INFORMATION TO USERS

This material was producad from a microfilm copy of the original document. While the moit advanced technological meant to photograph and reproduce thii document have been used, the quality it heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help ou understand markings or patterns which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to ob tain the mining page(s) or section, they are spliced into the film along w ith: adjacent pages, This may have necessitated cutting thru an image and dupli eating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image, fou will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning” the material. It is customary to begin photoi ig at the upper left hand comer of a large dieet and to continue photoi ig from left to right in equal sections with a small overlap. If necessary, sectioning is continued agein — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, titls, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed received.

Xorox Univerelty Microfilms S00 North Zoot>Road Ann Aibor, MicNgan 48106 KEHAGIAS, Christos Haralapos* 1945- COMMERCIAL CASEIN AS A SOURCE OF EDIBLE SIALIC ACID AND A GROWTH PROMOTING FACTOR FOR LACTOBACILLUS BIFIDUS VARIANT PENNSVCVWnttJS'. The Ohio State U niversity* Ph.D., 1976 Food Technology

Xerox University Microfilms # Ann Arbor, Michigan 4B106

© 1976

CHRISTOS HARALAPOS KEHAGIAS

ALL RIGHTS RESERVED COMMERCIAL CASEIN AS A SOURCE OF EDIBLE SIALIC ACID AND A GROWTH PROMOTING FACTOR FOR LACTOBACILLUS BIFIDUS VARIANT PENNSYLVANICUS

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of Tho Ohio State University

By

Christos II. Kehagias, B.Sc.

*****

Tho Ohio State University

1976

Reading Committee: Approved By P .M .T , IlnnBcn E.M. Micolojcik T. Krlstoffersen

Adviser Department of Food Science and Nutrition This investigation was supported in part by Public Health Service

Research G rants, Numbers FD-4G2 and FD-117 from tho Food and Drug Ad­ ministration, Washington, D. C,, and by IIATCIi funds. . Dedicated to

My Family ACKNOWLEDGMENTS

I would like to express my sincere appreciation to Dr. P. M. T, Hansen for his patience and encouragement throughout my graduate program, for his enthusiasm and guidance in the preparation of this dissertation. It has been an exciting experience for me to do research under his guidance. Dr. Hansen's affectionate concern for the students will always be remembered and will serve as a guide to mo in the fuluro. 1 would also liko to thank the Ilanscn family for their hospitality during my stay at Ohio State.

Special appreciation is expressed to Dr. E. M. Mikolajcik and Dr, Y. C.

Jao for their guidance with respect to microbiological studies.

I am greatly indebted to Dr. T. Kristoffcrsen, Chairman, and to the

Department of Food Science and Nutrition for giving mo the opportunity as a graduate research nssociato.

iv VITA

April 4, 1945 Born, Pteleos, Volos, Greece

1968 Diploma in Agriculture, Aristotelian Univer­ sity of Thessaloniki, Greece

1971-1972 (a) Certificate in Dairy Technology, Agri­ culture College, Wageningcn, Netherlands

(b) lies ear ch with the Nether I and Institute for Dairy Research (NIZO), Edc, Netherlands

1972-1973 Employment by the Ministry of Agriculture, Research Section - Dairy Science Research Station, DIabata, Thessaloniki, Greece

1973 -1970 Graduate Research Associate, Department of Food Scicnco and Nutrition, The Ohio State University, Columbus, Ohio

1970 Ph.D., Tho Ohio Stato University, Columbus, Ohio

PUBLICATIONS

Kelmgins, C. and Radema, 1., 1973: "Storage of Buttor Oil Under Various

Conditions," Ncih. Milk and Dairy J . , 27:379.

Kehagias, C. and Hansen, P. M. T ., 1975: "Properties of Casein After Limited

Acid Hydrolysis," J. Dairy Soi. , 58;798. (Presented at the American

Dairy Science Assoc, Meeting, Juno 1975.)

V FIELDS OF STUDY

Major Field: Food Science and Nutrition

Studios in Food Science; Professors J. L. Blaisdell, P. M, T. Hansen, W.J. Harper, R. V. Joscphson, T. Kristoffersen, and E.M. Mikol- ajcik

Studies In Biochemistry: Professors E. L. Gross and J, F. Snell

Studies in Chemistry: Professors D. Horton, M. H. Klapper, and Q. Vnn Winkle

Studies in Meat Chemistry: Professor II, W. Ockerman

Studies in Microbiology: Professors A.M. Ackormann and G. J. Bamvnrt

Studies in Nutrition: Professor V.M. Vivian

Studies in Statistics: Professor R. V, Skavaril

vi TABLE OF CONTENTS

Page

Dedication Hi

Acknowledgments iv

Vita v

Fields of Study vi

Tablo of Contents vli

List of Tables x

List of Figures xi

I. INTRODUCTION 1

II. LITERATURE REVIEW 3

A. Sialic Acids 3 1. Sialic Acids in Milk 5 a. Free (Dialy stable) Sialic Acids 6 b. Colloidal (Non-Dialystable) Forms 9 2. Isolation 13 a. Sinlyllnctoso (NANA-Lac) 14 b. Sialic Acids IS 3. Stability Toward Acid and Heat 17 4. Identification 20 5. Biosynthesis 23 6. Significance of Sinlic Acids 28 a. In Various Glycoproteins 28 b. In Casein 33 7. Nutritional Aspects 34

vii Page

B. Growth Promoting Factors of Lactobacillus bifidus variant Pcnnsylvanicus 36 1. Low Molecular Weight Compounds 36 2. Iligli Molecular Weight Compounds 40

HI. SCOPE OF INVESTIGATION . 42

IV. EXPERIMENTAL PROCEDURES 43

A. Materials 43

B, Methods 43 1. Liberation, Isolation and Purification of Sialic Acid 44 2. Direct Ehrlich Test 46 3. Colorimetric Determination of Sialic Acid 46 4. Paper and Thin-Layer Chromatography (TLC) 48 6* Crystallization of Sialic Acid 50 6. Growth of L. bifidus 51 7. Total Sugar Determination 54 8. Qualitative Test for Fructose 54 0. llexosnmincs 54 10. N-Acetylhoxosamines 54 11. Biuret Tost 54 12. Phosphorus 54 13. Ultra centrifugation-Sedimentation Analysis 55 14. Ultrafiltration 55 15. Urea Starch Gel Electrophoresis 55 16. Rennet Sensitivity 66 17. Calcium Sensitivity 56 18. Solubility Studies 56 19. Viscosity Measurements 57 20. Flavor 57

V. EXPERIMENTAL RESULTS 58

A. Liberation, Isolation and Purification of Sialic Acid 58 from Casein 1. Casoin 61 2. Hydrolyzing Agent 66 3. Neutralizing Agent 66 4. Methods of Purification 69

viii Pago

B. Identification and Crystallization of the Sialic Acid 70 1. Paper and Thin-Layer Chromatography 70 2. Crystallization 75 a. From Solution 76 b. By Solvent Evaporation 76

C. Characterization of the Casein Bifidus Factor (CBF) 82 1. Stimulation of L. bifidus by CBF and N-AcGluc 82 2. Stimulation of L. bifidus of Ultrafiltrates from CBF 83 3. Comparison of the Bifidus Activity of CBF and Casein 90 4. Chemical Composition 90 5. Physico-Chemicnl Properties 96

D. Properties of the Residual Sialic Acid-Free Casein 99 1. Yields of Residual Casein 102 2. Urea Starch Gel Electrophoresis 102 3. Sialic Acid and Phosphorus Content 110 4. Rennet and Calcium Sensitivity 111 5. Solubility Studies 112 6. Viscous Properties 121 7. Flavor 121

E. Simplified Production of a Crude Preparation 123 Containing Sialic Acid and Casoin Bifidus Fraction (SA/CBF)

VI. DISCUSSION 127

VII. SUMMARY 132

BIBLIOGRAPHY 134

ix LIST OF TABLES

Title Page

1 Nomenclature and Structure of Sialic Acids 4

2 Distribution of Sialic Acid in Milk 5

3 Oligosaccharides of Human Milk and Bovine 9 Colostrum Containing Sialic Acid

4 Sugar Content of Casein Fractions 10

5 Carbohydrate Composition of Sugar Fraction 11

6 Structure of the Polysaccharide Part of Cow 12 k -Caseinoglycopeptido

7 Sialic Acid Content of Bovine Whey Proteins 12

8 Ingredients of ATCC Medium 70 53

9 Chemical Composition of CBF 95

10 Effect of Hydrolysis on the Sialic Acid and 110 Phosphorus Contont of Casein

11 Rennet Sensitivity of Intact and Hydrolyzed Caseins 111

12 Viscosities of Hydrolyzed Casein 122

x LIST OF FIGURES

Title •

Structure of Hexasaccharide Isolated from Human 7 Milk and Containing Two NANA Residues

Pyrrole Derivatives Arising from NANA Under the 18 Action of Acid

Metabolic Pathways for the Interconversion of Amino 24 Sugars

Metabolism and Biosynthesis of NANA 29

Liberation, Isolation and Purification of Sialic Acid 59 from Casein

Elution Diagram of Dry Isolcctric Casein Hydrolysate 62 on DOWEX-l, X 8(2 x 23 cm) Column in Acetate Form

Elution Diagram of Fresh Isolectrlc Casein Hydrolysate 64 on DOWEX-l, X 8 (2 x 23 cm) Column in Acetate Form

Elution Diagram of Formic Acid Hydrolysate on 67 DOWEX-l, X 8 (2 x 23 cm) Column in Formate Form

Thin-Layer Chromatography of the Isolated Sialic Acid 71

Paper Chromatography of the Isolated Sialic Aoid 73

Microphotographs of Crystalline Sialic Acids 76

Stimulation of L» bifidus by CBF and N-AcGluc 84

UV Spectrum of FHtor-Sterilizcd and Autoclaved CBF Fractions

xi Figure Title Page

14 Growth Effect of Ultrafiltrate of CBF on L. bifidus 88

15 Comparison of Bifidus Activity of CBF with k -C asein 91

16 Comparison of Bifidus Activity of CBF with Casein 93 Preparations

17 Thin-Layer Chromatography for Amino Acids in CBF 97

18 Elution Diagram of CBF on Ccllcx D (2 x 23 cm) Column 100

19 Urea Starch Gel Electrophoresis of Sialic Acid-Free 103 Casein (from dry isolcctric casein)

20 Urea Starch Gel Electrophoresis of Sialic Acid-Free 105 Casein (from fresh isolcctric casein)

21 Urea Starch Gel Electrophoresis of Sialic Acid-Frco 108 Casein (from dry isolcctric casein)

22 Calcium Sensitivity of Hydrolyzed Caseins 113

23 Solubility of Hydrolyzed Caseins at 2®C and pH 115 4.0 to 5,6

24 Solubility of Hydrolyzed Caseins at 25°C and pH 117 4 .0 to 5 ,G

25 Solubility of Na and Ca Caseinates 119

26 Production Scheme for a Combined Sialic Aoid/CBF 124 Fraction from Casein

xii I. INTRODUCTION

The purpose of this study has been to explore the possible isolation of sialic acid from casein in a manner which would be practical and economical.

Sialic ncids arc a class of amino sugars which arc present in a wide variety of biological tissues, but for which there is only limited knowledge about any nu­

tritional role. For example, it is known that treatment of c c II b by neuramini­ dase, on enzyme which removes sialic acid from glycoproteins, has a profound effect on cellular functions.

The high content of sialic acid in bovine colostrum and human milk is in­ direct evidence that this amino sugar may be important in infant nutrition. How­

ever, as yet thoro is no clinical ovldence to prove this point. One of the reasons

might bo the scarcity of the material for nutritional studies. Although sialic

acids can be synthesized in tho organism from other sugars, it is quite possible

that tho demand by growing Infants cannot bo fulfilled by the biosynthetic routes

known to exist, and by that supplied in the diet.

Until now, sialic acid and other amino sugars have not been deliberately

incorporated into commercially avnllable Infant formulas. Therefore, any

attempt to make readily available edible sialic acid could be of value. Casein,

I which Is a commercially available dairy product, contains significant quantities of sialic acid and could serve as a source for in vitro production of sialic acid.

In the present study, wholecasein was exposed to a limited acid hydrolysis to liberate sialic acid, which was subsequently purified and its properties studied.

The reclamation of the sialic acid leaves the bulk of casein as a by-pro­ duct. For practical and economic reasons, it would be desirable to find a use for this rcsidunl casein. Therefore, the physical and chemical properties of the residual casein were also examined, although not from a nutritional point of view. This product is essentially casein which has received limited acid hydro­ lysis for the purpose of removing sialic acid. Furthermore, there might oven bo advantages of modifying casein in this manner. For example, the removal of sialic ncld might increase the digestibility of the casein, as well as making

more available for tho utilization by Lactobacillus bifidus the other amino sugars which roslde on this protein. Tho development of the L. bifidus flora in the in­

testines of infants is known to bo of importance to the well-being of the infants.

During the separation of tho constituents resulting from the isolation of

tho slnlic ncld, a mntorial was found which stimulated the growth of L. bifidus

to a very high degree. Tho nature and significance of this material was also

explored. II. LITERATURE REVIEW

A. Sialic Acids

Sialic acid is the name given to a group of amino sugars which con­ tain the nine carbon sugar, neuraminic acid, but vary in certain other sub­ stituents (Tabic 1).

Sialic acids represent a class of compounds which, over the years, have been isolated from a considerable number of animal secretions, excretions, tissues and cells as well as from certain bacteria. Sialic acids appear to be absent or irregularly distributed among organisms of a low evolutionary devel­ opment with the possible exception of bacteria. When tho evolutionary stage is more advanced, this substance is present almost universally from the cchino- derma (star fish, etc.) to the vertebrates. The reported occurrence of sialic acids in higher plants (Cabczas, 1973) is problematic because the results pub­ lished are generally based on determinations by the Thiobarbituric Aoid (TBA) method, which unfortunatoly lacks in absolute specificity for sialic acids. It

is necessary to isolate or crystallize these materials before certifying tho

general existence of sialic acids in higher plants.

Comprehensive reviews on the occurrence and distribution of sialic

acids have been published (Dlix and Jeanloz, 1969; Cabezas, 1973; Gottchalk, TABLE 1 NOMENCLATURE AND STRUCTURE OF SIALIC ACIDS

TRIVIAL NAME SYNTHETIC NAME STRUCTURE

N-Acetylneuraminic 5-Acetamido-3,5- acid dideoxy-D-glycero- (NANA) D-gulacto- .MH nonulosonic acid

OH

N-Glycolylneuram inic 3,5-Did coxy-5- acid glycolamido-D- (NGNA) glyccro-D-galacto- nonulosonic acid

N-Acetyl-4.-0- S-Acctamido-4-O- acctylncuraminic acetyl-3,5-dideoxy- C H f COHH acid D-glycero-D-galacto- nonulosonic acid ocean, ^

N-Acetyl-7-0- 5-Acetamido-7-0- tlfjC O H W acetylncuraminic aeetyl-3,8-dideoxy- acid D-glyccro-D-galacto- nonulosonio acid

N-Acctyl-7,8(or 9)- 5-Acctamido-7,8(or 9)- di- O- ac cty 1 n eura- di-O-acetyl-3,5-dldcoxy- minic acid D-glycero-D-galacto- (NODANA) nonulosonic acid

N-Glycolyl-8-0- 3,5-Dldeoxy-5-glycol- mcthyl ncu ram in Jc am ido- 8-O-methyl-D- ;o o n acid glycero-D -galacto- nonulosonic acid

4 lOGOb). Up to now, six naturally-occurring sialic acids have been isolated and ,

• * described. The trivial and systematic names, as well as the structures of those mailt sialic acids arc given in Tabic 1. The structure of other sialic acids

(N-accLyl-4,x-d!-0-acctylncuraminic acid and N-acctyl-x,y, z-tri-O-acetylncur- aminic acid) has sLill not been well established. Other forms of sialic acid, such as N-glycolyl-4(?)-0-acetylneuraminic acid, N-glycolyl-8-O-mcthylneuraminic acid, have been found in a few cases. In this review, the abbreviated trivial names will be used as identified in Table 1.

1. Sialic Acids in Milk

Generally, sialic acids arc found in nature as constituents of oligosaccha­ rides or polysaccharides, which in turn form the prosthetic group of glycopro­ teins or glycolipids. However, the sialic acid which is present in milk is an exception and can ho found both in free (dialyzable) and colloidal (non-dialyzable) forms. Kim ct nl. (19G8) have reported the following distribution of sialic acid in milk (Table 2).

TABLE 2. DISTRIBUTION OP SIALIC ACID IN MILK (concentration in /xg/ml of skim-milk) (Kim ct a l ., 1968)

Skim-milk 158.0 /jte/ml 100% Acid Casein 89.0 /xg/ml 56.3% Acid Whey 68,5 fig/ml 43.4% Protein ppt by 12% TCA 104,5 #Ag/ml 66. 1% TCA-Soluble Fraction 53.5 /tg/ml 33.8% a. Free (Dialyzablc) Sialic Acids

Gumlnskl (19G7a, 19G7b) separated on Sephadcx G-25 a fraction con­ taining low molecular, non-protoin compounds of NANA and a fraction con­ taining high molecular protein compounds from an ethyl ether extract of human milk. The content of NANA in both fractions decreased during lac­ tation.

Cnbczas, ct al. (19G8) estimated the variation over the first nine days of lactation of tho NANA content in human colostrum. The NANA

In a faction soluble in 80% ethanol, equivalent roughly to the oligosaccha­

rides, increased from 35 to 45 m g/100 ml, while the NANA in the insol­

uble fraction, containing the glycoproteins, decreased from 100 to 25 mg

per 100 m l.

Barker and Stacey (1963) reviewed studies related to the. soluble

carbohydrates present in milk and colostrum. Kuhn and Gnuhc (19G5) de­

termined tho location of sinlic acid residues in the oligosaccharides pre­

sent in human milk and bovine colostrum. These results are given in

Tablo 3.

