CA1243887A

CA1243887A - REDUCING .beta.-LACTOGLOBULIN CONTENT IN WHEY PROTEIN CONCENTRATES FOR INFANT FORMULA USE

Info

Publication number: CA1243887A
Application number: CA000458033A
Authority: CA
Inventors: Shuryo Nakai
Original assignee: CANADIAN DAIRY COMMISSION/COMMISSION CANADIENNE DU LAIT
Current assignee: CANADIAN DAIRY COMMISSION/COMMISSION CANADIENNE DU LAIT
Priority date: 1984-07-03
Filing date: 1984-07-03
Publication date: 1988-11-01

Abstract

TITLE
REDUCING .beta.-LACTOGLOBULIN CONTENT IN WHEY PROTEIN
CONCENTRATES FOR INFANT FORMULA USE

INVENTOR
Shuryo Nakai ABSTRACT OF THE DISCLOSURE
A low .beta.-lactoglobulin whey protein concentrate, suitable for incorporation into infant formula, can be made by controlling the pH of whey to about 3, adding an amount of aqueous ferric chloride suitable to precipitate most of the soluble proteins apart from .beta.-lactoglobulin, centrifuging and separating precipitate and supernatant, redissolving the precipitate in dilute acid, removing excess ferric ions, e.g. by ion exchange, and subsequently neutralizing and freeze-drying the product to yield the low .beta.-lactoglobulin whey protein concentrate. Such a whey protein concentrate may be employed in low .beta.-lactoglobulin infant formula before freeze-drying or after reconstitution with water by mixing about 2 parts of such a whey protein concentrate with 1 part of casein, such as modified bovine casein together with suitable amounts of fat, lactose, minerals and vitamins, to yield an infant formula with low .beta.-lacto-globulin content. Such a formula may be spray-dried for storage.

Description

~ 3~
.

Background to the Invention Whey protein has been used in a number of dietary preparativn6.
Malaspina and Moretti in United States Patent 3,896,241, describe the preparation o~ a whey protein concentrate which includes the step of passlng whey through a diatomaceous earth filter, ultrafiltering and contacting wlth a strongly acidic cationlc resin. Roberts' U.S. Patent 4,112,123, descrlbes a whey-based food composition including undenatured whey protein obtained by ultrafiltration. Gruette and Schul~e's U.S.
Patent 3,896,240, concerns the preparation of simulated human milk and includes the step of treating whey with an anion-exchange resin to remove substantially all anlol~ of weak acids from the whey.
In commercial human infant formulas, the ratio between casein and whey is ad~usted to simulate the composition of human milk.
~umanization of cow's milk may be accomplished by the addition of demineralized whey to skim milk. However, ~-lactoglobulin, the main whey protein in cow's milk, is not present in human milk, which instead contains o-lactalbumin, lactoferrin and immunoglobulins as its main whey proteins. Thusl a method to produce low ~-lactoglobulin whey would result ln an infant formula that i8 closer in composition to human milk than the currently available commercial formulas.
Summary oi the Invention According to the present invention there is provlded a method of preparing low ~-lactoglobulin whey protein concentrate (low ~-lg WPC) comprising:
(a) controlling the pH of whey to a value in the range 2.5 to 3.5, (b) adding suEEicient aqueous ferric chloride to the mixture resulting from step (a) to selectively precipitate most of the soluble proteins, apart from ~-lactoglobulin, present in whey, (c) separatLng the precipitate and the supernatant that result from step (b), (d) redissolving the precipitate, separated ln step (c)~ in dilute acid, (e) removing e~cess Eerrlc ions from the acld solution resulting from step (d) and neutrali~ing, and (f) concentrating the product oE step (e) to yield a low ~-lg WPC.

