(A «W- v———‘

I .. .. . . T .r~., I . 4 I , .. .I \ I...
I...” . I -Tu...4' ""CIII.T.I .!,:, ,, T. _, ,.-,-y..'.._.,.u.\,., , IT‘....‘... I, A‘. , _
b‘v . . II'I.I._. ,"‘|‘an|-... ' Ju~u.~ ., I-‘n ‘ ‘ - .. ’., -

“\‘l‘I‘b-Ifi‘nptut-LAIIM. l‘xhlllhi‘: . ,T. ,H \"'I:~ u .[y , I. -.~. ,. ‘, I. , .

>3:~'_\:v.>:.\.'.-:<e;-:~;\'I;-‘.:T‘>;~:-;

 

 

V3933

w

Imam“ tbtékffl. ZII‘

muggy.”

. “WM:
“331.“:‘219'v‘.~.'_' ,1. "_: A

 

THE EFFECTS or: pH mo uracmommsns on THE
BINDING or VITAMIN 8 12m ILACTOGLOBUUN
AND ASSOCIATED PEPTIDES

Thesis for the Degree of PIT. DI
MICHIGAN STATE UNIVERSITY
JAMES KIRK
1971

 

v-‘
-(,.-

I. If? 3?; A R Y
Michigan State

‘3 University r

This is to certify that the

thesis entitled

THE EFFECTS OF pH AND ELECTRODIALYSIS ON THE
BINDING 0F VITAMIN B12 BY B-LACTOGLOBULIN

AND ASSOCIATED PEPTIDES

presented by

JAMES KIRK

has been accepted towards fulfillment
of the requirements for

Ph.D. Food Science

degree in

 

 

& Human Nutrition

@améa/a MSM' =

We: profm

Date April 7, 1971

 

0-7630

ABSTRACT

THE EFFECTS OF pH AND ELECTRODIALYSIS ON THE
BINDING OF VITAMIN B12 BY B-LACTOGLOBULIN

AND ASSOCIATED PEPTIDES

BY

James Kirk

Milk proteins have been shown to bind vitamin
812 in excess of that inherently associated with the milk
proteins. Beta-lactoglobulin (B-Lg) was chosen for this
work, since past research has shown B-Lg to exhibit the
highest vitamin 812 binding capacity of the milk proteins.

The B-Lg used in this research was prepared in a
crystalline state and its parameters of purity, as deter-
mined by chemical and physical analyses, agreed with
published values. The vitamin B12 binding capacity of
the protein was shown to be pH dependent, exhibiting its
greatest association with the vitamin in the pH range of
6.6 to 6.8 where approximately 460 Hug of vitamin B12 per
mg of protein was bound. However, following electro-
dialysis of B-Lg, no measurable binding of 312 by the
protein could be detected at pH 6.6 and 6.8.

Polyacrylamide gel electropherograms, sedimenta-
tion coefficients and chemical analyses of B-Lg and

electrodialyzed B-lactoglobulin [(E)B-Lg] showed no

James Kirk

differences between the two proteins. However, the rela-
tive hydrolysis rates of B-Lg and (E)B-Lg by trypsin and
chymotrypsin indicated a conformational change in the
(E)B-Lg which was believed responsible for the loss in
binding capacity.

At pH 9.0 B-Lg exhibited a lower binding capacity
than was measured at pH 6.6 and 6.8. This was attributed
to a reduction in polar binding sites and a change in
conformation resulting from dissociation of the protein
to its monomeric state. At pH 9.0 the (E)B-Lg and
(E)B-Lg-peptide mixture exhibited binding capacities for
the vitamin which varied with the buffer system used in
the model system. The binding of vitamin 312 by (E)B-Lg
may have resulted from the secondary binding to the
solvent ions associated with the protein monomers and/or
the dissociation of electrodialyzed beta-lactoglobulin.

Polyacrylamide gel electr0phoresis of the B-Lg
and (E)B-Lg-vitamin B12 complexes indicated that the
vitamin was only loosely associated with the protein
molecules, possibly through electrostatic bonds.

Three peptides were recovered from purified
B-Lg by electrodialysis. These peptides were shown to
be electrostatically associated to the protein and were
arbitrarily designated as "negative," "+2" and "+3,"
according to their mobility characteristics on thin-

layer high-voltage electrOpherograms. The calculated

James Kirk

molar binding ratios of peptides to B-Lg indicated a
random association between the protein and peptides.
The molecular weights of the isolated peptides were
estimated at 1,500 to 3,600 by gel filtration chroma-
tography, whereas minimum molecular weight calculations
from residue weights of the limiting amino acids gave
values of 4,000 for the negative peptide and 6,000 for
the +2 and +3 peptides. These data may be interpreted to
suggest that the peptides may havesecondary structures.
Experiments designed to assess the binding of
vitamin B12 by the peptides in model systems revealed
that, under the conditions employed, the peptides did

not bind vitamin B Therefore, it is believed that

12'
the loss of peptides from the protein was not solely

responsible for the loss of binding capacity by (E)B-Lg.

THE EFFECTS OF pH AND ELECTRODIALYSIS ON THE

BINDING OF VITAMIN B BY B-LACTOGLOBULIN

12
AND ASSOCIATED PEPTIDES

BY

q“
’— :\{1r 1 —

JamesiKirk

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Food Science

1971

ACKNOWLEDGMENTS

The author expresses sincere gratitude to Dr.

J. R. Brunner and Dr. C. M. Stine, Professors of Food
Science, for their inspiration, counsel and patience
during this study and for their assistance in the prep-
aration of this manuscript. .

Appreciation is also extended to Dr. L. G.
Harmon, Dr. H. A. Lillevik and Dr. E. J. Benne, who re-
viewed the manuscript.

Grateful acknowledgment is due to Dr. B. S.
Schweigert and the Department of Food Science for making
arrangements so that this study could be made and for
making research facilities available. Acknowledgment is
also due to Miss Ursula Koch for her help in obtaining
amino acid analyses and ultracentrifugation data.

The author expresses his most sincere apprecia-
tion to his wife, Elaine, and family who have helped in

far too many ways to enumerate.

ii

TABLE OF CONTENTS

Page
ACKNOWLEDGMENTS . . . . . . . . . . . . ii
LIST OF TABLES . . . . . . . . . . . . . Vi

LIST OF FIGURES O O O O O O O O O O O O Vii

INTRODUCTION . . . . . . . . . . . . . 1
Vitamin B12: History . . . . . . . . . . 1
Vitamin B12: Physical PrOperties . . . . . . 4
Vitamin B12: Structure . . . A. . . . . . 5
Vitamin B12: Presence in Milk . . . . . . . 5
Characteristics of Bovine Beta-Lactoglobulin . . 7

Beta-Lactoglobulin: Characteristics . . . . 7
Beta-Lactoglobulin: Effect of pH . . . . . 10
Beta-Lactoglobulin: Binding Complexes . . . l3

EXPERIMENTAL PROCEDURE . . . . . . . . . . 16

Preparative Methods . . . . . . . . . . l6
Beta-Lactoglobulin: Isolation . . . . . 16
Beta-Lactoglobulin: Recrystallization . . . 16

Chemical Methods . . . . . . . . . . . l6
Nitrogen . . . . . . . . . . . . . l6
Tryptophan . . . . . . . . . . . . . l6
Sulfhydryl . . . . . . . . . . . . . l7
Amino Acid Analyses . . . . . . . . 17
N- -Terminal Amino Acid Analyses . . . . . . l7
Colorimetric Determination of Protein . . . . 17

Physical Methods . . . . . . . . . . 17
Method of Electrodialysis . . . . . . . . l7
Polyacrylamide Gel Electrophoresis . . . . . 19
Staining Polyacrylamide Gels . . . . . . . 23
High-Voltage Electrophoresis . . . . . . . 23
Gel Filtration Chromatography . . . . . . 26
Determination of Protein Binding Capacity . . 27
Determination of Peptide Binding Capacity . . 28
Enzymatic Hydrolysis of Proteins . . . . . 30
Ultracentrifugation . . . . . . . . . . 30

iii

RESULTS 0 O O O O O O O O O O O O O O O

Beta-Lactoglobulin . . . . . . . . .
Nitrogen Content . . . . . . . . .
Sulfhydryl Content . . . . . . . .
Gel ElectrOphoresis of Beta-Lactoglobulin
Sedimentation Coefficients . . . . .

Electrodialyzed Beta-Lactoglobulin . .
Sulfhydryl Content . . . . . . .
Gel Electrophoresis . . . . . . .
Sedimentation Coefficient . . . .

Amino Acid Analyses of Beta- -Lactoglobu1in and
Electrodialyzed Beta-Lactoglobulin . . . . . .

N- Terminal Amino Acids of Beta-Lactoglobulin and
Electrodialyzed Beta-Lactoglobulin . . . . .

Rate of Enzymatic Hydrolysis of Beta-Lactoglobulin
and Electrodialyzed Beta-Lactoglobulin . . . .

Paper High-Voltage ElectrOphoresis of the Tryptic
Digest of Beta-Lactoglobulin and Electrodialyzed
Beta-Lactoglobulin . . . . . . . . . . .

Peptides from Beta-Lactoglobulin . . . . . . .

Separation and Recovery of the Peptide Mixture . .

Physical and Cehmical Properties of Isolated
Peptides . . . . . . . . . . .

Minimum Molecular Weights of Isolated Peptides . .

Amino Acid Analyses of Isolated Peptides . . . .

Separation of Unbound CoGOBlz from Protein and
Peptide-Bound C050B12 Using Gel Filtration . . .

Vitamin B12 Binding Capacity of Beta-Lactoglobulin
and Electrodialyzed Beta-Lactoglobulin . . . .

Discontinuous Polyacrylamide Gel Electrophoresis of
Beta-Lactoglobulin and Electrodial zed Beta-
Lactoglobulin Equilibrated with Co 7B12 . . . .

Vitamin B Binding Capacity of Peptides Isolated by
Electroéialysis of Beta-Lactoglobulin . . . . .

DISCUSSION 0 O I O O O O O O I O O O O 0

Association of Vitamin B12 with Peptides . . .
Association of Vitamin B12 with Beta- Lactoglobulin .
Vitamin B12 Binding at pH 6. 6- 6. 8 . . . . . .
Vitamin B12 Binding at pH 9. 0 . . . . . . .
Vitamin Bl; Binding at pH 2.0 . . . . . . .
Effects of Electrodialysis on Beta-Lactoglobulin . .
Association of Vitamin B12 with Electrodialyzed Beta-
Lactoglobulin and Electrodialyzed Beta-
Lactoglobulin Peptide Mixture . . . . . . .
Vitamin B12 Binding at pH 6.6-6.8 . . . . . .
Vitamin B12 Binding at pH 9.0 . . . . . . .

iv

Page

43
49
52

53
53
55

57
59

62
62
72

72
78
78
79
80
81

83
83
84

Page

Mode of Association Between B12 and Beta-

Lactoglobulin . . . . . . . . . . . . . 88
Peptides Isolated from Beta-Lactoglobulin . . . . 88
Estimated Molecular Weight of Peptides . . . . . 90
Mode of Binding Between Peptides and Beta-

Lactoglobulin . . . . . . . . . . . . . 92

SUMMARY AND CONCLUSIONS . . . . . . . . . . . 94
LITERATURE CITED . . . . . . . . . . . . . 97

APPENDIX 0 O O O O O O O O O C O I O O O 107

Table

1.

LIST OF TABLES

Page

The apparent sedimentation constants of B-Lg

in 0.1DI potassium chloride and borate

buffer at 20 C . . . . . . . . . . 37
The apparent sedimentation constants of

(E)B-Lg in 0.1M ammonium acetate and borate

buffer at 20 C . . . . . ,. . . . . 39
Amino acid composition of B-Lg and (E)B-Lg . 42
Amino acid composition of isolated peptides

from B-Lg by electrodialysis . . . . . 56
Elution pattern of C060B12 from Bio-Rad P—2

gel filtration column versus counts per

minute of 5 ml eluate fractions . . . . 58
Vitamin B12 binding capacities of B-Lg,

(E)B-Lg and (E)B-Lg recombined with the

peptide mixture in various buffers and pH

values 0 O O O I I O O O C O O O 61
Vitamin B12 binding capacities of the peptides

after separation of unbound C057B12 from the

peptides by thin-layer high-voltage

electrOphoresis . . . . .. . . . . . 65

vi

LIST OF FIGURES

Figure

l. Polyacrylamide gel electropherogram of
recrystallized B-Lg and whey standard at
pH 8.3 O C O O O O O O O O O O O

2. Polyacrylamide gel electropherogram of
recrystallized B-Lg and B-Lg A, B and AB
standards at pH 8.3 .. . . . . . . .

3. Polyacrylamide gel electrOpherogram of B-Lg
and (E)B-Lg at pH 8.3 o o o o o o o o

4. Hydrolysis of B-Lg and (E)8-Lg with trypsin .

5. Hydrolysis of B-Lg and (E)B-Lg with
chymotrypsin . . . . . . . . . . .

6. Rate curves for tryptic hydrolysis of B-Lg
and (E)B-Lg per 60 seconds . . . . . .

7. Rate curves for chymotryptic hydrolysis of
B-Lg and (E)B-Lg per 60 seconds . . . .

8. High-voltage paper electropherogram of the
tryptic hydrolysate of B-Lg and (E)B-Lg in
pyridine-acetic acid-water buffer, pH 3.5 .

9. Elution pattern of electrodialyzed peptides
from Bio-Rad P-2 column with deionized water

10. Thin-layer high-voltage electropherogram of the
peptide mixture electrodialyzed from B-Lg in
pyridine-acetic acid-water buffer, pH 3.5 .

ll. Elution pattern of electrodialyzed peptides
from Bio-Rad P-4 column with 0.1M NH4OAc .

12. Verticle polyacrylamide gel electrOpherogram of
B-Lg and (E)B-Lg at pH 8.3 o o o o o 0

l3. Thin-layer high-voltage electropherogram of the
peptide mixture equilibrated with C057B12 .

vii

Page

34

35

40
44

45

46

47

48

50

51

54

63

66

Figure Page

14. Thin-layer chromatographic separation of
unbound C057B12 from the electrodialyzed
peptide mixture . . . . . . . . . . . 67

15. Thin-layer chromatographic separation of
unbound C057B12 from the electrodialyzed +2
peptide . . . . . . . . . . . . . 68

16. Thin-layer chromatographic separation of unbound
C057B12 from the negative and +3 peptides . . 69

17. Thin-layer chromatographic separation of unbound
C057Bl2 from the peptide mixture . . . . . 71

viii

INTRODUCTION

Vitamin B12 is unique among the vitamins because it
is not synthesized by higher plants and animals. Wherever
it is found in nature, its origin can be traced back to
bacteria or other micro-organisms, growing in soil, water
or in the rumen or intestine of some animals.

Humans are entirely dependent on dietary B12, since
their intestinal flora either do not synthesize 312’ or it
is not released from the cells of synthesizing organisms
in the region of the gut from which absorption occurs.

In view of this metabolic deficiency, nature has
provided for the initial vitamin B12 needs of the mammalian
offspring by its presence in the mammary secretions.

In the bovine, the vitamin B12 activity of milk is
attributed entirely to cobalamin because of a highly
selective absorption in the stomach. It is also known that
95% of the naturally occurring vitamin B12 in cow's milk is
associated with the milk proteins; however, these protein-
B12 complexes do not exhibit intrinsic factor activity.

Beta-lactoglobulin (B-Lg), the major whey protein
in milk, shows the highest vitamin B12 content of the milk
proteins. In addition to the B12 inherently associated to

B-Lg, additional B can be bound by the B-Lg when the two

12
components are equilibrated in model systems.
1

Previous research indicated that the major factor
involved in the binding of vitamin B12 by B-Lg was the
association of B12 with peptides bound to the B-Lg.

It was the intention of this research, with the
aid of previously published information, to investigate
the mode of vitamin B12 binding with the peptides and the

peptide carrier protein.

LITERATURE REVIEW

Vitamin B12: History

 

In 1920 Whipple gt_§l. found that feeding liver
to dogs made anemic by bleeding accelerated the regenera-
tion of their red blood cells. Minot and Mulphy (1926)
followed this discovery with the observation that liver
contains a nutritional element required for the cure of
pernicious anemia. Investigations were immediately begun
to determine the identity of the "anti-pernicious anemia
factor" present in liver. Early attempts to purify the
anti-pernicious anemia factor were greatly hampered by
the minute quantities present in liver, and the consequent
need to use pernicious anemia cases in relapse to assay
the liver extractions.

In 1948 crystalline vitamin B12 (5,6 dimethyl-
benzimidazolylcobamide cyanide) was identified as the anti-
pernicious anemia factor by Rickles gt_§l. (1948), and
Smith and Parker (1948).

Proof that cyano B functions in biological

12
reactions is credited to Barker et a1. (1958), and Barker
et a1. (1960) folloWing the isolation of deoxy-adenosyl-
312’ Vitamin B12, itself, has not been identified as an

active cofactor for any known enzymatic reaction. Rather,

3

a related family of analogs, the B12 coenzymes, have been

shown to function in this capacity.

Vitamin B12: Physical PrOperties

 

Cyanocobalamin crystallizes as dark red needles or
prisms with a melting point of 300 C. Whether cyano-
cobalamin is crystallized from aqueous acetone or from
water, a considerable but variable amount of loosely held
water (usually 10 to 12 per cent) is associated with the
crystals. This can be removed by heating under reduced
pressure. Vitamin B12 is soluble in water up to 1.25 per
cent, relatively soluble in lower alcohols and phenols,
but insoluble in most other organic solvents. Aqueous
solutions of cyanocobalamin are neutral and show absorption
maxima at 278, 361 and 555 nm, which do not shift markedly
with changes in pH as shown by Hodgkin gE_gl. (1949).
Electrometric titrations and conductivity measurements of
Hodgkin et_gl. (1948) also have shown the absence of any
strongly ionizing groups.