More recently, Grimmonproz, ct nl, (1970) determined the complete

structure of an hcxnsaccharide isolated from human milk. The structure

of this sugar is given in Figure 1. FIGURE 1

STRUCTURE OF HEXASACCHARIDE ISOLATED FROM HUMAN MILK AND CONTAINING TWO NANA RESIDUES (Grimmonprcz ct a l., 1970)

7 8

COOH

•H OH

HO

;ooH A cN H 9

TABLE 3. OLIGOSACCHARIDES OF HUMAN MILK AND BOVINE COLOSTRUM CONTAINING SIALIC ACID (Compiled from Kuhn and Gauhc, 1965)

SOURCE - OLIGOSACCHARIDE

(a) NANA 2 “►3 Gal l-*-3 N-AcGluc l-*-3 Gal 1-^4 Glue

Human Milk

(«) NANA 1—*-3 Gal

While the dinlyzablc oligosaccharides of human milk contain only NANA,

bovine dtalyznblc oligosaccharides contain NANA, NGNA and NODANA. In

goat milk, both NANA and NGNA arc present among tho dialyzable oligosnccha-

ridcs (Cnbezas, 1973).

b. Colloidal (non-dialyzablo) Forms

Casein Is the major protein in milk and contains significant quantities

of sialic acid. Much information has been accumulated in regard to the pre­

sence of sialic acid us u constituent of cnsein (Gupta and Ganguli, 1965; Jolles,

I960; Mnlpress, 1961; Marier, et al,, 1963), Table 4 gives the sugar content,

including sialic acid, of whole casein and casein fractions of various species. 10

Contrary to bovine and human caseins, which contain only NANA, sheep and goat caseins contain NGNA as well.

TABLE 4. SUGAR CONTENT OF CASEIN FRACTIONS (% of dry substance) (Adapted from Jollcs, 1960)

FRACTION Gal. N-AcGal NANA

Cow Whole casein 0.38 0.35 0.36 0.22-0.24 0.15-0.20 0.31-0.51

0.385 0.22 n .d . t t -Casein 0.40 0.36 0.38 0.15° 0.21 n .d . -C asein 0.6 0.11 0.09 n.d. n .d . 0 .1 -0 .3 5 K -C asein 1.4 1.22 2 .4 n .d . n.d. 1.8-1.2 n.d. n .d . 0.79-2.20

1. 00b 0.39 0.79-0.80

Whole casein without K fraction 0.13 0.10 0.07 para-K-Casein 0.80 0.80 0.55

Sheep Whole casein 0.33 0.24 0,09

Goat Whole casein 0,39 0.31 0.13°

Human Wholo casein 2.53 2.21 0.80 n.d. n .d . 2 .1 n.d, k not determined 11 Traces of mannoso (0.05) Traces of fueose (0. 11) 0 Mixture of N-aeetyl- and N-glycoIylnouramtnic acids 11

Among the colloidal forms of oligosaccharides present in milkt the car­ bohydrate moiety of k -casein has received the most attention. Tran and Baker

(1970) studied the glycopcptidcs which w ere liberated from k -casein. They found, although their molecular weights were different, that all contained equi- molar amounts of galactose (Gal), N-acetylgalactosamine (N-AcGal) and NANA.

Finnlly, they represented tho trl saccharide units as Cl-NANA 2-*-6-£-Gnl l«*-3 or 0-N-AcGal. The above units are attached to the peptide chain through the hydroxyl groups of serine or throontno or both. In all known eases, sialic acid occupies the terminal position.

The sugar composition of bovine k -casein is not constant at the different stages of lactation (Guerin et a l,, 1974; Sinldnson and Wheelock, 1970). Recent­ ly, Cournot et al. (1975) demonstrated the variable nature of the carbohydrate part of the K-cascinoglycopcptides (CGP) from bovine milk and colostrum. After alkaline borohydride treatment and filtration on Bio-Gel P-4, different carbo­ hydrate pnrts were obtained as summarized in Table 5.

TABLE 5. CARBOHYDRATE COMPOSITION OF SUGAR FRACTION ISOLATED FROM BOVINE MILK AND COLOSTRUM CASEINO- ______GLYCOPEPTIDES (CGP) (Fournet et al.. 19751 ______SUBSTANCES RESIDUES/ MOLE Gal N-AcGalltal* N-AcGluc NANA Fractions from normal CPG S1 1 0.43 2.15 s2 1 0.40 2.00 s, _ 1 0.45 1.00 Fractions from colostrum CPG Sc0 2 0.45 0,54 2.60 Scj 2 0.52 0.90 ' 2.00 Sc2 2 0.72 0.77 1.50 Sc„ _ 2 0.82 0 2.90 * The authors reported 50% recovery of this derivative. 12

From these results, it would appear that the three sugars are not al­

ways present in cquiinolar amounts and that the sugar moiety may not always

be a Irisnccharidc but could occur as a tclrasaccharide, such as illustrated

dtngrnmmnlically in Table G (Fournet, et a l., 1975) on the basis of the molar

ratios from Table 5.

TABLE G. STRUCTUHli OF THE POLYSACCHARIDE PART OF COW K-CASEmOGLYCOPEPTIDE (Fournet, ct al., 1975)

Isolated Substance FORMULA (From Table 4) a-NANA 2-i'(2,3,4,G)/?-gnl ]-+3N-AcGnl l-*0thr. 131 S ✓ * «■» ^ '

2 * I — -* NANA ---

S3 Ot-NANA 2-*(2,3,4,G)/3-Gnl l-*-3N-AcGal l->0 thr. 131

Besides casein, oilier millc proteins may also contain sialic acid to a

varying degree. In particular proteose-peptonc, a minor fraction of milk, has

been observed lo bo considerably rich in Sialic acid (Brunnor and Thompson,

1959). Kim et al. (19GS) determined the sialic acid in bovine whey proteins.

These results are summarized in Tabic 7.

TABLE 7. SIALIC ACJD CONTENT OF BOVINE WHEY PROTEINS (% of Slnlic Acid) (Kim, ctn l., 19G8)

Total whey proteins 1.33 Immunoglobulins 0.78 Crude lactalbumin 0,72 Proteose-pcptono 0.70 0 -lactoglobulin 0,09 13

Gangull ct al. (19G7) evaluated the relative contents of NANA In pro- teosc-pcptonc and other milk proteins like casein and whey proteins. The above authors, considering the distribution in sialic content by the protein fractions, found that the sialic acid due to casein was 71 to 80%, due to pro- teose-peptone 10-15%, and whey proteins were contributing 8- 12% of sialic acid. On the other hand, an evaluation of the content of sialic acid in these protein components in milk revealed that proteose-peptone was the richest fraction, having 24 mg/g , and next was casein. The carbohydrate fraction

of immunoglobulin (IgA) was studied by Clamp and Putman (1967) and was found to contain sialic acid and glucosamine.

In addition to the sialic acid which is bound to milk proteins, fat glo­ bule membrane has also been shown to contain sialic acid. Harrison ct al.

(1975) isolated n number of sfaloglycopoptidos after treatment of intact fat globules by pronnso and fractionation on Scphndex G-50. This study, along with some earlier findings (.Jackson, et nl., 1962; Newman and IInrri6on,

1973) proved that the slnlic acid 1 b contained in intrinsic membrane components and not in absorbed serum glycoproteins.

2. Isolation

Selection of (he methods for the isolation of sialyloligosncchnrldes or free sialic acids depends upon the type or form of sinlic acid in question and on the nature of the starting material.

Deprototnlznlion has been achieved by heat dennturntion, TCA-treat- mont or addition of organic solvent. TCA treatment gives an efficient depro- 14 tctnization, but probably causes some degradation of the sugars (Zillikcn and

O'Brien, 1959). Although a partial removal of ltptd can generally be obtained by centrifugation of the extracts at low temperature, chloroform-mcthanol mix­ tures are well known to be more efficient for this purpose.

a. Sialyl lactose (NANA-Lac)

Although NANA-Lac is only one of a scries of oligosaccharides contain­

ing NANA, it has received much attention over the years because it is one of the substrates for the enzyme neuraminidase (Rnfelson et al., 1963).

NANA-Lac is contained in the dlalyzable fraction of milk and colostrum.

Schncir et al. (1962)1solalcd NANA-Lac by passing the dialyzatc from fresh cow's colostrum through DOWEX-1 ,XS column. The elution took place with pyridine-

acetic acid, Tho elution system allowed the separation of NANA-Lac from sialic acid, which is usually eluted about 4-6 tubes later. Tho fractions containing

NANA-Lac nro freeze-dried and solubilized in methanol. Insoluble material is

removed by centrifugation and the NANA-Lac is precipitated by the addition of

ether. Tho white flocculcnt solid is dried, dissolved in water, and passed through

a column of DOWEX-50,X8 to remove traces of pyridine. The effluent is freeze-

dried and tho yield of NANA-Lac is 20-40 mg per 100 ml of cow's colostrum.

Schnolr and Rnfelson (19GC) further examined the NANA-Lac prepara­

tion and found it to bo holorogcncous, consisting of 3-NANA-Lac, 6-NANA-

Lac, and a none-identified NANA containing oligosaccharide. The authors

also studied the stability of the NANA-Lac bond (cither 2.- » 3 or 2' »G)

and found that buffered solutions of the isomer at pH values 5, 7 and 9 were 15

' stable whether frozen, refrigerated, or maintained at 37 6C, In unbuffered

solutions containing 200 g/0.2 ml of either Isomer (pH 3.2), extensive

autohydrolysis was observed — 2%, 20% and 60% hydrolysis respectively

after 2, 24 and 06 hours. Drying at 78eC, prolonged storage in a desicca­

tor at room temperature, or passage of an aquacous solution through a col­

umn of DOWISX-50 (H+) resin resulted in degradation to lactose and NANA.

Hygstedt (1908) isolated from cow's colostrum three fractions contain­

ing NANA-Lac, two of which were the same as those identified by Schneir and

Rafelson (I960) and termed as mono-NANA-Lac while the third one, containing

two NANA molecules, was termed di-NANA-Lac. Ho avoided the use of dialysis

and used two-plmse extraction in combination with a gel filtration step as a fast

and mild technique for combined dclipidization, deproteinization and desalting

of the biological material.

b. Sialic Acids

A common pructiso in use to liberate sialic acids is mild acid hydro­

lysis at temperatures ranging from 40 to 80°C (Zilliken and Whltehouse, 1958).

Warren (1959) used hydrolysis in sulfuric acid solution of pH = 1.5 at 80#C for

1 hour. Inherent problems in dissolving casein in sulfuric or hydrochlorio acid

prompted Aprahamian (1973) to use formic acid as a solubilizing and hydrolyzing

agent. The analytical results suggested that this medium offered some protec­

tion on the sialic acid against degradation in comparison with the hydrolysis in

sulfuric acid. 16

There are several reasons for the easy removal of sialic acids from glycoproteins by mild acid hydrolysis. Neuberger and Marshall (1966) have re­ ported that: "First of all, NANA resembles other 2-deoxysugars, the glyco­ sides of which are hydrolyzed between 500 and 1,000 times faster than the cor­ responding derivatives of glucose. Secondly, glycosides of N-acylneuraminic acid are kctosides, which might be expected to be hydrolyzed more rapidly thnn aldopyrnnosidcs. Finally the pH of N-acetylneuraminic acid is 2.6 and thus in 0. IN H2SO4 about 2% of the acid is ionized. The unionized carboxyl group may bo expected to inhibit protonation on the glycosidic oxygen which is very important in tho hydrolytic renction."

After removal of proteins and lipids, the sialic acids may be isolated with the aid of column chromatography. In most cases, the hydrolysates are neutralized with barium hydroxide, and Lhe precipitate is removed. The solu­ tion is then passed through a cation exchange resin in H+ form, usually DOWEX

50 or Amberlite IR-120. Tho purposo of using the cation exchange column is to remove residual cations and small amounts of ninhydrin positive compounds

(Whitehousc and ZilliUen, 1960). Howovor, material precipitated by tricloro- ncctic acid can pass through both the Ambcrlito and DOWEX resins without ab­ sorption (Clark et a l., 1962).

After tho cation oxchango column treatment, the solution is passed through an anion-exchange resin, commonly the DOWEX-1 In the form ate or ace­ tate form. Anion exchanges are very useful in separating N-acetyl and N-glyco- 17

' lyl-ncuraminic acids. However, higher acylated forms, such as N-acetyl-O-

acetylneuraminic acid, arc not isolated since O-deacylation reportedly takes

place.

3. Stability Toward Acid and Heat

The number and position of the functional groups of the sialic acids

are responsible for the difference in the chemical reactivity of these substances.

Notably, instability becomes an outstanding characteristic of the sialic acids

(Blix and Jcanloz, 1909; Noubergor ct a l,, 1966; Tuppy and Gottchalk, 1972).

Hie sensitivity of sialic acids to mild acid treatment, contrasting with

the acid resistance of D-glucosamino and D-galactosaminc, can be rationalized

if thoir nature as dcoxysugars with tho dcoxy group vicinal to the carbonyl (keto)

is taken into consideration (Noubergor and Marshall, 1966). For example, 2-

Deoxyhcxoscs arc known to be very sensitive toward mild acid treatment, and

have a much greater tendency to change from tho cyclic forms to tho open chain

form than have ordinary nldohcxoses and it Is this change which initiates sugar

degradation by ncid or alkali.

As to the mechanism of acid degradation of NANA, Gottchalk (1960)

(Figure 2) has proposed as first steps cycllzatlon to form the A -pyrrole deriva­

tive (b) (tho internal Schiff base of neuraminic acid) and dehydration of (b) to

form 5- tclrnhydrohybutylpyrrole-2-cnrboxylic acid (c). Gottchalk based his

proposition on the observation that the methylglycosido of neuraminic acid (meth-

oxylnouraminic neid) on heating in 0. IN If Cl for 5 min. at 80°C is changed to a FIGURE 2

PYRROLE DERIVATIVES ARISING FROM NANA UNDER THE ACTION OF ACID (Gottchalk, I960)

18 HooC-C. CH*CHoH-CM0H-CHOK” CH|PH

acid

I-CH d H-CH o k -CHOH-CH j W

acid

H ooo-H ^ ^/JJ-Ck1-e6-CHoH'CK1OK 2 0 compound with an ultraviolet absorption spectrum closely resembling that of pyrrolc-2-carboxylic acid. The compound also gave with p-dimethylamino- bcnstaldchyde (P-DAB) tho purple color characteristic for pyrroles (Ehrlich reaction). Compound (c) is further dehydrated to compound (d).

The rate of decomposition rises with increasing acidity and tempera­ ture. Kurkas and Chargaff (10G4) found no degradation at pH 3 upon reaction at 100°C for 1 hour. Under normal conditions of hydrolysis at 80°C and pH 1.5, other researchers found that sonic degradation can occur (Gibbons, 1962; Gib­ bons, 19G3). Heating with IN HC1 at 100°C for 3 hrs. results in the release of approximately 25% of tho nitrogen content of N-acetyleneuraminic acid in the form of ammonia (Graham et a l., 19G3; Spiro and Spiro, 19G2). Hum in forma­ tion, always associated with the treatment of sialic acid in IN HC1 at 100°C for longer periods, is most likely due to condensation and/or polymerization of un- snturatcd degradation products of the pyrrole derivatives formed as outlined above. Upon boiling with 12% HC1, sialic acids, like other 2-ketoaldonic acids, undergo decarboxylation (Tuppy and Gottchalk, 1972). Investigations on the de­ struction of sialic acid (Jordan and Lohr, 1962; Kicrmcier and Freisfeld, 19G5) in milk resulted in tho conclusion that even after sterilization under industrially ♦ • common conditions, the sialic acid content was not reduced.

4. Identification

Chemical and physico-chemical methods can be used to identify the sialic acids. 21

An unambiguous identification of sialic acids can only be done if enough material is available for crystallization. Sialic acids can be crystallized from purified samples (Blix, 1058; Zillikcn et a l,, 1956). However, methylation during crystallization and very small amounts of impurities can prevent crystal­ lization. If crystallization is achieved, the dried crystals can be analyzed for total acetyl, O-acctyl (Hestrin, 1040) and glycollyl (Tettamanti ct a l., 1965) groups. The decomposition point and specific optical rotation may also be de­ termined (Ncubergcr ct a l., 1066).

Tho infrared spectra of sialic acids show differences which, although small, arc sufficiently distinct for purposes of identification (Hakomori and

Terunobu, 1969). Nuclear magnetic resonance spectra of N-acetylneuraminic acid have also been published (Blix and Jcanloz, 1969), X-ruy powder diagrams may also be used for the identification of the individual sialic acids. A short­ coming of the latter method is that its resolving power in the case of mixed crystals (for instance, mixture of N-acetyl and N-glycollylncuraminic acids) is limited (Noubergor and Marshall, 1966).

For tho qualitative estimation and for assessment of purity of small a- mounts of sinlic acids, extensive use has been made of paper chromatography

Svcnnerholm and Svennorholm, 1358; Warren, 1960; Warren and I'elscnfeld, 1962;

Whitchousc and Zilliken, 1960; Zilliken and Whitchouse, 1958) and thin-layer chromatography (Granzor, 1962; Fuillard and Cabezas, 1963; Schaucr and

Faillard, 1968). However, separation and identification of mixtures containing 22 sialic acids, amino sugars and simple sugars by the above methods might not be successful. Gas chromatography is feasible after prior conversion of sialic acids or their methyl kotos ides to trimethylsilyl derivatives (Clamp et a l.,

1967; Craven and Gchrke, 1968; Sweeley et a l., 1966).

Several color reactions arc availablo which can bo employed to detect sialic acids In biological sumplos. The Ehrlich, orcinol, recorctnol and pcr- iodnto-TBA reactions (l\ippy and Gottchalk, 1972) have proven particularly val­ uable. However, the periodate oxidation Is tho most valuable because tho con­ sumption of the oxidant by N-acylncuraminic acids is strongly Influenced by tho presence and site of attachment of O-acctyl groups. The mechanism of tho reaction involves the rapid consumption of 2 moles of oxidant, which is due to scission between C-8 and C-9 and between C-7 and C-8 of the sialic acid mole- culo.

In neutral and in strongly acidic solutions, three moles of periodate are quickly taken up. Tho final oxidative breakdown of NANA by periodate occurs between C-6 and C-7, and results in the formation of tho chromogen of the per­ iodate-TBA reaction. In agreement with this mechanism, N-acetyl-7-O-acetyl and 7,8-di-O-acctyl neuraminic acids do not give positive TBA reactions in con­ trast with the N-Acetyl-4-O-aeetylneuraminic acid. Only sialic acids with an unsubstituted glycosidic hydroxyl group are rcactivo in tho poriodato-TBA test.

This assay is, therefore, suitable for the differential determination of free sinlio acid in the presence of bound sialic acid. It is notable that the TBA-assay 23

for sialic acids is affcclcd by even slight modifications in the structure, as can

be brought about by acid or heat, and generally results in decreased intensity of

the color reaction.

