jurnal dislipid

Upload: christine-francis

Post on 03-Apr-2018

221 views

Category:

Documents


0 download

TRANSCRIPT

  • 7/28/2019 jurnal dislipid

    1/7

    The Lim iting Am ino Acid Sequence in Raw andRoasted Peanut Protein

    DON E. M cOSKERT he P ro cte r a nd Gamble Compa ny , M iami V alle y L ab ora to rie sC in ci nnati, Oh io

    A number of reports have appeared inthe literature on the nutritive value of thepeanut (M itchell et al., '49; Buss and God-dard, '48; Reugamer et al., '50; RutgersU niversity B ureau of B iologica l R esearch,'46-50); but none of these concern measurements of the nutritive value of thewhole peanut after it has been roasted the form in which most peanuts areconsumed. As roasted peanuts are an important constituent of the diet, an evaluation of the nutritive value of roasted peanuts w as ne eded.In the experiments reported here, thenutritive value of several peanut pasteswas determ ined. The effect of roastingunder conditions that sim ulate practicalroasting with respect to time and temperature was determ ined. The lim itingamino acid sequence in roasted peanutpaste was shown to be lysine, threonine,and methionine. In unroasted peanutpaste lysine, methionine, and threoninew ere equally lim iting.

    EXPERIMENTALPeanut pastes. A ll peanut pastes usedwere prepared from U. S. no. 1 Spanishpeanuts. The peanuts were roasted in acommercial, radiant heat roaster to theend point tem perature given below. Afterroasting, the nuts were reduced to pasteby grinding. During grinding, 2% of hy-drogenated peanut oil w as added to reduceo il separa tion.The raw peanut paste was preparedfrom nuts roasted to an end point temperature of 95C which required 14 minutes. This treatment resulted in nutsthat were essentially unroasted butblanched sufficiently that the testa couldbe removed. A roasted peanut paste wasprepared from nuts heated to an end point

    J. N UT RIT IO N, 76: '62

    of 170 Cw hich required 36 m inutes. The1 70Ce nd p oin t temp era tu re w as se le cte dbecause it approximates that sometimesused in the practical roasting of peanuts.Feeding experim ents. All experim entswere carried out using 15 m ale, w eanlingSprague-Dawley rats per group and an experimental period of 28 days. The composition of the diets is shown in table 1.The amino acid supplements were mixedwith a portion of the sucrose to assure uniform distribution in the diet. AU dietswere made isonitrogenous through theaddition of glycine to the amino acid supplement.The anim als were housed in individualcages and given feed and w ater ad libitum .Body weights were recorded once eachweek and feed consumption three timeseach week. Body weight gain and feedefficiency were used as criteria of anim alresponse for the various diets. The feedefficiency and body weight gain data w ereana lyzed for statistical sig nificance usin gthe analysis of variance. M inim al significant differences between means were determ ined by the m ethod of Tukey ('52).The lysine, threonine, and m ethioninecontent of the peanut protein was determ ined as foUows. A sample (about 500mg) of fat-free peanut meal was hy-d rolyzed at reflux w ith 20 m l o f red istilledconstant boiling HC1 for 18 hours. Thhydrolysates were filtered and made to aknown volum e. A sam ple was withdrawnadjusted to pH 6.8, and diluted to an as-sayab le am ino acid conc entration.L ysin e w as m ea su re d u sin g L eu co no sto cmesenteroides and the medium of Kuikenet al. ('43). Streptococcus faecalis andthe above m edium w ere used to determ inem ethionine. Threonine was determ ined

    R eceived for publication Septem ber 22, 1961.45 3

  • 7/28/2019 jurnal dislipid

    2/7

    454 DON E . M cO SK ERT A BL E 1

    Compo sitio n o f die tsC o n tr ol d ie t E x per im e n ta l d iet

    CaseinP ea n ut p a st eP ean u t o ilF at -s olu b le v it am i n m i x 1W a te r-s olu b le v it am i n m i x 2Cellulose3Salt m ix USP 14Am in o a cid su p p le m en t 4Sucrose