Brink §E_al, (1949) determined the molecular
weight of cyanocobalamin, by the boiling point method, to
be 1,490 i 150; these data agree with the molecular weight
of 1,360 to 1,575 as reported by Hodgkin gt_31. (1949)
from their x-ray crystallography data. The variance of the
molecular weight is attributed to the variable hydration of

the B12 crystals.

Vitamin B12: Structure

 

The chemical structure of cyanocobalamin, as it is
isolated from the liver, was identified by Hodgkin gt_§l.
(1955). The cyanocobalamin molecule can be divided into
two major portions: the four reduced pyrrol rings forming
the macro "corrin ring" and the nucleotide, which unlike
the nucleotides of nucleic acids, contains 5,6 dimethyl-
benzimidazole as its base with ribose-3-phosphate linked
by an a-glycosidic bond. The corrin ring contains tri-
valent cobalt chelated to the four nitrogen atoms of this
ring, a nitrogen atom of the 5,6 dimethylbenzimidazole
ring, and a cyanide ion, the latter being an artifact of
the isolation procedure (Smith et_al., 1952).

Other functional moieties of the cyanocobalamin
molecule are: l-amino-Z-prOpanol to which the 2-OH group
of the phosphate is esterified, and propionic acid.

Cobalamin coenzymes, the functional forms of
vitamin B12, differ structurally from cyancobalamin in
that the cobalt is divalent in nature and the cyanide group

is replaced by a molecule of 5-deoxyadenosine.

Vitamin 312: Presence in Milk

The average vitamin B12 content of bovine milk has
been reported by Collins et a1. (1951) and Gregory (1954)
to be 6.6 micrograms B12 per liter, with a range of 2.0 to

24.0 micrograms B12 per liter. Findings of several

workers indicate the vitamin B12 content of milk did not
vary significantly with breed, season of the year, or
stage of lactation (Collins et_gl,, 1951; Collins gt_gl.,
1953; Hartman $11., 1955; and Gregory 3331., 1958).

However, heating does decrease the vitamin B 2 content in

1
milk, depending upon the amount of heat and the method of
processing employed (Ford, 1957; and Kon, 1961).

Gregory (1954) and Gregory and Holdsworth (1955a

and 1955b) reported nearly all of the B present in milk

12
was in the cobalamin form and that in sow's milk the B12
is bound to peptides and proteins, thus making it unavail-
able for test organisms. These data were confirmed by

Kim gt_al. (1965), who reported approximately 95% of the
total vitamin B12 content of cow's milk associated with
the milk proteins. Gizis §E_al. (1965) determined the
binding capacities of the individual milk proteins, in

pH 9.0 borate buffer, using Lactobacillus leichmannii

and COGOB

 

. They found B to be ubiquitously distributed

12 12
among the various milk proteins, with higher concentra-
tions associated with the whey proteins. Gizis gt_al.
(1965) also reported the isolation of two peptides (3,000
and 9,000 M.W.) from dialyzed cow's milk by an application
of electrodialysis, which absorbed vitamin B12 in the

order of 1 x 105 micromicrograms (pug) B12 per mg of

protein.

Although milk has been shown to have a considerable
vitamin B12 binding capacity, Ungley and Childs (1950) have
shown that these vitamin-protein complexes do not exhibit
intrinsic factor activity.

Characteristics Of Bovine
Beta-Lactoglobulin

 

 

Beta-Lactoglobulin:
Characteristics

 

 

B-Lg, the major whey protein of mature bovine milk,
is synthesized in the epithelial cells of the mammary gland
(Twarog and Larson, 1962) and corresponds to about 7 to 12%
of the total milk protein. B-Lg was first isolated and
crystallized by Palmer (1934). In 1946 Li indicated the
possibility that B-Lg from mixed milk was a heterogeneous
protein. Aschaffenberg and Drewry (1955 and 1957) were
the first to distinguish genetic variants A and B in B-Lg
by their difference in electrophoretic mobilities. Later,
Bell (1962) demonstrated the presence of a third genetic
variant, B-Lg C, in the milk of Jersey cows. Kiddy gt_al.
(1965) and Townend and Basch (1965) have shown that the
presence of B-Lg A, B and C is genetically controlled by
codominant autosomal alleles LgA, LgB and Lgc. Kiddy
gt_§l. (1965) demonstrated that cows can be homozygous or
heterozygous in their production of B-Lg, and heterozygous
cows give approximately equal amounts of the phenotypic

B-Lg. The genetic variance of B-Lg has been shown to

manifest itself in the amino acid composition of the
protein. Gordon gt_gl. (1961) and Piez gt_al. (1961)
found B-Lg A to differ from B-Lg B by the substitution
of two aspartic and two valine residues for two glycine
and two alanine per 35,500 molecular weight. Beta-
lactoglobulin C has been differentiated from B-Lg A or B
by two additional histidine residues and two less
glutamine residues per 36,000 molecular weight (Kalan
g£_§l, (1965). Kalan gt_§l. (1965) also determined by
partial amino sequencing that the C-terminal (leucine)
and N—terminal (isoleucine) residues of the three genetic
species were the same. Thus genetic amino acid substitu-
tion does not occur at, or near, the terminal portions of
the peptide chain.

Values reported for the molecular weight of
native B—Lg range from 35,000 to 37,000 (Pedersen, 1936;
Bull and Currie, 1946; Ogston and Tilley, 1955; Green and
Aschaffenburg, 1959; Wirtz gt_al., 1964; and Aschaffenburg
gt_gl., 1965). The molecular weight generally assigned
to native B-Lg is 36,000 at a pH between 5.5 and 7.5
where associative and dissociative reactions are at a
minimum. This corresponds to a sedimentation coefficient
of 2.85 S.

The first evidence that native B-Lg is a molecular
dimer composed of two 18,000 molecular weight polypeptides

is attributed to Bull (1946) and Bull and Currie (1946).

Later studies by Timascheff (1964), Townend and Timascheff
(1957), Green and Aschaffenburg (1959), Aschaffenburg
gt_gl. (1965) and Wirtz et_gl. (1964), using small-angle
x-ray scattering and ultracentrifugation, have shown that
native B-Lg at pH 5.7, where minimum aggregation occurs,
is composed of two identical spherical subunits (molecular
weight 18,000) with a radius of 17.0A, impinging on each
other by 2.3A at their surface contact. Timascheff (1964)
has reported the two subunits form a stable dimer as a
result of hydrOphobic interactions.

Nozaki et_al. (1959) showed the isoionic point
of mixed B-Lg in pure water as pH 5.39, while B-Lg A and B
exhibited isoionic points of 5.13 and 5.14 respectively.
Nozaki §E_al. (1959) pointed out that these values were
lowered as potassium chloride or calcium chloride was
added, owing to the binding of potassium and calcium,
which could suggest centers of unusually high negative
potential on the isoionic B-Lg.

Confirmational studies of B-Lg have been a center
of great controversy over the last ten years. Urnes and
Doty (1961) and McKenzie (1967) have offered the best
interpretation of Optical rotary dispersion data, con-
cluding that B-Lg is composed of approximately 0.35 alpha-
helix, 0.34 beta-conformation and 0.33 disordered form.
Tanford's review (1962) of the titration curves of native

B-Lg have shown that the native protein appears to contain

10

six imidazole groups per 36,000 molecular weight, compared
to four for the analytical figure. Investigation of this
phenomenon revealed that two carboxyl groups of the native
protein are titrated with pH characteristics of imidazole
groups because they are buried in the hydrophobic interior
of the protein molecule, where dimer attraction forces are
known to exist.

Beta-Lactoglobulin:
Effect of_pH

 

 

Below the Isoelectric Point of Beta-Lactoglobulin.--

 

Townend and Timascheff (1957) were the first to show the
dissociation of mixed B-Lg into monomeric subunits at pH
values below 3.5. Nozaki et_al. (1959) have eXplained the
dissociation of B-Lg as nonspecific electrostatic repulsion
resulting from a progressive increase in positive charge
as the pH is lowered. McKenzie and Sawyer (1966) reported
that the extent of protein dissociation is dependent upon:
pH, protein concentration and genetic variant. Townend
eE_§l. (1960) studied the reversible dissociation below
pH 3.5, using ultracentrifugal and light-scattering
techniques. They reported the subunits to be identical
and corresponded to the two polypeptide chains of the
protein dimer. Townend et_al. (1961) using radioactively
labeled B-Lg A and B showed that no hybrid B-Lg dimers are
formed when a mixture of the two genetic variants are

acidified and reneutralized. This would indicate a

11

specific structural difference in the area of subunit
contact of the genetic variants. However, Basch and
Timascheff (1967) found no difference in the titration
curves or maximum hydrogen ion binding capacity (40 H+
per molecule) of B-Lg A, B, and C.

Above pH 4.5 the titration curves of B-Lg are
dependent upon the genetic variant and can be accounted
for in terms of normal ionization of all groups, with the
exception of two histidine residues in the C variant and
the two hidden carboxyl groups in all three variants.
These data in conjunction with other physical prOperties
of B-Lg suggest that sufficient attractive forces exist
between ionizable groups to cause the genetic variants of
B-Lg to behave differently between pH 3.5 and 5.2.

Townend et_al. (1960) using ultracentrifugation
and electrOphoresis, Kumosinski and Timascheff (1966)
using x-ray crystallography, and McKenzie and Sawyer (1967)
using Optical rotary dispersion, have shown B-Lg A, B and
AB to tetramerize (a cubic array of eight dimer Spheres
(molecular weight 144,000) in the pH range of 3.5 to 5.2
with maximum association occurring at pH 4.6. McKenzie
(1967) found that tetramerization involves hydrogen bonding
of the carboxyl groups in the aspartic-glutamic rich areas
of the protein dimers, which in the case of the A variant
is enhanced by the additional aspartic residues. Townend

(1965) reported that approximately 30 carboxyl groups of

12

the native protein are buried by the formation Of the
tetramer.

At present no associative interaction has been
observed for B-Lg C in the pH range 3.5 to 5.2 (Bell and

McKenzie, 1964 and 1966).

Above the Isoelectric Point Of Beta-Lactoglobulin.--

 

Wirtz gt_gl. (1964) have carried out small-angle x-ray
scattering investigations Of B-Lg at pH 5.7 and found no
evidence Of molecular aggregation. The first evidence Of
physical change in B-Lg above its isoelectric point was
reported in the original ultracentrifugation studies Of
Pedersen (1936). Tanford et_al, (1959) observed a change
in Optical rotation while studying the titration curves
of B-Lg. They Observed a reversible change in configura-
tion near pH 7.5 at 25 C which parallels the titration of
imidazole groups and the two buried carboxyl groups from
the interior of the molecule. Tanford gt_§l. (1959)
regarded this reversible change in Optical rotation as a
refolding of part Of the polypeptide chain rather than an
unfolding of the protein molecule. Basch and Timascheff
(1967) support these data with their hydrogen ion equi-
librium experiments on B-Lg A, B and C.

Georges e£_al. (1962) have shown that at alkaline
pH (above 8) the B-Lg dimers are reversibly dissociated

into their monomeric subunits.

13

Studies by McKenzie and Sawyer (1966 and 1967)
involving zone electrOphoresis and Optical rotary disper-
sion have also shown a non-reversible, time dependent,
conformational transition in B-Lg at pH 8.6 and above.
The reaction involves polymerization to B-Lg polymers Of
200,000 molecular weight. The association reaction was
believed to involve oxidation Of sulfhydryl grOups and
rupture Of disulfides, since polymerization was nullified
in the presence of N-ethylmaleimide. This would support
the finding Of Dunhill and Green (1965), who observed
the N conformation Of B—Lg at low pH shielded the sulf-
hydryl groups, while the transition to the R-state in the
region of pH 7.4, caused by the release of the two hidden
carboxyls, made the sulfhydryls more accessible by the
refolding of the polypeptide chain.

Beta-Lactoglobulin:
Binding Complexes

 

Various workers have reported the binding Of non-
protein anions to B-Lg and modified B-Lg. Ray and
Chatterjee (1967) studied the binding Of methyl orange
and dodecyl-pyridine to B-Lg by equilibrium dialysis.

Ray (1968) and Seibles (1969) have demonstrated the
binding of dodecylsulfate to B-Lg, showing a complex of
two moles of dodecylsulfate per mole of protein. All
research concerning the binding Of anions to B-Lg has
shown the binding capacity Of the protein to increase by

a factor of 3-10 as the pH is raised from its isoionic

14

point to pH 7.5. However, no new binding sites were
exposed by the conformation change occurring at pH 7.5
(Ray and Chatterjee, 1967). The exact nature of the
binding sites on the protein is as yet undetermined,
although Seibles (1969) suggested the influence Of
histidine residues as a possible factor.

B-Lg has also been shown to bind inorganic
cations. Carr (1953) found B-Lg to exhibit a binding
capacity for calcium ions, which increased with a rise in
pH. These findings were supported by the work Of Zittle
gt_al. (1957). King et_al. (1959) showed the COpper and
iron inherently associated with skimmilk protein was not
affected by a decrease in pH from 6.7 to 3.0. However,
the ability Of the skimmilk and whey proteins to bind
additional copper and iron decreased as the pH was lowered
from 6.7 to 3.0. Aulakh (1967) reported a binding capacity
Of 2.5 moles Of COpper per mole Of B-Lg at pH 6.5. Barker
and Saroff (1965) have shown that B-Lg, one of the few
milk proteins that binds sodium ions, exhibits no binding
of the monovalent cations at its isoelectric point (IEP).
However, the binding Of sodium ions is known to increase
as the pH is raised above the IEP of B-Lg and reaches a
maximum at pH 9.48. Baker and Saroff (1965) have also
indicated that the binding Of sodium and the configuration
changes associated with B-Lg appear to be controlled by

the same ionization reaction.

15

Other B-Lg binding complexes have been reported.
Kim et a1. (1965) have reported B-Lg releases 203 micro-

micrograms vitamin B per mg Of protein, when autoclaved

12
in the presence Of acid and cyanide. In a later report
Gizis gt_31. (1965) indicated that native milk proteins
do not exhibit their maximum vitamin B12 binding capacity.
They Observed B-Lg to bind a total Of 850 ng/mg Of protein
Of vitamin B12 in borate buffer, pH 9.0. Dorris (1968)
reported that electrodialysis Of B-Lg released a peptide
mixture with an average vitamin B12 binding capacity of
5,410 pug/mg Of protein.

Recently Ford et_al. (1969) have confirmed the
presence of a reversible folate-B-Lg complex in milk,
which exhibited a pH dependence. At pH values of 8.8,
7.1 and 6.0 the folate was wholly associated with the B-Lg.
At pH 5.0 only 61 per cent was bound and at pH 3.6 all

folate was dissociated from the protein.

EXPERIMENTAL PROCEDURE

Preparative Methods

 

Beta-Lactoglobulin: Isolation

 

The isolation of B-Lg was performed essentially
as described by Aschaffenburg and Drewry (1957) with minor
modifications. Details are presented in the appendix.

Beta-Lactoqlobulin:
Recrystallization

 

Details are presented in the appendix.

Chemical Methods

 

Nitrogen

 

Nitrogen determinations were performed using a
micro-Kjeldahl technique. Details are presented in the

appendix.

Tryptophan

 

TryptOphan, an acid labile amino acid, must be
determined apart from the rest Of the amino acids.
TryptOphan was determined spectrOphotometrically as
described in Procedure W of Spies (1967). Details are

presented in the appendix.

16

l7

Sulfhydryl

 

The method used for the determination Of sulf-
hydryl groups was that Of Ellman (1959). Details are

presented in the appendix.

Amino Acid Analyses

 

Amino acid analyses were carried out on 22 h
hydrolysates of the B-Lg, (E)8-Lg, and isolated peptides,
employing a Beckman Amino Acid Analyzer Model 120 C
according to the method Of Moore, Speckman and Stein

(1958). Details are presented in the appendix.

N-Terminal Amino Acid Analyses

 

The n-terminal amino acids of B-Lg and (E)B-Lg
were determined as dansyl chloride derivatives (Gray,
1967). Details are presented in the appendix.

Colorimetric Determination
of Protein

 

 

Protein determinations were made using the Folin-
Lowry procedure described by Layne (1955). Details are

presented in the appendix.

Physical Methods

 

Method Of Electrodialysis
The equipment employed for electrodialysis con-
sisted Of an ice bath to insure low temperature during

electrodialysis, a power source reading voltage and

l8

amperage, and the electrodialysis cell. The electro-
dialysis cell consisted Of three parts: a glass
cylinder (45 cm in length and 4.8 cm in diameter) to
hold the dialysis water; a U-shaped glass rod to form
an internal frame for the dialysis membrane; and two
platinum foil electrodes.

Visking cellulose membrane was used as the semi-
permeable dialysis membrane and was treated with
ethylenediaminetetraacetic acid (EDTA) and thoroughly
washed with deionized water prior to use in the electro-
dialysis experiments. The membrane was knotted at one
end and slipped over the U-shaped glass frame, which
served as a place of stable attachment for the platinum
foil electrode and ensured the membrane was held apart.

The B-Lg to be electrodialyzed was suspended in
deionized water and poured into the dialysis sac. .The
membrane being held by the frame was positioned in the
glass cylinder, which was filled with deionized water.
The second platinum electrode, also supported by a glass
frame, was positioned in the dialysis water outside the
membrane and the entire electrodialysis cell was placed
in an ice bath. To ensure even protein suSpension during
electrodialysis a glass stirring rod was placed inside-
the membrane and attached to a variable-Speed Lightning
Mixer. The power source was connected to the platinum

electrodes of the dialysis system with the electrode in

19

the dialysate serving as the cathode and the electrode
inside the membrane serving as the anode.

