Journal of Clinicat InvestigationVol. 44, No. 2, 1965

Effect of Fatty Acid Saturation on the Distribution of theCholesterol Moiety of Very Low Density Lipoproteins *

NORTONSPRITZ t(From the Lipid Metabolism Laboratory of the Second [Cornell] Medical Division, Bellevue

Hospital, and the Department of Medicine, Cornell University, Medical College,New York, N. Y.)

The nonesterified cholesterol of each plasmalipoprotein fraction is in rapid exchange with thatof red cells, liver, and other circulating lipopro-teins (1-5). Although this dynamic equilibriumhas been demonstrated in many species, includingman, the factors determining the dynamics of thisexchange and the resultant distribution of labeledcholesterol have not been elucidated. The presentexperiments, designed to test the possibility thatthe degree of saturation of the esterified fatty acidsof lipoproteins could be such a factor, indicatethat isotopic cholesterol associated with very lowdensity lipoproteins composed largely of unsatu-rated fatty acids appears in other cholesterol poolsmore rapidly than that of saturated lipoproteinsduring both in vivo and in vitro equilibrations.

traperitoneally in a small volume of 50% ethanol. Theanimals were exsanguinated 12 to 14 hours later and thevery low density lipoproteins (< 1.019 fraction) har-vested from their sera. This fraction from a safflower-fed animal was combined with one from an animal fed co-conut oil and the mixture injected intravenously into arecipient rabbit. The fractions were selected so that ineach study the cholesterol of one donor was labeled withC" and the other with tritium. The recipients weresacrificed 30 to 210 minutes after injection, and the ratioof the two isotopes was determined in the free and esteri-fied fractions of their red cells, liver, whole serum, and< 1.019 and > 1.019 lipoprotein fractions. These isotoperatios were then compared to that injected. Changes inratios were considered to indicate differences in dynamicsof the cholesterol moiety of each type of lipoprotein.Each of these steps is discussed in more detail in thefollowing paragraphs.

Methods

Female New Zealand white rabbits weighing 1,800 to2,500 g were utilized. Diets consisted of standard rab-bit Purina chow to which 10% safflower or coconut oils 1

and 1% cholesterol 2 had been added. 4-C"- and 7a-H'-cholesterol3 were purified by preparative thin layer chro-matography before use. Thin layer chromatoplateswere prepared with binder-free silicic acid 4 and were 0.5mmin thickness (6).

I. In vivo studies

Summary of general experimental design (Figure 1).Three lipemic rabbits that had received diets containingcoconut oil and three that had received safflower oil weregiven 0.1 mc of tritiated or .03 mc of C"-cholesterol in-

* Submitted for publication June 15, 1964; acceptedNovember 5, 1964.

Supported by U. S. Public Health Service grant NB-03346-03.

t Established Investigator of the Health ResearchCouncil of New York (I-128).

1 E. F. Drew Co., Boonton, N. J.2 Nutritional Biochemicals Corp., Cleveland, Ohio.8 New England Nuclear Corp., Boston, Mass.4Silica gel H, Merck & Co., Rahway, N. J.

Safflower-feddonorrabbit

7-a H' cholesterol*

Very low densitylipoproteins

(<1.019)

I

Coconut-feddonorrabbit

4-Cl4-cholesterol*

Very low densitylipoproteins

(<1.019)

Injection mixture

Recipientanimal

Sacrificed30-210 min.

Isotope ratiosFC and CE

FIG. 1. OUTLINE OF THE EXPERIMENTALDESIGN. FC =nonesterified cholesterol; CE = cholesterol ester.

* In experiments 2 and 3 H' and C"4 isotopes werereversed.

339

NORTONSPRITZ

FIG. 2. FATTY ACID COMPOSITIONOF TRIGLYCERIDE AND CHOLESTEROLESTER FRACTION OF IN-JECTED< 1.019 FRACTION. The predominance of polyunsaturated fatty acids in both lipid classesobtained from the < 1.019 fraction of the safflower-oil-fed rabbits is illustrated. This effect ismore marked in the triglyceride than cholesterol ester fraction. Very similar patterns were

obtained in material from all donor animals. Phospholipid fatty acid composition similarly was

affected by dietary fat but less markedly so. The symbols for the fatty acids refer to the num-

ber of carbon atoms and the number of double bonds (11).