5. Biosynthesis

There 1ms been a tremendous increase in our knowledge with respect

to the metabolism of amino sugars from studies conducted in recent years. The

results of these studies have been impressive and considerable detailed infor­

mation is available regarding the pathways for the synthesis and tho enzymatic

reactions involved in the degradation of tho amino sugars. There are, however,

still a number of questions that need to bo answered in order to understand more

about their specific function. Figure 3 is a summary of the reactions by which

amino sugars are derived from glucose (Warren, 1972).

In respect to tho biosynthesis of sialic acid, Comb and Roseman (1958,

1960) speculated that tho enzymatic cleavage of NANA by aldolase from Vibrio

cholerne. which can bo reversed, might signify a possible route of synthesis.

However, the biosynthetic function of NANA aldolaso remains in question be­

cause the enzyme is present in both V. cholerae and Clostridium perfringens

and NANA has not been dcmonsti'utcd in tho above two organisms (Brunetti ct a l., 19G2).

Warren and Fclsonfcld (19G2) later studied tho reactions involved in

tho biosynthesis of NANA in rat liver and bovine submaxillary gland. They

found the following three reactions (a, b, c) to be involved in the biosynthesis of NANA: FIGURE 3

METABOLIC PATHWAYS FOR THE INTERCONVERSION OF AMINO SUGARS (Warren, 1972)

24 25 Glucose

+H20 ATP -P*

± P i Glucosc-G-P* ►Glucose-1-P » - • Glycogen

Fruclose-G-P

Glutamine NH3

+H20 -P< Glucosomlno-l-P- GlucoBnmine-G-P . V " m Glucosamine ATP HUTP Ac Go A Ac Co A -PP +H20 - P i UDP-glucosaminc N-Acctylglu osamine-6-P » N-Aoetylglucosnmine ATP

N- Acetyl glu cosam Inc-1- P

+UTP -PP

UDP-N-acctylglucosaniino

UpP-N-acctylgnlaclosnmlne

+ATP N-Acetylmanno- N-Acetylmannosamine samfno-G-P -P i +PEP -Pi Sialic acid-9-P

+PEP

Sialic acid Mg+V N-Acctyl-D-mannosaminc + ATP - — (a) N-acctyl-D-mannosamine-G-P + ADP

N-Acelyl-D-mannosamine-G-P + phosphoenolpyruvate (b) GSH, KCN + HgO '■» N-acelylneuramtnic-9-P + Pj Mg4"**

M g** N-Acctylncuraminlc acid-9-P + HgO -- '•* (c) GSH

N-acetylncuramtnic acid + Pj

N-Acotyl-D-mannosamine + phosphoenolpyruvate + ATP

+ 21^0 ■■ — ♦■N-ncctylneuramlnlc acid + ADP (d)

+ 2Pj

In Mils casot Uic equilibrium of tlio three reactions strongly favors the forma­ tion of NANA. More detailed studies on tho enzymes involved in the above reac­ tions followed later (Kundig et a l., 1966; Watson et al,, 1966).

Another mechanism by which NANA is synthostzed in bacterium Neis­ seria meningitis, which produces a capsule consisting largely of NANA, is given by roaction (c) below (BlncUlow and Warren, 1962; Warren and Blacklow, 1962).

Hero, preliminary phosphorylation of N-acctyl-D-mannosamine does not take place and indeed the U-phosphate is inactive. The reaction is essentially irreversible. 27

N-acctyl-D-mannosnmine + phosphoenolpyruvate

Mn+4 (e) + 1J20 ------►•NANA + P£ -SH

In vivo experiments have demonstrated that ^C-galactose, ^C-glucose and ^C-glucosamine can be incorporated into tissue sialic acids (Brunneti et a l., 1962; Kohn ct al, f 19G2). Eichberg and Karnovsky (19G3) studied the over­ all pattern of biosynthesis of the sialic acids Jn vitro by a system that might still retain sonic functional relation to the situation Jn vivo. The tissue chosen by the above investigators was bovine submnxillnry gland, which is rich in sialic acid. It was found that tho specific activity of the free, dialyzable sialic acid formed was 50 times greater than that of the bound, and was most probably its precursor. They supported tho concept that the Incorporation of NANA into macromolecules involves dialyzable NANA containing intermediates of high meta­ bolic activity.

Cyttdtno G'-monophospho-N-acetylneuramtnic acid (CMP-NANA) w a B first isolated from JE. coli K-235 by Comb et al. (1959). The biosynthesis of

CMP-NANA takes placo according to reaction (f) in hog submoxillary gland

■(Koan and Rosoman, 19GG) and in N. meningitis (Blncklow and Warren, 19G2;

W arren and Blncklow, 19G2).

CTP + NANA ►CMP-NANA + P P (f) -SH The purified enzyme required Mg**ions and gluthalhlone for activity. No other nucleotide can replace CTP in this reaction. The NANA of CMP-NANA can be transferred to appropriate acceptors to form the non-reducing terminal sugar of

NANA-Lac (Bartholomew ct a l., 1973; Carlson et a l., 1973) and glycoproteins

(O'Brien et a l., 19GG; Spiro and Spiro, 1968). Since NANA appears to be linked to CMP through n 0 linkage and is linked to other sugars through a neuramini­ dase sensitive ct -linkage, transfer of the NANA x'csidue must involve inver­ sion at C-2. In addition to CMP, UDP is another nucleotide which may be of im­ portance in the incorporation of sugars into glycoproteins. Figure 4 illustrates the biosynthesis and metabolism of NANA (Warren, 1972).

Although considerable work has been accumulated on the biosynthesis of sialic acids in recent years, little work has been done on assessing the relative importance of various pathways in the cell. At tho present time, wo can only con­ struct a scheme of reactions from incomplete information derived from studies on bacteria and animals where the coll has been destroyed and .much of the meta­ bolic balance has been sacrificed. Very little can also be said nbout the control mechanisms (e.g., feedback inhibition) of amino sugar metabolism or about pool sizes of free NANA.

6. Significance of Sialic Acids

a. In Various Glycoproteins

Glycoproteins are widely distributed in tho cells of vertebrates, inverte­ brates, plants and microorganisms, and have a variety of biological functions.

However, little is known nbout the significance of the carbohydrate moiety. FIGURE 4

METABOLISM AND BIOSYNTHESIS OF NANA (Warren, 1972)

29 30

Glycoproteins +Ho0 Gangliosidcs, etc.

-UDP

UDP-heterosaccharide-NANA'

-CMP

/ -CMP-nCMP Colominic acid (NANA). CMP-NANA Heterosacclmrido-NANA

+CTP -PP

NANA-9-P NANA

+PEP +PEP +Pyruvntc -P i -P i

+ATP -UDP N-Acetylmnnno- • N-Acetylmannosamine UDP-N-acetylglucosamino samine-G-P +H2O -P[

+ATP N-Acetylgluco- t—^ N-Acetylglucosamine samine-G-P +li2° "Pi 31

Winterburn and Phelps (1972) have stressed the need for using inductive reason­ ing in order to explain the occurrence of glycosylation of certain proteins. If we arc to ascribe n function for the carbohydrate moiety of the protein, it is a prerequisite that wo understand the complete biological function of the entire molecule. For example, carbohydrate residues can bo removed from plasma glycoproteins without affecting transporting functions (Blumbcrg and Warren,

1961; Muldoon and Wcstphnl, 1907) or tetrary structure (Schmid and Kamtyama,

1963; Oshlro and Eylar, 19G9). Nevertheless, the doubt still romalns whether tho only biological function of a protein such as transferrin is to transport iron.

Romovnl of NANA from k -casein docs not destroy its stabilizing ability; how- evor, we do not know if stabilization is tho only purpose of k -casein in milk.

Recent studios (Gyorgy et a l., 1974) have attributed a protective role of NANA by which this constituent protocts glycoproteins from degradation by certain bacteria.

Tho rolo of sialic acids has been studied in a number of biological materials, including the gastric juico. The glycoproteins of tho digestive tract are involved in coating the foodstuffs to facilitate tho transport of tho digest and in protecting tho cpilholia against meohanical and chemical injury. In the intestine, the protection of glycoproteins against attack by proteolytic enzymes, such as trypsin and chymotrypsin, seems to be duo to the constituent sialic acid residues (Gottchalk el al, lOGOa), It has also been shown that, after release of sialic acid from epithelial cells, their susceptibility to proteolytic enzymes is 32 increased, which in turn results in necrosis (Vclican et a l., 1970).

Sialic acid has been found to be a component of the zona pellucida mem­ brane of the mammalian ovum. After enzymatic removal of the sialic acid resi­ dues from the surface of the rabbit ovum, a loss of elasticity of the ovum was observed, and the passage of the spermatozoa through the zona pellucida was inhibited (Soupart and Clewc, 1965).

In most cases, the activity of enzyme glycoproteins is not impaired by the enzymatic removal of their terminal neuraminic add residues. However, sialic acid wns found to have a stabilizing effect on tho activity of a -galactosl- daso (Falllnrd and Schaucr, 1972). The activity of tho native enzyme decreased over a period of 6 days at 4°C by no more than 10%, the activity of tho neura­ minidase treated enzyme dropped within 42 hours to about 25% of the original, and within 6 days to a few percent of the original.

Falllnrd and Schaucr (1972) reviewed a number of biological functions attributed to sialic acid in abnormal cells. Glick and Glthcns (19G5) found that sialic acid at the surface of L 1210 mouse leukemia cells participates in the transport of potassium ions. Removal of sialic acid from the cells Inhibited specifically the trnnsport of I<+, regardless of tho direction of flow. The develop­ ment of massive tumors is prevented when methylcholnnthrene-induced sarcoma colls are treated with neuraminidase prior to the intrnperitoncal injection into rats (Currie and Bngshawc, 1909).

There is a belief that the main role of sialic acids is the contribution 33 to the charge of the molecules where they are bound. Although this might be true for Birmll size molecules, this is not a sufficient explanation for their pre­ sence in large molecular weight molecules where the contribution to charge may be minimal (Winterburn and Phelps, 1972).

b. In Casein

K-casein is the main glycoprotein fraction of casein and contains rela­ tively large quantities of sialic acid (1 - 2%), The behavior of k -casein in electrophoresis shows that it is not a single substance since after treatment to break disulfide bonds, 7-8 discrete bnnds have been found (MncKinlay and Wake,

1964; Schmidt, 19G4). MncKinlay and Wake (1965) found that tho concentration of sialic acid present in each fraction increased with the degree of electrophoretic mobility. The fraction that gave tho strongest band on electrophoresis was the slowest moving and wns devoid of carbohydrates. Schmidt et al. (1966) also found that the removal of sialic acid altered tho mobility of the faster-moving fractions but it has yet to be shown whether the presence of other carbohydrates may contribute to mobility.

That carbohydrates are not essential for rennln action was demonstrated

(MncKinlay and Wake, 1965) by the finding that both SCM- k -casein containing carbohydrates and SCM- k -casein devoid of carbohydrates were sensitive to rennin. There is still controversy in respect to the cxnct role that sialic acid plays on rennet sensitivity of casein. Many workers have found that the clotting time is increased when sialic acid is removed by neuraminidase (Gibbons and 34

* * Chccscman, 1962; Schmidt ct a l ., 1966; Whccloclc and Kntght, 1969). On the

other hand, other workers have shown that removal of sialic acid or carbohy­

drates did not delay rennet action (Armstrong et a l., 1967; MacRlnlay and Wake,

1965).

There arc some indications that sialic acid has some role in the forma­

tion of cnscin m icelles. Thompson and Pepper (1962) removed NANA from k -

casein by neuraminidase and found that the stabilization capacity (amount of 01 s-

ensein stabilized by a given amount of k - casein) was considerably reduced.

The sialic acid presumably also gives rise to differences in the repulsion forces

surrounding exposed k -casein molecules of different carbyhydrate content. Thus,

It seems possible that random accumulation of cither more k - casein or k -

casein of nbovo average carbohydrate content on a micelle would restrict fur­

ther growth of that micelle (Rose , 1969).

7. Nutritional Aspects

Bovine colostrum and human milk are well known to be rich in sialic

acid (Cabczns ct nl., 1968; Guminsld, 19G7b; Kumar, 1972; Zilliken ct al.,

1956). Because of the possible biological significance of sialic acid in infant

nutrition, Cabezns, ct al. ( 1069 ) , studied the nature and tho concentration

of the sialic acids existing in eight dictctlcal products (matornalizcd milks)

generally employed ns baby food in Western Europe. The yield of sialic acids,

expressed as NANA, varied from one product to another, between 22 and 62

ing/100 g of product,. If their mean concentration (expressed as mg/100 ml 35 of reconstituted milk) is compared with that of cow, it may be observed that it is relatively poor. The difference is bigger if their concentration is compared with that of human milk. On account of the physiological importance of sialic acid in the development of human infants, studies were also conducted (Jordan and Lohr, 19G2; Kicrmcicr and Frcisfeld, 19G5) in order to examine whether or not milk processing treatments actually in use would affect the sialic acid content of market milk. Although there is not absolute ngreement in these re­ sults, bolti studies indicate that under industrlally-common conditions, the sialic acid content is not reduced. In contrast, recent studies (Kumar and Hansen,

19V2) have shown that free N-acctylhexosamines in milk are adversely affected by heat treatment.

Tho high content of sialic acid in bovine colostrum and human milk, as well as the wide-ranging effects exhibited by various coll typos after neuramini­ dase treatment is evidence that sialic ncids might well be Important in infant nutrition. Thcroforo, any attempt to make roadility available, edible sialic acid could be of value. Mncno ct al. (1968) devised a method for obtaining a product containing NANA from the roots of "devil’s tonguo," Tho product was intended as an additive to cow's milk for feeding babies. However, there is a need to know more about the fate of sialic acid after ingestion. Although free sialic acid may serve ns a precursor for incorporation Into polymers, it is not certain that dietary sialic acid may survive the intestinal tract. 36

B, Growth Promoting Factors of Lactobacillus bifidus variant Pennsyl-

varicus

Lactobacillus bifidus var. PennsylvanicUB is classified by Tissier

(1900) as a non-motile, gram-posltlve, strictly anaerobict non-spore-forming rod. It has been reported (Brown and Townsley, 1970; Weiss and Rettgei;

1934) that after primary isolation, L. bifidus grows under ordinary aerobic conditions in pure culture, and it Is suggested that it should be classified as a facultative anaerobe, However, Vries and Stouthamer (1968) reported that the organism cannot grow aerobically on agar plates. Tho unique predominance of

L. bifidus flora in the intestinal tract and in the feces of breast-fed infants fol­ lowed by the relatively low pH of tho large bowel hns been associated with the natural resistance to cntoric infections exhibited by the breast-fed infants in contrast to those fed cow's milk formulas (Bullon and Willis, 1971). It is sug­ gested that the responsible factors in breast milk include its high lactose, low protein, low phosphate content, together with tho presenco of some growth- promoting substances, the so-called "bifidus factors", Jao (1974) has reviewed studies related to tho growth-promoting factors of L. bifidus. Most growth fac­ tors so far known represent low molecular weight compounds. However, higher molecular weight undialyxnble oligosaccharides and glycoproteins have also been described,

1. Low Molecular Weight Compounds

It has been shown by Gyorgy et al. (1954a) that human milk contains an essential growth factor for L. blfidus. The activity of human milk for the growth of L. bifidus was compared by Gy orgy et al. (1954b) with that of other species.

The average relative activity was highest for human colostrum, closely followed by rat colostrum, thou by humnn milk, rat milk and cow's colostrum. The milk of ruminants (cow, ewe, goat) showed only very slight activity. Somewhat higher activity was found in the milk of cat, monkey, dog, donkey, rabbit, mare and sow. The bifidus-active components of human milk and colostrum can be sepa­ rated (Gyorgy ct a l., 1954c) into a dlalyzablc fraction representing from 40 to

75% of the overall activity and a relatively non-dialyzablo fraction of 25 to 60%.

The growth promoting substances have been identified as N-acctyl-D-glucosamlno containing compounds, preferably in /3-D-glycosidic linkage (Rose et a l., 1954).

Zillikcn ct a l., (1955) confirmed those observations and bcsidos found that the ethyl and n-propyl N-acetyl-^ -D-glucosamtnides wore more active than the methyl compound. It has boon shown that N-substitutcd D-glucosamino acts as a direct precursor for the biosynthesis of muramtc acid in the cell walls of the bacterium (O'Brien ct a l., 1960), Obviously, L. blfidus var, Ponnsylvanicus is not capable of supplying sufficient amounts of D-glucosamlne from its ana­ bolic processes to synthesize its cell wall. For this, it has to rely on external sources.

Gyorgy (1957), in his review on the N-contatning saccharides in human milk, has tabulated compounds isolated from humnn milk according to their acti­ vity. Lnctoso and N-freo trlose (fucosldo-lactoso) wore microbiologically 38 inactive, while there was great variation among the N-containing oligosaccha­

rides, Lambert and Zillikcn (1965) described 24 new growth factors and drew useful conclusions as to the nature of a "blfidus factor". The above investiga­ tors found that other N-acctyl derivatives of D-glucosamlnc also promote growth.

The growth promoting activity increases with chain length of the aliphatic N- acyl residue. The high activity of N-bcnzoyl-D-glucosamtne is either diminished or abolished upon substitution in the benzene ring. Glycosides of N-acetyl-D- glucosnmine exhibit similar activity as the previously described yj -O-glycosides, but £ -glycosides with N-substitutlon other than acetyl do not promote growth.

The requirement for a highly active growth factor for L. blfidus vnr. Pennsyl- vanicus may, therefore, be specified as follows:

(a) the substance should be derivative of D-glucosaminc;

(b) the substance should be able to penetrate the bacterial

coll membrano readily; and

(c) the substance should be susceptible to the enzyme system

present injj. blfidus vnr. Pennsylvanicus. which is re­

sponsible for hcxosamino metabolism and cell wall synthesis.

Different types of "blfidus factors" not containing glycosamine deriva­ tives have also been reported. Hirata (1958) has studied the effects of lactose and lactulose on the microbinl flora in tho digestive tract of the infant with in

vivo experiments. He found that the blfidus flora in the stool of the infant is remarkably Increased w'licn lactulose was present. Although the mechanism of lactulose utilization In vivo was not established, it seems reasonable to con­

clude that this carbohydrate plays an important role in promoting growth of L,

blflduB in infants. Lactulose (£-galactosidofructose) Is formed in heated milk

products and Is n degradation product of lactose, Adachi and Patton (1061) have

reviewed studies on the presence and significance of lactulose in milk products,

Neimann et al. (I960) have made nn experimental feeding of infants with half

skimmed milk containing lactulose. Most of the experimental infants had feces

resembling U i o b c of breast-fed infants. Capsules containing lactulose have been

used successfully to treat infants carrying pathogenic Strains of Escherichia colt

which hnd failed to respond to antibiotic treatment (Schncegans et a l., 1067).