Two hundred volts Of direct current were applied
to the system. Initially, there was zero amperage.
However, as charged particles began diffusing through the
membrane, the current began to increase and reached a
maximum. At maximum amperage, the dialysate water was
changed and the electrodialysis continued-. The current
maximum was attributed to a charge equilibrium across
the membrane. Therefore, to force the dissociation
reaction, it was necessary tO change the water whenever
the amperage reached a maximum. The dialysate water was
changed until little or no amperage could be detected in
the system. This implies that few if any charged particles
were being transferred across the dialysis membrane to
complete the electrical circuit.

The electrodialysate was concentrated by per-
vaporation in EDTA treated cellulose membranes to approxi-

mately 200 m1, shell-frozen and lyophilized.

Polyacrylamide Gel ElectrOphoresis

The procedure for preparing and performing dis-
continuous polyacrylamide gel electrophoresis was
essentially similar to that described by Melachouris
(1969). The discontinuous flat-bed gel system required

two solutions: a running gel solution and a spacer gel

20

solution. The running gel solution (9% gel) was prepared
by dissolving 45 g Cyanogum 41 in 0.38014 tris-
hydroxymethylaminomethane (TRIS)-HC1 buffer, pH 8.9, and
made to a volume of 500 ml. TO this solution 0.5 m1 of N,
N, N', N'—tetramethylethylenediamine (TEMED) was added.
The spacer gel solution (5% gel) was prepared by adding
25 g Of Cyanogum 41 to 0.062M. Tris-HCl buffer, pH 6.7,
and made to a volume Of 500 ml. TO the latter solution,
0.5 m1 TEMED was added. The running and spacer gel
solutions were stored at 5 C and brought to room tempera-
ture before use.

The running and spacer gels were poured into a
flat Plexiglas gel bed (26 x 12 x 0.4 cm) as described.

A Plexiglas spacer was placed 15 cm from one end of the
gel bed to form a front for the running gel. The large
area of the gel bed was filled with 190 ml Of running gel
solution containing 2 m1 Of freshly prepared 10% (w/w)
ammonium persulfate solution. The running gel was poly-
merized under a nitrogen atmosphere.

The divider was removed and excess moisture was
blotted from the gel bed. The smaller vacated area Of
the gel bed was filled with 90 ml of spacer gel solution
containing 1 ml of a 10% ammonium persulfate solution. A
slot former was placed in the spacer gel solution, 0.5 cm
from the interface with the running gel. The spacer gel

was also polymerized under a nitrogen atmosphere.

21

Following polymerization of the polyacrylamide gel,
the slot former was carefully removed and the gel covered
with Saran Wrap to prevent drying of the gel bed.

The protein samples were dissolved in spacer-gel
buffer at a concentration Of 2%, and BromOphenol Blue
added to each sample as a marker dye. Varying amounts Of
protein solution ranging from 15 tO 35 ul, were applied to
the gel slots which were covered with Plexiglas cover.
Again the entire gel was covered with Saran Wrap to reduce
evaporation. The gel was connected to the buffer tanks
by gel-filled legs which rested in the electrode vessels,
each of which was filled with approximately 1600 m1 of
0.046M. Tris-glycine buffer, pH 8.3, and fitted with
platinum electrodes. Electrophoresis was carried out at
approximately 15 C at a voltage of 180 to 200 volts
(Heathkit Power Supply) until the buffer front had migrated
13 cm from the sample slots.

A modification Of the Melachouris (1969) procedure
was used for verticle polyacrylamide gel electrOphoresis
in an E—C Verticle Gel ElectrOphoric Chamber. ‘The gel
bed was formed with 7% Cyanogum 41 in 190 m1 0.380M
Tris-HCl buffer pH 8.9 containing 2 ml Of freshly prepared
10% (w/w) ammonium persulfate. The buffer tanks were
filled with 0.046M Tris-glycine buffer at pH 8.3 and were

fitted with platinum electrodes.

22

The protein samples were dissolved in running-
gel buffer at approximately 2% protein concentration.
Sucrose was added to the protein samples to insure a
density differential so that they could be introduced
into the sample slot without diffusing into the buffer.
BromOphenol blue was added to the samples as a marker dye.
Electrophoresis was started with 40 mA of current until
samples entered the gel (15 minutes); then increased to
80 mA (190-200 volts) and electrophoresis continued until
“the buffer front had migrated approximately 13 cm from
the sample slots.

The technique of Melachouris (1969) was also
adapted to polyacrylamide disc gel electrOphoresis. B-Lg
and whey standards were dissolved in Spacer gel buffer
at approximately 2% protein concentration. Sucrose was
added to the protein samples to insure a density differ-
ential.

The samples were electrOphoresed at 2 mA per disc.
Discs were removed at 10 min intervals during the 30 min
electrophoretic run.

Discs removed at 10 to 20 min intervals were
stained with ninhydrin reagent (see high-voltage
electrOphoresis, p. 24) to check for the presence of

fast moving peptides.

23

Staining Polyacrylamide Gels

Upon completion of the electrophoretic runs, the
polyacrylamide gels were removed from their Plexiglas
frames or glass discs and stained for 10 min in a dye
solution consisting Of 250 ml Of deionized water, 250 ml
Of methanol, 50 ml of glacial acetic acid and 5 g Of
Amido Black 10 B (napthol blue black). The excess dye
in the gel was removed in an electrolytic destaining cell
containing 7% acetic acid solution.

The electrOpherograms were photographed with a

Polaroid MP-3 Camera.

High-Voltage Electrophoresis

 

Papg£.--The high-voltage paper electrOphoretic
(HVPE) technique employed was a modification of the
technique originally described by Smith (1960). HVPE is
ideal for the separation of small molecular components,
with rapid diffusion rates and a net charge. Peptides
meet these requirements.

Peptides isolated by the electrodialysis Of 4X
recrystallized B-Lg were streaked on Whatman 3 MM filter
paper (20 x 56.5 cm) in 30 ul aliquots. The electro-
phoretic apparatus consisted Of a Plexiglas cell in which
the filter paper could be draped Over a horizontal sup-
porting bar, dipping each end Of the filter paper into
separate buffer compartments. The buffer tanks were

filled with 500 ml Of a volatile buffer (pyridine-acetic

24

acid-water, P l:AA 10:H O 189, pH 3.5). Varsol (mineral

2
spirits) was added to the Plexiglas cell to overlay the
buffer and completely cover the paper. The electro-
phoretic chamber was fitted with two glass cooling coils
for circulating water, which cooled the Varsol and
minimized heat build-up on the filter paper during
electrophoresis.

A field strength of 100 volts/cm was applied to
the system for l h. Upon completion of the electro-
phoretic run, the filter paper was removed from the
electrOphoretic chamber and dried at 90 C in a forced-
air oven. The peptides were located by spraying the
paper with ninhydrin reagent (1 g ninhydrin, 700 m1
absolute ethanol, 29 m1 of 2,4,6-trimethylpyridine and
210 ml glacial acetic acid) and drying at 90 C until the
spots became visible. The HVPE chromatograms were also
sprayed with anisaldehyde-sulfuric acid reagent according
to the procedure of Stahl and Kaltenbach (1961) to

determine the presence of carbohydrates.

Thin-layer.--High-voltage electrophoresis was used

 

primarily as a preparative procedure for separating the

isolated peptide mixture into its components. Subsequent
separations were performed on thin-layer cellulose plates,
rather than paper, to facilitate the recovery of peptides.

Thin-layer plates were prepared by suspending 15 g of MN

300 cellulose in 90 m1 of deionized water and homogenizing
the suspension in a Waring Blender. The slurried cellulose
was applied to 20 x 20 cm thin-layer plates at a thickness
of 500 microns with a Desaga Brinkmann thin-layer spreader.
The plates were air-dried for 12 h. No activation of the
MN-300 cellulose plates was required before TLHVE if the
ambient relative humidity Was under 75%.

Approximately 600 pl of the peptide mixture
(peptides dissolved in 0.1M ammonium acetate) were
applied to the thin-layer plates with Cordis disposable
microapplicators and thoroughly dried. The plates were
placed in a Reco Model E-800-2 water-cooled electro-
phoretic migration chamber and wicked to the buffer tanks
with filter paper. The plates were Sprayed with the same
buffer used in the carbon electrode buffer tanks, i.e.,
pyridine-acetic acid-water, pH 3.5 (see paper high-voltage
electrophoresis). A field strength of 40 V/cm was applied
to the thin-layer plates for 25 min. Upon completion of
electrOphoresis, the plates were dried at 90 C in a
forced air over. Location of the peptides was accom-
plished by spraying narrow strips (1.5 cm wide) on the
plate edges in the direction of migration with ninhydrin
reagent (see paper high-voltage electrophoresis). The
plates were dried at 90 C until the peptide spots were
evident. A ruler was used to mark off the areas contain-
ing the individual peptide streaks and the cellulose layer

was scraped from these areas.

26

The peptides can be eluted from the MN-300
cellulose with various solvent systems. However, 5%
acetic acid solution was the most suitable for this
purpose. The eluted peptides were shell-frozen, lyophi-

lized and stored in a dessicator over CaSO4 until needed.

Gel Filtration Chromatography

 

Gel filtration chromatography, using Sephadex
G—25 Fine, exclusion limit 5,000 M.W., was used to separate

unbound vitamin B from the B—Lg and (E)B-Lg vitamin B

12 12

complexes.

The Sephadex G-25 was rehydrated in deionized
water for 4 h prior to pouring the column. A Sephadex
chromatographic column (2.5 x 45 cm) was filled with the
G-25 slurry. Following the formation of a 1 inch layer
of beads at the bottom, the column outlet was Opened to
allow the bed to complete packing. The final column
height was 27 cm, having a void column of 62 ml. A flow
rate of 120 ml per h was used.

Prior to its use, the column was equilibrated
with the same buffer used in the cyanocobalamin equilibrium
binding experiment. This was accomplished by eluting
approximately 300 ml of buffer through the column.

The same procedure was used in preparing Bio-Rad
P-2, P-4 and P-10 polyacrylamide bead resins, and Sephadex

G-15.

27

Determination of Protein
Binding Capacity

 

 

To ensure consistency throughout binding experi-
ments, a standard procedure was established for determin-

ing protein-bound vitamin B The protein whose binding

12'
capacity was to be determined was dissolved in the
appropriate buffer, at a concentration of 1 to 4 mg per
m1 of buffer. One milliliter of the protein solution was
chromatographed on a Sephadex G-25 column previously
equilibrated with the same buffer. An Isco Model UA-2
Recording Ultraviolet Analyzer with an operating wave-
length of 254 nm was used to monitor the column effluent.
Thus, an absorption profile of the protein prior to the
addition of vitamin B12 was obtained.

The binding capacity of the protein was determined
by adding 1 ml of a standard B12 solution (either 51,
600 uug COGOB12 per ml, or 12,000 pug C057B12 per ml) to
1 m1 of protein solution. The mixture was allowed to
equilibrate for l h at room temperature with intermittent
mixing. The unbound B12 was separated from the protein
using Sephadex G-25 gel filtration chromatography. The
protein was excluded from the gel beads and was recovered
in the void volume effluent. The unbound COGOB12 per-
meated the gel beads and exhibited a characteristic

elution volume, which was determined by counting eluent

fractions with a Tri-Carb Scintillation Spectrometer.

28

The effluent containing the protein was collected,
shell-frozen and lyOphilized.

The lyophilized sample was dissolved in either 2
m1 of volatile sample buffer or 2 ml of deionized water.
A 1 ml aliquot was then counted in a Tri-Carb Scintilla-
tion Spectrometer and the average counts per minute (all
counting done for 10 min intervals) were converted to
vitamin B12 content by reference to previously prepared

57

standard curves for Co60 and Co B An aliquot of

B12 12'
this sample was also used for protein determination by
the Folin-Lowry procedure.

All binding capacities were expressed as micro-
micrograms (uug) vitamin B12 per mg protein.

Determination of Peptide
Binding_Capacity

 

 

Sample preparation and binding equilibrium pro-
cedures were the same as those described for proteins.
However, peptides were separated from unbound vitamin B12
by thin-layer chromatography (TLC) or TLHVE. The TLHVE
system used was the same system as previously described.
The TLC separation of unbound vitamin B12 was carried out
on prepoured Silica Gel-G plates (250 microns). Two
mobile phases were employed to effect the separation of
the unbound vitamin B12 from the peptides: 30 m1 glacial
acetic acid, 10 ml acetone, 50 ml methanol and 110 m1 of

benzene; and deionized water.

29

After the chromatograms were developed they were
dried in a forced-air oven at 90 C, sprayed with ninhydrin
reagent and redried at 90 C until color developed.

The peptide bands were then scraped from the thin-
layer plates and the bound vitamin B12 determined by
counting gamma emmissions with a Tri-Carb Scintillation
Spectrometer as described for proteins.

The migration characteristics of unbound C060B12
and COS7B12 on the TLC plates were determined by spotting
a 0.05% standard solution of crystalline B12 on each

plate. The vitamin B standard was readily apparent

12
because of its red color. This was confirmed by placing
the plates in a chlorine atmosphere for 5 to 10 min.
Following chlorination of the cyanocobalamin, the excess
chlorine vapor was removed from the plates by directing

a current of air over the plates. The plates were then
sprayed with o-tolidine-potassium iodide reagent (160 mg
o-tolidine dissolved in 30 m1 glacial acetic acid, made

up to 500 ml with deionized water and l g KI added
(Bollinger, 1965). Under these conditions, cyanocobalamin
is visible as a violet spot which corresponded exactly
with the Visual identification of B12. The corresponding
area where peptide-COS7B12 had been Spotted was then

scraped from the plates and the solid support counted with

a Tri-Carb Scintillation Spectrometer.

30

Enzymatic Hydrolysis of Proteins

 

Enzymatic hydrolysis experiments were performed
on B-Lg and (E)B-Lg using trypsin and chymotrypsin.
Hydrolyses were performed at 37 C. A Sargent Recording
pH Stat was employed to measure peptide cleavage by
automatic titration of protons with standardized 0.05N
triethylamine. The procedure and conditions used are

described by Fasold and Gundlach (1963).

Ultracentrifugation

 

The moisture content of the protein was determined
by drying samples over phosphorus pentoxide at reduced
pressure for 48 h. The protein samples were weighed on a
Mettler Type H 16 balance which was sensitive to 0.01 mg.
The samples were dissolved in the running buffer and
dialyzed for 24 h against the solvent.

Solvent densities were determined by pycnometry
at 20 C i l C. The viscosities of the solvent after
equilibrium dialysis were measured in a standard
Ubbelohde Dilution Suspended Level Type ASTM D 445
viscometer. The relative viscosities were calculated

according to the following equation:

JIJ
II
”If?

9—
po

0 0

where n, t and p are the viscosity, efflux time and

31

density of the solution and no, t and po are the

0
corresponding values for water at 20 C.
The partial specific volume Of the protein

solution was obtained by reference to literature values.

Sedimentation coefficient.--The sedimentation-

 

velocity experiments were carried out at 59,780 rpm
(259,700 x g) in a Beckman Model B Analytical Ultra-
centrifuge with two double sector synthetic boundary
cells.

The sedimentation coefficient is defined as the
velocity of the sedimentating molecules per unit field

as shown by the following formula:

S=T——l '%
w ° 60 dt

where x is the distance of boundary from the center of
rotation, g is the angular velocity in radians per sec,
and E is the sedimenting time in seconds.

The actual equation used in calculating the

sedimentation constant is:

2.303 lo
w ° 60
By plotting the log of distance (x) against time (2), the
sedimentation coefficient may be obtained from the slope

by the following formula:

32

S = -%4222—-° lepe ° 1 x 1013
w x 60

Sedimentation coefficients are usually reported as 820,w
values, which is the sedimentation coefficient the
protein would exhibit in a solvent with a density and
viscosity equivalent to water at infinite protein

dilution. The following is the formula containing the

terms necessary for these corrections:

 

 

s = s (BY—LL) (nw't) (1 - :pw't)
20,w Obs ”w,20 nw,t 1 - Vps,t
. nw t . .
the first term (-—4——) 15 the correct1on factor for the
w,20

viscosity of water at eXperimental temperature to that at
20 C. At the experimental conditions, this term had a
value of unity since all runs were performed at 20 C.

The second term (1543) corrects the relative viscosity of
the solvent to thaz’gf water at the same temperature. In
the last term, V, the partial specific volume of the

protein was obtained from literature values and was

assumed to be the same in all solvent systems. The
pw,t

s,t
density of the solvent at any temperature to that of

 

density factor ( ) is the correction ratio for the

water at 20 C.

RESULTS

Beta-Lactoglobulin

 

Nitrogen Content

 

The nitrogen content of the isolated B-Lg as
determined by micro-Kjeldahl was 15.53% which compared to
the literature value of 15.6% reported by Larson and
Jenness (1955). The nitrogen content was converted to
per cent protein using a conversion factor of 6.25.

The concentration of all B-Lg solutions used
during this study was based on this experimentally

determined nitrogen content.

Sulfhydryl Content

 

The number of available sulfhydryl groups present
in the isolated B-Lg was determined by Ellman's procedure.
In the presence of urea, the protein exhibited two
available sulfhydryl groups per mole of B-Lg (36,000 M.W.).
No sulfhydryl groups were detected in the absence of urea.

Gel ElectrOphoresis of
Beta-LactoglobulIn

 

Horizontal polyacrylamide gel electrOpherograms
of the isolated B-Lg, whey standards and pure B-Lg A and

B are Shown in Figures 1 and 2. The electrOphoretic

33

34

 

l23456

Figure l.--Polyacry1amide gel electropherogram of re-

crystallized p—Lg and whey standard at pH 8.3; left to right:
slot 1 - 5 p1 B-Lg; slot 2 - 10 pl B-Lg; slot 3 - 20 pl g-Lg;
slot 4 — 20 pl whey; slot 5 - 25 pl whey; slot 6 - 35 pl whey.

