Response of Rhesus Serum High Density …atvb.ahajournals.org/content/atvbaha/4/2/154.full.pdfResponse of Rhesus Serum High Density Lipoproteins to Cycles of Diet-Induced Hypercholesterolemia

Response of Rhesus Serum High DensityLipoproteins to Cycles of Diet-Induced

HypercholesterolemiaGunther M. Fless, Dawn Juhn, Joan Karlin,

Arthur Rubenstein, and Angelo M. Scanu

Two male rhesus monkeys underwent cyclical feeding of a hypercholesterolemlcdiet (2% cholesterol, 25% coconut oil) and a low-fat Purina monkey chow diet. Duringthe latter diet, high density lipoprotein (HDL) exhibited two components with peakdensities of d = 1.081 g/ml and 1.109 g/ml named HDLL and HDLH, respectively. Duringthe Initial hypercholesterolemic stage, except for apo A-ll which remained unchanged,there was a transient rise In HDL (mainly HDLJ as well as in HDL cholesterol and apoA-l, all reaching maximal values after about 2 weeks from the onset of the diet. The twoHDL species changed neither In size nor density as compared to their baseline coun-terparts, but had a comparatively higher content In cholesteryl ester and lesseramounts of triglycerides and phospholipids as compared to the normocholester-olemic animal. With the development of overt hypercholesterolemia (plasma choles-terol levels above 400 mg/dl), both HDL particles increased in density due to the lossof surface components (phospholipids and unesterlfled cholesterol) and core triglyc-erides with only minor changes in protein and cholesteryl ester contents. At thisstage, the same two animals exhibited significant changes in the size and buoyantdensity of LDL. When returned to a normal Purina chow diet, the animals' serumcholesterol levels declined rapidly to normal levels; normalization of the HDL distribu-tion also occurred but at a comparatively later time (26 weeks).

Our studies Indicate that the two HDL subsets characteristic of the normocholester-olemic rhesus monkey undergo significant changes in buoyant density as a functionof the stage of hypercholesterolemia and that changes In concentration and sizemainly affect the HDLL subspecies. At levels of plasma cholesterol below 400 mg/dl,this cholesterol Increment Is reflected by a significant increase in the number of theHDL subspecies without the overt participation of the low density lipoprotein classescharacteristic of the advanced hyperlipldemlc stage. Since we previously reportedthat greatly increased levels of cholesteryl esters enriched low density lipoproteins,(J-VLDL (very low density lipoprotein) and pre-B-VLDL during overt diet-Induced hy-percholesterolemia, it is apparent that cholesterol is distributed differently amonglipoprotein particles containing either apo A-l, apo B, or apo E depending on itsconcentration In plasma. (Arteriosclerosis 4:154-164, March/April 1984)

Rhesus monkeys fed high cholesterol diets devel-op altered concentrations of plasma lipopro-

teins. Since there may be a causal relationship be-tween the level of serum low density lipoprotein(LDL) and atherosclerosis, most attention has beenfocused on LDL. However, the concentration and

From the Departments of Medicine and Biochemistry, Universi-ty of Chicago, Chicago, Illinois.

This work was supported by NIH Grant Number HL 15062.Address for reprints: Dr. Gunther M. Fless, Department of Medi-

cine, Box 81, University of Chicago, 950 East 59th Street, Chica-go, IL 60637.

Received July 7,1983; revision accepted November 18, 1983.

density distribution of HDL species are also affectedby cholesterol-containing diets. Thus, it was ob-served that high density lipoprotein (HDL) levelswere temporarily increased in rhesus and sootymangabey monkeys on a high cholesterol diet butfell below baseline values with continued feeding.12

This absolute decrease in concentration of HDL wasmore pronounced for HDL, than for HDL3.23 Hypo-responding rhesus monkeys with relatively low plas-ma cholesterol levels had increased HDL cholester-ol, whereas hyper-responding monkeys with highplasma cholesterol levels had lower HDL cholesterolvalues.3 When squirrel monkeys were fed increas-ingly large amounts of cholesterol (0.05 mg/kcal-2.0

154

by guest on May 14, 2018

http://atvb.ahajournals.org/D

ownloaded from

http://atvb.ahajournals.org/

DIETARY ALTERATION OF HDL OVER TIME Fless et al. 155

mg/kcal), both HDL, and HDL, increased in concen-tration; but at a cholesterol intake of 2.0 mg/kcal,HDL, decreased and HDLj increased in concentra-tion so that the ratio of HDLj, to HDL, changed from3.0 to O.2.4 However, in most studies where the con-centrations of HDL, and HDLg were determined, theclassical density cut of 1.125 g/ml for separatingHDL, from HDLj was used although there is no guar-antee that the density distribution of monkey HDL issimilar to its human counterpart. We report here theresults of a longitudinal study concerning the effectsof a high cholesterol, high fat diet on HDL structureand density distribution in two rhesus monkeys. Areport on dietary alteration of LDL over time in thesame animals has appeared previously.5

Methods

Two male rhesus monkeys were maintained eitheron regular Purina primate chow or a modified low fatPurina primate chow supplemented with 25% coco-nut oil and 2% cholesterol.5 Blood was collected frortithe femoral vein of monkeys anesthetized with keta-mine and the serum was separated by centrifugingthe blood at 4°C for 30 minutes at 1000 g.