Peptones produced by the peptic or tryptic digestion of insulin, trypsin and cortatn other proteins also contain a factor recognized for the growth of var­

ious strains of Lactobacilli (Haurowitz, 1950). Tills factor, called strcpogenin,

is possibly formed in the intestines and might be of importance for the growth of s ■ intestinal bacteria. The composition of strcpogenin haB not become known, but

it can bo rcplnccd to a certain extent by seryl-glycyl-glutamic acid (Woolley,

1946a; Woolley, 194Gb). It is assumed, therefore, that strepogenln has a com­

position similar to that of tho tripeptide.

These findings explain the fact that the biological value of amino acid

mixture can bo different from the value of a protein containing an equivalent

amount of these amino acids, since intermediate degradation products of protein

could be of biological significance. Kisza and Ziajka (1973) studied the growth 40

•of five L. blfidus strains using three different media: cow's milk, cow's milk

with 20% pepsin digested milk added, nnd cow's milk with 2% lactulose syrup

added. All strains studied showed the best growth in media with pepsin digested

milk added, nnd their activity was highest in that medium. The slowest growth

was observed in milk alone.

2. High Molecular Weight Compounds

Gyorgy ct al, (1074) found that N-acetyl-D-glucosam ine containing

acidic oligosaccharides and glycoproteins become growth factors for L. blfidus

var, Penn, only after treatment with neuraminidase. This indicates that the

sialic acid residue linked to a carbohydrate chain exerts a protective effect

against degradation of such compounds by certain bacteria. In addition, it sug­

gests that L. blfidus vnr. Penn, lacks the neuraminidase that cleaves definite

glycoBidic linkages between sialic acid and the carbohydrate chains of glycopro­

teins or oven cell surfaces and membranes. Hence, the microorganism is un­

able to liberate the N-acetyl-D-glucosamine residues from these sources for

use in its coll wall synthesis. However, the amount of carbohydrate present in

the glycoprotein is critical and the transferrin with low carbohydrate content

(5.5%), treated or untreated, was completely innctive,

Tomazelli ct nl. (1954) tested a wide variety of substances and found

hog gastric mucin to be the most active sourcet for growth of L. blfidus. The

microbiological activity was attributed to a disaccharido which resides on the

mucin and which has been tentatively identified as galnctose-acetyl-glucosamide,

Inove and Nngayamn (1971) used pig gastric mucin in infant formulas made from cow's milk and found that the addition of the mucin increased the weight of low birth weight infants, as well as the count of L. blfidus in their feces. In the above studies, the authors do not comment about the presence of stalic acid in the mucin. III. SCOPE OF INVESTIGATION

The atm of this study lias been to devise a method for production of biochcmicals of high biological value from commercial casein and to determine the physico-chemical properties of the residual casein. The specific objectives were:

1. To develop a practical method for isolation of edible stallc add from commercial casein nnd to study its properties.

2. To characterize a substance from casein which was obtained during the isolation of sialic acid nnd which promoted tho growth of Lactobacillus blfidus.

3. To examine the solubility characteristics nnd viscous behavior of the residual casein and to ns'scsB its potential use as n casein replacement in food manufacturing.

42 IV. EXPERIMENTAL PROCEDURES

A, M aterials

1. In this study, a supply of commercial acid casein of French manu­ facture was obtained from Ross Laboratories, Columbus, Ohio. The product had been ground to 30 mesh particles and had a moisture content of approximate­ ly 7 .8%. Tills type of casein was used in most of the studies. However, in a few instances, two other types of casein were used;

a. Whole casein was made in the laboratory from raw

cow's milk by the usual isolcctrie precipitation which

was repeated twice to insure freedom from other

milk components. The acid casein was stored frozen

until required,

b. Acid casein was also purchased from Fischer Scientific

Company.

2. The hydrolysis of casoin and the subsequent concentration of the hydrolysntes were conducted in a 10 liter capacity all-glass rotary evaporator

{Buchi H-10, Rinco Industry Co.). The vacuum was created by a water aspira­ tor. For concentration of small amounts of hydrolysate, a laboratory model rotary evaporator was used.

43 44

3a. TJie UV absorption of the effluents was continuously monitored by an LK13 Unicord Typo 4701A control Unit.

3b. Tho conductivity was measured by a conductolyzer LKB Type

5300 B using a dicell.

3c. For the collection of the fractions, the following were used: fraction collector controller Type 3403 B, distributor Type 3402 B, and a lube rack holder Type 3415 A. All these pieces of equipment were products of LKB-Productor AB - Stockholm 12, Sweden.

4. Spcctrophotometric measurements were made with a Perkin-

Elmer Coleman 124 double-beam spectrophotometer.

B. Methods

1. Liberation. Isolation and Purification of Sialic Acid

a. Five hundred g of casein were hydrolyzed with 5 liters of dlluto U2SO4 at pll 1.3 - 1,0 for 1 hour at 80°C.

b. Tito pH of the cooled hydrolysate was adjusted to 4.6 with

Ca(OH)2.

c. The hydrolysate was filtered through a cheesecloth and tho residual casein was reclaimed.

d. Tho residual casein was washed twice with water to reclaim any sialic acid trapped in tho curd.

e. Tho washings were combined with the filtrate and concen­ trated by rotary evaporation to 100 - 200 ml. 4 5

f. Hie concentrated hydrolysatc was centrifuged and the preci­ pitate was discarded.

g. The pH of the supernatant was adjusted to 7.0 with Ca(OH)2 in order to remove sulfate in the form of CnSC>4.

h. The neutralized hydrolysate was concentrated to ~ 30 ml. and clarified by centrifugation.

i. Tho hydrolysatc was passed over a DOWEX-1, X 8 (2 x 23 cm) column In the acetate form, followed by washing with *^150 ml of H2O.

j. The NANA was eluted with a gradient of 2M C^COONa at a conductivity of 2 x 10"** mhos.

k. Hie "direct Ehrlich" positive oluates were combined and passed through a cution exchange column, Duolite C-20 (2.5 x 20 cm) in the hydrogen form.

The column was then washed with water until the effluents were negative to "direct

Ehrllch".

1. The effluents containing tho NANA were freeze-dried.

2, Direct Ehrllch Tost

Tho "direct Ehrlich" reaction was used as a qualitative test, and was car­ ried out by U10 W erner and Odin (1952) method,

a. Hcagcnt

5% - p - dimothylamlnobenzaldchydc in 50% HCl (V/V in H2O).

b. Procedure

i. To the test substance (>*.2 ml) in a screw-oap test tube, 46 add H20 to volume of 5 ml.

ii. Add 1 ml of Ehrllch*s reagent.

ill. Heat In boiling water bath for 30 m in., cool in running w ater.

iv. A violet color was an indication of sialic acid. Tho test was not positive for other sugars.

s 3. Colorimetric Determination of Sialic Add

The W arren (1959) TDA method was used in this study. Whenever the sialic acid was not in free form, e.g. whole casein, acid hydrolysis was employed to liberate It. However, in a few instances, enzymatic hydrolysis was also used.

a. Reagents

i. Periodate Solution - prepared by dissolving 4.28 g of sodium mcta periodate in 40 ml of distilled water, and adjusting the volume to

100 ml with 85% HgPC^.

it. Arsonito Solution - prepared by dissolving 10 g of sodium arscnite, and 7.1 g of sodium sulfate in 0.1N HjjSO^, and adjusting the volume to 100 m l.

iii. TBA Solution - prepared by dissolving . 6 g of 2-thiobar-

♦ bituric acid and 7.1 g of sodium sulfate in distilled water and adjusting the vol-

t ume to 100 ml.

b. Procedure

i. Sufficient sample (5-10 mg) was dissolved in 50 ml of water to obtain a solution containing . 05 moles of sialic acid in a volume of . 2 ml,

ii. To a .2 ml sample, .1 ml of periodate solution was added, the tubes were shaken and allowed to stand at room temperature for 20 min.

ill. At the end of the incubation period, 1.0 ml of arscnite solution was added to destroy excess periodate.

iv. After the addition of 3.0 ml of freslily-prepared TBA solu­ tion, the tubes were capped and placed in a boiling water bath for 15 min.

v. Tho tubes were removed from tho water bath and cooled in ice water for 15 min.

vi. Hie chromophore was extracted with 4.3 ml of eyclohexa- none, and cooled in ice water.

vii. The solvent layer was clarified by centrifugation for 5 min. at 1000 x g.

viii. 'Hie tubes were allowed to reach room tompernturo and tho absorbance of tho chromophore was recorded against a blank containing water treated in the same way.

o. Hydrolysis In Sulfuric Acid

i, • 100 mg of dry casein or casein bifidus factor (CBF) was suspended in 4.5 ml II20 and , 5 ml of 1. ON HgSO^..

ii. The tube was cooled and 8 ml of 1. ON NaOH solution was added to bring tho pH to 4.3 and precipitate most of the protein. 48

iti. The tube was mixed thoroughly and its contents filtered thoroughly through Watman No. 43 filter paper.

iv. Sialic acid content was determined on ,2 ml aliquots of the supernatant by the TBA assay.

d. Enzymatic Hydrolysis by Neuraminidase

i. Insoluble neuraminidase (from C. pcrfringens attached to agarose suspension in 2.0M NII4SO4 solution, pH 7.0, Sigma Chemical Co., St.

Louis) was suspended in 10 ml buffer solution containing 60 mg of CBF.

ii. The suspension was incubated at 37°C for 60 min.

iii. The sialic acid content was then determined by TBA assay.

4. Paper and Thin-Layer Chromatography (TLC)

a. Standards (purchased from Calbiochem, Los Angeles, Calif,)

i. N-acetylneuraminic acid, A grade

ii. N-acclyl-D-glucosamine, A grade

iii. N-acctyl-D-galactosamine, A grade

iv. Neuramlne-lactoso (beef colostrum), A grade

v. Dextrose, General Chemical Co., New York, N. Y.

vi. Lactose (U.S.P.), Allied Chemical, Morristown, N.Y.

vii. Galactose (C .P,), Pfantiehl Labs Inc., Waukegan, HI.

viii. Fructose, Fisher Scientific Co., Fairlawn, N.J.

ix. D-fucose, Nutritional Biochemicals Co., Cleveland, O.

Hie standard solutions of sugars were prepared by dissolving 5 mg of m aterial in . 5 ml

b. Paper Chromatography of Sugars

1. One-dimensional paper chromatography was carried out on a Watman No. 1 paper,

.ii. The solvent system was cthyl-acetate:pyridine: water

(2:1:1 V/V).

iii. Hie visualization of sialic acid and other sugars was made by spraying with . 5% 4-DimclhyIaminobcnzaldchyde, EM Laboratories Inc., 500

Executive Blvd., Elmsford, N.Y,

c . TLC

i. One-dimensional TLC was used on silica gel G plates activated at 100°C / 60 min.

ii. The plates were coated with a slurry of silica gel-G, prepared by suspending 25 g of silica gel G in -vCO ml H20.

iii. The plates were heated for 10 min at 100°C for color development.

d. Sugars

i. Tho solvent system employed for sugar determination had tho following composition: ethyl acctate:acctio acid;methanol:water (60:15:15:10

V/V).

ii. The spray reagent was freshly prepared each time and was a mixture of . 5 ml unlsaldehyde, 9 ml ethanol, 5 ml concentrated sulfuric acid and . 1 ml acetic acid,

c. Amino Acids

i. One-half g of sample was hydrolyzed with 5.0 ml of 8N

H2S04 at 110'C for 5 hr.

ii. After cooling the pH of the hydrolysate was brought to 4.0 with Ba(OII)2 and the precipitate was removed by centrifugation.

iii. The volume of the supernatant was then adjusted to 25 ml witli lf20.

iv. Sufficient amount of this solution was applied on TLC plates.

v. Various amino acids were used as standards at concentra­ tions o f. 4%

vi. The developing solvent had the following composition: ethyl acetate ; butanono : formic acid : water (5;3:1:1 V/V).

vii. Hie plates were developed with .1 % ninhydrln spray re­ agent (EM Laboratories, Inc., Elmsficld, N. Y,),

5, Crystallization of Sialic Acid

a. Crystallization From Solution

The method described here was based on the principles given by

Dlix (1958).

i. 305 mg of purified sialic acid (—95%) were dispersed in

1 ml H2O and 1 ml of mellmnol was added. 61

11. Upon standing in the freezer, the material was fully solu­ bilized in the II2O - methanol mixture.

iii. After solubilization, 1.5 ml ethyl ether was added and an amorphous precipitate was immediately filtered off.

iv. The preparation was again set aside in the freezer and crystallization took place with 5 hr.

b. Crystallization by Solvent Evaporation on Microscopic Slides

i. Freeze-dried, purified sialic acid preparations were dis­ solved in minimum amounts of water and allowed to dry on microscopic slides.

ii. Tlie slides were then examined under a Loitz DIalux micro­ scope, equipped with polarizing filters as well as dark field illumination.

C. Growth of L. bifidns

Tho procedures were adapted from the dissertation by Jao (1974).

a. L. bifidns vnr. Pennsylvanlcus was obtained from American

Typo Culture Collection (ATCC) with code number 118G3.

b. Media for Active Cells

i. Semi-synthetic broth-ATCC Medium 76 (containing peptone), was used. The composition is given in Table 8.

ii. Culture broth was Elliker broth (Dlfco, 0974).

0. Preparation of Stock Sample Solutions

i. Twenty five mg sample was dissolved in 15 ml double-distilled H20 .

ii. Hie pH was adjusted to 6.5

iii. Tho final volume was adjusted to 20 ml. iv. The samples were filter sterilized or autoclaved. 5 2

d. Young Active Cell Preparation

14 - 16 hour culture in Elliker Broth centrifuge (2000 rpm /10 m in.)

Cell scdimcntcd

wash with sterile saline (.85% NaCI) Centrifuge (2000 rpm /10 m in.) Repeat once.

Cell scdimcntcd

Suspend in semi-synthetic broth to 75% Transmittance at GC0 nm

Stock cell solution

e, Assay Method

i. Four ml of coll stock solution was pipetted into sterile Spectronlc-cuvctte.

11. One ml. of sample solution was then added.

iii. The tubes were incubated at 37°C and transmittance at 660 nm was read at selected time intervals. The reference blank contained 4 ml semi­ synthetic broth plus 1 ml sample solution.

iv. Tho percent transmittance was convortcd into absorbance unit.

Tho initial optical density (O. D.) value was subtracted from tho observed O. D. readings to obtain tho corrected O. D. values. 53

TABLE 8 INGREDIENTS OF ATCC MEDIUM 7G

For one liter of double-strength base medium:

K2HPO4 5.0 g Na-Acetate (anhydrous) 50,0 g Nz Case peptone (Sheffield) 10.0 g Lactose 70.0 g Alanine, Cystine, Tryptophan, each 0,4 g Asparagine 0,2 g Xanthine, Adenine, Guanine, Uracil, each 0.2 g Salt B (see below) 10,0 ml Pyrldoxtnc-HCl 2,4 mg Thtamtnc-HCl 0.4 mg Riboflavin 0.4 mg Nicotinic acid 1,2 mg Ca-Pantothenate 0.8 mg Btotin 8.0 g Folic acid 20.0 g p-aminobcnzolc acid 20,0 g Tween 80 1,0 g Adjust pH to 0.8

Ascorbic acid, 10 mg/ml, pH adjusted to 0.5, flltor-storillzcd added after medium is sterilized, 10/100 ml.

Salt B:

MgS04 7 II20 10.0 g FeS04 7 II20 0.5 g NaCl 0.5 g MnSO(1 2 H20 0.337 g Distilled water 250.0 ml 7, Total Sugar Determination

Quantitative sugar determination was mnde according to Dubois et a l,,

(1056) using phenol sulfuric acid. The values were read from a standard curve prepared with glucose,

8, Qualitative Tests for Fructose

Tho a - naphthol and urea reactions wero used for detection of fruc­ tose, according to the procedures given by Dische (1062),

9, Hoxosamtnes

The presence of hexosaminos was checked by Lcwandowski's (1973) method.

10, N-Acctylhcxosamines

Kumar's nnd Hansen's (1072) method was used for detection of N-ncetyl- hexosnmincs. In order lo eliminate any reactivity of p-DMAB with protclnous material, a sample was also used without heating it undor alkaline conditions.

11, Biuret Test

TIUs test was carried out according to the method described by John­ son and Swanson (1052),

12. Phosphorous

The ester-bound phosphate was determined by the Fiske and Subbarow method (1025), In order to climinato nny possibility of the presence of inorganic 55 phosphate, the samples were dlalyzcd against 0.5M NaCl In the presence of

EDTA.

13. Ultracontrifugatton - Sedimentation Analysis

Beckman, model E, analytical ultracentrifuge was used, with!

a . Rotor An-D, no. 3X57.

b. Temperature at 25°C.

c. Tho speed at 52,000 rpm.

d. Tliree percent sample solutions were prepared in phosphate buffer A* = 0.12 and pH G.7 (NaCl 44.5 g + Nal^PC^ • H20 15.2 g + Na2HP04

15.6 g in 10 liters H2O).

e. Sample solutions were used with or without dialysis (mem- ■ brano tubing M. W. cut off«- 3,500) for 48 hours against tho buffer solution. t

14. Ultrnflltrntion

A sample solution containing 25 mg in 20 ml of H O was passed through a a Diaflo Ultrafiltcr, UM-2 (43 mm), Amico Corporation.

15. Uron Starch Gel Electrophoresis

Tho electrophoretic procedure followed was that described by Morr

(1074). Whenever tho snmplcs wore trcntcd with rennet prior to electrophore­ sis, these were exposed to rennet for twice the length of the time required to coagulate reconstituted skiin-milU, 56

16. Rennet Sensitivity

a. To 3% sample of casein solutions (pH ~7.0), ,01M CaCl2 was added.

b. Appropriate amounts of rennet were then added to the sam ples,

c. The clotting time was recorded,

17. Calcium Sensitivity

a. Three sample concentrations of casein were prepared (1,2,3% - pH *>'7.0).

b. To each was added . 01M CaCl2.

c . Incubation was nt 32°C for 15 min.

d. Centrifugation was at 5,000 rpm for 10 min.

e. The O.D. of the supernatant was measured at 290 nm, after addition of one drop of 00% NaOII to clarify the solution. Maximum absorption occurred nt 290 nm rather than 280 nm, when the pH was higher than 11. 0 due to ionization of tyrosine (Bingham, 1971).

f. Hie supernatant protein was determined from a separate standard curve.