Preparation of donor lipoproteins. Sera obtained fromthree lipemic donor animals that haJ received safflowerand three that had received coconut oil were adjusted todensity of 1.019 with KBr and centrifuged at 105,000 g

for 16 to 20 hours at 5' C (7). The contents of the top1 cm of the tube were recovered and dialyzed for 48

hours at 5' C with nitrogen bubbling used for agitation.The dialysis fluid was a solution of 0.01 M phosphatebuffer at pH 7.4, 0.01% N-ethyl maleimide (NEM), and0.01 M Na ethylenediamine tetraacetate (EDTA) in0.15 N NaCl. NEMwas used to prevent in vitro trans-esterification of free cholesterol as described by Glomset

TABLE I

Characteristics of the two components of the injection mixtures in each experiment*

%of %radio- Cholesterol/ Cholesterol/cholesterol activity in Specific activ- triglyceride phospholipid

nonesterified FC fraction$ ity FC ratio ratioExp. Dura- Saff _ _-no. tion isotope Saff* Cocof Saff* Cocot Saff* Cocot Saff* Cocot Saff* Cocot

min - cPm/mg1 30 H' 29 35 78 72 117 94 26.0 12.0 2.4 1.82 40 C14 40 42 42 60 92 127 .9 1.7 1.1 1.53 60 C14 40 42 42 60 92 127 .9 1,7 1.1 1.54 70 H' 32 40 72 56 148 123 17.0 3.9 2.5 1.45 90 HI 29 40 78 55 117 123 26.0 3.9 2.4 1.46 180 Hs 29 35 78 72 117 94 26.0 12.0 2.4 1.87 210 Hs 32 40 72 55 148 123 17.0 3.9 2.5 1.4

<1.019 lipoprotein fraction of safflower-oil-fed donor rabbit.t <1.019 lipoprotein fraction of coconut-oil-fed donor rabbit.I FC = nonesterified cholesterol.

340

ISOTOPIC CHOLESTEROLDISTRIBUTION

(8) and EDTA to bind heavy metal oxidation catalysts(9). Figure 2 illustrates the marked difference in fattyacid composition of the cholesterol ester (CE) and tri-glyceride classes of a representative safflower- and coco-nut-oil-fed donor group. These data were obtained byan F & Mmodel 400 gas liquid chromatograph standard-ized with N.I.H. solutions D and E (10).

Before mixing, the lipid of each donor lipoprotein solu-tion was extracted and the CE, nonesterified cholesterol(FC), glyceride, and phospholipid fractions were iso-lated and quantitated by preparative thin layer chroma-tography (12, 13). Table I depicts data obtained fromboth components of each injection mixture in the sevenexperiments. Distribution of isotope between esterifiedand nonesterified cholesterol; the specific activity of theisotopic cholesterol; and cholesterol: phospholipid, cho-lesterol: triglyceride ratios varied widely among the do-nors. The values for each of these variables were higherin the lipoproteins from the safflower-fed donors in someexperiments and from the coconut-oil-fed donors inothers.

Administration of <1.019 lipoprotein fraction to recipi-ent animal. The recipient animals, which had been main-tained on fat-free diets for at least 4 days before the ex-periment, were lightly anesthetized with intravenous pen-tobarbital injection. The < 1.019 lipoprotein fractionsfrom the two donors were mixed immediately before ad-ministration and were injected into an ear vein over a2- to 3-minute period. The total volume injected variedfrom 15 to 30 ml and contained from 20 to 50 mg oflipid. Total radioactivity injected varied from approxi-mately 7,000 to 10,000 cpm of each isotope.

