Effects of Dietary Proteins and Amino Acid Mixtureson Plasma Cholesterol Levels in Rabbits1

MURRAY W. HUFF2 AND KENNETH K. CARROLL3

Department of Biochemistry, University ofWestern Ontario,London, Ontario, Canada N6A 5C1

ABSTRACT Amino acid mixtures corresponding to either casein or egg yolkprotein when fed to rabbits in a low fat, cholesterol-free semipurified diet,produced the same degree of hypercholesterolemia as the proteins themselves.Amino acid mixtures corresponding to soy protein isolate or sunflower proteingave low plasma cholesterol levels, but not as low as those obtained with theintact proteins. Experiments were designed to identify amino acids responsiblefor the differential in plasma cholesterol produced by casein or soy proteinamino acid mixtures. Plasma cholesterol levels obtained with a casein aminoacid mixture containing half the amounts of essential amino acids were significantly higher than the corresponding soy protein mixture. Feeding trials basedon systematic modifications to the casein and soy protein mixtures indicated thatinteraction between essential and non-essential amino acids was important indetermining plasma cholesterol levels. In further experiments, protein components were formulated by adding amino acids to casein to give a mixturecorresponding to soy protein, or by adding amino acids to soy protein to give amixture equivalent to casein. These diets failed to reverse plasma cholesterollevels which suggests differences in digestion and absorption of proteins relativeto amino acid mixtures. J. Nutr. 110, 1676-1685, 1980.

INDEXING KEY WORDStures •plasma cholesterol

casein soy protein amino acid mix-

inThe prominent role of nutritionhypercholesterolemia and atherosclerosisis generally accepted. Although most ofthe emphasis to date has been on dietaryfat and cholesterol, investigators haverecently recognized the significance ofother dietary compounds (1, 2).

The nature of the dietary protein hasbeen shown to have a significant effecton plasma cholesterol levels in bothexperimental animals and humans (3, 4).Rabbits become hypercholesterolemic anddevelop atherosclerosis when fed cholesterol-free diets containing casein or otherproteins from animal sources, whereasthese effects are prevented when thedietary protein requirement is suppliedby soy protein isolate or other plantproteins (5-7). In human experiments,substitution of animal protein in the diet

by soy protein mean analogues results ina significant lowering of plasma cholesterol levels in both normal subjects (8)and hospitalized hypercholesterolemicpatients (9).

Initial experiments with rabbits demonstrated that the observed effect wasat least partly due to the dietary aminoacid composition (6). An enzymatic digestof casein, or a mixture of amino acidscorresponding to casein, produced thesame degree of hypercholesterolemia asthe protein itself. An enzymatic digestof soy protein or an amino acid mixturecorresponding to soy protein, gave low

Received for publication 3 January 1980.1Supported by the Ontario Heart Foundation and the Medical

Research Council of Canada.1 Present address: The Baker Medical Research Institute, Com

mençaiRoad, Prahran, Victoria, Australia, 3181.3Career Investigator of the Medical Research Council of Canada.

1676

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

DIETARY PROTEIN, AMINO ACIDS & PLASMA CHOLESTEROL 1677

TABLE 1

Composition of diet

Ingredient1

Casein(vitamin-free)DextroseCelluflourSalt

mixMolasses(50%v/v)Vitamin

mixture3Cornoil27.0260.05.04.03.00.21.0

1The vitamins, salt mixture (Phillips and Hart saltmixture IV, plus cobalt) and "vitamin-free" casein

were obtained from ICN Life Sciences Group,Nutritional Biochemicals Division, Cleveland, OH.Dextrose monohydrate was obtained from TekladTest Diets, ARS/Sprague-Dawley Division of theMogul Corporation, Madison, WI., and Celluflourfrom Chicago Dietetic Supply House, Chicago, IL.Molasses was obtained from a local feed mill.*Casein was added in this amount in order to supplythe diet with 25% (w/w) protein. 3Composition(mg/kg): thiamin-HCl, 7.5; pyridoxine-HCl, 7.5;biotin, 0.3; myo-inositol, 150; riboflavin, 15; nicotinicacid, 75; Ca pantothenate, 80; folie acid, 0.75; cholinechloride, 1,500; vitamin A as retinyl acetate,20,000 IU; a-tocopheryl acetate, 110 IU; mena-quinone, 30. The water soluble vitamins wereprepared and added to the diet as describedpreviously (6). The fat soluble vitamins were dissolved in corn oil to provide a source of essentialfatty acids.

levels of plasma cholesterol similar to thatobtained with the intact protein. Theseexperiments suggested that the aminoacid composition of the diet could explainthe different effects of casein and soyprotein on the concentration of plasmacholesterol.

