UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Separation of antimicrobial protein fractions from animal resources for potential use in infant feeding Al-Mashikhi, Shalan Alwan Edan 1987

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1987_A1 A42.pdf [ 7.45MB ]
Metadata
JSON: 831-1.0097117.json
JSON-LD: 831-1.0097117-ld.json
RDF/XML (Pretty): 831-1.0097117-rdf.xml
RDF/JSON: 831-1.0097117-rdf.json
Turtle: 831-1.0097117-turtle.txt
N-Triples: 831-1.0097117-rdf-ntriples.txt
Original Record: 831-1.0097117-source.json
Full Text
831-1.0097117-fulltext.txt
Citation
831-1.0097117.ris

Full Text

SEPARATION OF ANTIMICROBIAL PROTEIN FRACTIONS FROM ANIMAL RESOURCES FOR POTENTIAL USE IN INFANT FEEDING by SHALAN ALWAN EDAN AL-MASHIKHI B . S c , University of Baghdad M . S c , University of Baghdad A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Food Science) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA August, 1987 (c) Shalan Alwan Edan Al-Mashikhi, 1987 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 DE-6(3/81) i i ABSTRACT In the f i r s t part of this study, a non-ferr ic method for selective elimination of p-lactoglobulin from cheese whey was investigated. A new method was developed based on hexametaphosphate treatment of cheese whey. When Cheddar cheese whey was treated under the optimized condit ions, i . e . , 1.33 mg/mL sodium hexametaphosphate at 22°C and pH 4.07 for 1 hr, more than 80% of B-lactoglobulin was removed by prec ip i ta t ion . Almost a l l of the immunoglobulins and the major portion of a-lactalbumin were retained in the supernatant as indicated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immunochemical assays. By d ia lys is against d i s t i l l e d water 72.2% of the phosphorus was removed from the supernatant. In the second and the third part of the thes is , chromatographic methods were used for isolat ion of immunoglobulins and lactoferr in from whey proteins. By using gel f i l t r a t i o n on Sephacryl S-300, 99, 83.3 and 92.1% bio logica l ly active immunoglobulin G were obtained for colostral whey, acid and Cheddar cheese whey, respectively. Lactoferr in , se lect ive ly adsorbed to the heparin-attached Sepharose, was eluted with 5 mM Veronal-HCl containing 0.5M NaCl, at pH 7.2. 1,4-Butanediol d iglycidyl ether-iminodiacetic acid on Sepharose 6B, or so-cal led metal chelate-interaction chromatography (MCIC), was loaded with copper ion and used for the same purpose. Of the two peaks obtained, the f i r s t yellowish peak was rich in l ac to fe r r in , while the second peak was r ich in immunoglobulins. Some of the physical and chemical properties of the proteins in these peaks, including immunochemical propert ies, isoelectr ic points, binding to bacterial 1ipopolysaccharides, and the mechanism of protein-metal interaction via h ist id ine modif icat ion, and the capacity of the i i i method were studied. The p o s s i b i l i t y of isolat ing immunoglobulins and lactoferr in from electrodialyzed whey was also investigated. In the fourth, f i f t h and sixth parts of the thesis , the method developed for isolat ion of immunoglobulins and lactoferr in from whey protein was applied to isolate these b io log ica l ly important proteins d i rec t ly from skimmilk, blood and egg white. The casein in skimmilk was found to compete with immunoglobulins for binding to copper ion in MCIC column when skimmilk was loaded in presence of 0.05 M Tr is-acetate buffer containing 0.5 M NaCl, pH 8.2; however, this problem was solved by changing the equi l ibrat ing buffer to 0.02 M phosphate buffer containing 0.5 M NaCl, pH 7.0. When blood was d i rec t ly applied to MCIC column, the y ie ld of b io log ica l ly active IgG was more than 95%. Ovotransferrin, strongly adsorbed to the MCIC column, was eluted with two-step elution protocols which suggests i t exists in two forms. The hist id ine residues in immunoglobulins, caseins, t ransferr in and ovotransferrin were found to be involved in the mechanism of the interaction with the MCIC column. iv TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS iv LIST OF TABLES v i i LIST OF FIGURES v i i i APPENDIX xvii ACKNOWLEDGEMENTS . x v i i i INTRODUCTION 1 LITERATURE REIVEW 4 A. History 4 B. Human milk vs. cow's milk 5 C. Humanizing infant formula 9 D. Al lergencity of whey proteins 10 E. Antimicrobial system in human milk and milk substitutes 13 1. Immunoglobulins 13 2. Lactoferrin 18 3. Lactoperoxidase and lysozyme 19 4. Bif idobacteria 20 5. Ovotransferrin 22 F. Isolation of bioactive proteins 23 1. General methods 23 2. Metal chelate interaction chromatography 24 MATERIALS AND METHODS 26 A. Materials 26 B. Acid whey preparation 26 C. Sodium dodecyl sulfate polyacrylamide gel electrophoresis 27 1. Discontinous SDS-PAGE 27 2. Gradient SDS-PAGE 28 D. Immunochemical analysis 28 1. Immunoelectrophoresis and immunodiffusion 28 2. Enzyme linked immunosorbent assay for anti-1ipopolysacchardies ac t iv i ty determination 29 3. Sandwish enzyme linked immunosorbent assay for IgG assays 30 E. Sodium hexametaphosphate treatment of cheese whey 30 F. Optimization procedure 32 1. Mapping super simplex 32 2. Centroid mapping optimization and simultaneous factor sh i f t . . . . 32 G. Evaluation of separation ef f ic iency (response value) 33 H. Surface plot 34 I. Phosphorus determination 34 V Page J . Fractionation procedures of bioactive components 34 1. Gel f i l t r a t i o n chromatography 34 2. S i l i c a adsorption chromatography 35 3. Heparin-Sepharose chromatography 35 4. Metal chelate- interaction chromatography 35 K. Determination of capacity of MCIC 36 1. Immunoglobulins 36 2. Ovotransferrin 39 3. Transferrin 40 L. Preparation of egg white 40 M. Preparation of apo, d i f e r r i c and dicupric ovotransferrin 40 N. Hist idine modification of proteins 41 0. Ise lectr ic focusing 41 P. Preparation of antisera 41 Q. Measurement of bacter iostat ic ac t iv i ty 42 R. Extraction of 1ipopolysaccharides 42 S. Lactoperoxidase assay 43 T. Separation of heavy and l ight chains of immunoglobulins 44 RESULTS AND DISCUSSION 45 PART I. Reduction of B-lactoglobulin content of cheese whey by using sodium hexametaphosphate 45 A. Optimum conditions for separation of immunoglobulins and B-lactoglobulin 46 B. Three dimensional i l l us t ra t ion of effects of pH and hexametaphosphate on separation ef f ic iency 48 C. Elimination of phosphorus 53 D. Proposal of new infant formula 55 PART II. Separation of bovine immunoglobulins and lactoferr in from whey proteins by gel f i l t r a t i o n techniques 57 A. Gel f i l t r a t i o n of Sephacryl S-300 58 B. Gel f i l t r a t i o n on TSK HW-5S 61 C. Isolation of immunoglobulins from whey proteins 67 D. Isolation of lactoferr in from whey proteins 67 E. Anti-1ipopolysacchardies ac t iv i ty of isolated immunoglobulins 75 PART III. Separation of immunoglobulins and lactoferr in from cheese whey by adsorption and chelating chromatography techniques 79 A. Adsorption chromatography methods 80 B. Metal chelate-interaction chromatography 83 1. Acid whey 83 2. Cheddar cheese whey 86 3. Electrodialyzed and sweet whey powders 90 a. IgG ac t iv i ty and protein content 90 b. Effect of pH adjustment 93 c. Isolation of immunoglobulins by controlled pore glass 97 d. Isolation of immunoglobulins by metal chelate- interaction chromatography 99 C. Binding capacity and recovery of immunoglobulins from metal chelate-interaction chromatography column 104 vi Page D. Anti-1ipopolysaccharides ac t iv i ty of immunoglobulins r ich fract ion 106 E. Lactoperoxidase content of lactoferr in r ich fract ion 108 F. Identi f icat ion of glycoproteins in lactoferr in r ich fract ion 108 G. Isoelectr ic points of lactoferr in and immunoglobulins rich fract ions 110 H. Hist idine modification and metal chelate- interaction chromatography properties 110 I. Separation of heavy and l ight chains of immunoglobulins 112 J . Separation of lactoferr in and lactoperoxidase in lactoferr in r ich fract ion 117 1. Gel f i l t r a t i o n method 117 2. Stepwise pH elution 117 3. Imidazole gradient elution 121 PART IV. Metal chelate- interaction chromatography of skimmilk 124 A. MCIC with Tris-acetate buffer 125 B. MCIC with phosphate buffer 129 C. Mechanism of casein-metal interaction . . 132 1. a-casein 133 2. a s j and 8-casein 133 3. K-casein 137 PART V. Separation of immunoglobulins and transferr in from blood serum and plasma by metal chelate- interaction chromatography 141 A. Metal chelate- interaction chromatography 142 1. Blood serum on Cu-loaded MCIC 142 2. Blood serum on MCIC columns loaded with other metal ions 142 3. Blood plasma on MCIC column 147 B. Immunochemical assays 149 C. Bacter iostat ic ac t iv i ty of blood immunoglobulins and transferr in 149 D. Anti-1ipopolysaccharides ac t iv i ty of blood immunoglobulins 149 E. Capacity of MCIC column for t ransferr in 152 F. Mechanism of protein-metal interaction 155 PART VI. Separation of ovotransferrin from egg white by metal chelate- interact ion chromatography 157 A. Metal chelate- interaction chromatography of egg white 158 B. Capacity of MCIC column for ovotransferrin 162 C. Mechanism of ovotransferrin separation by MCIC 164 CONCLUSIONS AND RECOMMENDATIONS 168 REFERENCES CITED 170 APPENDIX 184 v i i LIST OF TABLES Page Table 1. Gross Composition of Human and Cow's Milk (Grams per 100 Grams of Fluid Product) 3 6 Table 2. Protein Composition of Human and Cow's Mi lk a 8 Table 3. Concentration of bovine immunoglobulins in serum and secretions (mg/mL)a 15 Table 4. Biochemical character ist ics of bovine immunoglobuinsa 17 Table 5. Phosphorus distr ibut ion in supernatant and precipitate obtained by SHMP treatment. P, phosphorus 54 Table 6. Protein composition of human and cow's milks and whey-based and proposed B-lactoglobulin (B-Lg)-free infant formula 56 Table 7. Immunoglobulin G contents* of f ract ions, obtained from gel f i l t r a t i o n on Sephacryl S-300 and crude Ig prepared by ammonium sulfate treatment 62 Table 8. IgG ac t iv i ty of peak and unbound fractions from MCIC treatment of whey 87 Table 9. Whey proteins d is t r ibut ion* in acid whey, fractions obtained from MCIC of bovine acid whey, and the unbound materials to MCIC column 88 Table 10. IgG and protein contents of reconstituted ED and sweet whey powders compared to l iqu id cheese whey 92 Table 11. IgG ac t iv i ty of pH 4.5 and pH 8.2 supernatant (S) and precipitate (P) fractions from sweet whey and ED wheya 96 Table 12. IgG content of di f ferent stages of the isolat ion of IgG from skimmilk on MCIC column 128 Table 13. Binding of casein fractions to MCIC column before and after modification of hist id ine groups 135 Table 14. IgG contents* of blood serum or plasma fractions obtained from MCIC column loaded with dif ferent metal ions 150 LIST OF FIGURES Flow diagram of the procedure for elimination of B-lactoglobulin from Cheddar cheese whey with SHMP Preparation of metal chelate agarose Flow chart of MCIC process of cheese whey treatment for isolat ion of Ig , Approximate response surface patterns for (A) pH and (B) SHMP concentration obtained by mapping accumulated data from simplex optimization (Vertices 1-9) and centroid optimization (Vertices 10-15). T, target values of pH and SHMP SDS-PAGE of supernatant (S) and precipitate (P) obtained after treatment with 1.33 mg/mL SHMP at pH 4.07. CCW, Cheddar cheese whey; a -La , a-lactalbumin; B-Lg, B- lactoglobul in; IgG-HC, immunoglobulin G heavy chain; IgG-LC, immunoglobulin G l ight chain; BSA, bovine serum albumin; OVA, ovalbumin Immunoelectrophoretic pattern of Cheddar cheese whey. S, supernatant;CCW, Cheddar cheese whey; P, prec ip i ta te; IgG, immunoglobulin G; B-Lg, B- lactoglobul in; abwp, antibovine whey proteins Contour (A) and 3-dimensional (B) surface plots of relationship between pH, SHMP and Separation effeciency (SE) of cheese whey treatment. ("about" angle=60 and "above" angle=35 for 3-dimensional plot) Gel f i l t r a t i o n of bovine colostral whey on Sephacryl S-300 Superfine column (94 x 2.5 cm) eluted with 0.1 M Tris-HCl buffer pH 8.0 containing 0.5 M NaCl. Flow rate, 12 mL/hr. 1 and 3 are fractions 1 and 3, respectively Gel f i l t r a t i o n of crude Ig obtained from ammonium sulfate treatment on Sephacryl S-300 Superfine column (94 x 2.5 cm), eluted with 0.1 M Tris-HCl buffer pH 8.0 containing 0.5 M NaCl, flow rate 12 mL/hr. 1 and 3 are fractions 1 and 3, respectively SDS-PAGE of fractions obtained from gel f i l t r a t i o n on Sephacryl S-300. F l , fract ion 1; F3, fract ion 3 (Figure 8); Ig, crude immunoglobulin; LF, l ac to fer r in ; CW, untreated colostral whey; HC, LC, immunoglobulin heavy and l ight chains, respectively ix Page Figure 11. Immunoelectrophoretic analysis against anti-whole bovine antiserum of fractions obtained from Figure 8. F3, f ract ion 3; F l , f ract ion 1; BSA, bovine serum albumin; TF, transferr in 64 Figure 12. Figure 13. Figure 14, Figure 15. Gel f i l t r a t i o n pattern of colostral whey on TSK column (40 x 2.6 cm) eluted with 0.07 M imidazole-0.05 M KC1 buffer, pH 6.5; flow rate, 50 mL/hr. 1 and 2 are fractions 1 and 2, respectively SDS-PAGE of fractions (Fl and F2) obtained from Figure 12 as compared to standards. Lane 1, a-lactalbumin; Lane 2, B- lactoglobul in; Lane 3, bovine serum albumin; Lane 4, t ransfer r in ; M, standard mixture; AW, acid whey, CW, colostral whey; HC and LC, immunoglobulin heavy and l ight chains, respectively 65 66 Immunoelectrophoresis of fractions obtained by Sephacryl S-300 and Fractogel TSK column against anti-whole bovine serum antiserum. (Fl-S and F3-S, fract ion 1 and 3 of Figure 8, respectively) (Fl-T and F2-T fract ion 1 and 2 of Figure 12, respectively) IgG, immunoglobulin G 68 Gel f i l t r a t i o n of acid whey on Sephacryl S-300 Superfine column (94 x 2.5 cm), eluted with 0.1 M Tris-HCl buffer, pH 8.0 containing 0.5 M NaCl. Flow rate 18 mL/hr. 1, 2, 3 and 4 are the fractions obtained 69 Figure 16. Figure 17. Figure 18, Figure 19. Gel f i l t r a t i o n of Cheddar cheese whey on SepharylS-300 Superfine column (94 x 2.5 cm), eluted with 0.1 M Tris-HCl buffer, pH 8.0 containing 0.5 M NaCl. Flow rate 18 mL/hr. 1, 2, 3 and 4 are the fractions obtained , SDS-PAGE of fractions (1, 2, 3, 4) obtained from Figure 15. AW, acid whey; IgG, immunoglobulin G; HC and LC, heavy and l ight chains of immunoglobulins, respectively SDS-PAGE of fractions (1, 2, 3, 4) obtained from Figure 16. IgG, immunoglobulin G; CCW, Cheddar cheese whey; HC and LC, heavy and l ight chains of immunoglobulins, respectively Immunoelectrophoresis of fractions obtained from gel f i l t r a t i o n of whey protein against anti-whey protein antiserum. AW, acid whey; F l - H , fraction 1 from Figure 20; F2-A, fract ion 2 from Figure 15; F2-C, fraction 2 from Figure 16; IgG, immunoglobulin G 70 71 72 73 X Page Figure 20. Heparin-Sepharose chromatography of Cheddar cheese whey. Cheese whey (400 mL) dialyzed against 0.05 M NaCl in 5 mM veronal-HCl, pH 7.4 was applied to the column (10 mL sett led ge l ) . The column was washed with the same buffer and then eluted with a l inear gradient of NaCl (....) as indicated. The flow rate was 50 mL/hr. UB, unbound proteins; 1, fract ion 1 74 Figure 21. SDS-PAGE of fractions obtained from Figure 20, CCW, Cheddar cheese whey; UB; unbound whey proteins to Heparin-Sepharose column; F l , lactoferr in r ich f rac t ion; LF, lactoferr in 76 Figure 22. Anti-1ipopolysaccharide ac t iv i ty of colostral IgG isolated by gel f i l t r a t i o n on Sephacryl S-300. U-U , E. col i LPS; D-n , S_;_ typhi murium LPS; O - O , parapertussis LPS 77 Figure 23. Elution prof i les of adsorbed proteins from s i l i c a (S), controlled pore glass (C) and alumina (A) chromatographic treatment of Cheddar cheese whey. One l i t r e of Cheddar cheese whey in 0.005 M Na 2HP0 4, pH 8.2 was passed through 1.3 x 7.0 cm column of (S), (C) or (A) equil ibrated with 0.005 M phosphate buffer at pH 8.2. After washing with 30 mL of equi l ibrat ing buffer, the adsorbed proteins were eluted with El (50 mL 0.1 M acetic acid pH 2.77 containing 0.5 M NaCl), then E2 (60 mL 0.1 M Tris-HCl pH 9.0 containing 0.5 M NaCl). The flow rate was 1 mL/min 81 Figure 24. SDS-PAGE prof i les of cheese whey and fractions obtained from Figure 23.Lane 1, untreated Cheddar cheese whey; Lane 2, acetic acid fraction from s i l i c a sand; Lane 3, acetic acid fract ion from controlled pore g lass; Lane 4, acetic acid fract ion from alumina; Lane 5, Tris-HCl fract ion from s i l i c a sand; Lane 6, Tris-HCl fract ion from controlled pore g lass; Lane 7, Tris-HCl fraction from alumina; Lane 8, unbound fract ion from s i l i c a ; Lane 9, unbound fract ion from controlled pore g lass; Lane 10, unbound fraction from alumina; Lane 11, untreated Cheddar cheese whey; Lane 12, acet ic acid fract ion from alumina, LF, l ac to fe r r in ; HC and LC, immunoglobulin heavy and l ight chains, respectively 82 Figure 25. Elution prof i les of adsorbed proteins from MCIC on Sepharose 6B treatment of 1 L Cheddar cheese whey (CCW) and 1 L acid whey (AW) (obtained from raw mi lk) , using l inear gradient elution of 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.0 to 2.8. Flow rate was 0.8 mL/min. CCW, Cheddar cheese whey; AW, acid whey 84 xi Page Figure 26. SDS-PAGE prof i les of acid whey and fractions obtained by MCIC and gel f i l t r a t i o n . Lanes 1 and 2, are fractions obtained by MCIC; Lane 3, unbound material to MCIC column; Lanes 4 and 5, peak 1 and 2 obtained by gel f i l t r a t i o n on Sephacryl column; Lane 6, mixture of standard proteins ( t ransferr in , bovine serum albumin, B-lactoglobulins and a-lactalbumin); Lane 7, acid whey, LF, l ac to fe r r in ; HC and LC, heavy and l ight chains of immunoglobulins, respectively . . . . 85 Figure 27. Immunoelectrophoretic analysis of fractions obtained by MCIC of acid whey. IgG, immunoglobulin G; P2 and PI are Peak 2 and 1, respectively of Figure 25; TF, t ransfer r in ; BSA, bovine serum albumin 89 Figure 28. SDS-PAGE of Cheddar cheese whey and fractions obtained by MCIC on Sepharose 6B treatment. Lane 1, control cheese whey; Lanes 2, 3 and 4, areunbound f rac t ion; Lanes 5 and 7, f i r s t eluted peak; Lanes 6 and 8, second eluted peak; LF, l ac to fe r r in ; HC and LC heavy and l ight chains of immunoglobulin, respectively 91 Figure 29. SDS-PAGE prof i le of l iquid cheese whey after pH adjustment and centr i fugat ion. Lanes 1, 3, 5, 7, 9 and 11 are the precipitate and Lanes 2, 4, 6, 8, 10 and 12 are the supernatant of samples treated at pH 8.5, 8.0, 7.0, 6.0, 5.0 and 4.5, respectively, LF, l ac to fer r in ; BSA, bovine serum albumin; C, casein 94 Figure 30. SDS-PAGE prof i les of fractions from ED whey after pH adjustment and centriguation. Lanes 1 and 3, are precipitate at pH 8.2 and 4.5 respectively; Lanes 2 and 4, are supernatant at pH 8.2 and 4.5 respectively; Lane 5, electrodialyzed whey; LF, l ac to fe r r in , HC and LC, heavy and l ight chains of immunoglobulins, respectively 95 Figure 31. Elution pro f i l e of adsorbed proteins from CPG (10 mL) treatment of 250 mL of electrodialyzed whey (EDI), 250 mL of sweet whey (SW), and 1 L of ED whey (ED2). Arrows indicate start of elution with El (0.1 N acet ic acid pH 2.8 containing 0.5 M NaCl) and E2 (0.1 M T r i s - H C l , pH 9.0 containing 0.5 M NaCl) buffers 98 Figure 32. Elution pro f i l e of adsorbed proteins from MCIC on Sepharose 6B treatment of 960 mL electrodialyzed whey (EDW) and 720 mL sweet whey (SW)powders reconstituted in water, using l inear gradient elution of 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2 to 2.8. Flow rate was 0.8 mL/min. 1 and 2 are fractions obtained 100 xi i Page Figure 33. SDS-PAGE prof i les of ED whey and fractions obtained by MCIC on Sepharose 6B treatment. Lane 1, control untreated whey; Lanes 2, 3 and 4, are unbound f ract ions; Lane 5, wash f rac t ion; Lane 6, f i r s t eluted peak; Lane 7, second eluted peak; BSA, bovine serum albumin; HC, heavy chain of immunoglobulins 101 Figure 34. SDS-PAGE prof i les of sweet whey and fractions obtained by MCIC on Sepharose 6B treatment. Lanes 1, 12, control untreated whey; Lane 2, precip i ta te; Lanes 3, 4 and 5; unbound f ract ions; Lanes 6, 7, 8 and 9, wash f ract ions; Lane 10, f i r s t eluted peak; Lane 11, second (shoulder) eluted peak; BSA, bovine serum albumin 102 Figure 35. Saturation point for adsorption of crude Ig (prepared from colostrum by ammonium sulfate method) on Cu-loaded IDA-BGE Sepharose 6B (SROSE) and Sephacryl S-300 (SACRYL). 0.3% crude Ig was passed through a 10 mL column (7.0 x 1.4 cm) equil ibrated with 0.05 M Tris-acetate 0.5 M NaCl, pH 8.2. W, washing with the start ing buffers; E, elution with 0.05 M Tr is-acetate /0 .5 M NaCl, pH 4.0. The flow rate was 20 mL/hr 105 Figure^36. Anti-1ipopolysaccharide ac t iv i ty of Ig isolated from cheese whey by MCIC method. • - • , E^ col i LPS; o - a S^ typhi murium LPS; O - O . i L parapertussis LPS 107 Figure 37. SDS-PAGE of whey proteins (1), lactoferr in r ich fraction (2), lactoferr in (3) and lactoperoxidase (4). (A) stained with Commassie Br i l l an t Blue and (B) stained with periodic acid Schif f (PAS). HC and LC heavy and l ight chains respectively 109 Figure 38. Elution prof i les of control (Ig) and diethyl pyrocarbonate treated immunoglobulins (DEP Ig). Samples (30 mg/5 mL 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2) were applied to the column (1.4 x 7.0 cm) and washed (W) with the start ing buffer then eluted (E) with 0.01 M imidazole. Flow rate was 30 mL/hr I l l Figure 39. Elution prof i les of Fl-MCIC fract ion before (Fl) and after diethyl pyrocarbonate treatment (DEP F l ) . Samples (30 mg/5 mL 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2) were applied to the column (1.4 x 7.0 cm) and washed (W) with the start ing buffer then eluted (E) with 0.01 M imidazole. Flow rate was 30 mL/hr 113 xi i i Page Figure 40. Figure 41. Figure 42. Figure 43, Elution prof i les of reduced and alkylated heavy and l ight chains of immunoglobulin on Sephadex G-75 eluted with 1 M propionic ac id . 1 and 2 are fractions obtained 114 SDS-PAGE prof i les of heavy and l ight chains of immunoglobulins isolated by gel f i l t r a t i o n . Lanes 1 and 2, crude immunoglobulins; Lane 3 and 4, immunoglobulin l ight and heavy chains, respectively obtained from Figure 40 115 Elution prof i le of reduced and alkylated heavy and l ight chains of Ig on Ultrogel ACA 54 eluted with 0.1 M Tris-HCl buffer containing 4 M Guanidine-HCl and 1 mM iodoacetamide, pH 8.2. 1 and 2 are fractions obtained 116 Sephacryl S-300 column chromatography of lactoferr in r ich fract ion obtained by MCIC of acid whey. 100 mg sample was applied to Sephacryl column (83 x 2.5 cm) and eluted with 0.05 M potassium phosphate buffer, pH 7.4 containing 0.01 M NaCl. 1 and 2, are fractions obtained. The flow rate was 30 mL/hr 118 Figure 44. Stepwise elution prof i le of acid whey on MCIC eluted by decreasing pH values. Arrows indicate pHs 7, 6, 5 and 4 of 0.05 M Tris-acetate containing 0.5 M NaCl. 1 and 2 are fractions obtained 119 Figure 45. Elution prof i le of bound proteins of acid whey on MCIC column, eluted (E) by using pH gradient (5-2.8) of 0.05 M Tris-acetate containing 0.5 M NaCl. 1 and 2 are fractions obtained 120 Figure 46. Elution pro f i l e of lactoferr in r ich fract ion on MCIC column. E l , elution with l inear gradient of 0-10 mMimidazole solution ( . . . . ) ; E2, elution with 0.05 M Tris-acetate containing 0.5 M NaCl, pH 2.8. 1, 2 and 3 are fractions obtained 122 Figure 47. Figure 48. SDS-PAGE prof i les of fractions obtained from Figure 46. Lane 1, whey proteins; Lane 2, control Fl-MCIC; Lane 3, unbound f rac t ion; Lanes 4, 5 and 6 are peak 1, 2 and 3 of Figure 46; LF, l ac to fe r r in , LP, lactoperoxidase 123 Elution prof i le of skimmilk on MCIC column. 100 mL skimmilk undiluted (SM) or 50% diluted (DSM) with 0.05 M Tr is-acetate /0 .5 M NaCl) was passed through Cu-loaded Sepharose 6B (1.4 x 7.0 cm), and washed (W) with same buffer. E l , elution with the same buffer at pH 4.0; E2, elution with 0.01 M imidazole solut ion. 1 and 2 are eluted f ract ions. The flow rate was 21 mL/hr 126 xiv Page Figure 49. SDS-PAGE of fractions obtained in Figure 48. Lanes 1 and 2, skimmilk; Lane 3, unbound skimmilk to MCIC column; Lane 4, washing f ract ion; Lanes 5 and 6, are peak 1 and 2, respectively; Lane 7, standard IgG; Lane 8, a-casein 127 Figure 50 Elution prof i les of skimmilk (SM), Ig and LF mixture on MCIC column. 60 mg of Ig and LF was mixed with 1 mL skimmilk (SM+Ig+LF) and 30 mg Ig was mixed with 1 mL skimmilk (SM+Ig) and passed through MCIC column (1.4 x 7.0 cm). W, washing with 0.02 M phosphate buffer containing 0.5 M NaCl, pH 7.0; E l , elution with 0.01 M imidazole; E2, elution with Tris-acetate containing 0.5 M NaCl, pH 3.0 130 Figure 51 SDS-PAGE of fractions obtained in Figure 50. Lane 1, skimmilk; Lane 2, skimmilk and Ig mixture; Lanes 3 and 4 are unbound and peak 1 of SM-Ig mixture appl icat ion, respectively; Lane 5, SM-Ig-LF mixture; Lanes 6 and 7 are unbound and peak 1 of SM-Ig-LF mixture appl icat ion, respectively; Lane 8, immunoglobulins; Lane 9, lactoferr in and Lane 10, a-casein 131 Figure 52. Metal chelate interaction chromatography of a-casein. 3 mL of protein (10 mg/mL) before (a-CAS) and after diethylpyrocarborate modification (DEP a-CAS) equil ibrated with 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2 and applied to copper chelate Sepharose 6B (1.4 x 7.0 cm). W, washing with the same equi l ibrat ing buffer; E, elution with 0.01 M imidazole. Flow rate was 30 mL/hr 134 Figure 53. Metal chelate interaction chromatography of a s i - c a s e i n before (as-CAS) and after diethylpyrocarbonate modification (DEP as-DAS). See Figure 52 for conditions of separation 136 Figure 54. Metal chelate interaction chromatography of 8-casein before (8-CAS) and after diethylpyrocarbonate modification (DEP B-CAS). See Figure 52 for conditions of separation 138 Figure 55. Metal chelate interaction chromatography of (A) polymer K-casein ( K - C A S ) (B) monomer K-caseins (MK-CAS) before and after diethylpyrocarbonate modification (DEP K - C A S ) . See Figure 52 for separation conditions 139 Figure 56. Immobilized copper a f f in i t y chromatography of blood serum. Blood serum (1 g in 10 mL 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2) was applied to the column (1.4 x 7 cm). The column was washed with the starting buffer and then eluted with E l , 0.05 M Tris-acetate 0.5 M NaCl, pH 4.0, and with E2, 0.1 M imidazole. The flow rate was 30 mL/hr. Fl and F2 are fractions obtained 143 X V Page Figure 57. SDS-PAGE prof i les of blood fractions from MCIC on Sepharose 6B column. Sample ident i f ica t ion: Lanes 1, 2 and 3 are Fl of Figure 59 from Zn, Ni and Co loaded columns respectively; Lane 4, plasma protein eluted from Cu-loaded column with pH 4 buffer; Lanes 5, 6 and 7 are unbound, F l , and F2 in Figure 59, respectively; Lanes 8, 9 and 10 are standard transferr in(TF) , bovine serum albumin (BSA), and immunoglobulins ( Ig)respectively; Lane 11, blood plasma 144 Figure 58. Immunoelectrophoresis of fractions obtained in Figure 56.P, blood plasma; Fl and F2 fractions obtained in Figure 56; TF, t ransfer r in ; Ig, immunoglobulins, BSA, bovine serum albumin; abws, rabbit antibovine whole serum 145 Figure 59. Immobilized Zn- , N i - and Co- a f f i n i t y chromatography of blood serum. Blood serum (2 g in 20 mL 0.05 M Tr is-HCl /0 .15 M NaCl, pH 8.0) was applied to the column (2.8 x 8.5 cm). The column was washed with the start ing buffer then eluted (E) with 0.1 M Na-acetate/0.8 M NaCl, pH 4.6. The flow rate was 30 mL/h. 1, is fract ion obtained 146 Figure 60. Elution prof i les of adsorbed hemoglobin from MCIC columns (1.4 x 7.0 cm) loaded with Zn, N i , Co and Cu. 2 mL of hemoglobin (3 mg/mL in 0.05 M Tr is-HCl /0 .15 M NH«C1, pH 8.0) was applied to the column and washed (W) with 2-3 times bed volumes of the start ing buffer. E l , 0.1 M Na-acetate /0.8 M NaCl, pH 4.5; E4, 50% ethanol 148 Figure 61. Bacter iostat ic ac t iv i ty of isolated immunoglobulins and transferr in against E^ c o l i . C, control ; TF, transferr in (10 mg/ml); M, mixture of TF (5 mg/ml) and Ig (5 mg/ml); Ig, immunoglobulins (10 mg/ml) 151 Figure 62. Figure 62. Anti-1ipopolysaccharide ac t iv i ty of blood IgG isolated by metal chelate interaction chromatography method. E^ col i LPS; o - a , typhimurium LPS; O - O » JL parapertussis LPS 153 Figure 63. Saturation point of adsorption of standard TF on Cu-loaded IDA-BGE Sepharose 6B. 0.2% TF was passed through 10 mL column (7 x 1.4 cm) equil ibrated with 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2, W, wash with start ing buffer , E l , elution with 0.05 M Tris-acetate containing 0.5M NaCl, pH 4.0. Elution with 0.01 M imidazole 154 Figure 64. Elution prof i les of control (TF) and diethyl pyrocarbonate treated transferr in (DEP-TF). Samples (30 mg/5 mL 0.05 M Tr is-acetate /0 .5 M Nacl, pH 8.2) were applied to the column (1.4 x 7.0 cm) and washed (W) with the start ing buffer then eluted (E) with 0.01 M imidazole. The flow rate was 30 mL/hr 156 xvi Page Figure 65. Metal chelate- interaction chromatography of egg white. 2 mL of undiluted blended egg white was passed through Cu-loaded Sepharose 6B MCIC column (7 x 1.4 cm). UB, unbound proteins; FW, fraction eluted with washing step; E l , elution with start ing buffer, pH 4.0; 1, fraction eluted with E l ; E2, elution with 0.01M imidazole; 2, fract ion eluted with E2 159 Figure 66. SDS-PAGE of fractions of egg white obtained by MCIC column shown in Figure 65. Lanes 1, 2 and 3, are unbound f ract ions; Lane 4, washing f rac t ion; Lane 5, peak 1; Lane 6, peak 2; Lanes 7 and 8, standard ovotransferrin and ovalbumin, respectively; Lanes 9 and 10, control egg white 160 Figure 67. Immunoelectrophoresis against anti whole egg white antiserum of ovotransferrin fract ion (F) prepared by the MCIC method as compared to commercial ovotransferrin (OVT), and egg white (EW) 162 Figure 68. Saturation prof i le of commercial ovotransferrin on Cu-loaded Sepharose 6B column. 