Method of determining isotope ratios in tissues. Therecipient animals were killed by exsanguination from theinferior vena cava 30 to 210 minutes after injection.Fifty ml of blood was clotted and the serum obtained.In three experiments 10 ml of serum was adjusted todensity of 1.019 and a < 1.019 and > 1.019 lipoprotein frac-tion obtained as described for the donor sera. Red cellswere obtained from approximately 20 ml of blood drawnin a heparinized syringe. These were washed three timesin 0.15 M NaCl. The liver was freed from the gall-bladder and was cut into numerous small pieces andrepeatedly washed with 0.15 M NaCl until the washingswere clear. It was then homogenized in a Waringblendor. Each of these materials was extracted intochloroform: methanol, 2: 1, and the FC and CE fractionswere obtained by preparative thin layer chromatography.In one experiment, bile, aorta, kidney, and heart werealso obtained, but these contained insufficient radioactivityfor accurate counting and were not extracted in subse-quent experiments.

Determination of isotope ratios. A Packard dual chan-nel Tri-Carb scintillation counter was used. Scintillatorsolution was 5% 2,5-diphenyloxazole and 0.8% 1,4-bis-2-(4-methyl-5-phenyloxazolyl)benzene in toluene. In allinstances the activity of both isotopes was at least threetimes that of its background, and counting was continuedat least long enough for the lower counting isotope to be

4.1

r.

go

*oan

0o

-

UI

uPoW4

.4+0

A

0v0

V

U

pq

0'

0

A

0'

0

V.v

SX

l-

o40oY-

-- v

0040O%

00

_ux10m to -.1rr

V.q 4 el i

4 0U,)) -4

0 0 00

06 V.;_l

('-4m~

-0 _143

.d' -.4 e4 0%N M -

VO --4 --4 C'4 - -4

o6 ~ ci -

-4- C

C4 1 U)14 I

341

u

0u0

0

0a

0

-o4.J

u0

0

0r

0

4-

0

0-o

4)

8

0

4)j

0

a4-)

co

4-J

Cd

0

0Co4)

o cU -4)e. 4.Q~5=

a4)

-

._ 08 i

X_-5

4U

NORTONSPRITZ

0

*30-90 mino 180-210 min

0

l l lSerum <1.019 >1.019 Liver

2.60j7

I1.401-

1.301-

1.20

0

* 30-90 mino 180-210 min

I

1.101-

1.00

0.90

_* 0

0

0

0 0

I I I I I

Injected <1.019 >1.019 Liver RBC

FIG. 3. A) COMPARISONOF SAFF/COCOOF RECIPIENT TISSUES TO THAT IN-JECTED. The saff/coco of the injection mixtures has been adjusted to 1.00,and comparative values for this ratio in recipient serum, < 1.019 and > 1.019-lipoprotein fractions, red cells, and liver FC (nonesterified cholesterol) are

depicted. B) COMPARISONOF INJECTION MIXTURE ANDRECIPIENT TISSUES TO

342

2.16{

1.30j-

1.201-

1.10 -

00

C-4-

co(A

.00

0.90 -

0.80[-D

0.701RBC

0

coUX

L.a1)0

0

ISOTOPIC CHOLESTEROLDISTRIBUTION

counted with an accuracy of 1%o. Isotope ratios, even inthose specimens with the least activity, were reproduciblewith an error of 2% when separate portions of the tis-sue to be studied were processed in duplicate.

After sample counting was completed, an H' and a C1'internal standard was counted in every specimen. Thiswas done to detect quenching for each isotope as wellas to determine the ratio of C' counts between the twochannels so that correction of total counts in the H' chan-nel for C' could be individualized for each specimen.

II. In vitro studiesThe sera of two rabbits that had received coconut-oil

feedings and two that had received safflower oil were ob-tained as described above, except that isotopic cholesterolwas not administered to the animals. The sera from thesafflower-fed animal were incubated with 4-C' cholesterol,and those from the coconut-fed animals with 7a-HI cho-lesterol as described by Porte and Havel (14). The-< 1.019 fraction was isolated as described above andpassed through an 8.0-A Swinney filter.

A C14- and HW-labeled lipoprotein fraction was mixedand added to a solution of unlabeled washed human redcells. A sample of the lipoprotein fraction was removedat the beginning of the incubation. At 1-, 2-, and 4- or 1-,2-, and 3-hour intervals, a sample of the incubation mix-ture was removed, and the lipoprotein and red cells wereseparated as described above. The C"/H' ratio was thendetermined separately for the lipoproteins and red cellsat intervals up to 4 hours.