35

 

| 2 3 4

Figure 2.—-Polyacrylamide gel electropherogram of recrystal—
lized B—Lg and B-Lg A, B and AB standards at pH 8.3: left
to right; slot 1 - 10 pl B-Lg B standard; slot 2 - 10 pl

B—Lg AB standard; slot 3 - recrystallized B—Lg AB; slot 4 -
10 pl B-Lg A standard.

36

patterns confirm the presence of both genetic variants
(A and B). No contaminating proteins were apparent.
Polyacrylamide disc gel electrOphoresis of B-Lg in a
discontinuous buffer system did not show the presence of
any fast moving zones (i.e., peptides) in gels removed
at 10 and 20 min intervals and developed with ninhydrin.
Gels removed after 30 min showed typical B-Lg

patterns when stained with Amido Black 10B.

Sedimentation Coefficients

 

Sedimentation velocity studies of the isolated
B-Lg in 0.1N KCl, pH 5.1, revealed one molecular species,
i.e., $20,w = 2.92. These data agree with the sedimenta-
tiOn characteristics reported by Cecil and Ogston (1949)
and Ogston and Tombs (1957) for B-Lg.

The apparent sedimentation coefficients (Sapp)
of the B-Lg exhibited a slight concentration dependency
at pH 5.1 in 0.1N KCl increasing slightly with increasing
protein concentration (Table 1). This behavior is
typical for a slow association-dissociation interaction
(Gilbert, 1963).

The sedimentation coefficient of B-Lg was also
determined in borate buffer, pH 9.0, I = 0.24. At this
pH B-Lg should be dissociated to its monomeric state
(i.e., 18,000 M.W.) and evidence of this was apparent

by the decrease in the sedimentation coefficient to

37

.cofipsane muncflpafi on

coflumnucoocoo aflououm .m> mmmm mo uoam mcflumaommuuxo an oocflmuno mosam>d

 

 

 

m.m II o.oa

h.m mm.m m.h

m.m mm.m o.m

IT no.m m.m

H.N mm.m mo.o :HHOQOHmouooqlmuom
mumuom HUM

Aasxmev
mDSOHoammmou coflumupcoocou aflououm aflopoum

coaumucoEwoom ucoummmfl

 

O on no Howmso oumuon pom

oofluoaco Esflmmmuom SH.o CH mqnm mo muscumcoo sowumncoaflomm ucoummmm chIT.H mqmda

38

820 w = 2.16. This value compares with the value of
I

820 w = 2.4 reported by McKenzie and Sawyer (1967) for
I

B-Lg at pH 8.6.

Electrodialyzed Beta-Lactoglobulin

 

Sulfhydryl Content

 

Sulfhydryl determinations on the (E)B-Lg, using
Ellman's procedure, showed no loss or addition of -8H
groups. Sulfhydryl determinations carried out on the
(E)B-Lg in the absence of urea showed no available -SH

groups.

Gel ElectrOphoresis

 

Discontinuous polyacrylamide gel electrOphoresis
of the (E)B-Lg together with a whey standard and B-Lg A
and B standards, showed identical electrophoretic gel
patterns. As shown in Figure 3, the (E)B-Lg exhibited
electrOphoretic mobilities identical to the B-Lg prior to

electrodialysis.

Sedimentation Coefficient

 

Sedimentation velocity studies performed on the
(E)B-Lg in 0.1 M ammonium acetate, pH 6.8, showed a

single component with an S = 3.0. This value was in

20,w

agreement with the sedimentation coefficient determined
for native B-Lg. However, apparent sedimentation
coefficients calculated for (E)B-Lg did not exhibit the

same concentration dependency as was shown for B-Lg

(Table 2).

39

.coflusaflp ouflcwmcw ou
coflumuucoocoo cHououm .m> mmmm mo uon ocflumHommuuxo mo oocflmuno mosam>m

 

 

 

m.m mo.m o.oa
mo.~ H.m m.h
mm.~ oo.m o.m cflasoonouomqnmumm
H.m mo.m mo.o OONmHmHoouuooam
mumuom Hos
AHE\mEv .caououm
mDSOA0flmmooo coflumuucoocou cwououm .

coflumucoefloom acoummmd

 

0 cm um nommso cannon can oumuoom
EdflcoEEm 2H.o SH mqumhmv mo mucmumcoo coflumucofiflpom ucoummmm ocaln.m mqm<9

40

 

l 2 3 4

Figure 3.—-Polyacry1amide gel electropherogram of B-Lg and
(E)B—Lg at pH 8.3: slots 1 and 2 - (E)B-Lg at l and 2%
protein concentration; slots 3 and 4 — B-Lg at l and 2%
protein concentration.

41

The $20,w value of (E)B-Lg in borate buffer, pH 9.0
was 2.1. Apparent sedimentation coefficients for the
(E)B-Lg (Table 2) corresponded exactly with those calcu-
lated for the native protein (Table l) in the same buffer
system with the exception of the 0.75% protein concentra-
tion.

Apparent sedimentation velocity constants for the
electrodialyzed and native protein preparations indicate
that both preparations exhibit a concentration dependent
association-dissociation equilibrium.

Amino Acid Analyses of Beta-Lactoglobulin
and Electrodialyzed Beta-Lactoglobulin

 

 

Amino acid analyses were carried out on the 22 h
hydrolysate of B-Lg and (E)B-Lg. These data are presented
in Table 3 and are expressed as grams of amino acid per
100 grams of protein.

N-Terminal Amino Acids of Beta-Lactoglobulin
and Electrodialyzed Beta-Lactoglobulin

 

 

N-Terminal amino acid analyses, using the dansyl
chloride procedure confirmed iso-leucine as the N-terminal

amino acid for both B-Lg and (E)B-Lg.

42

TABLE 3.--Amino acid composition of beta-lactoglobulin and
electrodialyzed beta-lactoglobulin

 

Electrodialyzed

Amino Ac1d Beta-Lactoglobul1n Beta-Lactoglobulin

Residue

 

9 amino acid residues/100 g protein

 

 

LYS 12.58 11.41
HIS 1.71 1.53
ARG 2.30 2.45
ASP 10.06 10.20
THR 4.39 4.41
SR 3.18 3.10
GLU 17.00 16.58
PRO 4.14 4.34
GLY 1.07 1.15
ALA 5.66 5.85
1/2 CYS 2.31 2.51
VAL 6.15 6.21
MTH 1.24 0.76
I-LEU ' 5.77 5.82
LEU 14.15 14.33
TYR 3.47 3.76
PHEN 2.65 3.43
TRYP 2.16 2.16
TOTAL* 99.99 100.00

 

*Total 9 amino acid residues/100 g protein

43

Rate of Enzymatic Hydrolysis of Beta-
FLactoglObuIIn and EIecErodiaIyzed
Beta-Lactoglobulin

 

The relative enzymatic hydrolysis rates of B-Lg
and (E)B-Lg preparations provided a means of assessing the
possible effects of electrodialysis on B-Lg.

The terms used to compare proton release from the
substrates examined were percentage hydrolysis versus
hydrolysis time as shown in Figures 4 and 5. This method
of expression eliminates inconsistencies that may have
resulted from the proteolytic cleavage of a structured
protein and the subsequent appearance of new cleavage sites
heretofore unavailable due to conformational character-
istics. As evidenced from Figures 4 and 5, no major dif-
ferences can be seen in the tryptic or chymotryptic
hydrolysis of B-Lg and (E)B-Lg. However, the calculated
relative Km values for tryptic and chymotryptic hydrolyses
showed a lower Km for both enzymatic hydrolyses when
(E)B-Lg was the substrate (see Figures 6 and 7).

Paper High-Voltage ElectrOphoresis of
the Tryptic Digest of Beta-

Lactog16bulin and Electrodialyzed
Beta-LactoglObulin

 

 

Figure 8 shows the results of high-voltage paper
electrophoresis of the tryptic hydrolysates of B-Lg and
(E)B-Lg. Similar peptide patterns were obtained for the

B-Lg and (E)B-Lg preparations.

44

I00 -

O
O
I

fl
0
I

. X-(E) B-Lq
A-B-Lo

(I 0’
O O
I I

A
O
I

Per Cent Hydrolysis

so-

20-

IO-

 

O l l l I l l I

2 4 6 8 IO l2 l4
Time (min)

Figure 4.--Hydrolysis of B-Lg and (E)B-Lg with trypsin.

45

IOO -

80-

 

X-(E) B-Lq
A- B-Lq

40

Per Cent Hydrolysis

01
O

N
0

IO

 

O l 1 1 l 1
2 4 6 8 IO l2 l4

Time (min)

Figure 5.--Hydrolysis of B-Lg and (E)B-Lg with chymotrypsin.

(mM alkali I min)

Rate

46

 

 

 

.5 -
°"' x-(E) B-Lq
A- B-La .
.3 -
X
.2 - ‘
A

.l '-
O l I 4 I l l

2 4 6 8 IO l2

mg Substrate

Figure 6.--Rate curves for tryptic hydrolysis of B-Lg and
(E)B-Lg per 60 sec.

47

 

 

.6 - X-(E) B-Lo
A-B-Lo
.5 -
X .
g .4 ~ .
s
‘\
=5 A .
as
° .3 -
2
5
II
‘6
a: .2 -
.l -
I I I L I I
0 20 40 so so IOO l20

ma Substrate

Figure 7.--Rate curves for chymotryptic hydrolysis of B-Lg
and (E)B-Lg per 60 sec.

48

 

Figure 8.--High-voltage paper electropherogram of the tryptic
hydrolysate of B-Lg and (E)B—Lg in pyridine—acetic acid-water
buffer, pH 3.5: left to right; (Els-Lg hydrolysate and

B-Lg hydrolysate.

49

Peptides from Beta-Lactoglobulin

The peptide mixture Obtained by the electrodialysis
of B-Lg was light brown and had a sticky consistency
following lyophilization of the electrOdiffusate. Further
dehydration of the peptides was accomplished by drying
over phosphorous pentoxide under reduced pressure.

Initially, each electrodiffusate change during
electrodialysis was pervaporated and lyophilized separately.
The electrodiffusate fractions were rehydrated with deion-
ized water and passed over a Bio-Rad P-2 column using
deionized water as the eluate in an effort to remove salts
or free amino acids that may have been electrodialyzed from
the B-Lg. The electrodialyzed peptides were recovered in
the void volume from the column. A second peak was eluted
shortly after the peptides but did not exhibit a positive
test for protein or free amino acids (see Figure 9).,

The electrodiffusate fractions were analyzed by
high voltage electrOphoresis at pH 3.5 on both paper and
thin layer supporting media. All fractions exhibited the
same characteristic peptide pattern (Figure 10). One
peptide migrated to the anode, indicating that it was
negatively charged at pH 3.5 and was arbitrarily designated
as the "negative peptide." Two other peptides were
Observed, both of which migrated to the cathode. These

were arbitrarily designated as "+2" and "+3" peptides.

 

50

.GHE\HE v .mumn 30am
.muamm .m xmmm “musuxwa OUHHQOQ .H xmom .Hmum3 omuflcofimp EDHB GEOHOO
mum oomIOHm Eoum mmpflummm oonhamfloouuomam mo cumuumm cowusHMI|.m musmwm

Es 2.:

 

 

 

 

 

 

 

 

 

(mu 1792) aouoqlosqv

51

 

Figure 10.-—Thin—1ayer high-voltage electrOpherogram of the
peptide mixture electrodialyzed from B-Lg. Support; MN 300
cellulose:Buffer; pyridine-acetic acid-water, pH 3.5.

52

TLHVE confirmed that the peptide concentration was
the only variable in the individual electrodiffusate frac-
tions. The highest peptide concentrations were Obtained
between 2 and 4 h of electrodialysis. Subsequent to these
findings, the total electrodiffusate was combined following
pervaporation and lyOphilized. The recovery rate of pep-
tides from B-Lg was calculated as 7-8 mg of dried peptide
mixture per g of B-Lg.

Separation and Recovery of the
Peptide Mixture

 

 

The lyophilized peptides were separated using thin-
layer high voltage electrOphoresis as described in the
experimental procedure. Excellent resolution of the
peptides was obtained using this technique as Shown in
Figure 10.

Recovery of the isolated peptides from the cellulose
(MN 300 cellulose) support was accomplished by elution with
a solution of 5% acetic acid in deionized water, or a 1:1
(v/v) mixture of pyridine-deionized water. The latter
system was discarded because of the difficulty of removing
residual pyridine from the sample. Of the two systems,
although both effected the extraction of some cellulose
along with the peptides, the 5% acetic acid gave a more
complete recovery of the peptides with fewer extractions.

The per cent recovery, calculated on the basis of the

53

weight of peptides recovered from electrodialysis and the
calculated weight of peptides present in the extracted
material from the thin—layer support, was about 60-70%.

Physical and Chemical Properties of
Isolated Peptides

 

 

The molecular weights of the isolated peptides
appear to be greater than 1,500, but less than 3,600 on
the basis of the gel filtration experiments performed.
The peptide mixture was chromatographed over Bio-Rad P-2
polyacrylamide beads and was eluted from the column in the
void volume at 62 m1 (Figure 9). This indicated a
molecular weight of the peptides above the exclusion
limit of Bio-Rad P-2 (approximately 1500 M.W.). Gel
filtration chromatography of the peptide mixture over
Bio-Rad P-4 (Figure 11) showed an elution volume of 110
ml (Bio-Rad P-4 column VO = 63 ml). This behavior
indicated that the peptides may have a molecular weight
of less than 3,600.

Gel filtration chromatography of the individual
peptides on P-2 and P-4 gave elution patterns similar to
those exhibited by the peptide mixture.

Minimum Molecular Weights of
IsOlated Peptides

 

 

Minimum molecular weights of the isolated peptides
were calculated from the amino acid composition of indi-

vidual peptides according to the following formula:

54

.swE\HE o.m mmz much 30am

.oaoemz 2H.o sues season

Vim

pmmnowm Scum onsuxwfi moflummm omnaamwponuomao mo cumuuwm coHusHmll.HH wusmwm

Es 2.:

 

 

 

_

60:

.>

 

 

(um 1793) eouoqiosqv

55

 

= Conc. A.A.
M '100
w
where:
Mw A.A. = residue weight of amino acid residue
Mw = estimated minimum molecular weight of
peptide
Conc. A.A.= concentration of amino acid residue

in sample (g/lOOg protein).

TryptOphan, phenylalanine and tyrosine were not
used in minimum molecular calculations of the peptide
because the concentration of these amino acid residues
was so low that they could not be determined accurately.
All other amino acid residues present in the peptides
were used in minimum molecular weight calculations and
the least common denominator of these values was recorded
as the estimated molecular weight of the peptide.

The estimated molecular weights of the peptides
were calculated to be: negative peptide 4,000; +2

peptide 6,000; +3 peptide 6,000.

Amino Acid Analyses of Isolated Peptides

Amino acid analyses were performed on acid
hydrolysates of the isolated peptides. These data are
presented in Table 4 as grams of amino acid per 100 grams

of protein.

56

TABLE 4.--Amino acid composition of isolated peptides from
beta-lactoglobulin by electrodialysis

 

 

 

 

Amino "Negative" Peptide "+2" Peptide "+3" Peptide
Rgégdue g amino acid residues/100 g protein

LYS 6.66 6.79 28.37
HIS 2.10 2.26 4.65
ARG 2.46 0.75 3.26
ASP 7.37 6.79 6.98
THR 2.81 5.28 3.26
SER 6.32 26.42 9.77
GLU 11.58 8.30 9.30
PRO 2.81 3.02 1.86
GLY 30.53 22.64 22.79
ALA 4.21 9.06 5.12
1/2 CYS -- -- --

VAL 27.02 6.42 4.65
MTH -- -- --

I-LEU 2.81 1.89 3.72
LEU 3.51 4.53 4.65
TYR 1.75 1.51 2.79
PHEN 1.40 1.51 2.79
TRYP -- <0.38 <0.47
TOTAL* 113.34 107.55 114.43

 

*Total g amino acid residue/100 g protein

57

It should be noted that sulfur containing amino
acids were not detected in any of the peptides. Tryptophan
was not determined in the "negative" peptide because of

insufficient sample.

Separation of Unbound COGOB12 from

Protein and Peptide-Bound C060312

 

Using Gel Filtration

Bio-Rad P-2 polyacrylamide beads were unsuitable

for separating CO60

bound C0603

B12 from the protein or from peptide-
12 because the exclusion limit of the P-2 was
too close to the molecular weight of hydrated cyanocobala-
min. Data contained in Table 5 illustrate the elution

pattern of CO6OB12 from a P-2 column as monitored with a

Tri-Carb Scintillation Spectrometer. COGOB12 begins its
elution from the P-2 column in the void volume (63 m1) and
continues until approximately 130 ml of eluate is passed
from the column. The summation of net counts per minute
(opm) for all fractions counted in Table 5 totaled 5293.
One milliliter of the stock COGOB12 solution gave a total
of 5333 cpm, indicating a recovery of 99.2%.

Similar elution volumes were exhibited by B-Lg
and the peptides which have molecular weights above the
exclusion limit of the polyacrylamide beads.