0.2% ovotransferrin was passed through 3 mL of a Cu-loaded column (7 x 1.4 cm). E l , elution with 0.05 M Tris-acetate /0.5 M NaCl, pH 4.0; E2, elution with 163 Figure 69. Metal chelate interaction chromatography of apo-ovotrans-fer r in (APO-OVT), Fe-ovotransferrin (Fe-OVT) and Cu-ovotransferrin (Cu-OVT). 3 mL (8 mg/mL) was applied to Cu-loaded Sepharose 6B (7 x 1.4 cm) after equi l ibrat ion with 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2. W, Washing with the equi l ibrat ing buffer; E, elution with 0.01 M imidazole; flow rate was 30 mL/hr 165 Figure 70 Metal chelate interaction chromatography of control ovotransferrin (OVT) and diethyl pyrocarbonate treated ovotransferrin (DEP-OVT). 3 mL (8 mg/mL) was applied to Cu-loaded Sepharose 6B (7 x 1.4 cm) after equi l ibrat ion with 0.05 M Tr is-acetate /0 .5 M Nacl, pH 8.2. W, washing with the equi l ibrat ing buffer; E, elution with 0.01 M imidazole; flow rate was 30 mL/hr 166 X V I 1 APPENDIX Page Non-ferric methods for B-lactoglobulin removal from Cheddar cheese whey 184 \ xvi i i ACKNOWLEDGEMENTS I would l ike to express my sincere gratitude and appreciation to my supervisor Dr. Shuryo Nakai for his unlimited encouragement, invaluable advice and enthusiasm throughout the course of this investigation and in the preparation of this thesis . I also would l ike to thank the members of the committee, Drs. Brent Skura and William Powrie, Department of Food Science, and Dr. Robert Fitzsimmons, Department of Animal Science for their valuable suggestions and assistance. Thanks also to Sherman Yee who eased the lab work. Thanks are also extended to my parents for their constant support throughout the years of study. A special thanks to my wife for her encouragement and patience during the study. F i n a l l y , I wish to express my deep appreciation and gratitude to the Iraqi Government for giving me the opportunity to study abroad and for the f inancial support provided to me throughout the study. 1 INTRODUCTION In developed soc ie t ies , many infants are fed infant formula rather than human milk. National and international organizations such as the American Academy of Pediatrics (Committee on Nutr i t ion, 1980), the European Society for Pediatr ic Gastroenterology and Nutrit ion (ESPGAN, 1982) and the World Health Organization (1981), recommend breast feeding of infants whenever possible. There are many cases, however, where i t is d i f f i c u l t for women to nurse infants due to short supply of breast milk, insuf f ic ient nutr i t ion and health conditions of nursing mothers, and the necessity of some mothers having to work. Thus, continuing effort has been made to decrease nutrit ional and protective ab i l i t y differences between human milk and cows' milk (Kuwata et a l . , 1985). Although, there are many dif ferent infant formulas on the market today, a l l of them have not reached the required standard, i . e . human milk (Wilkinson, 1981). Nutrit ion and prevention of disease both have to be considered when dealing with infant formulae. It is believed that the premature infant can benefit from breast milk to an even greater extent than can the ful l - term infant (Workshop Part ic ipants, 1976). The most compelling argument in favor of breast feeding i s , however, the contention that breast milk contains factors that protect the infant against both systemic and gastrointestinal infect ions. The most dramatic testimonials in this regard are provided by the results of nursery epidemics where otherwise normal infants quickly succumb to a variety of infections unless they are fed raw human milk (Dortmann, 1967; Gerrard, 1974). The protection accorded the newborn by breast milk is also c r i t i c a l for premature infants where a re lat ively common fatal syndrome cal led necrotizing enterocol i t is can be prevented by breast feeding (Mizrawi et a l . , 1965; Touloukian et a l . , 1967; Barlow et a l . , 1974). It also has been reported that 2 the incidence of sudden death syndrome in infants is lower in breast fed infants than i t is in bott le- fed infants (Mobbs, 1972). The resistance of breast-fed infants against infection is summarized in several review a r t i c les (Hanson and Winberg, 1972; Goldman and Smith, 1974; Bezkorovainy, 1977; Rieter , 1978; Gurr, 1981; Packard, 1982; Friend et a l . , 1983; Lonnerdal, 1985; Rieter , 1985a) which ascribe the beneficial effects of human milk to factors that are therein but are absent from bovine milk. These factors are, for the most part , proteins in nature, the most important of which are immunoglobulin A, l ac to fe r r in , lysozyme, lactoperoxidase, leucocytes, the bif idus factor and interferon. Manufacturers of infant formulae must consider the factors mentioned above and not just the nutrient content, i . e . , carbohydrate, protein, fa t , mineral and vitamins. Therefore, one possible way to humanize cows' milk is to enrich cow's milk with these bioactive components. Prevention of pathogenic infections is only part of the problem. One must consider the prevention of immune response reaction in infants. The newborn is par t icu lar ly vulnerable. For a few days the stomach is porous, even to whole proteins. The milk of any non-human species is more l i ke ly to induce an allergy in sensit ive infants than mother's milk. In general, the occurrence of a l lergenic reactions due to infant formulae is much more common than human milk (Packard, 1982). Animal experiments suggest that B-lactoglobulin is more antigenic than either casein or a-Lactalbumin found in cow's milk. Compiled data from f ive studies of cows' milk protein intolerance in infancy showed sens i t iv i ty to p-lactoglobulin in 82% of the cases (Wilkinson, 1981). 8-Lactoglobulin has been regarded as a major allergen for bott le-fed infant in several outstanding papers (Lebenthal, 1975; Wharton, 1981; Moneret-Voutrin et a l . , 1982; Heppell et a l . , 1984; Kurisaki et a l . , 1985; Otani et a l . , 1985; 3 Pahud et a l . , 1985). A l lergenic i ty of B-lactoglobulin may be attributed to the fact that human milk, based on immunological reactions, contains only a trace of p- lactoglobulin (Brignon et a l . , 1985). Therefore, the elimination of B-lactoglobulin is one way to humanize the protein composition of cow's milk for infant feeding. Several attempts have been made for separating B-lactoglobulin from bovine whey (Forsum, 1974; Mathur and Shahani, 1979; Amundson and Watanawamchakorn, 1982). In our laboratory, a f e r r i c chloride method was established for se lect ive ly precipitat ing 3- lactoglobulin from bovine whey (Kaheko et a l . , 1985; Kuwata et a l . , 1985). However, because of the potential loss of the antimicrobial ac t iv i ty of lactoferr in when i t is saturated with i ron , non-ferr ic methods have been recently investigated. The objectives of this study were (1) to establish a new non-ferr ic method for eliminating p-lactoglobulin from cheese whey; (2) to investigate different chromatographic methods for the isolat ion of immunoglobulins and lactoferr in from cheese whey; (3) to investigate the p o s s i b i l i t y of using skimmilk d i rect ly for isolat ion of immunoglobulins; (4) to u t i l i z e the developed technique for isolat ion of immunoglobulins and transferr in from blood and (5) to apply the same technique for isolat ion of ovotransferrin from egg white. The biological ac t iv i ty and the p o s s i b i l i t y of using these proteins in f o r t i f i c a t i o n of infant formula were investigated. 4 LITERATURE REVIEW A. HISTORY Infant formula is required for infants whose mothers do not breast feed for socioeconomic, physical or psychological reasons. Before the industrial revolution, the only alternatives to mothers breast-feeding were starvation of the infant or the services of a wet-nurse (direct ly supplied donor breast mi lk) . However by the end of the 18th century, a conical baby bottle was invented. Thus modern technological developments for the f i r s t time made possible a widespread shi f t away from breast-feeding. The modern infant formula industry found i ts origins in the mid-19th century when Nestle in Switzerland and Borden in the U.S. began producing sweetened condensed milk. In the 1880's, canned evaporated milk made i ts appearance (Mi l ler , 1983). Until the 20th century, however, there was l i t t l e understanding of the nutr i t ional needs of infants or of sanitary requirements. But medical sc ient is ts recognized the need for safe and effect ive alternatives to breast-feeding in order to deal with problems of childhood disease and malnutrition resulting from the fa i lure of many mothers to breast-feed adequately. Most present-day infant formulas in the United States market are adaptations of the product designed by H.J . Gerstenberger and co-workers (1915). On a f l u i d basis this early formula consisted of 4.6% fat (a homogenized mixture of both plant and animal f a t ) , 6.5% carbohydrate and 0.9% protein and was given the name Synthetic Milk Adapted (SMA). Over the years the composition of infant formulas has been altered and adjusted, mostly in response to s c i e n t i f i c evidence of need (Packard, 1982). 5 B. HUMAN MILK VS. COW'S MILK Milk, whether from the human or other mammals, is an exceptionally complex mixture of more than 200 fat -soluble and water-soluble components. The milk components may originate from direct transfer from the blood, from biosynthesis from blood precursors or from a combination of both. Consequently, the composition of milk is affected by biochemical, physiological and hormonal factors which influence the composition of blood, factors which influence the rate of transfer of nutrient from blood to milk and factors which influence the rate of biosynthesis of compounds in the mammary gland (Blanc, 1981). Human and cow's milk consist mainly of water, fa t , carbohydrate, protein and minerals; however, this is where the s imi la r i t i es end. Table 1 gives the gross composition of human and cow's milk (Gurr, 1981). This is not very surprising i f one looks at mammalian animals in general. Milk from a l l mammalian species consists of the constituents l is ted above but the relative proportions of each vary as much as the species (Jenness, 1982). The proteins, f a ts , e tc . also vary in their biochemical and physical properties as well as their detai led composition. The fat concentration in human and bovine milk is not comparable i . e . 4.4 and 3.7% respectively (Packard, 1982). It should be noted that fat is the most variable constituent of milk both in absolute quantity and in composition. Human milk has a lower level of short chain fatty a d d s along with a higher concentration of polyunsaturated fatty acids than cow's milk. Another notable difference is that human milk has approximately a 50:50 ratio of saturated to unsaturated fatty acids while in bovine milk that ratio is 65:35. Unsaturated vegetable o i l s may be used to adjust the ratio in cow's milk. 6 Table 1. Gross Composition of Human and Cow's Milk (Grams per 100 Grams of Fluid Product) a . Component Human milk Cow's mi 1k Fat 4.4 3.7 Protein 1.0 3.3 Milk sugar (lactose) 6.9 4.7 Mineral matter (as ash) 0.2 0.7 Water 87.5 87.6 Energy, KJ/100ml_D 290 273 a Packard, 1982.. b Gurr, 1981. 7 The principal carbohydrate in milk is lactose, a disaccharide speci f ic to this secret ion. It contributes about 40 and 29% of the total energy of human and cow's milk, respectively. A carbohydrate is generally added to infant formula to increase the concentration, so that i t is in the range of human milk. There are some problems in deciding which carbohydrate to add, since the gut of some babies would be unable to handle this amount of lactose, resulting in lactose malabsorption and fermentative diarrhea. Other sources of carbohydrates should be added, i . e . , maltodextrin. The major differences in the mineral content between human milk and cow's milk is not so much in kind as amount. Human milk contains, on the average, about 0.2% ash, the mineral matter remaining after incinerat ion. Cow's milk averages about 0.70%. The higher percentage of ash, or rather minerals, is due primarily to much higher concentrations of calcium and phosphorus in cow's milk. High mineral content should be reduced, i f cow's milk is to be considered for use in infant feeding (Wharton, 1981). Cow's milk contains well over three times as much protein as human milk i . e . 3.3 vs. 1.00, respectively. Table 2 gives a breakdown of the different protein fractions in cow's milk and human milk (Gurr, 1981). Cows' milk contains over eight times as much casein as human milk. Casein is a complex protein and can be broken down into three major f ract ions: a s - , B- and K-case ins . Cow's milk consists of approximately 45% a s -case in while human milk consists of very l i t t l e of this component. Human milk on the other hand, has mainly p-casein. The coagulation properties of a s - and p-casein are quite d i f ferent . While ac id i f i ed p-casein forms a soft curd, a s -case in is thought to be more digest ible for the infant (Packard, 1982). Other protein fractions that d i f f e r between human and cow's milk are a-lactalbumin, p- lactoglobul in , lactoferr in and the immunoglobulins f ract ion. Table 2. Protein Composition of Human and Cow's Mi lk a 8 Human Milk Cow's milk Protein g/lOOmL Tota l , % g/lOOmL Total,% Total 0.88 100 3.30 100 Caseins 0.31 35 2.6 79 Total whey: 0.57 65 0.7 21 a-lactalbumin 0.15 17 0.12 3.5 B-lactoglobulin t r — 0.30 9.0 lactoferr in 0.15 17 t r — serum albumin 0.05 6 0.03 1.0 lysozyme 0.05 6 t r — immunoglobul ins 0.10 11 0.10 3.0 others 0.07 8 0.15 4.5 a Gurr, 1981. tr=trace. 9 B-lactoglobulin (B-Lg) makes up a high percentage of the protein in cow's milk while human milk contains very l i t t l e . The concentration of B-Lg in human milk is so low that for years i t was thought that B-Lg was exclusive to cow's milk (Packard, 1982; Brignon et a l . , 1985). Lactoferrin (LF) is found in human milk but only trace amounts are found in cow's milk. The immunoglobulin f rac-tion is re la t ive ly similar in concentration in the two milk sources but human milk contains mostly IgA while in cow's milk IgG is the predominant immunoglobul i n . From the previous discussion on compositional di f ferences, i t can be seen that i t is important to look beyond the gross make-up and investigate human and cow's milk on a more def in i t ive basis . The differences in composition between the two milk sources cause currently available infant formulae to be less than optimal substitutes for human milk. C. HUMANIZING INFANT FORMULA Due to many compositional differences between human and cow's milk, many attempts have been made to humanize infant formula. There have been three major ways in which cow's milk has been modified to bring i ts composition closer to human milk: added carbohydrate, substitute fat and using a combination of whey and skimmilk (Gurr, 1981; Wharton, 1981). The simplest change involves the addition of carbohydrates. The purpose for adding carbohydrates is to increase the overall carbohydrate content and to di lute the concentration of protein, fat and mineral per unit of energy intake. Maltodextrin or sucrose are examples of two carbohydrates that have commonly been used in the infant formula industry (Wharton, 1981). In some cases the fat from cow's milk has been removed and substituted with a mixture of plant and animal fa ts . The purpose for the combination of fats from plant and animal sources is to obtain a fatty acid composition similar to 10 that found in human milk. Studies have indicated that absorption is much more e f f ic ien t in the substituted fat formulas and becomes close to that found in breast milk (Wharton, 1981). Formulas based on demineralized whey have been used increasingly in recent years. Demineralized whey, containing whey protein and low concentrations of minerals, is used as the base and to this is added a small amount of skimmilk. The range of casein/whey protein ratios of human milk and cow's milk are 0.4 -0.7 and 3.0 - 4 .7, respectively. To humanize this ra t io , some of the commercial infant formulae are f o r t i f i e d with whey proteins. For instance, mixing skimmilk with 4 volumes of whey changes the ratio to 0.75. However simple mixing of bovine whey with skimmilk does not minimize the compositional differences bet-ween the two mi lks, i . e . higher contents of a-lactalbumin, l ac to fe r r in , lysozyme and immunogobulins (Ig) in human milk than in cow's milk (Hambraeus, 1977). While B-lactoglobulin is the dominant whey protein in cow's milk (approxi-mately 60% of the total whey proteins) , i t is completely lacking or very low in human milk (Liberatori and Napolitano, 1980; Brignon et a l . , 1985). D. ALLERGENICITY OF WHEY PROTEINS In adapted formulas, the mass balance of casein versus whey proteins of bovine or igin can be corrected to a 40:60 ratio in favor of the whey proteins (Anderson et a l . , 1982; Theuer, 1983). This adjustment adapts the cow's milk formula closer to. the protein composition of human milk. It also increases the content of some essential amino acids (threonine, tryptophan, and lysine) as well as cystine in comparison with normal cow's milk; thus, adaptation by addition of whey protein corresponds to nutrit ional improvement (Forsum, 1974). Presumably any of the individual proteins in cow's milk may induce speci f ic antibody and provoke al lergy in a susceptible c h i l d , but some proteins seem more antigenic than others (Wharton, 1981). The newborn is par t icu lar ly vulnerable. 11 For a few days the stomach is porous, even to whole proteins. The milk of any non-human species is more l ike ly to induce an al lergy in sensit ive infants than mother's milk (Sav i laht i , 1981). A protein, or an antigenic fragment of protein wi l l simply be absorbed whole. The body reacts to the offending agent by simply producing IgE. The IgE is specia l ly designed to recognize and bind to this spec i f ic antigen. There are, however, two ends to an antibody: the Fc, or the base of the Y-shaped antibody, and the Fab, the two protruding fingers of the Y. The fingers are designed to bind to the antigen. The base, the F c segment, binds in this instance to mast ce l ls or basophils (Packard, 1982). When an infant has been exposed to an al lergen, IgE is produced. It attaches i t s e l f to mast ce l l s of body tissue and at the same time issues a chemical command to the mast c e l l . In essence, the command ca l l s for release of histamines which in turn cause the various disorders that accompany an a l le rg ic reaction. Animal experiments suggest that B-lactoglobulin is more antigenic than either casein or the small amount of a-lactalbumin found in cow's milk (Ratner et a l . , 1958; Goldman et a l . , 1963 a,b; Moneret-Vautrin and G r i l l i a t , 1979). Compiled data from f ive studies of cow's milk protein intolerance in infancy showed sens i t iv i ty to p-lactoglobulin in 82%, casein in 43%, a-lactalbumin in 41%, bovine serum globulin in 27%, bovine serum albumin in 18% of the patients (Lebenthal, 1975). In view of the above resul ts , the well-known al lergencity of p- lactoglobulin has to be considered when manufacturing infant formula. Different approaches have been considered to decrease P-Lg al lergencity: heat denaturation, enzymatic hydrolysis or selective elimination of P -Lg . Heat denaturation of the protein was suggested by Ratner et a l . (1958), Anderson et a l . (1979), McLaughlan et a l . (1981), and more recently by Kilshaw et a l . (1982) 12 and Heppell et a l . (1984). However, heat treatment of whey proteins may destroy the biological ac t iv i ty of bioactive proteins present in the whey i .e . immunoglobulins, lactoferr in and lactoperoxidase. Tryptic and chymotryptic hydrolysis of whey proteins suggested by Pahud et a l . (1985) and Asselin et a l . (1986) are not recommended for the same reasons. Selective elimination of the a l l e rg ic compound may therefore be the method of choice to humanize the protein composition of cow's milk for infant feeding. Several attempts have been made for separating p-Lg from bovine whey. Sephadex G-75 gel f i l t r a t i o n was suggested as a method for whey protein fractionation to humanize infant formula (Forsum, 1974; Mathur and Shahani, 1979). A lso , p-Lg was preferent ia l ly precipitated at pH 4.65 from cheese whey after concentration by u l t r a f i l t r a t i o n and demineralization by e lectrodia lys is (Amundson and Watanawamchakorn, 1982; Slack et a l . , 1985). Pearce (1983) used heat treatment and pH treatment for separating p-Lg and an a-La rich fract ion from bovine Cheddar cheese whey. Recently, a f e r r i c chloride method was established for se lect ive ly precipi tat ing P-Lg from bovine whey (Kaneko et a l . , 1985; Kuwata et a l . , 1985). However, the use of f e r r i c chloride may saturate lactoferr in and abolish i ts antimicrobial a c t i v i t y . Antigenic react iv i t ies of chemically modified p-Lg was studied by Otani et a l . (1985). These researchers found that modification of arginine residues, tryptophan residues, or sulfhydryl groups had l i t t l e effect on the antigenic reac t iv i ty . However, a s igni f icant decrease in the react iv i ty was noted when P-Lg was acetylated, succinylated or modified with diethyl pyrocarbonate or coupled with glycine amide. These results suggest that there is a poss ib i l i ty that the amino group, hist id ine residue and carboxyl group may play an important role in the antigencity of bovine p-Lg. 13 E . A N T I M I C R O B I A L SYSTEM I N HUMAN M I L K AND M I L K SUBSTITUTES Several substances in human milk provide resistance to infant diseases, par t icu lar ly diseases of the intestinal tract (Cunningham, 1977; Larsen, 1978; Cunningham, 1979; Fa l lot et a l . , 1980; France et a l . , 1980; Pullan et a l . , 1980). Cunningham (1979) noted that f i r s t year mortality increased with the extent of formula feeding and was two-fold higher in a r t i f i c i a l l y fed infants; during the f i r s t four months the difference was 16-fold. S imi la r ly , breast feeding reduced the incidence of infection by Salmonella (France et a l . , 1980), respiratory syncytial virus (Pullan et a l . , 1980), and the incidence of hospital admissions for infection in infants (Fal lot et a l . , 1980). The factors in human milk thought to be responsible for the breast fed infant 's increased resistance to diseases have been reviewed extensively (Gothefors and Winberg, 1975; Bezkorovainy, 1977; Reddy et a l . , 1977; McClelland et a l . , 1978; Reiter , 1978; P i t tard , 1979; Welsch and May, 1979; Blanc, 1981; Gur.r, 1981; Hanson and Soderstrom, 1981; Packard, 1982). Some of the antimicro-bial factors are associated with ce l lu la r components present in human milk and thus are not present in human milk substitutes. Other antimicrobial factors are present in cow's milk and other substitutes in low or trace amounts. Only lactoperoxidase is present in larger quantities in cow's milk compared to human milk. The function of the antimicrobial factors wi l l be discussed. 1. Immunoglobulins Immunoglobulins or antibodies are a class of proteins that are comprised of four polypeptide chains. Each immunoglobulin unit is formed from two identical heavy chains and two identical l ight chains. The peptides are linked by d isu l f ide bridges (Butler, 1983). In humans, there are f ive classes of antibodies and they are named for their heavy chains. They consist of 14 immunoglobulin G (IgG), immunoglobulin M (IgM), immunoglobulin A (IgA), immunoglobulin D (IgD)and immunoglobulin E (IgE). The heavy chains for each class are designated by the appropriate Greek le t ter : y, y , a , S and e, respectively (Atassi et a l . , 1984). The l ight chains are either kappa (K) or lambda (A). Light chains have one constant region and one variable region. Heavy chains consist of three constant regions and one variable region. There are also hypervariable regions within the variable regions (Brock, 1979; Nisonoff, 1982). The simplest immunoglobulin is IgG, as i t is comprised of only one basic unit of four polypeptide chains and has a low molecular weight of 150,000 daltons. When the molecule is treated with a reducing agent, i t dissociates into two heavy and two l ight chains. The molecule may also be s p l i t by proteolyt ic enzymes such as papain, to give two types of fragments: F ab (antigen binding) and Fc (c rys ta l l ine ) . Proteolysis by pepsin at pH 4 yields an F(ab')2 (Fab-dimer) and various peptides derived from fragment Fc (Whitney et a l . , 1976; But ler , 1983). In bovine colostrum, there are three major immunoglobulins ( i . e . , IgG, IgM and IgA) with IgG comprising about 80% in both milk and colostrum (Butler, 1983). From early investigations i t is already known that di f ferent subclasses of IgG exist in bovine colostral IgG (Murphy et a l . , 1964). IgG2 and IgGi have been ident i f ied by some groups of workers (Butler, 1969; Duncan et a l . , 1972). Concentration of Ig in colostrum is high at parturi t ion and decreases with each successive postpartum milking (Butler, 1983). Table 3 summarizes the concentration of various species of bovine immunoglobulins in serum and secretions. However, these values may provide a rough guide for investigators, since there are dif ferent factors affecting the immunoglobulins concentration in 15 Table 3. Concentration of bovine immunoglobulins in serum and secretions (mg/mL)a. Body Fluid Samples IgGi IgG 2 IgA igM Serum 11.2 9.2 0.37 3.05 Colostrum (whey) 46.4 2.87 5.36 6.77 Milk 0.58 0.055 0.081 0.086 Nasal secretion 1.56 — 2.81 0.04 Sal iva 0.034 0.016 0.34 0.006 Tears 0.32 0.01 2.72 .176 Urine 0.009 t r 0.0013 tr Bi le 0.10 0.09 0.08 0.05 Vaginal secretion 0.23 0.13 0.90 — a But ler , 1983. 16 milk i . e . age, breed di f ferences, differences in the techniques used for measurement and immunization. The amino acid and carbohydrate composition of bovine IgGi and IgG2 have been studied by several investigators (Groves and Gordon, 1967; Lisowski et a l . , 1975). The carbohydrate content, sulfhydryl content and other physicochemical character ist ics of bovine Igs are shown in Table 4 (Butler, 1983). The human fetus and infant receives antibodies in utero, where IgG is absorbed through the placenta, and from human milk. Packard (1982) noted that the milk antibodies function as speci f ic host resistance factors by aggregating bacteria in the intestine, to f a c i l i t a t e their removal, by interfer ing with bacterial colonization of the intest inal l i n i n g , by assist ing other host resistance factors , f ix ing complement, by neutral izing toxins and by k i l l i n g viruses. Thus they provide crucial immunological protection unti l the newborn infants' defence systems can be established (Ogra and Ogra, 1978; Hanson and Soderstrom, 1981). Although classes of immunoglobulins ( i . e . , IgA, IgG and IgM) are present in human milk throughout the period of lactat ion, the highest concentrations are in colostrum (Reddy et a l . , 1977; Hambraeus et a l . , 1978; McClelland et a l . , 1978; Ogra and Ogra, 1978; Ones, 1979; Goldman et a l . , 1982). Small amounts of colostrum IgG are absorbed from the intestine during the f i r s t 18-24 hours after birth (Ogra et a l . , 1977), although the signif icance of this is not known. Intact secretory IgG, the major milk antibody, has been found in the intestine of breast-fed infants (Hambraeus et a l . , 1978), where i t appears to function in the passive transfer of spec i f ic immunity to mucosal surfaces to prevent penetration by microorganisms, viruses and antigens (Walker and Isselbacher, 1977; Hanson and Soderstrom, 1981). Antibodies against enteropathogenic Escherichia col i and col i enterotoxin have been postulated to play a role in Table 4. Biochemical character ist ics of bovine immunoglobulins 3. 17 Character ist ics IgGi IgG 2 IgA IgM igE SC Heavy chain 1 2 — S20,w 6.9 6.9 10.9 19.5 — 4.1 EiP 13.5 12.3 — 11.8 — — Carbohydrate (%) Total 2.8 2.6 8 11 — 5.9 Sulfhydryl groups Half-cysteine/100 residues 3.1 Total SH (mole/mole) — Free SH (mole/mole) S-S linkages/mole 2.6 — 43.6 0.9 21 — — Molecular weight 162K 152K 408K 1.030K — 74K Heavy chain (mol. wt.) 57K 54K 62K 76K — — Light chain (mol. wt.) 25K 23K 23K 22.5K — — a But ler , 1983. S20,W= Svedberg values EH8= Absorbance of 1% protein solution at 278 nm. 18 preventing infant diarrhea (Gindrat et a l . , 1972; S to l ia r et a l . , 1976; Rogers and Synge, 1978). IgM also provides protection against gram negative pathogens (Packard, 1982). The importance of oral administration of immunoglobulins in both animals and human infants is well documented. In animals, supplementing a milk replacer with Ig separated from porcine blood maintained 86% survival of piglets compared to no survival of the control group fed with the milk replacer alone ( E l l i o t , 1978). Similar survival rates are obtained by feeding piglets with bovine colostrum (McCallum et a l . , 1977). Improved survival and immunity of chicks by feeding whey containing Ig has also been reported from USSR (Kuznetsov and Rebrova, 1983). In humans, the importance of Ig in infant feeding has been well demonstrated in c l i n i c a l test results from India (Narayanan et a l . , 1983). Of s ix ty -s ix low birth weight infants s p l i t into two equal groups only 7 infants in the group fed with human colostrum developed infection compared to 18 in the infant formula-alone group, which is s ign i f icant ly di f ferent (P<0.01). Signi f icant results were also shown by Hilpert et a l . (1974/1975) where infants fed colostrum from sensit ized cows had an infection rate of 24% compared to 67.4% for the control group who were not fed colostrum. Rotavirus infection was also prevented by ora l ly administering cow's colostrum (Ebina et a l . , 1985). Diarrhea developed in only 1 of 6 infants given Rota colostrum, while 6 out of 7 infants given milk developed diarrhea. 