.i ~

~3~

There is further provided a ~ethod of preparing lnfant formula with low ~-lactoglobulin content comprising the method of preparing low ~-lg WPC outlined above, at least to step (e) with the additlonal step ~f:
S (g) mixlng in the range of 50-75% of the low ~-lactoglobulin whey pro-tein solution or concentrate produced by the method of step (e) or step (f) with, in the range of 25-50X caseln, and fat, lactose, minerals and vitamins in proportions such that the infant formula resulting resembles human milk in its protein composition.
The method of preparing dried infant formula with low ~-lacto-globulin content outlined above may be extended with the additional step of:
(h) concentrating the product of step (g) to yield a dried infant form-ula with low ~-lactoglobulin content~
The whey of step (a) may be selected from HCl-whey; lactic whey, cheddar cheese whey and cottage cheese whey.
The pH in step (a) above ls preferably controlled to be~ween

2.8 and 3.2, preferably about 3.
In step (b), sufficient aqueous ferric chloride iB usually added to the mixture resulting from step (a) to give a concentration between 3 mM and 6 mM, preferably about 4 mM, of ferric chloride in the mixture.
In step (d), the dilute ac~d employed i8 preferably hydro-chloric acid.
In step (e), excess ferric iona are removed by passage of the liquid resulting from ~tep (d), e.g. through a suitable ion exchange column, preferably a strong cation exchange column such as an Amberlite IR-112 ¦trademark] ion exchange column. Other techniques ~uch as ultra-filtration could be used for this step.
In step (f), a preferred method of concentrating the product of step (e) comprises spray drying.
A preferred embodiment of the method of preparing low ~-lg WPC
comprises:
- in step (a), controlling the pW of the whey to about 3~
- in step (b), adding sufficient aqueous ferric chloride to the mixture resulting Erom step (a) to give about a 4 m~ solution of ferric chloride in said mixture and then holding the n~xture resulting for about 2 hours at room temperature, - in ~tep (c), separatlng the precipitate and supernatant that result from step (b) bg precipitation, - in step (d), water-washing the precipitate obtained in step (c) and redissolving sald precipLtate in 3 N HCl at a pH of about 1.5, - in step (e), filtering the product of step (d) to remove undissolved matter, pas~ing the resuLting filtrate through a column containing a sultable strong cation exchange resin at a velocity of about 1.3 ml/ml resin hr to remove excess ferric ions, neutrali~ing the resul~ing efflu-ent with an aqueous solution of about I N sodium hydroxide and dialysing the resulting liquid against water, and - in step (f), spray drying the product of step (e) to yield a low ~-lactoglobulin whey proteln concentrate.
A preferred embodiment of step (g) comprises mixing about 65%
of low ~-lactoglohulin whey protein with about 35~ of casein, and fat, lactose, mlnerals and vitamins, in proportions such tha~ the infant formula resultlng resembles human milk in its protein composition.
A preferred embodiment of step (h) comprises spray drying the product of step (g) to yield a dried infant formula with low ~-lactoglo-bulin content.
A low ~-lactoglobulin whey protein r~y be prepared according to the methods of steps (a) to (e) outlined above, or their chemical equiva-lents.
A low g-lg WPC may be prepared according to the methods of steps (a) to (f), outlined above, or theLr chemical equivalentr~O
Accordlng to the present invention, there i6 further provided infant formulcl comprising in the range of 50-75% low ~-lactoglobulin whey protein with in the range of 25-50% casein together with suitabLe Amounts of fat, lactose, mlnerals and vLtamins. Dry infant formula comprislng ln the range o~ 50-75% low ~-lg WPC ~lth in the range o 25-50% casein to-gether with suitabLe amounts of fat, lactose, minerals and vitamins may also be made.
Description of the Figures Figure I shows electrophoretograms of whey protelns in I, , 38~