18. Solubility Studies

a. The solubility of casein between pH 4.0 to 5,6 at 2°C and 25*C was determined by Bingham's method (1971),

b. Tho solubility of Na and Ca caseinates in the pH range 6.0 to 57

9.0 was determined by modifying the method used by Hayes and co-workers (1969), which was developed to simulate commercial dispersion conditions of co-precipi- tatcs. The modification included the use of a Waring blender instead of Ultra-

Turrox mixer, and the calculation of supernatant protein was made spectropho- tometricnlly instead of the sediment volume.

19. Viscosity Measurements

Viscosity measurements on solutions of varying concentrations (4, 8 and 16%) were carried out by a Brookfield Synchro-lectric Viscometer, Model

RV, at different tem peratures (5, 25 and G0°C). The solutions w ere prepared by dissolving the required amount of sample by adding NaOH to bring the pH to 7.0.

20. Flavor

The flabor trials were conducted by a panel of four judges. Tho concen­ tration of casein in the solution was 3%. To mask the bitter flavor, 4% lactose was incorporated. V. EXPERIMENTAL RESULTS

The present study has been concerned with (a) ltberation of sialic acid from casein nnd its isolation nnd purification; (b) identification and crystallization of the isolated sialic acid; (c) characterization of a casein

"blfidus factor" (CBF) which was obtained during the isolation of sialic acid;

(d) determination of the properties of the residunlt sialic acid-free casein; and finallyy (c) production of crude preparation containing sialic acid and casein blfidus factor (SA/CBF).

A. Ltberation, Isolation and Purification of Sialic Acid from Casein

In the scnrch for a procedure for obtnining purified sialic acid from casein, tho basic nppronch involved acid hydrolysis of casein to liberate sialic add, stepwise purification of the soluble sinlic add in tho hydrolysate, and an examination of tho isolated product to determine its purity.

Variables investigated wore; (a) casein; (b) hydrolyzing agent;

(c) neutralizing agent; and (d) method of purification. The procedure wliich was evcntinlly selected has been described in detail in the experimental pro­ cedures section (pp. 44-45) nnd is diagrammed in Figure 5.

58 FIGURE 5

LIBERATION, ISOLATION AND PURIFICATION OF SIALIC ACID FROM CASEIN

59 6 0

500 g of casein in HYDROLYZE 5 liters of dilute 1I2S04 (pH 1.3-1.5) for 1 hour a t 80°C

BRING pH TO 4.6 WITH Ca(OH)2

[FILTER THROUGH CHEESE CLOTH — t z r 1------wash with HoO [f i l t r a t e RESIDUAL combine washings

[CONCENTRATE TO 100-200 ml |

CENTRIFUGE

PPTSUPERNATANT D iscard

BRING pH TO 7 .0 WITH Ca(OII), CONCENTRATE TO ^ 30 ml

SUPERNATANT PPT * Discard

PASS THROUGH DOWEX-1, X 8 (2 x 23 cm)

COMBINE POSITIVE DIRECT EHRLICH TUBES

| PASS THROUGH DUOLITE C-20H (2.5x20 cm)

FREEZE DRY 61

The elution diagram of the concentrated hydrolysate on DOWEX-1, X 8 is shown in Figure 6. The leading UV-absorblng material was a peptide frac­ tion which was not retained by the ion exchange column. The properties of this fraction are described in a later section (p. 82). Sialic acid was eluted at an ionic strength corresponding to a conductivity of 2 x 10 mhos. The yield by tho outlined procedure for two experiments was:

Experiment I Experiment ii

Total solids 800 mg/500g casein 1000 mg/500g casein Total NANA 760 mg 850 mg Puri ty 95 % 85 %

1. Casein

The choico of commercial casein as starting material was made, after it wns observed that fresh isoelectric casein curd remained strongly hydrated after hydrolysis and neutralization to pH 4.6. This had a negative effect on both f'c yield nnd purity of the final product because a part of the sialic acid was trapped by the wot casein and the recovered hydrolystnte contained more of the

* soluble peptides (Figure 7) than hydrolysntos from dry isoelectric casein. This contamination tntorfei'ed with the final purification. The yield of sialic acid from fresh isolcctric curd was:

Totnl solids 3,056 mg/1,OOOg casein Total NANA 1,100 mg Purity 36% FIGURE C

ELUTION DIAGRAM OF DRY 1SOLECTRIC CASEIN HYDROLYSATE ON DOWEX-1, X 8 (2x23 cm) COLUMN IN ACETATE FORM (Elution with a gradient of 2M CIl3COONn; UV absorption at 270 nm ; direct Ehrlich reaction at 865 nm .)

62 WATER SODIUM ACETATE

NANA

CBF

-TUBE NO. FIGURE 7

ELUTION DIAGRAM OF FRESH ISOLECTRIC CASEIN IIYDROLYSATE ON DOWEX-l, X 8 {2x23 cm) COLUMN IN ACETATE FORM {Elution with a gradient of 2M CIl3COONa; UV absorption at 270 nm; direct Ehrllch reaction at GG5 nm .)

64 O.D. (565 nm) P .P . (270 nm)

TUBE NO 66

2. Hydrolyzing Agent

The hydrolysis conditions in Figure 5 arc essentially those used tn

W arren's (1969) analytical procedure for determination of sialic acid in proteins.

These conditions wore chosen, although Aprahamlan (1973) has reported that for analytical purposes, IICOOH is less damaging to sialic acid during hydrolysis than H2SO4 and, furthermore, that the yields would be higher because of the more complete solubilization of casein during hydrolysis.

Experiments with IICOOH as the hydrolyzing reagent showed that this approach would be impractical for production purposes. Treatment of casein with IICOOH resulted in a viscous solution from which it was difficult to sepa­ rate sialic acid (from casein) without having to resort to large amounts of neutral­ izing agents. Removal of IICOOH by evaporation and subsequent extraction of sialic acid by solvents or water was attempted, but was deemed inefficient be­ cause of tho compact plastic nature of the dried casein. The isolation of the sialic ncid was also problematic, as shown by the much more complicated nature of the elution pattern in Figure 8 which resulted in a low purity of the final pro­ duct. The yield of the sialic ncid from IICOOH hydrolyzed casein was:

Total solids : 100 mg/lOOg casein T otal NANA : 47 mg Purity : 47 %

3, Neutralizing Agent

Tho choice of Ca(OII)Q as neutralizing agent, instead of Ba(OH)2 used FIGURE 8

ELUTION DIAGRAM OF FORMIC ACID IIYDROLYSATE ON DOWEX-1.X8 (2 x 23 cm ) COLUMN IN FORMATE FORM (Elution with a gradient made from 2N NH4COOII, adjusted to pH 4 .0 with IICOOI-I, and 2N NH4COOH, pH 7.0; UV absorption at 270 nm; direct Ehrlich reaction at 565 nm.)

67 O.D. (270 nm) UE NO TUBE NANA O DNIID MATERIAL IDENTIFIED NON

00 j c O.D. (565 nm) 69 in Warren’s procedure, was based on the consideration given to the possible edible uses of the sialic acid and the residual casein, because Ba is known to be toxic. Calcium hydroxide was preferred over NaOH, because the major part of the sulfate was removed in tho form of CaSOg, while Na2SO^ remained in a soluble form, placing a greater load on the ion exchange columns.

4. Methods of Purification

Alternative methods to the anion and cation exchange column chroma­ tography for purification of sialic ncid have been examined.

a. Extraction of the sialic acid from dry hydrolysate with 200 ml of McOII and subsequent drying gave the following results:

Totnl solids : 900 mg/500g casein Total NANA : 351 mg Purity : 39%

b. Extraction of sialic ncid with saturated Na2SO^ followed by pre­ cipitation of Na2SO^ with twice tho volume of EtOH, and precipitation of the sialic acid with ethyl ether, gave tho following results:

Total solids : 635 mg/500g casein Totnl NANA : 337 mg Purity : 53 %

c. Precipitation of impurities from the concentrated hydrolysate with 20 ml of 5% uranyl acetate and further purification by cation exchange col­ umn, gave tho following results; 70

Totnl solids 2,800 mg/500g casein Total NANA 924 mg Purity 33%

None of these methods was deemed comparable to the adopted method ‘

for the purpose of purifying NANA. However, solvent extraction may be a

useful approach in developing a large-scale extraction technique.

An additional consideration in favor of the anion and cation exchange

column chromatography was the separation of the casein bifidus factor (CDF)

which was found to stimulate the growth of L. bifidus to a high degree. So, in

ndditlon to sialic acid which was obtained in relatively high yield and high purity

by tho adopted method, another material was also separated which might possess

significant biological value.

B. Identification and Crystallization of the Isolated Sialic Acid

Although the Warren's TBA test was used to determine tho yleldB of

NANA In the isolated products, additional characterization was necessary be­

cause this colorimetric test lacks absolute specificity,

1. Paper and Thin-Layor Chromatography

Tho chromntographlo results are shown in Figures 9 and 10. TLC

gave a somewhat smeared pattern while paper chromatography gave more dis­

tinct spots. However, both methods confirmed the similarity between the pat­ terns of tho isolated sialic acid and tho synthetic NANA (N-acetyl - neuraminic FIGURE 9

THIN-LAYER CHROMATOGRAPHY OF THE ISOLATED SIALIC ACID. (Solvent: ethyl acetate/acetic acld/methanol/water, G0/15/16/10 V/V. Spray reagent: a mixture of .5 ml anisaldehyde, 9 ml ethanol, ,5 ml concentrated sulfuric acid and . 1 ml acetic acid.)

1. Lactose 2. NANA-Lac 3. NANA 4. Isolated Sialic Acid 6. CBF

71 72 FIGURE 10

PAPER CHROMATOGRAPHY OF THE ISOLATED SIALIC ACID. (Solvent: ethyl-acolato/Pyrldtnc/water, 2/1/1 V/V. Spray reagent: . 5% 4-Dimethyl- aniino bcnzaldchyde.)

1. Lactose 2. D-galacloso 3. Dextrose 4. Fucose 5. NANA 6. N-AcMnn 7. N-AcGal 8. N-AcGluc 9. Isolated Sialic Acid o - 7 5 acid). These findings do not exclude the possibility of NGNA (N-glycolyl- ncuramlnic ncid being pi’csent since no standard for NGNA was obtained. How­ ever, it is known that bovine casein contains only NANA (Cabezas, 1973).

2. Crystallization

a. From Solution

Crystallization of sialic acid from solution in the presence of McOH was achieved. Tho crystals had tho same appearance as those given by Svcnner- holm (195G) for NANA. Similar crystals were obtained for NANA isolated from bovine casein by Dumas and Alnls (19G1). However, tho crystals obtained in this experiment were very fragile and were not used for microscopic studies.

b. By Solvent Evaporation

Direct examination of the isolated freeze-dried material with polarized light revealed the presence of some material showing bircfrigence, among amor­ phous particles. The material showing birlfrigenco was of crystalline nature and this was of importance, since it gave indication of the high purity of the iso­ lated sinlic ncid.

Improved crystallization of these preparations of sialic add was achieved by evaporating a water solution on a glass slide, and Figure 11 shows two types of btrefrigent crystals. One type had the shape of needles clustered in the form of sphnorilitos, while the other had the form of plates. The needle type crystals were moro abundant than tho plate type crystals, which were less common,

* Tho synthetic NANA (Cnlbiochem) showed the snmo typo of crystals. FIGURE I I

MICROPHOTOGR API IS OF CRYSTALLINE SIALIC ACIDS (Mag. x 175). (All micrographs, with the exception of E, were viewed with polarized light. E was not blrefrigcnt and was viewed with dark field illumination.)

A, D Isolated Sialic Acid

C, D Synthetic NANA

E Synthetic NANA-Lac

Zfi. 80 f 82 although the plates were predominant over the needles. Apparently, the different conditions under which synthetic NANA was prepared compared to the NANA de­

rived from casein might be the reason for this difference. The tendency of NANA to crystallize in the form of needles is in agreement with the illustrations given by Dumas and Alats (19G1), and Svcnncrholm (1956). However, Gottchalk (1960b)

reported in a personal communication with Svennerholm that half esters of sialic acids crystallize in plates. Although the presence of half esters in the prepara­ tions cannot be excluded in this experiment, it is possible that the two types re­ present polymorphs.

The crystals of NAN A-Lac (Calbiochcm) were completely different from those observed for NANA, and were lacking bircfrigcnce. The lack of btrefrl- gence indlcntcs Isotropic behavior and is a common property of the crystals that crystallize in the cubic system.

C. Charnctcrizntion of the Casein Blfidus Factor (CDF)

During the preparation of casctn-derived sialic acid by ion-exchange chromatography (Figures 5 and G), a UV-absorbing fraction (CBF) was obtained which wns the object of further studios. This wnB done because previous studies have shown that this frnction stimulated the growth of Ij. blfidus to a high degree (Jao, 1974),

1, Slim illation of L. blfidus by CBF and N-AcGluo

N-acetylglucosamine has been known to be one of the growth-promoting 83 substances for L. blfidus ( Jao , 1074; Hose ct a l., 1954). In this experiment,

N-AcGluc served as a control for determining the growth-promoting activity of the CBF.

The results of this experiment are shown in Figure 12. It was

apparent that the CBF indeed stimulated the growth of L. blfidus to a

greater degree than did N-AcGluc. The results also showed that the fil­

ter sterilised samples promoted the growth of L, blfidus more than the auto-

clavcd samples.

Figure 13 gives the UV spectrum of both the filter-storilized and auto- clavcd Bnmplcs of CBF. Both spectra exhibited their maxima at 280 nm. How­ ever, the autoclnvcd sample showed more intense absorbance at lower wave lengths compared to the filter-sterilized sample. It is possible that interactions between mntno acids and sugars may be responsible for these changes in the spectra. However, these interactions introduced by heat wero not completely destructive since the autoclnvcd sample lost only partly its stimulatory effect on L. blfidus (Figure 12),

2. Stimulation of L. blfidus of Ultrnfiltrntes from CBF

Figure 14 illustrates the effect of adding ultrafiltrate from CBF to the growth media for L. blfidus. Tho UM-2 membrane retained material with mole­ cular weight larger than 1,000. The growth-promoting activity of the ultrafil­ trate was lower compared to the original solution. However, some residual activity was left in tho ultrafiltrate, indicating that tho growth-promoting activity 4

FIGURE 12

STIMULATION OF L. DIFIDUS BY CBF AND N-AcGIuc

A N-AcGluc (filter-sterilized) □ CBF (filter-sterilized) ■ CBF (autoclavcd) O Control (filter-sterilized)

84 O.D. (6.60 nm) NUAIN IE (h) TIME INCUBATION FIGURE 13

UV SPECTRUM OF FILTER- STERILIZED AND AUTOCLAVED CBF FRACTIONS

Filter - sterilized

------Autoclnvcd

86 200 O.D AEEGH (nm) WAVELENGTH 3o0 FIGURE 14

GROWTH EFFECT OF ULTRAFILTRATE OF CBF ON L. BIFIDUS

a CBF (filter-sterilized) A CBF Ultrafiltrate (filter-sterilized) O Control (filter-sterilized)

88 O.D. (660 nm) INCUBATION TIME (h) TIME INCUBATION 90 of the CBF fraction was due partly to molecules with molecular weight higher than 1,000 and partly to molecules with molecular weight lower than 1,000, No measurements were made on the retained material which was entrapped by the mombrano,

3. Comparison of the Bifidus Activity of CBF and Casein

Since CBF was derived from casein, it was of interest to compare the bifidus activity of CBF with vnrious casein preparations.

Figures 15 and 10 show that CBF has higher activity compared to intact isolcctric casein, hydrolyzed caseins, and x-casein. In fact, all the types of casein which were used in this experiment exhibited somo degree of inhibition on

L. bifidus. The rosults suggested that the stimulating factor was not available or might not be in sufficient quantities to stimulate the growth of the organism.

4, Chemical Composition

The CBF was not a single compound. Analysis indicated the presence of 1% enrbohydrnte, mainly lactose (Table 9 and Figure 9), which might result in CBF from any lactose that was trapped in the isolcctric casein during its preparation. The possible presence of lactulose (4-O-0-D-Galactopyranosyl-

D-Ot-Fructose) which is known to stinfulate the growth of L. bifidus (Hlrata,

1958; Niemann ct al., 19GG) was discounted because after treatment with - galactosidase, no fructose was detected either by TLC or by qualitative tests.

The CBF gave positive reaction in tho "direct Ehrlich test", often used FIGURE 15

COMPARISON OF BIFIDUS ACTIVITY OF CBF WITH k -CASEIN

D CBF (filter-sterilized) A Casein (autoclavcd) O Control (filter-storilizcd) O.D. (660 nrn) NUAIN IE (h) TIME INCUBATION 16 92 FIGURE 16

COMPARISON OF BIFIDUS ACTIVITY OF CBF WITH CASEIN PREPARATIONS

□ CBF (filter-sterilized) A Intact Casein (autoclavcd) ■ CaBoin hydrolyzed in H2SO4 reclaimed at pH 4.6 (autoclavcd) • Casein hydrolyzed in H2S04 reclaimed at pH 1.5 (autoclaved) O Control (filter-sterilized)

93 O.D. (660 nm) NUAIN IE (h) TIME INCUBATION 95

TABLE 9 CHEMICAL COMPOSITION OF CBF

Type of reaction or Chemical Tests or Compounds Tested percent of material present

Direct Ehrltch reaction before or after + dinlysis

Biuret reaction +

Free or bound sialic acid -

Hexosamines -

Free or bound N-acctylhcxosainines^ -

Phosphorus 0.15 %

Total carbobydrntes (lactose) 7.0 %

Total Nitrogen 9.0 % 96 to demonstrate sialtc acid. However, specific carbohydrate tests (Table 9) demonstrated that hexosamine, N-AcGluc and NANA were not present in any detectable amounts in free or bound form. It was assumed, therefore, that the positive "direct Ehrlich reaction" was caused by the presence of tryptophan-

containing peptides which have been reported to yield an unstable positive reac­ tion by this test (Gottchalk, 1960). The characteristic fading of the tryptophan- induced color was confirmed for the exhaustively dialyzed CBF material.