Llpoproteln Preparation

HDL was isolated by a combination of rate-zonaland isopycnic equilibrium density gradient ultracen-trifugation. Total lipoproteins were floated by adjust-ing serum to d = 1.21 g/ml with solid NaBr andcentrifuging 20 hours in the Ti-60 rotor (Beckman,Palo Alto, California) at 59,000 rpm. All solutionscontained 0.01% disodium ethylenediaminetetra-acetate (Na;, EDTA) and 0.01% NaN3 and were ad-justed to pH 7.0. The background density of the iso-lated total lipoproteins was raised to 1.4 g/ml byadding more NaBr (0.29 g/ml lipoprotein solution),and the solution was layered under a linear 7.5% to30% NaBr gradient in a SW-40 tube. Sample vol-umes were usually 2 ml or less. HDL was separatedfrom LDL at 20°C by spinning the SW 40 rotor ateither 20,000 rpm for 16 hours or at 35,000 rpm for 4hours. Density gradient centrifugation of HDL wasthen carried out in a 7.5% to 20% NaBr gradient inthe SW 40 rotor at 39,000 rpm, at 20°C for 48 hours,at which time isopycnic equilibrium was reached.

Single-Step Density Gradient Ultracentrlfugatlon

The lipoprotein profile of mesus serum was ob-tained by "single-step" density gradient ultracentrifu-gation according to the slightly modified procedure ofForeman et al.6 The discontinuous gradient was pre-pared by weighing 0.5 g of sucrose into an emptySW-40 tube, then layering in sequence 5 ml 4 MNaCI, 0.5 ml serum, and 0.46 M NaCI to the top of thetube. Centrifugation was carried out at 39,000 rpmfor 66 hours at 20°C at which time isopycnic equilibri-um was reached. The tubes were pumped out at a

rate of 1 ml/min through an ISCO UA-5 monitor (In-strumentation Specialties Co., Lincoln, Nebraska)set at 280 nm. Densities of fractions from a controlgradient were determined with a Precision DensityMeter, DMA-02 (Anton Paar, Graz, Austria) as pre-viously described.7 Densities of HDL determinedwith this method are apparently less dense than thetrue buoyant density as measured in the analyticalultracentrifuge probably because of preferentialbinding of sucrose to HDL. The areas for HDI_L andHDLH were determined digitally using an Apple IImicrocomputer by dividing the HDL species at thedensity representing the local minimum between thepeak densities of HDLL and

Analytical Ultracentrlfugatlon

The molecular weights and the buoyant densities(pb = 1/v) of HDL fractions were determined simulta-neously by high speed sedimentation and flotation atthree different densities at 20°C in a Model E ultra-centrifuge (Beckman, Palo Alto, California) equippedwith a photoelectric scanner as previously de-scribed.58 Before analysis, HDL samples were dia-lyzed against three NaBr solutions of varying densityand concentration (1.5%, 20%, and 30%) each con-taining 0.01 % Naj EDTA (pH 7.0). The partial specif-ic volume and therefore the molecular weight of HDLdetermined in sodium bromide is essentially anhy-drous, because Patsch et al.9 have shown that vmeasurements obtained in that salt are identical tothose executed with a Mettler-Paar precision densi-meter in conjunction with dry weight concentrationmeasurements.

Chemical Analysis

Lipoprotein composition was determined bymeasuring protein,10 phospholipid,11 cholesterol,both free and esterified,12'13 and triglyceride14 as pre-viously described.5 HDL cholesterol was quantitatedafter the precipitation of the apo B containing lipopro-teins from 1 ml serum with 40 IJL\ heparin (5 x 106

units/I) and 50 ^l 2 M MnCI2.15 The cholesterol con-

centration of the supernatant was determined withthe Liebermann-Burchard reaction using the AutoAnalyzer II.16 Serum cholesterol was determined onthe Auto Analyzer II.16

Immunoassays

Rhesus serum apo A-l and apo A-ll were meas-ured in a double antibody radioimmunoassay usingantirhesus HDL antibodies and purified iodinatedrhesus apo A-l and apo A-ll. The method was basedon published procedures for human apo A-l'7 andhuman apo A-ll.18 Apo A-l and apo A-ll were purifiedfrom rhesus HDL as described previously.19 Apo Ewas determined by electroimmunoassay20 usingspecific antisera raised in the goat against rhesusapo E which was prepared from rhesus VLDL asdescribed previously.21



ownloaded from


156 ARTERIOSCLEROSIS VOL 4, No 2, MARCH/APRIL 1984

Electrophoresis

The method of Weber and Osborn22 was used forsodium dodecyl sulfate gel electrophoresis with 10%polyacrylamide gels (SDS-PAGE).

Results

HDL Density Gradient Profiles

Initially, when the animals were on the control diet,HDL consisted of two subspecies, one having a low-er density (HDLJ than the other (HDLJ. Both werelocated mainly in the HDLj density interval (Figure1). No detectable peak, but only a trailing shoulderon the HDLH could be observed in the density regionusually ascribed to human HDLg. The feeding of ahigh cholesterol-coconut oil-supplemented diet totwo male rhesus monkeys resulted in progressivealterations of their HDL density gradient profiles.With the rise in serum cholesterol, there was an im-

mediate change in the HDL profile caused by anincreased concentration of HDL^. This increase wasapparent in both monkeys after 3 to 4 days whenserum cholesterol levels had risen to between 220and 240 mg/dl. Maximal values of HDL^ werereached in 10 to 11 days, at which time serum cho-lesterol in Monkey 6 was 410 mg/dl and in Monkey24 was 332 mg/dl. During this time interval, the ratioof HDLL to HDI_H changed from 0.25 to 0.39 in Mon-key 6 and from 0.58 to 0.91 in Monkey 24 and repre-sented a 56% increase for both. This ratio was deter-mined by integration of the respective areas underthe bimodal HDL peak. However, the increase wasnot permanent and returned to the basal values inboth animals in 3 to 4 weeks while they were still onthe high fat diet.