ResultsI. In vivo studies

Isotope ratios of free cholesterol. HS/C14 orC14/H' ratio was determined for the donor in-jection mixture and for the tissues obtained fromthe recipient animals at the end of each experiment.Table II lists the ratio of the isotopic cholesterolfrom the safflower-fed donor to that from thecoconut-fed. This ratio is designated as "saff/coco." In experiments 1, 4, 5, 6, and 7 it wasH3/C14, and in the other two it was C14/H3. Fig-ure 3A indicates the ratio of the recipient tissues'saff/coco to that injected, and Figure 3B illustratesthe relationship between the whole serum saff/coco to that injected and the recipient tissues.

In each of the experiments lasting from 30 to90 minutes, the loss from the serum of the saff-fed

donor's isotopic FC was more rapid than that ofthe coco-fed. This selective loss was indicated bylower saff/coco ratios in the sera of recipients thanin the injected mixtures in all five experiments(Figure 3A). This finding was most marked inexperiment 1 in which the serum ratio was 74%of that injected. In experiments 2 and 3, the ob-served fall in serum FC saff/coco reflected that ofits very low density (< 1.019) lipoprotein frac-tions. Although the whole serum ratio had fallenonly to 86 and 93%o of the injected, that of the< 1.019 lipoprotein fraction in these two studieswas 82 and 81%o, respectively. Conversely, liverFC saff/coco exceeded that injected, indicatingselective accumulation of the isotopic cholesterolfrom the safflower donor lipoprotein in that organ.Red blood cell FC saff/coco exceeded that in-jected in four of the five experiments lasting 30to 90 minutes.

Figure 3B more strikingly illustrates this selec-tive distribution of isotopic FC in the 30- to 90-minute experiment. In every instance the > 1.019lipoprotein, liver, and red cell FC saff/cocoexceeded that of whole serum obtained simultane-ously. The saff/coco of the < 1.019 fraction, thematerial injected, was even lower than that of thewhole serum, further indicating selective loss ofisotopic cholesterol from the donor lipoprotein andits selective accumulation in recipient tissues.

In experiment 6, lasting 180 minutes, the FCsaff/coco of the < 1.019 fraction in the recipientrabbit equaled that injected, whereas the higherdensity lipoproteins (> 1.019) continued to showselective accumulation of the isotopic cholesterolfrom the saff-fed donor. In the 210-minute ex-periment (no. 7) selective distribution was nolonger evident; whole serum, liver red cell, and in-jected FC saff/coco now being equal.

Esterified cholesterol ratios. It is evident fromTable II that considerable change from the in-jected saff/coco ratio occurred in the plasma andliver CE. Unlike the FC, however, no consistentpattern for this change in plasma could be estab-lished; the ratio in plasma exceeded the ratio in

SERUMSAFF/COco. The saff/coco of the recipient serum obtained at the endof each experiment has been adjusted to 1.00, and comparative values forthis ratio in the donor mixture and the recipient < 1.019 and > 1.019 lipo-proteins, red cells, and liver FC are depicted. "Saff/coco" refers to the ra-tio of the isotopic cholesterol from the safflower-fed donor to that from thecoconut-fed.

343

NORTONSPRITZ

.25 'I

FIG. 4. COMPARISONOF INJECTED AND LIVER CE SAFF/COCO TO INJECTED FC SAFF/coco. The saff/coco of theinjected FC has been adjusted to 1.00 in all instances.Comparison of the injected CE (cholesterol ester) to in-jected FC saff/coco appears on the left vertical line andcomparison between liver CE saff/coco and injected FCon the right. The diagonal lines connect values obtainedin a single experiment. In all instances, the liver CEsaff/coco reached a value intermediate between the in-jected CE and FC.

the injected lipoprotein in three experiments, was

equal in one, and less in the other three.Similarly liver CE saff/coco varied widely from

that injected and that of plasma obtained at theend of the experimental period. A constant inter-relationship among the saff/coco of the injectedCE, liver CE, and the injected FCwas noted, how-ever, and is illustrated in Figure 4. In all in-stances, the liver CE saff/coco was intermediatebetween the ratio for injected CE and FC. Thisrelationship was noted whether the injected FCsaff/coco was greater or less than that of the CEand suggests that hepatic CE in these studies re-

ceived contributions both directly from circulatingCE and from esterification of circulating FC.