A number of human studies havedemonstrated that dietary amino acidcomposition can influence the level ofplasma cholesterol (10-13). Olsen andco-workers (10, 11), using amino acidmixtures, demonstrated that the sourceof non-essential nitrogen and the ratio ofessential to non-essential amino acidswere important with respect to plasmacholesterol levels.

This paper describes further experiments designed to investigate in moredetail the relationship between dietaryamino acid composition and plasma cholesterol levels in rabbits. Two additionalamino acid mixtures, one correspondingto egg yolk protein and the other to sun

flower seed protein, were tested. A seriesof experiments were conducted in whichsystematic modifications were made tothe casein and soy protein amino acidmixtures in an attempt to identify aminoacids or groups of amino acids which areimportant in determining the level ofplasma cholesterol. The effects of addingamino acids to casein and soy protein, inorder to alter the dietary amino acidpattern, were also investigated.

MATERIALS AND METHODS

The experimental diets were fed tomale, New Zealand White rabbits weighing initially 1.0-1.5 kg. The rabbitswere housed individually in metal cageswith wire mesh bottoms, in an air-conditioned room with controlled temperatureand lighting. All rabbits were fed a groundcommercial diet (Master Feed RabbitPellets, Maple Leaf Mills Ltd., Toronto,Ontario) for 4-5 days after arrival, and

were transferred gradually to the experimental diets over a 3-4-day period. Dietsand water were provided ad libitum,animals were weighed weekly and feedintake was measured over the 28-dayperiod of the experiment. Non-fastingblood samples were obtained by cardiacpuncture from unanaesthetized rabbits.Plasma total cholesterol was extractedas described by Mann (14) and assayedby the method of Zlatkis and Zak (15).

The basal, low fat, cholesterol-free,semipurified diet was similar to that usedin earlier experiments (5, 6) and is shownin table 1. The proteins and amino acidswere incorporated into this diet in amountscalculated to provide dietary nitrogenequivalent to 25% protein. Amino acidmixtures were formulated with L-aminoacids (Ajinomoto U.S.A. Inc., New York).In all diets using soy protein, the isolatePromine-R (donated by Dr. E. W. Meyer,

Central Soya, Chicago, IL) was used.Amino acid mixtures corresponding to soyprotein were formulated using the aminoacid composition of the isolate (table 2).

The diets in this study were fed torabbits for a 28-day period. The experiments were grouped into four series, andthe diets in each series were formulatedas follows.

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

1678 MURRAY W. HUFF AND KENNETH K. CARROLL

liifures1fea

§oCW

1CQO1"**&H

ooexouCE1X4*i8•eu

CO011•eJw>I

'CC/JX

| •< 03.£

2 oSU

+'*

"î"<S*lîf'E

^ o"Es< "1<

._5£•.=-p-5 gc«u)

Õ/3 £ ''o

+<

¿ =-5-U

< O1x £g*31/5

+<<

à a'Su<5|S o3g^ £y.uS

¿F=affix

S £«£•?<=!<<u

u+x

<-°<<«hl+ï

^JÕJw

+|<3IÕ

a uc3*Ç

<IIa1*eullo.E

sïe«oSlUC

t)o

1-if> QQQ,--SSSSSSS5SSSn-«c,nr-oc,oc,ocjiri«^<oiriei>»o-'iriE3a5SS2S535ininooœNcnœin«««»MMaiit>«i^F«>}«ir>inoo<o<so<oiB<N«Ninrieieiiriodi^M'V'W-i«S3S5SSHSS25nmootoc4rt(cinoio4inwc4oiiriooi-:ci-9i'*-(oPSS5S§^S525SSSSSSS532SP3S5"S^S525mmao(OC4c^(otnc4(Mmooj««cef:«^»-<oÕ3S"Õ5§^SS25ÊSsassaSasaIß

Ift IOWS§-Sririö3-S3t-'n««oo^«-»«--»i^jc-tirjcxci--r-r-raoioei'a«tocr>e4-*r)^iftW>MOD«e4COtOMC4(Mlftricleiiriooi^ei'»-»-»t^piftootDq-jnt-;'*^1i

ö S s «Sì

illlllii,1î1111! î111 ! 1

iIO

OÕr

9NinS

£a>

t~ri-4.