The increased concentration of light HDLL was notcaused by the formation of an HDI^-like particle be-cause electroimmunoassay of rhesus serum (Mon-key 24) indicated that the concentration of apo E at

8

UJ

o<CQ

enCD<

102 L03 L05 LO9 I.I6 I.O2 L03 LO5

BUOYANT DENSITY, g/ml

109 U6

Figure 1 . Alteration in the HDL density gradient profile of Monkeys 6 and 24 due tothe feeding of monkey chow supplemented with 2% cholesterol and 25% coconut oil asa function of time. In each animal the data were collected during the last progressionperiod. The Iipoprotein profiles were recorded with an absorbance monitor set at 280nm at a flow rate of 1 ml/min. The vertical line at d = 1.08 g/ml is included as a visualaid for easier observation of changes in HDL density.



ownloaded from



baseline was only 0.045 mg/ml and in the hyper-lipidemic state ranged between 0.15 and 0.25 mg/ml.Since apo E and apo A-l are the major apoproteins ofHDLC, the rise in concentration of apo E should havebeen similar to that of apo A-l, which was 10 to 20times higher than apo E throughout the study. Fur-thermore, apo E was found mainly over a densityinterval ranging from Lp(a) to LDL with only minorquantities in the HDLL fraction (Figure 2), which indi-cated that the density distribution of HDLC differedfrom HDL,..

After the concentration of HDI_L had reached maxi-mal values (in 10 to 11 days), the buoyant density ofHDLH began to increase. In the 3 weeks during whichthe serum cholesterol had risen to over 500 mg/dl,both HDLL and HDL ,̂ were considerably denser thanunder control conditions. Although the serum totalcholesterol values of Monkey 6 continued to rise to745 mg/dl after 6 weeks on the hypercholesterolemicdiet and the values for Monkey 24 rose to 600 mg/dlafter 7 weeks, density values increased only slightly.HDLL reached a plateau at an apparent density of1.082 g/ml and HDLH at densities ranging between1.104 to 1.110 g/ml.

When the two monkeys were returned to a normalchow diet, their serum cholesterol regressed to nor-mal values within 2.5 to 4 weeks. However, the HDLdensity gradient profile took much longer to normal-

ize (Figure 3). In Monkey 6 a gradual decrease inHDL density was apparent 1 week after the regres-sion started but did not become significant until the26th week. In the case of Monkey 24, a permanentdecrease in HDL density was evident after the fourthweek. Normalization, however, was only reachedafter 26 weeks on the Purina Chow diet.

Effect of Diet on Serum Apo A-l and Apo A-llLevels

The concentration of serum apo A-l and apo A-llwas determined by radioimmunoassay as a functionof time over two cycles of progression and regres-sion and is shown in Figures 4 and 5. These figuresalso include the serum cholesterol values taken fromreference 5 to illustrate the induced cycles of hyper-cholesterolemia in these two animals and to serve asa point of reference. Serum apo A-l levels were gen-erally much more variable than apo A-ll. The latterappeared to be unaffected by the test diet and fluctu-ated approximately 30% about a value of 0.52 mg/mlfor Monkey 6 and 0.46 for Monkey 24 over the courseof the study.

Serum apo A-l levels were clearly altered by theadministration of the test diet and its substitution withnormal monkey chow during the regression period.Both monkeys exhibited similar behavior in that the

apo B

apo E

apo A-l

2 3 4 5 6 7 8 9 I 0 D I 2 3 4 5B

Figure 2. SDS-PAGE (10%) of lipoproteln fractions obtained from Monkey 24 on the8th day and 12th week after the start of the last progression. The numbers refer tofractions obtained from the "single-spin" gradients, which are indicated on the respec-tive profiles in Figure 1. A. These gels are from the 8th day and are loaded with 20 /xgprotein per gel. B. This set is from the 12th week and is loaded with 40 /ig protein pergel.



ownloaded from



RtaMMOHHOMY •

LOT L05 L09 U6

BUOYANT

L02 103 L05

DENSITY, g/hnl

L09 U6

Figure 3. HDL density gradient profiles of Monkey 6 and Monkey 24 during the lastregession phase. The experimental conditions are identical to those given in Figure 1.

I

<0 K> 20 30 40 SO 60 70

TIME (weeks)

Figure 4. The response of serum apoA-l, apoA-ll, andcholesterol over two cycles of diet-induced hypercholes-terolemia with subsequent regressions in Monkey 6. Ar-row P indicates the start of the progression period andArrow R, the regression period.

concentration of apo A-l immediately increased 30%to 70% and reached maximal values in about 1.5weeks upon the start of each progression period.The increase in apo A-l was temporary and lasted 6weeks and 4 weeks, respectively, during the twoprogression periods that were initiated in Monkey 6,and 3 weeks for the progression periods of Monkey24 before returning to preprogression apoproteinlevels. After the initial transient rise, apo A-l levelsincreased at a slower rate over 10 to 20 weeks. Themiddle progression period of Monkey 24 differed inthat apo A-l continued to decline in concentration for8 weeks until the animal was regressed with monkeychow. The animal was not bled during this regres-sion period because of treatment for a severe shoul-der injury.

At the start of each regression period there was animmediate drop of 40% to 70% in the concentrationof apo A-l that reached a minimum within 1 week in



ownloaded from



MONKEY 24

10 20 30 40 50 60 TO 80

TIME (weeks)

Figure 5. The response of serum apo A-l, apo A-ll, andcholesterol over three cycles of diet-induced hypercholes-terolemia with subsequent regressions in Monkey 24. Ar-row P indicates the start of the progression period andArrow R, the regression period.

both monkeys. However, the decrease was onlytransient and the apo A-l concentration returnedwithin 1 to 2 weeks to preregression values. With thedecrease in apo A-l, there was also a parallel drop inthe ratio of HDLL to HDLH, especially in Monkey 24.Again the decrease was only transient; however, in-stead of returning to normal values, the ratio roseabove preregression values between 2 to 3 weeksafter regression. Stable ratios were achieved after 3to 5 weeks.