2. Lactoferr in Lactoferr in , an iron-binding protein of breast milk, is considered to be of great importance for the breast-fed infant. It has been shown (Kirkpatrick et a l . , 1971; Bullen et a l . , 1972, Reiter, 1983) that lactoferr in can bind iron In 19 v i t ro and in v ivo, thereby preventing the growth of iron-requiring microorganisms. Lactoferrin is a single-chain glycoprotein with an approximate molecular weight of 75,000-85,000 (Blackberg•and Hernel l , 1980). The structure of lactoferr in consists of two largely independent domains, each carrying i ts own iron-binding s i t e . This proposed structure has received strong support from studies demonstrating that cleavage of l ac to fe r r in , usually by proteases, could under certain conditions y ie ld half-molecules capable of binding just a single iron atom (Brock, 1985). Human milk contains from 3 to 100 times as much iron-binding protein lactoferr in as cow's milk (Packard, 1982), and a trace amount of the serum iron-binding protein t ransferr in . Lactoferrin is active in v i t ro against enteropathogenic E. col i (Rogers and Snyge, 1978; Spik et a l . , 1978; Dolby and Honour, 1979; Samson et a l . , 1979; Samson et a l . , 1980; ), Vibrio cholerae (Arnold et a l . , 1977), Streptococcus mutans (Arnold et a l . , 1977), and Candida  albicans (Kirkpatrick et a l . , 1971), presumably by chelating iron and making i t unavailable for microbial growth (Packard, 1982). Lactoferr in , together with secretory IgA from human milk, have a considerable bacter iostat ic effect against human enteropathogenic strains of E. col i (Stephens et a l . , 1980; Dolby and Stephens, 1983). S imi lar ly , bovine colostral IgG, together with lac to fe r r in , is found to be active against a strain of E i co l i pathogenic to calves (Stephens et a l . , 1980). Since the ant i -E. col i ac t iv i ty of lactoferr in is not destroyed by proteolyt ic digest ion, lactoferr in may also play an antibacterial role in vivo (Samson et a l . , 1980). 3. Lactoperoxidase and lysozyme Lactoperoxidase (LP), which catalyzes the oxidation of the thiocyanate by 20 hydrogen peroxide to hypothiocyanate, serves as a major antimicrobial agent in cow's milk (Reiter et a l . , 1976; Reiter, 1978; Pruit t and Tenovuo, 1985). Lactoperoxidase is active against c o l i , Pseudomonas f luorescens. Salmonella  typhimurium and strains of K lebsie l la aeroqenes (Pruitt and Tenovuo, 1985). Bovine LP behaves d i f ferent ly from other protective proteins. Its concentration is low in bovine colostrum and increases rapidly to reach a peak at 4-5 days post-partum. Human peroxidase is highest in colostrum and declines rapidly within 1 week (Gothefors and Marklund, 1975; Reiter , 1985b). Bjorck (1978) reported a concentration l imit of 0.5 ug/mL human milk which is far below the average of 10-30 ug/mL present in bovine milk. It is important to note that lactoperoxidase is not inactivated by the gastr ic juice from an infant (pH 5) (Gothefors and Marklund, 1975), whereas pepsin at pH 2.5 inactivates lactoperoxidase (Paul and Ohlsson, 1985). Lysozyme cleaves the cel l wall peptoglycan of a number of gram posit ive and gram negative microorganisms, and appears to potentiate the ac t iv i ty of IgA against E. co l i (Adinolfi et a l . , 1966; H i l l and Porter, 1974), and with peroxide and with ascorbate to lyse g± col i and Salmonella (Mi l ler , 1969). Human milk contains approximately 3000 times as much lysozyme as cow's milk (Chandan et a l . , 1964; Chandan et a l . , 1968). Human milk lysozyme possesses a molecular weight, amino acid composition, spec i f ic a c t i v i t y , thermal s t a b i l i t y and ant igenici ty which are quite di f ferent from the lysozyme in cow's milk (Eitenmil ler et a l . , 1974; Eitenmil ler et a l . , 1976). 4. Bif idobacter ia The bi f idobacter ia (previously designated Lactobaci11 us bif idus) protect the infant against disease by producing vo la t i l e acids which inhibi t the 21 pro l i fera t ion of pathogenic microorganisms in the gut (Friend et a l . , 1983). Within 3-4 days after b i r t h , the intestinal tract of breast fed infants contains up to 99% Bifidobacterium bifidum type IV (Gyllenberg and Raine, 1957; Haenel, 1970). The f lora of formula fed infants does, not contain type IV, but rather 30-40% B_j_ bif idium Type II (Haenel, 1970). A number of substances or factors in human milk have been reported to stimulate the prol i ferat ion of the b i f idobacter ia . These factors include: buffer capacity, lactulose, lac to fer r in , pantothenic ac ids , oligosaccharides and glycoproteins. Infants fed either human milk or a test formula with low buffering capacity have a re la t ive ly low fecal pH (5.1-5.4) and a higher proportion of bi f idobacter ia than fecal coliforms or streptococci (Wi l l is et a l . , 1973; Bullen et a l . , 1977). Those infants who were fed highly buffered cow's milk formula had a s ign i f icant ly higher fecal pH (5.9-8.0) and a mixed fecal f lo ra (Wil l is et a l . , 1973; Bullen et a l . , 1977). It has been suggested (Bullen et a l . , 1976) that E_j. co l i and S .^ faecium i n i t i a l l y colonize the gut and produce ac id . In breast fed infants the pH of the intestine drops and growth conditions become favorable for bi f idobacter ia and unfavorable for other organisms. The crucial drop in pH and thus the prol i ferat ion of the bi f idobacter ia is prevented when infants are fed highly buffered formulas (Friend et a l . , 1983). Even though lactulose is not present in human milk or cow's milk, i t has been reported that supplementation of prepared formulas with lactulose (formed during heating of s te r i l i zed formula) increases the proportion of intestinal bi f idobacter ia (Mendez and Olano, 1979; Shvedova, 1981). Lactoferrin may indirect ly promote the growth of the bif idobacteria by inhibi t ing the growth of competing E± co l i (Spik et a l . , 1978). Pantothenic acid derivatives have also been shown to stimulate one strain of B^ infantis (Tamura et a l . , 1972). A nitrogen-containing oligosaccharide (Bezkorovainy and 22 Topouzian, 1981) and glycoproteins (Hirano et a l . , 1968; Bezkorovainy and Nichols, 1976) have been shown to stimulate the growth of bif idobacteria (Gyorgy, 1953). The greatest stimulatory ac t iv i ty is found in human colostrum, followed by human milk, cow's colostrum and cow's milk. A recent report (Ashoor and Monte, 1983) noted that human milk contains two d is t inc t B^ bifidum var.  Penn. bi f idus stimulating factors , the ac t iv i ty of which varies from sample to sample. 5. Ovotransferrin Ovotransferrin (OVT), also cal led conalbumin, is a glycoprotein with a molecular weight about 76,000 and contains no free sulfhydryl groups or phosphorus. The protein moieties of ovotransferrin of egg white and transferrin of chicken blood serum are iden t ica l , but the carbohydrate prosthetic groups are di f ferent (Powrie and Nakai, 1986). Schade and Caroline (1944, 1946) f i r s t reported that ovotransferrin and serum transferr in inhibited the growth of E. co l i and other bacterial species. This antibacterial effect was destroyed by the addition of Fe + 3which saturated the iron binding si tes of the proteins (Brock, 1985). Valenti et a l . (1983) concluded that the antimicrobial ac t iv i ty of hen's ovotransferrin was quantitatively and qual i ta t ive ly similar to that of human lac to fe r r in . These proteins demonstrated a similar protective effect on experimentally-induced bacterial infections in newborn guinea pigs. These observations have led to the concept of "nutrit ional immunity" (Weinberg, 1977). A greater resistance against enterobacterial infection of human infants fed with breast milk than those fed with a r t i f i c i a l formula is well documented (Packard, 1982). This has been attr ibuted, to a great extent, to the presence of a large quantity of lactoferr in in human milk compared to cow's milk. The 23 s imi la r i t i es in structure and biological ac t iv i ty between ovotransferrin and lactoferr in jus t i f y the antimicrobial effect of ovotransferrin added to infant formula. In addit ion, ovotransferrin did not sensit ize ovotransferrin fed infants (Giacco-Del et a l . , 1985). F. ISOLATION OF BIOACTIVE PROTEINS 1. General methods With a view to the isolat ion of immunoglobulins as well as of a l l other biopolymers, one must make use of those physicochemical properties or parameters that are peculiar to the polymer in question, and that are quantitatively di f ferent from the physicochemical properties of the other accompanying (but unwanted) polymers (Van Oss, 1982-83). There are f ive fundamentally different physicochemical parameters of biopolymers: s o l u b i l i t y , e lec t r i c change, surface tension, size and shape, and ligand s p e c i f i c i t y . Some of these parameters wi l l be reviewed in isolat ion strategies of immunoglobulins, lactoferr in and ovotransferrin from dif ferent biological sources. Methods currently available for the isolat ion of bovine immunoglobulins and their subclasses based on s o l u b i l i t y , e lec t r ic charge and size and shape are batch processes, which are d i f f i c u l t to mechanize (Butler and Maxwell, 1972; Fey et a l . , 1976; Kanamaru et a l . , 1977; Butler et a l . , 1980; Kanamaru et a l . , 1980; Kanamaru et a l . , 1981; Kanamaru et a l . , 1982a; Kanamaru et a l . , 1982b; Shimazaki and Sukegawa, 1982; Butler , 1983; Brooks and Stevens, 1985; Bokhout et a l . , 1986). Other methods based on a f f i n i t y chromatography using protein A-Sepharose (Ey et a l . , 1978; Martin, 1982) or immuno-adsorbents (Bokhout et a l . , 1986) are quite expensive for large scale pur i f i ca t ion . Lactoferrin was f i r s t isolated from human milk by Groves (1962) and from cow's milk by Gordon et a l . (1962) by using ammonium sulfate precipitat ion 24 and/or ion exchange column. Several methods for i ts isolat ion have been described but most of them are rather laborious (Johansson, 1969; Querinjean et a l . , 1971; Law, and Reiter , 1977; Kawakata, 1984). However, Lbnnerdal et a l . (1977) used metal chelate a f f in i t y chromatography and Blackberg and Hernell (1980) used heparin-Sepharose for lactoferr in i so la t ion . The most frequently used isolat ion methods for ovotransferrin are based on sal t precipi tat ion and ion exchange (Warner and Weber, 1951; Williams, 1962; Azari and Baugh, 1967; Antonini , 1977). These methods, however, are labor-intensive and d i f f i c u l t to mechanize. 2. Metal Chelate-Interaction Chromatography (MCIC) A l i t t l e more than a decade ago, a novel pur i f icat ion technique for proteins using "immobilized metal a f f in i t y chromatography" was introduced by Porath et a l . (1975). This technique was later cal led metal chelate-interaction chromatography (MCIC) by Rassi and Horvath (1986). Since i ts introduction, the technique has gained wide acceptance and was recently reviewed (Lonnerdal and Keen, 1982; Sulkowski, 1985). The application of MCIC has been reported for separation of human serum proteins (Porath et a l . , 1983; Andersson, 1984; Ramadan and Porath, 1985), human lactoferr in (Lonnerdal et a l . , 1977), lysozyme (Torres et a l . , 1979), human f ibroblast interferron (Edy et a l . , 1977) and human serum albumin (Hansson and Kagedal, 1981). As explained by Lonnerdal and Keen (1982), the binding of proteins is believed to be the result of the a b i l i t y of e lectron-r ich l igands, such as h is t id ine , cysteine and tryptophan, to substitute weakly bonded l igand, such as water or buffer ions, in the complexes. When a protein, with surface exposed amino acids having electron-donating capacity, is exposed to a metal, a strong multipoint attachment can resul t . This binding is stable even in 1M NaCl ruling 25 out the p o s s i b i l i t y of ionic interaction being the principal force in the binding. It is important to real ize that this 'k ind of interaction is independent of whether the protein is iron binding or not or , in the case of an iron-binding protein, whether the protein is in Fe-saturated form or in the apo-form. Fe-saturated lactoferr in can bind to a copper-loaded gel as strongly as the apo-form of lactoferr in (Lonnerdal et a l . , 1977; Lonnerdal and Keen, 1982). 26 MATERIALS AND METHODS A. MATERIALS Sodium hexametaphosphate (SHMP, pur i f ied grade) was purchased from Fisher S c i e n t i f i c Company (Fairlawn, NJ); 8- lactoglobul in , a-lactalbumin, bovine immunoglobulins, rabbit anti-bovine IgG, bovine serum albumin, lac to fer r in , a lkal ine phosphotase conjugated rabbit anti-bovine IgG, lactoperoxidase, ovalbumin, ovotransferr in, a-casein and diethyl pyrocarbonate were purchased from Sigma Chemical Company (St. Louis, MO). B-Casein was obtained from Chemalog (South P l a i n f i e l d , NJ). K-Casein was prepared according to the method of Z i t t l e and Custer (1963). a s i -Case in was a g i f t from Dr. R. Yada and sulfhydryl blocked K-casein (sss-K-casein) was a g i f t from Dr. S. Nakai. Materials for column chromatography were: s i l i c a (sand) (fine granular type No. S-150 from Fisher Laboratory Chemical, Fairlawn, NJ), controlled pore glass 80-120 mesh, PG 1400-120 and fumed s i l i c a S-5055.(from Sigma Chemical Company, St . Louis, MO); alumina (neutral AG7, 100-200 mesh, No. 132-1140 from BioRad Laboratories, Mississauga, ON). Electrodialyzed and sweet whey powders were from Mead Johnson and Company (Evansvi l le , IN). Bovine blood plasma was obtained from Intercontinental Packers L t d . , Vancouver, BC. Skimmilk was purchased from a local market. Cheddar cheese whey was obtained from Dairyland Foods (Burnaby, BC). Escherichia co l i Serotype 0142:K86(B):H6 (ATCC No. 23985), Salmonella typhimurium (ATCC No. 13311) and Bordetella parapertussis (ATCC No. 15311) were supplied by American Type Culture Col lect ion (Rockvi l le , MD). Al l other chemicals were of analyt ical reagent grade. B. ACID WHEY PREPARATION Raw milk and colostrum were obtained from the University Animal Science Farm. Acid whey was prepared from raw milk and colostrum. The milk was centrifuged at 4,000 x g for 30 min at 5°C for cream separation. Acid whey was 27 prepared from the skimmilk by adding 50% acetic acid solution to pH 4.6 at 25°C and centrifuging at 10,000 x g for 15 min to remove casein precipi tates. C. SODIUM DODECYL SULFATE - POLYACRYLAMIDE GEL ELECTROPHORESIS The method of Laemmli (1970) was used after modif ications. Polyacrylamide gel electrophoresis in the presence of 0.2% sodium dodecyl sulfate (SDS-PAGE) was performed with a slab type vert ical gel system using the Atto SJ 1060 DSH Electrophoresis unit (Atto C o . , Tokyo, Japan). 1. Discontinous SDS-PAGE A whole gel was composed of separating gel (lower gel) 0.2 cm thick, 11 cm long, and 13.5 cm wide, and stacking gel (upper ge l ) . Ten and 3% polyacrylamide gels were used as the separating and stacking ge ls , respectively, of which the ratio of acrylamide to N.N'-methylene-bis-acrylamide was 25. Polymerization of both gels was catalyzed by 0.02% ammonium persulfate. One mL of whey solution (2-4 mg protein/mL) was treated with 5% SDS and 0.2 mM 2-mercaptoethanol in boi l ing water for 1.5 min, followed by the addition of 200 mg sucrose and 50 jiL of 0.05% bromphenol blue tracking dye solut ion. Twenty f ive yL of the treated whey solution was applied to the sample slot after the sample slots and upper electrode chamber were f i l l e d with Tr is -g lyc ine electrode buffer (3g Tr is + 14.4g glycine + lg SDS in 1 L, pH 8.3). Electrophoresis was performed at room temperature with a constant voltage of 90 volts unti l the tracking dye marker migrated to 1 cm from the gel bottom, in approximately 4.5 hr. The gel was then removed, placed on a net p last ic f loater (a gel supporter), immersed in 0.25% Coomassie B r i l l i a n t Blue R-250 dye solution (Weber and Osborn, 1969), and stained for 1.5 hr. The gel was rinsed with water, transferred to a di f fusion destainer (model 172A, Bio Rad 28 Laboratories, Richmond, CA), and destained ver t i ca l l y for 18 to 20 hr with a c i rculat ion of destaining solution (a mixture of 10% acetic acid and 7.5% methanol) through a cartridge of activated carbon. Glycoproteins were stained by using periodic ac id-Schi f f (PAS) technique described by Zacharius et a l . (1969). 2. Gradient SDS-PAGE Solutions containing 3% and 20% acrylamide were prepared from a stock solution of 60% acrylamide with 4% crossl ink ing. The solutions were made in 0. 25 M Tris-HCl (pH 8.3) containing 0.2% SDS and 0.125% tetramethyl ethylenediamine (TEMED). Ammonium persulfate for polymerization was added to the solutions immediately before mixing. Gradients were generated using a two chamber device containing 20 ml of 20% acrylamide in the mixing chamber and 20 ml of 3% acrylamide in the reservoir chambers. The mixture was then pumped into a vert ical slab mould of the Atto SJ 1060 SDH Electrophoresis Unit (Atto Co . , Tokyo, Japan), at a flow rate of 2 mL/min. Sample preparation, electrophoresis condit ions, staining and destaining were performed as described in Section ( C - l ) . D. IMMUNOCHEMICAL ANALYSIS 1. Immunoelectrophoresis and immunodiffusion Immunoelectrophoresis and immunodiffusion analysis were carried out according to the method of Williams and Chase (1971) with modif ications. Nine mL of 1% agarose in 0.05 M Na-barbital acetate buffer, pH 8.3 was gelatinized over Gelbond f i lm (0.02 x 7.5 x 10 cm, FMC Corporation Marine Col lo id Division Bioproducts, Rockland, MA). Three yL of whey sample was applied to a punched sample well with a diameter of 2 mm and immunoelectrophoresis was performed at room temperature for 45 min with a constant voltage of 60 vo l ts . Sixty uL of 29 antibody (Miles Laboratories Inc.) was added to the trough and dif fusion was performed overnight in a cold room. After deproteinization by shaking in 0.3 M and 0.15 M NaCl solutions and then in water, each for 1 day, the gel was a i r - d r i e d , and stained with Amido Black 10B dye solut ion. Quantitative immunochemical analysis of IgG was carried out by radial immunodiffusion (R.I.D.) with a R.I.D. k i t (Miles Laboratories Inc. ) . Whey or protein samples were dialyzed against 20 mM sodium phosphate buffer pH 7.0 for 2 days and freeze-dr ied. Whey samples were then dissolved in 0.05 M barbital acetate buffer pH 8.3 to give a concentration within the range of the kit used for determination of immunoglobulins. After deproteinizat ion, the gel was a i r -dr ied and then stained with Amido Black 10B dye solut ion. 2. Enzyme linked immunosorbent assay for anti-1ipopolvsaccharide act iv i ty  determination The method of Stephens (1984) was used with s l ight modif ications. Immulon 2 flat-bottomed microti tre plates were coated with 100 uL of 0.01% 1ipopolysaccharides (LPS) in coating buffer (0.05 M sodium carbonate, pH 9.6) for 2 hr at room temperature and washed three times in 0.01 M sodium phosphate bufferred saline (PBS) containing 0.05% Tween, pH 7.2. Serial di lut ions of immunoglobulins, made in PBS/Tween, were dispensed in 100 uL volumes and the plates incubated for 2 hr at room temperature. After further washing with PBS/Tween (3 times), 100 uL of alkal ine phosphatase conjugated rabbit antibovine IgG (1:750 d i lu t ion in PBS/Tween) were added and incubated for 2 hr. Plates were washed and 100 uL of substrate (p-nitrophenyl phosphate, 1 mg/mL in I'M diethanolamine buffer at pH 9.8 containing 0.5 mM magnesium chloride and 0.2% sodium azide) was added. After 30 min the reaction was stopped by addition of 20 uL 5N NaOH and the absorbance change was read on an ELISA plate reader 30 (Titertek Multiscan, Flow Laboratories, Scotland) with a 405 nm f i l t e r . Corrections were made for non-specif ic adsorption of Igs. 3. Sandwich enzyme linked immunosorbent assay for IgG assays The method of Troncone et a l . (1986) was used with modif ications. Immulon 2 flat-bottom microt i tre plates were coated with 100 uL rabbit anti-bovine IgG (1:100 d i lu t ion in PBS/Tween) and incubated for 2 hr at room temperature and washed three times in PBS containing 0.05% Tween. Serial di lut ions of immunoglobulin samples, made in PBS/Tween, were dispensed in 100 uL volumes and the plates incubated for 2 hr at room temperature. After further washing, 100 uL of a lkal ine phosphatase conjugated rabbit anti-bovine IgG (1:750 di lut ion in PBS/0.05% Tween) were added and incubated for 2 hr. Plates were washed and 100 uL of substrate (p-nitrophenyl phosphate disodium 1 mg/mL in 1 M diethanolamine buffer at pH 9.8 containing 0.5 mM magnesium chloride and 0.2% sodium azide) was added. After 30 min the reaction was stopped by the addition of 20 uL 5 N NaOH and the absorbance was read on an ELISA plate reader with a 405 nm f i l t e r . Corrections were made for non-specif ic adsorption of Igs. E. SODIUM HEXAMETAPHOSPHATE TREATMENT OF CHEESE WHEY An al iquot of 10% sodium hexametaphosphate (SHMP) solution was added to 25 mL of pH adjusted cheese whey while maintaining the pH by dropwise addition of 3 N NaOH or 3 N HC1. The mixture was held for 1 hr, then centrifuged at 10,000 x g for 15 min. The precipitate was dispersed in 5 mL of 0.5 M Tris-HCl buffer, pH 6.8, and made up to 25 mL after further pH adjustment to 6.8. The supernatant was neutralized to pH 6.8 with 3 N NaOH. The samples were dialyzed against d i s t i l l e d water for 48 hr and then freeze dried (Figure 1). 31 CHEDDAR CHEESE WHEY pH 4.0-4.5 SHMP 1.0-1.4 mg/mL at room temp, hold for lhr centrifuge at 10,000 X g PREC PITATE SUPERNATANT disperse in Tris-HCl buffer, pH 6.8 dialyze for 48 hr centrifuge at 10,000 X g adjust, pH 6.8 dialyze for 48 hr centrifuge at 10,000 X g SUPERNATANT freeze dry PRECIPITATE (discard) SUPERNATANT freeze dry PRECIPITATE (discard) p-LG RICH POWDER IG RICH POWDER Figure 1. Flow diagram of the procedure for elimination of B-lactoglobulin from Cheddar cheese whey with SHMP. 32 F. OPTIMIZATION PROCEDURE The mapping super simplex optimization (MSO) of Nakai et a l . (1984) was used to f ind the most suitable conditions for the polyphosphate treatment of cheese whey which would give the maximum separation ef f ic iency of Igs and. a minimum amount of B-lactoglobulin in the supernatant. An IBM PC computer was used for computation for the MSO and centroid mapping optimization (CMO) by the method of Aishima and Nakai (1986). The experimental conditions (factors) used in MSO and CMO were within the following ranges: pH 4.0-4.5, SHMP concentration 1.0-1.4 mg/mL. A l l experiments were carried out at room temperature (22°C) . 1. Mapping super simplex Mapping super simplex introduced by Nakai et a l . (1984) and written for the IBM-PC was used in order to speed up the i terat ive optimization procedure and graphical ly i l l us t ra te the experimental response surface. After doing nine experiments, the level values for each factor used in the optimization were divided into four groups based on their locations on the scale within large, medium and small l i m i t s . The large and small l imits were determined from individual plots of response value (Separation eff ic iency) vs. each factor level ( i n i t i a l and f inal concentration of SHMP and pH). The medium l imit was an average of both large and small l im i ts . These l imits were used for grouping the data. Data points for one factor which belonged to the same groups of other factors were joined together thus giving an estimate of the response surface. The maps for a l l factors provided new level values for each factor . 2. Centroid mapping optimization and simultaneous factor sh i f t Centroid mapping optimization (Aishima and Nakai, 1986) was used in order to improve the optimization ef f ic iency and to allow for a series of experiments 33 to be run simultaneously. After doing another six experiments, the map for each factor was generated in similar manner as in ( F - l ) . The maps provided target values where the high separation e f f ic ienc ies were located. A Simultaneous Factor Shif t Program (Nakai et a l . , 1984) written for an IBM-PC was used. Target values (estimated best separation eff ic iency) were determined from the graphs. The program is designed to sh i f t a l l factor levels obtained from the mapped graphs simultaneously one f i f t h the distance between the present best value and the target value. The new experimental conditions (vertices) result ing from the Simultaneous Factor Shif t Program were investigated and their response values were calculated. G . EVALUATION OF SEPARATION E F F I C I E N C Y (RESPONSE VALUE) Peak areas of whey proteins on the electrophoretograms were analysed using a Kontes f iber optic scanner (Model K-494800, Kontes S c i e n t i f i c Instruments, Vineland, NJ) together with a Varicord variable response recorder (Model 42 B, Photovolt Corp, NY). Separation ef f ic iency (SE) was expressed as the "Igs to p-Lg ratio" calculated from peak area of Igs (PAi g s ) and P-Lg (PAp_|_g) on the densitometric patterns as: SE= P A I g s / ( P A I g s + PAp-Lg) P A i g s was estimated by multiplying the heavy chain peak area by a coef f ic ient of 1.4 since the determination of l ight chain peak area was d i f f i c u l t due to overlapping with other minor proteins. The coeff ic ient 1.4 was derived from analysis of IgG standards. For quantitative analys is , the variation of staining and destaining conditions during electrophoresis was standardized using an internal standard of ovalbumin. Ten microl i ters of 0.1% ovalbumin solution treated with SDS and 2-mercaptoethanol (similar to the treatment of sample) was added to each whey 34 sample solution and analyzed simultaneously. The ovalbumin peak area measured for every run was compared with the peak areas measured for a series of ovalbumin standard. The ratio of ovalbumin values, thus obtained, was used as a correction factor . H. SURFACE PLOT Contour and 3-dimensional surface plots were obtained using the UBC Surface Visual izat ion Routines program (Mair, 1982) on an Amdahl 470 V/8 computer. The 3-dimensional plot was rotated and t i l t e d for the best view of the surface: (a) The "about" angle was the angle of turn, in degrees, of rotation about the z -ax is , measured clockwise from the posit ive x -ax is ; (b) the "above" angle was the angle of t i l t , in degrees, of rotation about the y - a x i s , measured above the xy plane. In this work, x=pH, y=SHMP and z=SE (Separation e f f i c iency ) . I. PHOSPHORUS DETERMINATION The phosphorus distr ibut ion of the fractions obtained by SHMP treatment was determined according to the method of Morrison (1964). J . FRACTIONATION PROCEDURES OF BIOACTIVE COMPONENTS 1. Gel f i l t r a t i o n chromatography Immunoglobulins were isolated from colostral whey, acid whey and cheese whey using Sephacryl S-300 (Pharmacia Fine Chemicals, Uppsala, Sweden) (94 x 2.5 cm) and Fractogel TSK HW-55 (EM Science, Gibbstown, NJ) (40 x 2.6 cm). The column of Sephacryl S-300 was equil ibrated with 0.1 M Tris-HCl buffer, pH 8.0 containing 0.5 M NaCl (Pharmacia Fine Chemicals, 1978), while the column of Fractogel TSK was equil ibrated with 0.07 M imidazole - 0.05 M KC1 buffer, pH 6.5. 35 2. S i l i c a adsorption Chromatography Chromatographic conditions were based on the process recommended by Spring and Peyrouset (1982) for s i l i c a , controlled pore glass and alumina. Small columns (1.3 x 7.5 cm) were equil ibrated with 0.005 M sodium phosphate solut ion, pH 8.2, then whey containing 0.005 M phosphate was passed through. After washing off the unbound materials, the bound lactoferr in was eluted with 0.1 M acetic acid containing 0.5 M NaCl, and the immunoglobulin fract ion was eluted with 0.1 M Tris-HCl buffer containing 0.5 M NaCl, pH 9.0. 3. Heparin-Sepharose Chromatography Lactoferr in was isolated from cheese whey by using heparin-Sepharose column after equi l ibrat ion with 0.005 M Veronal-HCl containing 0.05 M NaCl, pH 7.4 (Blackberg and Hernel l , 1980). 4. Metal Chelate-interaction Chromatography Sepharose 6B was activated according to the method of Sundberg and Porath (1974). One hundred grams of suction-dried Sepharose 6B was washed on a glass f i l te r - funne l with water and then mixed with 100 mL of 1,4-butanediol diglycidyl ether (BGE) and 100 mL of 0.6 M sodium hydroxide solution containing 2 mg of sodium borohydride per mil l n i t r e . The suspension was mixed by rotation for 8 hr at 25°C and the reaction stopped by washing the gel on a glass f i l ter - funnel with large volumes of water. Epoxyactivated gel obtained above was coupled to iminodiacetic acid according to the method of Porath and 01 in (1983). To 100 grams of epoxyactivated gel 250 mL of 2 M Na2C03, 12.5 g of disodium iminodiacetate, and 0.15 gram of sodium borohydrate were added. The suspension was kept at 60°C overnight with slow s t i r r i n g . The gel was washed thoroughly on a Blichner funnel 36 with water, with di luted acetic acid (5%) and again with water until the washings were neutral (Figure 2). Iminodiacetic acid 1,4-butanediol d iglycidyl Sepharose 6B (IDA-BGE Sepharose Sepharose 6B) was packed into glass columns with d i s t i l l e d water. The upper one-half to two-thirds of the chelating Sepharose was saturated with copper ions as indicated by their blue color , followed by washing off with d i s t i l l e d water and equi l ibrat ion with the start ing buffer (0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2). Liquid whey containing 0.05 M Tris-acetate and 0.5 M NaCl, pH 8.2 was passed through the copper loaded column. After washing off the unbound whey protein fractions with the start ing buffer, l inear gradient of elution was used to elute the bound proteins; the pH gradient was formed using equal volumes of start ing buffer at pH 8.2 and a l imit buffer at pH 2.8. A l ternat ive ly , bound proteins were eluted with the same buffer at pH 4.0 and 0. 