Per cent of total radioactivity in FC and CEfractions. Table III depicts mean values andranges for the per cent of isotope in the CE frac-

TABLE III

Mean and range of per cent of isotope in cholesterol esterfraction of injected material, serum, and liver

TotalSaff isotope Coco isotope isotope

Injected 34 38 36lipoprotein (22 to 58) (28 to 45)mixture

Recipient 65 64 65serum (32 to 84) (54 to 84)

Recipient 15 17 16liver (3 to 21) (4 to 30)

tion of the injection mixture, serum, and liverfor each isotope and the total. The per cent esteri-fied was higher in serum than that injected in allexperiments, and in the liver the per cent esterifiedwas considerably below both that injected and theserum.

II. In vitro experiments

Table IV compares the C14/H3 ratios in the< 1.019 fraction and red cells after 1 to 4 hoursof incubation at 37° C. In both experiments theisotopic cholesterol from the safflower-fed animal'slipoproteins (C14) was higher in the red cells andreciprocally lower in the lipoproteins than in theoriginal mixture (adjusted to 1.00).

TABLE IV

Incubation of mixture of in vitro-labeled lipoproteinswith unlabeled red cells

C'4/H' ratio

RBC Lipoprotein mixtureMixture

Exp. of saff Hours Hoursno. and coco 1 2 3 4 1 2 3 4

1 1.00 12.3 3.8 3.4 0.51 0.77 0.642 1.00 4.1 4.5 2.0 0.84 0.83 0.89

Discussion

The experiments described in this study weredesigned to permit the determination of the effectof fatty acid composition on the distribution oflipoprotein cholesterol. The use of double isotopetechnique made it possible to compare the fate ofinjected isotopic cholesterol associated with highlyunsaturated to saturated lipoproteins within asingle animal. These experiments indicate thatisotopic free cholesterol associated with unsatu-

344

ISOTOPIC CHOLESTEROLDISTRIBUTION

rated very low density lipoproteins leaves theselipoproteins and appears in red cells, liver, andlower density lipoproteins more rapidly than thatof the saturated lipoproteins.

These findings could result from any or a combi-nation of the following events: 1) Exchange ofisotopic cholesterol between the injected lipo-proteins and the recipient tissue was withoutchange in the cholesterol content of either. Inthis case, both the transfer of cholesterol from theinjected lipoproteins as well as its replacement bynonlabeled recipient cholesterol would have pro-ceeded more rapidly for the safflower lipoproteinsthan for the lipoproteins from coconut-fed animals.2) Net transfer of FC from the injected unsatu-rated lipoproteins to recipient tissues was morerapid than from the saturated lipoproteins. 3)Distribution of intact injected lipoproteins withtheir component isotopic FC molecules at selectiverates was influenced by fatty acid composition.

The relative role of each of these mechanismscannot be determined from this study. Exchangeof FC molecules is the most likely explanation forthe appearance of radioactivity in the > 1.019lipoprotein in vivo and for the red cells in both thein vivo and in vitro studies, although net transferhas not been excluded.

On the other hand, previously reported studiesin which the glyceride portion of chylomicra wasisotopically labeled (15-17) suggest that the iso-tope appearing in the liver in the present experi-ment may represent, to some extent, cholesterolstill a component of the lipoprotein with which itwas injected. Several findings in the present ex-periments indicate, however, that isotopic liver FCarose primarily from exchange or net transfer ofplasma FC rather than from the transport of in-tact labeled lipoproteins from plasma to liver. Inall instances, the per cent of esterified isotope inthe liver was far lower than that injected or thatof the serum (Table III). This finding, togetherwith an increase in the per cent of isotope esteri-fied in the serum over that injected, is consistentwith transfer of isotopic circulating FC (either asexchange or net transfer) while labeled CE re-mains a part of the plasma lipoprotein. Further-