HDL Cholesterol

The sera of the two test monkeys (6 and 24) werealso analyzed for HDL cholesterol over an 18-25week period at the start of the experiment (Figure 6).In Monkey 6, at the start of the progression period,there was a close association between the concen-tration of apo A-l and the level of HDL cholesterol.Both underwent a transient increase which occurredat the same time of the increase in the light HDL bydensity gradient centrifugation (Figure 1). In Monkey24, these parameters also followed a parallel behav-ior during the first regression period. Both the apo A-land HDL cholesterol decreased immediately afterthe return to normal monkey chow. Thus, HDL cho-lesterol values closely correlate with serum apo A-llevels in both monkeys so that the observed changesin apo A-l probably reflect alterations in the concen-trations of HDL.

0 20 40 60 SO 100 ICO 140 160

TIME (days)

Figure 6. A comparison of serum apo A-l concentra-tions to HDL cholesterol during the first progression andregression period in Monkey 24 and the first progressionperiod in Monkey 6.

Chemical and Physical Properties of HDL

The chemical composition of the two forms ofHDLL and HDLH of Monkey 24 were analyzed before,after 2 weeks, and after 3 months of the test dietduring the last progression cycle (Table 1). In addi-tion, molecular weights and buoyant densities weredetermined simultaneously by equilibrium centrifu-gation in solutions of NaBr (1.5% NaBr, 20% NaBrand 30% NaBr). The particle size and density ofHDLL was 3.73 x 10s and 1.081 g/ml respectivelywhile HDLH had a molecular weight of 2.47 x 105

and density of 1.109 g/ml. Two weeks of feeding thehypercholesterolemic diet to the animal did notchange these values significantly, although there ap-peared to be a small increase in the buoyant densityof the two HDL particles. After 3 months of the regi-men, the size of HDLL was significantly reduced to2.80 x 10s with a concomitant increase in density to1.101 g/ml. On the other hand, the molecular weightof HDLH was not altered in spite of a significant rise inparticle density to 1.129 g/ml.

Although HDL particle size and buoyant densitywere not greatly affected after 2 weeks of the testdiet, the chemical composition was significantlychanged and serum HDL cholesterol and apo A-lwere maximal. Both HDLs lost phospholipids andtriglycerides and gained cholesteryl esters. Freecholesterol increased only in HDl^ whereas theHDLH maintained its allotment of free cholesterol.The protein content remained relatively unchangedwith just a slight increase for the HDI_L particle.

The sum of free and esterified cholesterol mole-cules increased 31% in HDI_L but only 6% in thedenser HDLH. The greater cholesterol content of theHDL^ particle accounted for only part of the in-creased serum HDL cholesterol. A further 33% incre-ment was due to a 56% rise in the number of HDI_L



ownloaded from



Table 1. Physlcochemlcal Parameters of HDL Subfractions as a Function of Time

Molecular weight x 10~5

Buoyant density (g/ml)

Equivalent radius* (A)

Protein (g/mol)

Phospholipid}:(mol/mol)

Free cholesterol(mol/mol)

Cholesteryl ester(mol/mol)

Triglycerides(mol/mol)

Baseline

3.73

1.081

51.7

123000(33.1 )t183(38.0)

50(5.2)

98(17.0)

29(6.6)

HDL,.

2 weeks

3.78

1.087

51.7

129000(34.1)

166(34.1)

56(5.7)

138(23.7)

11(2.4)

3 months

2.80

1.101

46.6

113000(40.5)

111(30.6)

30(4.1)

98(22.7)

6.6(2.0)

Baseline

2.47

1.109

44.6

101000(40.7)

113(35.3)

22(3.4)

66(14.6)

17(6.0)

HDLH

2 weeks

2.37

1.119

43.8

103000(43.3)

96(31.5)

20(3.2)

73(20.1)

5.3(1-9)

3 months

2.37

1.129

43.7

111000(46.7)

81(26.6)

17(2.8)

76(20.7)

6.6(3.3)

*The radius of an equivalent sphere r was calculated from the molecular weight and the density of HDL based on theassumption that these particles are spherical.

tNumbers in parentheses refer to percentage composition.i"The following molecular weights were used in calculating the number of lipid molecules per mole HDL: phospholipid 775;

free cholesterol 387; cholesteryl ester 650; triglycerides 850.

particles as determined from the area of the HDLspecies on the density gradient (Figure 1) whereasthe concentration of the HDLH was constant in thistime period. Therefore, the increase in HDL choles-terol during the transient rise was accounted for bytwo roughly equal processes: an expansion of thefree and esterified cholesterol pool per HDL^ particleand an increased number of HDL^ molecules. Fur-ther feeding of the test diet resulted in additionalchanges in composition of HDL. Thus, HDL^ under-went a small loss in protein and large decreases of alllipid components relative to the 2-week composition.When compared to baseline values, only the choles-teryl ester content was unchanged, whereas phos-pholipid and free cholesterol decreased by 40%, tri-glyceride by 77% and protein by 8%. HDLH alsounderwent further changes after 3 months of the testregimen. Phospholipid, free cholesterol, and triglyc-erides were reduced relative to control values by28%, 23%, and 46% respectively, whereas the con-tents of cholesteryl ester and protein increased by15% and 10%, respectively. The total cholesterolcontent of HDLH was therefore slightly higher andthat of HDI-L, lower when compared to the baselinecomposition. Although the total cholesterol carriedby the two HDL species was slightly less than duringthe period when normal monkey chow was fed, HDLcholesterol did not fall in most cycles due to in-creased concentrations of serum apo A-1. We wereable to confirm this for only one cycle. Because thedesign of this experiment required too much serumvolume that was committed to the analysis of otherparameters, we could not afford to continue measur-ing HDL cholesterol.