Experimental data have Shown that Sephadex G-25

with an exclusion limit of 5,000 can be used to separate

58

TABLE 5.--Elution pattern of CO6OB12a from Bio-Rad P-2

gel filtration column versus counts per minute of 5 m1
elute fractions

 

 

Gross Counts Gross Countg
Fraction per minuteb Fraction per minute
(10 min basis) (10 min basis)
1 52 26 71
2 53 27 59
3 41 28 80
4 50 29 55
5 51 30 59
6 51 31 52
7 53 32 61
8 50 33 45
9 47 34 56
10 44 35 54
ll 84 36 60
12 96 37 41
13 (void volume)657 38 48
14 653 39 37
15 776 40 42
16 773 41 42
17 499 42 49
18 503 43 31
19 326 44 47
20 324 45 37
21 208 46 48
22 191 47 43
23 173 48 31
24 132 49 31

25 (125ml V8) 114

 

al ml of CO6OB12 stock solution (51,600 pug BlZ/ml)

diluted to 4ml with 0.1N ammonium acetate.

bBackground gamma radiation count was 37 counts
per minute (10 min basis).

59

60 60

free CO 312 from the protein-Co B 2 complex by gel

1
filtration chromatography. However, separation of unbound

60

Co B or C057B12 from the peptide mixture or isolated

12
peptides could not be accomplished by gel filtration
chromatography using Bio-Rad P-2, P-4, P-6, P-10;

Sephadex G-15 or G-25.

COGOB12 or C057B12, at the concentrations being
used, do not absorb at 254 or 280 nm. Consequently, the
presence of cyanocobalamin that is not bound to proteins
or peptides cannot be detected in the elution volume by
monitoring with an Isco UA-2 ultraviolet analyzer. Rather,
the elution pattern must be obtained by counting gamma
emission in the eluate from the column.

Vitamin B12 Binding Capacity of Beta-

LactoglobuIIn and Electrodialyzed
Beta-LactogIObulin

 

 

 

The results of vitamin B12 binding studies with
B-Lg and modified B-Lg are shown in Table 6. The binding
capacity of B-Lg in 0.1M ammonium acetate, pH 6.8, and
Jenness and KOOps buffer, pH 6.6, where the pH and ionic
strength is equivalent to that of normal milk, were 460
ppg vitamin Biz/mg of protein. (E)8-Lg and recombined
(E)B-Lg and peptides showed no measurable vitamin B12
binding capacity in either the ammonium acetate or Henness
and KOOps buffer systems.

At pH 9.0 the binding capacities of all three

forms of B-Lg for vitamin B12 varied with the buffer

60

system being used (see Table 6). In borate buffer the
recombined (E)B-Lg and peptides showed a binding capacity
of 200 ppg vitamin BIZ/mg of protein, whereas B-Lg was
shown to bind only 150 pug vitamin BIZ/mg of protein. In
TRIS-HCl, pH 9.0, all three forms of B-Lg exhibited bind-
ing capacities that were not evidenced at any other pH or
in any other buffer system. The binding capacities ex
exhibited by B-Lg and (E)B—Lg in ammonium acetate-sodium
hydroxide solvent, pH 9.0 were essentially equivalent (100
ppg vitamin BlZ/mg of protein), whereas the re-equilibrated
(E)8-Lg and peptides showed only a slightly lower vitamin
B12 content.

In contrast to the results obtained at pH 9.0,
B-Lg and the modified forms of B-Lg showed no measurable
binding capacity at pH 2.0.

Confirmation that vitamin B12 was bound to the
B-Lg was obtained by dissolving the protein-vitamin B12
complex (460 ppg CO6OBlz/mg of protein) in 0.1M ammonium
acetate and dialyzing the solution against the solvent for

24 h. The vitamin B bound to the B-Lg in the dialysate

12

was redetermined and showed no loss of vitamin B12.

61

.ocHEHouop on 30H OOH.o

.EMOM HoEocoE
mufi SH mp3 mqnm chance on Aammav poocsoe can mmmnommEHB mo poms ucm>HOmQ

.mHm Seamufi> canon gamuoum
one Eoum mam.ocsoocs oumnmmom O» can: mm3 GEOHOO mmnw xoomcmomm

 

gene case case o.~ meow oeaoaaoouamn
nuaeeuoHBo sneeom

om OOH OHH o.a meexouema sensor
cuppoom EOaGOEEd

omm omm omm o.m HomeHme
oom ow oma o.m mumpom
mega QBAB owe m.m mmoom pom mmoccon
QBQB UQBAB owv m . m OUMDOOMISHSOEEAN

 

mooflumom cam

moumxmv eacensooam mqumxmv more

 

mm Hmmmsm

Aswanoum ma pom mam Nam 00 ommnm>¢

oo mnnv ocsom co

m om

 

mosam> mm.ocm mummmso msoflum> ca mooflumom was
marmamv pocHnEOOou pom maimamv .mnnm mo mowuwommmo mcflocflo Nam cflsmufl>nu.m mqmde

62

Discontinuous Polyacrylamide Gel Electro-
phoresis of Beta-Lactoglobulin and
Electrodialyzed Beta-Lactgglobulin

Equilibrated with CO57B12

 

 

 

 

B-Lg and (E)B-Lg with 460 and 10 ppg/mg protein of

bound C057B12, respectively, were subjected to vertical

gel electrOphoresis at pH 8.3. The results of this experi-
ment are shown in Figure 12. As Shown in the electro-

pherogram, both genetic variants (A and B) were present.

57

The Co B initially bound to either B-Lg or (E)B-Lg did

12

not appear to affect the electrOphoretic mobilities of the
proteins. The zones containing B-Lg A and B were cut

from the gel and counted in a Tri-Carb Scintillation
Spectrometer. The C057812 was not found associated with
either genetic variant, nor was there any evidence that

the vitamin remained in the sample slots of the gel.

Vitamin B12 Binding Capacity of

 

Peptides Isolated by Electro-
dialysis of Beta-Lactgglobulifi

 

 

Following the equilibration of the peptide mixture

60

with Co B the peptide-C0608 mixture was chromato—

12’ 12
graphed on a Bio-Rad P-4 gel filtration column. NO

absorption at 254 nm was monitored with elution of the
column's void volume, indicating that the peptides had

not formed a complex with vitamin B12 having a molecular

weight in excess of 3,600.

63

 

l 2 3 4

Figure 12.—-Vertica1 polyacrylamide gel electropherogram of
B-Lg and (E)B—Lg at pH 8.3. Column 1- (E)B—Lg—B12 complex;

Column 2- (E)B—Lg; Column 3- B—Lg—B12 complex; Column 4-
B-Lg.

64

Free COGOB12 was separated from the peptide B12
equilibration mixture by thin-layer high-voltage electro-
phoresis. The results of this experiment (Figure 13 and
Table 7) showed only background radiation associated with
the negative and the +3 peptides, while the +2 peptide
exhibited a gamma radiation count equivalent to 200 ppg
of vitamin 812. However, close inspection of the

crystalline B standard (column 1 of TLHV electrOphero-

12

gram) indicated that vitamin B migrated in the direction

12
of the cathode at pH 3.5 in the field of 40 V/cm.

Thin-layer ascending chromatography was also
employed in the separation of unbound C057B12 from the
peptide mixture as well as the separated peptides. The
results of these experiments are shown in Figures 14, 15
and 16. The peptide remaining at the origin was identi-
fied as the +3 peptide, whereas the +2 peptide and
negative peptide exhibited Rf values of 0.12 and 0.23,
respectively. Determinations of the vitamin B12 binding
capacities of the negative and +2 peptides showed only
background radiations.

The crystalline B12 standard remained at the
origin in this system. Gamma radiation accounting, per-

formed on the C057B

57
CO B12

the origin. Thus, the binding capacity of the +3 peptide

for C0578

12 reference spot, indicated that the

applied to the thin-layer plate also remained at

12, if any, could not be determined by this TLC

technique.

66

- I86

— 230

— 6'0
206

—
i3” _ — — 206

Figure l3.—-Thin-1ayer high—voltage electropherogram of the
peptide mixture equilibrated with Co 7B 2. Support - MN 300
cellulose; buffer - pyridine—acetic acié-water, pH 3.5.

Left to right: column 1— cyrstalline B12; column 2—
electrodialyzed peptides; column 3— plectrodialyzed
peptides plus Co 7B12; column 4— C05 B12 standard.

(Numbers refer to radiation level in cpm-—background
radiation 183 cpm.)

65

mp3 mcfipcsoo Ham

.cHE\mucsoo wma

.mflmon CHE ea c co coma
um pondmmofi mm3 coflumfipmn pcooumxommm

 

 

ocflEuouoo op 30H 00» mm cflmwuo
ocHEHouoo on 30H oou IT opflumom m+

on: com ems meeudmm ~+
OSHEuouoo Op 30H oon m ooflumom m>wummmz
ocsom mam cflEmuH> :flE\mucsoo umz mamemm

M

 

mflmouosmouuomao ommuao> amen Howmaicflnu an mooflumom on» Scum

ocooocs mo coflumummom noumm ooflumom may mo moapflommmo mcapcfla

O

mam ceemue>ua.n mamas

66

- '86

- 230

- 6'0
206

__ ‘
ISM - 206

 

Figure 13.--Thin-layer high-voltage electropherogram of the
peptide mixture equilibrated with 0057B 2. Support - MN 300
cellulose; buffer - pyridine—acetic aci -water, pH 3.5.

Left to right: column 1- cyrstalline B12; column 2-
electrodialyzed peptides; column 3- glectrodialyzed

peptides plus Co 7812; column 4- C05 B12 standard.

(Numbers refer to radiation level in cpm--background
radiation 183 cpm.)

67

 

Fi ure l4.--Thin-layer chromatographic separation of unbound
CO 7B12 from the electrodialyzed peptide mixture. Support —
Silica Gel G; mobile phase - 55% benzene, 25% methanol,

5% acetone and 15% glacial acetic acid. Left to right:
column 1- crystalline B12 standard; column 2- electrodialyzed
peptide plus Co 3127 column 3- electrodialyzed peptide;
column 4- C057B12. (Numbers refer to radiation level in
cpm--background radiation 108 cpm.)

68

 

I 2 3

Figpre 15.--Thin-layer chromatography separation of unbound
Co 312 from the negative, +2 and +3 peptides. Support -
Silica Gel G; mobile phase - 55% benzene, 25% methanol, 5%
acetone and 15% glacial acetic acid. Left to right: column
1- +3 peptide plus C057B12 (cpm 2450); column 2- +2 peptide
plus C05 B12 (193 cpm); column 3- negative peptide plus
C057Blz (188 cpm). Background radiation 186 cpm.

69

 

I 2 3

F1 ure 16.--Thin-layer chromatographic separation of unbound
Co 7B12 from the +2 peptide. Support - Silica Gel G; mobile
phase - 55% benzene, 25% methanol, 5% acetone and 15%
glacial acetic acid. Left to right: column 1- crystal-
line 312 standard; column 2- +2 peptide plus C057B12 (192
cpm); column 3- C057B12 standard. (Origin 2360 cpm; area

corresponding to +2 peptide 185 cpm). Background radiation
187 cpm.

70

TLC technique using deionized water as the mobile
phase was then employed to effect the resolution of the
free C057812 from the peptide. The results of this
experiment (Figure 17) showed that ideal resolution
between the peptides and vitamin B12 was not obtained.
However, the absence of gamma radiation at the trailing

edge of the peptide band indicated that peptides did not

bind V1tam1n B12.

71

     

   

I 2

F1gure l7.--Thin-1ayer chromatographic separation of unbound
Co 7B12 from the peptide mixture. Support - Silica Gel G;
mobile phase - deionized water. Left to right: column 1—
peptide mixture equilibrated with CO57B12 (radiation level
at center of plate 194 cpm; trailing edge of peptide mixture
197 Cpm; center of peptide band 2160 cpm); column 2-

crystalline B12 standard. Background radiation 188 cpm.

DISCUSSION

In View of previous data concerning the binding
of vitamin B12 to the peptides electrodialyzed from B-Lg
reported by Gizis (1965) and Dorris (1968), the rationale
developed for this thesis was to identify the binding Site
on the peptides for the ig,yip£9_binding of vitamin B12
in model systems. Because B-Lg was reported to have the
highest concentration of associated peptides of all the
milk proteins, it was chosen as the experimental system.

The B-Lg used in this research was isolated from
mixed herd milk and was purified by repeated recrystalli-
zation. Reported chemical analyses, sedimentation veloci-
ties and polyacrylamide gel electrOpherograms (see pages
33-48 Results) are in agreement with comparable values
published by Tilley (1960), Bell and McKenzie (1964) and

Melachouris (1969).

Association of Vitamin B12 with Peptides

 

Following electrodialysis of the crystalline
B-Lg as described under experimental methods, the desorbed
peptides were recovered from the electrodiffusate. The
vitamin B12 binding capacity of these peptides was

determined by equilibration with CO6OB12 for l h in

72

73

various buffer systems. The unbound COGOB12 was sepa-

rated from the peptides by gel filtration chromatography

using Bio-Rad P-2 as previously described by Dorris

(1968). The results of this experiment indicated that

the peptides exhibited an apparent binding in the range

of 5,000 pug COGOBlz/mg of peptide. This value

agreed with the binding value reported by Dorris (1968).
Because a maximum B12 binding value by the pep-

tides was not previously reported, an experiment was

designed to evaluate this parameter. The peptides were
60

equilibrated with varying levels of Co B12 and the free
COGOB12 separated by gel filtration chromatography using
60

Bio-Rad P-2. Increasing levels of a Co B12 solution
added to the peptide equilibration mixture provided
data which indicated that the binding capacity of the
peptides was unreasonably high. These experiments pro-
vided the first indication that this procedure for
determining the B12 binding capacity of the peptides in
model systems may be erroneous.

The unreasonably high B12 binding capacities
measured for the peptides were only explainable by assum-
ing that there was an incomplete separation of free
COGOB12 from the peptide-CO6OB12 equilibration mixture
on the P—2 column. To confirm this postulate, a standard

60

CO B 2 solution was chromatographed over the P-2 column

1
and the eluate monitored with a scintillation

74

spectrometer to determine the elution volume of the

60

Co B standard. The results of this experiment showed

12
that COGOB12 exhibited no UV absorbance at 254 nm as

previously assumed by Dorris (1968), and that the COGOB12

was eluted from the P-2 column in the void volume along

with the peptide mixture. Further study of the elution

60B

profile of the Co standard over the Bio-Rad P-2

12
column indicated that the adsorbance monitored during the

chromatographic separation of the B12 standard was the

result of the methyl and propyl esters of para-hydroxy

60B

benzoic acid added to the Co solution as an anti-

12
microbial agent. The results obtained from these experi-

60

ments indicated that the Co B binding to the peptides

12
as reported by Dorris (1968) was due to the elution of

free CO6OB12 with the peptides.

In subsequent experiments Bio-Rad P-4 and other

gel filtration supports were employed in an effort to

effect a separation of the unbound C057B12 (C057B

60 57

12 was

used in place of Co B12) from the peptide-Co B12 mix-

ture. However, when column eluates from the peptide

57B

mixture and Co standard systems were monitored by

12
both UV absorbance and scintillation spectrometry, the
peptides and C057B12 exhibited Similar elution character-
istics on all gel filtration supports employed. This
observation indicated that the peptides and C057B12

behaved like molecules having similar molecular weights.

75

Although separation of free C057B12 from the

peptide mixture with Bio—Rad P-4, a procedure reported
by Gizis 33.223 (1965), proved unsuccessful, the fact
that no UV absorbance at 254 nm was monitored at or near
the elution of the void volume (peptide-COS7B12 complex
would have a molecular weight near 3,600) gave the first

indication that vitamin Bl2 was not bound to the peptides.

It was not immediately ascertained that a peptide-COS7B12

complex, if present, could not elute with the free pep-

. 57
t1des and Co B12.

In view of these experimental findings, it was

apparent that a different method for separating free

57

Co B from the peptides was needed before the presence

57B

12

of a peptide-Co 12 complex could be established.

Gizis and Meyer (1968) reported that peptides
electrodialyzed from human blood bound vitamin B12 by

non-electrostatic association. Therefore, TLHVE was

employed in an attempt to separate the unbound COS7B12

from the peptides to determine if the peptides bound

57

Co B by a mechanism other than electrostatic attrac-

12
tion. The results of these experiments (Figure 13)

Showed no trace of the C057B12 associated with either
the +3 or negative peptides. The TLHVE data were in-
conclusive with regard to the binding capacity of the

+2 peptide. The center of the ninhydrin stained +2

peptide zone, when removed and counted in the

76

scintillation spectrometer, showed a radiation level
equivalent to 200 ppg of vitamin B12. Although these
data indicated that the +2 peptide had bound B12, close

inspection of the crystalline B standard, which had

_ 12
been spotted on the same plate, showed it migrating

toward the cathode under the electrophoretic conditions

employed. Thus, the vitamin B12 which was associated

with the +2 peptide could have resulted from the con-

comitant migration of the free COS7B12 mixture. This

re mm“: HfiflEI‘V

W)”: .
L1.

postulate was further supported by the observation that
the radioactivity determined at the leading front of the
+2 peptide band showed only a background level of radia-
tion.

These inconclusive results led to the development
of an ascending TLC technique in an attempt to separate
the unbound COS7B12 from the peptide. In this manner
the migration of peptide and labeled B12 in an electro-
static field would be avoided. The mobile phase consist-
ing of 55% benzene, 25% methanol, 15% glacial acetic
acid and 5% acetone produced a migration of the negative
and +2 peptides, whereas, vitamin B12 remained at the
origin. As shown in Figures 14, 15, and 16, the migrat-
57B

12 associated with them.

Hence, it was established that the +2 peptide did not

ing peptide zones had no Co

bind B A possible electrostatic binding of E12 by

12'
the +3 peptide could not be determined with this mobile

77

phase as a result of its inability to migrate. There-
fore, a different mobile phase was employed to determine

if the +3 peptide was forming a binding complex with the

57
CO B12.

In subsequent TLC separations, deionized water

was used as the mobile phase. The resolution between

the peptide mixture and the C057B12 obtained with this

system was not ideal (Figure 17). However, the C057B12

standard indicated sufficient resolution between the
peptides and free C057B12 that the trailing edge of the
peptide band could be analyzed for radioactivity. The
cellulose was removed and monitored, showing no COS-[B12
present. Subsequent TLHVE of the peptides isolated from
the trailing edge of the peptide zone confirmed the
presence of all three peptides in the cellulose. Thus,
it was concluded that vitamin B12 is not bound by any

of the identified peptides electrodialyzed from B-Lg.