precipitate treated at dlfferent pHs. In figure 1, C iB cottage cheese whey, Sup is the sup2rnatant from pH 3.0 treatment, Ig i9 immunoglobu-lins, PP is proteose peptone, BSA is bovine serum albumin, o-La is ~-lactalbumin and ~-Lg is ~-lactoglobulin.
Figure 2 shows the effect of pH and ferric ion concentration (the figures at the right side of the graph) upon separation efficiency.
Figure 3 shows densitometric traces of stained polyacrylamide gel for (a) supernatant from treatment at pH 2.9 wlth 4 ~M FeC13, (b) precipitate, and (c) untreated whey.
Figure 4 shows a flow diagram of procedures for the removal of ~-lactoglobulin in the manufacture of low ~-lg WPC, together with the procedures for the recovery of ~-lactoglobulin and the preparation of crude lactose~
Flgure 5 shows a flow chart suitable for the reduction of the proportion of ~s -casein present in bovine casein.
Detailed Description of the Invention This invention is concerned with the production of low ~-lacto-globulin whey protein, low ~-lactoglobulin whey protein concentrate, and the incorporation of the same into infant formula to yield infant formula with a protein profile substantially similar to that of human mllk.
(1) Preparation of low ~-lactoglobulin whey proteln i Whey protein precipitation and determination of ~ sultable conditions_for separation of ~-lacto~lobulln The pH of the whey was ad~usted to between 2.5 and 5.5 and then 0.5 M ferrlc chloride was added to give a final Fe concentration of between 3 and 6 mM. The p~l of the reaction mixture was read~usted back to the desired pH with 1 N NaOII. After holding for 2 hr at room tempera ture, the mixture was then centrifuged at 10,000 x g for 15 min. The precipitate was subsequently analysed by polyacrylamide gel electropho-resis (PAGE). The nitrogen content was also determined using the micro-K~eldahl technique.
Electrophoresis demonstrates that between pH 2.8 and 3.2 almostall whey proteins, except for ~-lactoglobulin, were precipitated, see flgure 1. The supernatant contains mainly ~-lactoglobulin without con-tamination of immullogloblllil-s. Between pH 3.4 and 4.3, ~-lactoglobulin l1ad a tendency to precipttate. Above pU ~.5, the quantity of precipi-tated protein was small and there was no specificity.
Within the p~1 of 2.6 to 3.1, the effect of ferric ion concen-tration on separation eEfLciency was studied in more detail. The highest separation efficiency was obtained at pH 2.9 and a ferric ion concentra-tion of ~i.O m~, as is shown ln figure 2. Temperature was found not to affect the precipitation below p11 3.2 significantly but was found to be important above pH 3.~.
It was found that there was no tendency for the whey proteins to precipitate from demineralized whey (obtained by mixed ~ed ion exchange treatment) in the p11 range of 3.0 to 4.8 even r~hen ferric chloride was added to give a final ferric ion concentration of lO mM.
From the above research the following conditions were selected for non-demineralized whey: p~1 of the whey 3.0 ~ 0.1; final ferric ion concentratLon of 4.0 mM; and room temperature for 2 hr~ Under these con-ditions, the yield of low ~-lg ~PC was 28.7~ of the total whey nitrogen.
~ensitometric traces of stalned polyacrylamide gels of supernatant and precipltate obtained under very similar conditions are sl-own in rig. 3.
By adding increasing amounts of K2HP04 or KCl to aliquots of demineraliæed whey (pH adjusted and ferric chloride added), it was found that approximate1y 20-40 mg o~ phosphate/dl whey was the optimum level for separation efficiency. No difference in separation efficiency was found between the KCl-added samples and the deminerallzed sample3, Solubilization of the ferric-protein complex Above p~1 5.6 or below p~1 1.8, the solubllity of the precipitate was aLmost 100~. On the other l1and, between pH 3.0 and 4.5, the precipi-tate was gradually ~olubilized with increa6ing NaCl concentratlon. It was found that hydrochloric acid at a pll of about 105 provided suitable ; conditions for solubilization of the ferric-protein complex.
;~ 30 Removal of the Eerric ion from the ferr c-protein complex A 2/'o (w/v) suspenston of ferric-protein complex was acidified with 3 N 11Cl to between pl1 1.5 to 1.6 and then Eiltered using Whatman No.
1 filter paper. As tryptophan residues of protein are readily destroyed under low p11s, the ferric-protein complex was passed through a column containing Amberlite IR-112 resin at a ~pecific velocity of 1.3 ml/ml gel hr, then neutraLized. This resin can effectively retain as much as !