Compositional Analysis of HDLThe precipitous decline in the content of phospho-

lipid, and to a lesser degree of free cholesterol, perHDL particle with increasing severity of hypercholes-terolemia raised the question whether these HDLspecies still conformed with the common lipoproteinstructure in which a spherical core of cholesteryl es-ters and triglycerides is surrounded by a monolayerof cholesterol and phospholipid fatty acyl chains withprotein and phospholipid head groups occupying theouter surface of the particle. The compositional dataof the various rhesus HDLs (Table 1) was analyzedaccording to Shen et al.23 and was compared to thecorrelations between size and chemical compositionof normal human lipoproteins obtained by theseauthors.

The space and surface fitting of the hydrophobiccore of rhesus HDL are shown in Figure 7 A and Band conform reasonably well with those determinedfor human lipoproteins. This indicates that the vari-ation in the total number of hydrophobic cholesterylesters and triglycerides, on the one hand, and that ofphospholipid and free cholesterol on the other, is notarbitrary but is related to the size of the particle. Thepacking of phospholipid and protein at the lipopro-tein-water interface for rhesus HDL is shown in Fig-ure 7 C. The slope of a line through the experimentalpoints is identical to that obtained for human lipopro-teins and indicates that the packing of phospholipidin HDL is identical in both species. However, theintercept of this line gave a molecular area of 17.6 A2

per amino acid instead of 15.6 A2 for human lipopro-teins, which indicates that the apoproteins in rhesusHDL may be packed somewhat more loosely than



ownloaded from



20 30 40 50 60

r, A

2.4 -

0 0.04 0.08 0.I2 0.I6

Figure 7. A compositional analysis of rhesus HDLs ac-cording to Shen et al.23 A. Space-fitting of the hydropho-bic core with cholesteryl esters and triglycerides. B. Sur-face-fitting of the hydrophobic core with the hydrophobictails of the phospholipids and free cholesterol. C. Pack-ing of phospholipids and protein at the lipoprotein-waterinterface. The solid line represents correlations betweensize and chemical composition of human lipoproteins andwas taken from reference 23 to serve as comparison to therhesus HDL data. Open symbols refer to HDLL; closedsymbols refer to HDL+,. o = control; D = 2 weeks, A = 3months. The terms nM, n^ and n^ represent the

h^ c ^, ^ p

number of amino acid, phospholipid, free cholesterol, cho-lesteryl ester, and triglyceride molecules, respectively, perHDL particle; r is the equivalent radius. The number ofamino acid residues per particle was obtained from theprotein content given in Table 1 by assuming an averageresidue weight of 100 as was done in reference 23.

their human counterpart. Taken altogether, the com-positional analysis indicates that the observedchanges in chemical composition of rhesus HDL dur-ing hypercholesterolemia are consistent with thestructural organization of lipoproteins. Thus, thechanges in the number of cholesteryl esters or tri-glycerides in the hydrophobic core of rhesus HDLare compensated for by equivalent alterations in thephospholipid, free cholesterol, and protein content inthe outer surface layers of the lipoprotein particle.

Discussion

Feeding diets rich in fat and cholesterol to nonhu-man primates changes not only the concentration,but also the density distribution of the serum lipopro-teins.1'5 Although alterations in the ratio of HDLj toHDLj were reported in rhesus monkeys,324 theseobservations were based on the use of density 1.125g/ml to separate human HDL, from HDLj. However,the extension of this operational definition to nonhu-man primates has not been documented and couldlead to misleading estimations of HDL subspe-cies.32425 In the present report, we attempted to cir-cumvent this problem by subjecting the sera to iso-pycnic density gradient ultracentrifugation. In thisway we were able to sequentially follow the changesin the HDL profile occurring in two rhesus monkeysbefore, during, and after a hypercholesterolemic di-etary regime. However, we encountered a problemin the nomenclature of the HDL subspecies becausethey both floated at a density less than 1.125 g/mland exhibited large density shifts as a result of thediet-induced hypercholesterolemia. Since HDL con-sisted of two subspecies throughout the study, weresorted to naming the one having a lower density,HDI_L, and the other with higher density, HDI^. Thetwo rhesus HDL subtractions may be the rhesusequivalent of human HDL^ and HDLjg,26'27 althoughour particles contained less protein and more choles-teryl esters and triglycerides than the human speciesand have lower buoyant densities and molecularweights. In a previous study28 we examined 22 maleand 27 female normal rhesus monkeys by densitygradient centrifugation and found that their HDL ex-hibited a bimodal behavior similar to that describedin the present study. In a recent report, Morris et al.29

found a unimodal, rather than a bimodal, distributionfor HDL in chow-fed monkeys. The reason for thisdifference in results is unclear; however, their HDLhad a mean density (1.1105 g/ml) and a molecularweight (207,000) which were similar to those of theHDLH subspecies observed in our studies. Duringthe more advanced stage of the hyperlipidemia (e.g.,after 3 months of feeding the test diet), the physicaland chemical properties of our two HDL species alsobegan to approach those of human HDL, and HDL,(see Table 1). Relative to the control HDL species,the increased density observed in both particles andthe reduction of the HDI-L size were caused by an



ownloaded from



almost parallel loss of surface components, namely,phospholipid and free cholesterol and core triglycer-ides and by only minor changes in the content ofcholesteryl ester and protein. Although there weregreat differences in the chemical makeup of the HDLspecies, depending on the severity of the hypercho-lesterolemia, compositional analysis according toShen et al.23 indicated that accepted lipoproteinstructure was maintained.