01 M imidazole as a two step e lut ion. The eluent was collected in fractions and protein peaks were detected by UV absorbance at 280 nm using Cary 210 Spectrophotometer (Varian Instrument D iv is ion , CA). Following gradient e lut ion, the chelating gel was regenerated with 0.05 M Na2EDTA solution to s t r ip off the copper ions, followed by 6 M urea to remove any remaining bound proteins. After washing with d i s t i l l e d water, the chelating Sepharose was ready for the next cycle of copper loading and whey treatment (Figure 3). K. DETERMINATION OF CAPACITY OF MCIC 1. Immunoglobulins Crude Ig was isolated form bovine colostral whey by ammonium sulfate prec ip i ta t ion , according to the method of Fey et a l . (1976). A solution of the 37 Agarose-OH + CH2CH-CH2-0-(CH2)4-0-CH2-CH-CH2 0 V 1,4-butanediol d iglycidyl ether (BGE) Agarose-0-CH2-CH0H-CH2-0-(CH2)4-0-CH2-CH-CH2 V BGE-Agarose HN' ,CH2-COOH v CH 2 -C00H iminodiacetic acid (IDA) Agarose-0-CH2-CH0H-CH2-0-(CH2)4-0-CH2-CH0H-CH2-NN ,CH2-C00H IDA-BGE-Agarose Cu2+ CH2-COOH Agarose-0CH2-CH0H-CH2-0-(CH2)4-0-CH2-CH0H-CH2-N Jpu ,CH 2?-Q  SCH2C Cu-IDA-BGE-Agarose Figure 2. Preparation of metal chelate agarose 38 • IDA-BGE Sepharose 6B (1) H20 wash (2) Copper (0.05 M CuCl2) loading (3) H2O wash (4) start ing buf fer 3 equi l ibrat ion Cu-IDA-BGE Sepharose 6B (1) Whey (2) Starting buffer wash > Unbound Fraction ( P - L g , a-La) T Protein bound Cu-IDA-BGE Sepharose 6B pH gradient e lu t ion b > Fraction 1 (Lf, Ig, BSA) > Fraction 2 (Ig) Cu-IDA-BGE Sepharose 6B (1) EDTA wash (-> Cu 2 + ) (2) 6 M urea (-> remaining proteins) a start ing buffer = 0.05 Tr is acetate, pH 8.2, 0.5 M NaCl b l inear gradient: start ing buffer as in Footnote a , l imit buffer = 0.05 M Tr is acetate, pH 2.8, 0.5 M NaCl Figure 3. Flow chart of MCIC process of cheese whey treatment for isolat ion of Ig 39 crude Ig (roughly 0.3% w/v) with A28O of approximately 2.1 was passed through a small column (1.4 x 7 cm) containing 2.2 mL copper-loaded chelating Sepharose 6B equil ibrated in start ing buffer (0.05 M Tris-acetate pH 8.2, 0.5 M NaCl). While the crude Ig solution was being applied to the top of the column, the eluted fractions were continually monitored with respect to A280« Saturation of the column ( i . e . , no further binding of protein) was indicated when A28O of the eluted fractions became equal to A28O o f t n e original crude Ig solut ion. The binding capacity of the copper-loaded gel was calculated, at this saturation point, as the amount of applied crude Ig minus the amount of unbound Ig (This amount of Ig was calculated from the A28O and volume of each f ract ion) . After washing off unbound protein with the start ing buffer, the bound proteins were eluted with 0.05 M Tris-acetate buffer at pH 4.0 containing 0.5 M NaCl. The recovery of Ig was calculated as the percentage of eluted protein compared to bound protein. 2. Ovotransferrin A solution (0.2% w/v) of commercial OVT with A28O ° f approximately 1.94 was passed through a small column (1.4 x 7 cm) containing 3 mL copper-loaded chelating Sepharose 6B equil ibrated with the start ing buffer. While the OVT solution was being applied to the top of the column, the eluted fractions were continually monitored by measuring A280* Saturation of the column ( i . e . , no further binding of protein) was indicated when A28O o f the eluted fractions became equal to 1.94. The binding capacity of the copper-loaded gel was calculated as the difference between the amounts of OVT applied and unbound. After washing off the unbound protein with the start ing buffer, the bound proteins were eluted f i r s t with 0.05 M acetate-Tris buffer at pH 4.0 containing 0.5 M NaCl, and then with 0.01 M imidazole. The recovery of OVT was calculated as the percentage of eluted protein compared to bound protein. 40 3. Transferrin A solution of 0.2% (w/v) transferr in in 0.05 M Tr is -ace t i c acid/0.5 M NaCl buffer , pH 8.2, was passed through the column charged with copper ion. The absorbance at 280 nm (A280) o f t n e eff luent from the column was monitored. The binding capacity was calculated from the differences of the absorbance of the eluted protein as compared to the absorbance at the saturation point of the column. The percentage of protein eluted was calculated by comparing the amount of protein bound to the column with the protein eluted from the column. L. PRETREATMENT OF EGG WHITE Eggs were obtained from the University of Br i t ish Columbia Experimental Farm. To obtain a homogeneous and less viscous sample with suitable flow propert ies, the separated egg whites were blended (2000-2500 rpm, 7-10 sec) in a Lourdes MM-1A MultiMixer (Lourdes Instrument Corporation, Old Bethpage, NY) as reported by Li-Chan et a l . (1986). M. PREPARATION OF APO, DIFERRIC AND DICUPRIC OVOTRANSFERRIN Iron free (apo) OVT was prepared by dialyzing standard OVT f i r s t against 0.1 M c i t r i c ac id , pH 2-3 for 36 hr at 4°C, then against deionized water prior to lyophi l i za t ion . D i fer r ic OVT was prepared by dialyzing the apo-OVT against 1.7 mM ferrous ammonium sulfate for 36 hr, then excess iron was removed by gel f i l t r a t i o n on a Shephadex G-25 column equil ibrated with 0.05 M Tr is-acetate/0 .5 M Nacl , pH 8.2 (Cole et a l . , 1976). In a similar manner, dicupric OVT was prepared by dialyzing against 0.01M cupric chlor ide. 41-N. HISTIDINE MODIFICATION OF PROTEINS Hist idine residues of Ig, TF, OVT and casein fractions were modified according to the method of Rogers et a l . (1977) with modifications. Diethyl pyrocarbonate (DEP) to make the f inal concentration of 20 mM was added d i rec t ly to a 5-10 mg/mL protein solution in 0.05 M phosphate buffer pH 6.6 containing 8 M urea while s t i r r i n g . After 20 min s t i r r ing the extent of ethoxyformyl h ist id ine formation was determined by an increase in A 2 4 o of the reaction solution (E240 = 5.9xl0 3 L M - 1 cm - 1 ) (Roosemont, 1978). The purity of DEP used was determined according to Holbrook and Ingram, (1973). 0. ISOELECTRIC FOCUSING Analytical horizontal polyacrylamide gel i soe lec t r ic focusing (IEF-PAGE) was carried but in a Bio-Rad Model 1415 electrophoresis c e l l , according to the manufacturer's instruct ions. Gel slabs were 45 mm x 125 mm and 0.8 mm thick. Bands were located by means of a Coomassie Blue protein s ta in . P. PREPARATION OF ANTISERA Antiserum to egg white and other proteins were produced by immunizing adult female New Zealand white rabbits (UBC, Animal Care Unit) each time with 1-10 mg of antigen emulsified in Freund's complete adjuvant (FCA). Immunizations were given in multiple subcutaneous s i t e s , and repeated intravenously (I.V.) in a two to six week period by replacing FCA with phosphate buffered sa l ine , pH 7.2, as a carr ier unti l a sat isfactory response was obtained. Serum was tested by double di f fusion in gel against egg white proteins as reported by Garvey et a l . (1977). 42 Q. MEASUREMENT OF BACTERIOSTATIC ACTIVITY The method of Dolby and Stephens (1983) was used for the determination of bacter iostat ic ac t iv i ty of the isolated proteins. Immunoglobulin and transferr in prepared by the MCIC method were added to 5 mL of Trypticase Soy Broth (TSB) at concentrations of 10 mg/mL each or in a mixture of 5 mg/mL each. A half mL of 0.05 M NaHC03 was added to each broth and s te r i l i zed by f i l t r a t i o n (Millex-HA, 0.45 um, Mi l l ipore Corp. Bedford, MA). The broths including the control (TSB only) were inoculated with 10 4 colony forming units (cfu)/mL of the test culture; samples of the broth cultures were taken after 1, 3, and 5 hr. Serial d i lut ions of the bacteria in the broth cultures was accomplished by plating on Trypticase Soy Agar (TSA) with the spiral plater (Anonymous, 1985). The inoculated plates were incubated at 37°C for 18 hr. R. EXTRACTION OF LIPOPOLYSACCHARIDES Lipopolysaccharides of JE^ . col i , S. typhi murium and B^ parapertussi s were extracted by using phenol/water according to the method of Jann (1985). After cul t ivat ion in TSB, the bacteria were k i l l e d by the addition of 1% phenol, centrifuged at 5000 x g for 30 min and washed with 0.15 M saline and centrifuged again. They were then freeze-dr ied. One gram of dry bacteria was suspended in 20 mL of water at 68°C. Twenty mL of 90% phenol, prewarmed to 68°C were added to the bacterial suspension and the mixture was kept at 68°C with vigorous s t i r r ing for 15 min. After cooling to about 10°C in an ice-bath, the suspension was centrifuged at 5000 x g for 30 min. This resulted in the formation of two phases. A precipitate was formed between the layers and a bacterial pel le t in the lower phase. The upper aqueous phase was collected by suct ion, then the lower phenol phase together with the pel le t was treated with 20 mL water at 68°C as described above. The combined aqueous phases were 43 dialyzed against d i s t i l l e d water for 48 hr in the cold room to remove phenol and low-molecular weight material . The solution was then freeze-dried to give a white powder. S. LACTOPEROXIDASE ASSAY The lactoperoxidase content of fractions obtained by MCIC method were analyzed by using the procedure described by Sigma Chemical Co. (Bu l l . No. 8-84 for peroxidase prod. No. p.8250). A substrate mixture of 0.1 M potassium phosphate buffer, pH 6.0 (0.32 mL), 0.147 M hydrogen peroxide (0.16 mL), 5% (w/v) pyrogallol (0.32 mL) and d i s t i l l e d water (2.1 mL) was mixed by inversion. The i n i t i a l A420 was monitored unti l constant with a cuvette containing H£0 as the reference. To this mixture, at zero time, 0.1 mL of a lactoperoxidse containing fract ion (10 mg lactoperoxidase per mL of 0.1 M potassium phosphate buffer, pH 6.0) was added. The solution was mixed by inversion and the increase in the in A420 was recorded every 10 seconds for about two minutes, using a Cary 210 Spectrophotometer (Varian Canada Inc. ) . The i n i t i a l l inear rate of increase in absorbance was determined using l inear regression, and was used to determine the units of lactoperoxidase ac t iv i ty per mg s o l i d : A420/20 second Units/mg sol id = (12)* x (mg enzyme as solid/mL reaction mix) 12* = Extinction coeff ic ient as determined by Sigma Units obtained were compared to that of bovine lactoperoxidase (80 units/mg protein using pyrogallol as substrate). Lactoperoxidase was calculated as the percentage of lactoperoxidase content in the pooled f ract ions. 44 T. SEPARATION OF HEAVY AND LIGHT CHAINS OF IMMUNOGLOBULINS Reduction and alkylat ion of S-S groups in the Ig r ich fraction were performed according to the method of Garvey et a l . (1977). A 10 mL solution of 2% Ig in 0.55 M Tris-HCl buffer pH 8.2 was bubbled with N 2 for 15 min. Five mL of 0.15 M di th iothre i to l in 0.55 M Tris-HCl buffer were added, and the mixture was allowed to react for 1 hr under a posit ive N 2 atmosphere. Ten mL of 0.25 M 2-iodoacetamide (IAA) in 0.55 M Tris-HCl buffer pH 8.2 were then added to the reduced Ig sample and the mixture was kept in the cold room for 1 hr. The reduced and alkylated Ig was then equil ibrated with 1 M propionic acid or 0.1 M Tris-HCl buffer containing 4 M guanidine-HCl and 1 mM IAA, pH 8.2 and fractionated on a Sephadex G-75 and an Ultrogel ACA 54 column (40 x 2.6 cm), respectively. RESULTS AND DISCUSSIONS REDUCTION OF 0-LACTOGLOBULIN CONTENT OF CHEESE WHEY BY USING SODIUM HEXAMETAPHOSPHATE 46 Since elimination of B-lactoglobulin by f e r r i c chloride methods (Kaneko et a l . 1985; Kuwata et a l . 1985) can saturate iron binding proteins present in whey and may result in loss of the antimicrobial ac t iv i ty of l ac to fe r r in , non-ferric methods were investigated (Appendix 1). Polyphosphates have been extensively used as additives in food processing. Gordon (1945) in his patent used polyphosphate to extract whey proteins from cheese whey. In this part , polyphosphates were investigated as a possible means for the select ive precipitat ion of B-lactoglobulins from cheese whey leaving immunoglobulins in the supernatant. A. OPTIMUM CONDITIONS FOR SEPARATION OF IMMUNOGLOBULINS AND B-LACTOGLOBULIN Mapping simplex optimization with two factors generated three experiments cal led the " i n i t i a l simplex". After obtaining the response values which were, in our case, the separation ef f ic iency (SE) of immunoglobulins, the values were reported back to the computer to obtain new vertices (experimental conditions) in a form of the repetit ive sequences of centroid, re f lec t ion , and curve- f i t t ing . After performing nine vertices for the MSO, mapping was done by plott ing the response values against the factor leve ls . A crude approximation of the response surface appeared to direct the search for higher SE towards more acidic conditions (pH 4.0-4.2); therefore, the range for lower and upper l imit of pH was narrowed down to 4.0-4.2. In a similar manner, higher SE could be expected at higher polyphosphate concentrations (1.2-1.4 mg/mL); therefore, the range of concentration of SHMP was restr icted to 1.3-1.4 mg/mL. With the new lower and upper l imits for the pH and SHMP, the CMO program was appl ied. After entering the new ranges, a new i n i t i a l simplex (three vertices) was created. The experiments were carried out and the response values were reported. Upon improvement in the response values, simultaneous shi f t was implemented after six experiments in the centroid search. Figures 4A and 4B show gure 4. Approximate response surface patterns for (A) pH and (B) SHMP concentration obtained by mapping accumulated data from simplex optimization (Vertices 1-9) and centroid optimization (Vertices 10-15). T target values of pH and SHMP. 48 the approximate response surface for both factors from which, based on the present best values of 4.07 and 1.33, the target values (T) of 4.03 and 1.36 were set for pH and SHMP, respectively. However, because of the fa i lure to achieve further improvement in response values, further experimentation was discontinued. Examining the best response value (Vertex 15 Figures 4A and 4B), i t was found that pH 4.07 and 1.33 mg SHMP/mL yielded about 80% elimination of B-lactoglobulin from Cheddar cheese whey (into the prec ip i ta te ) , with almost complete recovery of immunoglobulins in the supernatant. The majority of a-lactalbumin was found in the supernatant as indicated by SDS-PAGE (Figure 5). However, most of the bovine' serum albumin was precipitated along with the B-lactoglobul in with only a small amount remaining in the supernatant. Recovery of Igs in the supernatant was evident immunochemically as shown in Figure 6. The supernatants showed a long, outer precipi tat ing l ine corresponding to the standard IgG arc; while the precipitates showed no such precipi tat ing l ine . The inner precipi tat ing l ines (by the sample application well) might represent l ac to fe r r in , transferr in and IgM. Serum albumin and 8- lactoglobul in arcs are indicated by the inner and the outer precipi tat ing l i n e s , respectively, formed toward the anode, far from the application wel l . The response of anti-bovine whey protein antiserum toward a-lactalbumin was rather weak (not shown), probably due to i ts low molecular weight and consequently lower ant igenic i ty . Approximately 90 % of Igs in the supernatant were determined to be IgG by means of R.I.D. (not shown). B. THREE DIMENSIONAL ILLUSTRATION OF EFFECTS OF pH AND HEXAMETAPHOSPHATE ON SEPARATION EFFICIENCY Contour and 3-dimensional surface plots were generated by computer to aid in v isual izat ion of the relationship between pH, SHMP and SE of immunoglobulins in Cheddar cheese whey. Figures 7A and 7B showed that, in general, combinations of 49 C C W p s = BSA m IgG-HC OVA * IgG-LC + m B-Lg 4ft<*-La Figure 5. SDS-PAGE of supernatant (S) and precipitate (P) obtained after treatment with 1.33 mg/mL SHMP at pH 4.07. CCW, Cheddar cheese whey; a-La, a-lactalbumin; B-Lg, B- lactoglobul in; IgG-HC, immunoglobulin G heavy chain; IgG-LC, immunoglobulin G l ight chain; BSA, bovine serum albumin; OVA, ovalbumin. 50 s CCW p IgG 8-L9 abwp abwp abwp abwp abwp abwp Figure 6. Immunoelectrophoretic pattern of Cheddar cheese whey. S, supernatant; CCW, Cheddar cheese whey; P, precipi tate; IgG, immunoglobulin G; B-Lg! 8- lactoglobul in; abwp, antibovine whey proteins. 51 (A) Figure 7. Contour (A) and 3-dimensional (B) surface plots of relationship between pH, SHMP and Separation effeciency (SE) of cheese whey treatment, ("about" angle=60 and "above" angle=35 for 3-dimensional p lo t ) . 52 53 low pH and high SHMP concentration resulted in good separation ef f ic iency. Separation ef f ic iency was improved by decreasing pH values below 4.25 and by increasing SHMP concentration above 1.2 mg/mL. By lowering the pH below the isoe lec t r ic point of whey proteins, the posit ive side chain amino groups could interact with the negative groups surrounding phosphate molecules by which polyphosphates act as cross- l inking agents. However, di f ferent proteins may interact d i f ferent ly via polyphosphates to form aggregates. Surface exposed amino groups and unfolding or expansion of protein molecules when the polyphosphate is bound may play an important role in that interact ion, leading to preferential precipi tat ion of p-lactoglobulins (Melachouris, 1972). The contour plot (Figure 7A) shows three humps with SE of 0.508, 0.577 and 0.302, which can be seen also in Figure 7B. Multiple peaks might have been caused by the absence of data points between SE 0.508 and the other two peaks with SE's of 0.302 and 0.577; in other words, more experiments under the conditions with closer intervals between pH 4.2 and 4.08 might be required in order to obtained a smoother surface. C. ELIMINATION OF PHOSPHORUS The phosphorus distr ibut ion in the supernatant and the precipitate are shown in Table 5. Removal of 72.27. and 45.3% of the total phosphorus from the supernatant and the precip i ta te , respectively, was achieved by d ia lys is against d i s t i l l e d water for 48 hr. Faci le removal of polyphosphate from whey preparations by d ia lys is might indicate weak binding of phosphorus with whey proteins. Since no precipi tat ion of whey proteins by SHMP was observed at pH values higher than 5, i t was assumed that ionic interaction might be involved in the interaction between SHMP and proteins. At pH values lower than the isoe lec t r ic point, posit ively charged groups (basic amino acid residues) in protein molecules might interact Table 5. Phosphorus distr ibut ion in supernatant and precipitate obtained by SHMP treatment. P, phosphorus. Fraction mg P/100 mL mg P/100 mL % removed after by d ia lys is d ia lys is Supernatant 65.0 18.1 72.2 Precipitate 16.0 8.75 45.3 Cheese whey 50.0 8.6 82.8 55 with each other via SHMP and cause the aggregation of the proteins. By increasing the pH above the isoe lec t r ic point and increasing net negative charges, this interaction might be disrupted and the free SHMP could be removed by a simple d ia lys is process, thereby separating a whey protein fract ion containing a phosphorus content within the level recommended for infant formulas (0.033%) (Friend et a l . , 1983). D. PROPOSAL OF NEW INFANT FORMULA Table 6 compares the composition of the new infant formula to that of human and cow's milk and the current commercial SMA (whey-based) formula. In the commercial SMA formula, the ratio of casein/whey proteins of cow's milk (79/21) has been changed to 40/60 in order to mimic the ratio found in human milk. However, simple adjustment of casein/whey protein ratio does not minimize the compositional differences between cow and human milk proteins, i . e . higher contents of a-lactalbumin, l ac to fe r r in , immunoglobulins and lysozyme in human milk as compared to cow's milk. By eliminating B-Lg completely with fu l l retention of other whey proteins, the whey protein composition of the new B-Lg-free formula would be as shown in Table 6. In this proposed formula, immunoglobulins and lactoferr in are much closer in their quantities to that found in human milk. In addit ion, lysozyme separated from egg white can be incorporated in this new infant formula as proposed by Friend et a l . (1983), in order to improve the therapeutic value of infant formula. Thus, the B-Lg-reduced supernatant which is rich in immunoglobulins when incorporated into infant formulae, may be an additional benefit to infant feeding. 56 Table 6. Protein composition of human and cow's milks and whey-based and proposed p- lactoglobulin (P-Lg)-free infant formula. Whey-based P-Lg free Human3 Cow formula formula Protein Total % Total % Total % Total % Total 100 100 100 100 Caseins 35.0* 79.0 3 40. OC 40.0 Total whey 65.0 3 21.0 3 60.0 C 60.0 a-lactalbumin 17.0 3 2.8 b 8.0 17.2 d p-lactoglobulin — 11.2 b 32.0 — immunoglobul ins 11.0 3 2.3 6.6 d 1 4 . l d serum albumin 6.0 3 1.8 5. id 10.9 d lactoferr ins 17.0 3 1.7b 4 .9 d 10.5 d lysozyme 6.0 3 — — — others 8 .0 3 1.2 b 3.4 d 7.3d 3 Gurr, 1981 b Calculated from electrophoretic scanning of Cheddar cheese whey c Friend et a l . , 1983 d Whey protein composition is calculated based on our data (b) 57 PART II SEPARATION OF BOVINE IMMUNOGLOBULINS AND LACTOFERRIN FROM WHEY PROTEINS BY GEL FILTRATION TECHNIQUES 58 Although benefits expected by feeding infants with non-immunized immunoglobulins compared to feeding hyperimmunized Ig (Hilpert et a l . , 1974/ 1975; Bal labr iga, 1982; Ebina et a l . , 1984, 1985) are unknown, some uses may be j u s t i f i e d in collaboration with other antimicrobial components in cow's milk as discussed by Packard (1982). Lactoferrin together with secretory immunoglobulin A from human milk showed a considerable bacter iostat ic effect against human enteropathogenic strains of E^ col i (Dolby and Stephens 1983; Stephens et a l . , 1980). In this part of the thesis , gel f i l t r a t i o n techniques were assessed for immunoglobulins and lactoferr in f ract ionat ion. A. GEL FILTRATION ON SEPHACRYL S-300 Gel f i l t r a t i o n is the method of choice for the preparative isolat ion of proteins (Van Oss, 1982-1983). Bovine colostrum contains proteins with a wide range of molecular sizes from a-lactalbumin (MW ca. 14,500) to IgM (MW ca. 1X10 6). Gel f i l t r a t i o n of high molecular weight proteins requires a highly porous and r ig id gel to provide a good s t a b i l i t y and short separation times, therefore, Sephacryl S-300 was chosen for this experiment. The y ie ld of immunoglobulin isolated by gel f i l t r a t i o n can be as high as 90% (Van Oss, 1982-1983). Figures 8 and 9 show the separation patterns of the untreated bovine colostral whey and the ammonium sulfate treated whey, respectively. Results indicated that two major peaks were obtained in the elution pro f i l e of treated whey (Figure 9) which were also in the elution prof i le of untreated whey (Figure 8). The f i r s t fract ion (Fl) appeared in the void volume of the column which indicated that the molecular weight was in the v ic in i t y of a m i l l i o n , and the second peak indicated a lower molecular weight protein. Fraction 3 (F3) of both untreated bovine colostral whey and the ammonium sulfate treated whey contained 59 ELUTION VOLUME, ml Figure 8. Gel f i l t r a t i o n of bovine colostral whey on Sephacryl S-300 Superfine column (94 x 2.5 cm) eluted with 0.1 M Tris-HCl buffer pH 8.0 containing 0.5 M NaCl. Flow rate, 12 mL/hr. 1 and 3 are fractions 1 and 3, respectively. 60 Figure 9. Gel f i l t r a t i o n of crude Ig obtained from ammonium sulfate treatment on Sephacryl S-300 Superfine column (94 x 2.5 cm), eluted with 0.1 M Tris-HCl buffer pH 8.0 containing 0.5 M NaCl, flow rate 12 mL/hr. 1 and 3 are fractions 1 and 3, respectively. 61 99% IgG when analyzed by radial immunodiffusion (Table 7) and double sandwich ELISA. Figure 10 shows the electrophoretic patterns of colostral whey and fract ions obtained by gel f i l t r a t i o n fractionation (Figure 8: F l , F3). It was observed that F3, which represented IgG, was a highly pur i f ied fraction containing IgG. However, the less pure Fl fract ion contained some contaminants which moved the same distance as lac to fer r in . Lactoferrin may have bound firmly to the high molecular weight compound (IgM) through hydrogen-bonding since both proteins are glycoproteins. Immunoelectrophoresis is one of the most important techniques for the ident i f icat ion and characterization of immunoglobulins and their c lasses, because of i t s sens i t iv i ty and spec i f i c i t y (Ohtani and Kawai, 1981). Interpretation of immunoelectrophoretic patterns depends upon the spec i f ic i ty and the potency of the ant i -sera employed. Immunoelecrophoretically, F l and F3 showed a single prec ip i t in arc against anti-whole bovine antiserum, attesting to the purity of these fractions (Figure 11). Since F l diffused a short distance, i t represents a high molecular weight (IgM) while F3 which diffused a longer distance may represent the lower molecular weight IgG. The di f fusion rate of antibody or antigen is inversely proportional to the molecular weight (Atassi et a l . , 1984). B. GEL FILTRATION ON TSK HW-55 Figure 12 is an elution pattern of colostral whey on a TSK HW-55 column which is basica l ly similar to the elution prof i le of colostral whey on Sephacryl S-300. SDS-PAGE indicated that the major portion of F l and F2 were immunoglobulins. However, F l and F2 were not as pure as Fl and F3 obtained from Sephacryl S-300 (Figure 13). Immunoelectrophoresis c lear ly showed that the Table 7. Immunoglobulin G contents* of fractions obtained from gel f i l t r a t i o n on Sephacryl S-300 and crude Ig prepared by ammonium sulfate treatment. Sample Protein content** (mg/mL) IgG content (mg/mL) Purity % F3 (Figure 8) 10 9.90 99 F3 (Figure 9) 10 9.90 99 Crude Ig 10 5.10 51 AW-fraction 2 (Figure 15) 7.2 6.00 83.3 CCW-fraction 2 (Figure 16) 7.6 7.00 92.1 * Determined by radial immunodiffusion analysis . * * Determined by BioRad reagent. 63 F1 F3 Ig LF CW Figure 10. SDS-PAGE of fractions obtained from gel f i l t r a t i o n on Sephacryl S-300. F l , f ract ion 1; F3, fract ion 3 (Figure 8); Ig, crude immunoglobulin; LF, l ac to fe r r in ; CW, untreated colostral whey; HC, LC, immunoglobulin heavy and l ight chains, respectively. 64 a F3 F3 F1 F1 BSA TF Figure 11. Immunoelectrophoretic analysis against anti-whole bovine antiserum of fract ions obtained from Figure 8. F3, fract ion 3; F l , fraction 1; ; BSA, bovine serum albumin; TF, t ransferr in . 65 0 120 240 ELUTION VOLUME, ml Figure 12. Gel f i l t r a t i o n pattern of colostral whey on TSK column (40 x 2.6 cm) eluted with 0.07 M imidazole-0.05 M KC1 buffer, pH 6.5; flow rate, 50 mL/hr. 1 and 2 are fractions 1 and 2, respectively. 66 F1 F2 1 2 3 4 M AW CW - HC n * m W LC Figure 13. SDS-PAGE of fractions (Fl and F2) obtained from Figure 12 as compared to standards. Lane 1, a-lactalbumin; Lane 2, B-lactoglobulin; Lane 3, bovine serum albumin; Lane 4, t ransferr in ; M, standard mixture; AW, acid whey, CW, colostral whey; HC and LC, immunoglobulin heavy and l ight chain, respectively. 67 fract ions obtained by TSK column contained some impurit ies, mainly bovine serum albumin, not found in the fractions obtained by Sephacryl S-300 column (Figure 14). By comparing fractions obtained from gel f i l t r a t i o n on Sephacryl S-300 ( F l , F3) with those obtained on TSK HW 55 ( F l , F2), i t is obvious that the former contains immunoglobulins of higher purity than those obtained from TSK HW 55 (Figure 14). Therefore, the Sephacryl S-300 type column was chosen to isolate Ig from acid and cheese whey. C. ISOLATION OF IMMUNOGLOBULINS FROM WHEY PROTEINS Figure 15 shows the elution prof i le of acid whey on Sephacryl S-300 while Figure 16 is the elution prof i le of Cheddar cheese whey on the same column. The f i r s t f ract ion (Fl) which eluted at the void volume, was a turbid solution containing high molecular weight components. The Fl fract ion was sharper for acid whey than that for the cheese whey. The F l fract ion may represent the l ipoprotein fract ion in both wheys. The second fract ion (F2), a shoulder of fract ion 3, contained immunoglobulins from acid whey and cheese whey as indicated by SDS-PAGE (Figure 17 and 18). The immunoglobulin fract ion contained some impurities which were probably lactoferr in and bovine serum albumin (Table 7). Immunoelectrophoresis showed that fract ion F2 of both acid whey and cheese whey contained mainly immunoglobulins with smaller amounts of bovine serum albumin (Figure 19). D. ISOLATION OF LACTOFERRIN FROM WHEY PROTEINS Figure 20 shows the elution prof i le of Cheddar cheese whey when eluted from a haparin-Sepharose column with a l inear NaCl gradient. Lactoferrin was selec-t ive ly adsorbed to the column from cheese whey and was eluted at about 0.5 M 68 Figure 14. Immunoelectrophoresis of fractions obtained by Sephacryl S-300 and Fractogel TSK column against anti-whole bovine serum antiserum. (Fl-S and F3-S, f ract ion 1 and 3 of Figure 8, respectively) (Fl-T and F2-T fraction 1 and 2 of Figure 12, respectively) IgG, immunoglobulin G. 69 Figure 15. Gel f i l t r a t i o n of acid whey on Sephacryl S-300 Superfine column (94 x 2.5 cm), eluted with 0.1 M Tris-HCl buffer, pH 8.0 containing 0.5 M NaCl. Flow rate 18 mL/hr. 1, 2, 3 and 4 are the fractions obtained. 70 ELUTION VOLUME, ml Figure 16. Gel f i l t r a t i o n of Cheddar cheese whey on Sepharyl S-300 Superfine column (94 x 2.5 cm), eluted with 0.1 M Tris-HCl buffer, pH 8.0 containing 0.5 M NaCl. Flow rate 18 mL/hr. 1, 2, 3 and 4 are the fractions obtained. 71 IgG AW 4 3 2 1 AW Figure 17. SDS-PAGE of fractions (1, 2, 3, 4) obtained from Figure 15. AW, acid whey; IgG, immunoglobulin G; HC and LC, heavy and l ight chains of immunoglobulins, respectively. 72 Figure 18. SDS-PAGE of fraction (1, 2, 3, 4) obtained from Figure 16. IgG, immunoglobulin G; CCW, Cheddar cheese whey; HC and LC, heavy and l ight chains of immunoglobulins, respectively. 