5 A decrease in the degree of esterification of isotopiccholesterol in liver compared to that in plasma has alsobeen observed by Brot, Lossow, and Chaikoff (18) inin vitro experiments in rats. They interpreted this to

more, Nestel and Scow (16) demonstrated thatwhen the chylomicron glyceride was isotopicallylabeled, the isotope from those chylomicra thatwere highly saturated disappeared from plasmaand appeared in liver more rapidly than that fromchylomicra with relatively unsaturated fatty acids.This finding, in association with those of the pres-ent experiments, lends further support to the con-cept that the FC constituent of lipoproteins ex-changes with or is transferred to tissues inde-pendently of other lipid components (2, 4) andthat fatty acid composition may affect the dy-namics of this fraction differently from that of thewhole lipoprotein or its glyceride.

In the experiments of 180 and 210 minutes inthe present study, the selectivity of the distributionof the injected double-labeled FC was largelylost. The most likely explanation for this phe-nomenon is that cholesterol had been redistributedafter its initial exchange. Once the isotopic mole-cules had left the lipoproteins with which they wereinjected, subsequent exchanges from the tissue towhich they had migrated would no longer be se-lective, and a return of the ratios between isotopeto that injected would be expected. The occur-rence of a second exchange of cholesterol mole-cules in the period of time of these experiments isconsistent with the known rapid rates of free cho-lesterol exchange among lipoproteins (4, 5), liver,and red cells (1-3).

The role of a second exchange in these findingsis further supported by the findings of Naidoo,Lossow, and Chaikoff (19) that isotopic chylomi-cron cholesterol injected into rats and removed pri-marily into liver begins to reappear in plasma 50minutes after injection. In experiment 6, thepersistence of a saff/coco ratio greater than thatinjected in the recipient's > 1.019 lipoproteinfraction when liver, red cells, and < 1.019 frac-tion had all returned to the injected ratio also sup-ports the contention of these authors that choles-terol transferred from lower to higher density lipo-proteins then has a turnover rate slower than thatremaining in the lower density lipoproteins.

indicate hydrolysis of circulating CE as isotopic cho-lesterol entered hepatic tissue. This interpretation doesnot explain the findings of the present in vivo study inview of the similarity of hepatic, serum, and injected FCsaff/coco and the marked dissimilarity between these ra-tios and those of the CE fractions (Table II).

345

NORTONSPRITZ

Although the saff/coco ratio of the plasma, redcell, and liver FC fractions showed a consistentpattern, that of the plasma CE fraction did not.Plasma CE may arise either from hepatic CE orby direct esterification of circulating nonesterifiedcholesterol (8). Furthermore, the turnover ofindividual cholesterol esters may vary with thefatty acid to which it is esterified (20). In theexperiments reported in this study the relative con-centrations of the two isotopic forms of cholesterolwere influenced, in addition to these factors, bythe relationship between the amount of each iso-tope free and esterified in the injection mixturesand the degree to which the much more rapidflux of the free cholesterol had been selective. It isnot surprising in view of this complex of factorsinfluencing isotopic CE ratios in these animalsthat diverse results were obtained.

Several factors indicate, however, that the se-lective changes in FC saff/coco observed in thestudies of 30 to 90 minutes were not due to iso-tope originally in the CE fraction of the injectedmaterial: 1) In experiments 1, 2, and 3 the saff/coco of the injected CE exceeded that of the FC,whereas in 4 and 5 it was lower. Yet in all in-stances, the same relationship between donorand recipient FC fractions was noted at the endof study. 2) As noted above, at 180 and 210 min-utes, re-equilibration of the FC fraction occurredresulting in close agreement between the injectedsaff/coco and that of whole serum (no. 7) or the< 1.019 fraction (no. 6) obtained at the end of theexperiments. Had CE isotope made a significantcontribution to FC saff/coco in these experiments,one would expect variation from this original FCratio, since in one experiment CE saff/coco ex-ceeded and in the other was less than that of theFC. 3) The in vitro studies indicate selectivedynamics of the lipoprotein FC independent ofCE, since in vitro labeling was restricted to thenonesterified cholesterol.