A transient rise in HDL concentration was pre-viously reported on feeding high fat, high cholesteroldiets to rhesus,1 sooty mangabey,2 and spider mon-keys.30 The increase in HDL that we observed wasnot due to the generation of HDLc particles, whichhave been well described by the work of Mahley etal.,31 but was caused by increased HDLL and wasattended by a parallel rise of apo A-l, but not apo A-ll.Thus, serum apo A-l is much more sensitive to di-etary manipulation than apo A-ll. This conclusion isin keeping with previous human studies32 and thework of Parks and Rudel33 who found that neithersynthesis nor degradation of serum apo A-ll in ver-vets was influenced by diet. Egg consumption in hu-mans was shown to lead to increased plasma HDLcholesterol which is entirely due to increased HDL,accompanied by an increase in apo A-l, but not apoA-ll.34

One possible explanation for the rise in apo A-l inour rhesus monkeys is an increased synthesis (in-testinal, hepatic, or other). With regard to the intes-tine, chylomicrons contribute a significant portion ofapo A-l to plasma HDL.3536 Since Parks and Rudel33

found that chylomicrons of vervets fed a high fat(either saturated or unsaturated), high cholesteroldiet contained apo A-l, an increased chylomicroninflux into the circulation could account for the plas-ma elevation in apo A-l. However, these authors alsofound that chylomicrons contained equivalentamounts of apo A-ll. Thus, chylomicron metabolismalone may not explain the divergent behavior of apoA-l and apo A-ll that we observed in our animals.Obviously, we need to assess the contributions byboth liver and intestine.

Alternatively, the rise in serum apo A-l level orHDI^ could be caused by a decreased degradationof apo A-l. Schaefer et al.37 established recently thatthe residence time of apo A-l in humans was posi-tively correlated with the level of HDLjg which ap-proaches the size and density of HDI_L in our rhesusmonkey. Another factor which favors this concept isthe fact that HDLjg contains particles that have apoA-l as the main apolipoprotein with very little or noapo A-ll.3839 We may speculate that the removal ofthese HDLs occurs either wholly or in part via the apoA-l receptor that have been described in the liver,40

lymphoblastoid cells,41 and steroidogenic tissues.40

This may no longer be true when with continuation ofthe high fat and cholesterol-enriched diet serum cho-lesterol levels approach 400 mg/dl. If these HDL spe-cies, and especially HDI^, become enriched withcholesterol to the degree that they acquire apo E,

these particles would be removed by the cE receptor, which is a relatively fast process com-pared to that ensured by the apo A-l receptor, andlower levels of HDI^ may be expected.

It has been shown that pronounced hypercholes-terolemia with serum cholesterol levels above 400mg/dl is marked by the appearance of significantamounts of cholesteryl ester enriched p-VLDL (verylow density lipoprotein) and pre-p-VLDL42 and theend of the transformation of LDL into the larger andless dense cholesteryl ester-enriched hyperlipidemicLDL.5 These particles are structurally suited for up-take by B, and B, E receptors. At this stage, the HDLspecies finished their progressive shift to higher den-sity and are relatively unimportant in comparison toLDL, p-VLDL, and pre-p-VLDL as vehicles for cho-lesterol transport. Why denser HDLs accumulate inthe plasma during overt hypercholesterolemia is un-clear. Cholesteryl esters may now be synthesized inincreased amounts by either the liver or intestine andappear in the circulation as members of p-VLDL andpre-p-VLDL.42-•" As a consequence, there is lessgeneration of surface components (apoproteins,phospholipids, free cholesterol) through the action oflipoprotein lipase on these relatively triglyceride-poor particles to insure the transformation of denseto light HDL particles.

Based on the above and previous observations,we envisage that animals exhibiting a progressiveelevation of plasma cholesterol as a response to ahypercholesterolemic diet undergo an early HDLstage where HDL particles assume a major role inthe transport of LCAT-produced cholesteryl estersand their removal from plasma (total cholesterol be-low 400 mg/dl). With the continuing influx into thecirculation of cholesteryl esters from liver and intes-tine, apo B and apo E containing particles take careof the increased cholesterol mass. In animals thatare low responders, we would expect to see only theearly HDL stage. Recent studies conducted in thislaboratory on rhesus monkeys fed a 15% lard, 0.25%cholesterol, diet support this conclusion (unpub-lished observations). Thus, it is apparent that there isa close interplay between levels of plasma cholester-ol and functional participation of plasma lipoproteinsin cholesterol transport. Rudel et al.44 investigatedthe effect of dietary cholesterol and saturated fat onthe HDL cholesterol concentration in two subspeciesof African green monkeys. They found that animalsresponding with a moderate hypercholesterolemicserum (200-400 mg/dl) often exhibited a significantelevation of HDL cholesterol. In turn, animals thatwere maximally responsive to their dietary regimehad most of their plasma cholesterol transported inthe low density lipoprotein class.

Our observation that serum cholesterol levels re-turn to baseline before normalization of the HDL pro-file deserves a comment. Either the factors control-ling the interconversion among plasma lipoproteinsare not fully operative in the immediate reversalstage of the diet-induced hypercholesterolemia or a



ownloaded from



complete equilibration between plasma and tissuecholesterol pool has not yet taken place. A moredetailed investigation of this stage of regession isworth undertaking.