73. Figure 19. Immunoelectrophoresis of fractions obtained from gel f i l t r a t i o n of whey protein against anti-whey protein antiserum. AW, acid whey; Fl-H fract ion 1 from Figure 20; F2-A, fract ion 2 from Figure 15; F2-C, fraction 2 from Figure 16; IgG, immunoglobulin G. 74 ELUTION VOLUME, ml Figure 20. Heparin-Sepharose chromatography of Cheddar cheese whey. Cheese whey (400 mL) dialyzed against 0.05 M NaCl in 5 mM veronal-HCl, pH 7.4 was applied to the column (10 mL sett led ge l ) . The column was washed with the same buffer and then eluted with a l inear gradient of NaCl ( ) as indicated. The flow rate was 50 mL/hr. UB, unbound proteins; 1, fraction 75 NaCl i s o c r a t i c a l l y . The isolated lactoferr in gave only one precipitant l ine with Immunoelectrophoresis (Figure 19), which indicated the high purity of this f rac t ion . SDS-PAGE (Figure 21) also yielded a single band, confirming the purity of this f rac t ion . E. ANTI-LIPOPOLYSACCHARIDES ACTIVITY OF ISOLATED IMMUNOGLOBULINS Figure 22 shows the anti-1ipopolysaccharide ac t iv i ty of immunoglobulins isolated from colostrum using a gel f i l t r a t i o n technique. Lipopolysaccharides were extracted from the pathogenic bacter ia , E^ col i , S_;_ typhimurium and Bordetella parapertussis and their binding with isolated Ig was measured using an enzyme linked immunosorbent assay (ELISA). The isolated IgG was shown to have binding ac t iv i ty against LPS isolated from E^ col i and to a lesser extent against S_j. typhimurium. Higher recognition of colostral Ig to LPS from E± col i as compared LPS from S_j. typhimurium may indicate that the dairy cow was infected more often by E^ col i than by S. typhimurium. Surprisingly isolated Ig showed ac t iv i ty and recognition of LPS isolated from EL parapertussis which causes whooping cough in infants, and which may indicate the presence of s imi lar i t ies in the antigenic structure of LPS between this bacterium and those extracted from E_j_ col i . The strain of col i used in this study is known to cause diarrhea in infants by producing one or both of two classes of enterotoxins. E^ col 1 enterotoxin mainly affects f lu id transport processes of the small intest ine, therefore, any changes in either or both absorption and secretion may result in diarrhea (Holmgren, 1985). Holmgren (1985) discussed possible approaches for the prevention of and treatment of E_j. col i pathogenic act ion. One approach was to use receptor blockade by using a non-toxic binding agent. This approach was to prevent binding of toxin to the epithelium. Therefore, high binding act iv i ty 76 CCW UB F1 CCW LF Figure 21. SDS-PAGE of fractions obtained from Figure 20, CCW, Cheddar cheese whey; UB; unbound whey proteins to Heparin-Sepharose column; F l , lactoferr in r ich f rac t ion; LF, l ac to fer r in . 77 2.0r Pg Ig Figure 22. Anti-1ipopolysaccharide ac t iv i ty of colostral IgG isolated by gel f i l t r a t i o n on Sephacryl S-300. m-m , E i col i LPS; D-O , S i typhimurium LPS; o - 0 > l i parapertussis LPS. 78 of isolated Ig toward LPS extracted from g± col i may interfere with the binding of bacterial toxin to the epithelium thus preventing diarrheal diseases when Ig is used for infant feeding. These studies demonstrate that Cheddar cheese whey could be an important source for the isolat ion of immunoglobulins and lac to fe r r in . Importance of lactoferr in as a bacter iostat ic agent has been well established (Reiter, 1983). Immunoglobulins isolated from cheese whey as an immunoglobulin rich fraction (Fraction 2) can be incorporated into infant formulae. The 1 ipopolysaccharide binding ac t iv i ty of immunoglobulins may give a clear evidence to add this fract ion to commercial infant formulae in order to protect infants from gastrointestinal in fect ion. SEPARATION OF IMMUNOGLOBULINS AND LACTOFERRIN FROM CHEESE WHEY BY ADSORPTION AND CHELATING CHROMATOGRAPHY TECHNIQUES 80 When considering the treatment of a large amount of cheese whey, the d i f f i c u l t i e s in mechanization of gel " f i l t rat ion techniques for extracting immunoglobulins and lactoferr in renders a large scale operation less feasible and more d i f f i c u l t . Therefore, adsorption and metal chelating chromatographic methods were investigated to extract these proteins by an easy and ef f ic ient method. A. ADSORPTION CHROMATOGRAPHY METHODS In searching for a method to isolate immunoglobulins and other bioactive compounds from whey protein, the economic dr ive , ease to mechanize and capacity of the method were used as c r i t e r i a . Figure 23 shows the elution prof i les of adsorbed proteins from the chromatographic treatment of 1 l i t r e of Cheddar cheese whey (adjusted to pH 8.2) on s i l i c a , controlled pore glass and alumina, eluted with acet ic acid solution followed by Tris-HCl buffer. The amount of protein in these fractions from the s i l i c a column was very small , as indicated by the low A28O values and small peak areas. SDS-PAGE (Figure 24) shows that the acetic acid eluted fract ion contained primarily lactoferr in (Lane 2). The amount of protein in the Tris-HCl eluted fraction was too small to be ident i f ied (lanes 5 and 12). The s i l i ca - t rea ted cheese whey ( i . e . , unadsorbed protein f rac t ion , lane 8) showed l i t t l e difference from the electrophoretic pattern of untreated cheese whey (lane 1 & 11). Similar results were obtained when the chromatography was carried out at the start ing pH of 7.5 instead of 8.2. As shown by SDS-PAGE prof i les (Figure 24), the large A28O P e a k eluted by acetic acid from controlled pore glass contained mainly lactoferr in and bovine serum albumin (BSA) (lane 3), while the peak eluted by Tris-HCl contained a mixture of proteins including lac to fer r in , BSA and Ig (lane 6). 81 Figure 23. Elution prof i les of adsorbed proteins from s i l i c a (S) (close to the basel ine) , controlled pore glass (C) and alumina (A) chromatographic treatment of Cheddar cheese whey. One l i t r e of Cheddar cheese whey in 0.005 M Na2HP04, pH 8.2 was passed through 1.3 x 7.0 cm column of (S), (C) or (A) equil ibrated with 0.005 M phosphate buffer at pH 8.2. After washing with 30 mL of equi l ibrat ing buffer, the adsorbed proteins were eluted with El (50 mL 0.1 M acetic acid pH 2.77 containing 0.5 M NaCl), then E2 (60 mL 0.1 M Tris-HCl pH 9.0 containing 0.5 M NaCl). The flow rate was 1 mL/min. 82 1 2 3 4 5 6 7 8 9 10 11 12 Figure 24. SDS-PAGE prof i les of cheese whey and fractions obtained from Figure 23. Lane 1, untreated Cheddar cheese whey; Lane 2, acetic acid fraction from s i l i c a sand; Lane 3, acetic acid fract ion from controlled pore glass; Lane 4, acetic acid fraction from alumina; Lane 5, Tris-HCl fraction from s i l i c a sand; Lane 6, Tris-HCl fraction from controlled pore glass; Lane 7, Tris-HCl fract ion from alumina; Lane 8, unbound fract ion from s i l i c a ; Lane 9, unbound fraction from controlled pore g lass; Lane 10, unbound fraction from alumina; Lane 11, untreated Cheddar cheese whey; Lane 12, acetic acid fract ion from alumina, LF, lac to fer r in ; HC and LC, immunoglobulin heavy and l ight chains, respectively. 83 The acetic acid elution of alumina column yielded a small peak followed closely by a second larger peak. The larger peak contained mainly lactoferr in and BSA according to SDS-PAGE (Figure 24 lane 4). Tris-HCl elution yielded only a very broad peak with low A280; t n i s f ract ion contained mainly lactoferr in (Figure 24, lane 7). The Tris-HCl eluant emerging from the column was observed to be turbid , start ing at the 80 mL elution volume, which may be due to l i p i d fractions eluted from the alumina. Of the adsorption chromatographic support materials investigated, controlled pore glass appeared to be the most promising for isolat ion of Ig. However, reducing the amount of cheese whey sample to 250 mL and analyzing the unbound fract ion indicated that appreciable quantities of Ig were not adsorbed to the CPG column. This suggested that the capacity of CPG for Ig adsorption was not very high. Increasing the CPG bed volume from 10 to 20 mL improved recovery somewhat, but the capacity was s t i l l insuf f ic ient for e f f ic ien t removal of Ig from whey. B. METAL CHELATE INTERACTION CHROMATOGRAPHY 1. Acid whey Acid whey (obtained from ac id i f i ca t ion of raw skimmilk) was applied to a column containing copper ions immobilized on Sepharose 6B. After washing unbound proteins with the start ing alkal ine buffer, the adsorbed proteins were eluted using a l inear gradient. Figure 25 shows the elution pro f i l e of adsorbed proteins from MCIC treatment of 1 l i t r e of acid whey. Two major peaks were detected by monitoring A28O of the ef f luent . The fractions comprising the f i r s t peak were yel lowish, while the second peak was co lor less . SDS-PAGE (Figure 26 lane 1) indicated that the f i r s t peak was the lactoferr in rich f rac t ion ; however, i t also contained other proteins which were adsorbed to the 84 Figure 25. Elution prof i les of adsorbed proteins from MCIC on Sepharose 6B treatment of 1 L Cheddar cheese whey (CCW) and 1 L acid whey (AW) (obtained from raw mi lk) , using l inear gradient elution of 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.0 to 2.8. Flow rate was 0.8 mL/min. CCW, Cheddar cheese whey; AW, acid whey. 85 Figure 26. SDS-PAGE prof i les of acid whey and fractions obtained by MCIC and gel f i l t r a t i o n . Lanes 1 and 2, are fractions obtained by MCIC; Lane 3, unbound material to MCIC column; Lanes 4 and 5, peak 1 and 2 obtained by gel f i l t r a t i o n on Sephacryl column; Lane 6, mixture of standard proteins ( t ransferr in , bovine serum albumin, B-lactoglobulins and a-lactalbumin); Lane 7, acid whey. LF, l ac to fe r r in ; HC and LC, heavy and l ight chains of immunoglobulins, respectively. 86 column and eluted in the same f rac t ion; these proteins were Ig and BSA. The second smaller peak was mainly composed of Ig (Figure 26 lane 2). Figure 26 (lane 3) also indicated that the unbound fract ion contained no Ig but mainly a-La and B-Lg. Immunochemical analyses (Table 8) showed that the majority of active IgG ( i . e . , over 77%), was present in the second pooled peak, over 20% of active IgG was present in the f i r s t pooled peak and was undetectable in the unbound f ract ions. Table 9 shows the distr ibut ion of whey proteins in comparison to the pooled fractions obtained by the MCIC process as calculated from the densitometric scanning of SDS-PAGE. Lactoferrin and immunoglobulins were predominant proteins in peak one (Fl) (48% and 20%, respectively) while bovine serum albumin represented 27% of this f rac t ion . This fraction was almost free from a-La and P-Lg. Immunoglobulins were the major proteins in peak 2, representing 88.4% of the total proteins in this f ract ion . Immunoglobulins and lactoferr in in both fractions represented more than 75% of the total proteins. Immunoelectrophoresis conducted against anti-bovine whole serum antibodies (Figure 27) also indicated the presence of IgG (the predominant type of immunoglobulins in bovine milk) as well as the presence of BSA in both f ract ions. 2. Cheddar cheese whey Figure 25 shows an elution prof i le of adsorbed proteins from MCIC treatment of 1 L Cheddar cheese whey at pH 8 .2. A similar elution pattern was obtained from acid whey. Two major peaks were detected by monitoring A28O o f the ef f luent . The fractions comprising the f i r s t peak were s l igh t ly yellowish, while the fractions from the second peak were c lear . Table 8. IgG ac t iv i ty of peak and unbound fractions from MCIC treatment of whey. Concentration mg/mL IgG protein IgG purity Acid whey unbound ND 80.3 ND 1st peak 5.0 20.3 24.6 2nd peak 18.0 23.4 77.2 Cheddar cheese whey unbound ND 79.3 ND 1st peak 1.25 51.4 2.4 2nd peak 40.0 75.4 53.0 ED whey powder unbound ND 65.2 ND 1st peak 5.00 31.3 16.0 2nd peak 20.00 70.7 28.3 ND not detectable a IgG ac t iv i ty of dialyzed fractions was determined by RID D Protein concentration was determined by Kjeldahl N x 6.38 88 Table 9. Whey proteins d is t r ibut ion* in acid whey, fractions obtained from MCIC of bovine acid whey, and the unbound materials to MCIC column. Protein components Acid whey Fl F2 Unbound % % % % Materials % a-lactalbumin 13.9 27.9 B-lactoglobulin 56.9 4.8 2.1 67.8 Immunoglobulins 11.4 20.1 88.4 — Bovine serum albumin 9.2 27.1 9.5 3.3 Lactoferrin 8.6 48.0 — — * Calculated from peak area of the electrophoretic patterns. 89 Figure 27. Immunoelectrophoretic analysis of fractions obtained by MCIC of acid whey. IgG, immunoglobulin G; P2 and PI are Peak 2 and 1, respectively of Figure 25; TF, t ransfer r in ; BSA, bovine serum albumin. 90 SDS-PAGE (Figure 28) indicated that the f i r s t large peak was composed of lactoferr in and bovine serum albumin (lanes 5 and 7) while the second smaller peak was mainly composed of immunoglobulins (lanes 6 and 8). Some B-Lg and a-La were also adsorbed on Cu 2 + - immobil ized Sepharose, as indicated in the unbound fract ion at the early stage of whey appl icat ion; no whey proteins were present in the unbound fract ion after 120 mL of whey had been applied (lane 2). However, after 460 mL of applied whey, p-Lg and a-La were not adsorbed (lane 3 and 4) indicating that the column had lower a f f i n i t y for these components than for LF, BSA and IgG. The absence of p-Lg and a-La in the eluted peaks (lanes 5 to 8) can be explained by the fact that they were weakly bound and were either displaced by LF, BSA and IgG during later stages of whey application or were eluted during the washing stage pr ior to the elution of adsorbed proteins. Immunochemical analysis (Table 8) showed that the majority of active IgG was present in the second peak with very minor ac t iv i ty in the f i r s t peak and undetectable amount in the unbound f ract ions. The IgG ac t iv i ty of the second peak (Ig r ich fraction) of Cheddar cheese whey was lower than that of acid whey obtained from raw skimmilk. This may be due to pasteurization (73°C, 15 sec) of milk used for cheese manufacturing. 3. Electrodialyzed and sweet whey powders a. IgG ac t iv i ty and protein content IgG ac t iv i ty and protein content of reconstituted electrodialyzed (ED) whey and sweet whey (after reconstitution to 6.5% total sol ids solution) are compared to those of l iquid cheese whey (Table 10). Similar values for % protein were obtained for l iquid whey and the reconstituted wheys. However, the IgG act iv i ty was much lower for the reconstituted wheys (especial ly the sweet whey) than 1iquid cheese whey. 91 1 2 3 4 5 6 7 8 w c «Pi ff C . h e d d , a r C h e e s e w h e y a n d fractions obtained by MCIC on Sepharose 6B treatment. Lane l , control cheese whey; Lanes 2, 3 and 4 are unbound f rac t ion; Lanes 5 and 7, f i r s t eluted peak; Lanes 6 and 8 second ? ™ t 6 d , P K a i ; - L F * l a c t o f e r r i n ; HC and LC heavy and l ight chains of immunoglobulin, respectively. * . » a m a ur Table 10. IgG and protein contents of reconstituted ED and sweet whey powders compared to l iquid cheese whey. g IgG/lOOg prote in 3 % prote in 0 Liquid cheese whey 4.0 0.88 Sweet whey 1.2 0.84 ED whey 2.2 0.87 a IgG content was determined by R.I.D. b Protein content was determined as Kjeldahl N x 6.38; units for % protein are g protein/100 mL for l iquid cheese whey, and for sweet and ED whey powders after reconstitution to 6.5% (total sol ids) solut ion. 93 b. Effect of pH adjustment Figure 29 shows the SDS-PAGE prof i le of the supernatant and precipitate fractions of fresh l iquid cheese whey after pH adjustment in the pH 4.5 - 8.5 range. No precipitates were observed and only small amounts of pel lets were obtained after centrifugation of the pH adjusted l iquid whey. The precipitates contained trace amounts of B-Lg and an unidentif ied component "C" (presumed to be a casein f rac t ion , from i ts posit ion on the electrophoretic p r o f i l e ) . Figure 30 shows the SDS-PAGE prof i le of the supernatant and precipitate fractions of ED whey at pH 4.5 and 8.2. As indicated, the precipitate at pH 4.5 (lane 3) from ED whey contained a large amount of immunoglobulins, some BSA, B-Lg, a-La and a l l of the LF. At pH 8.2, the precipitate from ED whey (lane 1) contained some LF, BSA, B-Lg and only a small amount of Igs. It was found from immunochemical analyses (RID, Table 11) that the Ig in the precipitates from sweet whey at both pH values were inactive while those in the supernatant fractions showed very low immunochemical a c t i v i t y . L i t t l e difference in IgG ac t iv i ty was observed between the two pH treatments of sweet whey. In fac t , supernatant and precipitate fractions from sweet whey adjusted to varying pH values from 4.5 to 8.5 showed l i t t l e effect of pH, indicating that pH may not be the only factor causing prec ip i ta t ion . Denaturation of the proteins during processing of the whey powder may be the other factor causing prec ip i ta t ion . On the other hand, for ED whey the amount of total precipitate as well as IgG ac t iv i ty in the precipitate was dependent on the pH. SDS-PAGE prof i les show much greater tendency for precipi tat ion of Ig from ED whey at pH 4.5 to 5.5 than at higher pH. 94 Figure 29. SDS-PAGE prof i le of l iquid cheese whey after pH adjustment and centr i fugat ion. Lanes 1, 3, 5, 7, 9 and 11 are the precipitate and Lanes 2, 4, 6, 8, 10 and 12 are the supernatant of samples treated at pH 8.5, 8.0, 7.0, 6.0, 5.0 and 4.5, respectively, LF, l ac to fer r in ; BSA, bovine serum albumin; C, casein. 95 1 2 3 4 5 Figure 30. SDS-PAGE prof i les of fractions from ED whey after pH adjustment and centriguation. Lanes 1 and 3, are precipitate at pH 8.2 and 4.5 respectively; Lanes 2 and 4, are supernatant at pH 8.2 and 4.5 respectively; Lane 5, electrodialyzed whey; LF, l ac to fe r r in , HC and LC, heavy and l ight chains of immunoglobulins, respectively. 96 Table 11. IgG ac t iv i ty of pH 4.5 and pH 8.2 supernatant (S) and precipitate (P) fractions from sweet whey and ED whey a. Concentration, mg/mL IgGD p rote in 0 glg/lOOg protein Sweet whey pH 8.2 P NDd 18.3 NDd S 1.25 77.0 1.6 pH 4.5 P NDd 16.9 NDd S 1.25 70.2 1.8 ED whey pH 8.2 P NDd 12.4 NDd S 5.00 80.6 6.2 pH 4.5 P 1.10 29.9 3.7 S 1.41 57.6 2.5 a No data were obtained for l iquid cheese whey, which showed almost no precipi tat ion over the entire pH 4 - 8 range. b IgG ac t iv i ty of dialyzed fractions (10X concentrated) was determined by RID. c Protein concentration was determined by Kjeldahl N x 6.38. d ND = not detectable. 97. The supernatant fract ion from ED whey at pH 8.2 showed higher immunochemical ac t iv i ty than the precipitated f rac t ion , which showed no detectable IgG ac t iv i ty due to the low amount of total protein as well as IgG. However at pH 4.5, ac t iv i ty was higher in the precipitated fract ion and less ac t iv i ty was found in the supernatant fract ion (Table 11). These results suggest that at pH 4.5 IgG is preferent ia l ly precipitated from the ED whey. Unlike the precipitated Ig from sweet whey which was immunochemically inact ive, precipitated Ig from ED whey remained act ive . c. Isolation of immunoglobulins by controlled pore glass Figure 31 shows the elution prof i les of adsorbed proteins from controlled pore glass (CPG) (10 mL bed volume) chromatography of 250 mL of sweet whey (SW) and EDI whey, respectively. In each case, a single peak was eluted by buffer El (0.1 N acet ic acid pH 2.8 - 2.9 containing 0.5M NaCl) and a single peak was eluted by buffer E2 (0.1 N Tris-HCl pH 9.0 containing 0.5 M NaCl). The size of the f i r s t peak was decreased for EDI and SW compared to ED2, while the size of the second peak remained f a i r l y constant, despite the four- fo ld reduction in volume of applied whey (250 mL in EDI and SW vs. 1 l i t r e in ED2, Figure 31). Electrophoretic analysis of the unbound fract ion of whey from CPG indicated that appreciable quantities of Ig were not adsorbed to the CPG but were eluted in the unbound fract ion (Figure not shown). This suggests that the capacity of CPG for Ig adsorption is not very high. Although the quantity of whey applied in these studies had already been reduced to 250 mL (EDI), compared to 1 l i t r e (ED2), Ig recovery was s t i l l far from quantitat ive. Increasing the CPG bed volume from 10 to 20 mL improved recovery somewhat, but the capacity was s t i l l insuf f ic ient for quantitative removal of Ig from whey. 98 Figure 31. Elution prof i le of adsorbed proteins from CPG (10 mL) treatment of 250 mL of electrodialyzed whey (EDI), 250 mL of sweet whey (SW), and 1 L of ED whey (ED2). Arrows indicate start of elution with El (0.1 N acetic acid pH 2.8 containing 0.5 M NaCl) and E2 (0.1 M T r i s - H C l , pH 9.0 containing 0.5 M NaCl) buffers. 99 d. Isolation of immunoglobulins by metal chelate- interaction chromatography Figure 32 is an elution prof i le of adsorbed proteins from MCIC treatment of 960 mL ED whey reconstituted in water (pH adjusted to 8.2). An elution prof i le s imi lar to that from fresh l iquid cheese whey was obtained. However, the two peaks were broader and smaller than those from fresh l iqu id cheese whey. SDS-PAGE (Figure 33) showed that the f i r s t large peak was composed of BSA, some Ig and P-Lg while the second smaller peak was composed mainly of Igs and a trace amount of BSA (lanes 6 and 7, respect ively) . The prof i les of the unbound whey fractions (Figure 33, lanes 2, 3 and 4) indicated that some Igs were not adsorbed. Even at the early stage with 120 mL of applied whey (lane 2), some BSA, LF, P-Lg and trace amounts of Igs were found in the unbound fractions suggesting that MCIC may have less a f f i n i t y for these proteins from the ED whey than from fresh l iqu id cheese whey. Radial immunodiffusion showed the purity of Igs recovered from ED whey by MCIC were not as good as from fresh cheese whey (Table 8). The f i r s t peak contained about 16% IgG while the second peak was composed of about 28% IgG on a protein bas is . Unbound fractions showed no detectable IgG a c t i v i t y , despite the presence of Ig band in the SDS-PAGE prof i le (Figure 33). The elution prof i le of adsorbed proteins from MCIC treatment of 720 mL sweet whey reconstituted in start ing buffer at pH 8.2 is shown in Figure 32. Possibly two peaks were eluted, with the second peak appearing as a shoulder from the f i r s t larger peak. Both the large and the shoulder peak were composed of Igs, BSA, LF, some p-Lg and a -La , based on SDS-PAGE (Figure 34, lanes 10 and 11 respect ively) . The difference between the two peaks was that the shoulder peak (lane 11) contained less P-Lg and more a-La than the larger peak (lane 10). The wash fractions (lanes 6, 7, 8 and 9) were composed mainly of p-Lg, a-La, some BSA and trace amounts of LF and Igs. Unbound fractions (lanes 3, 4, 5) 100 Figure 32. Elution pro f i l e of adsorbed proteins from MCIC on Sepharose 6B treatment of 960 mL electrodialyzed whey (EDW) and 720 mL sweet whey (SW) powders reconstituted in water, using l inear gradient elution of 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2 to 2.8. Flow rate was 0.8 mL/min. 1 and 2 are fractions obtained. 101 1 2 3 4 5 6 7 55 = = - - - BSA • « r HC •••• • mm Figure 33. SDS-PAGE prof i les of ED whey and fractions obtained by MCIC on Sepharose 6B treatment. Lane 1, control untreated whey; Lanes 2, 3 and 4, are unbound f ract ions; Lane 5, wash f rac t ion; Lane 6, f i r s t eluted peak; Lane 7, second eluted peak; BSA, bovine serum albumin; HC, heavy chain of immunoglobulins. 102 1 2 3 4 5 6 7 8 9 10 11 12 Figure 34. SDS-PAGE prof i les of sweet whey and fractions obtained by MCIC on Sepharose 6B treatment. Lanes 1, 12, control untreated whey; Lane 2 precip i ta te; Lanes 3, 4 and 5; unbound f ract ions; Lanes 6, 7, 8 and 9, wash f ract ions; Lane 10, f i r s t eluted peak; Lane 11, second (shoulder) eluted peak; BSA, bovine serum albumin. 103 were composed mainly of p-Lg and a-La with some BSA, LF and Ig at a l l stages of whey appl icat ion, with the exception that at the early stage there was re la t ive ly less p-Lg and a-La (lane 3). Precipi tat ion occurred in the sweet whey which was made up in the starting buffer (pH 8.2 containing 0.5 M NaCl) pr ior to application to MCIC. Figure 34 (lane 2) shows the pro f i l e of the precipitate containing a l l major whey proteins including Ig. Another problem that arose during the chromatography of both ED and sweet reconstituted wheys was the build up of pressure. The Sepharose gel became t ight ly packed during whey appl icat ion, shrinking the bed height from 9.0 cm to 8.0 cm. The bed height increased and swelled back only s l igh t ly during the subsequent washing and eluting stage. Unbound proteins from l iquid whey application were normally washed from the column with start ing buffer of half the volume of applied whey to obtain effluent A28O o f ° « 3 o r l ess ; whereas with sweet whey i t required almost twice the volume of applied whey in order to reduce A28O t 0 ° « 9 « This indicates that sweet whey proteins have less a f f i n i t y to bind to the Cu-loaded column than l iqu id cheese whey and can easily be removed during the washing step. IgG ac t iv i ty of ED and sweet whey powders were much lower than l iquid Cheddar cheese whey. Adjustment of the pH of reconstituted ED and sweet whey powders resulted in large amounts of precip i ta te , par t icu lar ly under acidic condit ions. The precipitates contained large quantities of LF and Ig, as well as other proteins. Even at a higher pH (8.2), where no precipitat ion was observed unless the whey was centrifuged, the reconstituted ED and sweet whey samples had a tendency to clog the MCIC columns. These results suggest that these powders ( i . e . , ED and sweet whey) are not good start ing materials for isolat ion of active Ig. It is possible that some process during the manufacture of these powders (e .g. heating or drying) caused 104 denaturation and aggregation of the whey proteins. It is therefore recommended that either l iqu id whey or powders produced by a milder process should be used. C. BINDING CAPACITY AND RECOVERY OF IMMUNOGLOBULINS FROM METAL CHELATE-INTER-ACTION CHROMATOGRAPHY COLUMN Crude Ig (0.3%) separated from colostral whey by ammonium sulfate precipi tat ion was passed through a 2.2 mL Cu-loaded Sepharose 6B column as shown in Figure 35. It was found that the capacity of the column was 101 mg Ig/mL column. The proportion of immunoglobulin eluted from the column under acidic conditions was 94% of the amount bound to the column. Figure 35 also shows that the capacity of 62 mg Ig/mL of Cu-Loaded IDA-BGE Sephacryl S-300 column was lower than that of the Sepharose 6B counterpart. However, the proportion of immunoglobulins eluted from this column was 100% of the amount adsorbed to the column. Thus, the binding capacity for bovine Ig appears to f a l l in the same range as that reported by Lonnerdal et a l . (1977) for human lac to fer r in , which was 70 mg LF/mL gel containing 50 umole copper ions. Using a 2.5 X 9 cm column containing a 50 mL bed volume of chelating Sepharose g e l , approximately half (25 mL) of which was copper-loaded ge l , near-quantitative recovery of IgG was obtained in the peak containing Ig, from application of 750 mL of l iquid whey (6.5% total s o l i d s , containing 250 mg IgG). Radial immunodiffusion indicated that peak 1 and peak 2 contained 6 and 200 mg IgG respectively, while the IgG content of the unbound fraction was not detectable (below the lower detection l imit of R. I .D.) . The recovery of IgG was 82.2% for l iquid whey. The purity of IgG was also increased from about 4% in cheese whey to 53% in the IgG rich fraction (F2). These results indicate that MCIC capacity for Ig isolat ion from cheese whey is at least 200 mg IgG per 25 mL of copper-loaded gel or about 8 mg IgG/mL copper-loaded g e l ; thus, approximately 105 0 100 200 300 400 500 ELUTION VOLUME, ml Figure 35. Saturation point for adsorption of crude Ig (prepared from colostrum by ammonium sulfate method) on Cu-loaded IDA-BGE Sepharose 6B (SROSE) and Sephacryl S-300 (SACRYL). 0.3% crude Ig was passed through a 10 mL column (7.0 x 1.4 cm) equil ibrated with 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2. W, washing with the start ing buffers; E, elution with 0.05 M Tris-acetate /0.5 M NaCl, pH 4.0. The flow rate was 20 mL/hr. 106 1 litre of whey could be treated with about 25 mL of copper-loaded g e l . The discrepancy between the apparent capacity of MCIC for crude Ig (101 mg/mL) and for Ig in whey proteins may be due to the presence of other proteins in cheese whey. In addition to isolat ion of Ig, both LF and BSA from cheese whey were also bound to the copper-loaded g e l . Binding of these two proteins may have lowered the apparent binding capacity for Ig from whey. However, LF and BSA are probably less strongly bound to the immobilized copper than Ig, since they were eluted before Ig during pH gradient e lut ion. D. ANTI-LIPOPOLYSACCHARIDE ACTIVITY OF IMMUNOGLOBULINS RICH FRACTION Figure 36 shows the anti-1ipopolysaccharide ac t iv i ty of an IgG rich fraction isolated from Cheddar cheese whey by MCIC method. Immunoglobulins isolated from cheese whey recognized and bound to LPS extracted from col i , S^ typhimurium and B^ parapertussis. Among the three 0 antigens, immunoglobulins showed higher recognition and binding a b i l i t y to the LPS isolated from E^ col i indicating that this antigen is f a i r l y common to the dairy cow's from which the wheys were produced. Lower recognition was obtained for LPS isolated from S .^ typhimurium. The isolated Ig showed binding a b i l i t y to LPS isolated from B± parapertussis. LPS, which is serological ly cal led as 0 antigen and pharmacologically known as endotoxins, are the integral components of the outer membrane of Gram-negative bacter ia . They contain protein, l ip ids and 1 ipopolysaccharides. Potential ly a l l of these components are more or less exposed on the ce l l surface and thus can interact with l iv ing ce l ls or substances in the environment. Some proteins form pores through which substances can be exchanged across the cel l wa l l . Other proteins part icipate in iron scavenging or ce l l adhesion. With respect to pathogenicity of Gram-negative bacter ia , the cel l wall 107 2.0 Figure 36. Anti-1ipopolysaccharide ac t iv i ty of Ig isolated from cheese whey by MCIC method. • - • , g± co l i LPS; • - • S .