01»1

i^i

0°1^1

§i

11l111

1O)

inS

P•"*in

oin

oo•*oir-

«citeO>

ï^;Non-essential

aminoacids

L-Alanine

L-Asparticacii32cioe,«111•1111r~Cs**<p^_0O]L-Cystinein

o (C 04i* •<•t^•*-•

r- inrta ^ ^*So

in tri-HOí

Oí "Vt-Q

IO IOCOci

04in1 3 3in

o o.S sá*I04

OO511*1"4*

(O «9|lilililio»§111i

111•*CO 0404gj

2 «a•«

oo o inos t* oóinHco

n coqco^ oo^O

O5 !*••*gì

^ Voiri

c xxoö kriin-l

r i -r l-o

« «rt-oL-Glutamic

ac

L-proline1Serinel^Tyrosines-aä

'SÌ|EOPto

^i

<_go«£SSO

ed8|fl

e—

"0ax=s||h.

e11u

3SO*•!.¿Iil•a

Hi-i1

"i11!s^

1ìliWDi•^ jSfB**

"Õ1

"5T>»SeiliI

S»•2s-aSS-aill*

iecrtÜc£*^^

Qft.9JC** uu&

g-SHlÃŽo-

O'-M-ë*J|•S

"2aHJOS

g:JSu¿'Si

¿»•il*ï

TSS

a?1Expressed

proteinwerepro-determinedesse

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

DIETARY PROTEIN, AMINO ACIDS & PLASMA CHOLESTEROL 1679

Series 1. Two additional protein sources,sunflower and egg yolk protein weretested along with their correspondingamino acid mixtures. The dietary nitrogenin the basal diet was supplied by thefollowing: a) sunflower seed protein isolate (donated by Dr. F. W. Sosulski; CropScience Dept, University of Saskatchewan, Saskatoon, Canada); b) an aminoacid mixture corresponding to sunflowerseed protein; c) egg yolk protein [TekladMills, Madison, WI; delipidated as described by Hamilton and Carroll (5); finalpreparation contained 81% protein andless than 1% lipid], and d) an amino acidmixture corresponding to egg yolk protein.

Series 2. This series of experiments wasdesigned to investigate which aminoacids or groups of amino acids influencedthe level of plasma cholesterol. In orderto simplify the approach, systematic modifications were made to the casein and soyprotein amino acid mixtures. The effectsof these modifications on the level ofplasma cholesterol were determined. Inmost cases, the results of one experimentdetermined the modification used for thefollowing experiments.

The mixtures of amino acids wereformulated as follows: a) a mixture corresponding to soy protein isolate exceptthat the essential amino acids (EAA) werereduced by one-half and the non-essentialamino acids (NEAA) increased proportionately; b) a mixture corresponding tocasein except that the EAA reduced byone-half and the NEAA increased proportionately; c) a mixture in which the EAAwere in the same proportions as in soyprotein isolate and the non-essentialnitrogen was supplied entirely by glu-tamic acid; d) a mixture in which the EAAwere in the same proportions as in caseinand the non-essential nitrogen was supplied entirely by glutamic acid; e) thesame as in (c) except that alanine suppliedthe non-essential nitrogen; f) the sameas in (d) except that alanine supplied thenon-essential nitrogen; g) soy proteinEAA plus the NEAA serine, aspartic acid,proline and glutamic acid in the sameproportions as in soy protein; h) caseinEAA plus the NEAA serine, aspartic acid,proline and glutamic acid in the same

proportions as in casein; i) soy EAA plusthe NEAA serine, aspartic acid, prolineand glutamic acid in the same proportionsas in casein, and j) casein EAA plus theNEAA serine, aspartic acid, proline andglutamic acid in the same proportions asin soy protein.