.. . .. . ....

c ~ 3~7 1.5 mg Fe/m1 resin. If the break point was ~et a~ 98% ferrlc ion removal, the processing capacity of the resin would be 100 g protein/3.1 lltres resln. On the other hand, Rexyn 101 ~ could not remove th~ ferric ion from the ferric-protein complex under the same experimental condi-tions.
~'~ Compared to Rexyn 101, Amberlite IR-112 has fewer crosslinks:
2~ against 8% of Rexyn 101. Since the ferrlc-protein co~plex has a large molecular weight, the complex may have been unable to penetrate the interior of Rexyn 101.
A chelating ion exchange resin, uch 8S Chelex 100 G9, i8 reported to have an extraordinarlly high affinity for copper, i}on and other heaYy metals compared to alkali and alkall earth metals, such as sodium, potassium and calcium. The Eerric-protein complex was di~olved ,~ in a NaOH solution at pH 6.8 - 8.0 and then treated with Chelex 100. At - 15 this p~l range, however, Chelex 100 did not re~ove the ferric ion fro~ the ferric-protein complex.
A flowchart (figure 4) outlines procedures for the removal of ~-lactoglobulin in the ~anllEacture of low ~-lg WPC together with proce-dures for the recovery of ~-Lactoglobulin and the preparation of crude lactose. Another method of recovering ~-lactoglob~llin from the super-natant comprises adJusting the pH of the supernatant to 4.0 ~ 0.3 and adding ferric chloride to give a concentration of 10 n~l and then centri--'; fuging.
After ion-excllange or the equivalent treatment to remove excess ferric ions, the low ~-lg WPC solution was neutralized with 1 N NaOH, dialy~ed agaLnst distllLed water and then freeze drled. The 1% (w/v) low ~-lg WPC solution did not precipitate in the neutral p~l range when heated to 90C for 30 min, while ordlnary WPC prepared by ultraflltratiorl was - qulte heat unstable. According to J.N. DeWlt (Neth. Mllk Dairy J. 35, 30 47, 1981) the thermal behaviour of the whey proteins is mainly governed by the properties of ~-lactoglobulln. The ~-lactoglobulin content of the iron-re~noved low ~-lg WPC was much less than that of the ultrafiltered WPC; 3.3-11% compared to 50% in the total protelns according to the PAGE
patterns.

i .. . . . . .

i The am~no acid composition of the ultraEiltered WPC, low ~lg WPC, and ~-lactoglobulin rich fraction is given below in Table 1. These results show that low ~-lg WPC and the ~-lactoglobulin rich fraction have ~n excellent amino acid composition. Compared to the ultrafiltered WPC, the limiting amino acid of low ~-lg WPC i6 valine and, in the case of ultrafiltered WPC ls an aromatic amino acid due to high ~-lactoglobulin content.
(2) Preparation o Infant Formula A new infant formula was formulated to meet the nutritional requirements recommended by the European Society for Pediatric Gastro-enterology and Nutrltion ~ESPGAN) and is shown in Table 2. The protein composition of human and COW~8 milk, commercial humanized formula and the new lnfant formula are shown in Table 3. The whey proteln composition of the new formuLa ls considerably closer to human milk than commercial humanlzed formula.
Modlflcation of bovine casein _ The main component of bovine casein is ~ -casein, which is, however, only a trace component of human milk~ which contalns ~- and k-casein as its ma~or components. Under physiological conditions, bovine ?0 casein form~ a hard curd and it is believed that the o -casein fraction rl may be responsible. In order to modify the ~ and ~-casein proportions in bovine casein, calcium ions and rennet were used. Figure 5 shows a flow chart of a procedure suitable for the reduction in the proportlon of C~8l-case1n present in bovine ca~ein.
The rennet actlvity test was performed to determine the appro priate rennet level for casein modification. Based on the rennet activi-ty obtained 2.0 ml of 2% rennet per 40 ml casein solution was chosen Eor casein modificatlon. Calcil1m ion concentration for the final casein treatment was determined on the basis that about 0.44 g of casein protein 30 is present in 100 g of human milk. To simulate human milk, 30% recovery i is desired from 2X casein giving 0.6% whlch was chosen inste2d of 0.44%
in the hope that the modified casein may possess similar physical proper-ties to human milk a6 well as a lligher nutritional quality brought about by a higher protein content. After experimentation, 10.6 mM calcium was selected as a sultable level to be employed in casein modification. It ~3~l~7 Amino Acid Composition (g/100 g protein) of ultrafiltered . WPC, lo~-Lg WPC and ~-~Lg rich fraction - AminoUltrafilteredLow-l.gP-Lg Rich ~-Lg ~Lact~
Acids WPC WPC Fractlon _Albumin ASP 10.60 13.18 13.66 11.4 18.7 THR 7.01 6.50 6.22 5.0 5.5 S~R 5.35 6.27 5~00 4.0 4.8 GLU 17.59 15.77 17.62 19.3 12.9 PR0 7.96 4.88 5.33 5.1 1.5 .' GLY 2.04 3.18 2.45 1.4 3.2 ALA 4.24 3.82 5.08 7.0 2.1 VAL 5.68 4.72 5.09 6.1 4.7 MET 1.91 2.51 3.20 3.2 `1.0 ; ILEU4.06 5.63 5.22 6.9 6.8 LEU 9 .45 10.72 9.72 15.5 11.5 . TYR 2.94 3.98 3.34 3.7 5.4 - PHE 3.58 4.07 3.30 3.5 4.5 LYS 8.77 10.30 7.29 11.8 . 11.5 HIS 1.64 2.45 2.59 1.6 2.9 TRP 2.50 2.98 2.65 2.7 5.3 ARG 2.43 3.02 2.76 2.8 1.2 CYS 2.60 3.28 3.39 3.4 6.4 TEAA250.14 57.14 52.01 63.4 65.5 Llmiting AATYR+P~IE _ VAL TYR~PFI~ TYR-tP~lE VAL
Chemical . Score76.9 65.5 75.5 67.2 56 9 Amlno ~cid compositions of ~-Eg and ~-lactalbumin are quoted from "Fundamentals of ~airy Chemistry" edited by B.H. Webb, A.H. Johnson and J~A. Alford, 1974, p 94, the Avi Publlshing Co.
2 TEAA: Total essential amino acid ~f~L?/~