^ typhimurium LPS; O O , EL parapertussis LPS. 108 1ipopolysaccharides are not only dominant antigens but also mediators of a great many biological a c t i v i t i e s (Jann and Jann, 1985, Rietschel et a l . , 1982). Even though i t is hard to speculate on what follows the binding of Ig with LPS, the binding may interfere with the process by which bacteria adhere to and colonize the intest inal l in ing (Packard, 1982). The interaction of Ig with LPS may also disturb the biological processes which are involved in the transport of materials across the bacterial cel l wal l . E. LACTOPEROXIDASE CONTENT OF LACTOFERRIN RICH FRACTION The lactoferr in r ich fract ion obtained by MCIC treatment was yellowish in color and became greenish on freezing. The color was thought to be due to the presence of lactoperoxidase which constitutes about 1% of the whey proteins. The color of lactoperoxidase is due to the iron content ( i t contains 0.071% iron which represents one atom of iron per molecule) (Paul and Ohlsson, 1985). The lactoperoxidase containing fraction lost i ts. color when dialyzed against 0.1 M c i t r i c acid for 36 hr. Moreover, the lactoperoxidase assay indicated the presence of less than 2% lactoperoxidase in LF f rac t ion . The presence of lactoperoxidase in the LF rich fract ion would give extra antimicrobial ac t iv i ty to the isolated fraction (Pruitt and Tenovuo, 1985). F. IDENTIFICATION OF GLYCOPROTEINS IN LACTOFERRIN-RICH FRACTION Figure 37 shows the SDS-PAGE of cheese whey, the F l fract ion obtained from MCIC column, standard lactoferr in (LF) and lactoperoxidase (LP). Samples were stained with Coomassie B r i l l i a n t Blue (Figure 37A) as well as periodic ac id -Schi f f (PAS) stain (Figure 37B). Using Coomassie s ta in , LF rich fraction was shown to contain Ig, LP and lac to fer r in . According to the SDS-PAGE of 109 Figure 37. SDS-PAGE of whey proteins (1), lactoferr in r ich fraction (2), lactoferr in (3) and lactoperoxidase (4). (A) stained with Commassie Br i l l an t Blue and (B) stained with periodic acid Schif f (PAS). HC and LC heavy and l ight chains, respectively. 110 standard LF and LP, lactoferr in has s l igh t ly higher molecular weight than LP. PAS technique, which was applied to detect glycoproteins following electrophoresis on SDS-PAGE, indicated that LF, LP and the heavy chain of Ig contained covalently bound carbohydrate. This staining method also indicated that the majority of carbohydrate in Ig was located in the heavy chains of the molecule. G. ISOELECTRIC POINTS OF LACTOFERRIN AND IMMUNOGLOBULINS RICH FRACTIONS The isoe lec t r ic point of fractions obtained by MCIC treatment of cheese whey were found to be in the range of 5.2-6.6 which covered the isoe lect r ic points of standard lac to fe r r in , 6.0, and immunoglobulins 5.5-6.8 (Josephson et a l . , 1972). However, the isoe lec t r ic point of lactoperoxidase was 8.63 which was lower than the reported data of 9.16 - 9.8 (Righetti and Caravaggio, 1976). H. HISTIDINE MODIFICATION AND METAL CHELATE-INTERACTION CHROMATOGRAPHY Among a l l of the amino acids comprising proteins, Rassi & Horvath (1986) found that h ist id ine and cysteine gave the highest retention factors when passed through a Cu-IDA column. However, studying the retention behavior of free amino acids might not represent the real behavior of that amino acid when i t is found in protein. The role of h ist id ine in immunoglobulins and the lactoferr in rich fract ion isolated by MCIC treatment was investigated using MCIC. When control Ig was applied to MCIC column, almost no protein was eluted in the washing step (Figure 38), indicating that a l l the sample applied was adsorbed to the column. This fraction was subsequently eluted with 0.01 M imidazole solut ion. However, modification of h ist id ine groups of Ig by diethyl pyrocarbonate (DEP-Ig) greatly inhibited the interaction with the copper ion. Most of histidine-modified Ig was found in the washing solut ion, while only a small amount of protein was Ill Figure 38. Elution prof i les of control (Ig) and diethyl pyrocarbonate treated immunoglobulins (DEP Ig). Samples (30 mg/5 mL 0.05M Tris-acetate containing 0.5 M NaCl, pH 8.2) were applied to the column (1.4 x 7.0 cm) and washed (W) with the start ing buffer then eluted (E) with 0.01 M imidazole. Flow rate was 30 mL/hr. 112 adsorbed. These data strongly suggest the involvement of h ist id ine groups of Ig in the interaction with copper ion immobilized on agarose. The small amount of DEP-Ig adsorbed on the column may suggest the involvement of other amino acids in the interact ion, especial ly cysteine which is considered to be the second major force contributing to the interaction (Rassi & Horvath, 1986). Figure 39 shows the elution prof i les of the lactoferr in r ich fraction obtained from cheese whey before and after h ist id ine modif icat ion. Before hist id ine modif icat ion, a l l proteins in the sample applied were adsorbed to the column of MCIC, however, after blocking hist id ine groups most of the proteins applied were desorbed by washing the column with the start ing buffer. Appreciable amounts of protein were eluted with 0.01 M imidazole. This may indicate that some other forces are involved in the interaction or the modification process of h ist id ine was not complete (since this fraction was mixture of Ig, LF and BSA, i t was d i f f i c u l t to calculate mole modified hist idine per mole protein) . I. SEPARATION OF HEAVY AND LIGHT CHAINS OF IMMUNOGLOBULINS A typical elution pattern of reduced and alkylated Ig from Sephadex G-75 column is shown in Figure 40. The f i r s t peak which eluted at the void volume represents the heavy chain while the second peak consists of l ight chains. The third peak is eluted at a volume corresponding to the total bed volume, and therefore represents small molecules, spec i f i ca l l y the reducing and alkylating agents. The homogeneity of the heavy and l ight chain preparation was demonstrated by SDS-PAGE (Figure 41). Better resolution of heavy and l ight chains was obtained when Ultrogel ACA 54 was used instead of Sephadex G-75 as indicated in Figure 42. Based on the total absorbance unit ca lculat ion, the percentage of heavy chains ( f i r s t peak) and the l ight chains (second peak) obtained from these figures were 70-75% and 25-30% respectively. These values 113 Figure 39. Elution prof i les of Fl-MCIC fract ion before (Fl) and after diethyl pyrocarbonate treatment (DEP F l ) . Samples (30 mg/5 mL 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2) were applied to the column (1.4 x 7.0 cm) and washed (W) with the start ing buffer then eluted (E) with 0.01 M imidazole. Flow rate was 30 mL/h. 114 Figure 40. Elution prof i les of reduced and alkylated heavy and l ight chains of immunoglobulin on Sephadex G-75 eluted with 1 M propionic ac id . 1 and 2 are fractions obtained. 115 1 2 3 4 Figure 41. SDS-PAGE prof i les of heavy and l ight chains of Immunoglobulins Isolated by gel f i l t r a t i o n . Lanes 1 and 2, crude Immunoglobulins; Lanes 3 and 4, Immunoglobulin l ight and heavy chains, respectively obtained from Figure 40. 116 3.5 100 200 300 ELUTION VOLUME, ml Figure 42. Elution prof i le of reduced and alkylated heavy and l ight chains of Ig on Ultrogel ACA 54 eluted with 0.1 M Tris-HCl buffer containing 4 M Guanidine-HCl and 1 mM iodoacetamide, pH 8.2. 1 and 2 are fractions obtained. 117 are within the range of reported data for Ig of di f ferent species (Fleischman et a l . , 1962; Small and Lamm, 1966). J . SEPARATION OF LACTOFERRIN AND LACTOPEROXIDASE IN LACTOFERRIN RICH FRACTION In addition to the gel f i l t r a t i o n process used to isolate the LF from LF rich f rac t ion , several attempts were made in order to obtain better resolution of LF from other protein contaminants. 1. Gel f i l t r a t i o n method Figure 43 shows the elution pattern of lactoferr in r ich fract ion (peak 1) from MCIC of bovine acid whey on Sephacryl S-300 column. Two major peaks were obtained, the f i r s t peak was colorless while the second peak was yel lowish. SDS-PAGE analysis (Figure 26) indicated that the f i r s t peak consisted mainly of immunoglobulins while the second peak contained predominantly l ac to fe r r in . 2. Stepwise pH elution Figure 44 represents a stepwise elution process for separating LF rich fract ion on MCIC column loaded with copper. Elution with the start ing buffer, 0.5 M NaCl in 0.05 M Tris-acetate at pH 7 and 6 did not remove any of the bound proteins from the column; however, elution with the same buffer at pH 5 removed some of the bovine serum albumin. Further elution with the same buffer at pH 4.0 gave two peaks, i . e . , LF and Ig, which were not wel1-separated. Using pH 5 as a cleaning step for eluting bovine serum albumin, then elution with a pH gradient (5.0-2.8) s l igh t ly improved the separation of LF and Ig as shown in Figure 45. 118 1.4 0 ' 0 60 120 180 240 300 ELUTION VOLUME, ml Figure 43. Sephacryl S-300 column chromatography of lactoferr in rich fraction obtained by MCIC of acid whey. 100 mg sample was applied to Sephacryl column (83 x 2.5 cm) and eluted with 0.05 M potassium phosphate buffer, pH 7.4 containing 0.01 M NaCl. 1 and 2, are fractions obtained. The flow rate was 30 mL/hr. 119 280 560 840 1260 ELUTION VOLUME, ml Figure 44. Stepwise elution pro f i l e of acid whey on MCIC eluted by decreasing pH values. Arrows indicate pHs 7, 6, 5 and 4 of 0.05 M Tris-acetate containing 0.5 M NaCl. 1 and 2 are fractions obtained. 120 2.5 0 0 70 140 210 280 350 ELUTION VOLUME, ml Figure 45. Elution prof i le of bound proteins of acid whey on MCIC column, eluted (E) by using pH gradient (5-2.8) of 0.05 M Tris-acetate containing 0.5 M NaCl. 1 and 2 are fractions obtained. 121 3. Imidazole gradient elution Fraction 1 obtained by MCIC process contained Ig and BSA in addition to LF. This fract ion was pooled and rechromatographed on MCIC column and eluted with a 0-0.01 M imidazole gradient. After washing off the unbound proteins two wel1-separated peaks were obtained (Figure 46). The f i r s t peak contained mainly Ig, while the second peak was mainly lactoperoxidase as indicated by SDS-PAGE (Figure 47). However, subsequent elution with the start ing buffer at pH 2.8 gave an extra peak which was highly pur i f ied lactoferr in according to SDS-PAGE analys is . It i s , therefore, possible to separate these three b io logica l ly active proteins. The results of this study demonstrate that cheese whey can be a rel iable source for extracting immunoglobulins and lac to fe r r in . Of the absorption chromatography techniques investigated, metal chelate- interaction chromatography is the best method as i t is simple in operation for separating b io logica l ly important immunoglobulins, lactoferr in and lactoperoxidase. Furthermore, this method has advantages of high capacity, quantitative recovery without detectable damage to the immunological ac t iv i ty of the proteins and easy regeneration. Based on anti l ipopolysaccharide ac t iv i ty of isolated immunoglobulins and the well known bacter iostat ic ac t iv i ty of l ac to fe r r in , (Packard, 1982) the separated bioactive proteins may, therefore, be useful in f o r t i f i c a t i o n of infant formulae or infant feeding. 122 Figure 46. Elution prof i le of lactoferr in r ich fract ion on MCIC column. E l , elution with l inear gradient of 0-10 mM imidazole solution ( . . . . ) ; E2, elution with 0.05 M Tris-acetate containing 0.5 M NaCl, pH 2.8. 1, 2 and 3 are fractions obtained. 123 6 5 4 3 2 1 Figure 47. SDS-PAGE prof i les of fract ions obtained from Figure 46. Lane 1, whey proteins; Lane 2, control Fl-MCIC; Lane 3, unbound f rac t ion; Lanes 4, 5 and 6 are peak 1, 2 and 3 of Figure 46; LF, l ac to fe r r in , LP, lactoperoxidase. PART IV METAL CHELATE INTERACTION CHROMATOGRAPHY OF SKIMMILK 125 In skimmilk there are two major fractions of milk protein, caseins and whey or serum proteins, which are the pH 4.6 insoluble and soluble f ract ions, respectively. In Part III, immunoglobulins and lactoferr in were isolated quite easi ly and e f f i c i e n t l y from whey proteins by Cu-chelate chromatographic supports; however, i f skimmilk can be used as a start ing material , broader u t i l i za t ion of these anti-microbial compounds may be feas ib le . Two types of buffer were used for MCIC column chromatography in this study. A. MCIC WITH TRIS-ACETATE BUFFER The p o s s i b i l i t y of u t i l i z i n g skimmilk for d i rec t ly recovering immunoglobulins and lactoferr in was investigated. Figure 48 shows the elution prof i le of skimmilk before and after 50% di lut ion with 0.05 M Tr is-acetate/0 .5 M NaCl, pH 8.2. After washing the unbound proteins with the start ing buffer, the adsorbed proteins were eluted with 0.05 M acetate-Tr is /0 .5 M NaCl pH 4.0. However, the amount of proteins eluted under ac id ic conditions (peak 1) was small , and the flow rate became quite slow indicating precipi tat ion of casein result ing in clogging of the column. This behaviour was observed regardless of whether skimmilk was di luted or undiluted. Subsequent elution with 0.01 M imidazole recovered the bound proteins which appeared as the main fraction (peak 2) in the p r o f i l e s . SDS-PAGE analysis (Figure 49) indicated that proteins eluted in the washing step were immunoglobulin, l ac to fe r r in , a-lactalbumin, B-lactoglobulin and casein, while proteins eluted under ac id ic conditions were mainly immunoglobulin and lac to fer r in . The fraction eluted with 0.01 M imidazole, however, was mainly casein. Table 12 shows the IgG distr ibut ion of fractions obtained at different stages of the isolat ion process of IgG from skimmilk. Immunochemical analysis showed that the majority of active IgG was present in the fract ion eluted at 126 0 70 140 210 280 ELUTION VOLUME, ml Figure 48. Elution prof i le of skimmilk on MCIC column. 100 mL skimmilk undiluted (SM) or 50% diluted (DSM) with 0.05 M Tr is-acetate /0 .5 M NaCl) was passed through Cu-loaded Sepharose 6B (1.4 x 7.0 cm), and washed (W) with same buffer. E l , elution with the same buffer at pH 4.0; E2, elution with 0.01 M imidazole solut ion. 1 and 2 are eluted f ract ions. The flow rate was 21 mL/hr. 127 Figure 49. SDS-PAGE of fractions obtained in Figure 48. Lanes 1 and 2, skimmilk; Lane 3, unbound skimmilk to MCIC column; Lane 4, washing f rac t ion ; Lanes 5 and 6, are peak 1 and 2, respectively; Lane 7, standard IgG; Lane 8, a-casein. 128 Table 12. IgG content of di f ferent stages of the isolat ion of IgG from skimmilk on MCIC column. Sample Prote in 3 Cone, of selected fract ion mg/mL IgGD Cone. mg/mL IgG Purity % Skimmilk (control) 37.8 0.562 1.49 Unbound skimmilk 18.0 0.246 1.36 Washing fract ion 4.0 0.828 20.70 Peak 1 (Figure 48) 1.08 0.911 84.35 Peak 2 (Figure 48) 19.6 0.272 1.38 a Determined by Bio-Mississauga, Ont.) D Determined by R.I.D. •Rad Protein Assay Kit (Bio-Rad Laboratories, 129 acid ic pH (pH 4.0) . This fraction was more than 84% pure. However, the amount of IgG bound to the column was quite small (less than 10%) as compared to the unbound fract ion of skimmilk. More than 20% pure IgG was detected in the washing fract ion indicating that IgG of skimmilk bound rather weakly to the column. By eluting the strongly bound material with a strongly competing electron-donor solution (0.01 M imidazole) the fract ion obtained contained almost no IgG but was rich in casein fractions (Figure 49). These results suggest that there was a competition between IgG and caseins to bind copper ions linked to agarose, and the casein fract ion bound more strongly than immunoglobulins to the column under these condit ions. B. MCIC WITH PHOSPHATE BUFFER An attempt was made to f ind conditions under which casein could be eluted while retaining immunoglobulins and lactoferr in on the column. Figure 50 shows the elution pro f i l e of a mixture of skimmilk and immunoglobulins on copper loaded column of MCIC after equi l ibrat ion with 0.02 M phosphate buffer containing 0.5 M NaCl, pH 7.0. After washing with the start ing phosphate buffer, the unbound proteins were removed. The bound proteins were then eluted with 0.01 M imidazole and 0.05 M Tr is-acetate /0 .5 M NaCl, pH 3.8 to obtain Fl and F2, respectively. Electrophoretic analysis (Figure 51) indicated that the unbound proteins (turbid fraction) (lane 3) were casein f ract ions, while proteins eluted with 0.01 M imidazole (lane 4) were immunoglobulins. Peak 2 eluted with Tr is-acetate containing 0.5 M NaCl, pH 3.0 was too small to detect by SDS-PAGE. Figure 50 shows the elution prof i le of a mixture of Ig and lactoferr in in the presence of caseins in skimmilk, under the same conditions on MCIC column. A similar pattern to that of skimmilk-Ig mixture was obtained; however, the 130 Figure 50. Elution prof i les of skimmilk (SM), Ig and LF mixture on MCIC column. 60 mg of Ig and LF was mixed with 1 mL skimmilk (SM+Ig+LF) and 30 mg Ig was mixed with 1 mL skimmilk (SM+Ig) and passed through MCIC column (1.4 x 7.0 cm). W, washing with 0.02 M phosphate buffer containing 0.5 M NaCl, pH 7.0; E l , elution with 0.01 M imidazole; E2, elution with Tr is-acetate containing 0.5 M NaCl, pH 3.0. 131 10 9 8 7 6 5 4 3 2 1 Figure 51. SDS-PAGE of fractions obtained in Figure 50. Lane 1, skimmilk; Lane 2, skimmilk and Ig mixture; Lanes 3 and 4 are unbound and peak 1 of SM-Ig mixture appl icat ion, respectively; Lane 5, SM-Ig-LF mixture; Lanes 6 and 7 are unbound and peak 1 of SM-Ig-LF mixture appl icat ion, respectively; Lane 8, immunoglobulins; Lane 9, lactoferr in and Lane 10, a-casein. 132 amount of protein bound to the column was increased. SDS-PAGE pro f i l e (Figure 51) indicated that the unbound fract ion (lane 6) was basica l ly casein while proteins eluted with 0.01 M imidazole were a mixture of Ig and lactoferr in (lane 7). The bound protein mixture (Ig and LF) may be separated by using gel f i l t r a t i o n or 0-0.01 M imidazole gradient. Results suggest that the type of ions in the buffer used for i n i t i a l column equi l ibrat ion and washing had a great influence on whether or not proteins were unbound or bound to the MCIC column. Since caseins are c lass i f i ed as phosphoproteins (Whitney et a l . , 1976), i t is believed that using phosphate buffer in the equi l ibrat ion step of the column can form a complex with the copper ion (with di f ferent color from that formed with T r is -ace t ic acid buffer) which may prevent the phosphoproteins from binding to the column. These results also suggest the involvement of phosphoserine groups in the interaction with the copper ions. C. MECHANISM OF CASEIN-METAL INTERACTION The pr inc ip le of protein separation by MCIC process l i es in the different a f f i n i t i e s of proteins to bind to immobilized metal ions. It is suggested that this binding is dependent on the a v a i l a b i l i t y of h is t id ine , cysteine and tryptophan residues of the proteins to form stable coordination complexes with metal ions (Sulkowski, 1985). In general, adsorption of protein to MCIC is performed at a s l igh t ly alkal ine pH with high ionic strength solutions to decrease non-specif ic e lectrostat ic interact ions. Elution is commonly accomplished by lowering the pH, which reverses protein coordination to the metal-chelate and results in protein displacement. A l ternat ive ly , a competing electron donor as a mild chelating agent ( e . g . , imidazole) or a strong chelating agent (e .g. EDTA) may be used to purge bound proteins. Since hist id ine has been 133 suggested to exhibit the strongest retention factor on a copper chelate column at pH 6.0, (Rassi and Horvath, 1986), the elution prof i le of histidine-modified casein fract ions was investigated. 1. q-Casein a-Casein represents more than 62.5% of total casein fractions in cow's milk and is composed of a s and K-casein in ratio of 4:1 (Whitney et a l . , 1976). Figure 52 shows the elution prof i le of control a-casein and hist idine-modif ied a-casein. After loading control a-casein on Cu-chelate gel and washing with the start ing alkal ine pH buffer, only 10.7% of total applied protein could be eluted under these condit ions. The rest of protein (89.2%) was displaced by using a mild chelating agent i . e . 0.01M imidazole (Table 13). However, blocking 3.7 hist id ine residues by diethyl pyrocarbonate out of the total of four hist id ines per mole a-casein (Webb et a l , 1974) increased the amount of unbound protein from 10.7% to 82.2% (Table 13) indicating the involvement of h ist id ine groups in the interaction with the copper ion. Less than 18% of the protein was bound and could be eluted with 0.01 M imidazole, which might indicate the involvement of other amino acids i . e . , Trp, Cys, Tyr in the interact ion. 2. a^ | - and B-casein a s i -Casein is a subfraction of the whole casein which represents a s l igh t ly less than 50% of the total casein. Figure 53 indicates the elution prof i les of a s i -casein and hist id ine blocked a s i - case in . Without modification of h ist id ine residues of a s i - case in , only a small amount (6.1%) of the protein applied was eluted off the column in the washing step and 93.9% of protein interacted with the copper under alkal ine pH and was subsequently 134 Figure 52. Metal chelate interaction chromatography of a-casein. 3 mL of protein (10 mg/mL) before (a-CAS) and after diethylpyrocarborate modification (DEP-a-CAS) equil ibrated with 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2 and applied to copper chelate Sepharose 6B (1.4 x 7.0 cm). W, washing with the same equi l ibrat ing buffer; E, elution with 0.01 M imidazole. Flow rate was 30 mL/hr. 135 Table 13. Binding of casein f rac t ions 3 to MCIC column before and after modification of h ist id ine groups. Proteins 0 Washing step (W) % Eluting step (E) % Control a-casein 10.7 89.2 DEP-a-casein 82.2 17.8 Control a s i - c a s e i n 6.1 93.9 DEP-a s i -casein 94.3 5.7 Control B-casein 13.6 86.4 DEP-8-casein 54.7 45.3 Control polymer K-casein 87.9 12.1 DEP-Polymer-K-casei n 94.3 5.7 SSS-K-Casein 38.5 61.5 DEP-SSS-K-Casein 98.9 1.1 a Calculated based on total D See Figures 52, 53, 54 and absorbance units 55 for abbreviation identity 136 Figure 53. Metal chelate interaction chromatography of a s i - c a s e i n before (a s-CAS) and after diethylpyrocarbonate modification (DEP a s -CAS) . See Figure 52 for conditions of separation. 137 eluted with the eluting solution (Table 13). However, blocking 4.2 hist idine residues by DEP out of a total of 5 hist id ine residues per mole a s i - case in (Whitney et a l , 1976) destroyed the protein's a b i l i t y to bind to copper ion immobilized on agarose. Based on the calculat ion of total UV absorbance, 94.3% of applied protein was recovered in the washing step and a small amount of the applied protein (5.7%) was bound and subsequently eluted with 0.01 M imidazole. The second major protein of bovine casein is p-casein which represents 30% of total casein fract ion (Whitney et a l . , 1976). Figure 54 shows the elution prof i les of p-casein before and after h ist id ine modification with DEP. Compared to 13.6% of unbound control p-casein, 54.7% was unbound after hist idine modification (Table 13). This indicates that even though 4.4 h ist id ine residues were modified out of the total 5 h ist id ine residues per mole p-casein (Whitney et a l , 1976), 45.3% of protein applied was bound to the copper ion and subsequently eluted with 0.01 M imidazole solut ion. The reason for this high binding rate after h ist id ine modification compared to other casein fractions (Table 13) may be due to the temperature-dependent associa t ion-d issocia t ion properties of P-casein. 3. K-casein K-casein which represents 12.5% of the total casein in cow's milk occurs in the form of a mixture of aggregates of ic-caseins held together by intermolecular d isu l f ide bonds (Whitney et a l , 1976). K-Casein prepared by the method of Z i t t l e and Custer (1963) was considered to assume an aggregated form. Figure 55A shows the elution behaviour on Cu-chelate agarose of this preparation with and without hist id ine modif icat ion. The amount of unbound K-casein for control and hist id ine modified K-casein were 87.9% and 94.3% respectively (Table 13). This indicated that the amount of K-casein adsorbed on Cu-chelate and recovered 138 Figure 54. Metal chelate interaction chromatography of p-casein before (p-CAS) and after diethylpyrocarbonate modification (DEP p-CAS). See Figure 52 for conditions of separation. 139 Figure 5 5 . Metal chelate interaction chromatography of ( A ) aggregated K-casein ( K - C A S ) (B) monomer K-caseins (MK-CAS) before and after diethylpyrocarbonate modification ( D E P K - C A S ) . See Figure 5 2 for separation condit ions. 140 by 0.01 M imidazole was quite low for both control K-casein and hist idine modified K - c a s e i n . This behavior of aggregated K-casein on the MCIC column may be due to i ts aggregated structure which probably restr icted the access to metal ions. Whether the aggregated structure of K-casein blocked the copper-K-casein interaction was determined. Figure 55B shows the elution prof i les of K-casein monomers formed by reducing the d isu l f ide bonds and blocking them with sodium tetrathionate. The amount of bound protein for reduced K-casein (MK-CAS) was more than 61% of the total protein applied compared to 12.1% of aggregated K-casein adsorbed onto the same column (Figure 55B). Modification of 2.7 h ist id ine residues out of 3 hist id ine residues per mole monomer K-casein (Whitney et a l . , 1976) decreased the amount of bound modified K-casein to 1.1% indicating the involvement of h ist id ine residues in the interaction of K-casein with copper ions. This suggests that d isu l f ide bonds of the aggregated K-casein had l i t t l e or no effect on the interaction with metal ions. In conclusion, use of skimmilk d i rec t ly as a start ing material for Ig and lactoferr in separation is feas ib le . Using Tris-acetate buffer as an equi l ibrat ing buffer induced competition between Ig and caseins to bind copper ions and decreased the capacity of the MCIC column in binding of immunoglobulins. However, using phosphate buffer as an equi l ibrat ing buffer prevented phosphoproteins from binding to the column and allowed them to be col lected in the unbound fract ion while immunoglobulins and lactoferr in are retained on the column. Chemical modification studies with Tris-acetate buffer equil ibrated MCIC columns, indicated the involvement of h ist idyl residues of some casein fractions in the interaction with copper ions. SEPARATION OF IMMUNOGLOBULINS AND TRANSFERRIN FROM BLOOD SERUM AND PLASMA BY METAL CHELATE INTERACTION CHROMATOGRAPHY 142 It was reported that in 1982, 800,000 tonnes of blood were dumped into sewage systems throughout Europe. This is equivalent to 140,000 tonnes of potent ia l ly valuable protein that was l i t e r a l l y flushed down the drain (Alexander, 1984). If one could extract these proteins they could be used for food processing and animal feeding. However, most of the available methods involve select ive precipi tat ion methods which are batch processes and thus d i f f i c u l t to mechanize. The p o s s i b i l i t y of using MCIC method to extract immunoglobulins and transferr in from blood plasma and serum was assessed in this part of the thes is . A. METAL CHELATE INTERACTION CHROMATOGRAPHY 1. Blood serum on Cu-loaded MCIC When blood serum (obtained by incubating blood samples overnight at 5°C) in 0.05 M Tr is-acetate /0 .5 M NaCl buffer, pH 8.2, was applied to a Cu-loaded IDA-BGE Sepharose 6B column (Figure 56), the major portion of the blood proteins (mainly albumin) did not bind and were recovered when the column was washed with the start ing buffer. The bound proteins were eluted with the pH gradient buffer (Fl) and ident i f ied to be mainly immunoglobulins. Subsequent elution with 10 mM imidazole (F2) recovered the major portion of transferr in (Figure 57; lanes 5, 6 and 7). Fraction 1 gave a long arc by Immunoelectrophoresis which was similar to that of standard IgG, while fract ion 2 yielded arcs around the well similar to that of standard transferr in and immunoglobulins (Figure 58). 2. Blood serum on MCIC columns loaded with other metal ions Figure 59 represents the elution patterns of blood serum from columns packed with Zn- , Ni - and Co-loaded IDA-BGE Sepharose 6B. The capacity of Zn-loaded column to adsorb protein was found to be higher than that of Ni - and 143 Figure 56. Immobilized copper a f f i n i t y chromatography of blood serum. Blood serum (1 g in 10 mL 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2) was applied to the column (1.4 x 7 cm). The column was washed with the start ing buffer and then eluted with E l , 0.05 M Tris-acetate 0.5 M NaCl, pH 4.0, and with E2, 0.1 M imidazole. The flow rate was 30 mL/hr. Fl and Fz are fractions obtained. 144 11 109 8 7 6 5 4 32 1 — — TF BSA Ig-HC Ig-LC Figure 57. SDS-PAGE prof i les of blood fractions from MCIC on Sepharose 6B column. Sample ident i f ica t ion: Lanes 1, 2 and 3 are F l of Figure 59 from Zn, Ni and Co loaded columns respectively; Lane 4, plasma protein eluted from Cu-loaded column with pH 4 buffer; Lanes 5, 6 and 7 are unbound, F l , and F2 in Figure 56, respectively; Lanes 8, 9 and 10 are standard t ransferr in(TF) , bovine serum albumin (BSA), and immunoglobulins (Ig) respectively; Lane 11, blood plasma. 145 Figure 58. Immunoelectrophoresis of fractions obtained in Figure 56. P, blood plasma; F l and F2 fractions obtained in Figure 56; TF, t ransferr in; Ig, immunoglobulins, BSA, bovine serum albumin; abws, rabbit antibovine whole serum. 146 ELUTION VOLUME, ml Figure 59. Immobilized Zn- , Ni - and Co- a f f in i t y chromatography of blood serum. Blood serum (2 g in 20 mL 0.05 M Tr is-HCl /0 .15 M NaCl, pH 8.0) was applied to the column (2.8 x 8.5 cm). The column was washed with the start ing buffer then eluted (E) with 0.1 M Na-acetate/0.8 M NaCl, pH 4.6. The flow rate was 30 mL/h. 1, is fraction obtained. 