The present study may provide an insight intothe mechanism by which the ingestion of un-saturated dietary fat produces an increase in tissuecholesterol content (21-23). In at least one studyin which this hypocholesterolemic effect was ob-served in rats (24), a fall in plasma concentra-tion occurred as tissue content rose. These find-ings suggest that the hypocholesterolemic effectof unsaturated dietary fat could result from an al-

teration in lipoprotein structure such that theequilibrium between plasma and tissue cholesterolpools is altered to favor the latter. Differences inthe dynamics of free cholesterol between lipopro-teins with different fatty acid composition demon-strated in the present study support this postula-tion if the observed selective shifts of isotoperepresent net transfer of cholesterol from the in-jected to recipient tissues.

The evidence for this proposition is further lim-ited, however, by the extent to which the resultsof the present and other studies utilizing animalsare relevant to the mechanism by which humancirculating ,8-lipoprotein cholesterol is lowered byunsaturated dietary fat. First, the hypocholes-terolemic effect regularly produced in man (25,26) is frequently not seen in animals. In thepresent study and others previously reported (21,22) higher plasma cholesterol values were oftenfound in animals receiving unsaturated fat. Sec-ondly, both thoracic duct chylomicra and the lipo-proteins characteristic of the induced lipemias inanimals usually studied in these experiments arequite different from those of human 8-lipoproteins.In the present study, the very high cholesterolcontent relative to that of other lipid classes (TableI) is peculiar to the < 1.019 lipoprotein fractionof rabbits made lipemic by a high fat, cholesterol-containing diet. It is important to point out, how-ever, that in spite of a wide range of cholesterol/triglyceride and cholesterol/phospholipid ratios inthe injected materials, selective distribution wasregularly observed.

Summary

An experiment devised to test the effect of fattyacid composition on the dynamics of very lowdensity lipoprotein free cholesterol is described.The < 1.019 lipoprotein fraction of rabbits madelipemic with safflower or coconut oil was labeledwith either 4-C14- or 7a-H8-cholesterol. This frac-tion from a safflower-fed rabbit was mixed withone from a rabbit fed coconut oil; each was la-beled with a different isotope and the mixture in-jected into a recipient animal. The free choles-terol isotope ratio in the whole serum, < 1.019and > 1.019 lipoprotein fractions, red cells, andliver-free cholesterol was compared to that in-jected. The disappearance of isotopic cholesterol

346

ISOTOPIC CHOLESTEROLDISTRIBUTION

from < 1.019 fractions obtained from safflower-fed donors and its appearance in the recipient's redcells, liver, and > 1.019 lipoprotein fraction wasmore rapid than that from coconut-fed donors.Similarly, in in vitro studies, exchange of freecholesterol between lipoproteins and red cells wasmore rapid for the safflower-fed animals.

Acknowledgments

The author is indebted to Mrs. G. Rabkin, Mrs. V.Leonard, and Mr. D. Ganz for their excellent technicalassistance.

References1. Hagerman, J. S., and R. G. Gould. The in vitro

interchange of cholesterol between plasma andred cells. Proc. Soc. exp. Biol. (N. Y.) 1951, 78,329.

2. Lossow, W. J., N. Brot, and I. L. Chaikoff. Distri-bution of the cholesterol moiety of a chylomicron-containing lipoprotein fraction of chyle in the rat.J. Lipid Res. 1962, 3, 207.

3. Eckles, N. E., C. B. Taylor, D. J. Campbell, andR. G. Gould. The origin of plasma cholesteroland the rates of equilibration of liver, plasma, anderythrocyte cholesterol. J. Lab. clin. Med. 1955,46, 359.

4. Minari, O., and D. B. Zilversmit. Behavior of doglymph chylomicron lipid constituent during incu-bation with serum. J. Lipid Res. 1963, 4, 424.

5. Roheim, P. S., D. E. Haft, L. I. Gidez, A. White,and H. A. Eder. Plasma lipoprotein metabolismin perfused rat livers. II. Transfer of free andesterified cholesterol into the plasma. J. clin.Invest. 1963, 42, 1277.

