The effect of oxidized lipids in the diet on serum lipoprotein peroxides in control and diabetic rats

The effect of oxidized lipids in the diet on serum

lipoprotein peroxides in control and diabetic

rats.

I Staprans, … , X M Pan, K R Feingold

J Clin Invest. 1993;92(2):638-643. https://doi.org/10.1172/JCI116632.

The levels of oxidized serum lipoproteins are increased in humans and animals with diabetes. We have examined the contribution of dietary oxidized lipids on the levels of oxidized lipoproteins. In both control and streptozocin induced diabetic rats, the oxidized lipid content of mesenteric lymph chylomicrons (CM) increased when increasing quantities of oxidized lipids were administered intragastrically. However, at all levels of administered oxidized lipids, the quantity of oxidized lipids in CM was greater in the diabetic animals. These results indicate that oxidized lipids are absorbed and packaged into CM and suggest that there is increased absorption of oxidized lipids in diabetic animals. In nondiabetic rats fed a fat-free diet, the levels of oxidized lipids in their serum lipoproteins were very low. When oxidized lipids were added to the diet, the quantity of peroxides in serum lipoproteins increased about fivefold. In diabetic animals fed a fat-free diet, there were also very low levels of oxidized lipids in their serum lipoproteins, and there was no difference between control and diabetic rats. However, when diabetic animals were fed a diet containing oxidized lipids, the quantity of oxidized lipids in their serum lipoproteins increased 16-fold and were significantly greater than in controls. Thus, in both control and diabetic rats the quantity of oxidized lipids in the diet largely determines the levels of oxidized lipids […]

Research Article

Find the latest version:

http://jci.me/116632/pdf

The Effect of Oxidized

Lipids in the Diet

on

Serum

Lipoprotein Peroxides

in

Control

and Diabetic

Rats

IlonaStaprans,Joseph H. Rapp, Xian-Mang Pan, and Kenneth R. Feingold

TheDepartmentofVeteransAffairs MedicalCenter and theDepartmentsofSurgery and Medicine, Universityof California, SanFrancisco,California94121

Abstract

Thelevelsof oxidizedserumlipoproteinsare increased in hu-mans andanimalswith diabetes. We haveexaminedthe

contri-butionof dietary oxidized lipidson the levelsof oxidized

lipo-proteins.Inbothcontroland streptozocin induced diabetic rats,

the oxidized lipid content of mesenteric lymph chylomicrons (CM) increased whenincreasing quantitiesof oxidizedlipids

were administered intragastrically. However, at all levels of

administered oxidized lipids, the quantity of oxidized lipids in CMwasgreaterin the diabetic animals.Theseresults indicate thatoxidized lipidsare absorbed andpackaged intoCM and suggest that thereis increasedabsorption of oxidizedlipids in

diabetic animals. In nondiabetic rats fed a fat-free diet, the

levelsof oxidized lipids in their serum lipoproteinswerevery

low. Whenoxidized lipidswereaddedto thediet,thequantity ofperoxides inserumlipoproteins increasedaboutfivefold.In

diabetic animals fed a fat-free diet, therewere also very low

levels of oxidized lipids in theirserumlipoproteins, and there

was nodifferencebetweencontrol and diabeticrats. However, whendiabetic animals werefedadiet containing oxidized lip-ids, the quantity of oxidized lipidsintheirserumlipoproteins increased 16-fold and were significantly greater than in

con-trols. Thus, in both control and diabeticrats thequantity of oxidizedlipidsin thediet largely determines the levels of oxi-dized lipids in circulating lipoproteins. However, in diabetic animals the effect of diet is morepronounced. Togetherwith

theCMstudies,theseresultsdemonstratethatdietaryoxidized lipids makeamajor contributiontothe levels of oxidizedlipids

incirculating lipoproteins and indicate that increased absorp-tion of oxidizedlipidsindiabeticanimalsmayplayarole in the elevation of oxidizedlipoproteinsobservedinthisdisorder.(J.

Clin. Invest. 1993.

92:638-643.)

Keywords:

lipid

peroxides*

oxidizeddietary fat*serum

lipoproteins

*

chylomicron

Introduction

Atherosclerosisisamajorcauseofthe

morbidity

and

mortality

associatedwithdiabetes. TheFraminghamand other studies

have shown that theincidence of cardiovascular disease is

sig-nificantly increased in diabetes even when adjusted for the

otherknownrisk factors for atherosclerosis( 1, 2). This

indi-cates thatdiabetic patientshaveanincreased risk of vascular diseasethatisnotsimplytheresultofanincreasedprevalence

of riskfactors,but rather thatdiabetes isanindependent risk

AddresscorrespondencetoIlonaStaprans,Ph.D.,LipidResearch Lab-oratory(151L),VAMedicalCenter,SanFrancisco,CA94121.

Receivedfor publication29April1992andinrevisedform2 De-cember1992.

The Journal ofClinicalInvestigation,Inc. Volume92,August1993, 638-643

factor forvascular disease.Themechanismbywhich diabetes

increasestherisk of atherosclerosisremainsspeculativedespite

intensive investigativeefforts.

Recentstudies have shown that the oxidation of lipopro-teinsmayplayakey roleinatherogenesis(3-5). Lipid

peroxi-dationconverts lipoproteinsto a moreatherogenicform (3-5).

Oxidized LDLand other lipoproteins, including oxidized

f-VLDL (6), are taken up more rapidly bymacrophages,leading tofoam cell formation.Inaddition,these oxidizedlipoproteins

cause increased cytotoxicity to endothelial cells in culture,

which could result in endothelial injury (7). Moreover,

oxi-dizedLDLhas been showntostimulatethe release of

chemo-tactic proteins, cytokines,and growth factors from endothelial and other cells (3-5, 8-10), resultingin migrationof mono-cytes, maturation of monocyte/macrophages, foam cell

for-mation, and proliferation ofsmooth muscle cells.