147 Co-loaded columns. F l fraction obtained from these columns are compared in Figure 57 (lanes 1, 2 and 3). They 'are composed of Igs, TF and other high molecular weight proteins and appear to contain a small amount of albumin. 3. Blood plasma on MCIC column The plasma supplied had a red hue which was probably due to the partial hemolysis of red blood ce l l s during centrifugation of blood after addition of sodium c i t ra te as anticoagulant. When blood plasma in 0.05 M Tris-acetate buffer, pH 8.2, was applied to a Cu-loaded IDA-BGE Sepharose 6B, the major portion of the proteins eluted with 0.05 M acetate-Tris buffer at pH 4.0 were Igs as indicated by SDS-PAGE, R.I.D. and ELISA analysis (with more than 957. biological a c t i v i t y ) . The upper part of the Cu-loaded column, however, became strongly reddish. To identify this colored material , standard bovine hemoglobin was applied to the column. Figure 60 shows the elution patterns of hemoglobin from Zn- , N i - , Co- and Cu-loaded IDA-BGE Sepharose 6B columns. It was found that 0.1 M acetate/0.8 M NaCl buffer , pH 4.5, 0.1 M acetic ac id / . 5 M NaCl, pH 2.8, or 0.01 M imidazole were not effect ive for eluting the adsorbed hemoglobin from Cu-loaded IDA-BGE Sepharose 6B; however, elution with 50% ethanol was ef fect ive for hemoglobin removal. It is interesting to note that chicken hemoglobin bound to a copper-loaded column was eluted quite easi ly with acidic buffers (our unpublished data) which may indicate structural differences between chicken and bovine hemoglobins. Figure 60 also shows the behavior of hemoglobin towards other metal ions, z inc , cobalt and nickel ions. It was found that hemoglobin was readily eluted by using 0.1 M acetate/0.8 M NaCl buffer, pH 4.5, from the columns loaded with metals other than Cu. These results indicated that hemoglobin-metal interactions are dependent on the kind of metal ions and the source of hemoglobin. 148 Figure 60. Elution prof i les of adsorbed hemoglobin from MCIC columns (1.4 x 7.0 cm) loaded with Zn, N i , Co and Cu. 2 mL of hemoglobin (3 mg/mL in 0.05 M Tr is-HCl /0 .15 M NH4C1, pH 8.0) was applied to the column and washed (W) with 2-3 times bed volumes of the start ing buffer. E l , 0.1 M Na-acetate /0.8 M NaCl, pH 4.5; E4, 50% ethanol. 149 B. IMMUNOCHEMICAL ASSAYS Table 14 compares the IgG contents of di f ferent fractions obtained from MCIC columns loaded with dif ferent metal ions. The copper-loaded column gave the highest IgG purity in F l (Figure 56). However, F2 (Figure 56) also contained some IgG which was evident in the immunoelectrophoretogram (Figure 58). This indicated that not a l l IgG was eluted at pH 4.0 (Figure 56); some IgG subclasses (IgGi, IgG2) may bind more strongly than others. By comparing Zn- , N i - , and Co-loaded columns, F l obtained from the Ni-loaded column gave the highest IgG purity while the Zn-loaded column gave the lowest IgG content. Even though Co-loaded column gave high IgG purity i t is not recommended for the isolat ion of Igs since i ts capacity to bind Igs was low and decreased with repeated use. C. BACTERIOSTATIC ACTIVITY OF BLOOD IMMUNOGLOBULINS AND TRANSFERRIN Figure 61 shows that immunoglobulins and transferr in isolated by MCIC method had inhibitory effects on the growth of E^ col i during the f i r s t 3 hr as compared to the control . This results were in agreement with that reported by Stephens et a l . (1980) who found that the bacter iostat ic ac t iv i ty of IgGi against E i co l i could be enhanced by addition of lactoferr in isolated from human milk. The inhibitory effect of transferr in alone was higher than that of Igs. However, mixing Igs with TF may have a synergist ic effect on the inhibi t ion of Ei c o l i . Transferr in , by virtue of i t s high a f f in i ty for i ron , can retard microbial growth by making this element re la t ive ly unavailable (Harrison, 1985). D. ANTI-LIPOPOLYSACCHARIDE ACTIVITY OF BLOOD IMMUNOGLOBULINS It is well known that antibodies are considered to be the architecture of the immune system (Packard, 1982). Human and animals are defenseless without 150 Table 14. IgG contents* of blood serum or plasma fractions obtained from MCIC column loaded with di f ferent metal ions. Fract ion** * Protein (mg/mL)** IgG (mg/mL) %IgG Cu (Figure 56 Fl) 26.5 26.0 98.1 Cu (Figure 56 F2) 23.3 10.2 43.8 Cu (plasma) 26.0 25.0 96.2 Zn (Figure 59 Fl) 21.6 5.0 23.2 Ni (Figure 59 Fl) 24.6 20.0 81.3 Co (Figure 59 Fl) 25.2 20.0 79.4 * Radial immunodiffusion was used for the determination. * * Bio-Rad Protein Assay Kit (Bio-Rad Laboratories, Mississauga, Ont) was used for the determination. * * * Serum fract ion unless otherwise noted. 151 Figure 61. Bacter iostat ic ac t iv i ty of isolated immunoglobulins and transferrin against E. c o l i . C, control ; TF, transferr in (10 mg/mL); M, mixture of TF (5 mg/mL) and Ig (5 mg/mL); Ig, immunoglobulins (10 mg/mL); CFU, cel l forming unit , 152 them. The key character ist ics of the immune system are s p e c i f i c i t y , memory and a b i l i t y to recognize foreign bodies. Figure 62 shows the ac t iv i ty of bovine IgG isolated from blood by MCIC method toward LPS extracted from c o l i . 5^ typhimurium and EL parapertussis. Blood IgG recognition and binding to LPS from ^ typhimurium and E .^ col i may indicate that dairy cows had experienced these bacterial in fect ions, however, IgG binding to JL parapertussis which causes whopping cough in infants may indicate s imi la r i t i es in the surface exposed antigens to those of Enterobacteriaceae family. The binding of blood IgG with these antigens may interfere with some of the physiological ac t iv i ty or may change the adhesion properties of these bacteria with the intest inal surface and thus prevent, to some extent, the infection caused by these bacteria (Packard, 1982). E. CAPACITY OF MCIC COLUMN FOR TRANSFERRIN Figure 63 shows the A28O elution prof i le from the application of 0.2% transferr in solution to a copper loaded chelating Sepharose 6B column. Saturation of the column was reached at an elution volume of about 184 mL indicating no further protein was adsorbed. At this point, the capacity for TF was calculated to be approximately 167 mg/mL copper-loaded g e l . Thus, the binding capacity for bovine TF appears to be higher than that reported by Lonnerdal et a l . (1977) for human lac to fe r r in , which was 70 mg LF/mL gel containing 50 ymole copper ions. Subsequent elution with buffer at pH 4.0 recovered only 15% of the applied protein; however, the residual TF was recovered by using 0.01 M imidazole as an eluent. This behavior of TF on MCIC column may indicate the biphasic nature of TF. One phase was eluted simply by protonation (acidic pH) and the second phase was eluted by using a stronger eluent. 153 1.0r 10 100 1000 ^9 ig Figure 62. Anti- l ipopolysaccharide ac t iv i ty of blood IgG isolated by metal chelate interaction chromatography method. E_j. col i LPS; • - • , S. typhimurium LPS; O - O , i L parapertussis LPS. 154 Figure 63. Saturation point of adsorption of standard TF on Cu-loaded IDA-BGE Sepharose 6B. 0.2% TF was passed through 10 mL column (7 x 1.4 cm) equil ibrated with 0.05 M Tris-acetate containing 0.5 M NaCl, pH 8.2, W, wash with start ing buffer, E l , elution with 0.05M Tris-acetate containing 0.5M Nacl , pH 4.0; E2, elution with 0.01 M imidazole. 155 F. MECHANISM OF PROTEIN-METAL INTERACTION Porath et a l . , in their pioneering work (1975), postulated that h is t id ine , cysteine and tryptophan residues of a protein were most l i ke ly to form stable coordination bonds with metal ions at a neutral or alkal ine pH. However, there have been no detailed data published on the role of these amino acid residues in the interaction with metal ions. Figure 64 shows the elution prof i le of standard bovine transferr in on Cu-loaded column before and after modification of h ist id ine groups with DEP. By washing the column, with three bed volumes of the start ing alkal ine buffer, no TF bound to the column was eluted; the bound TF was eluted with 0.01 M imidazole. Modification of 16.6 h ist id ine residues out of the total 18 residues per mole TF (Sutton and Jamieson, 1972) almost completely inhibited the interaction of the protein with metal ion. Almost a l l of the protein introduced into the column was restored in the washing buffer without binding to the column. The decreased binding of hist idine-modif ied TF with copper ion would c lear ly indicate the importance of h ist id ine residues in the coordination binding with the metal ions. Similar behavior was observed when hist idine-modif ied immunoglobulins were passed through Cu-chelate g e l . These results would support the theory of Porath et a l . (1975). These studies demonstrate that immunoglobulins can be isolated from blood serum and plasma. Radial immunodiffusion analysis indicated that Cu-loaded column yielded the highest IgG ac t iv i ty (higher than 95%), indicating the mildness of this method for isolat ion of Igs. However, i t is recommended to regenerate the Cu-loaded column with 50% ethanol when blood plasma is used as start ing material . 156 1.2 T E e O oo CM < CD G£ O in CD < .8 • T 9 6 a. 30 60 ELUTION VOLUME. Figure 64. Elution prof i les of control (TF) and diethyl pyrocarbonate treated transferr in (DEP-TF). Samples (30 mg/5 mL 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2) were applied to the column (1.4 x 7.0 cm) and washed (W) with the start ing buffer then eluted (E) with 0.01 M imidazole. The flow rate was 30 mL/hr. SEPARATION OF OVOTRANSFERRIN FROM EGG WHITE BY METAL CHELATE INTERACTION CHROMATOGRAPHY 158 The greater resistance against enterobacterial infection of human infants fed with breast milk than those fed with a r t i f i c i a l formula has been attr ibuted, to a great extent, to the presence of a large quantity of lactoferr in in human milk compared to cow's milk (Packard, 1982). The s imi lar i ty in structure and biological ac t iv i ty between ovotransferrin and lactoferr in jus t i fy the antimicrobial effect of ovotransferrin being added to infant formula (Valenti et a l . , 1983; Giacco-Del et a l . , 1985). In addi t ion, i t has been found that ovotransferrin does not sensit ize infants (Giacco-Del et a l . , 1985). This part ofthe thesis deals with the separation and the mechanism of the binding of ovotransferrin by using MCIC method. A. METAL CHELATE-INTERACTION CHROMATOGRAPHY OF EGG WHITE It is known that the transi t ion metals can form a complex with compounds r ich in electrons (Porath et a l . , 1975). These compounds may be aromatic or heterocycl ic , including proteins due to their contents of Cys, His , and Tyr. However, the binding of these groups to metal ions depends on the ava i l ab i l i t y of these groups or the topography of the protein molecule (Sulkowski, 1985). Figure 65 represents the elution prof i le of undiluted, blended egg white on copper-loaded Sepharose 6B. As indicated by SDS-PAGE (Figure 66), at an early stage the unbound material was mainly ovalbumin, however, at the later stage, most of egg white proteins passed through suggesting the column had reached i ts saturation point. During the washing step, another peak (FW) appeared in the eluant which, based on electrophoretic pattern, could be a mixture of ovotrans-ferr in and lysozyme . The bound material eluted at pH 4.0 (peak 1) appeared to be pure ovotransferr in, purer than the commercial ovotransferr in. The protein eluted with 0.01M imidazole (peak 2), was l i ke ly to be ovotransferrin. Immunoelectrophoresis conducted against anti-egg white proteins antiserum 159 Figure 65. Metal chelate- interaction chromatography of egg white. 2 mL of undiluted blended egg white was passed through Cu-loaded Sepharose 6B MCIC column (7 x 1.4 cm). UB, unbound proteins; FW, fract ion eluted with washing step; E l , elution with start ing buffer, pH 4.0; 1, fraction eluted with E l ; E2, elution with 0.01M imidazole; 2, fract ion eluted with E2. 160 1 2 3 4 5 6 7 8 9 10 Figure 66. SDS-PAGE of fractions of egg white obtained by MCIC column shown in Figure 65. Lanes 1, 2 and 3, are unbound f ract ions; Lane 4, washing f rac t ion; Lane 5, peak 1; Lane 6, peak 2; Lanes 7 and 8, standard ovotransferrin and ovalbumin, respectively; Lanes 9 and 10, control egg white. 161 indicated that the ovotransferrin fract ion (peak 1) obtained by MCIC was f a i r l y pure being free from contaminants which were observed in the commercial ovotransferrin (Figure 67). B. CAPACITY OF MCIC COLUMN FOR OVOTRANSFERRIN Figure 68 shows the A28O elution prof i le when 0.2% commercial OVT solution (A28O = 1*94) was applied to a copper-loaded chelating Sepharose 6B. The column was saturated at elution volume of about 60 mL when A28O o f t n e eluted fract ion reached 1.94. At this point , the capacity for OVT was calculated to be approximately 20 mg OVT/mL copper-loaded g e l . Subsequent elution with buffer at pH 4.0 (El) and with 0.01 M imidazole (E2) recovered 65% and 30% of the bound OVT, respectively. Thus, the binding capacity of OVT appeared to be lower than that reported by Lonnerdal et a l . (1977) for human lactoferr in (LF), which was 70 mg LF/mL gel containing 50 umol copper ions. The desorption of OVT in two steps may indicate the presence of at least two forms of OVT. One form can be eluted by lowering the pH. The acidi ty weakens the binding of proteins with the metal ions by protonation of the protein electron donor groups which are responsible for binding with the metal ions. The second form can be eluted by using a strong competitor, i . e . imidazole. Imidazole can e f f i c i en t l y compete with exposed His groups on the protein for metal binding. As reported by Sulkowski (1985), the presence of one hist id ine residue is suf f ic ient for retention while the presence of 2 or 3 His results in multipoint attachment to IDA-Cu gel and a stronger retention. The presence of two forms of conalbumin reported by Clark et al.(1963) and Feeney et a l . (1963) was discussed by Powrie and Nakai (1986). Rogers et a l . (1977) reported two forms of hist id ine in OVT, one was reactive and the other was DEP-non-reactive. 162 Figure 67. Immunoelectrophoresis against anti whole egg white antiserum of ovotransferrin fract ion (F) prepared by the MCIC method as compared to commercial ovotransferrin (OVT), and egg white (EW). 163 Figure 68. Saturation pro f i l e of commercial ovotransferrin on Cu-loaded Sepharose 6B column. 0.2% ovotransferrin was passed through 3 mL of a Cu-loaded column (7 x 1.4 cm). E l , elution with 0.05 M Tris-acetate /0.5 M NaCl, pH 4.0; E2, elution with 0.01 M imidazole. 1 and 2 are eluted ovotransferr in. .164 C. MECHANISM OF OVOTRANSFERRIN SEPARATION BY MCIC The binding of OVT to metal-chelate gel is believed to be the result of the a b i l i t y of e lectron-r ich groups such as h ist id ine and tryptophan to substitute weakly bonded water or buffer in the metal complex (Lonnerdal and Keen, 1982). The s t a b i l i t y of the binding even in 1M NaCl would rule out the poss ib i l i t y of ionic interaction being the principal force in the interact ion. To demonstrate whether the metal binding a b i l i t y of OVT has any role in the mechanism of OVT separated by MCIC, metal bound OVT was applied to the MCIC column. Figure 69 represents the elution prof i le of metal free OVT (apo-OVT), f e r r i c - and copper-saturated OVT from the MCIC column. Evidently, an i ron- or copper- containing OVT was adsorbed on a Cu-chelate gel and subsequently eluted. Figure 69 also shows that metal-saturated OVT bound with MCIC column as strongly as the apo-form of OVT. Before and after chromatography of Fe-OVT and Cu-OVT, the protein contained two atoms of F e 3 + and C u 2 + per molecule, respectively (Lonnerdal and Keen, 1982). After addition of excess F e 3 + or C u 2 + and subsequent d i a l y s i s , OVT has two bound metal atoms per mole protein (Brock, 1985). However, on Cu-chelate g e l , surface-exposed groups on OVT may have a tendency to bind copper ions. To investigate i f h ist id ine groups were responsible for the copper-chelate gel and OVT interact ion, DEP modified OVT was applied to the column. Figure 70 shows the elution prof i le of OVT before and after hist id ine group modification by DEP. Without modif icat ion, OVT was adsorbed strongly to the Cu-chelate gel and no OVT was eluted in the washing step at alkal ine pH. Bound OVT could subsequently be recovered by 0.01M imidazole. Hist idine modified OVT (DEP-OVT) did not bind to Cu-chelate gel and was almost completely washed out during the washing step at alkal ine pH. This behavior of DEP-OVT on Cu-chelate gel clearly indicated that modification of 11.7 hist id ine residues out of 13 hist idine 165 o 1 • 0 30 60 ELUTION VOLUME, ml Figure 69. Metal chelate interaction chromatography of apo-ovotransferrin (APO-OVT), Fe-ovotransferrin (Fe-OVT) and Cu-ovotransferrin (Cu-OVT). 3 mL (8 mg/mL) was applied to Cu-loaded Sepharose 6B (7 x 1.4 cm) after equi l ibrat ion with 0.05 M Tr is-acetate /0 .5 M NaCl, pH 8.2. W, Washing with the equi l ibrat ing buffer; E, elution with 0.01 M imidazole; flow rate was 30 mL/hr. 166 o • • • • • • , , 0 30 60 ELUTION VOLUME, ml Figure 70. Metal chelate interaction chromatography of control ovotransferrin (OVT) and diethyl pyrocarbonate treated ovotransferrin (DEP-OVT). 3 mL (8 mg/mL) was applied to Cu-loaded Sepharose 6B (7 x 1.4 cm) after equi l ibrat ion with 0.05 M Tr is-acetate/0 .5 M Nacl, pH 8.2. W, washing with the equi l ibrat ing buffer; E, elution with 0.01 M imidazole; flow rate was 30 mL/hr. 167 residue per mole OVT dras t ica l ly inhibited the proteins' a b i l i t y to bind Cu-chelate matrix. Considering two hist id ine residues are involved in the metal binding in the Fe- or Cu- saturated OVT, there are s t i l l 11 hist id ine groups le f t free to interact with immpbi1ized. metal on Sepharose 6B. However, blocking hist id ine groups would destroy the a b i l i t y of OVT to bind MCIC column. In conclusion, this part demonstrates for the f i r s t time a method to separate ovotransferrin from egg white by a single chromatographic step. The s p e c i f i c i t y and capacity of MCIC column for ovotransferrin are high, and i t should be easy to adapt the method for isolat ion of ovotransferrin to a larger scale operation. The isolated OVT from egg white may be incorporated in infant formula, s ince, i t has not had sensit iz ing effects on OVT-treated babies (Giacco-Del et a l . , 1985). Giacco-Del (1985) reported that when ovotransferrin-enriched milk was fed to 15 babies for 60 days the values of total IgE, as determined by the radioimmunoassay method, remained within the normal range. 168 CONCLUSIONS AND RECOMMENDATIONS A hexametaphosphate method was developed for minimizing p-Lg and maximizing Ig in whey. The new method avoids saturation of iron binding proteins which abolishes their bacter iostat ic ac t iv i ty and also avoids u l t i l i z a t i o n of non-food grade chemicals which would require regulatory approval, compared to SHMP which is already approved as a food grade chemical. For isolat ion of bioactive proteins, MCIC treatment of cheese whey, skimmilk, blood plasma and serum and egg white is recommended, based on the following f indings: (1) Ig of almost 90% purity can be recovered from cheese whey using a simple process, with pract ica l ly no pre-treatment of whey. (2) The MCIC has high capacity for Ig isolat ion from cheese whey, at least 1 l i t r e of whey per 25 ml copper-loaded g e l . (3) Lactoferrin and bovine serum albumin may also be separated from cheese whey. (4) The treated whey or unbound fraction contains mainly p-lactoglobulin and a-lactalbumin in concentrations similar to the untreated wheys. (5) This method can be used for extraction of Ig and TF from blood, and ovotransferrin from egg white. The preliminary experiments indicate the potential of this method to isolate Ig d i rec t ly from skimmilk. (6) The fact that chemical modification of h ist id ine groups in Ig, TF, OVT, LF rich fract ion and casein f ract ions, destroys coordinate binding to MCIC columns supports the idea of the involvement of h ist id ine groups in the interaction with copper immobilized on g e l ; however, changing the elution conditions may activate the other mechanism of the interact ion. (7) The isolated immunoglobulins from colostrum, cheese whey and blood 169 recognize LPS isolated from EL col i , S_j_ typhimurium and B_j_ parapertussis. These results may encourage the addition of the bioactive proteins to infant formula to give similar performance as human milk. (8) The MCIC column can be easi ly regenerated for re-use. FUTURE WORK In order to apply the MCIC process for large scale recovery of Ig, further studies are required, (1) to search for cheaper and/or mechanically more stable material for the chelating support which would be more suitable for scale-up process; (2) to determine the l i fe - t ime of the column; (3) to investigate the inhibi t ion of pathogens with the separated Ig by animal assays; (4) to study the structure-function relationship of Ig ac t iv i ty and improve the funct ional i ty of immunoglobulins. 170 REFERENCES CITED A d i n o l f i , M., Glynn, A . A . , Lindsay, M. and Milne, C M . 1966. Serological properties of YA antibodies to Escherichia col i present in human colostrum. Immunol. 10:517. Aishima, T. and Nakai, S. 1986. Centroid mapping optimization: a new ef f ic ient optimization for food research and processing. J . Food S c i . 51:1297. Alexander, A. 1984. Making a recovery. Food 6:29. Amundson, C.H. and Watanawamchakorn, S. 1982. Production of enriched protein fractions of 3- lactoglobulin and a-lactalbumin from cheese whey. J . Food Proc. Preserv. 6:55. Anderson, K . J . , McLaughlan, P . , Devey, M.E. and Coombs, R.R.A. 1979. Anaphylactic sens i t iv i ty of guinea pigs drinking di f ferent preparations of cow's milk and infant formulae. C l i n . Exp. Immunol. 35:454. Anderson, S . A . , Chinn, H.I. and Fisher , K.D. 1982. History and current status of infant formulas. Am. J . C l i n . Nutr. 35:381. Andersson, L. 1984. Fractionation of human serum proteins by immobilized metal a f f i n i t y chromatography. J . Chromatogr. 315:167. Anonymous. 1985. User Manual Spiral Plater , Model D. Spiral Systems, Inc. Bethesda, MD. Antonini , E. 1977. Composition for the treatment of bacterial intestinal in fect ions. U.S. Patent 4,029,711. Arnold, R.R., Cole, M.F. and McGhee, J .R. 1977. A bacteriocidal effect for human lac to fe r r in . Science 197:263. Ashoor, S.H. and Monte, W.C. 1983. High performance l iqu id chromatographic determination of Bifidobacterium bifidium growth factors in human milk. J . AOAC. 66:135. Asse l in , J . , Amiot, J . , Gauthier, S . F . and Mourad, W. 1986. Al lergenic i ty of whey proteins according to enzymatic hydrolysis. CIFST, 29th Annual Conference, June 29 - July 2, 1986. A t a s s i , M.Z., van Oss, C . J . and Absolom, D.R. (Eds.) 1984. Molecular  Immunology A Textbook. Marcel Dekker, Inc. , New and Basel . Atha, D.H. and Inghams, K.C. 1981. Mechanism of precipi tat ion of proteins by polyethylene g l y c o l s , Analysis in terms of excluded volume. J . B i o l . Chem. 256:12108. Azar i , P. and Baugh, P.F. 1967. A simple and rapid procedure for preparation of large quantities of ovotransferrin. Arch. Biochem. Biophys. 118:138. Bal labr iga, A. 1982. Immunity of the in fant i le gastrointestinal tract and implications in modern infant feeding. Acta Paediatr. Japonica. 24:235. 171 Barlow, B . , S a n t u l l i , T . V . , Heird, W.C., P i t t , J . , Blanc, W.A. and Schull inger, J . N . 1974. An experimental study of acute neonatal enterocol i t is - the importance of breast milks. J . Fed. Surg. 9:587. Bezkorovainy, A. 1977. Human milk and colostrum proteins: A review. J . Dairy S c i . 60:1023. Bezkorovainy, A. and Nichols, J . H . 1976. Glycoproteins from mature human milk whey. Pediatr. Res. 10:1. Bezkorovainy, A. and Topouzian, N. 1981. Bifidobacterium bif idus var. Pennsvlavanicus growth promoting ac t iv i ty of human milk casein and i ts der ivat ives. Eur. J . Biochem. 13:585. Bjorck, L . , 1978. Antibacterial effect of the lactoperoxidase system on psychrotrophic bacteria in milk. J . Dairy Res. 45:109. Blackberg, L. and Hernel l , 0. 1980. Isolation of lactoferr in from human whey by a single chromatographic step. FEBS Letters 109:180. Blanc, B. 1981. Biochemical aspects of human milk. Comparison with bovine milk. World Rev. Nutr. and Dietet. 36:1. Bokhout, B .A . , van Asten-Noordijk, J . J . L . and Stok, W. 1986. Porcine IgG. Isolation of two IgG-subclasses and anti-IgG c lass - and subclass-speci f ic antibodies. Mol. Immunol. 23:675. Brignon, G . , Chtourou, A. and Ribadeau-Dumas, B. 1985. Does P-lactoglobulin occur in human milk? J . Dairy Res. 52:249. Brock, J . H . 1985. Transferr ins. In: Metalloproteins. Part 2 Metal Proteins  with Non-redox Roles. Harrison, P.M. (Ed.) Verlag Chemie, GmbH , Weinheim. Brock, T.D. 1979. (Ed.) Biology of Microorganisms. Prentice Ha l l , Inc., Englewood C l i f f s , N.J. p.526-538. Brooks, T . L . and Stevens, A. 1985. Preparative HPLC pur i f icat ion of IgG and IgM monoclonal antibodies. Amer. Lab. Oct. p.54. Bul len, C . L . , Tearle, P.V. and W i l l i s , A .T . 1976. Bif idobacteria in the intest inal tract of infants: and in vivo study. J . Med. Microbiol . 9:325. Bul len, C . L . , Tearle, P.V. and Stewart, M.G. 1977. The effect of humanized milks and supplemented breast feeding on the fecal f lo ra of infants. J . Med. Microbiol . 10:403. Bul len, J . J . , Rogers, H.J . and Leigh, L. 1972. Iron-binding proteins in milk and resistance to Escherichia col i infection in infants. B r i t . Med. J . 1:69. Butler , J . E . 1969. Bovine immunoglobulins: A review. J . Dairy S c i . 52:1895. But ler , J . E . 1983. Bovine immunoglobulins: An augmented review. Vet. Immunol, and Immunopath. 4:43. 172 But ler , J . E . and Maxwell, C .F . 1972. Preparation of bovine immunoglobulins and free secretory component and their spec i f ic ant isera. J . Dairy S c i . 55:151. Butler , J . E . , Peterson, L. and McGivern, 1980. A re l iable method for the preparation of bovine secretory immunoglobulin A (SIgA) which circumvents problems posed by IgG1 dimers in colostrum. Mol. Immunol. 17:757. Chandan, R . C . , Shahani, K.M. and Hol ly , R.G. 1964. Lysozyme content of human milk. Nature 204:76. Chandan, R . C . , Parry, R.M. and Shahani, K.M. 1968. Lysozyme, l ipase and ribonuclease in milks of various species. J . Dairy S c i . 51:606. Clark, J . R . , Osuga, D.T. and Feeney, R.E. 1963. Comparison of avian eggwhite conalbumins. J . B i o l . Chem. 238:3621. Cole, M.F. , Arnold, R.R., Mestecky, J . , Prince, S . , Kulhavy, R. and McGhee, J .R . 1976. In: Microbial Aspects of Dental Caries Vol . II. S t i l e , H.M., Loesche, W.J. and O'Brien, T .C . (Eds). Information Retrieval Inc. , Washington, D.C. and London. Committee on Nutr i t ion, American Academy of Pediat r ics . 1980. Encouraging breast-feeding. Pediatrics 65:657. Cunningham, A . S . 1977. Morbidity in breast-fed and a r t i f i c i a l l y - f e d infants. J . Pediatr. 90:726. Cunningham, A . S . 1979. Morbidity in breast-fed and a r t i f i c i a l l y - f e d infants, II. J . Pediatr. 95:685. Dolby, J .M. and Honour, P. 1979. Bacteriostasis of Escherichia col i by milk IV. The bacter iostat ic antibody of human milk. J . Hyg. Camb. 83:255. Dolby, J .M. and Stephens, S. 1983. Antibodies to Escherichia col i O antigens and the in v i t ro bacter iostat ic properties of human milk and i ts IgA. Acta Paediatr. Scand. 72:577. Dortmann, A. 1967. Bringt die Muttermilch dem SSugling einen Gewinn? Arch. f. Gyn. 204:213. Duncan, J . R . , Wilkie, B .N . , Hiestand, F. and Winter, A . J . 1972. The serum and secretory immunoglobulins of cat t le : characterization and quantitation. J . Immunol. 108:965. Ebina, T . , Sato, A . , Umezu, K., Aso, H. , Ishida, N., Seki , H. , Tsukamoto, T . , Takase, S . , Hoshi, S. and Ohta, M. 1984. Treatment of multiple sclerosis with antimeasles cow colostrum. Med. Microbiol . Immunol. 173:87. Ebina, T . , Sato, A . , Umezu, K., Ishida, N. , Ohyama, S . , Oizumi, A . , Aikawa, K., Katag i r i , S . , Katsushima, N. , Imai, A . , Kitaoka, S . , Suzuki, H. and Konno, T. 1985. Prevention of rotavirus infection by oral administration of cow colostrum containing antihumanrotavirus antibody. Med. Microbiol . Immunol. 174:177. 173 Edy, V . G . , B i l l i a u , A. and Somer, P.D. 1977. Pur i f icat ion of human f ibroblast interferron by zinc chelate a f f i n i t y chromatography. J . B i o l . Chem. 252:5934. Ei tenmi l ler , R.R., Fr iend, B .A . , Shahani, K.M. and B a l l , E.M. 1974. Immunochemical studies on human and bovine milk lysozymes. J . Food S c i . 39:930. Ei tenmi l ler , R.R., Fr iend, B.A. and Shahani, K.M. 1976. Relationship between composition and s t a b i l i t y of bovine milk lysozyme. J . Dairy S c i . 59:834. E l l i o t , J . I . 1978. Productivi ty. Canada Agr ic . 23:27. ESPGAN Committee on Nutr i t ion. 1982. Guidelines on infant nut r i t ion . Acta. Paediatr. Scand. Suppl. 302:1. Ey, P . L . , Prowse, S . J . and Jenkin, C.R. 1978. Isolation of pure IgGi, IgG2a and IgG2b immunoglobulins from mouse serum using protein A-Sepharose. Immunochem. 15:429. F a l l o t , M.E. , Boyd, J . L . and Oski , F.A. 1980. Breast feeding reduces the incidence of hospital admissions for infection in infants. Pediatr. 65:1121. Feeney, R . E . , Abplanalp, H. , Clary, J . J . , Edwards, D .L . , and Clark, J .R . 1963. A genetical ly varying minor protein constituent of chicken egg white. J . B i o l . Chem. 238:1732. Fey, H. , P f i s t e r , H. , Messer l i , J . , Sturzenegger, N. and Grolimund, F. 1976. Methods of i so la t ion , pur i f icat ion and quantitation of bovine immunoglobulins: a technical review. Zb l . Vet. Med. B . , 23:269. Fleischman, J . B . , Pain, R.H. and Porter, R.P. 1962. Reduction of y-Globul ins. Arch. Biochem. Biophys. Suppl. 1:174. Forsum, E. 1974. Nutrit ional evaluation of whey protein concentrates and their f ract ions. J . Dairy S c i . 