Series 3. In this series the dietarynitrogen component was formulated bycombining amino acid mixtures with theintact proteins in order to alter the overalldietary amino acid pattern: a) an aminoacid mixture (54% w/w of the final mixture) was added to soy protein isolateto give a mixture equivalent in aminoacid composition to casein, and b) anamino acid mixture (47% w/w of the finalmixture) was added to casein to give amixture equivalent in amino acid composition to soy protein isolate.

Series 4. The diets in this series wereformulated on the basis of data derivedfrom previous experiments in which animal proteins were found to be hyper-cholesterolemic relative to plant proteins(5, 6). A computer program was used tocorrelate the amino acid composition ofall the proteins tested with the plasmacholesterol level obtained after 4 weeksof feeding. No single amino acid showedeither a significant positive or negative correlation with plasma cholesterol. Strongercorrelations were obtained by combininggroups of amino acids. When the sum ofthe amounts of isoleucine, leucine, thre-onine, tyrosine and serine was divided bythe sum of arginine, glycine and glutamicacid, a higher value was obtained foranimal proteins (1.0-1.5) than for plantproteins (0.5-1.0). This value was highlycorrelated with plasma cholesterol levels(r = 0.82).

To test this correlation the diets forseries 4 were formulated as follows: a) anamino acid mixture (26% w/w of the finalmixture) was added to soy protein isolateto give a ratio:

2ile, leu, thr, tyr, serSarg, gly, glu

= 1.4;

b) an amino acid mixture (21% w/w of thefinal mixture) was added to casein to givea ratio:

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

1680 MURRAY W. HUFF AND KENNETH K. CARROLL

TABLE 3

Effect of Amino Acid Mixtures on Plasma Cholesterol Levels and Growth Performance (Series 1)'

No. ofDietanimalsControlsSoy

proteinisolateSoyproteinaminoacid

mixture2CaseinCasein

aminoacidmixture*Experimental

dietsa)Sunflowerproteinb)Sunfloweraminoacid

mixturec)Egg yolkproteind)Egg yolk amino

acid mixture610696656Initial

wtg1,305

±1,254

±1,287±1,206

±1,308

±1,242

±1,482±1.302

±52826969754216051Weight

gaing/day15

±213

±214±313

±215

±211

±211±210

±3Feed

consumptionPlasma

cholesterolInitialg/day61

±61

±63±60±64±51

±67±53

±32422141066±56±60±54±48

±69

±70±63±14271587122914

days28daysmg/dl79

±127

±185±159

±57

±152

±192±127

±122047391416251668

±14124

±30221±38"213

±42«53

±12140

±28"286±35«225

±39a

' Results expressed as mean ±SEM. 2These experiments are the same as reported in Huff et al. (6)and are presented here for comparative purposes. •Significantly different from soy protein isolate(P < 0.05).

Sile, leu, thr, tyr, serSarg, gly, glu

= 0.54.

The amino acid compositions of all thediets are listed in table 2.

Analysis of variance was performedaccording to a one-way classification andtests of significance were based on Duncan's New Multiple Range test which

allowed for unequal numbers in differentgroups (16).

RESULTS AND DISCUSSION

The results of this study provide furtherevidence that mixtures of amino acids caninfluence the concentration of plasmacholesterol in rabbits. Earlier experiments (6) demonstrated that amino acidmixtures corresponding to casein or soyprotein, gave similar cholesterol levelswhich were not significantly differentfrom those obtained with the intact proteins. The experiments in series 1 weredesigned to determine whether the hy-percholesterolemia obtained with eggyolk protein and the low plasma cholesterol levels obtained with sunflower seedprotein could also be duplicated by aminoacid mixtures. Table 3 shows that the amino

acid mixture corresponding to egg yolkprotein produced elevated plasma cholesterol levels as did egg yolk protein itself.The amino acid mixture corresponding tosunflower seed protein gave a significantly higher value than the intact protein(P < 0.05). Although the difference wasnon-significant, an amino acid mixturecorresponding to soy protein also gavehigher levels of plasma cholesterol thanthe intact protein. These results suggestthat amino acid composition can partiallyexplain the differential in plasma cholesterol produced by animal and plantproteins, but may not be the only factorinvolved.