Severalstudies have demonstrated thatthelipoproteins iso-lated from both humans (11, 12) and experimentalanimals

with diabetes (13-15) are oxidized to a greater extent than

lipoproteinsfrom controls.Inrats,oxidized lipoproteinswere

found inVLDL+ LDLfraction( 15 ). Thus,diabetesis

asso-ciated with the increasedpresenceof oxidizedserum

lipopro-teins,and theselipoproteinshavethepotential for causing

tis-suedamage and lipid accumulation inthe arterial wall. The

mechanismby which diabetesresultsinincreased levels oflipid

peroxidesinserumlipoproteinsis unknown. Leading

hypothe-sesthat couldaccountfor the increased oxidation of

lipopro-teinsinclude (a) diabetes stimulates peroxide-producing sys-temssuchaslipoxygenase ( 16),(b)indiabetes there isa

de-creasein the antioxidantcontentof certain tissues, resulting in

adecreased

protection against

oxidation (

17-19), (c)

second-ary todiabetes there isanincrease in tissuedamage,

resulting

in increasedoxidation(20,21),and

(d)

indiabetesmodification

of

lipoproteins

by glucoseenhances theoxidationof

lipopro-teins(22).

Anotherpossibilityis that oxidized

lipids

in the diet could bea

contributing

factortotheincrease ofoxidized

lipoproteins

found indiabetes.Numerousstudieshave shown that diabetes alters thestructureandfunction ofthesmall intestineand that

thesechanges resultinincreased

absorption

ofa

_variety

ofsub-stances including

lipids (23-25).

We havetherefore investi-gated theeffectofoxidized

dietary

lipids

onthelevels of oxi-dized

lipoproteins

in both the

lymph

andserumof control and diabeticrats.

Methods

Materials. Streptozocin,2-thiobarbituric acid,

1,1,3,3-tetramethoxy-propane,vitamin E, linoleicacid,andD20wereobtained from

Sigma

ChemicalCo., St.Louis, MO.Ketodiastixwereobtained fromAmes Division,MilesLaboratory,Elkhart,IN. Vitamin

E-depleted

and natu-ralcornoilwerepurchasedfrom ICNPharmaceuticals,CostaMesa,

Co., ThousandOaks, CA. 1-'4C-linoleicacid andglycerol tri(9-3H) oleatewerepurchasedfrom NewEngland Nuclear, Boston,MA.

Animals.FemaleSprague-Dawleyrats( 180-200g)werepurchased fromBantin andKingmanInc.(Fremont, CA) andweremaintained on reverse 12-hlight cycle (0300-1500dark, 1500-0300 light). Dia-beteswasinduced inratsbyinjecting,afteranovernight fast,40mg/kg streptozocin i.p. in 1 M sodium citratebuffer, pH4.5. The urine of these animalswasanalyzedwithKetodiastix,and animals that didnot have at least 1%glycosuriaatalltimeswereeliminated from thestudy. Control animalswereinjectedwith 1 M sodium citrate andwere main-tained under similar conditions. Animalswerestudied 14 d after strep-tozocin treatment and werefastedfor 6 h before all experiments. All animal studies wereperformed in accordance with institutional poli-cies, and all procedureswereapproved by the Subcommittee on Ani-malStudies.

Diets. Control anddiabetic rats were maintained on regular rat chowsuppliedby Simonsen Laboratories, Gilroy, CA. Alternatively, 10 d afterstreptozocin injection, diabetic rats and their controls were fed for 4-5 d either alipid-freediet consisting of 65% sucrose, 20% casein,4%salt, 1 1% alpha cell mix andvitaminsorthe samediet to whicheither natural corn oil (low peroxide; 4.7 nmol peroxide/mg oil) orvitamin E-depleted cornoil (high peroxide; 82.5 nmol peroxide/mg oil)wasadded.

Isolation ofserum lipoprotein fractions. Previously it has been shown that indiabeticratserum, lipid peroxides are in VLDLand LDL,butnotin the HDLfractions ( 15 ). Since rats contain very low amounts of LDL, all serumlipoproteinoxidation measurements in our studies weredoneoncombinedfractions containing VLDL+ LDL. Foreach measurement, lipoproteins were obtained from 9 ml serum (pooled fromtworats) afteradjustingthedensity to 1.063 withKBr andcentrifugingat 5 X 106gfor 18 hat 14'C. Thelipid layer was removed,dialyzed against 0.15MNaCl containing 0.5 mM EDTA and 0.5gMbutylatedhydrohytoluene(BHT),'and analyzedfor oxidation immediately.When serum samples were examined for chylomicrons bycentrifugingfor 20 minat5X 196gminbeforeKBraddition, no significantamountsof chylomicronswereisolated.

Chylomicron preparation. Chylomicrons were isolated from rat mesentericlymphusingtheprocedure describedpreviously (26).Rats were intragastrically administered a single bolus ofalipid emulsion consistingof 3 ml ofoil andnonlipidmilk(1:2, vol/vol) dispersedwith aPolytron homogenizer(Brinkmann Instruments, Westbury, NY). Thefollowingoilswere usedforgastric intubation: naturalcornoil with lowperoxide content (4.7 nmol peroxide/mg oil); commercially obtained vitamin E-depletedcornoil with mediumperoxidecontent (82.5 nmolperoxide/mg oil);andvitaminE-depleted cornoil that had been heatedfor 1 h at 100°C(high peroxidecontent, 120.0 nmol peroxide/mg oil).Therats wererestrained,andchylewascollectedon icefor8 h.Duringthisperiodratswereallowedtodrinkwaterandwere periodically injected subcutaneouslywith 0.15 M salinetomaintain hydration. Chylomicronswereisolatedfrom lymph byflotationin an SW-40rotor(BeckmanInstr.Inc.,Fullerton, CA)for SX106gfor 1 h at 12°C.Allsampleswerestored in 1mM EDTAundernitrogen. In someexperiments, 0.1 mCiglyceroltri(9-3H)oleate andoxidized or nonoxidized ['4C]linoleicacid(0.02 mCi)wasadded to the vitamin E-depletedcornoil. Oxidation of linoleic acid was carried out as de-scribed by Boeynaemsetal. forpreparation of oxidized arachidonic acid(27).['4C]Linoleicacidwasoxidized for2 hin methanol in the presenceof 5AMCu2SO4and5-10 mg linoleic acid carrier. Measure-mentby GLC (28) indicated that more than 95% of the linoleic acid wasoxidized. For partial purification, the oxidized linoleic acid was subjectedtothin layer chromatography, reisolated, and then used im-mediatelyin the intubation mixture (28).