    to 15% protein (N X 6.25)to 28% fat1.05.05.04.85.0t o 1 00%

    to 15% protein (N X 6.25)to 28% fat1.05.05.03.75.0t o 1 00%i C om posit ion in gram s: v it am in A , 6.25; v itam in D (400,000 IU/gm ), 1.56; a-tocop h erol, 5 .0;^ zM en ad ione '60- th iam in e-H Cl, 80; r ib of lav in , 100; n iacin , 400; C a p an toth en at e, 400; folie acid(1% t r it u rat ion w ith su crose), 500; p y r id ox in e-HC l, 80 m g. V it am in B iz (0.1% tr it u rat ion w ithm ann it ol), 3; ch olin e, 60; in ositol, 40; ascorb ic acid , 2; p -am in ob en zoic acid , 2 gm . B iotm (100u g/m l), 60 m l. S u crose to 1,000 gm .s C ellu Flou r, C hicago Dietet ic S up ply C om p an y, C hicago.Aminocid con t en t as d escr ib ed in th e follow in g tab les m ad e u p as a m ix tu re w ith su crose.

    with S. faecalis and the medium of H enderson and Snell ('48) as modified bySirny et al. ('54). In all assays anincubation period of 72 hours at 37Cwasused. A cid production, determined bytitrating with 0.05 N potassium hydroxide,served as the measure of bacterial growth.RESULTSData showing the ef fect roasting has onthe nutri tive value of peanut protein is presented in table 2. The feed and proteinefficiency data point up the inferiority ofboth peanut preparations compared withcasein as a protein source for the rat. A lthough less obvious, the differences infeed and protein efficiency obtained between the raw and roasted peanuts wereT A B L E 2E ffect of roasting on the nutritiveva lu e o f pea nu ts

    ProteinsourceCaseinRaw

    peanut4Roastedpeanut4Average

    Average Averagefeed proteineightefficiency1efficiency2ain3gm36.82.4450.827.1 1.815.622.5 1.50 70.5

    bod y w eigh t gain ,_.Feed ef f icien cy = - x 100.feedF = 3 6.18; P = 0 .05; d (least sign if ican t d if fer en ce)~ 4 5 b od y w eigh t gaini Protein ef f icien cy = - p rofein ^ 5 ,- 'F = 3 4.16; P = 0 .05; d (least sign if ican t d if fer en ce)= 0 .29 .3F = 3 186- P = 0 .05; d (least sign if ican t d if fer en ce)= 26 .7 .4 R oast in g con dit ion st ated in tex t.

    shown to be significant by statisticalanalysis.T he low nutritive value of the raw peanut protein was assumed, on the basis ofreports in the literature, to be due primarilyto a def iciency of methionine. As shown intable 3, however, the addition of methionine in the absence of lysine (diet 3) orthreonine (diet 5) did not improve thefeed ef ficiency or body weight gain significantly. The addition of lysine and threonine, individual ly or in combination (diets4, 11, 13), showed no improvement infeed ef ficiency. The combined addi tion oflysine, threonine, and methionine (diets6, 7, 8) produced a significant increase infeed ef ficiency and body weight gain. Theaddition of either tryptophan (diet 7) orhistidine (diet 6) or tryptophan and histi-dine (diet 8) to diets containing lysine,threonine, and methionine failed to produce a further improvement in feed efficiency. H owever, body weight gain wassomewhat better with diet 8 than withdiets 6 and 7. The addition of the threemost limiting amino acids to the raw peanut protein supported better growth thanthe diet that contained 15% of casein. Theomission of threonine (diet 5) produceda significant decrease in feed effici encyand body weight gain relative to the un-supplemented peanut protein diet (diet2). Because it appeared that an aminoacid imbalance may have been produced,a second experiment (experiment 2 , table3) was carried out to determine what

  • 7/28/2019 jurnal dislipid

    3/7

    LIM ITING AM INO ACIDS IN RAW AND ROASTED PEANUTS 455TAB LE 3D etermination of the limiting amino acids in raw peanut protein

    Supplem ental am ino acids (percentage ofiet)mentDlet'1

    123456782

    910111213141516L-Lysine0.710.710.710.710.710.710.710.710.71DL-Methio-

    nine0.610.610.610.610.610.610.610.610.61DL-Threo-ine0.380.380.380.380.380.38Di--Trypto-han0.090.090.090.090.090.090.09L-Histi-ine20.130.130.130.130.130.13Glycine31.531.530.720feedefficiency445.029.529.530.524.540.543.543.046.230