The experiments in series 2 weredesigned to investigate which aminoacids or groups of amino acids couldinfluence plasma cholesterol levels. Theseexperiments were all based on variationsof the amino acid mixtures correspondingto either soy protein or casein. The resultsobtained were variable and somewhatdifficult to explain (table 4).

It was initially thought that the essential amino acids might be important in thiseffect. This idea was investigated byreducing the essential amino acids by

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

DIETARY PROTEIN, AMINO ACIDS & PLASMA CHOLESTEROL 1681

TABLE 4Effects of amino acid mixtures on plasma cholesterol levels and growth performance (series 2)'

Plasma cholesterol

DietInitial Weight Feed

wt gain consumption Initial 14 days 28 days

Controls«Soy protein amino acid

mixture 1,254 ±82Casein amino acid mixture 1,206 ±69

g/day

13 ±213 ±2

glday

61 ±260 ±2

56 ±2754 ±8

mg/dl

127 ±20159 ±39

124 ±30213 ±42

Series 2—Modificationstoaminoacidmixtures3a)

Soy V4EAAb)Cas ViEAAc)

Soy EAA +Glud)Cas EAA +Glue)

Soy EAA +Alaf) Cas EAA +Alag)

Soy EAA +Ser,Asp,-Pro.Glu(Soy)h)

Cas EAA +Ser.Asp,-Pro.Glu(Cas)i)

Soy EAA +Ser,Asp,-Pro.Glu(Cas)j)

Cas EAA +Ser,Asp,-Pro.Glu(Soy)1,5421,4711,1971,1861,57467526551981,582

1051,353

±951,132

±841,556

±321,494

±717±215

±116

±212±113

±210±111

±117

±28±

111

±342666362655223269

656±

459

±351

±265

±279

±86±98

±60±58

±104±59

±68±51

±56±1211194911121758992541161192291701532508080±20±37±

17±24±35±

19±

18±25±

14±30116271195992271248a3014'42253

52120

±12206

±21"76

±7115

±22

1Results expressed as mean ±SEM. 2These results are the same as in table 3 and are presented in this

table for comparative purposes. The results from the soy protein amino acid mixture are from 10 rabbitsand those for the casein amino acid mixture are from nine rabbits. 3 Six rabbits per dietary group.EAA = essential amino acids. See text for a detailed explanation of the diets. * Significantly differentfrom the corresponding soy amino acid mixture (P < 0.05).

half and increasing the non-essentialamino acids proportionately. It was reasoned that, by significantly altering thequantity of essential amino acids, the roleof the essential amino acids as a groupcould be determined.

The plasma cholesterol levels obtainedwith the casein amino acid mixture containing half the amounts of essentialamino acids, were significantly higher(P < 0.01) than the corresponding soyprotein mixture. This suggested that thedifference in plasma cholesterol may bedue to the essential amino acids or aninteraction between the essential and thenon-essential amino acids.

Experiments in series 2 (c-f) weredesigned to examine these possibilities,by completely replacing the non-essentialamino acids in the two mixtures by either

glutamic acid or alanine. These modifications were based on experiments byOlsen et al. (IO, 11) who found that usingglutamic acid as the sole source of non-essential nitrogen produced a loweringof serum cholesterol in humans. Theresults of the present experiments usingglutamic acid produced surprising results. The soy essential amino acids plusglutamic acid gave high levels of plasmacholesterol whereas low levels wereobtained with the mixture of glutamicacid plus the casein essential amino acids.Use of alanine as the source of non-essential nitrogen produced a hypercholes-terolemia with both mixtures. Theseexperiments suggested several possibilities: a) the essential amino acids themselves are not responsible for the effectof casein and soy protein on plasma

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

1682 MURRAY W. HUFF AND KENNETH K. CARROLL

TABLE 5

Modification of the amino acid composition of casein and soy protein bythe addition of amino acid mixtures (series 3)1

Amino acidcomposi

tion equiv-InitialDiet2aient to: wtgWeightgaing/dayFeed

consumptiong/dayPlasma

cholesterolInitial

14daysmg/dl28 days

ControlsSoy protein isolateCasein

Soy protein 1,270 ±50 14 ±2 66±5Casein 1,192 ±45 14 ±1 58 ±1

51 ±7 58±9 71 ±492 ±3 145 ±16 210 ±41"