6. Mangold, H. K. Thin layer chromatography oflipids. J. Amer. Oil Chem. Soc. 1961, 38, 708.

7. Havel, R J., H. A. Eder, and J. H. Bragdon. Thedistribution and chemical composition of ultra-centrifugally separated lipoproteins in human serum.J. clin. Invest. 1953, 32, 428.

8. Glomset, J. A. Further studies of the mechanism ofthe plasma cholesterol esterification reaction. Bio-chim. biophys. Acta (Amst.) 1963, 70, 389.

9. Nelson, G. J. Studies on human serum lipoproteinphospholipids and phospholipid fatty acid com-position by silicic acid chromatography. J. LipidRes. 1963, 3, 71.

10. Horning, E. C., E. H. Ahrens, Jr., S. R. Lipsky,F. H. Mattson, J. F. Mead, D. A. Turner, andW. H. Goldwater. Quantitative analysis of fattyacids by gas-liquid chromatography. J. Lipid Res.1964, 5, 20.

11. Farquhar, J. W., W. Insull, Jr., P. Rosen, W.Stoelfel, and E. H. Ahrens, Jr. The analysis of

fatty acid mixtures by gas-liquid chromatography.Nutr. Rev. (suppl.) 1959, 17, 1.

12. Goldrick, B., and J. Hirsch. A technique for quan-titative recovery of lipids from chromatoplates.J. Lipid Res. 1963, 4, 482.

13. Amenta, J. S. A rapid chemical method for thequantification of lipids separated by thin-layerchromatography. J. Lipid Res. 1964, 5, 270.

14. Porte, D., Jr., and R. J. Havel. The use of choles-terol-4-C"-labeled lipoproteins as a tracer forplasma cholesterol in the dog. J. Lipid Res. 1964,2, 357.

15. Havel, R. J., and D. S. Frederickson. The metabo-lism of chylomicra. I. The removal of palmiticacid-i-C" labeled chylomicra from dog plasma.J. clin. Invest. 1956, 35, 1025.

16. Nestel, P. J., and R. 0. Scow. Metabolism of chylo-microns of differing triglyceride composition. J.Lipid Res. 1964, 5, 46.

17. Nestel, P. J., R. J. Havel, and A. Bezman. Metabo-lism of constituent lipids of dog chylomicrons. J.clin. Invest. 1963, 42, 1313.

18. Brot, N., W. J. Lossow, and I. L. Chaikoff. In vitrouptake and hydrolysis, by rat tissues, of cholesterolesters of a very low density, chyle lipoproteinfraction. J. Lipid Res. 1964, 5, 63.

19. Naidoo, S. S., W. J. Lossow, and I. L. Chaikoff.Disappearance of the cholesterol moiety of an in-jected chylomicron-containing fraction of chylefrom the circulation of the rat. J. Lipid Res. 1962,3, 309.

20. Goodman, D. S., and T. Shiratori. In vivo turnoverof different cholesterol esters in rat liver andplasma. J. Lipid Res. 1964, 5, 578.

21. Avigan, J., and D. Steinberg. Effects of saturatedand unsaturated fat on cholesterol metabolism inthe rat. Proc. Soc. exp. Biol. (N. Y.) 1958, 97,814.

22. Gerson, T., F. B. Shorland, and Y. Adams. The ef-fects of corn oil on the amounts of cholesterol andthe excretion of sterol in the rat. Biochem. J. 1961,81, 584.

23. Friedman, M., and S. 0. Byers. Effects of saturatedand unsaturated fat feeding on experimental throm-boatherosclerosis. Amer. J. Physiol. 1962, 203,626.

24. Bieberdort, F. A., and J. D. Wilson. Influence ofunsaturated fat on cholesterol-C" metabolism inthe isotopic study state in the rabbit (abstract).Clin. Res. 1964, 12, 264.

25. Ahrens, E. H., Jr., D. H. Blankenhorn, and T. T.Tsaltas. Effect on human serum lipids of substi-tuting plant for animal fat in diet. Proc. Soc.exp. Biol. (N. Y.) 1954, 86, 872.

26. Kinsell, L. W., J. Partridge, L. Boling, S. Margen,and G. Michaels. Dietary modification of serumcholesterol and phospholipid levels. J. clin. En-docr. 1952, 12, 909.

347