    eightgain'gm120.076.085.080.060.0140.014

    1 Diets 1 and 9 contained 15% of casein. All other diets contained 15% of raw peanut protein.2 A dded as the am ino acid hydrochloride.3 A mount added to keep diets isonitrogenous.For experiment 1 F =104.38; P = 0.05; d (least significant difference) = 3.4.For experiment 2 F= 35.72; P = 0.05; d (least significant difference) = 4.0.5 For experiment 1 F= 99.43; P = 0.05; d (least significant difference) = 15.7.For experiment 2 F= 32.39; P = 0.05; d (least significant difference) = 18.0.am ino acid(s), if any, was responsible forthe imbalance. The animal performancew ith the peanut protein (diets 2, 10) andthe casein control diets (diets 1, 9) in thetw o experim ents shows that these experiments are comparable. The feed efficiencies obtained for diets 11 through 16,w hich contain the im portant combination sof the 4 am ino acids (excluding threo-nine), are not significantly different. Infact, diet 16, which was made up from thesame components as diet 5, did not produce a significant decrease in either feedefficiency or body weight gain. In a thirdexperiment, the details of which are notreported, the om ission of threonine fromthe otherwise complete amino acid mixture did not depress feed efficiency. Thus,it seems likely that an am ino acid imbalance does not exist and that lysine, threonine, and m ethionine are equally lim itingin raw peanut protein.An initial experiment (not reported)using a peanut paste prepared fromroasted peanuts as the dietary proteinsource in dicated, contra ry to the pu blishedreports (Grau, '46; Cama et al., '55), thatlysine, and not methionine, was the mostlim iting amino acid under our experi

    mental conditions. Furthermore, it suggested that m ethionine w as not the secondmost lim iting acid in roasted peanut protein. Consequently, the lim iting aminoacid sequence for roasted peanut proteinwas determ ined, as shown in table 4 (experiment 3). The addition of lysine (diet3) improves the peanut control diet asm easured by the feed efficiency. This improvement was not sufficient to be reflected in a significant increase in bodyweight gain. Data concerning diet 4 showthat methionine is not even the secondmost lim iting acid, but rather diets 5 and6 dem onstrate threonine and m ethionineto be the second and third lim iting acids,respectively. The further addition oftryptophan (diet 7) or of tryptophan andhistidine (diet 8) produced som e increasein feed efficiency and weight gain, butneither was considered statistically significant.In experim ent 3, reported in table 4, theamino acids were added as mixtures of 4amino acids except for diet 3 which contained only lysine. Since the results ofthis experiment do not agree with thosereported in the earlier literature, it wasimportant to ascertain whether a se-

  • 7/28/2019 jurnal dislipid

    4/7

    45 6 DON E. M cOSK ERT AB LE 4

    D etermination of the limiting amino acids in roasted peanut proteinSupp lemental am ino ac id s ( pe rc en tag e o fiet)ErnlS"Die"3

    123456784

    9101112131415L-Lysine*0.710.710.710.710.710.71-0.710.710.710.710.71DL-Methio-

    nine0.610.610.610.61_0.610.610.61DL-Threo-ine3_0.380.380.380.380.380.380.380.38DL-Trypto-han0.090.090.090.090.090.09L-Hisa-ine20.130.130.13 0.130.13Glycine1 .531.feedefficiency43.922.927.427.733.943.847.eightgain*9m122.145.748.455

    1 Diets 1 and 9 contained 15% of casein. A ll other diets contained 15% of roasted peanut protein.2 A d ded as the am ino acid hy drochloride.3 E quiv alent to 0.19% of available threonine since the D- form is not utiliz ed. For experim ent 3 F = 1 42.9; P = 0.05; d (least signiEcant dif ference) = 3 .7.For experim ent 4 F= 98.5; P = 0.05; d (least signif icant dif f erence) = 3.7.5 For experim ent 3 F = 154.6; P = 0 .05; d (least signif icant dif f erence) = 16.6.For experim ent 4 F= 81.2; P = 0.05; d (least signif icant dif ference) = 1 9.9.quential addition of the lim iting am inoacids w ould y ield sim ilar results (ex perim ent 4, table 4). T he results generallyconf irm those of ex perim ent 3. T here are,how ever, tw o dif f erences w hich shouldbe pointed out. First, the addition ofthreonine to the ly sine supplem ented protein (diet 12) did not produce a signif icant increase in feed ef f iciency ov er thatproduced by supplem entation w ith ly sinealone (diet 11). T he inconsistency w ithrespect to threonine probably representsa dif ference in the peanut stock or a difference in the am ount of change in theprotein during roasting since the sam epeanut stock w as not used in the tw o experim ents. T his is further borne out bythe feed ef f iciency of the peanut controldiets 22.9 in experim ent 3 and 27.8 inexperim ent 4. T he latter feed ef f iciencyis v ery close to that obtained w ith the rawpeanut preparations and is thus indicativ e of a lesser am ount of alteration during roas ti ng .S econd, the addition of try ptophan andhistidine caused a signif icant increase inf eed ef f iciency . T his increase in feed eff iciency is of questionable im portancesince the dif ference in feed ef f iciency betw een diets 6 and 8 in experim ent 3 is