Analytic methods. Triglycerides in chylomicrons and the LDL +VLDL lipoprotein fraction were determined by the method of

1.Abbreviations used inthis paper: BHT, butylatedhydrohytoluene;

CM,chylomicrons;TBARS,thiobarbituric acid reactive substances.

Fletcher(29). Cholesterol levelsweredeterminedusing Sigma

diag-nostic kit(procedure351).Proteinwasdeterminedbythe method of Lowryetal.(30). Lipidoxidationwasestimatedby measuringboth

lipid peroxidesandlipid peroxide decomposition products. Lipid per-oxideswereestimatedby acolorimetric methodasspecified bythe manufacturerusingacommerciallyavailable kit(LPO)fromKamiya

BiomedicalCo. The method is basedonahemoglobin catalyzed

reac-tion ofperoxides with amethyleneblue derivative that specifically quantitates lipid peroxides by forminganequalmolar concentration of methyleneblue(31 ).Cumenehydroperoxidewasusedasastandard.

Lipid peroxide decomposition products,whichconsist ofavarietyof aldehydes,weremeasuredasthiobarbituric acid reactive substances (TBARS) (32).

Student'sttest wasusedtotestforthesignificancebetweenmeans.

Resultsarereportedasmeans±SE.

Results

Effect ofdiabetes on oxidizedserum lipoproteins. When fed

standard ratchow, serum lipoproteins (LDL+ VLDL)from

diabetic ratshad agreaterquantityof TBARS than

lipopro-teins from control rats (1.04±0.13 nmol/mg cholesterol in controlsvs. 1.75±0.31nmol/mgcholesterolindiabeticrats;P <0.05)

(Fig.

1 A).This observation is inagreement with the

published

results of Morel et al. ( 15) that demonstrated

in-creased TBARS in rats with streptozocin-induced diabetes.

ThemeasurementofTBARS is themostwidelyused method

forestimating lipid oxidation; however,it maynotbespecific

or quantitative since the generated aldehydes may be water

soluble andnotremain withthelipoproteins duringthe

isola-tionprocedure.Toconfirm thatlipoproteinsisolatedfrom

dia-beticrats areoxidizedto agreaterextentthan those from con-trol rats,weestimatedlipoproteinoxidationby directly

mea-suring lipid peroxides in serum lipoproteins using a

quantitative

andstoichiometriccolorimetricmethod(31_).As

showninFig. 1 B,with this method thelipoproteinsfrom

dia-beticratscontained 10-foldmorelipid peroxidesthan did those

from controlrats(1.97±1.57nmol/mgcholesterolincontrols vs. 19.62±6.82 nmol/mg cholesterol in diabetic rats; P

A. TBARS in serum VLDL + LDL

0 L S

.A

0

Eu

Em

E

0 E C

3.

2-

I1-0

4-p<O.05

control

animal

B. Peroxides in serur

L-

*o30-o 0

0

E 1

co

p<.05e

control

animal

.111,11,11,011.-1 . Figure 1. TBARS (A) diabetic andlipidperoxide(B)

I _{content of VLDL}

m VLDL + _LDL + LDL lipoprotein

frac-tion isolatedfrom the serumof control and diabeticratsfeda regu-larratchowdiet.All data

_means±SE

arerepresented_{and there}as

werefiveto

_six

serum

samples in each group. Eachsample represents

dia

bectig

c serumpooled from two

<0.05).Thus,usingtwodifferentmethods for estimating

lipo-protein oxidation thelipoproteins isolated from diabetic rats were more oxidized than lipoproteins isolated from control rats.

Effect

ofdietary oxidized lipids on the peroxide content of

serumlipoproteins. Todetermine the contribution of oxidized lipids in the diet to the content of oxidized lipids in serum

lipoproteins,control anddiabeticrats werefeddiets containing different quantities of oxidized lipid. Since the oxidized lipid content ofcommercially available standard rat chow is not

controlled,weusedsynthetic diets prepared inourlaboratory.

Fig. 2 shows serum peroxide concentrations in control rats. When rats werefedalipid-free diet(sucrosediet),the levelsof peroxides in serum lipoproteins were low (1.00±0.09 nmol/ mgcholesterol). Similarly, whencontrol rats werefedadiet containing 5% corn oil containing low peroxides (4.7 nmol

peroxide/g oil), the quantity of peroxides in the serum lipo-proteins were also low (0.87±0.11 nmol/mg cholesterol). However, whenthe lipid-containing diet had high quantities of peroxide (5% oil, 82.5 nmol peroxide/g oil),thequantity of peroxides in the lipoproteins increased almost fivefold (1.00±0.09 nmol/mg cholesterolvs.5.34±0.50 nmol/mg cho-lesterol;P <0.001 ). MeasurementsofTBARS confirmed these results. TBARS increased twofold when oxidized

lipid-con-tainingdietwasfedtotheanimals instead ofalipid-freesucrose

diet.Thus, in control animals the quantity of oxidized lipidsin

thediet influencesthe concentration ofperoxidesin theserum

lipoproteins.

In diabetic animals (Fig. 3) fed a lipid-free diet (sucrose

diet),which ofcoursecontainednooxidizedlipid,therewere

alsoverylow levelsofperoxidesin thelipoproteins.Infact,the concentration ofperoxidesin thelipoproteins ofdiabetic ani-malsfed thelipid-free dietwas not different from theserum

peroxideconcentration observedincontrols(Fig. 2). Thus,the increase in thequantity ofoxidized lipidsinserumlipoproteins

of chow-fed diabetic animals(Fig. 1) disappearswhen animals arefedadietcontainingnooxidizedlipid.