The positive biuret reaction given by CBF indicated the presence of pep­ tide bonds and in the absence of amino sugars, it was assumed that all of the nitrogen found was protein derived. The results of TLC for amino acids are given in Figure 17. Only valine was identified with certainty, the resolution of the remaining amino acids being relatively poor. However, at least 7 spots were detected.

The phosphorus content of CBF was much lower compared to that of intact ensein (Tables 9 and 10). This indicated that CBF was not a homogenous degradation product of casein. It should be noted that the hydrolysis conditions used were not sufficiently severe to hydrolyze the phosphorus bonds in casein.

5. Physico-Chomicnl Properties

The casein bifidus factor (CBF) was readily soluble in water and was very hydroscopic. Tho isolcctric point was found to be at 4.8 by turbidity measurements.

The CBF was not retained by tho cation exchange column Duolite C-20I1 FIGURE 17

THIN LAYER CHROMATOGRAPHY FOR AMINO ACIDS IN CBF (Solvent: ethyl acetate/Butanone/formic acid/water, 6/3/1/1 V/V. Spray reagent: .4% ninhydrin.)

1. Intact Casein 2. Alanine 3. Arginine 4. Aspartic Acid 5. Sorine 6. Tryptophan 7. Vnltno 8. CBF

97 98 99

nor by the anion exchange columns DOWEX-1, X 8 and Cellcx-D (DEAE), This

indicated the absence of strong charges on the CBF,

The electrophoretic pattern obtained in presence of urea and mercap-

tcthanol is shown in Figure 19, Two bands were observed with very low stain­

ing capacity compared with the statning capacity of the casein bands. The electro­

phoretic studies also showed that the material was inert to rennet treatment,

indicating the absence of a rennet sensitive bond. Although only two bands were

observed by electrophoresis, column chromatography on DEAE revealed the pre­

sence of five peptides, as it is illustrated in Figure 18. The absence of the snmo

number of bands in electrophoresis suggested that cither these peptides were not

resolved or that they were not strongly charged and did not stain by amido black.

Sedimentation analysts did not directly show multiplicity of scdtmenttng

bnnds but tho rapidly diffusing peak indicated tho presence of rathor low molecu­

lar weight material,

D. Properties of tho Residual Sialic Acid-Free Casein

The reclamation of the residual casein was made after adjustment of

pH of tho hydrolysnte to 4.G (Figure 5), The precipitated material was washed

twice wilh wntor and then freeze-dried. Reclamation of tho major part of casein

could also be made without adjustment of the pH of the hydrolysate. However,

such reclamation method was generally less favorable and yielded a less stable

product.

9 FIGURE 18

ELUTION DIAGRAM OF CBF ON CELLEX D (2 x 23 cm ) COLUMN (CBF dissolved in 0.01M tris buffer, pH 8, elution with a gradient of 2N NnCl in tris buffer; UV absorption at 270 nm.)

100 O.D, (270 nm) GRADIENT OF 2N NaCl 2N OF GRADIENT 1, Yields of Residual Casein

The following yields were obtained:

pH of Reclamation • g of residual casein % Recovery per 500g of casein

1.5 360 72

4 .6 420 84

These results show that most of the casein could be reclaimed in both eases, but more casein was reclaimed when the pH of the hydrolysate was ad­ justed to 4.6.

2. Urea Starch Gel Electrophoresis

Results of urea starch gel electrophoresis are shown in Figuros 19 and 20. The patterns for hydrolyzed caseins wore similar to those for intact caBcin with distinct bnnds for Cts, and K- caseins. These results showed that the limited acid hydrolysis employed did not attack the peptide bonds in the bulk of casein.

Removal of sialic acid and associated peptides did not affect the electro­ phoretic mobility of tho major K -casein fractions. This might be attributed to tho fact that tho major components of k -casoin, which were resolved here, have very little or no. enrbohydrntes (MacKinlny and Wake, 1965).

Following rennet treatment, both hydrolyzed and intact caseins showed evidence of the rennet action by the disappearance of tho k -casein zones and FIGURE 19

UREA STARCH GEL ELECTROPHORESIS OF SIALIC ACID-FREE CASEIN (from dry isolcctric casejn)

A. Intact Casoin B. Casoin hydrolyzed in H2SO4, reclaimed at pH 4.6 C. Casoin hydrolyzed in HCOOH, reclaimed at pH 4.6 D. CBF

R. Same samples following rennet treatment 104

V

6

<*5 FIGURE 20

UREA STARCH GEL ELECTROPHORESIS OF SIALIC ACID-FREE CASEIN (from frosh isolcctric casein)

A. Intact Casoin B. Casein hydrolyzed in H2SO4, reclaimed at pH 4.6 R. The samo samples following rennet treatment

105 106 mm® 107 and the simultaneous appearance of positively-charged zones, However, once again, there were no major differences In the electrophoretic profiles of hydro­ lyzed and intact caseins.

Judging by the electrophoretic- patterns in Figure 20 there seems to be no ninjor differences in the residual caseins resulting from laboratory-made fresh casein, and commercial dry casein. Therefore, in this respect, the source of casein appears to be inconsequential.

Although intact and hydrolyzed caseins exhibited an overall similar elec­ trophoretic profile with respect to number and mobility of bands, there were sig­ nificant differences in the staining intensity of the K - and para- K -casein bands.

These differences suggested that tho rennet action was more uniform and more complete In tho hydrolyzed products than in tho intact casein, possibly bccauso tho ronnct-scnsilivo bond may have been moro exposed by tho removal of sialic acid, at least for some of the k -casoin molecules. Another explanation might be that para- K -casoin (+) and (++) occurred ns tho result of tho loss of a seg­ ment of the para- /: -casein (+++).

Figure 21 shows tho electrophoretic results obtained with residual casein reclaimed at the low pH of the hydrolysis (1.5). IL was apparent that this material differed from the intact casoin moro than the residual casein reclaimed at pH 4. G. Tho differences included changes in tho electrophoretic mobility of tho a 8 and /7-caseins as well as a loss in differentiation of the K -casein bands.

For this reason, tho reclamation of casein at pH 4.G was favored over reclama­ tion at pH 1,5, Figure 21 also shows that more changes occurred In tho sialic FIGURE 21

UREA STARCH GEL ELECTROPHORESIS OF SIALIC ACID-FREE CASEIN (from dry isolcctric casein)

A. Intact Casein Casein hydrolyzed in II2SO4 (for 1 h), reclaimed at pH 1.5 Casein hydrolyzed in H2SO4 (for 2 h)t reclaimed at pH 1.5 R. Same samples following rennet treatment

108 109 110 acid-frcc casein when the hydrolysis was prolonged to two hours, •

3, Sialic Acid and Phosphorous Content

Chemical analysts for sialic acid and phosphorus arc shown in Table 10.

Whereas sialic acid was nearly completely removed (*'90%), the phosphorus was not affected. The easy rcmovul of sialic acid by mild acid hydrolysis was due to several reasons (Noubcrger and Marshall, 1966). First of all, sialic acid is a 2-deoxysugar, the glycosides of which are hydrolyzed much faster compared to the derivatives of glucose. Secondly, a glycoside of sialic acid is a ketoslde which is hydrolyzed more rapidly than an aldopyranoside. Finally, the pH of the sialic acid is 2.6 and thus, at pH 1.5, most of the acid is no longer ionized.

TABLE 10. EFFECT OF HYDROLYSIS ON THE SIALIC ACID AND PHOSPHORUS CONTENT OF CASEIN

SIALIC ACID PHOSPHORUS SIALIC ACID SAMPLE % % % Remaining

Intact Casein .3 -.3 5 .72 100 H^SO^-Cascin .03-.04 .72 12 HCOOII-Casoin .04-.05 .72 15

This is important because unionized carboxyl groups inhibit protonation on the glycosidlc oxygen in the hydrolytic reaction. The identical values of phosphorus in inacl and hydrolyzed caseins are of interest since it is known that casein phos­ phate can easily bo removed from casein by heat treatment (Belec and Jennoss,

1962; Shnrmn and Hanson, 1970b). I ll

4, Rennet and Calcium Sensitivity

The rennet sensitivity results arc given in Table 11. The clotting time showed that the caseins coagulated slower than skim-milk and the acid-hydro- lyzcd casein was the slowest. The curd of the sialic acid-free casein was soft in comparison with that of intact casein.

TABLE 11. RENNET SENSITIVITY OF INTACT AND HYDROLYZED CASEINS

SAMPLE CLOTTING CURD TIM E, SEC. CHARACTERISTICS

Reconstituted slcim- 400 Firm milk +0.01M CaCl2

3% intact casein h* 1,030 Firm 0.01M CnCl2

3% H2SO4 casein + 1,490 Soft 0.01M CaCI2

There is still controversy with respect to the exact role that sialic acid plays on rennet sensitivity of casoin. Many workers have found that the clotting tim e of k -cnscin or milk was increased when sialic odd was removed by neura­ minidase (Gibbons and Checsoman, 1962; Schmidt et a l., 1966; Whcelock and

Knight, 1969). On tho other hand, other workers have shown that removal of sialic acid or other enrbohydrates did not delay rennet action (Armstrong ct al. ,

19G7; MncKinlny and Wake, 1965). The above results show that casein is still sensitivo to rennet after removal of sialic acid and some soluble peptides by acid 112 hydrolysis. However, the clotting time was somewhat prolonged.

Figure 22 shows the Ca4*4* sensitivity of the caseins. Hydrolyzed casein was more Ca++ sensitive than intact casein. It is known that §c -casein loses its ability to stabilize a g and f} caseins against Ca4"4" precipitation when glycomacro- pcplide is removed by rennin. The removal of sialic acid and of soluble pep­ tides associated with CBF might have to do with the increased Ca‘H' sensitivity of the hydrolyzed casein observed here. However, the heat treatment should not be neglected since it has been found by Hansen (1068) that upon heating (100°C for 30 min), the ability of K -casein to stabilize Ofs-cnscln from precipitation by ChH' is decreased.

6. Solubility Studies 4 Figures 23 and 24 show the solubility of cascinB between pH 4.0 to 5.0 at 2°C and 25°C respectively. All three preparations exhibited similar solubil­ ity over the entire pH range at both temperatures. Howover, the percent of sol­ uble casein was somewhat lower at 25eC as compared to that at 2eC. This was in agreement with Bingham’s results (1971) and was probably a reflection that the hydrophobic interactions were more stable at room temperature than at 2°C

(Knuzmann, 1959). Apparently, the removal of sialic add and soluble peptides by ncid hydrolysis did not affect these interactions.

Figure 25 illustrates the solubility of and Ca++ salts of intact and hydrolyzed caseins. The supornntant casein was used as an index of solubility and was plotted against pH from 6,0 to 9.0, There was little difference in the FIGURE 22

CALCIUM SENSITIVITY OF HYDROLYZED CASEINS

O Intact Casein

□ Casein hydrolyzed in H2SO4 reclaim ed at pH 4,6

113 CASEIN (mg/ml) o SUPERNATANT C l col CASEIN (%) tc FIGURE 23

SOLUBILITY OF HYDROLYZED CASEINS AT 2°C AND pH 4.0 TO 5.6

O Intact Casein

A Casein hydrolyzed In H2SO4 reclaimed at pH 1.5

□ Casein hydrolyzed in H2SO4 reclaimed at pH 4. G

115 1 1 6 FIGURE 24 # SOLUBILITY OF HYDROLYZED CASEINS AT 25°C AND pH 4.0 TO 5.6

O Intact Casein

A Casein hydrolyzed in H2SO4 reclaim ed at pH 1 .5

□ Casein hydrolyzed in H2SO4 reclaim ed at pH 4.6

117 SOLUBLE CASEIN (%) C l o o

00 FIGURE 25

SOLUBILITY OF Na- AND Ca- CASEINATES

O Intact Caseins, Na-Caseinate

Intact Caseins, Ca-Caseinate

□ Casein hydrolyzed in H2SO4 reclaimed at pH 4.6, Na-Caseinate

■ Casein hydrolyzed in H2SO4 reclaimed at pH 4. C, Ca-Caseinate

119 SUPERNATANT to CASEIN (mg/ml) © o 120 121 solubility of the sodium suits of intact and hydrolyzed caseins at pH 6,5 to 8.0.

There was no difference at pH 9.0. This figure also illustrated that the sodium caseinates were more soluble than the corresponding calcium salts. The hydro­ lyzed caseins were least soluble in the presence of calcium, indicating their higher Ca** sensitivity.

6. Viscous Properties

In Table 12, the viscous properties of intact and the residual caseins are compared. The sialic acid-free casein still exhibited the viscous charac­ teristics of Intact caseins, although to a somewhat reduced degree. Thore is not yet agreement about the influence of sialic acid on the rhcological properties of glycoproteins. It is known that the release of sialic acid at pH 6.0 by neura­ minidase decreased the viscosity of ovine sub-maxillary mucin (Gottchalk, 1960a).

However, Ostchiro and Eylar (1969) found only little difference in the viscosity of fctuin and desinlized fctuin. The minor differences for casein observed here jpay be attributed to the removal of cither the sialic ncid or the soluble peptides.

7. Flavor

The food uses of casein and casein products have been greatly hampered by their tendency to develop an unpleasant, gluey, off-flavor. Rnmshaw and Dun- stane (1969) have suggested that the off-flnvor develops during the early stages of non-onzymntic browning, Hansen et al, (1970) have implicated mechanically induced free radicals as a trigger for the development, Sharma and Hansen 122

TABLE 12 VISCOSITIES OF HYDROLYZED CASEIN (in centipoise)

INTACT 4% HYDROLYZED 4% RPM OF SPINDLE G0°C 25*C 5®C 60*C 25°C 5° C

100.0 13.4 19.6 29.2 12.3 17.6 25.5 50.0 8.8 13.0 21.2 8.4 12.0 18.4 20.0 0.5 10.0 13.0 6.0 9.5 12.0 10.0 0.0 8.0 9.0 6.0 G.O 8.0 5.0 6.0 6.0 8.0 6.0 6.0 8 .0 2.5 8.0 8.0 8.0 8.0 8.0 8.0 1.0 10.0 20.0 10.0 10.0 20.0 10.0 0*5 10.0 30.0 10.0 10.0 20.0 10.0

INTACT 8% HYDROLYZED 8% RPM OF SPINDLE 60°C 25°C 5#C 60°C 25°C 5*C

100.0 23.0 43.6 67.5 18.7 40.2 60.6 50.0 15.4 29.8 50.0 13.6 26.4 42.4 20.0 11.0 19.5 39.5 9.0 18.0 31.5 10.0 7.5 16.0 37.0 7.0 14.0 24.0 5.0 10.0 16.0 36.0 8.0 12.0 28.0 2.5 16.0 20.0 30.0 16.0 18.0 26.0 , 1.0 20.0 20.0 30.0 20.0 20.0 30.0 0.5 20.0 30.0 40.0 20.0 30.0 40.0

INTACT 16% m fDROLYZED 16% RPM OF SPINDLE 60°C 25°C 5°C 60*C 25°C 5*C

100.0 ______8210 50.0 12900 123 8740 20.0 258 13750 118.5 9400 10.0 249 14400 ------116 9000 _—___ 5.0 246 14800 688000 122 9000 559000 2.5 244 16000 672000 124 10400 545000 1.0 250 17000 656000 140 11000 536000 0.5 300 18000 568000 160 12000 559000 123

(1070a) have related the possible homolytic cleavage of covalent linkages, such as ester/phosphorus bonds to the development of gluey off-flavor.

In a flavor trial comparing intact and sialic acid-free casein, it was found that the latter had the same flavor characteristics as intact casein. The gluey off-flavor was not removed by the treatment employed, suggesting that

* sialic acid is not associated with the gluey off-flavor.

E, Simplified Production of a Crude Preparation Containing Sialic Acid

and Casein Bifldus Fraction (SA/CBF)

For most food applications, and in cases where there is not a need to separate the sialic acid from CBF in a purified form, the manufacture of a crude preparation containing both sialic acid and CBF (SA/CBF) would appear to be advantageous. Figure 26 illustrates tho steps that were used to prepare such a crude fraction in the laboratory from 500 g of casein. The yield by the outlined procedure was;

Total SA/CBF : 17.00 & Total NANA : 1.32 S Purity of NANA ; 7 .8 % Total nitrogen 8.3 % Total residual casein : 420.0 g

Tho yields of sialic acid and CBF were higher compared to previous methods because losses involved in the purification steps wore avoided. The overall recovery was 88% compared to 54% by the other approach. The crude FIGURE 26

PRODUCTION SCHEME FOR A COMBINED SIALIC AC1D/CBF (SA/CBF) FRACTION FROM CASEIN

124

* 125

HYDROLYSIS OF 1 PART DRY CASEIN WITH 10 PARTS OF 0.18N II2S04 (pH 1.5) AT 80eC FOR 1 HOUR

BRING pH TO 4.6 WITH Ca(OH)2

DRAIN OFF THE HYDROLYSATE

DEPROTEINATED RESIDUAL CASEIN HYDROLYSATE

BRING pH TO 7 .0 WASH WITH WATER AND CONCENTRATE TO 1/10 VOLUME COMBINE WASHINGS WITH HYDRO­ LYSATE___

CLARIFY THE CONCEN­ DRY BY CON­ TRATED HYDROLYSATE VENTIONAL MEANS

FREEZE DRY SIALIC ACID-FREE (Expected yield: 3. 4%) CASEIN (Isolectric) (Expected yield: 84%)

NOTE: * 11)0 residunl casein can be further washed to remove any excess of calcium.

** The removed solids will contain most of tho calcium sulfate. 126 preparation obtained was pure white and readily soluble in H 20 .

The outlined approach to separating combined preparation of sialic acid and CBF (SA/CBF) from casein appears to be sufficiently uncomplicated for possible commercial exploration. It may bo noted that in this scheme, it may be possible to employ solvent extraction of the crude preparation (SA/CBF) for separating sialic acid from the CBF. VI. DISCUSSION

This study has been concerned with the isolation of sialic acid from ca­ sein in a manner which would be practical and economical. The results of the study have shown that commercial casein is a good source for manufacture of n sialic acid-rtch preparation which may be further purified to produce crystalline

N-acetylncuraminic acid (NANA).