As a final comment, we realize that the changesinduced in these animals were the consequence of adiet containing both saturated fat and cholesterol.Thus, we are not in a position to assess whether theeffects are due to either of these components or dueto their synergistic interaction. We tend to favor thelatter hypothesis based on data by Ershow et al.45

who showed that neither coconut oil (33%) nor cho-lesterol alone (300 mg/1000 Kcal) substantially al-tered rhesus plasma cholesterol concentration orlipoprotein profile. Studies to answer these ques-tions unequivocally are now being conducted in thislaboratory.

AcknowledgmentsWe thank Robert Wissler for his interest and advice in this work.

We gratefully acknowledge the services of Laura Harris, TimothyBridenstine, and Lance Lusk in the CORE Animal Facilities andChemistry Laboratories of the Specialized Center of Research inAtherosclerosis. We are grateful to K. Yamamoto for performingthe apo E measurements. We also thank Lili Salvador for herexcellent technical assistance and Rose E. Scott for preparing themanuscript.

References1. Morris MD, Greer WE. Hyperalphalipoproteinemia in choles-

terol-fed rhesus monkeys. Fed Proc 1972;31:727A2. Nelson CA, Morris MD, Greer WE. Effects of cholesterol

feeding on primate serum lipoproteins III. The change in highdensity lipoprotein components. Biochem Med 1983;29:214-226

3. Rudel LL, Lotland HB. Circulating lipoproteins in non-hu-man primates. In: Strong JP, ed. Primates in medicine, vol 9.Basel: Karger, 1976:245-246

4. Rudel LL, Lofland HB. Circulating lipoproteins in non-hu-man primates. In: Strong JP, ed. Primates in medicine, vol 9.Basel: Karger, 1976:245-246

5. Fless GM, Flscher-Dzoga K, Juhn DJ, Bates SR, ScanuAM. Structural and functional changes of rhesus serum lowdensity lipoproteins during cycles of diet-Induced hypercho-lesterolemia. Arteriosclerosis 1982;2:475-486

6. Foreman JR, Karlln JB, Edelateln C, Juhn DJ, RubenstelnAH, Scanu AM. Fractionation of human serum lipoproteinsby single-spin gradient ultracentrifugation: Quantification ofapolipoproteins B and A-l and lipid components. J Llpid Res1977:18:759-767

7. Fless GM, Scanu AM. Physiochemical characterization ofrhesus low density lipoproteins. Biochemistry 1975:14:1783-1790

8. Fless GM, Wissler RW, Scanu AM. Study of abnormal plas-ma low-density lipoprotein in rhesus monkeys with diet-in-duced hyperiipidemla. Biochemistry 1976;15:5799-5805

9. Patsch W, Schonfeld G, Gotto AM Jr, Patsch JR. Charac-terization of human high density lipoproteins by zonal ultra-centrifugation. J Biol Chem 1980:255:3178-3185

10. Lowry OH, Rosebrough NJ, Fair AL, Randall RJ. Proteinmeasurement with the Folin-phenol reagent. J Biol Chem1951:193:265-275

11. Bartlett GR. Phosphorus assay in column chromatography. JBiol Chem 1959;234-.466-^68

12. Allaln CC, Poon LS, Chan CS, Richmond W, Fu PC. Enzy-matic determination of total serum cholesterol. Clin Chem1974;20:470-475

13. Gallo LL, Atasoy R, Vahouny GV, Treadwell CR. Enzymat-ic assay for cholesterol ester hydrolase activity. J LJpid Res1978:19:912-916

14. Wahlefeld AW. Triglycerides: Determination after enzymatichydrolysis. In: Bergmeyer HU, ed. Methods of enzymaticanalysis, 2nd English ed. New York: Academic Press. 1974:1831-1835

15. Albers JJ, Warnlck GR, Wlebe D, et al. Multi-laboratorycomparison of three heparin-Mn+ + precipitation proceduresfor estimating cholesterol in hkjh-density lipoprotein. ClinChem 1978:24:853-856

16. Llpid Research Clinics Program. Manual of laboratory op-erations Vol 1: LJpid and lipoprotein analysis. DHEW publica-tions no (NIH) 75-628. Bethesda, Maryland: National Insti-tutes of Hearth

17. Karlln JB, Juhn DJ, Starr Jl, Scanu AM, Rubensteln AH.Measurement of human high density lipoprotein apolipopro-tein A-l in serum by radlolmmunoassay. J Lipid Res1976:17:30-37

18. Goldberg RB, Karlln JB, Juhn DJ, Scanu AM, Edelsteln C,Rubensteln AH. Characterization and measurement of hu-man apolipoprotein A-ll by radioimmunoassay. J LJpid Res1980:21:902-912

19. Edelsteln C, Llm CT, Scanu AM. The serum high densitylipoproteins of Macacus rhesus II. Isolation purification andcharacterization of their two major polypeptides. J Biol Chem1973:248:7653-7660

20. Laurell C-B, McKay EJ. Electrolmmunoassay. Methods En-zymol 1981:73:339-369

21. Pfafflnger D, Edelstein C, Scanu AM. Rapid isolation ofapolipoprotein E from human plasma very low density lipo-proteins by molecular sieve high performance liquid chroma-tography. J Lipid Res 1983:24:796-800

22. Weber K, Osborn H. The reliability of molecular weight deter-mination by dodecyl sulfate-polyacrylamide gel electrophore-sis. J Biol Chem 1969:244:4406-4412

23. Shen BW, Scanu AM, Kezdy FJ. Structure of human serumlipoproteins inferred from compositional analysis. Proc NatlAcad Sci USA 1977;74:837-841