    3.7 (d = 3.7) and in experim ent 4 thedif ference obtained betw een sim ilar diets(diet 13 and diet 15) w as 4.0 (d = 3.7).It is im portant, how ever, that supplem entation of either raw or roasted peanutprotein w ith ly sine, threonine, and m ethi-onine im prov ed the peanut protein to sucha degree that it w as better than the control diet w hich contained 15% of casein.T he am ount of each of the three m ostlim iting am ino acids in norm al roastedpeanut protein necessary to produce anoptim al feed ef f iciency is show n in f igure1. Four supplem ental lev els w ere chosenso that tw o lev els w ould be suboptim al andtw o lev els w ould be m ore than adequate.T he point of intercept betw een the tw ostraight lines should equal the m inim alrequirem ent of the am ino acid under testw hen adequate quantities of the otherlim iting acids are supplied. T hese lev elsw ere found to be 0.31, 0.19, and 0.21% ofthe diet f or ly sine, threonine, and m ethio-n ine , respect ive ly .T he ly sine, threonine, and m ethioninecontent of raw and roasted peanut pastes,as determ ined by m icrobiological assay ,is show n in table 5. T he am ino acid content is ex pressed as percentage of the protein and as percentage of the diet.

  • 7/28/2019 jurnal dislipid

    5/7

    LIM ITING AM INO ACIDS IN RAW AND ROASTED PEANUTS 45755

    50uI 4 5

    3530 J_O .I 0 .2 0 .3 0 .4 0 .5 0 .6 0 .7 0 .8% S u pp le m en ta l Ly s in e

    O .I 0 .2 0 .3 0 .4% S u pp le m en ta l Th re on in e55

    > , 5 0uC:145 35

    30 O O.I 0.2 0.3 0.4% S u pp le m e nta l M eth io nin eFig. 1 Determ ination of m inimal amounts oflysine, threonine, and methionine to produceo ptim al fee d efficien cy.

    T AB LE 5Lysine, threonine, and methionine contentof r aw and roasted peanut protein

    Raw peanu troteinLysine

    MethionineThreonine%

    ofprotein14.021.303.16%

    ofdiet0.600.200.48Roasted

    peanutprotein%ofprotein^3.431.162.79%

    ofdiet0.520.170.421 Average of duplicate determ inations on tw o diff er en t p ro te in p re pa ra ti on s.2 A verage of duplicate determ inations on 5 differe nt p ro te in p re pa ra ti on s.

    DISCUSSIONThe nutritive value and the lim itingamino acids of the protein in peanutsroasted under conditions that approximate those of practical roasting has notbeen determ ined previously. Based onprotein efficiency, the raw and roastedpeanut preparations used in these experiments compare well w ith the boiled and160C roasted products studied by Bussand Goddard ('48). Their preparations,fed at a 10% protein level in a diet containing about 28% of fat, yielded a protein efficiency ratio of 1.84 for the boiledprotein and 1.65 for the protein roasted40 minutes at 160C. Jones and D ivine('44) obtained a sim ilar protein efficiencyratio (1.81) with unroasted peanut protein in a diet containing 15% of proteinand 8% of fat.It was unexpected that lysine, threonine, and methionine were equally lim iting in the raw peanut protein, becausem ethionin e has been rep orted consistentlyas the m ost lim iting w ith lysine the secondmost lim iting. Two reasons are possiblefor this apparent discrepancy. First, thereis the possibility that the amino acid requirem ents un der these dietary con ditions(high protein and high fat) are differentfrom those previously published and, second, that the amino acids in peanut protein are not completely available to therat. An indication that the am ino acid requirem ents are not m arkedly increased bythe high level of fat used in these dietswas obtained by substituting whole-eggprotein for peanut protein in these diets.It was found that a level of 10% of eggprotein in the diet produced w eight gains,feed efficien cies, a nd protein efficienciesequal to or better than those obtained withthe amino acid supplem ented peanut protein diets. The egg protein diet supplies0.91% of lysine, 0.46% of threonine, and0.32% of methionine. Assuming that theam ino acids of whole egg are completelyavailable and that whole egg is a wellbalanced protein, the lysine, threonine,and m ethionine content of the w hole-eggprotein diet may be used as an estimateof the requirement for these amino acidsu nd er th ese d ie ta ry c on ditio ns .The roasted peanut diets supplem entedat the optimal level contained 0.83 and