Whendiabeticrats werefedadietcontainingoxidized lip-ids,thequantity ofperoxidesinserumlipoproteinsincreased

approximately 16-fold (0.70±0.10 nmol/mg cholesterol vs.

11.08±2.01 nmol/mg cholesterol; P < 0.005) (Fig. 3). As

noted above(Fig. 2), incontrol animals fed oxidizedlipidat the sameconcentration, the quantity of peroxides in the serum

Peroxides in serum VLDL+LDL Figure 2. Peroxide

con-0 6- tentof VLDL+LDL

lipoprotein fraction iso-latedfrom theserum

Z 4- of controlratsfed for 5

3-////dthefollowingdiets:

cm

2 // (a)fat-freesucrosediet,

".2- (b)sucrosedietto

1 i, i | .which 5% oilcontaining

E lowlipid peroxides (4.7

sucrose oil oxidized oil nmol/mgoil)was

diet added,and(c)sucrose

diettowhich 5% oil containing highlipidperoxides (82.5 nmol/goil)wasadded. All data arerepresentedasmeans±SE,and therewerefiveserumsamplesin each group. Eachsamplewasderived fromserumpooled fromtwo

rats.P<0.001 when control group fedafat-free dietwascompared

with the control group fed oxidized fat diet.

Peroxides in serum VLDL.LDL Figure3. Peroxide

con-Z 14- _tentof VLDL₊LDL

'1 2 T lipoprotein fraction

iso-@10 @ lated fromtheserum

= 10 _{Hof diabetic rats fed for 5}

o>//_{d thefollowingdiets:}

E

6 _(a)fat-freesucrosediet,

E

4- (b)sucrosedietto

o 2- which 5%oilcontaining

E _{o eyz=}

1B//////B1///// _

lowlipid peroxides (4.7

sucrose oil oxidized oil nmol/mg oil)was

diet added, and (c) sucrose diettowhich 5%oil containinghigh lipid peroxides (82.5 nmol/goil)wasadded. Alldata arerepresentedasmeans±SE,and therewerefiveserumsamplesin each group. Eachsamplewasderivedfromserumpooledfromtwo rats. P<0.005 whenadiabeticgroupfedafat-free dietwascompared with thediabeticgroupfed an oxidizedfatdiet;P<0.05 when control groupfedanoxidizeddiet(Fig. 2)wascompared withthediabetic groupfedthe same diet.

lipoproteins increasedonly byfivefold.Moreover, the quantity

oflipid peroxidesintheserumof diabetic animalswas twofold

higher than in controls fedthe samediet(P < 0.05). Similarly,

TBARSincreased threefoldwhen rats were fed

oxidized-lipid-containing diet instead ofa sucrosediet.

Indiabeticrats,theserumperoxidecontent was low when

fedafat-freesucrosedietandslightly increasedwhenoil

con-tainingsmallamountsof peroxideswas added to the sucrose

diet(Fig. 3).However, thisincrease(2.1±0.6nmolperoxide/

mgcholesterol) didnotachieve statisticalsignificance.

Addi-tionally,theincreasewasmuchless than thatobservedwitha

diet containing high quantities of peroxides in the oil

(11.08±2.01 nmol peroxide/mgcholesterol). Sincethe lipid contentin thedietwassimilar, theseresultsindicatethat the

increased peroxides in serum lipoproteins of diabetics result

from theingested oxidized lipids andnotfromthe total

quan-tityoflipidin thediet.

Additionally, it can be seen that while both control and

diabeticanimals demonstrateanincreasein theconcentration

ofperoxidesinlipoproteinsin response todietary oxidized

lip-ids, theextentoftheincrease issignificantlygreaterindiabetic

animals, suggestingthat theyare more susceptibleto dietary

oxidizedlipids.

Fig.4showsadose-responsecurveof the effect ofadding increasing quantities of oxidizedlipidstothedietonthe

con-centration ofperoxidesincirculating lipoproteins. Thereisa

Peroxides in serum VLDL + LDL Figure4. Dose response Z 3- _{I curveof the effect of}

increasingamountsof lipidperoxidesin the diet on the serum lipid

peroxidesinVLDL E_{_1- / r=0.953} + LDL lipoprotein

frac-~~~~tion.

Controlratswere

o fedfor5 da sucrosediet

E

- . , , . , . , to which increasing

0 2 4 6 8 amounts oflipid

perox-oxidized oil (% diet) ide containing oil (82.5 nmol/goil)wereadded. Alldataarerepresentedasmeans+SE,and therewerefourserum

strong correlation between thequantityofperoxidesin the diet andthe concentration of peroxides in thecirculating

lipopro-teins (r = 0.953, P <0.001), furthersuggesting thatdietary

peroxides are an important source of oxidized lipid inserum lipoproteins.Interestingly,there was no increase in serum

lipo-protein peroxides when the oil content in the diet wasincreased

from 0.0% to 2.5%,raisingthepossibilitythat there is a

thresh-old of oxidized lipid intake below which theperoxidecontent of serum lipoproteins is not altered. However it is alsopossible

that with a more sensitive assay procedure low quantities of

dietaryoxidized lipidmightbe shown to affect thequantityof oxidized lipoproteins.

Theeffect ofdietary oxidized lipidsonthe

peroxide

content

ofchylomicrons from

continuously

collected intestinallymph.