Protein plus amino acidmixtures

(a) Soy protein (46%)+ amino acids (54%) Casein 1,355 ±92 15 ±3 66 ±1 61 ±4 88 ±15 94 ±23

(b) Casein (53%)+ amino acids (47%) Soy protein 1,384 ±40 12 ±4 63 ±3 50 ±5 133 ±29 185 ±50"

1Results expressed as mean ±SEMfor six rabbits per dietary group. 2Added to the diet in amountscalculated to provide dietary nitrogen equivalent to 25% protein. •Signifiantly different from soy proteinisolate (P < 0.05).

cholesterol; b) the casein mixture requires a large amount of glutamic acid tolower plasma cholesterol; c) the soyprotein mixture requires non-essentialamino acids other than glutamic acid oralanine to maintain low plasma cholesterol levels, or d) the interaction betweenthe essential and non-essential aminoacids is of importance.

The diets in series 2 (g and h) wereformulated to determine which non-essential amino acids were required toproduce the differential in plasma cholesterol observed with the complete aminoacid mixtures. It was decided to includeglutamic acid and serine in the mixturessince these are found in similar amountsin casein and soy protein. The two non-essential amino acids, showing the greatest difference between the two proteins,aspartic acid and proline, were alsoincluded (table 2). The resulting mixturesgave values similar to those obtainedwith the complete amino acid mixtures,indicating that alanine, cystine and tyro-sine were not required to obtain thedifferential in plasma cholesterol levels.This finding strengthened the hypothesisthat interaction between essential andnon-essential amino acids was important.

Further modifications were made tothese mixtures by combining the casein

essential amino acids with the soy non-essential amino acids and vice versa(series 2, i and j). The casein-essential-plus-soy-non-essential amino acid mixture reversed the effect, as anticipated,giving low levels of plasma cholesterol.However, the mixture containing the soy-essential-plus-casein-non-essential aminoacids gave an even lower level of plasmacholesterol. This finding is significant inthat this mixture is the only one that hasbeen tested that gives plasma cholesterollevels as low as the soy protein isolateitself.

The experiments in series 3 (table 5)demonstrated that the addition of aminoacids to casein to give an amino acidcomposition equivalent to soy proteinhad no effect. Similarly the addition ofamino acids to soy protein to give anamino acid composition equivalent tocasein did not reverse the result.

The diets used in Series 4 were formulated to test the hypothesis that thedietary ratio:

Sue, leu, tyr, thr, ser

Sarg, gly, glu

was important in regulating plasma cholesterol levels. The addition of aminoacids to soy protein raised this ratio from

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

DIETARY PROTEIN, AMINO ACIDS & PLASMA CHOLESTEROL 1683

TABLE 6

Modification of the amino acid composition of casein and soy protein bythe addition of amino acid mixtures (series 4)1

Diet«Dietary

ratioof: SIle.Leu, Feed Plasmacholesterol2Arg,Gly,GIu

wt gain tion Initial 14 days 28 days

glday glday mgldl

ControlsSoy protein isolateCasein

0.781.1

1,270 ±50 14 ±2 66 ±5 51 ± 7 58 ± 9 71 ± 41,192 ±45 14 ±1 58 ±1 92 ± 3 145 ±16 210 ±41«

Protein plus aminoacidmixturesa)Soy protein(74%)+

amino acids(26%)b)Casein(79%)+

amino acids (21%)1.40.541,3471,298±56±429±212 ±251

±60±2374±60±11987±100

±1916109

±24162

±29-

1Results expressed as mean ±SEMfor six rabbits per dietary group. 2Added to the diet in amountscalculated to provide dietary nitrogen equivalent to 25% protein. ' Significantly different from soy proteinisolate (P < 0.05).

0.78 to 1.40 and the addition of aminoacids to casein lowered the ratio from 1.1to 0.54. As in series 3, the results ofthis experiment demonstrated that theaddition of amino acids failed to produceany marked alterations in the effects ofintact proteins on plasma cholesterol(table 6).