57:665. France, G . L . , Marmer, D .J . and Steele, R.W. 1980. Breast feeding and salmonella in fect ion . Amer. J . Dis . Ch i ld . 134:147. Fr iend, B .A . , Shahani, K.M. and Mathur, B.N. 1983. Newer advances in human milk substitutes for infant feeding. J . Appl. Nutr. 35:88. Garvey, J . S . , Cremer, N.E. and Sussdorf, D.H. (Eds.) 1977. Methods in  Immunology. W.A. Benjamin Inc. , Reading. Gerrard, J.W. 1974. Breast feeding: second thoughts. Pediatrics 54:757. Gerstenberger, H . J . , Haskins, H.D., McGregor, H.H. and Ruh, H.O. 1915. Studies in the adaptation of an a r t i f i c i a l food to human milk. Am. J . Dis. Chi ld . 10:249. 174 Giacco-Del , G . S . , Leone, A . L . and Ferlazzo, A. 1985. Total IgE in newborns treated prophylact ical ly with conalbumin. Int. J . T i s s . Reac. 7:535. Gindrat, J . J . , Gothefors, L . , Hanson, L.A. and Winberg, J . 1972. Antibodies in human milk against E. col i of the serogroups most commonly found in neonatal infect ions. Acta Pediatr. Scand. 61:587. Goldman, A . S . , Anderson, J r . D.W., Se l l a rs , W.A., Saperstein, S . , Kniker, W.T. and halpern, S.R. 1963a. I. Oral challenge with milk and isolated milk proteins in a l l e rg ic chi ldren. Pediatrics 32:425. Goldman, A . S . , S e l l a r s , W.A., Halpern, S.R. and Anderson, J r . D.W. 1963b. II. Skin testing of a l l e rg ic and normal children with pur i f ied milk proteins. Pediatrics 32:572. Goldman, A . S . and Smith, CW. 1974. Host resistance factors in human milk. J . Pediatr. 82:1082. Goldman, A . S . , Garza, C , Nichols, B.L. and Goldblum, R.M. 1982. Immunological factors in human milk during the f i r s t year of lacta t ion. J . Pediatr. 100:563. Gordon, W.G. 1945. Method for the separation of protein from animal matter containing protein in water-soluble form. U.S. Patent 2,377,624 (June 5, 1945). Gordon, W.G., Z iegler , J . and Basch, J . J . 1962. Isolation of an iron-binding protein from cow's milk. Biochim. Biophys. Acta. 60:410. Gothefors, L. and Marklund, S. 1975. Lactoperoxidase ac t iv i ty in human milk and in sal iva of newborn infant. Infect. Immunol. 11:1210. Gothefors, L. and Winberg, J . 1975. Host resistance factors . J . Trop. Pediatr. 21:260. Groves, M.L. 1962. The isolat ion of red protein from milk. J . Am. Chem. Soc. 82:3345. Groves, M.L. and Gordon, W.G. 1967. Isolation of a new glycoprotein-a and a v-globul in from individual cow milks. Biochem. 6:2388. Gurr, M.I. 1981. Review of the progress of dairy science: human and a r t i f i c i a l milks for Infant feeding. J . Dairy Res. 48:519. Gyllenberg, H. and Raine, P. 1957. The value of colony counts in evaluating the abundance of Lactobacil lus bif idus in infant feces. Acta Pathol. Microbiol . Scand. 41:144. Gyorgy, P. 1953. A hitherto unrecognized biochemical differences between human milk and cow's milk. Pediatr. 11:98. Haenel, H. 1970. Human normal and abnormal gastrointestinal f l o r a . Amer. J . C l i n . Nutr. 23:1433. 175 Haire, R.N. , T i s e l , W.A., White, J . G . and Rosenberg, A. 1984. On the precipi tat ion of proteins by polymers: The hemoglobin-polyethylene glycol system. Biopolymers 23:2761. Hambraeus, L. 1977. Proprietary milk versus human breast milk in infant feeding. Ped. C l i n . North Amer. 24:17. Hambraeus, L . , Lonnerdal, B . , Forsum, E . , and Gebre-Medhin, M. 1978. Nitrogen and protein components of human milk. Acta Pediatr. Scand. 67:561. Hanson, L.A. and Soderstrom, T. 1981. Human milk: defense against infect ion. In: Nutrit ion and Child Health: Perspectives for the 1980's. Alan, R. (Ed.) p.147-159. L i s s , Inc. , New York. Hanson, L.A. and Winberg, J . 1972. Breast milk and defense against infection in the newborn. Arch. Dis . Chi ldh. 47:845. Hansson, H. and Kagedal, L. 1981. Adsorption and desorption of proteins in metal chelate a f f i n i t y chromatography. J . Chromatogr. 215:333. Harrison, P.M. (Ed.) 1985. Metal!oproteins. Part 2 Metal Proteins with  Non-redox Roles. Verlag Chemie, Gmbh, Weinheim. Heppell, L .M. , Cant, A . J . and Kilshaw, P . J . 1984. Reduction in the antigenicity of whey proteins by heat treatment: a possible strategy for producing a hypoallergenic infant formula. Br. J . Nutr. 51:29. H i l l , I.R. and Porter, P. 1974. Studies on the bactericidal ac t iv i ty for Escherichia col i of porcine serum and colostral immunoglobulins and the role of lysozyme with secretory IgA. Immunol. 26:1239. Hi lper t , H. , Amste, H. , Pahud, J . J . and Gerber, H. 1974/1975. Tests on the immunological ac t iv i ty of spec i f ic bovine ant i -E^ col i milk antibodies. Nestle Research News 130. Hirano, S . , Hayashi, H. and Terabayashi, T. 1968. B io logica l ly active glycoprotein in human colostrum. J . Biochem. 64:563. Holbrook, J . J . and Ingram, V.A. 1973. Ionic properties of an essential h ist id ine residue in pig heart lactate dehydrogenase. Biochem. J . 131:729. Holmgren, L. 1985. Toxins affecting intest inal transport processes. In: The  virulence of Escherichia c o l i . Review and Methods. Sussman, M. (Ed.) Academic Press, New York. Jann, K. 1985. Isolation and characterization of 1ipopolysaccharides (0 and R antigens) from Escherichia c o l i . In: The virulence of Escherichia c o l i ,  Review and Methods (Ed.) Sussman, M. Academic Press Inc. New York. 176 Jann, K. and Jann, B. 1985. Cell surface components and virulence: Escherichia  co l i 0 and K antigens in relation to virulence and pathogencity. In: The  virulence of Escherichia c o l i , Reviews and Methods. (Ed.) Sussman, M. Academic Press, Inc. New York. Jenness, R. 1982. Inter-species comparison of milk proteins. In: Developments  in Dairy Chemistry - I. Fox, P.F. (Ed.) p.87. Applied Science Publishers. London and New York. Johansson, B.G. 1969. Isolation of crysta l l ine lactoferr in from human milk. Acta Chem. Scand. 23:683. Josephson, R.V. , Mikolajcik, E.M. and Sinha, D.P. 1972. Gel isoe lect r ic focussing of selected bovine immunoglobulins. J . Dairy S c i . 55:1508. Kanamaru, Y . , N i k i , R. and Arima, S. 1977. Suscept ib i l i ty of bovine colostral immunoglobulin G (IgG) subclasses to digestive enzymes and reducing agents. Agr ic . B i o l . Chem. 41:1203. Kanamaru, Y . , Fu j i t a , T . , Ohya, S. and Usui , H. 1980. Non-specific pur i f icat ion of bovine secretory IgA from normal bovine colostrum. Res. B u l l . Fac. Agr. Gifu Univ. 43:215. Kanamaru, Y . , Kuzuya, Y. and Tanahashi, T. 1981. Preparation of bovine free secretory component from colostrum. Res. B u l l . Fac. Agr. Gifu Univ. 45:223. Kanamaru, Y . , Kuzuya, Y. and Tanahashi, T. 1982a. Pur i f icat ion of secretory IgA from bovine colostrum. Agr ic . B i o l . Chem. 46:1531. Kanamaru, Y . , Kuzuya, Y. and Tanahashi, T. 1982b. Some properties of secretory IgA pur i f ied from bovine colostrum. Agr ic . B i o l . Chem. 46:2009. Kaneko, T . , Wu, B.T. and Nakai, S. 1985. Selective concentration of bovine immunoglobulins and a-lactalbumin from acid whey using FeCl3. J . Food S c i . 50:1531. Kawakata, N. 1984. The Isolation and characterization of human milk lactoferr in and the f luctuations of i ts concentration at di f ferent stages of lac ta t ion . J . of the Nippon Medical School. 51:426. Cited from Dairy S c i . Abs. 48:1586 (1986). Kilshaw, P . J . , Heppell, L .M.J , and Ford, J . E . 1982. Effects of heat treatment of cow's milk and whey on the nutri t ional quality and antigenic propert ies. Arch. Dis . Ch i ld . 57:842. Kirkpatr ick, C . H . , Green, I., Rich, R.R., and Schade, A . L . 1971. Inhibition of the growth of Candida albicans by iron saturated lac to fer r in : relation to host defense mechanisms in chronic mucocutaneous candidiasis . J . Infect. Dis . 124:539. 177 Kur isak i , J . , Nakamura, S . , Kaminogawa, S . , Yamauchi, K., Walanabe, S . , Hotta, K., Hat tor i , M. 1985. Antigenicity of modified B-lactoglobul in examined by three di f ferent assays. Agr ic . B i o l . Chem. 49:1733. Kuwata, T . , Pham, A .M . , Ma, C.Y. and Nakai, S. 1985. Elimination of B-lactoglobul in from whey to simulate human milk protein. J . Food S c i . 50:605. Kuznetsov, V .S . and Rebrova, G.V. 1983. Dried whey added to the diet of young birds to increase natural resistance to in fect ion . Sbornik Nauchnykh Trudov Moskovskoi Veterinarnoi Akademii 12. Cited from Dairy S c i . Abs. 47:2802. Laemmli, U.K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 227:680. Larsen, S.A. 1978. Relation of breast versus bottle feeding to hospital ization for gastroentr i t is in a middle class U.S. poluplat ion. J . Pediatr. 92:417. Law, B.A. and Reiter , B. 1977. The isolat ion of bacter iostat ic properties of lactoferr in from bovine milk whey. J . Dairy Res. 44:595. Lebenthal, E. 1975. Cow's milk protein a l le rgy . Pediatr. C l i n . N. Amer. 22:827. L ibera tor i , A . C . and Napolitano, L. 1980. Isolation and preliminary physiochemical characterization of human B-lactoglobul in. Milchwissenschaft 35:65. Li-Chan, E . , Nakai, S . , Sim, J . , Bragg, D.B. and Lo, K.V. 1986. Lysozyme separation from egg white by cation exchange column chromatography. J . Food S c i . 51:1032. Lisowski, J . , Janusz, M., Tyran, B . , Morawiecki, A . , Galob, S. and Bialkowska, H. 1975. Immunoglobulins of colostrum. III. Comparative studies of ovine serum and colostral IgGi and IgG 2 . Immunochemistry 12:159. Lonnerdal, B. 1985. Biochemistry and physiological function of human milk proteins. Am. J . C l i n . Nutr. 42:1299. Lonnerdal, B. and Keen, C .L . 1982. Metal chelate a f f i n i t y chromatography of proteins, J . Appl. Biochem. 4:203. Lonnerdal, B . , Carlsson, J . and Porath, J . 1977. Isolation of lactoferr in from human milk by metal-chelate a f f in i t y chromatography. FEBS Letters 75:89. Mair, S .G. 1982. Surface v isual izat ion routines. University of Br i t ish Columbia, Computing Centre, Vancouver, B.C. Canada. Martin, L.N. 1982. Separation of guinea pig IgG subclasses by a f f in i ty chromatography on protein A-Sepharose. J . Immunol. Methods 52:205. ) 178 Mathur, B.N. and Shahani, K.M. 1979. Use of total whey constituents for human food. J . Dairy S c i . 62:99. McCallum, I.M., E l l i o t , J . I . and Owen, B.D. 1977. Survival of colostrum-derived neonatal piglets fed gamma-globulins. Can. J . Animal S c i . 57:151. McClelland, D .B .L . , McGrath, J . and Samson, R.R. 1978. Antimicrobial factors in human milk. Acta Pediatr. Scand. Suppl. 271:3. McLaughlan, P . , Anderson, K . J . , Widdowson, E. and Coombs, R.R.A. 1981. Effect of heat on anaphylact ic-sensit iz ing capacity of cow's milk, goat's milk, and various infant formulae fed to guinea pigs. Arch. Dis . Ch i ld . 56:165. Melachouris, N. 1972. Interaction of B-lactoglobulin with polyphosphates. J . Agr. Food Chem. 20:798. Mendez, A. and Olano, A. 1979. Lactulose, A review of some chemical properties and applications in infant nutr i t ion and medicine. Dairy S c i . Abs. 41:531. M i l l e r , F.D. (Ed.) 1983. Out of the Mouths of Babes: The Infant Formula  Controversy. Social Philosophy and Policy Center, Bowling Green State Universi ty , OH, U.S.A. M i l l e r , T . E . 1969. K i l l i n g and lys is of gram-negative bacteria through the synergist ic effect of hydrogen peroxide, ascorbic acid and lysozyme. J . Bacter io l . 98:949. Mizrawi, A . , Barlow, 0 . , Berdon, W., Blanc, W.A. and Silverman, W.A. 1965. Necrotizing enterocol i t is in premature infants. J . Ped. 66:697. Mobbs, G.A. 1972. Breast feeding and cot deaths. Med. J . Austr. 2:794. Moneret-Vautrin, D.A., and G r i l l i a t , J . P . 1979. Relative importance of the di f ferent constituants of milk in cow's milk a l lergy . Cahiers de Nutrition et de Dietetique. 14:237. Moneret-Voutrin, A . , Humberto, G . , A l i a s , C. and G r i l l i a t , J . P . 1982. Recent data on the immunoallergic properties of dairy proteins. Le Lait 62:396. Morrison, W.R. 1964. A fas t , simple and re l iable method for the microdetermination of phosphorus in biological materials. Anal . Biochem. 7:218. Murphy, F . A . , Aalund, 0 . , Osebold, J.W. and Car ro l , E . J . 1964. Gamma globulins of bovine lactal secretions. Arch. Biochem. Biophys. 108:2320. Nakai, S. and Le, A . C . 1970. Spectrophotometric determination of protein and fat in milk simultaneously. J . Dairy S c i . 53:276. Nakai, S . , Koide, K. and Eugster, K. 1984. A new mapping super-simplex optimization for food product and process development. J . Food S c i . 49:1143. 179 Narayanan, I., Prakash, K., Verma, R.K. and Gujra l , V.V. 1983. Administration of colostrum for the prevention of infection in the low birth weight infant in a developing country. J . Trop. Pediatr. 29:197. Nisonoff, A. (Ed.) 1982. Introduction to Molecular Immunology. Sinauer Associates, Inc. , Sunderland, Massachusetts. Ogra, S .S . and Ogra, P.L. 1978. Immunological aspects of human colostrum and milk. I. Distr ibution character ist ics and concentrations of immunoglobulins at di f ferent times after the onset of lactat ion. J . Pediatr. 92:546. Ogra, S . S . , Weintraub, D. and Ogra, P.L. 1977. Immunological aspects of human colostrum and milk. III. Fate and absorption of ce l lu la r and soluble components in the gastrointestinal tract of the newborn. J . Immunol. 119:245. Ohtani, H. and Kawai, T. (Eds.) 1981. Immunoelectrophoresi s , Vo l . 3. Igaku-Shoin L t d . , Tokyo, New York, N.Y. Ones, S.U. 1979. Immunoglobulins of human colostrum and milk. J .Pediatr . 94:497. Otani, H. , Uchio, T. and Tokita , F. 1985. Antigenic react iv i t ies of chemically modified B-lactoglobulins with antiserum to bovine B- lactoglobul in. Agr ic . B i o l . Chem. 49:2531. Packard, V .S . (Ed.) 1982. Human Milk and Infant Formula. Academic Press, New York. Paul, K . -G . and Ohlsson, P.- I . 1985. The chemical structure of lactoperoxidase. In: The Lactoperoxidase System Chemistry and Biological Signif icance. Pru i t t , K.M. and Tenovud, J . O . (Eds). Marcell Dekker Inc. New York. Pahud, J . J . , Monti, J . C . , and Jost , R. 1985. A l lergenic i ty of whey protein: Its modification by t rypt ic in vivo hydrolysis of the protein. J . Pediatr. Gastroentrol. Nutr. 4:408. Pearce, R .J . 1983. Thermal separation of B-lactoglobul in and a-lactalbumin in bovine Cheddar cheese whey. Aust. J . Dairy Technol. 38:144. Pharmacia Fine Chemicals. Sephacryl S-300 Superfine and cal ibrat ion ki ts for gel f i l t r a t i o n . Printed in Sweden by Upplands Graf iska, A .B . October 1978-12. P i t ta rd , W.B. 1979. Breast milk immunology. Amer. J . Dis. Ch i ld . 133:83. Porath, J . and O l i n , B. 1983. Immobilized metal ions a f f i n i t y adsorption and immobilized metal ion a f f in i t y chromatography of biomaterials. Serum protein a f f i n i t i e s for gel-immobilized iron and nickel ions. Biochemistry 22:1621. 180 Porath, J . , Carlsson, J . Olsson, I., Belfrage, G . , 1975. Metal chelate a f f i n i t y chromatography, a new approach to protein f ract ionat ion. Nature. 258:598. Porath, J . , 01 i n , B . , and Granstrand, B . , 1983. Immobilized metal a f f in i ty chromatography of serum proteins on gel immobilized group IIIA metal ions. Arch. Biochem. Biophys. 225:543. Powrie, W.D. and Nakai, S. 1986. Chemistry of eggs and egg products. In: Egg  Science and Technology. Stadelman, W.J . , C o t t e r i l l , O . J . (Eds. ) . The AVI Publishing Company, Inc. pp. 97-139. Pru i t t , K.M. and Tenovuo, J . O . (Eds.) 1985. The Lactoperoxidase System  Chemistry and Biological Signi f icance. Marcel Dekker, Inc. New York. Pul lan, C .R . , Toms, G . L . , Martin, A . J . , Gardner, P . S . , Webb, J . K . G . and Appleton, D.R. 1980. Breast-feeding and respiratory syncytial virus in fect ion . B r i t . Med. J . 281:1034. Querinjean, P . , Masson, P.L. and Heremans, J . F . 1971. Molecular weight, single chain structure and amino acid composition of human lac to fe r r in . Eur. J . Biochem. 20:420. Ramadan, N. , and Porath, J . 1985. Separation of serum proteins on a Fe^ + -monohydroxamate adsorbent. J . Chromatogr. 321:105. Rassi , Z . E . , and Horvath, C. 1986. Metal chelate- interaction chromatography of proteins with iminodiacetic acid-bonded stationary phases on s i l i c a support. J . Chromatogr. 359:241. Ratner, B . , Dworetzky, M., Oguri, S . , Ascheim, L. 1958. Studies on the a l le rgenic i ty of cow's milk. Pediatrics 22:648. Reddy, V . , Bhaskaran, C , Raghuramula, N. and Jagdeeson, V. 1977. Antimicrobial factors in human milk. Acta Pediatr. Scand. 66:229. Reiter , B. 1978. Review of the progress of dairyscience: antimicrobial systems in milk. J . Dairy Res. 45:131. Reiter , B. 1983. The biological signif icance of l ac to fe r r in . Int. J . T i s s . Reac. 1:87. Reiter , B. 1985a. Protective proteins in milk - biological signif icance and exploi tat ion. IDF Bul let in No. 191. Reiter , B. 1985b. The lactoperoxidase system of bovine milk. In: The  Lactoperoxidase System Chemistry and Biological Signi f icance. Prui t t , K.M. and Tenovuo, J . O . (Eds.) . Marcel 1 Dekker Inc. New York. Reiter , B. Marshall , V.M.E. and Bjorck, L. 1976. Non-specific bactericidal ac t iv i ty of the lactoperoxidase-thiocyanate-hydrogen peroxide system of milk against Escherichia col i and some gram-negative pathogens. Infect. Immun. 13:800. . 181 Rietschel , E . T . , Schade, U . , Jensen, M., Wollenweber, H.W., Luden'tz, 0. and Greisman, S .G. 1982. Bacterial endotoxins: Chemical structure, biological ac t i v i t y and role in septicaemia. Scand. J . of Infect. Dis . Suppl. 31:8. R ighet t i , P.G. and Caravaggio, T. 1976. Isoelectr ic points and molecular weight of proteins. A Table. J . Chromatogr. 127:1. Rogers, H .J . and Synge, C. 1978. Bacteriostat ic effect of human milk on JL c o l i : the role of IgA. Immunol. 34:19. Rogers, T . B . , Gold, R.A. and Feeney, R.E. 1977. Ethoxyformulation and photo-oxidation of hist id ines in t ransferr ins . Biochemistry 16:2299. Roosemont, J . L . 1978. Reaction of hist id ine residues in proteins with diethyl pyrocarbonate: Dif ferent ial molar absorpt iv i t ies and r e a c t i v i t i e s . Anal. Biochem. 88:314. Roth, E .F . J r . , Bookchin, R.M. and Nagel, R.L. 1979. Deoxyhemoglobin S gelation and inso lub i l i t y at high ionic strength are d is t inc t phenomena. J . Lab. C l i n . Med. 93:867. Samson, R.R., M i r t le , C. and McClelland, D.B.L. 1979. Secretory IgA does not enhance the bacter iostat ic effects of iron-binding or vitamin B12 binding proteins in human colostrum. Immunol. 38:367. Samson, R.R., M i r t le , C. and McClelland, D.B.L. 1980. The effect of digestive enzymes on the binding and bacter iostat ic properties of lactoferr in and vitamin B l 2 binder in milk. Acta Pediatr. Scand. 69:517. S a v i l a h t i , E. 1981. Cow's milk a l le rgy . Al lergy 36:73. Schade, A . L . and Carol ine, L. 1944. Raw hen egg white and the role of iron in growth inhibi t ion of . Shigel la dysenteriae. Staphylococcus aureus. Escherichia col i and Saccharomyces cerevisiae. Science 100:14. Schade, A . L . and Carol ine, L. 1946. An iron binding component in human blood plasma. Science. 104:340. Shimazaki, K.- I . and Sukegawa, K. 1982. Chromatographic prof i les of bovine milk whey components by gel f i l t r a t i o n on Fractogel TSK HW55F column. J . Dairy S c i . 65:2055. Shvedova, L.V. 1981. Effect of v i ta lakt humanized milk formula with different carbohydrate compositions on the development of bifidobacterium in the intestine of infants. Voprosy Pitaniya 3:14. Slack, A.W., Amundson, C.H. and H i l l , C.G. J r . 1985. Production of enriched p-lactoglobulin and a-lactalbumin whey protein f ract ions. J . Food Proc. Preserv. 10:19. Small, P.A. J r . , and Lamm, M.E. 1966. Polypeptide chain structure of rabbit immunoglobulins. I. yG-Immunoglobul i n . Biochemistry 5:259. 182 Spik, G . , Cheron, A . , Montreuil, J . and Dolby, J .M. 1978. Bacteriostasis of a mi lk-sensi t ive strain of Escherichia col i by immunoglobulins and iron-binding proteins in associat ion. Immunol. 35:663. Spring, F. and Peyrouset, A. 1982. Process of extraction of lactoferr in and immunoglobulins of milk. UK Patent Applicantion, GB2098998A. Stephens, S. 1984. Development of a sensit ive solid-phase radioimmunoassay technique for quantif ication of c l a s s - s p e c i f i c Escherichia col i antibodies. J . Immunol. Meth. 71:203. Stephens, S . , Dolby, J . M . , Montreuil, J . and Spik, G. 1980. Differences in inhibi t ion of the growth of commensal and enteropathogenic strains of Escherichia co l i by lactoferr in and secretory immunoglobulin A isolated from human milk. Immunol. 41:597. S t o l i a r , D.A. , Pel ley, R .P . , Kaniecki-Breen, E . , Klaus, M.H. and Carpenter, C . C . J . 1976. Secretory IgA against enterotoxins in breast milk. Lacent 1:1258. Sulkowski, E. 1985. Pur i f icat ion of proteins by IMAC. Trends Biotechnol. 3:1. Sundberg, L . , and Porath, J . 1974. Prepration of adsorbents for b iospeci f ic a f f i n i t y chromatography. I. Attachment of group-containing ligands to insoluble polymers by means of bifunctional oxiranes. J . Chromatog. 90:87. Sutton, H.E. and Jamieson. 1972. Transferr ins, haptoglobin and ceruloplasmin. In: Glycoproteins - Their Composition, Structure and Function. Part A. Gottschalk, A. (Ed.) E lsev ier , NY. Taguchl, G. (Ed.) 1957. Experimental Designs. Maruzen Publishing C o . , Tokyo, Japan. Tamura, Z. Nakajima, T. and Samejima, K. 1972. Bif idus factor in carrot V: Isolation of new active substances. J . Jap. Med. Assoc. 67:182. Theuer, R.C. 1983. Milk Proteins in infant nutr i t ion: development and manufacture of infant formula. Kiel Milchwirtsch Fortschr i t tsber . 35:423. Torres, A . R . , Peterson, E.A. Evans, W.H. Mage, M.G. and Wilson, S.M. 1979. Fractionation of granule proteins of granulocytes by copper chelate chromatography. Biochim. Biophys. Acta 576:385. Touloukian, R., Berdon, W., Amoury, R., and S a n t u l l i , T .U . 1967. Surgical experience with necrotizing enteroco l i t i s . J . Ped. Surg. 2:389. Troncone, R., V i ta le , M., Donatiel lo, A . , F a r r i s , E . , Rossi , G. and Auricchio, S. 1986. A sandwich enzyme immunoassay for wheat g l i a d i n . J . Immunol. Meth. 92:21. Valent i , P. , Antonini , G . , Von Hunolstein, C , V isca, P . , O r s i , N. and Antonini, E. 1983. Studies on the antimicrobial ac t iv i ty of ovotransferrin. Int. J . T i s s . Reac. 1:97. 183 Van Oss, C . J . 1982-1983. Isolation and characterization of immunoglobulins. Separation and Pur i f icat ion Methods 11:131. Walker, W.A. and Isselbacher, K . J . 1977. Physiology in medicine: intestinal antibodies. N. Eng. J . Med. 297:767. Warner, R.C. and Weber, I. 1951. The preparation of c rysta l l ine conalbumin. J . B i o l . Chem. 191:173. Webb, B .H . , Johnson, A . H . , A l fo rd , J . A . (Eds.) 1974. Fundamentals of Dairy  Chemistry Second Ed. The AVI Publishing Company, Inc. Westport CN. P. 94. Weber, K. and Osborn, M. 1969. The r e l i a b i l i t y of molecular weight determinations by dodecyl sulfate-polyacrylamide gel electrophoresis. J . B i o l . Chem. 244:4406. Weinberg, E.D. 1977. Infection and iron metabolism. Am. J . C l i n . Nut. 30:1485. Welsch, J . K . and May, J . T . 1979. Ant i - in fect ive properties of breast milk. J . Pediatr. 94:1. Wharton, B. 1981. Immunological implications of alternatives to mother's milk. In: The Immunology of Infant Feeding. Wilkinson, A.W. (Ed.) Plenum Press, New York. Whitney, R. McL., Brunner, J . R . , Ebner, K . E . , F a r r e l l , J r . H.M., Josephson, R.V. , Morr, C.V. and Swaisgood, H.E. 1976. Nomenclature of the proteins of cow's milk: Fourth rev is ion. J . Dairy S c i . 59:795. Wilkinson, A.W. 1981. The Immunology of Infant Feeding. Plenum Press, New York. Will iams, J . 1962. A comparison of conalbumin and transferr in in the domestic fowl. Biochem. J . 83:355. Will iams, C A . and Chase, M.W. 1971. Methods in Immunology and Immunochemistry v o l . 3. Academic Press, New York, NY. W i l l i s , A . T . , Bul len, C . L . , Williams, K., Fagg, C . G . , Bourne, A. and Vignon, M. 1973. Breast milk substi tute. A bacteriological study. Br. Med. J . 4:64. Workshop Part ic ipants. 1976. Human milk in premature infant feeding: Summary of a workshop. Pediatrics 57:741. World Health organization 1981. Contemporary patterns of breast feeding. WHO, Geneva. Zacharius, R.M. and Z e l l , T . E . , Morrison, J . H . , Woodlock, J . J . 1969. Glycoprotein staining following electrophoresis on acrylamide ge ls . Anal. Biochem. 30:148. Z i t t l e , C A . and Custer, J . H . 1963. Pur i f icat ion and some of the properties of a s i - c a s e i n and K-case in . J . Dairy S c i . 46:1183. 184 APPENDIX: NON-FERRIC METHODS FOR B-LACTOGLOBULIN REMOVAL FROM CHEDDAR CHEESE WHEY 185 Different approaches were tr ied to select ive ly eliminate B-Lg from cheese whey. 1. Amundson and Watanawamchakorn approach Amundson and Watanawamchakorn (1982) claimed that adjustment of the conductivity of 80% volume reduced electrodialyzed whey to 100-200 uMHOS and adjustment of the pH to 4.65 would result in 53.3% protein removal. However, when this method was t r ied on acid whey (prepared from raw milk by ac id i f ica t ion) and cheese whey after extensive d ia lys is and adjustment of the conductivity (conductivity Bridge Model 31, Yellow Springs Instruments, Co . , Inc. , Ohio) by addition of 1 M NaCl, i t did not work. 2. So lub i l i t y differences of P-Lg and Ig at di f ferent pHs The pH of 1% p-Lg or Ig was adjusted with 0.1 N NaOH or 0.1 N HC1 to the range 4-8, and the protein content of the supernatant was determined according to the method of Nakai and Le (1970) after centrifugation at 10,000 x g. The lowest s o l u b i l i t y of p-Lg was found at pH 4.7 and i t was 46% of the total p-Lg in the original solut ion. The lowest so lub i l i t y of crude Ig was found around pH 5.8 and i t was about 50% of the total Ig in the original solut ion. To check the p o s s i b i l i t y of denaturation of P - L g , two tests were performed. F i r s t , gel f i l t r a t i o n of p-Lg on TSK HW-55 column (15 x 1.5 cm, flow rate 1 mL/min, eluted with 50 mM imidazole-KCl, pH 6.5) showed no evidence of aggregation in the elution patterns of standard p-Lg used, which suggested that P-Lg was native. The second denaturation test depended on the so lub i l i t y of globular proteins at their i soe lec t r ic point in the presence of 0.2 M NaCl. It was found that more than 90% of P-Lg was soluble as shown by absorbance at 280nm under the above conditions after centrifugation at 10,000 X g. 186 To study the the effect of pH on the s o l u b i l i t y , another attempt was made to dissolve 1% B^Lg and Ig at c i t rate buffer in dif ferent pH values 4.2, 4.6, 5.2, 5.6, and 6.0, with conductivity in the range of 4000 - 8000 uMHOS. However, the results indicated that no precipi tat ion of p-Lg occurred whereas minor turbidi ty was noticed in the Ig solut ions. 3. Combination of Amundson and Watanawamchakorn and Pearce methods According to Amundson and Watanawamchakorn (1982) pH, ionic strength and the concentration of whey are the major factors influencing the precipi tat ion of B-Lg, while according to Pearce (1983) the pH, temperature and the concentration of whey are the major factors influencing the preparation of enriched p-Lg f rac t ion . In our study an attempt was made to combine these factors in addition to other factors i . e . , cys, KI and CaCl£. The ranges of the values of these factors were as follows: Ionic strength (I) = 0.0-0.2, pH = 4.5-5.0, temperature (T) = 25-50°C, total so l id (TS) = 6.5-20%, cysteine (Cys) = 0.0-0.2%, potassium iodide (KI) = 0.0-0.2%, calcium chloride (CaCl2) = 0.0-0.2%. According to Taguchi's scheme (1957) for fractional factor ia l analysis , the possible interactions were chosen to be IXpH, TXpH, IXT, KIXI, IXCa, CaXCys, IXCys. Calcium free acid whey was prepared by the addition of potassium oxalate followed by d ia lys is for 48hr against d i s t i l l e d water. A set of experiments was performed (Taguchi, 1957, L15 for two leve ls ) . Experiments were carried out in random order and SDS-PAGE was used to evaluate the separation ef f ic iency of P-Lg and a-La in each experiment. 187 ANOVA was performed (using a Monroe 1880 Calculator) for factors and interact ions, and i t was concluded that I, TS, pH, Cys, and Ca were important factors while KI and temperature were nonsignif icant. Simplex optimization was then applied for the f ive factors with lower l imits and upper l imits as follows: 1(0.0-1.00), TS (6.5-20%), pH (4.5-5.0) and (0.0-0.2 M) for Cys and Ca. Six vert ices were experimented and SDS-PAGE showed that the optimization of these factors f a i l e d , therefore we could not select ively precipitate p-Lg. In an attempt to remove Ca-phosphate from whey under alkal ine condition, an Amber!ite anion exchanger was used to adjust the pH of cheese whey to 8.0. However,the amount of Ca-phosphate removed by this process was much lower than that obtained by adjusting the pH using sodium hydroxide. 4. Precipi tat ion of B-Lg by polyethylene glycol The behaviour of solutions containing globular proteins and polyethylene glycol(PEG) has been investigated by many protein chemists (Haire et a l . , 1984). PEG provides an attract ive alternative to other agents used for protein prec ip i ta t ion . It allows regulation of protein so lub i l i t y without observable effects on protein structure and function. However the only noticable effect of PEG is the shi f t of the pKa of the ionizable groups at high PEG concentrations (Atha and Inghams, 1981). In some s i tuat ions, the use of PEG has a def ini te advantage over the use of high sal t concentrations, where sal t ing in and salt ing out phenomena and speci f ic interactions with both cation and anion complicate both the interpretation and the extrapolation to a well defined, b io logica l ly relevant state (Roth et a l . , 1979). An aliquot of forty percent (W/V) PEG was added to 20 mL of cheese whey with good mixing to give f inal PEG concentrations of 6.7, 9.2, 11.4, 13.3 and 20%. Samples were then kept at room temperature for an hour and were then 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0097117/manifest

Comment

Related Items