  • 7/28/2019 jurnal dislipid

    6/7

    458 DON E. M cOSK ER0.38% of ly sine and of m ethionine, respectiv ely an am ount that agrees quitew ell w ith the requirem ent estim ated fromthe w hole-egg protein diet. T he peanutdiets contained 0.61% of threonine w hichis considerably higher than the abov e estim ated requirem ent of 0.46% . T his, inconjunction w ith the dem onstration thatthreonine m ust be added to the peanutprotein diets, suggests that about 30% ofthe threonine in peanut protein is biologically unavailable to the rat. R ice is another ex am ple of a rather w idely usedprotein w hich, based upon am ino acidanaly ses, appears to contain adequatethreonine. B ut Pecora and Hundley ('51)hav e repo rted th reonine as second lim itingaf ter ly sine. R osenberg et al. ('59) alsoreported threonine as the second m ostlim iting am ino acid in rice protein.L ittle ef fort has been directed tow ardsex plaining the low threon ine av ailability .It is not know n, for instance, w hether thethreonine ex ists in a peptide sequencew hich is resistant to enzym atic attack oras a threonine deriv ativ e, such as threonine phosphate, w hich the rat m ay not beable to utiliz e as threonine. T his problemw ould seem to be grow ing in im portancenow that tw o rather w idely used proteins,rice and peanut, are reported to containa signif icant am ount of biologically unavailab le threonine .T he change in the sequence of lim itingacids upon roasting is caused by eitheractu al alteratio n o f, o r d ec reased av ailab ility of , ly sine and threonine. Data in table5 show a decrease of alm ost 15% in thely sine content of the roasted peanut protein as com pared w ith the unroasted product, w hereas the m ethionine and threonine lev els are decreased by about 10 and11%, respectiv ely . T hus, it seem s that adestruction of ly sine as w ell as a probablelowe r b io log ical avai lab il it y are re sponsibl efor ly sine becom ing m ost lim iting in theroas te d p eanuts .B ressani and M ertz ('58) determ ined thely sine requirem ent of the rat at dif ferentdietary protein lev els. A t a lev el of 16%of protein they show ed that 0.8 to 0.9%of ly sine in the diet w as required to produce m ax im al body w eight gain. T hiscom pares favorably w ith the 0.8% determ ined in the present studies. How ever,

    in arriv ing at these values it has been assum ed the ly sine in the corn gluten usedby B ressani and in the roasted peanut protein used in these experim ents w as 100%available. S uch an assum ption m ay notbe jus ti f ie d.B alasundaram et al. ('58) hav e repo rtedon the ef fect of adding, as indiv idual supplem en ts, ly sine, m ethionine, isoleucine,threonine, cy steine, or try ptophan to dietscontaining 15% of raw peanut proteinand 9% of fat. T hese authors claim asignif icant increase in protein ef ficiencyw hen ly sine, m ethionine, isoleucine, andthreonine are added; no ef fect f romcy steine; and a signif icant decrease inprotein ef ficiency ratio w hen try ptophanis added. T he average grow th rate of1.58 gm per day obtained w ith their rawprotein is con siderably low er than the 2.71gm per day obtained in the present study .T he greatest im provem ent in w eight gainshow n by B alasundaram upon supplem entation w ith a single am ino acid(m ethionine) w as only 0.17 gm per day .T his is m ark edly less than the 2.90 gmper day obtained in these studies w hen theraw protein w as supplem ented w ith ly sine,threonine, m ethionin e, histidine, and try ptophan. It is im possible to accept theirinterpretatio n of their results w ithout completely changing the present concept oflim iting am ino acids. It is probable thatthe anim als used in their ex perim ents w ereresp onding to the am ino acid supplem en tsin a nonspecif ic m anner since the extrem ely poor grow th rate is indicativ e ofa poor diet.