T~aE 2 Chemic~l1 Compo6ition oi a New Type of Infant Formula CbmmerciaL ES~GAN~s New Formula Prcxluct a Recommendation b 5upplies UnitlicLdc1/10(~n1 pcwder/iOOg liquid/lOOiL ___ li4uid/lOCml _ Carbohydrate g 1.4 10.37 1.5 1.2 - 1.9 Fat g3.8 28.14 3.6 2.7 - 4.1 Carbbhydrate g 7.6 56.27 7.2 5.4 - 8.2 A~h g0.3 2.2 0.25 ergy Kea170.2 519.8 67.6 6&(64-72) Calcium mg 48 356 44.4 40 Phws~lorus mg 32 237 33.0 20-35 Magnesium mg 5 37 5.3 4 Sodi~n ~g 20 148 15 Potassium mg 60 444 56 ChLoride mg 50 370 37 Iron mg1.0 7.4 1.25 0.7 zinc mg0.4 3.0 0.37 0.2 Copper ~g 30 222 47.5 20 lodine ~g5.1 38 6.8 34 M~nganese ~g _5,1 38 15.8 _ 3.4 Vitamin A IU375 2780 264 167 - 333 Vitamin D -ru 50 450 43 27 - 533 Vitamln E IU1.2 8.9 0.95 0.47 ~itamin K ~g 6 45 5.8 2.7 T~-iamine ~g 60 450 71 n Ri.bo~lavin ~g go 675 105 40 Ascorbic acid mg 12 90 5.8 S.3 Pyrldoxine ~g52.5 390 42.3 23.3 Niacinam~ide ~g 375 2780 530 167 Pantothenic acid ~g 450 3340 210 200 FoLic acid ~g 6 45 5 2.7 Biotin ~ g2.5 19.0 1.5 1.0 Ca/P ratio 1.5 1.345 not less ~lan 1.2 not more than 2~0 KlNa rat:Lo 3.0 3-73 __ _. ... ..
a 5~ ~ humanized liquid forml1a minerals an~ vitamin~ are expressed æ minim~n -;