These experiments indicated that theintact protein component of the diet hadan overriding effect. However, this is notincompatible with the findings that plasmacholesterol levels can be markedly affected by amino acid mixtures. It is unlikely that digestion of an intact proteinwill provide amino acids in exactly thesame proportions and time sequence asfeeding an equivalent amino acid mixture(19). Therefore, the effects on plasmacholesterol obtained with the combinations of intact proteins and amino acidmixtures (Series 3 and 4) may be relatedto the way in which the amino acids arereleased during digestion. A summary ofthe results of these studies is shownin figure 1.

In a recently published abstract, Krit-chevsky et al.4suggested that the differingeffects of casein and soy protein on plasmacholesterol levels in rabbits may be dueto their different ratios of arginine tolysine (casein, 0.49; soy protein, 1.13).

Feeding casein plus arginine and soyprotein plus lysine for 8 months had noconsistent effect on plasma cholesterollevels, but the soy-protein-plus-lysinediet increased atherogenicity. The lack ofeffects of this amino acid ratio on serumcholesterol is consistent with the resultsof the present experiments. In series 3,amino acids were added to casein and tosoy protein, which reversed the arginine/lysine ratio, but had no effect on plasmacholesterol. This ratio was also testedusing the computer program mentionedearlier, which correlated the amino acidcomposition of all the proteins we havetested with plasma cholesterol. The argi-nine/lysine ratio gave a low negativecorrelation (r = —0.35).

The studies of Weigensberg et al. (20)indicated that diets high in lysine werehypercholesterolemic for rabbits. Theresults obtained with the correlationprogram did not show a strong correlationof dietary lysine with plasma cholesterollevels. Also, casein and soy protein have asimilar lysine content despite their different effects on plasma cholesterol.

The experiments presented in thispaper using amino acid mixtures clearly

•Kritchevsky, D., Tepper, S. A. & Story, J. A. (1978) Influenceof soy protein and casein on atherosclerosis in rabbits. Fed. Proc.37, 2801 (abs.).

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

1684 MURRAY W. HUFF AND KENNETH K. CARROLL

CASEMCU* sen, ASP

WW.SUJ WOT)

SUNfLOMwI ,,.,*.I "-»sor SM"AÕ£Õ>. \ „tu.

Jî, KftAA OLUISOYI «At,l>. «

SOTE*»» .CASEINSIPIA«mo

du [CASI

Fig. 1. Effects on plasma cholesterol of feeding low fat, cholesterol-free, semipurified diets containingamino acid mixtures. The diets were fed to rabbits for 28 days and results are given as mean ±SEM.

show that the plasma cholesterol level inrabbits can be modified by the amino acidcomposition of the diet. Interpretation ofthe experiments is difficult, but severaltentative conclusions can be made. It isdoubtful whether a single amino acidis responsible for the effect on plasmacholesterol levels. It is also unlikely thatthe essential amino acids or the non-essential amino acids themselves are afactor. The results suggest that interactionbetween essential and non-essential aminoacids is important in determining thelevel of plasma cholesterol. Interactionbetween other groups of amino acids suchas neutral, acidic, basic, branched chain,etc., could also be important. The aminoacid mixtures were based on the composition of casein and soy protein isolate,which were obtained by analysis of hydrochloric acid digests. Therefore, glutamineand asparagine were included with glu-tamic acid and aspartic acid, respectively.It remains possible that the glutamineand asparagine content of the two proteinsis a factor.

It is not clear how amino acids are ableto influence plasma cholesterol levels.Whether the effect is at the level of absorption or at other stages of amino acidmetabolism is not known. It is also difficult to speculate at which point in intermediary metabolism amino acids exerttheir effect on cholesterol metabolism.

It is also unclear whether amino acidmixtures affect plasma cholesterol levels

in exactly the same way as the intactproteins. Further experiments are required to determine the effects of aminoacid composition relative to the structureof the intact proteins. Also, it cannot beruled out that non-protein constituents ofthe protein preparations may be partiallyresponsible.

ACKNOWLEDGMENTS

Technical assistance by Ms. C. Millar,Ms. M. Cotéand Mr. R. Rasmussen isgratefully acknowledged. The assistanceof Dr. D. C. K. Roberts with the computerprogramming is also acknowledged.