To determinedirectly whether diabetic rats absorb a greater

quantity of oxidized lipid than do the controls, intestinal

lymphwascontinuously collected from control and diabetic rats for 8 h after the intragastric administration ofvarying

amounts of lipidperoxides. Duringthe 8-hperiodofcollection,

the quantity oftriglyceridessecreted in thelymphofdiabetic

and control rats was similar (506±67 mg in diabetics vs. 456±26 mgtriglycerideincontrols). Fig. S illustrates that in control rats the quantity oflipid peroxides in chylomicrons

isolated from the collectedlymphincreases withincreasing per-oxide content of theintragastrically administered oil. When

expressedasnmolperoxide/mgtriglyceride, theperoxidesin

lymphchylomicronswas0.380±0.06 nmol/mg triglyceridein

the low peroxide group vs. 0.580±0.01 nmol/mgtriglyceride in thehighperoxidegroup(P<0.01 ). The results indicatethat

dietary peroxides are absorbed and packaged into chylomi-crons.

In diabetic animals, a similar increase in thequantity of

peroxidesinchylomicrons in responsetoincreasingamounts

ofperoxideadministrationwasobserved(Fig. 5).When chylo-microns isolated from diabetic animalswere compared with

chylomicronsfrom controls, lipid peroxidesweresignificantly

increased in the medium and high peroxide groups: 0.640±0.04vs.0.490±0.03 nmol(P <0.005) inthemedium peroxide group and 0.830±0.02 vs. 0.580±0.01 nmol (P

Peroxides in CM

CD 1.0

I-E

0.8-"I, V

* 0.6

-CL

(n

!!0.2

E

C 0.0

low medium high

peroxide content In oil

Figure5.Lipid peroxide contentof chylomicrons (CM)from control and diabeticrats.Rats were administered oil contain-ing increascontain-ingamounts oflipid peroxides. Low, medium,and high per-oxide oil contained 4.7, 82.5,and 120.0 nmol peroxide/mg oil, respec-tively.Lymphwas col-lected for 8hfor chy-lomicron isolation.All data areexpressedasnmol peroxide/mg chylo-micron triglycerideand arerepresentedasmeans±SE.Therewerefive ratsin each group. P<0.005 when control group fed a medium perox-ide oil diet wascomparedwith thediabeticgroup fed the same diet; P

<0.001when control group fed a highperoxidediet wascompared withthediabetic groupfed the same diet; P < 0.01 whencontrol group fed high peroxide dietwascomparedwith the control group fed low peroxide diet;P<0.005whendiabeticgroup fed a highperoxidediet wascomparedwith thediabetic groupfed the lowperoxidediet. z control;udiabetic.

<0.001

)

in the

high peroxide

group. Similarresultswere ob-tained whenTBARSin

_chylomicrons

weremeasured:

high

per-oxide

_controls,

0.59±0.04

nmol/mg triglyceride; high

peroxide

diabetics,

0.81±0.02

nmol/mg triglyceride;

P<0.001. Because

thedifference betweenthe

_quantity

ofoxidized

_lipid

in

chylo-microns isolated fromcontrol anddiabeticratsis greatest when the

peroxide

contentinthe oil is the

highest,

it is

likely

that the

peroxidesinlymph chylomicrons areprimarilyderivedfrom the diet andnotfrom oxidationin the smallintestine.

Thus,

after the

intragastric

administration of the same

quantity

of total

_{lipid peroxides, lymph chylomicrons}

from

diabeticanimals contained

significantly

more

_{lipid peroxides}

than did

_chylomicrons

fromcontrols.

The

effect

of

'4C-labeled

dietary

oxidized linoleic acidonthe

'C-labeled

linoleic acidcontent

ofchylomicronsfrom

continu-ously

collected intestinal

lymph.

Tofurtherdemonstrate that diabetic ratsabsorb greater

quantities

of oxidized lipid than control _{rats, intestinal}

_lymph

wascollected fromcontrol and

diabeticratsfor 8 hafter theintragastricadministration ofoil

containing

oxidized['4C]linoleicacid. _[3H

_ITriolein,

which is

not

easily oxidized,

was _{simultaneously} added to the oil to

correctfor variations in the

quantity

of

lymph

collected.When the ratio of oxidized ['4C]linoleic acid to [3H ]triolein

('4C/3H)

recoveredinchylomicronswasexamined in control

and diabetic rats, we found that the ratio was increased

(2.67±0.37)

in diabetic rats when compared with controls

( 1.34±0.06),

P<0.01 (Table

I).

In contrast,the ratiowasthe

samein control

(

1.55±0.21)anddiabetic animals

(

1.50±0.12)

when nonoxidized [ '4C]linoleic acidwasadministered. These data further demonstrate that diabetic animals have an in-creased absorption ofoxidized _{dietary lipids}with more

oxi-dized

lipids

recovered in thelymph chylomicrons ofdiabetic animals.

Discussion

Severalstudies have shown that thequantityof oxidized

lipids

inserumlipoproteinsisincreased in diabetic patients (11, 12)

and inanimalswithexperimentally induced diabetes

( 13-15).

Numerousin vitro and in vivo studies suggestthatoxidized

lipoproteins playanimportantrole inatherogenesis(3-5). Ox-idizedlipoproteinsleadto anincrease in lipid uptake and

depo-sition inmacrophages, resultingin foam cellformation(3-5). Additionally,oxidizedlipoproteinsarecytotoxictoendothelial

cells inculture(7). Furthermore,oxidized lipoproteins

stimu-late therelease ofcytokinesand growthfactorsthat could po-tentiate theatherogenic process (3-5, 8-10). An increase in oxidized

lipoproteins

could therefore contribute through

sev-TableI. Ratioof['4C]LinoleicAcid['HiOleicAcidLabel Recovered in Chylomicrons

Nonoxidized linoleic acid Oxidized linoleic acid

Controlrat* 1.55±0.21 Control ratt 1.34±0.06 Diabetic rat* 1.50±0.12 (NS) Diabetic ratt 2.67±0.33§

Control and diabeticratswereadministered nonoxidized[3H]oleic acidsimultaneouslywith either oxidized or nonoxidized['4C]linoleic acid. Lymphwascollected for 8 h and the'4C/3Hin the lymph was determined. * Values representanaverage fromfourpreparations.

eralmechanisms to the increased risk of atherosclerosis that is associated with diabetes.