These B binding characteristics of the isolated

12
peptides do not agree with the findings of Gizis et a1.

(1965) and Dorris (1968), both of whom reported that B12
bound to the isolated peptides in model systems. How-
ever, their experiments were developed on the premise

that Bio-Rad P-2 (Dorris, 1968) and Bio-Rad P-4 (Gizis

et a1., 1965) could separate unbound C057B12 from the

57B

peptide-Co 12 complex.

78

The experiments reported herein, indicated that
what Gizis et a1. (1965) and Dorris (1968) interpreted
as a BIZ-peptide complex was an unfortunate artifact

resulting from the elution of free vitamin B with the

12
peptides. Thus, it must be concluded that the peptides

isolated from B-Lg by electrodialysis do not bind B12

when mixed in model systems. In view of these experi-
mental findings, the original rationale of establishing

the binding mechanism between B and the protein-

12

associated peptides was abandoned.

Association of Vitamin B12

with Beta-Lactoglobulin

 

 

Vitamin B12 Binding
atApH 6.6-6.8

 

 

Beta-lactoglobulin when dissolved in 0.1M NH4OAo,
pH 6.8, and in Jenness and Koops buffer, pH 6.6, exhibited
an average binding capacity of approximately 460
ppg COGOBlz/mg of protein. Using Avagadro's number, the
molecular weights of vitamin B12 and B-Lg (36,000), and
the amount of B12 bound per mg of protein, a molar binding
ratio of 10 moles of vitamin B12 per mole of B-Lg was
estimated. If the 460 pug of vitamin B12 bound/mg
protein as determined experimentally and the vitamin B12
content inherently associated with B-Lg (i.e., 870 ppg
BlZ/mg protein; Gizis ep_gl., 1965) are combined, the

amount of vitamin B12 bound to B-Lg would be in the range

79

of 1300 ppg B12/mg protein. This B12 binding value at
pH 6.6 and 6.8 represents approximately one-half that
reported by Gizis §E_§l, (1965) and twice that reported
by Dorris (1968) for experiments performed in borate
buffer at pH 9.0.

Vitamin B12 Binding
at pH 9.0

 

 

At pH 9.0 the B12 binding capacity of B-Lg
varied with the buffer system employed. In borate
buffer at pH 9.0, B-Lg bound 150 ppg of vitamin Blé/mg
of protein, less than half the binding capacity exhibited
by B-Lg in the pH range of 6.6 to 6.8. The molar bind-
ing ratio of vitamin B12 to B-Lg in borate buffer de-
creased to approximately one-fifth the value observed
at pH 6.6 to 6.8. This reduced binding capacity of B-Lg
was also exhibited in the experiments performed with an
ammonium acetate-sodium hydroxide solvent at pH 9.0
(Table 6). The calculated molar binding ratio indicated

that 1 mole of vitamin B 2 was bound per moleoof B-Lg.

1
Some insight into these binding characteristics
in both borate buffer and ammonium acetate-sodium
hydroxide solvent at pH 9.0 can be obtained from the
ultracentrifuge data reported in Tables 1 and 2. Sedi-
mentation coefficients for the B-Lg in a borate buffer
system was 2.1 820,w' When compared to a sedimentation

coefficient of 2.9 S at pH 5.1 (a dimeric unit of

20,w

80

36,000 MW) it is apparent that the B-Lg preparation
exists in its monomeric state, i.e. 18,000, in borate
buffer at pH 9.0. Thus, it is proposed that the loss of
vitamin B12 binding capacity by B-Lg at pH 9.0 could be
the result of a decrease in electrostatic binding sites
concomitant with the increase in pH and/or a change in
the conformation of the dissociated B-Lg.

The unusually high vitamin B12 binding capacity
of B-Lg in TRIS-HCl buffer system at pH 9.0 (Table 6)
is attributed to the direct association of COGOB12 with
the TRIS ions. The basis for this postulate arises from
the relatively more diffuse elution pattern of the

COGOB12 in TRIS buffer when compared to the elution of

COGOB12 in other buffer systems. Thus, the apparent
binding was considered to be a peculiarity associated
with the TRIS-HCl buffer system.

Vitamin B12 Binding

at_pH 2.0

 

 

The vitamin B12 binding capacity of B-Lg was
measured at pH 2.0 where the protein also exists in its
monomeric state. At this low pH, there was no apparent
association of B12 to the protein; a characteristic
attributed to the electrostatic repulsion of the pro-

tonated protein molecules. This binding characteristic

has also been reported by Ford et a1. (1969) concerning

81

the dissociation of the folate-B-Lg complex below pH
3.65.

These data concerning the binding of B12 to B-Lg
at various values of pH supplement the results reported
by Gizis gp_§l. (1965) who reported that B-Lg in borate

buffer, pH 9.0, bound B in greater quantities than

12
inherently associated with the native protein. However,
the data from these experiments indicate that the optimum
pH range for the ip 21239 adsorption of 312 by B-Lg was
6.6-6.8, the normal pH range of fresh milk.

Effects of Electrodialysis on
Beta-Lactoglobulin

 

 

The reduced vitamin B binding capacity ex-

12
hibited by (E)B-Lg was believed by Dorris (1968) to be

a result of the loss of the associated peptides which

he reported as having a high B12 binding capacity. How-
ever, the present investigation has established that
these peptides do not bind vitamin B12 in model systems.
Consequently, alternate explanations for the reduction

in B binding capacity as a result of electrodialysis

12
were sought.

Chemical analyses, sedimentation velocities and
polyacrylamide gel electrOpherograms reported for B-Lg
and (E)B-Lg (see Results section) were in agreement with

comparable values for B-Lg published by Tilley (1960),

Bell and McKenzie (1964) and Melachouris (1969). The

82

interpretation of these data suggests that electrodialy-
sis exhibits a subtle change in the protein.

In view of this observation an investigation was
initiated to determine the relative rates of enzymatic
hydrolysis of B-Lg and (E)B-Lg as an indication of molec-
ular change. The rate of proteolysis of B-Lg and (E)B-Lg
by trypsin and chymotrypsin (Figures 6 and 7), as esti-
mated by pH Stat-monitored proton release at pH 8.0,
illustrated that electrodialysis could have affected the
conformation of the B-Lg molecules. Relative Km values
calculated for the proteolytic hydrolysis of both B-Lg
and (E)B-Lg (Figures 6 and 7) showed a lower value for
(E)B-Lg with both trypsin and chymotrypsin, indicating
a more rapid rate of hydrolysis.

Although Monnot §E_gl. (1967) reported that
hydrolysis of B-Lg with trypsin did not follow the
Michaelis-Menten law because of allosteric kinetics, it
is proposed that the observed reaction velocity data are
significant if interpreted only as a difference between
the relative reaction rates of B-Lg and (E)B-Lg and not
as a means of determining specific reaction kinetics of
the proteolytic reactions. Therefore, within the limits
imposed upon these data, it was concluded that the
increased rate of hydrolysis observed for (E)B-Lg is
indicative of a change in its molecular conformation.

Although the experimental data do not contain the

“WI? '1." n'.'::t"..'f".-‘;'_.’ 51137”;

83

information required to discern the cause of this change,
it is postulated that the removal of the protein bound
salts and/or peptides from the B-Lg molecules resulted
in an increase in intramolecular bonding. This hypo-
thesis is supported by published findings of Tanford
(1961) concerning the increased hydrophobicity of pro-
tein molecules in low ionic strength solutions.
Association of Vitamin B12 with
Electrodialyzed Beta-LaOEOgIObuIIn
andFElectrodiaIyzed Beta-

Lactoglobulin Peptide
MIxture

 

Vitamin B12 Binding

Previous studies have shown that following
electrodialysis, B-Lg exhibited a large reduction in
vitamin B12 binding capacity in borate buffer at pH 9.0.
Dorris (1968) attributed this behavior of the protein to
the loss of its associated peptides. He supported these
findings by demonstrating that (E)B-Lg had a B12 binding
capacity equal to that of B—Lg when the electrodialyzed
protein was equilibrated with the peptide mixture prior
to the addition of C060B12. As previously stated the
results from the current research indicated that the
peptides recovered from (E)B-Lg did not bind to B12.
Therefore, the binding capacity of the (E)B-Lg was de-
termined at various levels of pH and in various buffer

systems to determine the effect of pH and specific

buffer ions on the B12 binding capacity of (E)B-Lg.

 

84

(E)B-Lg and (E)B-Lg in combination with the pep-
tides in 0.1M NH4OAc, pH 6.6 showed binding capacities
for COGOB12 that were less than 10 ppg BIZ/mg protein.
However, slight variations associated with the gamma
emissions of COGOB12 required that binding values of
10 ppg/mg protein or less be reported as too low to
determine.

The inability of the (E)B-Lg and the recombined
(E)B-Lg-peptide mixture at pH 6.6 and 6.8 to bind B12 is
attributed to a reduction in the number of primary bind-
ing Sites previously proposed. In view of the binding
of B12 by the electrodialyzed protein at pH 9.0 in both
borate and ammonium acetate-sodium hydroxide solvents,
the loss of protein bound salts and the subsequent loss
of secondary binding sites for the B12 was not believed
to be of principal importance in the loss of B12 binding
by the (E)8-Lg.

Vitamin B12 Binding
at pH 9.0

 

 

Whereas the vitamin B12 binding capacity of B-Lg
in bOrate buffer at pH 9.0 decreased, the binding capaci-
ties Of the (E)B-Lg and (E)B-Lg-peptide mixture increased
to 40 and 200 ppg CO6OB12/mg protein, respectively.

These data represent a molar binding ratio of 1 mole of
vitamin B12 for every 2 moles of (E)B-Lg and 3 moles of

vitamin B12 for every 1 mole of recombined (E)B-Lg—peptide

85

mixture in borate buffer at pH 9.0. The binding capaci-
ties of the (E)B-Lg and recombined (E)B-Lg-peptide mix-
ture reported here are lower than those reported by

Dorris (1968) (i.e., 133 ppg Co6O

BIZ/mg of (E)B-Lg and
244 ppg COGOBlz/mg of recombined mixture). The discrep-
ancies arising between the vitamin B12 capacities for
these two forms of modified B-Lg are attributed to dif-
ferences in the electrodialysis technique.

The electrodialysis system employed in this study
consisted of placing the anode inside the dialysis mem-
brane with the cathode positioned in the diffusate.
Placement of the electrodes in this manner together with
a constant stirring of the protein suspension assured
that the electrostatic potential, which is believed
responsible for the dissociation of the salts and pep-
tides from the B-Lg, was directed at the protein inside
the membrane. The electrodialysis technique described
by Dorris (1968) indicated that both electrodes Were
placed in the diffusate. Placement of the electrodes in
this manner would result in dissociation of salts and/or
peptides from the B-Lg when the electrostatic potential
was initially applied. However, as the charged particles
diffuse through the dialysis membrane, the maximum elec-
trostatic potential of the system would no longer be

directed toward the B-Lg inside the membrane.

86

A protein system electrodialyzed as described
by Dorris (1968) would be subjected to a lower electro-
static potential, contributing to a lower recovery of
peptides and salts from the protein. Consequently, the
protein would be exposed to less stress and possibly
undergo less change in conformation than if one elec-
trode had been placed inside the dialysis sac. This
hypothesis would be difficult to prove, but is offered
as a possible explanation for the higher vitamin B12
binding activities of the (E)B-Lg and the (E)B-Lg-peptide
mixture reported by Dorris (1968).

Although (E)B-Lg showed no binding at pH 6.6 and

6.8, a significant binding of B by the (E)B-Lg was

12
observed at pH 9.0. Ultracentrifugation data (Tables

1 and 2) show that the (E)B-Lg in this buffer system
dissociated to its monomeric state. Therefore, vitamin
B12 binding by (E)B-Lg may be a result of secondary
binding to the borate ions associated with the protein
monomers. An alternate explanation for the B12 binding
capacity of (E)8-Lg (1 mole B12 per 2 moles of (E)B-Lg)
is that the change in protein conformation resulting
from electrodialysis and subsequent monomerization of
the (E)B-Lg yielded only one monomer from the original

(E)8-Lg dimer (36,000 M.W.) with an available binding

site for vitamin B12.

87

The recombined system of (E)B-Lg and peptides in
borate buffer at pH 9.0 showed a regain in vitamin B12
binding capacity similar to that reported by Dorris
(1968). The binding capacity of 200 ppg vitamin BIZ/mg
of (E)B-Lg-peptide mixture resulted in a molar binding

ratio of 3 moles of C0608

12 per mole of (E)B-Lg (18,000
M.W.). This regain in the binding capacity of the
(E)B-Lg when equilibrated with the peptides in borate
buffer at pH 9.0 may be indicative of an interaction be-
tween the borate buffer ions and peptides, negating the
effects of the proposed conformational change in (E)B-Lg.

The extremely high binding capacities measured
for all three forms of B-Lg in TRIS-HCl buffer as pre-
viously discussed are attributed to TRIS-associated
COGOB12 and, as such, are considered artifactual.

The static binding capacities measured for the
(E)B-Lg and (E)B-Lg-peptide mixture in NH4OAc-NaOH sol-
vent at pH 9.0 (Table 6) as well as B-Lg in the same
solvent system are believed to reflect the actual elec-
trostatic binding sites available on the B-Lg molecules
at pH 9.0. The calculated molar binding ratio indicated

that 1 mole of vitamin B is bound per mole of B-Lg.

12
These data support the interpretation that the peptides
are not directly responsible for the higher binding

capacity of B-Lg and that the peptides alone are

88

responsible for the regain in the binding capacity of

(E)B-Lg at pH 9.0.

Mode of Association Between B12

 

and Beta-LactogIObulin

 

Although the binding of B12 by B-Lg has been re-
ported previously, the type of association involved in
the formation of the protein-B12 complex was not deter-

mined. In an effort to elucidate the type of binding

60 60
B12

protein) was dissolved in 0.1M NH4OAc and dialyzed for

suspected a B-Lg-Co complex (460 ppg CO BIZ/mg

24 h against pure solvent. Scintillation spectrometry

data Showed that the B was not dialyzed from the pro-

12

tein, indicating that the B 2 was associated with the

1
protein. Aliquots of both B-Lg and (E)B-Lg-vitamin
B12(C060) complexes were then subjected to polyacryl-
amide gel electrophoresis at pH 8.3 (see Figure 12).
At the conclusion of the electrophoretic run, the

60

CO B 2 was not associated with the protein, nor was the

1
B12 located anywhere in the running gel. These data
indicate a dissociation of the COGOB12 from the protein
prior to entering the gel, which would imply a loose
association of the B12 with the protein, possibly through

electrostatic or polar association.

Peptides Isolated from Beta-Lactoglobulin
HVPE and TLHVE showed three peptides present in

the electrodiffusate from B-Lg (see Figure 10). It is

89

assumed that these peptides were not contaminants in
the protein system, since non-associated peptides should
have been removed from the system during the protein
purification procedure. Furthermore, no evidence of pro-
tein decomposition was established from amino acid
analyses, n-terminal amino acid analyses, -SH determina—
tions, sedimentation velocities and gel electrOpherograms
of the B-Lg and (E)B-Lg as reported in the results.
These data in addition to the findings of Moretti gp_§l.
(1958) and Gordon (1960), who have shown that proteins
recovered by electrodialysis (500 V potential) of starch
gels following electrophoresis Showed no indications of
protein decomposition, support the hypothesis that the
peptides electrodialyzed from the B-Lg were originally
associated with the protein.

The recovery of peptides from B-Lg was in the
range of 7 to 8 mg/g of B-Lg (0.75% of the protein).
A more accurate determination of the peptide recovery
from the B-Lg by electrodialysis was difficult to obtain
because of the large volumes of electrodiffusate (8,000
to 9,000 ml) which had to be concentrated prior to
lyophilization. The percentage recovery reported,
herein, was higher than has been reported in previous
studies (Dorris, 1968), and is attributed to a more com-

plete electrodialysis of the B-Lg.

9O

TLHVE data established the relative concentra-
tions of the peptides in the electrodiffusate as: +2
peptide > +3 peptide > negative peptide. During pre-
liminary studies this pattern of peptide concentration
was reversed when the electrode inside the dialysis sac
was connected to the cathode of the power supply unit.
Placement of the electrodialysis electrodes in this
manner Significantly lowered the recovery of peptides
because the positively charged peptides did not traverse
the membrane as readily. In View of these results the
cathode was positioned outside the dialysis sac as a

standard procedure.

Estimated Molecular Weight of Peptides

 

Gel filtration chromatography Of the peptide
mixture over Bio-Rad P-2 (Figure 9) indicated that the
peptides possessed molecular weights greater than 1,500.
When the peptides were passed over the Bio-Rad P-4
column (VO of 63 m1, Figure 11) their elution volume
was 110 ml, indicating that the peptides appear to have
similar molecular weights. The shoulders exhibited on
the elution pattern (Figure 11) were not interpreted as
resolution of the peptides, since each collected eluate
fraction was shown by TLHVE to contain the peptide mix-
ture.

Following the separation of the peptides by

TLHVE, individual peptides were again chromatographed to

91

ascertain if molecular interactions between the peptides
were responsible for the elution characteristics ex-
hibited by the peptide mixture. The elution volumes of
the individual peptides indicated that each peptide ex-
hibited the elution patterns similar to that of the pep-
tide mixture, i.e., characteristics of a globular protein
with a molecular weight range of 1,500 to 3,600. However,
little is known concerning these peptides and it is pos—
sible that they may have molecular weights larger than

is indicated from gel filtration chromatography data
because of unusual composition or structural character-
istics.