    SUMMARYThe lim iting am ino acid sequence ofblanched but unroasted peanut proteinw as found to be ly sine equal to threonineequal to m ethionine. T ryptophan andhistidine do not appear to be lim itingin the unroasted protein.In roasted peanut protein the lim itingam ino acid sequence w as ly sine, threonine, and m ethionine. T ryptophan andhistidine som etim es produced an increasein feed ef f iciency w hen added in the presence o f ly sine, threo nine, and m ethionine.R o astin g u nd er th e co nd itio ns d esc rib edcaused a decrease in the am ount of ly sine,threonine, and m ethionine equal to 15, 11,

  • 7/28/2019 jurnal dislipid

    7/7

    LIM ITING AM INO ACIDS IN RAW AND ROASTED PEANUTS 45 9and 10% of the total, respectively. Thisdecrease was due to an actual destructionof the amino acids. No attempt was madeto quantify the loss of biologically available am ino acids.C alculations of the lim iting am ino acidsbased upon amino acid content and uponthe rat am ino acid requirements suggestthat methionine would be more lim itingthan lysine and that threonine should notbe lim iting. The discrepancy betw een thecalculated and determ ined lim iting sequence must be due to alteration of, or ad ec re ase d b io lo gic al a va ila bility o f, ly sin eand threonine, or to both.A feed efficiency equal to or better thanthat of a 15% casein diet can be obtainedby supplem enting roasted peanut proteinwith at least 0.31% of L-lysine, 0.19% ofDL-threonine, and 0.21% of DL-methio-nine. L IT ERATUR E C IT EDBalasundaram , S., H . R. Cama, D. A. Malik andC. Venkateshan 1958 Nutritive value of differently processed groundnut meals and theeffect of supplementation of the meals withamino acids, antibiotics and vitam in Bi. J.Nutrition, 66: 75.Bressani, R ., and E. T. Mertz 1958 Relationship of protein level to the minimum lysinerequirement of the rat. Ibid., 65: 481.Buss, L . W ., and V. R. Goddard 1948 Effect ofheat upon the nutritive values of peanuts. I.Protein quality. Food Res., 13: 506.Cama, H. R., S. Balasundaram and D. A. Malik1955 The effect of heat-treatment upon thenutritive value of groundnut protein. Congr.Intern. Biochim ., R sums,ommuns. 3rdC ongr., B russels, p. 113.

    Grau, C. R. 1946 Protein concentrates asamino acid sources for the chick: corn glutenmeal, cottonseed meal and peanut meal. J.N utrition, 32: 303.Henderson, L. M ., and Esmond E. Snell 1948A uniform medium for determ ination of aminoacids with various microorganisms. J. Biol.Chem., 372: 15.Jones, D . B., and J. P. Divine 1944 The protein nutritional value of soybean, peanut, andcottonseed flours and their value as supplements to wheat flour. J. Nutrition, 28: 41.Kuiken, K . A., W . H. Norman, C. M . Lyman, F .Hale and L. Blotter 1943 The microbiological determ ination of amino acids. I. Valine,leucine, and isoleucine. J. Biol. Chem ., 151:615.M itchell, H . H., T. S. Hamilton and J. R . Beadles1949 The nutritional effects of heat on foodproteins, w ith particular reference to com mercial processing and hom e cooking. J. N utrition,39: 413.Pecora, L . J., and J. M . Hundley 1951 Nutritional improvement of white polished rice bythe addition of lysine and threonine. Ibid.,44: 101.Rosenberg, H . R., R . Culik and R. E. Eckert 1959L ysine and threonine supplem entation of rice.Ibid., 69: 217.Ruegam er, W . R., C. E . Poling and H. B. Lockhart1950 An evaluation of the protein qualitiesof six partially purified proteins. Ibid., 40: 231.Rutgers University, Bureau of Biological Research 19461950 C ooperative determ inations of the am ino acid content, and of thenutritive value of six selected protein foodsources. New Brunswick, New Jersey.Sirny, R . J., O . R. Braekkan, M . Klungs0yr andC. A. Elvehjem 1954 Effects of potassiumand sodium in m icrobiological assay m edium s.J. B acteriol., 68: 103.Tukey, J. W . 1952 Method described by H.Scheffe. J. Am . Stat. Assn., 47: 38.