The origin of theincrease in oxidized lipoproteins in dia-betesis unknown. Inthepresent study we have examined the effect ofoxidized lipids in the diet on the levels of serum oxi-dized lipoproteinsand thepossibilitythat an increased

absorp-tion of lipid peroxides in the diabetic is responsible for the elevated levels of lipid peroxides in the serum of diabetics.

Incontrolanimals,thequantity of oxidizedlipid inthediet

directlycorrelatedwith the concentration of oxidized lipid in theserumlipoproteins (r= 0.953,P < 0.001). Animalsthat

werefed a diet containing no oxidized lipid had very little oxi-dizedlipidinserumlipoproteins, while animals fedadiet

con-taining increasing quantities of oxidized lipid had increasing concentrations of oxidized lipid in their serum lipoproteins. These results indicate that dietary oxidized lipids play an im-portantrole indetermining the concentration of oxidized lipid inserumlipoproteins. Moreover,the quantity of oxidized lipid

inchylomicrons isolated fromthe mesenteric lymphdrainage

ofthe smallintestine also closely correlated with the amount of

oxidized lipid administered to the animals (r = 0.989, P

<0.01 ). These resultsindicatethatoxidized lipids in the diet

areabsorbedby the smallintestineand are transported in chy-lomicronstothecirculation, where they contributeto the total

body pool of oxidized lipids. Giventhat ourstudies

demon-stratethatoxidizedlipids in the diet affect not only the concen-tration ofoxidizedlipid in chylomicrons but also the serum

VLDL + LDLlipoprotein fraction,one can speculatethatthe

oxidized lipid in chylomicrons is either delivered to the liver. whereit is repackaged and secreted in endogenous lipoproteins

ortransferred in the circulationtootherlipoprotein particles. The presenceofoxidizedlipid inratlymph has been

inves-tigated previouslywithcontradictoryresults. Several

investiga-torshavenotfoundanyoxidizedlipidsin the lymph and have concludedthatdeperoxidationtakesplace in the intestinal mu-cosa(33-35). Recently,however,Aw etal.(36), by

measuring

TBARS, have demonstrated lipid peroxidation in the rat

lymph. The degree oflipid peroxidation correlated with the GSHcontentof the intestinalmucosacells.

Decreasing

intes-tinalGSH levelsbyfeedinginhibitorsof GSH transferase

re-sulted in an increase in

absorption

and transport into the

lymph of oxidized

lipids.

Theseresultsindicatethat the

absorp-tion andtransportof oxidized

lipids by

the small intestinecan

vary dependingupon GSH levels in the small

intestine,

and such differences

might

accountfor the

conflicting

observations

regarding the presence of oxidized

lipids

in the

lymph.

Re-cently we havedemonstratedthat

dietary

lipid peroxides

are

incorporated intoratlymph

chylomicrons

and these

lipid

per-oxides have a significant influence on the metabolism of plasmachylomicrons inrats(37).

Asreportedpreviously byMorel et al. (15), wealso have observed that theconcentration of oxidizedlipidintheserum

lipoproteinfraction(d< 1.063) from diabeticratsfeda stan-dard chowwasincreasedin

comparison

withcontrols.Amain conclusion ofthepresentstudyis thatanincreased

absorption

ofdietary oxidized lipidsisprimarily responsible forthe ele-vatedlipid peroxidesinthe serum ofstreptozocin-induced dia-beticrats.This conclusion issupported bythree separate find-ings. First,intheabsenceoflipid peroxidesin thedietno in-crease in serumlipid peroxidesoccursin thediabeticanimals compared with controls.Second,theleveloflipidperoxidesin

the serum of control and diabetic ratsclosely correlates with

theleveloflipidperoxidesinthediet. However, inthediabetic

ratthere isagreaterelevation oflipid peroxides in theserum

than inthe controlratwhenfedadietcontaining lipid

perox-ides.Third,theincreasedabsorptionof oxidizedlipids bythe diabeticratis directlydemonstratedby experimentsshowing

that, following gastricadministration oflipidperoxides,

dia-beticratsabsorbedanincreasedamountof oxidizedlipidinto their intestinallymphcompared withnondiabeticrats.Taken

together,these results indicate that theincreased absorption of lipid peroxidescontributestotheelevated levelsofserumlipid

peroxidesthatcharacterize the diabeticstate.Additionally, dia-beticanimalsarehyperphagic (23, 38)andthereforewill con-sume increased quantities ofdietary oxidized lipids, which could also contributetothe elevated

quantities

of

lipid

perox-ides inserumlipoproteins.

Theincrease inabsorption of oxidizedlipidsby the small intestine in diabeticratscould be dueto anumberoffactors.As

discussed above, Aw et al. (36) have shown that decreased GSH levels in the intestinal mucosaleadto anincreasein

ab-sorption of oxidizedlipids by the small intestine. Studies in diabeticratshave shown that thereisadecrease in GSH

trans-ferasein the small intestine( 14)thatcould leadtodecreased GSH levels andtherebyincrease oxidizedlipid

absorption.

Fur-thermore,Lovenetal.(18)haveshown that the

activity

ofthe

superoxide

dismutase,

the

primary

enzymein

reducing

hydro-peroxide,is decreased in the smallintestine ofdiabeticrats.Itis

possible that this decrease could also contribute to the in-creasedabsorptionofoxidized

lipids

indiabeticrats.

Inadditiontoalterations in the GSHsystemandoxidative

stateof the small intestine in diabeticrats,otherfactorscould alsocontributetothe increased oxidized

lipid

absorption. Dia-betes results ina

variety

of

changes

inintestinal

lipid

metabo-lism. Both cholesterol and

fatty

acid

synthesis

are increased approximately twofold in the smallintestine ofdiabetic ani-mals(38, 39). Moreover, thetransportof newly

synthesized

cholesterol from the small intestine tothe circulation is in-creased fourfold indiabeticanimals

(40).