In the search for a procedure for obtaining purified sialic acid from ca­ sein, the basic approach involved acid hydrolysis of casein to liberate sialic add, stepwise purification of the sialic acid in the hydrolysate and finally, examina­ tion of the purity of the isolated product. A number of alternative approaches worn investigated and It was found:

(a) That the commercial isolectric casein was preferred over fresh isolcctric casein because the curd of the fresh casein remained strongly hydrated after hydrolysis and the recovered hydrolysate contained more soluble poptidcB which could interfere with the final purification;

(b) That H2S04 was preferred over HCOOII as tho hydrolyzing agent, because the removal of HCOOII by evaporation and subsequent extraction of sia­ lic acid by solvents or water was inefficient and, furthermore, the isolation of tho sialic acid was problematic;

127 128

(c) That Cn(OII)2 was preferred over Ba(OH)2 o r NaOH because Ba++ is toxic and Na+ fails to remove excess SO4 .

The procedure which was finally selected involved the use of anion and cation exchange column chromatography for the final isolation of sialic acid from the hyclrolysatc. With this approach, the average yield from 500 g of casein was

000 mg sialic acid of 90% purity. The recovery was 54%, considering 0.3% NANA in casein. This method was more efficient compared to various other methods used which involved solvent extractions. However, it may bo noted that the ease of solvent extraction makes this approach a possibility for use in a large-scale production.

Thin-layer, paper chromatography and crystal morphology confirmed the identity of the isolated sialic acid with N-acctyl neuraminic acid (NANA),

The information obtained from crystallizing various preparations of NANA sug­ gested that the casein-derived sialic acid was possibly more pure than a pur­ chased, synthesized product which contained a substantial amount of half esters.

A by-product of the isolation of sialic acid was the casein btfidus factor

(CBF), which stimulated the growth of L. btfidus to a high degree. In rospect to the stimulatory crfcct of CBF, it was found that most of the responsible fac­ tors were associated with a material possessing a molecular weight largor than

1,000 as judged by membrane exclusion data. The CBF factor was found to be somewhat hent labile, possibly because of an interaction between amino acids and free sugars. Although CBF was derived from casein, no evidence was found in this study for caseins possessing any stimulatory effect on L, bifidus. A

similar situation has been reported for sonic other glycoproteins by Gyorgy et

al. (1974). They found that Atj-acid glycoprotein from human scrum, trans­

ferrin, urinary glycoprotein, and submaxlllary mucin did not exhibit any growth

promoting activity on L. bifidus. However, after treatment with neuraminidase or after acid hydrolysis, tho growth activity of the glycoproteins, except that of transferrin, increased in close correlation to the content of N-AcGluc. The

authors suggested that the sialic acid might exert a protective effect for the gly- coprotein against degradation by some of the L. bifidus enzymes. Transferrin, treated 01* untreated with neuraminidase, was completely inactive and in this respect, behaved like tho bovine caseins in the present study. The reason for transferrin and casein not possessing significant bifidus activity mny be their low content of N-acotylhoxosamines. Gyorgy ct al. (1974) did not report if any

"bifidus factors" were released from these other glycoproteins by the treatment used to remove sinllc acid. Certainly, it is of importance that such a fraction was rclonsed from bovine casein.

Chemical analysis of CBF has revealed that it was not a pure compound but contained sovcral substances. The fraction contained no detectable amino

sugars, but 7% lactose in free form, .15% phosphorus and 9% nitrogen. Com­

plete amino acid nnalysis has not yet been completed. The possible presence of lactulose, a degradation product of lactose which contains fructose and galac- toso and which possesses bifidus activity was discounted because no fructose was 130

* found in the preparation after treatment with f i -galactosidase.

Some evidence was found to support the concept that CBF was the re­

sult of a degradation of a part of the casetnomacropcptide of -casein. This

part contains many residues of serine or threonine which arc polar and uncharged

amino acids (Hill and VValte, 19G9). It contains only one phosphorus, which is

located on a serine residue (No, 149)(Mercier ct a l., 1973). The possibility

that tills part is the origin of CBF would be consistent with the observed high sol­

ubility, neutral characteristics and the low phosphorus content of CBF. The only

slight changes observed In tho # -casein electrophoretic patterns after hydroly­

sis woro also consistent with this concept because tho loss of a predominantly

uncharged segment from a protein molecule would not be expected to cause sig­

nificant changes in tho electrophoretic mobility.

However, the lack of N-ncetylhcxosnmincs and the resistance to rennet

of CBF indicated that the segmont did not contain the rennet-sensitive bond. It

is of interest Hint the residual casein was shown to contain the rennet-sensitive

bond intact. Tho only aspect which was not fully explained by this concept was

the possible presence of tryptophan in CBF which seemed to be responsible for

the fading color given by tho "direct Ehritch reaction", since tryptophan is not

present in tho mncropeptide part of K -casein. However, it should be stressed

that CBF was not homogenous and may contain free or bound tryptophan from

other sources.

Any practical and economical application for the production of sialic 131 acid and the CBF from casein would logically require that the bulk of residual casein bo reclaimed for edible or technical utilization. Therefore, it was de­ sirable to study the properties of this modified casein in order to determine any possible uses. The removal of sialic ncid from casein would not necessarily reduce the food value but might possibly improve the digestibility of the protein.

Gottchalk et al. (I960) reported for ovine submnxlllnry gland glycoprotein that it was degraded by trypsin to a far greater extend after enzymatic removal of sialic acid. Although the nutritional aspects of the residual sialic acid-free casein were not studied here, electrophoretic patterns, solubility properties, rennet and calcium sensitivity, and viscous character revealed that the residual casein was remarkably similar to the intact casein. For most food applications, it might be expected that sialic acid-free casein would perform equal to regular casein in functional properties.

The approaches to isolation of sialic acid and the CBF from casein have, in this study, been principally concerned with separation of purified compounds.

However, it should be recognized that the dcprotetnatcd hydrolysate contains all of the liberated sinlic ncid and all of the CBF fraction. Without further purifica­ tion, it was possible to dehydrate this crude preparation which contained: 17 g total soltds, 1.32 g total NANA (7.8% NANA), and 8.3% total nitrogen. Infant formulas and other foods enriched with such crude preparation (SA/CBF) will havo the characteristics of being rich in sialic acid and will contain a factor which will stimulate tho growlh of L. bifidus. Therefore, such enriched formulas would be more similar to human milk. 0

VH. SUMMARY

In this study, a method was developed for isolation of pure static acid

and a casein bifidus factor (CBF) from casein. The CBF stimulated the growth

of L. bifidus. The stimulatory effect was measured by the growth response of

L. bifidus var. penn. on a semi-synthetic broth-ATCC medium 7G containing

0.25 mg CBF per ml.

The method involved the hydrolysis of one part of casein in ten parts of

0.18N H2SO4 at pH 1.5, and 80°C for 60 min. After the removal of protein,

the hydrolysatc from 500 g of casein contained; 17 g total solids, 1.32 g total

NANA, and 8.3% total nitrogen. The liberated sialic acid and CBF were sepa­

rated from the hydrolysate in a purified form by ion exchange chromatography.

The conccntrutcd hydrolysate was passed over a DOWEX-1, X 8 column in the

acetate form followed by washing with water, Tho leading UV absorbing peak

was the CBF containing; 2 g total solids, 9% total nitrogen, 7% carbohydrates

(mainly lactose), and .15% phosphorus. The peptide fraction of CBF was hctoro

genous and the L, bifidus activity was mainly associated with peptides of mole­ cular weights largor than 1,000. It was somewhat heat labile and exhibited ncu- trnl characteristics. The NANA was elutod from DOWEX-1, X 8 column with a

132 133 gradient of 2M C^COONa at a conductivity of 2 x 10“3 mhos. The "direct

Ehrlich" positive eluates were passed through cation exchange column Duoltte

C-20H (H+) and freeze-dried, The purified sialic acid material contained: .9 g total solids, .81 g total NANA (90% pure). The preparations of the sialic acid showed the crystalline shape of N-acetylncuraminic acid (NANA).

The removal of sialic ncid and CBF from casein produced a modified, sialic acid-free casein which exhibited similar electrophoretic patterns, solu­ bility properties and viscous characteristics to intact casein. Rennet sensiti­ vity studies showed that the sialic acid-free casein contained the rennet-sensitive bond intact. However, the calcium sensitivity of this modified casein was some­ what higher compared to whole casein. For most food applications, it might be expected thnt sialic acid-free casein would perform equal to regular casein.

i BIBLIOGRAPHY

Adachl, S. and Patton, S., 1961, ’'Presence and Significance of Lactulose in

Milk Products," J. Diary Sci. , 44:1375.

Aprnhnmian, V. B ., 1973, Liberation of Sialic Acid From Casein Hydrolyzed

in Formic Acid. Thesis, Ohio State University, Columbus, Ohio.

Armstrong, C, E., Macklnlay, A. G ., Hill, R.J. and Wake, R.G., 1967,

’’The Action of Rennin on k -Casein: The Heterogeneity and Origin

or tho Soluble Product," Biochim. Blophys. Acta, 140:123.

Barker, S.A. and Stacey, M ., 1963, "The Carbohydrates of Milk and Colos-

trum." Dairy Sci. Abstr. . 25:445,

Bartholomew, B.A., Jourdian, G.W. and Roseman, S., 1973, "The Sialic Acids.

XV. Transfer of Sialic Acid to Glycoproteins by a Sialyltransferaso

From Colostrum," J. Biol. Chcm. , 248:5751,

Belcc, J. and Jonncss, R ., 1962, "Depliosphorization of Casein by Heat Treat­

ment: I, hi Caseinate Solutions," J. Dairy Sci. , 45;12.

Binglmm, E. W., 1971, "Influence of Temperature and pH on the Solubility of

o S l, P , and K-Casein," J. Dairy Sci.. 54:1077.

134 Blacklow, R.S. and Warren, L ., 19G2, "Biosynthesis of Sialic Acids by

Neisseria meningitidis," J. Biol. Cliem. , 237:3520.

Blix, G., 1958, "Sialic Acids," Fourth International Congress of Biochemistry

Vienna, September. In CIBA Symposium on the Chemistry and Biology

of Mucopolysaccharides, London, p. 302.

Blix, G. and Joanloz, W .R., 1969, "Sialic Acids and Muramic Acid. IA:213.

In: The Amino Sugars, Vol. IA, edited by W. R. Jeonloz, Academic

Press, Now York.

Blumberg, B.S. and Warren, L ., 1961. "The Effect of Sialidase on Trans­

ferase and Olhor Serum Proteins," Biochim. Biophys. Acta, 50:90.

Brown, C.D. and Townsloy, P.M ., 1970, "Fermentation of Milk by Lacto-

bacillus bifidus," Canadian Inst, of Food Technol. J ., 3:121.

Brunctti, P ., Jourdian, G. W. and Roseman, S., 1962, "BI. Distribution

and Properties of Animal N-Acetylneuraminic Aldolase," J. Biol. Cliem

237:2452.

Brunner, J. R. and Thompson, M. P ., 1959, "Some Characteristics of the

Glycomncropcptido of Casein: A Product of the Primary Rennin Action,"

J. Dairy Soi. . 42;1881, 136

* Bullen, C.L. and Willis, A, T,, 1971, "Resistance of the Breast-Fed Infant

to Gastroenteritis," British Med. J ,, 3:338.

Cabczas, J.A ., 1973, "Hie Typo of Naturally Occurring Sialic Acids," Rev.

K b p . F ls io l., 29:307.

Cabezas, J. A. and Carrion, A., 1969, "Sialic Acids: X. Content in Acyl-

neuraminic Acids of Several Dictctical Products (Matcrnizcd Milks),"

Rev. Esp. Flsiol.. 25:141.

Cabezas, J.A ., Carrion, A. and Varguez-Porto, J ., 1968, "Sialic Acids:

VIL Composition and Variations of the Content of the Neuraminic

Derivatives in Human Colostrum," Rev. Esp. Flsiol. , 24:85.

Carlson, D.M., Jourdian, G.W. and Roseman, S., 1973, "Hie Sialic Acids:

XIV. Synthesis of Siolyl-lactoso by a Sialyltransferasc From Rat Mam­

mary Gland," J. Biol. Cliem., 248:5742,

damp, J. R., Jackson, R, H. and Hough, L ., 1967, "Hie Simultaneous Esti­

mation of L-Fucoso, D-Matuiose, D-Galaotoso, N-Acetyl-D-Glucosamine

and N-Acctylncuraminic Acid in Glycopeptides and Glycoproteins,"

Bioohim. Biophys. Acta. 148:342.

Clamp, J.R. and Putinnin, F.W ., 1967, "Glycopeptides of Immunoglobulins,"

Biocliom. J ., 103*225. 137

Clark, W ., Jackson, R. H. and Pallansch, M .J., 1962, "The Isolation of

Sialic Acid in High Yield from Colostrum," Biochim. Blophys. Acta,

58:129.

Comb, D. G. and Roscman, S., 1958, "Composition and Enzymatic Synthesis

of N-Acctylncuraminic Acid (Sialic Acid)," J. Am. Chem. Soc..

80:497.

Comb, D.G. mid Roscman, S., 1960, "The Sialic Acids: I. Hie Structure

and Enzymatic Synthesis of N-Acetylneuraminic Acid," J. Biol. Chem.,

235:2529.

Comb, D. G., Shimizu, F. and Roseman, S., 1959, "Isolation of Cytidine-51-

Monophospho-N-Acetylneuraminic Acid," J. Am. Chem. Soc., 81:5513.

Craven, D. A, and Gclirkc, C.W ,, 1968, "Quantitative Determination of N-

* * Acctylncuraminic Acid by Gas-Liquid Chromatography," J. Chromatog.

37:414.

Dische, Z ., 1962, "Color Reactions of Hoxoses." Methods in Carbohydrate

Chemistry, Vol. I, p. 491.

Dubois, M ., Gillcs, K.A., Hamilton, J.K ., Rebcrs, P. A. and Smith, F ., 1956,

"Colorimetric Method for Determination of Sugars and Related Substances,"

Anal. Chem., 28:350. 138

' Dumas, B.R. and Alvis, C., 1961, "Preparation De L'Acide N-Acetyl-Neura-

miniquo Crlstalliso; A Partir de la Caseine De Vachc," Bull. Soc.

Chlm. Biol.. 43:377.

Eichberg, J. and Karnovsky, M. L ., 1963, "The Formation of N-Acetylneura-

minic Acid in Ovine Submaxillary Gland," J. Biol. Chem., 238:3827.

Fatllard, H. and Cabezas, J ., 1963, "Cellulose Thin-Layer Chromatrography

of Amino Sugars," J. Physiol. Chem., 333:266.

Faillard, II. and Schauer, R ., 1972, lh Glycoproteins. "Glycoproteins as

Lubricants, Protective Agents, Carriers, Structural Proteins and as

Participants in Other Functions." A. Gottclialk (editor), Elsevier Pub­

lishing Company.

Fiskc, C.G. and Subbarow, Y., 1925, "The Colorimetric Determination of

Phosphorous," J. Biol. Chem.. 66:375.

Fournct, B., Flat, A. M., Montrcuil, J. and Jollcs, P ., 1975, "Hie Sugar

Part of K-Casoins From Cow Milk and Colostrum and Its Microhetcro-

goneity," Bloch imio, 57:161

Ganguli, N.C., Gupta, S.K., Joshi, V.K. and Ehalerao, V.R., 1967, "Sialic

Acid and Iloxose Contents of Protcose-Peptone of Milk in Relation to

Other Milk Proteins," Indian J. Dairy Sci., 20:96. 139

Gibbons, R.A ., 19G2, "The Decomposition of Sialic Acid.11 Biochem. J ., 82:32.

Gibbons, R.A ., 19G3, ’’The Sensitivity of Neuraminosidic Linkage in Mucosub-

stanccs Toward Acid and Toward Neuraminidase,” Biochem. J .. 89:380.

Gibbons, R. A. and Checscman, G. G ., 19G2, "The Actions of Rcnnin on Casein:

The Function of Neuraminic Acid Residues, ” Biochim. Biophys. Acta,

56:354.

Glide, L.J. and Githcns, S., 19G5, " Role of Sialic Acid in Potassium Trans--

port of L 1210 Leukemia Cells,” Nature, 208:88.

Gottscholk, A., 19G0a, "Sialic Acids: Their Molecular Structure and Biological

Function In Mucoproteins,” Bull. Soc. Chim. Biol., 42:1387.

Gottscholk, A., 19G0b, Tho Chemistry and Biology of Sialic Acids and Related

Substances. The University Press, Cambridge.

Gottscholk, A., 19GG, Glycoproteins. Elsevier Publishing Company, p. 190-3.

Gottschalk, A ., Fazetas Do St. Groth, S., I960, "Studies on Mucoproteins:

HI. The Accessibility to Trypsin of the Susceptible Bonds in Ovine

Submaxillnry Gland Mucoprotoin," Biochim. Biophys. Acta, 43:513. Graham, E.R,B., Murphy, W.II. and Gottscholk, A., 19G3, "DC. On the

Susceptibility to Alltali and to Hydroxylamino of the Predominant Carbo-

hydratc-Pcptido Linkage in Ovino Submaxillary Gland Glycoprotein,"

Biochim. Biophys. Acta, 74:222.

Granzcr, E ., 1962, "Thin-Layer Chromatography of Aminosugars with Silica-

Gel G." J. Physiol. Chem.. 328:277.

Grlmmonproz, L., Bouquelet, S., Bayard, B,, S|pik, G., Monsigny, M. and

Monlrcuil, J ., 1970, "Determination de la Structure du Lexaose isole

du lait do fern 1110 lo Dilactaminyl-Lncto-N-Tetraose," Eur. J. Biochem. ,

13-484.

Guerin, J ., Alais, C ., Jollcs, J. and Jolles, P ., 1974, " k - Casein From

Bovine Colostrum," Biochim. Biophys. Acta. 351:325,

Guminski, T .., 1967a, "Neuraminic Acid in Human Milk: I . Studies on High

and Low Molecular Compounds of Neuraminic Aoid Employing Separa­

tion of Milk on Dcxtran G d," Dairy Sci. Abstr.. 29:639.

Guminski, T. , 19G7b, "Neuraminic Aeid in Human Milk: I I . Content of High

and Low Molecular Compounds of Nouraminie Acid in Human Milk in

Relation to the Stage of Lactation," Dairy Sci. Abstr. , 29:639, Gupta, S. K. and Ganguli, N.C., 10G5, ’’Sialic Acid Content of Casein Prepara­

tions From Cow and Buffalo Milks, " Milchwissenschaft, 20:10.