~3~7 T~LE 3 Protein Cbmpc6ition of ~hnnan and C~'s MiIk, an~ lluma~ized and New Formula IhunEnized New Ccw's milkformula formula Human miIk ~/lOOml Total~O g/100ml Total~Og/100m1 TotalY _g/100ml Total~
Casein~ 2.60 79 0.60 40 0.49 35 0.3135 Total whey 0.70 21 0.90 60 0.91 650.57 65 G~La 0.12 3.5 0.1S 10 0.51 36.4 0.1517 ~-Lg 0.30 9.0 0.39 26 0.08 5.7 Lactoferrin trace - trace - - - 0.15 17 Serum albumin 0.03 1.0 0.04 2.7 0.06 4.3.S 6 Lyz~yme trace - trace - - - 0.05 6 Ig 0.10 3.0 0.13 8.70.21 15.0 0.1011 Others 0.15 4.5 0.19 12.60.05* 3.6 0.078 Total 3.30 100 1.50 100 1.40 100 0.88_ 00 ~hDnmnlzed formula: SM~
Cbw's and lu~man nilk oompositions were cited from }~ibraeus et al, 1977 auil 1978.
Protein composltloro of the h~u~mized formula and new formula were calculat~d ba~ed On ingredient proteins.
* thi~ fraction contains probably lactoEerrin, lysozyme, other ~inor whey CC~XX~3ntB3 peptides, free amino acids and ~heir metabolites.

, ~J
!l .

was Eound that coarseness and size of precipitated curd decreaaed ln the order of untreated casein, treated caseln and human casein in coagula~ion tests.
Infant formula preparation The ratio of whey protein ver6uE~ casein was adjusted by ~ixing 65% low ~-lg WPC and 35% modified casein. After adJustlng the proportion of the protein fraction in order to control mineral composition, several organic ions, such as calci~lm, magnesium, sodium, potassium, phosphate, citrate and chloride, were added. Table 4 details the amounts of salts added and the overall mineral composition.
Lactose was then mixed with the proteln solution and the pH was finally ad~usted. Thls solution was pasteurized at 75C or 15 min or 95C for 15 min and pH 6.8 to 7.2. Under these conditions, precipltation of protein dld not occurO After pasteuri~ation, Yegetable oil was added and the mixture was homogenized. The emulsion stabillty of thi~ solution w~s s1mllar to the product prepared uslng ultraflltered WPC. The homogeni ed solution was finally spray dried.

' ~3~t' TA~ 4 Salts Added to the ~ew Infant Formll~, per 200 ~1 Protein Ca ~ Na K P (~
In~redients ~Wei~t) ~ ng llgn~ ~g mg Iq~
~dified casein 1.167 (g) 0.98 20.53 9.30 ~-Lg WPC 2.068 (g)1.82 5~0 ~.18 Ca(Otl) 112.8 (ng) 61.0 CaCl 2tl 0128.2 (mg) 35.0 61.0 K2HI?04199 .3 (n~g) 84.47 35 .46 NaH2 439 3 (mg) 7.51 10.14 i6~120 21.1 (mg) 9 26.2 Na cltrate71.1 (mg) 19.2 ~tri~: a~Ld 21.0 (l~ 803 _ _ 12.8 TotAl 2 . 8 96 9 40 105 64 .1 100 ,

Claims

CLAIMS:

I. A method of preparing low .beta.-lactoglobulin whey protein concen-trate comprising:
(a) controlling the pH of whey to a value in the range 2.5 to 3.5;
(b) adding sufficient aqueous ferric chloride to the mixture resulting from step (a) to selectively precipitate most of the soluble proteins, apart from .beta.-lactoglobulin, present in whey;
(c) separating the precipitate and the supernatant that result from step (b);
(d) redissolving the precipitate, separated in step (c), in dilute acid;
(e) removing excess ferric ions from the acid solution result-ing from step (d) and neutralizing; and (f) concentrating the product of step (e) to yield a low .beta.-lactoglobulin whey protein concentrate.
2. A method of preparing infant formula with low .beta.-lactoglobulin content comprising the method of claim 1 at least to step 1(e) with the additional step of:
(g) mixing in the range of 50-75% of the low .beta.-lactoglobulin whey protein solution or concentrate produced by the method of step 1(e) or step 1(f) with in the range of 25-50% casein, and fat, lactose, miner-als and vitamins in proportions such that the infant formula resulting resembles human milk in its protein composition.
3. The method of preparing dried infant formula with low .beta.-lacto-globulin content comprising the method of claim 2 with the additional step of:
(h) concentrating the product of step 2(g) to yield a dried infant formula with low .beta.-lactoglobulin content.
4. The method of claim 1 wherein the whey of step 1(a) is selected from HCl-whey, lactic whey, cheddar cheese whey and cottage cheese whey.