Furthermore,

several

investigators

have shown that thetransportof

triglyceride

from

the small intestinetothecirculation in the

postabsorptive

state

is also increasedbydiabetes (41-43). Thesestudies indicate thatthe fluxof

lipid

from the smallintestinetothecirculation is enhancedby

diabetes,

andit is

possible

that the increase in oxidized

lipid

transport is related to thesealterations in the

lipidflux.

Last, diabetes has been showntoaffectthe

_absorption

ofa

varietyof nutrients. The increased

absorption

of amino

acids,

glucose, bile

acids, phosphate

and

fatty

alcohols have been dem-onstrated indiabetic animals

(24, 25).

Of greater relevance is the finding that the

absorption

of cholesterol is increased in diabetes(25, 44,

45).

A

_{single study}

alsoindicatesanincrease

infattyacid

absorption (46).

These increases in intestinal

ab-sorptionhave beenattributedtoincreased maximal transport

capacity, increased

passive permeability,

alterationsinthe in-testinalwaterlayer,andincreased mucosalmass.

_Regardless

of theexact

etiology,

itis

possible

that such

_changes

in the small intestine of diabetic animalsalsoleadtotheincreased

absorp-tion of oxidized

lipids.

In summary, the presentstudydemonstratesthat the

quan-tity ofoxidized

lipid

in the diet affects the concentration of oxidized

lipid

in

_{circulating lipoproteins}

in both control and diabetic rats. In diabetic

animals,

both food intake and the

account for the increase inthe concentration of oxidizedlipids

in serumlipoproteins.Given thewidespreaduse offried foods in both commercial and domesticsettings,the American diet is rich in thermally oxidized lipid, which could lead to an in-creasedquantityof oxidized lipid inserumlipoproteins, espe-cially in diabetics, and thereby contribute to atherosclerosis both in nondiabetic and diabetic individuals.

Acknowledgments

Theexcellent technical assistance of Ms. Kelly Kim and Mr. Arthur Moserisappreciated. Drs. Marvin D. Siperstein and Carl Grunfeld are gratefully acknowledged for thehelpfuldiscussions.

This workwassupportedby the Medical Research Service, Vet-eransAffairs, National Institutes of Health HL-4 1470 and Pacific Vas-cular ResearchFoundation.

References

1.Kannel, W.B., and D. L.McGee. 1979.Diabetesandglucose tolerance as

risk factors forcardiovascular disease: theFramingham study.DiabetesCare. 2:120-126.

2.Feingold,K.R., and M. D.Siperstein.1986.Diabeticvasculardisease. Adv.

Int.Med. 31:309-340.

3.Steinberg,D., S.Parthasarathy, T.Carew,J.C. Khoo, and J. L. Witztum. 1989. Beyond cholesterol.Modificationof low density lipoprotein that increases its atherogenicity. N.Engl. J.Med.320:915-924.

4.Steinberg, D.,and J. L.Witztum. 1990.Lipoproteinsandatherogenesis. Current concepts. JAMA (J. Am.Med.Assoc.).246:3047-3052.

5.Witztum, J. L., and D. Steinberg. 1991. Role of oxidized low density lipo-protein in atherogenesis. J. Clin. Invest. 88:1785-1792.

6. Parthasarathy, S., M. T. Quinn, D.C. Schwenke, T. E. Carew, and D.

Steinberg. 1989. Oxidativemodification of beta-very lowdensity lipoprotein. Arteriosclerosis.9:398-404.

7. Morel, D. W.,P.E.DiCorleto,andG. M. Chisolm. 1984. Endothelial and smooth muscle cells alter lowdensitylipoproteininvitrobyfreeradical oxida-tion. Arteriosclerosis.4:357-364.

8.Cushing,S.D.,J.A.Berliner,A.J.Valente,M.C.Territo,M.Navab,F.

Parhami, R.Gerrity,C. J. Schwartz, andA.M.Fogelman. 1990.Minimally modified lowdensity lipoprotein inducesmonocytechemotacticprotein 1 in human endothelial cells and smooth muscle cells. Proc. Natl.Acad.Sci. USA. 87:5134-5138.

9.Liao,F., J. A.Berliner,M.Mehrabian,M.Navab,L. L.Demer,A.J.Lusis,

andA.M.Fogelman.1991.Minimallymodifiedlowdensitylipoprotein isbiologi-cally activeinvivo.J.Clin.Invest.87:2253-2257.

10. Fox, P. L., G. M.Chisolm,and P. E.DiCorleto. 1987.

Lipoprotein-me-diatedinhibitionofendothelial cellproduction ofplatelet-derived growth factor-like proteindepends onfreeradicallipidperoxidation.J.Biol. Chem. 262:6046-6054.

11.Nishigaki, I.,M.Hagihara,H.T.Tsunekawa,M.Maseki,and K.Yagi. 1981.Lipid peroxidelevelsofserumlipoprotein fractions of diabeticpatients. Biochem. Med.25:373-378.

12.Sato, Y., N. Hotta, N.Sakamoto,S. Matsuoka, N.Ohishi,and K.Yagi.

1979.Lipid peroxidelevel inplasmaofdiabeticpatients. Biochem. Med. 21:104-107.

13.Higuchi,Y.1982.Lipid peroxidesandalpha-tocopherolin rat

streptozo-tocin-induceddiabetesmellitus.ActaMed.Okayama. 3:165-175.

14.Pisanti, F. A., Frascatore, S., and G. Papaccio. 1983. Superoxide

dismu-taseactivityin the BBrat: adynamic timecoursestudy.LifeSci. 43:1624-1632. 15.Morel, D. W., and G. M. Chisolm. 1989. Antioxidant treatment of

dia-beticratsinhibits lipoprotein oxidationandcytotoxicity.J.LipidRes. 30:1827-1834.

16.McNally,A.K., G. M.Chisolm,D. W.Morel, and M. K. Cathcart. 1990. Activated human monocytes oxidize low-densitylipoproteinbya lipoxygenase-dependent pathway.J.Immunol. 145:254-259.