Gurric, G. A. and Bagshawc, K. D ., 19.69, ’’TUmour Specific Immunogcnicity

of Methylcholanthrcne-Induced Sarcoma Cells After Incubation in Neura­

m inidase , 11 Brit. J. Cancer, 23:141.

Gyorgy, P ., 1957, Chemistry and Biology of Mucopolysaccharides. "N-

Contninlng Saccharides in Human Milk.” Edited by G.W. E. Wolstenholme

and M. O'Connor, Little, Brown & Co., Boston, p. 140.

Gyorgy, P ., Hoover, R. E ., Kuhn, R. and Rose, C. S., 1954c, "Blfldus Factor:

HI. The Rule of Dialysis, ” Arch. Biochem. Biophys., 48:209.

Gyorgy, P., Jcanloz, R.W., Nicolai, H. and Zillikcn, F., 1974, "Undialyli­

able Growth Factors for Lactobacillus blfidus var. ponnsylvanicus:

Protcctivo Effect of Sialic Acid Bound to Glycoproteins and Oligosaccha­

rides Against Bacterial Degradation,” Eur. J. Biochem.. 43:29.

Gyorgy, P., Kuhn, R., Rose, C.S. and Zillikcn, F ., 1954b, "Bifidus Factor;

II. Its Occurrence in Milk From Different Species and in Other Natural

Products. ” Arch. Biochem. Biophys.. 48:202.

Gyorgy, P ., Norris, R. F. and Rose, C, S., 1954a, "Bifidus Factor: I, A Var­

iant of Lactobacillus bifidus Requiring a Special Growth Factor,” Arch.

Biochem. Biophys., 48;193, Hakomori, S. and Torunobu, S., 1969, "Isolation and Characterization of a

Glycosphingolipid Having a New Sialic Acid," Biochemistry, 8:5082,

Iianscn, p. M. T ,, 1968, "Stabilization, of as-Casein by Carr age enan," J.

Dairy S c i., 01*192,

Hansen, P.M. T., Harper, W.J. and Sharma, K.K., 1970, "Formation of

Free Radicals in Dry Milk Proteins." J. Food Sci,, 35:598,

*

Harrison, R., Higginbotham, J.D, and Newman, R .t 1975, "Sialoglycopcptidcs

From Bovine Milk Fat Globule Membrane," Biochim. Biophys. Acta.

389:449.

Haurowitz, F ., 1950, Chemistry and Biology of Proteins. Academic Press Inc.

New York, p. 315.

Hayes, J. F ., Dunkorloy, J. and Muller, L. L., 1969, "Studies on Co-Precipi-

tates of Milk Proteins," Austr. J. Dairy Tech., 24:09.

Hestrin, S., 1949, "Hie Reaction of Acetylcholine and Other Carboxylio Deriva­

tives With Ilydroxylamino and Its Analytical Application." J. Biol. Chem.

180:249.

Hill, R.J. and Wake, R.G., 1969, "AmphiphUe Nature of k -Casein as the

Basis for Its Micelle Stabilizing Property," Nature. 221:635, 143

Hi rata i Y ., 1958, "On Nutrition of Infant With Special Reference to the Meta­

bolism of Lactose and Lactobacillus bifidus." Japan J. Pediat., 9:681.

Hygstedt, O ., 1968, "The Isolation of Sialyl-lactosides With the Aid of Gel

Filtration," Analyt. Biochem. , 23:391.

Inovc, K. and Nagayama, T., 1971, "Effects of Mucin Upon Growth of Low

Birth Woight Infants and Their Intestinal Microflora," Dairy Sci. Abst.

33:3563.

Jackson, R.JI., Coulson, E.J. and Clark, W.R., 1962, "The Mucoprotein of

the Fat/Plasma Interface of Cow's Milk: I. Chemical and Physical

Characterization," Arch. Biochem. Biophys.. 971:373.

Jao, Y. C., 1974, Growth Effect of N-Acetylhexosamines on Lactobacillus bifidus

var. pcnnsylvanlcus. Thesis, Ohio State University, Columbus, Ohio.

Johnson, B.C. and Swanson, A.M., 1952, "Milk Serum Proteins: I. A Quanti­

tative Biuret Test For Milk Serum Proteins," J. Dairy Sci.. 35:823.

Jollcs, P ., 1966, "Casein," hi Glycoproteins. Edited by A. Gottscholk,

■ » Elsevier Publishing Company, p. 335.

Jordan, J. and Lohr, H., 1962, "On the Presence of Neuraminic Acid in Human

Milk, Cow's Milk and in Dairy Products," Milchwiss. , 17;61. 144

Karkns, J. D, and Char gaff, E ., 1964, "Studies on the Stability of Simple De­

rivatives of Sialic Acid," J, Biol. Chem., 239:949.

Kauzmunn, W., 1959, "Some Factors.in the Protein Denaturation," Adv. Pro­

tein Chem., 14:11.

Kean, E. L. and Roscman, S ., 1966, "The Sialic Acids: X. Purification and

' Properties of Cytidine S'-Monophosphosialic Acid Synthetase," J. Biol.

Chem., 241;5643.

Kicrmcicr, F. and Frolsfcld, I., 1965, "On Sialic Acid Content of Cow's Milk:

m . Sialic Acid of Milk as Influenced by Technical Treatment,"

' MilchwiflB., 20:414.

Kim, Y.K,, Arima, S. and Hashimoto, Y., 1968, "Sialic Acid in Milk: m .

Sialic Acid In Normal Milk and in the Fractions of Milk Proteins Obtained

by Dethylaminoethyl-Cellulose Column," Dairy Sci. A bstr.. 30:443.

Kiszn, J. and Ziajka, S., 1973, "Studies on Lactobacillus bifidus Cultivation

in Cow's Milk," Milchwiss., 28:696

Kohn, P ., Winzlor, R.J. and Hoffmann, R.C., 1962, "Metabolism of D-Gluco-

samine and N-Acetyl-D-Glucosamino in the Intact Rat.” J. Biol. Chem. .

237:304, Kuhn, R. and Gauhe, A., 1965, "Determination of the Location of Sialic Acid

Residues in O ligosaccharides," Dairy Sci, A bstr. , 27:460*

Kumar, S., 1972, "Reclaiming Amino Sugars From Milk and Whey," Ohio

Report* 57:90.

Kumar, S, and Hansen, P. M. T ., 1972, "Now Reaction Mixture for Spectro-

photometric Determination of N-Acetylhexosamines," Anal* Chem*.

44:398.

Kundig, W,, Ghosh, S. and Roseman, S., 1966, "N-Acyl-D-Mannosamine

Kinase From Rat Liver," J. Biol. Chem.. 241:5619.

Iambert, R. and Zillikcn, F ,, 1965, "Novel Growth Factors for Lactobacillus

bifidusvar. Ponnsylvanicus." Arch. Biochem. Biophys*. 110:544.

Lewandowslci, J. ., 1973, "A Simple Determination of Hcxosamines in Column

Effluents," J. Anal. Biochem., 54:325.

Mackinlay, A.G. and Wake, R.G., 1964, "The Heterogeneity of k -C asein ,"

Biochim. Diophys. Acta. 93:378.

MacKinlay, A.G, and Wake, R.G., 1965, "Fractionation of S-Carboxymethy!

k - Casein and Characterization of the Components," Biochim. Biophys. Maeno, M ., Held, T. and Kavadzima, T ,, iocs, "Method for Obtaining Indus­

trial Phenothiazine," Dairy Sci. Abstr. , 30:589,

Malprcss, F. 11., 1961, "Sialic Acid Released by Rcnnin Action on Cow and

Human Milk," Biochem. J ., 80:19.

Maricr, J. B., Tcssicr, H. and Rose, D., 1963, "Sialic Acid as an Index of

the k -Casein Content of Bovine Skim-Milk," J. Dairy Sci., 46:373,

Mcrcicr, J. C., Ribadean-Dumas, B., and Grosclaudo, B,, 197V. "Amfno-

Acid Composition and Sequence of Bovine k -Casein," Neth. Milk Dairy J ..

27:313.

Morr, C. W., 1974, "Methods of Gel Electrophoresis of Milk Proteins," In

American Dairy Sci. Assoc. Nomenclature and Meth. of Milk Protein.

Committee, p. 12.

Muldoon, T.G. and Wcstphnl, U .J., 1967, "Steroid-Protein Interactions:

XV, Isolation and Characterization of Corticosteroid-Binding Globulin

From Human Plasma," J. Biol. Chem., 242:5636.

Neimonn, N., Loworgine, E ., De Manciaux, M., Stehlin, S. and Pcrccbois, G.,

1966, "Clinical and Bacteriological Study of a Milk Bifldigenic Because

of its Lactulose Content," Dairy Sci. Abstr. , 28:193. 147

* Ncuberger, A. and Marshall, R. D ., 19GG, In Glycoproteins: "Methods for

the Qualitative and Quantitative Analysis of the Component Sugars,"

Edited by A. Gottscholk, Elsevier Publishing Company, p. 190.

Ncuberger, A ., Marshall, R. D. and Gottschalk, A ., 19GG, In Glycoproteins:

"Some Aspects of the Chemistry of the Component Sugars of Glycopro­

teins," Edited by A. Gottschalk, Elsevier Publishing Co., 1959.

Newman, R. A. and Harrison, R ., 1973, "Characterization of the Surface of

Bovine Milk Fat Globule Membrane Using M icroelectrophoresis,"

Biochim. Biophys. Acta. 298:798.

O'Brien, P .J., Canady, M .R., Hall, C.W. and Neufeld, E. F ,, 19GG, "Trans­

fer of N-Acctylneuraminic Acid to Incomplete Glycoproteins Associated

With MIcrosomes," Biochim. Biophys. Acta. 117:331.

O'Brien, P .J., Glick, M.C. and Zillikcn, F., 19G0, "Acidic Amlnosugars

l 4 "I t1 - C -a, b-Methyl-N-Acetyl-D- Glucosaminidc into Muramic Acid." Biochim. Biophys. Acta. 37:357.

OshU*o, Y. and Eylar, E. H ., 19G9, "Physical and Chemical Studies on Glyco­

proteins: IV. The Influence on Sialic Acid on the Conformation of Fctuin,"

Arch. Biochem. Biophys.. 130:227.

Rafclson, M. E ., Schneir, M. and Wilson, W .V., 1963, "Studies on the Neura­

minidase of Influenza V irus," Arch. Biochem. Biophys., 103:424.

4 148

Ramshaw, E. II. and Duns tone, E.A ., 1969, "The Flavor of Protein.11 J. Dairy

Res., 36:203.

Rose, C.S., Kuhn, R., Zillikcn, F. and Gyorgy, P ., 1954, "Bifidus Factor:

V. The Activity of a and fi -Mcthyl-N-Acetyl-D-GIucosamlnldes,"

Arch. Biochem. Biophys., 49:123.

Roses, D., 1969, "A Proposed Model of Micelle Strucluro in Bovine Milk,"

Dairy Sci. Abstr. , 31:171.

Sclmuer, R, and Fnillard, H., 1968, "The Action Specificity of Neuraminidase:

The Action of Bacterial Neuraminidase on Isomeric N, O, -dincetylncura-

minic Acid Glycosides in the Submoxlllary Mucin of Horse und Cow,"

Z. Physiol. Chem., 349:961.

Schmid, K. and Kamiyama, S., 1963, "Studies on the Structure of otj-Acid

Glycoprotein: IV. Optical Properties." Biochemistry, 2:271,

Schmidt, D.G., 1964, "Starch-Gel Electrophoresis of k -Casein." Biochim.

Biophys. Acta. 90:411.

Schmidt, D.G., Both, P. and Konlng, P .J., 1966, "Fractionation and Some

Properties of k -Casein Variants." J. Dairy Sci.. 49:776. Schnccgans, E., Haarscher, A., Lutz, A. and Schmittbuhl, J ., 1967, "Lacto­

bacillus bifidus implantation in the Healthy Infant and in the Carrier of

Pathogenic Escherichia coll, "

Schneir, M. L. and Hafclson, M. E ., 1966, "Isolation and Characterization of

Two Structural Isomers of N-Acetylneuraminyl Lactose from Bovine

.Colostrum," Biochim, Biophys. Acta. 130:1.

Schneir, M ., Winzlcr, J. and Rafelson, M. G., 1962, N-Acetvlneuraminlac-

tosc. John Wiley and Sons, Inc.

Sharma, K.K. and Hansen, P.M. T ., 1970a, "Heat-Induced, Gluey Flavor In

Isolated Casein Fractions," J. Dairy Sci., 53:640.

Sharma, K.K. and Ilanscn, P.M .T., 1970b, "Heat-Induced Dephosphorlzation

of Dehydrated Caseins," Proceedings. XVm International Dairy Con­

gress, Sidney, Australia, 1:58.

Sinkinson, G. and Wheelock, J.V ., 1970, "Carbohydrates of the Glycopeptides

Released by the Action of Rennin of Whole Milk," Biochim. Biophys. Acta,

215:517,

Soupnrt, P. and Clowe, T. H., 1965, "Sperm Penetration of Rabbit Zona Pellu-

cida Inhibited by Treatment of Ova With Neuraminidase," Fertility

Sterility, 16:677. 150

Spiro, M .J. and Spiro, R.G., 1962, "Composition of the Peptide Portion of

Fctuin," J. Biol. Chem.. 237:1507.

Spiro, M .J, and Spiro, R.G., 1968, ’.'Glycoprotein Biosynthesis: Studies on

Thyroglobulin: Tlivroid Sialvtransfcrase." J. Biol, Chem.. 243:6520.

Svcnnerholm, E. and Svenncrholm, L ., 1958, "Quantitative Paper Partition

* Chromatography of Sialic Acids." Nature, 181:1154.

Svcnnerholm, L ,, 1956, "On the Isolation and Characterization of N-Acetyl-

Sialic Acid," Acta Soc. Med. Upsal., 61:74.

Swecley, C. C., Wells, W.W. and Bentley, R ,, 1966, "Gas Chromatography of

Carbohydrates," Methods Enzymol., VIII;95.

Tettamanti, G., Bertona, L., Berra, B. and Zambotti, V., 1965, "Glycolyl-

Ncuraminic Acid in Ox Brain Gangliosides," Nature. 206:192.

Thompson, M.P. and Pepper, L., 1962, "Effect of Neuraminidase on K-Casein,"

J. Dairy Sci.. 45:794.

Tissicr, II., Recherchor sur la (lore intestinate des nourrlssons. These de

Paris 1900. Cited by Schuler, V. R ., Ruppert, A . and Muller, F ., 1968,

"The Microorganisms of the Bifidus Group," Milchwiss. , 23:356. Tomarclli, R.M ., Hassincn, J. B., Eckliardt, E. R ., Clark, R. H. and

Bcrnliart, F.W ., 1954, "Tlio Isolation of Crystalline Growth Factor

for Strains of Lactobacillus bifidus," Arch. Biochem. and Biophys.,

48:225.

Tran, V. D. and Baker, B. E., 1970, "Casein: IX. Carbohydrate Moiety of

K-Cascin," J. Dairy Sci. , 53:1009.

Tuppy, It. and Gottschalk, A., 1972, In Glycoproteins. "The Structure of

Sialic Acids and Their Quantitation," Edited by A. Gottschalk, Elsevier,

Amsterdam, p. 403.

Vclican, C ., Marcus, N. and Tacorian, S., 1970, "Increased Susceptibility to

Proteolytic Action of Gastric Superficial Mucosa After Sialic Ach* Release,

Am. J. Digest. Diseases, 15:413.

Vries, N. D. and Stouthamcr, A .It., 1968, "Fermentation of Glucose, Lactose,

Galactose, Mannitol and Xylose by Bifidobncteria," J. Bact. , 96:472.

Warren, L., 1959, "The Thiobarbituric Acid Assay of Sialic Acids," J. Biol.

Chem. , 234:1971.

Warren, L., 1972, In Glycoproteins, "The Biosynthesis and Metabolism of

Amino Sugars and Amino Sugar Containing Heterosaccharides," Edited by

A. Gottschalk, Elsevier Publ. Co., Amsterdam, p. 1123. Warren, L. and Bloeklow, R, S., 1962, "Biosynthesis of N-Acetylneuraminic

Acid and Cytldine-G'-Monophospho-N-Acetylneuraminic Acid in Neisseria

meningitidis," Biochem. Biophys. Res, Commun., 7:433.

Warren, L. and Fclsenfeld, H., 1962, "The Biosynthesis of Sialic Acids,"

J. Biol. Chem., 237:1421.

Watson, D. R., Jourdian, G.W. and Roscman, S., 1966, "Sialic Add 9-Phos-

photnsc Synthetase," J. Biol. Chem.. 241:5627.

Weiss, J. E. and Rcttger, L. F ., 1934, "Lactobacillus bifidus," J. Bact., 28:501.

Worncr, I. and Odin, L ., 1952, "On the Presence of Sialic Acid in Certain Gly­

coproteins and in Gangliosides," Acta Soc. Med. Unsalien.. 57:230.

Whcelock, J. V. and Knight, D. J ., 1969, "The Action of Rennet on Whole Milk,"

J. Dairy Res.. 36:183.

Whltchousc, M. W. and Zillikcn, F ., 1960, "Isolation and Determination of

Neuraminic (Sialicl Acids." Methods Bioch. Anal.. 8:194,

Wintorburn, P.J. and Phelps, C. F., 1972, "The Significance of Glycosylated

Proteins." Nature. 236:147.

Woolley, D. W., 1946a, "Some Correlations of Growth Promoting Powers of

Proteins With Their Streptogenin Content," J. Biol. Chem. . 162:383. 153

Woolley, D. W., 1946b, "Streptogenin Activity of Scryl Glycyl Glutamic Acid,"

J. Biol. Chem,, 166:783.

Zillikcn, F ., Braun, G. A. and Gyorgy, P ., 1956, "Gynaminlc Acid and Other

Naturally Occurring Forms of N-Acetylneuraminic Acid," Arch. Biochem.

Biophys. , 54*564.

Zillikcn, F. and O'Brien, P .J., 1959, "N-Acetylneuraminic Acid," Biochem.

Preparations, 7:1, John Wiley and Sons, Inc.

Zillikcn, F,, Rose, C.S., Braun, A.G. and Gyorgy, P., 1955, "Preparation

of Alkyl N-Acetyl- a - and 0-D-Glucosaminidcs and Their Biological

Activity for Lactobacillus bifidus var. Penn.," Arch. Biochem. Biophys.

54:392.

Zillikcn, F. and Whitehouso, M. W., 1958, "The Nonulosaminic Acids." Adv.

Carboh. Chem.. 13:237.