C.K., R. C. Pittman,G.A. Keller,andD. Steinberg.

1983.Tissue sites ofdegradationofapoprotein A-Iintherat.J.Biol. Chem. 344:7161-7167.

27. Bilheimer, D. W., S. Eisenberg, and R. I. Levy. 1972. The metabolism of very lowdensity lipoprotein proteins.I.Preliminaryin vitro and invivoobservations. Biochim.Biophys.Acta.260:212-221. 28.Mahley,R.W.,H.Weisgraber,G. W.Melchior,T. L.

Innerar-ity,and K. S.Holcombe. 1980.Inhibition ofreceptor-mediated

clear-anceoflysineandarginine-modified lipoproteinsfrom theplasmaof

ratsandmonkeys.Proc.Natl.Acad. Sci. USA. 77:225-229.

29.Spady,D.K.,D.W.Bilheimer,and J.M.Dietschy.1983. Rates ofreceptor-dependentand-independentlowdensitylipoprotein up-take in the hamster. Proc.Natl. Acad. Sci. USA. 80:3499-3503.

30.Jeske,D.J.,andJ. M.Dietschy. 1980. Regulationofratesof cholesterolsynthesisinvivo in the liver andcarcassoftheratmeasured

using

[3H]water.

J.LipidRes. 21:364-376.

31.Spady,D.K.,and J. M.Dietschy. 1983. Sterolsynthesisinvivo in 18tissues of thesquirrelmonkey,guinea pig, rabbit, hamster,and

rat.J.LipidRes.24:303-315.

32.Andersen,J.M.,and J. M.Dietschy.1978.Relativeimportance

ofhighandlowdensitylipoproteinsintheregulationof cholesterol

synthesisin the adrenal gland, ovary, and testis of the rat. J. Biol. Chem. 253:9024-9032.

33.Folch, J.,M.Lees, and G.H.SloaneStanley. 1957.Asimple

methodfor the isolation andpurification of totallipidsfrom animal tissues. J. Bio. Chem. 226:497-509.

34. Schwartz, A. L., S. E. Fridovich, and H. F. Lodish. 1982. Kineticsof internalization and recycling ofasialoglycoproteinreceptor inahepatoma cell line.J.Biol. Chem. 257:4230-4237.

35.Pardridge, W. M.,A.J. VanHerle,R.T.Naruse, G.Fierer, and

A.Costin. 1983.Invivoquantificationofreceptor-mediateduptake of

asialoglycoproteinsbyratliver. J. Bio. Chem. 258:990-994. 36. Spady, D. K., M. Huettinger, D. W. Bilheimer, and J. M. Dietschy. 1987. Role ofreceptor-independentlowdensity lipoprotein

transport in the maintenance of tissue cholesterol balance in the

nor-mal andWHHLrabbit. J.Lipid Res. 28:32-41.

37. Brown, M.S., J. R. Faust, and J. L. Goldstein. 1975. Role ofthe lowdensitylipoproteinreceptor inregulatingthecontentof free and esterified cholesterol in human fibroblasts. J. Clin. Invest. 55:783-793. 38.Goldstein,J.L., and M. S. Brown. 1977. Thelow-density

lipo-protein pathway and its relation to atherosclerosis.Annu. Rev.

Bio-chem.46:897-930.

39.Chen, I. S., S. S. Hotta, I. Ikeda, M. M.Cassidy,A.J.Sheppard,

andG. V. Vahouny. 1987.Digestion,absorptionandeffectson

cho-lesterolabsorption of Menhaden oil, fish oil concentrate and corn oil by rats. J. Nutr. 117:1676-1680.

40. Wong, S. H., P. J. Nestel, R. P. Trimble, G. B. Storer, R. J. Illman, and D. L. Topping. 1984. Theadaptive effects of dietary fish andsafflower oil on lipid and lipoprotein metabolism in perfused rat liver.Biochim. Biophys. Acta. 792:103-109.

41. Nestel, P. J., W. E. Connor, M. F. Reardon, S. Connor, S. Wong, and R. Boston. 1984. Suppression by dietsrich in fish oil of very lowdensity lipoprotein production in man. J.Clin. Invest. 74:82-89. 42. Wong,S., and P. J. Nestel. 1987.Eicosapentaenoic acid inhibits thesecretion oftriacylglycerol and of apoproteinBandthebinding of LDL inHep G2 cells.Atherosclerosis. 64:139-146.

43. Nossen, J. O., A. C. Rustan, S. H.Gloppestad, S. Malbakken,

andC.A.Drevon. 1986.Eicosapentaenoicacid inhibits synthesis and secretion oftriacylglycerols by cultured rat hepatocytes. Biochim.

Biophys.Acta.879:56-65.