CLAIMS (cont.):
5. The method of claim 1 step (a) wherein the pH is controlled to between 2.8 and 3.2.
6. The method of claim 1 step (a) wherein the pH is controlled to about 3.
7. The method of claim 1 step (b) wherein sufficient aqueous fer-ric chloride is added to the mixture resulting from step (a) to give a concentration between 3 mM and 6 mM of ferric chloride in the mixture.
8. The method of claim 1 step (b) wherein sufficient aqueous fer-ric chloride is added to the mixture resulting from step (a) to give a concentration of about 4 mM of ferric chloride in the mixture.
9. The method of claim 1 step (d) wherein the dilute acid is hydrochloric acid.
10. The method of claim 1 step (e) wherein excess ferric ions are removed by passage of the liquid resulting from step (d) through a suit-able ion exchange column.
11. The method of claim 1 step (e) wherein excess ferric ions are removed by passage of the liquid resulting from step (d) through a suitable strong cation exchange column.
12. The method of claim 1 step (f) wherein concentrating the product of step (e) comprises spray drying.

13. The method of preparing low .beta.-lactoglobulin whey protein con-centrate of claim 1 wherein:
- step (a) comprises controlling the pH of whey to about 3;
- step (b) comprises adding sufficient aqueous ferric chloride to the mixture resulting from step (a) to give about a 4 mM solution of ferric chloride in said mixture and then holding the mixture resulting for about 2 hours at room temperature;

CLAIMS (cont.):
13. (cont.) - step (c) comprises separating the precipitate and super-natant that result from step (b) by precipitation;
- step (d) comprises water-washing the precipitate obtained in step (c) and redissolving said precipitate in 3 N HCl at a pH of about 1.5;
- step (e) comprises filtering the product of step (d) to remove undissolved matter, passing the resulting filtrate through a column containing a suitable strong cation exchange resin at a velocity of about 1.3 ml/ml gel hr to remove excess ferric ions, neutralizing the resulting effluent with an aqueous solution of about 1 N sodium hydroxide and dialysing the resulting liquid against water; and - step (f) comprises spray drying the product of step (f) to yield a low .beta.-lactoglobulin whey protein concentrate.
14. The method of claim 2 step (g) wherein about 65% of low .beta.-lactoglobulin whey protein is mixed with about 35% of casein, and fat, lactose, minerals and vitamins in proportions such that the infant formula resulting resembles human milk in its protein composition.
15. The method of claim 3 step (h) which comprises freeze drying the product of claim 2 step (g) to yield a dried infant formula with low .beta.-lactoglobulin content.
16. A low .beta.-lactoglobulin whey protein fraction.
17. A low .beta.-lactoglobulin whey protein fraction, the whey being selected from HCl-whey, lactic whey, cheddar cheese whey and cottage cheese whey.
18. A low .beta.-lactoglobulin whey protein concentrate wherein the valine content is about 4.7 g/100 g protein.

CLAIMS (cont.):
19. A low .beta.-lactoglobulin whey protein concentrate having the amino acid composition as in Table 1, column 3.
20. A low .beta.-lactoglobulin whey protein concentrate wherein the .beta.-lactoglobulin content has been reduced to below about 11% of the total proteins in the concentrate.
21. The low .beta.-lactoglobulin whey protein of claims 16 or 17 when incorporated into an infant formula to provide from 50-75% of the pro-tein content of the formula.
22. The low .beta.-lactoglobulin whey protein concentrate of claims 18, 19 or 20 in an infant formula composition, said concentrate being from about 50-75% of the total protein content of the formula.