The inadequacies of the gel filtration data in-
ferred that the minimum molecular weights of the isolated
peptides be estimated from their amino acid composition.
These estimations also generated problems in that the
limited amount of peptides available for amino acid
analysis necessitated the calculation of the limiting
amino acid residue by integration of its absorbance
curve. Consequently, only amino acid residues which
could be accurately measured were used to make these
calculations. The estimated minimum molecular weights
were: negative peptide, 4,000; +2 peptide, 6,000; +3
peptide, 6,000. Although these values did not agree with
the values determined by gel filtration chromatography,

they do compare favorably with values Of 3,000 to 9,000

92

Daltons reported by Gizis §E_Ei° (1965) for two peptides
Obtained in a similar manner from electrodialyzed skim—
milk. Dorris (1968) also reported an approximate molecu-
lar weight of 4,000 for a peptide mixture recovered from
electrodialyzed B-Lg from gel filtration chromatography
data. These data support the hypothesis that the pep-
tides may have a secondary structure causing them to
behave like much smaller molecules on gel filtration.

Mode of Binding Between Peptides
and Beta-LactogIObuIin

 

 

The question now arises, ”Does the B-Lg contain
a specific binding site for the associated peptides?”
The protein-peptide binding ratio as estimated from the
average molecular weight of the peptides (2,500 Daltons)
indicated approximately one mole of peptides per 10 moles
of B-Lg. If the binding ratio is based on the average
minimum molecular weight of the peptides as determined
from amino acid composition (5,000 Daltons) the molar
binding ratio is reduced to one mole of peptide per 20
moles of B-Lg. Regardless of which value is considered,
it appears certain that the binding of peptides to B-Lg
is non-Specific and is a random electrostatic associa-
tion between peptides and protein.

This smallnumber of B-Lg molecules involved in
the peptide-protein association, as well as the large

molecular weight differential between the peptide and

93

B-Lg, also answers the question of why the sedimentation
coefficients of B-Lg did not vary significantly from that

of (E)B-Lg.

SUMMARY AND CONCLUSIONS

The adsorption of vitamin B12 by B-Lg, (E)B-Lg
and the (E)B-Lg-peptide mixture was determined by equili-
brating a known CO6OB12 or COS7B12 standard solution with
the proteins for l h. The unbound vitamin was separated
from the protein by gel filtration chromatography. The
radioactive B12 associated with the protein was measured
by scintillation spectrometry. These experiments showed
that adsorption of the vitamin by B-Lg, (E)8-Lg and the
(E)B-Lg-peptide mixture was pH dependent.

B-Lg exhibited its greatest binding capacity at
pH 6.6 and 6.8 (460 ppg BlZ/mg protein). The reduction
in binding capacity by (E)B-Lg and the (E)B-Lg-peptide
mixture at this pH range was attributed to a reduction
in the number of primary binding sites on the electro-
dialyzed protein molecules.

At pH 9.0 B-Lg exhibited a binding capacity of
150 ppg vitamin B12 per mg of protein in a borate buffer
and 100 ppg per mg of protein in an ammonium acetate-
sodium hydroxide solvent. The experimental results
indicated that this reduction in the binding capacity of

B-Lg could have resulted from a decrease in its electro-

static binding sites and/or a conformational change in

94

95

the dissociated dimers. Whereas, B-Lg showed a decreased
adsorption of vitamin B12 at pH 9.0, (E)B-Lg and the
(E)B-Lg-peptide mixture showed an increase in vitamin B12
adsorption. This increase was attributed to the second-_
ary binding of the vitamin to the buffer ions associated
with the (E)B-Lg and/or the dissociation of the (E)B-Lg
molecules at pH 9.0 which may have negated the effects of
electrodialysis on the B-Lg.

The static binding capacities of B-Lg, (E)B-Lg
and the (E)B-Lg-peptide mixture for vitamin B12 in
ammonium acetate-sodium by droxide solvent support the
interpretation that the protein-associated peptides are
not directly responsible for the higher binding capacities
of B-Lg, or that the peptides alone are responsible for
the regain in the binding capacity of (E)B-Lg at pH 9.0.

The inability of the vitamin to bind to either
B-Lg or the (E)B-Lg model system at pH 2.0 was an indica-
.tion of electrostatic repulsion by the protonated protein

molecules.

High-voltage paper-electrophoresis and thin-
layer high-voltage electrophoresis showed that three
peptides were electrodialyzed from the purified B-Lg.
Experimental findings revealed that these peptides were
electrostatically associated to the protein. Calcula-
tions of the protein-peptide molar binding ratio showed

that the binding of peptides is non-specific and probably

96

a random association between peptides and protein. The
molecular weights of the isolated peptides were not
accurately determined. However, gel filtration chromato-
graphic parameters and minimum molecular weight calcula-
tions made from amino acid analyses indicated molecular
weights in the range of 2,500 to 5,000, which may indi-
cate the presence of an unusual secondary structure in
‘the peptides.

Vitamin B12 binding experiments carried out with
the peptide mixture and the individual peptides showed
no binding Of the vitamin by the peptides. Recombina-
tion experiments with (E)B-Lg and the peptide mixture

at pH 6.6 (0.1M NH OAc). PH 6.8 (J and K buffer) and

4
pH 9.0 (NH4OAc-NaOH) showed no regain of binding capac-
ity. From these data it was concluded that the decrease
in vitamin B12 binding capacity by electrodialyzed B-Lg

cannot be attributed solely to the loss of the protein-

associated peptides.

LITERATURE CITED

97

LITERATURE CITED

Aschaffenburg, R. and J. Drewry. 1955. Occurrence of
different B-lactoglobulins in cow's milk.
Nature 176:218.

Aschaffenburg, R. and J. Drewry. 1957a. Improved method
for the preparation of B-lactoglobulin and d-
lactalbumin from cow's milk. Biochem. J. 65:
273. _—

Aschaffenburg, R. and J. Drewry. 1957b. Genetics of
the B-lactoglobulins of cow's milk. Nature 180:
376.

Aschaffenburg, R., D. W. Green and R. M. Simmons. 1965.
Crystal forms of B-lactoglobulin. J. Mol. Biol.
13:194.

Aulakh, J. Singh. 1967. A study of lipid oxidation in
model systems of copper—proteins, milk lipid and
milk dialysate. Ph.D. dissertation, Michigan
State University, East Lansing, Michigan.

Baker, H. P. and H. A. Saroff. 1965. Binding of sodium
ions to beta-lactoglobulin. Biochem. 4:1670.

Barker, H. A., R. D. Smyth and H. Weissbach. 1958. A
coenzyme containing pseudovitamin B12. Proc.
Natl. Acad. Sci. 44:1093.

Barker, H. A., R. D. Smyth, H. Weissbach, A. Munch-
Petersen, J. I. Tookey, J. N. Ladd, B. E. Volcani
and R. M. Wilson. 1960. Assay, purification
and properties of the adenosylcobamide coenzyme.
J. Biol. Chem. 232:181.

Basch, J. J. and S. N. Timasheff. 1967. Hydrogen ion
equilibrium of the genetic variants of bovine
B-lactoglobulin. Arch. Biochem. Biophys. 118:
37.

Bell, K. 1962. One-dimensional starch-gel electro-
phoresis of bovine skim-milk. Nature 195:705.

98

And.-

99

Bell, K. and H. A. McKenzie. 1964. B-lactoglobulins.
Nature 204:1275.

Bell, K. and H. A. McKenzie. 1967. The whey proteins
of bovine milk; B-lactoglobulins A and B.
Biochem. Biophys. Acta. 147:123.

Bollinger, H. R. 1965. IV. Thin-layer chromatography
of water-soluble vitamins. In: Thin-Layer
Chromatography A Laboratory Handbook. p. 244:
Academic Press, Inc. Publishers. New York.

Brink, N. G., D. E. Wolf, E. Kaczka, E. L. Rickles,
E. L. Koniuszy, F. R. Wood and K. Folkers.
1949. Vitamin B12. IV. Further characteriza-
tion of vitamin B . J. Am. Chem. Soc. 71:
1854. 12 —

Bull, H. B. 1946. Monolayers of B-lactoglobulins.
II. Film molecular weight. J. Am. Chem. Soc.
68:745.

Bull, H. B. and B. T. Currie. 1946. Osmotic pressure

of B-lactoglobulin solutions. J. Am. Chem. Soc.

.22‘742'

Carr, C. W. 1953. Studies on the binding of small ions

in protein solutions with the use of membrane
electrodes. IV. The binding of calcium ions in
solution of various proteins. Arch. Biochim.
Biophys. 44:424.

Cecil, R. and A. G. Ogsten. 1949. The sedimentation
constant, diffusion constant and molecular
weight of lactoglobulin. Biochem. J. 44:33.

Collins, R. A., A. E. Harper, M. Schreiber and C. A.
Elvehjem. 1951. The folic acid and vitamin
B12 content of the milks of various Species.
J. Nutr. 42:313.

Collins, R. A., R. E. Boldt, C. A. Elvehjem and E. B.
Hart. 1953. Further studies on the folic acid
and vitamin B12 content of cow's milk. J.
Dairy Sci. §§;24.

Dorris, J. C. 1968. The effect of electrodialyzable
peptide fractions on the binding of vitamin B12
by beta-lactoglobulin and other proteins.
Ph.D. dissertation, Michigan State University,
East Lansing, Michigan.

100

Dunhill, P. and D. W. Green. 1965. Sulfhydryl groups
and the NZR configuration change in B- '
lactoglobulin. J. Mol. Biol. 4§:147.

Ellman, G. L. 1959. Tissue sulfhydryl groups. Arch.
Biochim. Biophys. 84:70.

Fasold, H. and G. Gundlach. 1963. Characterization of
peptides and proteins with enzymes. In: ‘
Methods of Enzymatic Analysis. Ed. Hans-Ulrich
Bergmeyer. Academic Press. New York.

Ford, J. E. 1957. Factors influencing the destruction
by heat of vitamin B in milk. J. Dairy Res.
24:360 12

Ford, J. E., D. N. Salter and K. J. Scott. 1969. The
folate—binding protein in milk. J. Dairy Res.
36:435.

Georges, C., S. Guinaud and J. Tonnelat. 1962. Etude
thermodynamique de la dissoccation reversible de
la B-lactoglobuline B pour des pH supérieurs a
5.5. Biochem. Biophys. Acta. £2:737.

Gilbert, G. A. 1963. Sedimentation and electrophoresis
of interacting substances. III. Sedimentation
of a reversible aggregating substance with con-
centration dependent sedimentation coefficients.
Proc. Roy. Soc. (London) A276:354.

Gizis, E., Young Kim, J. R. Brunner and B. S. Schweigert.
1965. Vitamin B12 content and binding capacity
of cow's milk proteins. J. of Nutrition 41:349.

Gizis, E., J. R. Brunner and B. S. Schweigert. 1965.
Vitamin B12 binding peptides in cows' milk.
J. Dairy Sci. 44:1704.

Gizis, E. and Leo M. Meyer. 1968. Small molecular
weight vitamin B12 binders in normal human serum.
Nature 217:272.

Gordon, A. H. 1960. A technique for elution of pro-
teins from starch gel. Biochem. Biophys. Acta
42:23. -

Gordon, W. G., J. Basch and Edwin B. Kalan. 1961. Amino
acid composition of B-lactoglobulin A, B and AB.
J. Biol. Chem. 236:2908.

101

Gray, W. R. 1967. Dansyl chloride procedure. In:
Methods in Enzymology Vol. XI. "Enzyme Struc-
ture." Academic Press, New York. p. 139.

Gregory, M. E. 1954. The microbiological assay of
vitamin B12 in the milk of different animal
species. Brit. J. Nutrition 4:340.

Gregory, M. E. and E. S. Holdsworth. 1955a. The
occurrence of a cyanocobalamin-binding protein
in milk and the isolation of a cyanocobalamin-
protein complex from sow's milk. Biochem. J.
59:329.

Gregory, M. E. and E. S. Holdsworth. 1955b. Some
properties of the cyanocobalamin-protein complex
from sow's milk, and the mode of linkage of
cyanocabalamin with protein. Biochem. J. gg:335.

Gregory, M. E., J. E. Ford and S. K. Kon. 1958. The
B vitamin content of milk in relation to breed
of cow and stage of lactation. J. Dairy Res.
25:447.

Green, D. and R. Aschaffenburg. 1959. Twofold symmetry
of the B-lactoglobulin molecules in crystals.
J. Mol. Biol. 4:54.

Hodgkin, Dorothy C., M. W. Porter and R. C. Spiller.
1949. Crystallographic measurements of the
antipernicious anemia factor. Proc. Roy.

Soc. B., 434:609.

Hodgkin, Dorothy, J. Kamper, M. Mackay, J. Pickworth,
K. N. Trueblood and J. G. White. 1965. Struc-
ture of vitamin B12. Nature 178:64.

Hartman, A. M., L. P. Dryden and G. H. Riedel. 1956.
Vitamin B 2 content of milk and milk products
as determ1ned by rat assay. J. Nutrition £2:77.

Kalan, E., Rae Greenberg and M. Walter. 1965. Studies
on B-lactoglobulins A, B and C. I. Comparison
of chemical properties. Biochemistry 4:991.

Kiddy, C. A., R. Townend, W. Thatcher and S. Timasheff.
1965. B-lactoglobulin variation in milk from
individual cows. J. Dairy Res. §£:209.

102

Kim, Young Park, E. Gizis, J. R. Brunner and B. S.
Schweigert. 1965. Vitamin B12 distribution in
cow's milk. J. of Nutrition 44:394.

King, R. L., J. R. Luck, I. I. Litman, W. G. Jennings,
and W. L. Dunkley. 1959. Distribution of
natural and added COpper and iron in milk. J.
Dairy Sci. 42:780.

Kon, S. K. 1961. Influence of heat treatment and light
on the composition and quality of milk. Fed.
Proc., 20Cl Pt. III:209.

Kumosinski, J. F. and S. N. Timasheff. 1966. Molecular
interactions of B-lactoglobulin. The Stoichio-
metry of the B-lactoglobulin Mixed Tetrameriza-
tion. J. Am. Chem. Soc. 44:5635.

Larson, B. L. and R. Jenness. 1955. B-lactoglobulin.
Biochem. Preparations 4:23.

Layne, E. 1955. Spectrophotometric and turbidimetric
methods for measuring proteins. In: Methods
in Enzymology. Vol. III. Academic Press, New
York. p. 447.

Li, C. H. 1946. Electrophoretic inhomogeneity of
crystalline B-lactoglobulin. J. Am. Chem. Soc.
68:2746.

Melachouris, N. 1969. Discontinuous gel electrOphoresis
of whey proteins, casein and clotting enzymes.
J. Dairy Sci. 52:456.

McKenzie, H. A. 1967. Advances in protein chemistry.
Vol. 22. Ed. C. Anfinsin, M. Anson, J. Edsall
and F. Richards. Academic Press, New York.

p. 176.

McKenzie, H. A. and W. H. Sawyer. 1966. Zone electro-
phoresis of B—lactoglobulins. Nature 212:161.

McKenzie, H. A. and W. H. Sawyer. 1967. Effect of pH
on B-lactoglobulins. Nature 214:1101.

Minot, G. R. and W. P. Murphy. 1926. Treatment of
pernicious anemia by a special diet. J. Am.
Med. Assoc. §Z;470.

103

Monnot, Monique, Hadassa Weintraub, and Jeannine Yon.
1966. Evidence of activation by substrate in the'
hydrolysis of B-lactoglobulins A and B by
Trypsin. Bull. Soc. Chim. Biol. 48:959.
Dairy Sci. Abstracts 24:110. 1967?

Moore, 8., D. H. Speckman and W. H. Stein. 1958.
Chromatography of amino acids on sulfonated
polystyrene resins. Anal. Chem. 24:1185.

Moretti, J. 1958. Application of electrophoresis on
starch gel to the study of haptoglobin. Bull.
Soc. Chim. Biol. 44;59. Chem. Abstracts 42:
1959.

Nozaki, Y., L. Bunville and C. Landord. 1959. Hydrogen
ion titration curves of B-lactoglobulins.
J. Am. Chem. Soc. 42:5523.

Ogston, A. G. and J. M. Tilley. 1955. Studies on the
heterogeneity of crystallized B-lactoglobulin.
Biochem. J. 42:644.

Ogston, A. G. and M. P. Tombs. 1957. The heterogeneity
of bovine B-lactoglobulin. Biochem. J. 44:399.

Pedersen, K. O. 1936. Ultracentrifugal and electro-
phoretic studies on the milk proteins. I.
Introduction and preliminary results with frac-
tions from skimmilk. Biochem. J. 44g948.

Piez, K., E. Davie, J. Folk and J. Gladner. 1961. B-
lactoglobulins A and B. I. Chromatographic
separation and amino acid composition. J. Biol.
Chem. 224:2912.

Ray, A. 1968. Association of proteins reduced by co-
operative binding of sodium dodecylsulfate.
Spec. Publs. Chem. Soc. 22:49.

Ray, A. and R. Chatterjee. 1967. Interactions of
B-lactoglobulin with large organic ions.
Conform. Biopolym., Pap. Int. Symp. Madras.
1:235.

Rickles, E. L., N. G. Brink, F. R. Koziusky, T. R. Wood
and K. Folkers. 1948. Crystalline vitamin B12.
Science 107:396.

104

Seibles, T. S. 1969. Interaction of dodecylsulfate with
native and modified B-lactoglobulin. Biochemistry
Easton 4:2949.