24. Baker HN, Eggen DA, Malcolm GT, Melchlor GW, RohelmPS, Strong JP. Divergent responses of HDL in hyper- andhypo-responding rhesus monkeys observed after increaseddietary cholesterol intake. Arteriosclerosis 1981 ;1:362a

25. Scanu AM, Edelsteln C, Vltello L, Jones R, Wissler RW.The serum high density lipoproteins of Macacus rhesus I.Isolation composition and properties. J Biol Chem 1973;248:7648-7652

26. Anderson DW, Nichols AV, Forte TM, Llndgren FT. Parti-cle dlstnbution of human serum high density lipoproteins.Blochim Biophys Acta 1977;493:55-68

27. Blanche PJ, Gong EL, Forte TM, Nichols AV. Characteriza-tion of human high-density lipoproteins by gradient gel elec-trophoresis. Biochim Biophys Acta 1981; 665:408-419

28. Ress GM, Scanu AM. Presence of high density lipoproteins(HDL) of different hydrated density in control rhesus monkeysand after dietary-induced hypercholesterolemia: Evidence fora denser HDL (HDL,)- Presented at the 6th International Con-gress of Atherosclerosis, Berlin 1982

29. Monis MD, Nelson CA, Greer WE. The physical character ofhigh density lipoprotein subtractions in gorillas and rhesusmonkeys. Fed Proc 1982;41:1020

30. Srlnlvasan SR, Da if ere 8 ER, Jr, Ruiz H, Pargaonkar PS,Radhakrishnamurthy B, Berenson GS. Rapid serum lipo-protein changes in spider monkeys on short term feeding ofhigh cholesterol-high saturated fat diet. Proc Soc Exp BiolMed 1972:141:154-160

31. Mahley RW. Alteractions In plasma lipoproteins induced bycholesterol feeding in animals including man. In: Dletschy JM,Gotto AM, Ontko JA, eds. Disturbances In llpid and lipoproteinmetabolism. Baltimore: Waverty 1978:181-197

32. Zech LA, Schaefer EJ, Bronzert TJ, Aamodt RL, Brewer



ownloaded from



HB, Jr. Metabolism of human apolipoproteins A-l and A-ll:compartmental models. J Lipid Res 1983;24:60-71

33. Parks JS, Rudel LL. Different kinetic fates of apolipoproteins 40.A-l and A-ll from lymph chylomicra of non-human primates.Effect of saturated versus polyunsaturated dietary fat. J LipidRes 1982;23:410-421 41.

34. Patsch JR, Karlin JB, Gotto AM, Jr. Egg consumption andplasma lipoproteins. Arteriosclerosis 1982;2: 443a

35. Schaefer EJ, Jenkins LL, Brewer HB, Jr. Human chylomi-cron apolipoprotein metabolism. Biochem Biophys Res Com- 42.mun 1978;80:405-412

36. Tall AR, Green PHR, Glickman RM, Riley JW. Metabolicfate of chylomicron phospholipids and apoproteins in the rat. 43.J Clin Invest 1979,64:977-989

37. Schaefer EJ, Zech LA, Jenkins LL, et al. Human apolipo-protein A-l and A-ll metabolism. J Lipid Res 1982; 23:850-862 44.

38. Kostner GM, Patsch JR, Sailer S, Braunsteiner H, Hola-sek A. Polypeptide distribution of the main lipoprotein densityclasses separated from human plasma by rate zonal ultra-centrifugation. Eur J Biochem 1974;15: 611-621 45.

39. Cheung MC, Albers JJ. Distribution of high density lipopro-tein particles with different apoprotein composition: particles

with A-l and A-ll and particles with A-l but not A-ll. J Lipid Res1982;23:747-753Scanu AM, Byrne RE, Mihovllovic M. Functional roles ofplasma high density lipoproteins. Crit Revbiochem1982:13:109-140Lefkin Al, Mihovilovic M, Scanu AM. High density lipopro-teins influence cholesterol homeostasis in cultured virus-transformed human lymphoblastoid cells. J Biol Chem1982:257:14280-14287Lusk L, Chung J, Scanu AM. Properties and metabolic fateof two very low density lipoprotein subfractions from rhesusmonkey serum. Biochim Biophys Acta 1982:710:134-142Fraser R, Dubien L, Musil F, Fosslien E, Wissler RW.Transport of cholesterol in thoracic duct lymph and serum ofrhesus monkeys fed cholesterol with various food fats.Atherosclerosis 1972:16:203-216Rudel LL, Reynolds JA, Bullock BC. Nutritional effects onblood lipid and HDL cholesterol concentrations in two sub-species of African green monkeys (Cercopithecus aethiops).J Lipid Res 1981:22:278-286Ershow AG, Nicolisl RJ, Hayes KC. Separation of the di-etary fat and cholesterol influences on plasma lipoproteins ofrhesus monkeys. Am J Clin Nutr 1981; 34:830-840

Index Terms: dietary hypercholesterolemia • rhesus monkeys • progression • regression •HDL heterogeneity • transient HDL response • density gradient ultracentrifugation • apoliprotein A-lapolipoprotein A-ll • HDL structure



ownloaded from


G M Fless, D Juhn, J Karlin, A Rubenstein and A M Scanuhypercholesterolemia.

Response of rhesus serum high density lipoproteins to cycles of diet-induced

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1984 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.4.2.1541984;4:154-164Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/4/2/154World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:



ownloaded from

http://atvb.ahajournals.org/content/4/2/154

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


Documents

Response of Rhesus Serum High Density …atvb.ahajournals.org/content/atvbaha/4/2/154.full.pdfResponse of Rhesus Serum High Density Lipoproteins to Cycles of Diet-Induced Hypercholesterolemia