LITERATURE CITED

1. Truswell.A. S. (1978) Diet and plasma lipids—areappraisal. Am. J. Clin. Nutr. 31, 977-989.

2. Glueck, C. J. & Connor, W. E. (1978) Dietary-coronary heart disease relationships recon-noitered. Am. J. Clin. Nutr. 31, 727-737.

3. Carroll, K. K. (1978) Dietary protein in relation to plasma cholesterol levels and atherosclerosis. Nutr. Rev. 36, 1-5.

4. Carroll, K. K. (1978) The role of dietaryprotein in hypercholesterolemia and atherosclerosis. Lipids 13, 360-365.

5. Hamilton, R. M. G. & Carroll, K. K. (1976)Plasma cholesterol levels in rabbits fed low fat,low cholesterol diets. Effects of dietary proteins, carbohydrates and fibre from differentsources. Atherosclerosis. 24, 47-62.

6. Huff, M. W., Hamilton, R. M. G. & Carroll,K. K. (1977) Plasma cholesterol levels inrabbits fed low fat, low cholesterol, semipurified diets: effects of dietary proteins, proteinhydrolysates and amino acid mixtures. Atherosclerosis 28, 187-195.

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from

DIETARY PROTEIN, AMINO ACIDS Sa PLASMA CHOLESTEROL 1685

7. Carroll, K. K., Huff, M. W. & Roberts, D. C. K.(1977) Dietary protein, hypercholesterolemiaand atherosclerosis. In: Atherosclerosis IV (Schet-tler, G., Goto, Y., Hâta,Y. & Klose, G., eds.),pp. 445-448, Springer-Verlag, Berlin.

8. Carroll, K. K., Giovannetti, P., Huff, M. W.,Moase, O., Roberts, D. C. K. & Wolfe, B. M.(1978) Hypocholesterolemic effect of substituting soy protein for animal protein in thediet of healthy young women. Am. J. Clin. Nutr.31, 1312-1321.

9. Sirtori, C. R., Agradi, E., Conti, F., Maniero,O. & Gatti, E. (1977) Soybean-protein dietin the treatment of type-II hyperlipoprotein-emia. Lancet /, 275-277.

10. Olsen, R. E., Nichaman, M. Z., Nittka, J. &Eagles,]. A. (1970) Effect of amino acid dietsupon serum lipids in man. Am. J. Clin. Nutr.23, 1614-1625.

11. Olsen, R. E., Bazzano, C. & D'Elia, J. A.

(1970) The effects of large amounts of glutamicacid upon serum lipids and sterol metabolismin man. Trans. Am. Physicians 83, 196-210.

12. Garlich, J. D., Bazzano, G. & Olsen, R. E.(1970) Changes in plasma free amino acidconcentrations in human subjects on hyper-cholesterolemic diets. Am. J. Clin. Nutr. 23,1626-1683.

13. Coles, B. L. & McDonald, I. (1972) Theinfluence of dietary carbohydrate: lipid interrelationships. Nutr. Metab. 14, 238-244.

14. Mann, G. V. (1961) A method for measurement of cholesterol in blood serum. Clin.Chem. 7, 275-284.

15. Zlatkis, A. & Zak, B. (1969) Study of a newcholesterol reagent. Anal. Biochem. 29, 143-148.

16. Steele, R. G. D. & Torrie, J. H. (1960 Principles and Procedures of Statistics, McGraw-Hill, New York.

17. FAO (1970) Amino-Acid Content of Foodsand Biological Data on Proteins, Food Policyand Food Science Service, Nutrition Division,Food and Agriculture Organization of theUnited Nations, Rome.

18. Adamson, I. & Fisher, H. (1971) The aminoacid requirements of the growing rabbit: anestimate of qualitative needs. Nutr. Rep. Int. 4,59-64.

19. Sleisenger, M.H.&Kim,Y. S. (1979) Proteindigestion and absorption. N. Eng. J. Med.300, 659-663.

20. Weigensberg, B. I., Stary, H. C. & McMillan,G. C. (1964) Effect of lysine deficiency oncholesterol atherosclerosis in rabbits. Exp.Mol. Path. 3, 444-454.

by on Septem

ber 20, 2010 jn.nutrition.org

Dow

nloaded from