17.Matkovics, B.,I.Varga,I.Szabo, and H.Witas. 1982. The effect of dia-betesontheactivities of theperoxidemetabolism enzymes. Horm. Metab. Res.

14:77-79.

18.Loven, D. P., H. P. Schedl, L. W. Oberley, H. D. Wilson, L. Brunch, and C.N.Niehaus. 1982. Superoxide dismutaseactivityin the intestinal mucosa of

streptozotocin-diabeticrat.Endocrinology. 11 1:737-742.

19.Carnovale,C.E.,J. A.Monti,V. A.Catania,and M. C.Carrillo. 1990.

Possibleroleof blood insulin levelsonglutathioneS-transferase activities from different tissues of malerats.Can. J.Physiol.Pharmacol. 68:170-173.

20.Harman, D. 1969.Prolongationof life: role of free radical reactions in aging. J.Am.Geriatr. Soc.17:721-735.

21. Halliwell, B., and J. M.C.Gutteridge. 1986. Oxygen free radicals and iron in relation tobiologyandmedicine: some problems and concepts. Arch. Biochem. Biophys. 246:510-514.

22. Hunt, J. V., C. C. T.Smith, and S. P. Wolff. 1990. Autoxidative glycosyla-tion andpossibleinvolvement ofperoxides and free radicals in LDL modification byglucose. Diabetes. 39:1420-1424.

23.Feingold, R. K. 1989. Importance of small intestine in diabetic hypercho-lesterolemia. Diabetes. 38:141-145.

24. Olsen, W. A., and I. H. Rosenberg. 1970. Intestinal transport ofsugars and amino acids in the diabetic rats. J. Clin. Invest. 49:96-105.

25. Thomson, A. B. R., and R.Rajotte. 1983. Effect of dietary modification ontheuptake of glucose, fatty acids and alcohols in diabetic rats. Am. J. Clin. Nutr.38:394-403.

26. Staprans, I., and J. M. Felts. 1989. A possible mechanism for accelerated atherogenesis in male versusfemale rats.Arteriosclerosis.9:224-229.

27. Boeynaems, J. M., A. R. Brash, J. A.Oates,and W. C. Hubbard. 1979. Preparation and assay ofmonohydroxy-eicosatetranoic acids. Anal. Biochem. 104:259-267.

28. Wisneski, J. A., E. W. Gertz, R. A. Neese, and M. Mayr. 1987. Myocardial metabolism of free fatty acids. Studies with14Clabeled substrates in humans. J. Clin. Invest. 79:359-366.

29.Fletcher, M. J. 1968. Acolorimetricmethodforestimatingserum

triglycer-ides. Clin. Chim. Acta. 22:393-397.

30.Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275.

31.Ohishi, N., H. Ohkawa, A. Miike, T. Tatano, and K.Yagi.1985.Anew assaymethod for lipid peroxides using a methylene blue derivative. Biochem. Int.

10:205-211.

32. Morel, D. W., J. R. Hessler, and G. M. Chisolm. 1983. Lowdensity

lipoprotein cytotoxicity induced by free radicalperoxidation oflipid.J.Lipid

Res.24:1070-1076.

33. Andrews, J. S., W. H. Griffith, J. F. Mead, and R. A. Stein. 1960. Toxicity ofair-oxidized soybean oil. J. Nutr. 70:199-210.

34.Bergan, F. G., and H. H. Draper. 1970.Absorptionandmetabolismof

I-'4C-methyllinoleatehydroperoxide.Lipids.5:976-982.

35.Glavind, J. 1970. Intestinal absorption andlymphatictransport ofmethyl linoleatehydroperoxide and hydroxyoctadecadinoeate in the rat. Acta Dermato-Venereol. 24:3723-3728.

36.Aw, T. Y., M. W. Williams, and L. Gray. 1992. Absorption and lymphatic transportofperoxidized lipids by rat small intestine in vivo: role ofmucosal GSH. Am. J.Physiol. 262:G99-G106.

37.Staprans, I., X. M. Pan, M. Miller, and J. H. Rapp. 1993. Theeffectof oxidation on thechylomicron metabolism in rats. Am. J.Physiol. 264:G561-G568.

38. Feingold, K. R., M. H. Wiley, G. MacRae, G. A. Kaysen, P. Y. Schoen-feld, and M. D. Siperstein. 1982. The effect of diabetesmellitusonsterolsynthesis in thediabetic rat.Diabetes. 31:388-395.

39. Feingold, R. K., A. Moser, S. Adi, M. Soued, and C.Grunfeld. 1990. Smallintestinefattyacidsynthesisisincreased indiabeticrats.Endocrinology.

127:2247-2252.

40.Feingold, R. K., G. Zsigmond, M. Hughes-Fullford, S. R. Lear, and A. H. Moser. 1985. The effectof diabetes mellitus on the lymphatic transport of intes-tinal sterols.Metab. Clin.Exp.34:1105-1109.

41.Steiner, G. S., M. Poapst, and J. K.Davidson. 1975. Production of

chylo-micron-likelipoproteins from endogenous lipid by the intestine and liver of dia-betic dogs.Diabetes.24:263-271.

42.Risser, T. R., G. M. Reaven, and E. P. Reaven. 1978. Intestinal very low densitylipoprotein secretion in insulin-deficient rats. Diabetes. 27:902-908.

43. Popper, D. A., Y. Shiau, and M. Reed. 1985. Role of small intestine in pathogenesis ofhyperlipidemiaindiabetic rats. Am.J.Physiol. 249:G16 1-167. 44.Nervi, F. O., A. Gonzalez, and V. D. Valdivieso. 1974. Studies on choles-terolmetabolism in the diabetic rat. Metab. Clin. Exp. 23:495-503.

45.Thomson, A. B. R., and R. Rajotte. 1983. Effect of dietary modification onthe enhanced uptakeof cholesterol in diabetic rats. Am. J. Clin. Nutr. 37:244-252.