Smith, I. Chromatographic and Electrophoretic Technique.
Vol. II. Zone Electrophoresis. Interscience
Publishers, Inc., New York. 1960. pp. 158-179.

Smith, E. L., B. Ball and D. M. Ireland. 1952. B12
vitamins. II. Netural, basic and acidic cobala-
mins. Biochem. J. 42:395.

Smith, E. L. and F. J. Parker. 1948. Purification of
anti-pernicious anemia factor. Biochem. J.
42:VIII.

Spies, J. R. and D. C. Chambers. 1948. Chemical
determination of tryptophan. Anal. Chem. 24:30.

Spies, J. R. 1967. Determination of tryptophan in pro-
teins. Anal. Chem. 24:1412.

Tanford, C. 1961. Electrostatic free energy and poly-
electrolytes. In: Physical Chemistry of Macro-
molecules. John Wiley and Sons, Inc., New
York. p. 485-521.

Tanford, C. 1962. Contribution of hydrophobic inter-
actions to the stability of the globular con-
formation of proteins. J. Am. Chem. Soc. 44;
4240.

Tanford, C., L. G. Bunville and Y. Nozaki. 1959. The
reversible transformation of B-lactoglobulin at
pH 7. 5. J. Am. Chem. Soc. 81: 4032.

Tilley, J. M. 1960. The chemical and physical proper-
ties of bovine B-lactoglobulin. Dairy Sci.
Abstr. 22:111.

Timasheff, S. N. 1964. The nature of interactions in
proteins derived from milk. In: Symposium on
Foods: Proteins and their reactions. Avi,
Westport, Conn. p. 179.

Townend, R. 1965. B-lactoglobulins A and B: The envir-
onment of the Asp/Gly difference residue. Arch.
Biochim. Biophys. 109:1.

Townend, R. and J. J. Basch. 1965. Amino acid analysis
of the B-lactoglobulins from individual cows of
two phenotypes. Arch. Biochim. Biophys. 109:325.

105

Townend, R., C. A. Kiddy and S. Timasheff. 1961.
Molecular interactions of B-lactoglobulin. VII.
The hybridization of B-lactoglobulins A and B.
J. Am. Chem. Soc. 44:1419.

Townend, R. and S. Timasheff. 1957. The molecular
weight of B-lactoglobulin. J. Am. Chem. Soc.
14:3613.

Townend, R., L. Weinberger and S. Timasheff. 1960.
Molecular interactions in B-lactoglobulin.
IV. The dissociation of B-lactoglobulin below
pH 3.5. J. Am. Chem. Soc. 42:3175.

Townend, R., R. J. Winterbottom and S. Timasheff. 1960.
Molecular interactions of B-lactoglobulin. II.
Ultracentrifugal and electrophoretic studies
of the association of B-lactoglobulin below its
isoelectric point. J. Am. Chem. Soc. 42:3161.

Twaroz, Judith and B. L. Larson. 1962. Cellular identi-
fication of B-lactoglobulin synthesis in bovine
mammary cell cultures. Expt. Cell Res. 24:350.

Ungley, C. C. and G. A. Childs. 1950. Absorption of
vitamin B12 in pernicious anemia. III. Failure
of fresh milk or concentrated whey to function as
Castle's Intrinsic Factor or to potentiate the

. action of orally administered vitamin B12. Brit.
Med. J. 2:911.

Urnes, P. and P. Doty. 1961. Optical rotation and the
conformation of polypeptides and proteins. Adv.
in Protein Chem. 16:401. Ed. C. Anfinsen, K.
Bailey, M. Anson 53d J. Edsall. Academic Press,
New York.

Whipple, G. H., C. W. Hooper and F. S. Robcheit. 1920.
Blood regeneration following simple anemia. IV.
Influence of meat, liver, and various extractives,
alone or combined with standard diets. Am. J.
Physiol. 44:236.

Wirtz, J., S. Timasheff and V. Luzzati. 1964. Molecular
interactions in B-lactoglobulin. VIII. Small-
angle x-ray scattering investigation of the
geometry of B-lactoglobulin A tetramerization.

J. Am. Chem. Soc. 44:164.

106

Zittle, C. A., E. S. Della Monica, R. K. Rudd and J. H.
Custer. 1957. The binding of calcium ions by
beta-lactoglobulin both before and after aggre-
gation by heating in the presence of calcium
ions. J. Am. Chem. Soc. 14:4661.

APPENDIX

107

 

APPENDIX

Isolation of Beta-Lactoglobulin

 

Ten liters of fresh, mixed-herd milk were col-
lected from the Michigan State Dairy. Without allowing
the milk to cool, it was heated to 40 C. Twenty grams
of anhydrous sodium sulfate were slowly added to each
100 ml of milk, stirring continuously until all the
sodium Sulfate was dissolved.

After the addition of sodium sulfate, the temp-
erature was lowered to 25 C, effecting the precipitation
of globulins, proteose peptones, casein and fat. The
precipitate was filtered out with E and D No. 515 fluted
filter paper. The filtrate, containing B-Lg, d-
lactalbumin and serum albumins, was collected for further
fractionation.

Concentrated hydrochloric acid was added to the
filtrate (1 ml conc. HCl per 100 ml filtrate) to adjust
the pH to approximately 2. The resulting precipitate,
containing all of the proteins except B-Lg, was removed
by centrifugation in an International High-Speed Refrig—
erated Centrifuge, Model HR-l with a No. 856 head,

operated at 8,000 rpm for 30 min.

108

i I ...I.II.IIIII. II 7 III! I I I'll ‘ all «Iii .Il." l Irvin I Ill '1:

109

The supernatant, containing B-Lg, was filtered
through E and D No. 515 fluted filter paper and refil-
tered through S & S No. 47841/2 analytical grade fluted
filter paper to insure the removal of precipitated pro-
teins from the B-Lg.

The filtrate was adjusted to pH 6 with concen-
trated ammonium hydroxide (approximately 0.6 ml per 100
ml of filtrate was required) and 20 g of ammonium sulfate
per 100 m1 of filtrate was slowly added with continuous
stirring. This resulted in the precipitation of the
B-Lg. When the ammonium sulfate had completely dis-
solved, the solution was allowed to sit quiescently for
one hour to ensure complete flocculation of the B-Lg.
Beta-lactoglobulin was recovered by filtering the solu-
tion under vacuum through a thick layer of Johns-Mansvill
Hyflo Super-Cel or Super Cel filter aid deposited on
S & S No. 597 filter paper in a Buchner funnel. The
filtrate was discarded. It was important to keep the
filter aid covered with solution once filtration had
begun or a gelatinized layer of B-Lg formed on the top

of the filter aid, dramatically reducing the flow rate.

Following filtration, the filter cake was slur-
ried with deionized water and placed in Visking cellulose
dialysis tubing which had been boiled in dilute
ethylenediaminetetraacetic acid (EDTA) solution and

dialyzed against deionized water. After 12 h of dialysis

110

the contents of the dialysis tubing were filtered with
suction through Whatman No. 1 filter paper in a Buchner
funnel.

The filtrate containing the B-Lg was collected
and the pH adjusted to pH 5.8 with 1N hydrochloric acid.
A few drOps of toluene were added as a preservative.

The solution was placed in EDTA treated dialysis tubing
and dialyzed for 48 h against deionized water (dialysate
was changed every 12 h).

The B-Lg solution was then removed from the
dialysis membrane and the pH adjusted to 5.2 with 1N
HCl, which produced some cloudiness, and the dialysis
continued. A few hours after the pH was lowered to
5.2, small B-Lg crystals began to form, having the
appearance of small platelets.

Previous researchers have reported the presence
of a heavy oily layer of B—Lg instead of crystals, a
phenomenon also encountered in this study. However, by
refiltering the supernatant after centrifugation to
insure removal of precipitated d-lactalbumin and serum

albunims, this problem was eliminated.

Recrystallization and Purification
The B-Lg crystals from five separate isolations
were suspended in approximately 300 ml of deionized
water and analytical grade sodium chloride was added

until all the crystallized B-Lg was dissolved. The

111

solution was filtered through analytical grade filter
paper and the pH adjusted to 5.17 with lN HCl. The
solution was placed in an EDTA treated cellulose mem-
brane along with one or two ml of toluene and dialyzed
against deionized water at 4 C until crystallization was
complete.

Beta-lactoglobulin was recrystallized four times

using this procedure.

Nitrogen

Nitrogen determinations were performed using a
micro-Kjeldahl technique. The digestion mixture con-
in 500 ml

sisted of 5.0 g CuSO4 - 5H 0 and 5.0 g SeO

2 2
of concentrated sulfuric acid. Approximately 6 to 8 mg
of dried protein samples were prepared in triplicate for
each analysis. The samples were digested with 4 ml of
the digestion mixture over an Open gas flame for l h.
After the samples had cooled, 1 ml of 30% H202 was added
to each digestion flask and the digestion continued for
l h. When the contents of the digestion flasks had
cooled, the sides were rinsed with approximately 10 ml
of deionized water.

Twenty-five m1 of 40% NaOH were added to each
digestion mixture and the ammonia released was steam
distilled into 15 m1 of a 4% boric acid solution contain-

ing 5 drops of methyl red-bromcresol green indicator.

The indicator consisted of 400 mg of bromcresol green

112

and 40 m1 of methyl red in 100 mg of 95% ethanol. The
distillation was continued until a total of 75 ml of
distillate were collected in the receiving flask. The
ammonium-borate complex was titrated with 0.0194N HCl.
Both a blank and recovery standard were determined by
the above procedure.

The average recovery for the ammonium sulfate

standard was 98.5%.

Tryptophan

 

A 15 mg sample was weighed directly into a small
vial. To each sample vial, 100 pl of freshly prepared
pronase solution was added (containing 10 mg pronase per
m1 of 0.1M phosphate buffer, pH 7.5), and a drop of
toluene. The vials were stoppered and incubated for
24 h at 40 C. After cooling, the sample vials were
placed in 50 ml Erlenmeyer flasks containing 9.0 ml of
21.2N sulfuric acid (387 ml of H20 to 500 ml concen-
trated H2804) and 30 mg of p-dimethylaminobenzaldehyde
which had been prepared immediately prior to use. To
each sample vial was added 0.9 ml of 0.1M phosphate
buffer, pH 7.5. The vials were tipped over and the con-
tents were quickly mixed by gentle swirling.

The Erlenmeyer flasks were then placed in the
dark at 25 C for 6 h. After adding 0.1 m1 of 0.045%

sodium nitrite the reaction mixtures were shaken and

113

the color allowed to develop for 30 min in the dark at
ambient temperature. Transmittance of the samples was
read at 590 nm.

Duplicate blanks were treated as described above.
One set of blanks contained everything but the protein
being analyzed, and the other contained everything but
the protein and pronase. Using this method, the trypto-
phan inherent in the pronase was not attributed to the
protein being analyzed.

A standard curve for tryptophan, having a range
of 0 to 120 mg of tryptOphan, was prepared according to

Procedure E of Spies and Chambers (1948).

Sulfhydryl

 

A 5-10 mg sample of protein was dispersed in
1 ml of deionized water. Nine m1 of pH 7.0 phosphate
buffer, which was 8M with respect to urea, was added
to the sample solution and allowed to stand for 10 min
at room temperature. To a 3 ml aliquot of this sample
solution was added 0.02 ml of 5,5' dithiobis 2-
nitrobenzoic acid (DTNB) reagent (39.6 mg of DTNB in
10 m1 of pH 7.0 phosphate buffer). The tubes were
shaken and the transmittance was read immediately.

Because it was very difficult to obtain a suit-
able standard curve, the sulfhydryl content was deter-
mined using the extinct coefficient as reported by

Ellman (1959).

114

Amino Acid Analyses

 

Five mg dried protein and 20 mg dried peptide
samples were weighed directly into 10 ml ampules. Five‘
ml 6N constant boiling HCl were added to each ampule.
The contents of the ampules were frozen in a dry ice-
ethanol mixture and air was evacuated with a high-vacuum
pump by allowing the samples to slowly thaw. Upon com-
plete removal of gases, the samples were refrozen and
sealed under vacuum.

The sealed ampules were placed in an oil bath
at 110 C for a hydrolysis period of 22 h. Upon com-
pletion of hydrolysis, the samples were removed and
cooled to room temperature.

The sealed ampules were opened and 1 m1 of
non-leucine standard was added to determine transfer
losses. The contents of the ampules were transferred
to a pear-shaped evaporation flask and evaporated to
dryness with a rotary evaporator at 50 C. After evap-
oration to dryness, each sample was taken up in deionized
water and re-evaporated. This procedure was continued
until all the remaining hydrochloric acid residue was
removed.

The dried hydrolysates were dissolved in 0.067M
citrate-hydrochloric acid buffer, pH 2.2, and brought
to a final volume of 5 ml in the case of B-Lg, and to

1 ml in the case of peptides. Varying aliquots of these

115

protein hydrolysates were required for amino acid analy-
sis which depended upon the protein content of the
samples.

Amino acid standards were chromatographed with
the same buffers and ninhydrin solution used for the
protein samples to ensure accurate quantitation of amino
acids present.

The amino acid composition of the protein samples

was expressed as gram residue of amino acids per 100 g

protein.

N-Terminal Amino Acid Analysis

 

Approximately 10 mp moles of protein were dis-
solved in 0.5 ml of 8M urea solution (ammonia and
cyanate free), buffered with 0.5M sodium bicarbonate.
To this was added 0.5 ml of a strong dansyl chloride
solution (20 mg l-dimethylaminonapthalene-S-sulfonyl
chloride per m1 of acetone). The mixture was allowed
to react for 12 h at ambient temperature or several
hours at 37 C.

Salts, urea and dansyl hydroxide were removed
by passing the reaction mixture over a small Sephadex
G-25 column. The dansyl-labeled, desalted protein was
collected in a small test tube and dried by lyophiliza-
tion. To the dried sample, 0.5 ml of 6.7N HCl (made

from constant boiling HCl) was added and the tube sealed

116

with a propane torch. The sealed tube was placed in a

105 C Oil bath and hydrolyzed for 18 h. Upon completion

of hydrolysis the tubes were opened and the HCl removed

by drying over sodium hydroxide pellets in a dessicator.
The dried samples were taken up in 10 pl of

50% (v/v) pyridine-water solvent. This solvent ensures

transfer of any insoluble dansyl-amino acid derivatives.

The samples were Spotted on 20 x 20 cm thin-layer chroma-

tography plates precoated with Silica gel G1? (250

microns). The plates were placed in a Reco Model E-800-2,

water-cooled electrOphoretic migration chamber, containing

pyridine/acetic acid/water (10/20/2500 v/v) buffer at

pH 4.4 in the buffer tanks. The plates were sprayed

with the same buffer and electrOphoresed for 30 min

at a field strength of 50 volts/cm. Upon the completion

of electrophoresis, the plates were dried in a forced

air oven at 90 C. The n-terminal dansyl amino acid

derivatives were visualized under long-wave (365 nm)

ultraviolet radiation.

Colorimetric Determination of Protein
The reagents required were: Reagent A, 2%
sodium carbonate in 0.1N sodium hydroxide; Reagent B,

0.5% CuSO4-5H O in 1% sodium or potassium tartrate;

2
Reagent C, alkaline copper solution prepared by mixing

50 ml of reagent A with 1 m1 of reagent B (this should

117

be prepared fresh each day); Reagent D, Folin-Ciocalteu
reagent diluted to 1N. Reagent D can be Obtained com-
mercially from Fisher Scientific Company.

One milliliter of protein solution was placed in
a test tube and 5 m1 of reagent C were added. The tube
was shaken and allowed to stand for 10 min. To this
solution 0.5 ml of reagent D was added and the contents
Of the tube were mixed and allowed to stand for
30 min. The transmittance was read at 500 nm with a
Beckman Model DU-2 Spectrophotometer. A control blank
was prepared containing 1 ml of water in place of the
protein solution.

'The protein content was determined by conversion
of transmittance to pg of protein per ml, using a stand-
ard curve. The standard curve was prepared with bovine
serum albumin having a range of 0 to 100 pg protein per

ml.

Enzymatic Hydrolysis of Proteins

 

The pH stat was standardized at pH 8.0 and 37 C
with phosphate buffer. B-Lg and (E)B-Lg samples were
prepared in deionized water and the pH raised to 8.0
with 2N triethylamine before adjusting the protein con-
centration to 10 mg per m1.

Enzyme solutions were prepared in 0.02M CaCl2

at a concentration of 2.5 mg of enzyme per ml. The

118

enzyme preparations were stored at 4 C after adjustment
of pH to 8.0 with triethylamine. Immediately prior to
each enzymatic hydrolysis a syringe containing 1 ml of
the enzyme preparation was placed in a 37 C water bath
to insure optimum temperature for enzymatic activity.
The substrate concentrations were varied by
quantitative dilutions of the standard protein solution.
The final volume of protein solution applied to the pH-
stat was 10 m1. 'Each 10 ml substrate sample was per-
mitted to equilibrate at 37 C and pH 8.0 in the pH-Stat
prior to adding enzyme. The electrode control was then
activated to the titration position and 1 ml of the
enzyme solution at pH 8.0 and 37 C was added. During
enzymatic hydrolysis in the pH-Stat reaction vessel,
0.05N triethylamine was automatically titrated into the
reaction mixture to maintain a constant pH Of 8.0. The
volume of 0.05N triethylamine titrant was automatically
plotted relative to time. A blank containing deionized
water with pH adjusted to 8.0 with 0.05N triethylamine
was employed to determine alkaline consumption due to
enzyme degradation and carbon dioxide absorption.
Hydrolysis data were analyzed by plotting per-
centage total hydrolysis versus time. Thus, the hydroly-
sis curves for each concentration of native and elec-
trodialyzed B-Lg could be compared and the relative Km
values forB-Lg and (E)B-Lg for both trypsin and chymo—

trypsin calculated.