University of Southern California Dissertations and Theses

The separation of complete and incomplete blood group antibodies

(USC Thesis Other)

The separation of complete and incomplete blood group antibodies

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE SEPARATION OF COMPLETE AND INCOMPLETE
BLOOD GROUP ANTIBODIES
by
m m James Meacham
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(Biochemistry)
June 1960
U NIVERSITY O F SO U TH E R N C A LIFO R N IA
G R A D U A T E S C H O O L
U N IV E R S IT Y PARK
LO S A N G E L E S 7 , C A L IF O R N IA
This dissertation, written by
............................ Jam ea .M each am .........................
under the direction of h%®....Dissertation C om
mittee, and approved by all its members, has
been presented to and accepted by the Graduate
School, in partial fulfillm ent of requirements
fo r the degree of
D O C T O R O F P H I L O S O P H Y
....
^ Dean
...........
DISSERTATI ON C O M M IT T E E
Chalrigan
..
..
TABLE OF CONTENTS
PAGE
INTRODUCTION ............................................ 1
Blood Groups.................................... 3
Rh Blood Groups . ........................... . 8
Blood Group Antibodies ........................... 13
Separation of Blood Group Antibodies........ 23
STATEMENT OF THE PROBLEM AND PLAN OF ATTACK..... 27
Statement of the Problem.................... 27
Plan of Attack............................. . . 29
MATERIALS............................... 31
METHODS........................................... 34
Serological Methods . . * . . ................ 34
Fractionation Methods ............................. 36
Starch block electrophoresis ........ 36
Ultracentrifugation ......................... 38
Anion-exchange cellulose chromatography . . 38
Solvent fractionation ....................... 39
Stroma-resin adsorption .................... 40
Concentration Methods ............................. 40
Pervaporation ............................... 41
Concentration by dialysis .................. 42
Lyophllization ............................... 43
Salting out ...................... 44
li
lii
PAGE
Protein Determination ..... 44
Lowry method............................... 44
Kjeldahl method.......... 46
Neseler method ............................. 49
Biuret method................ 49
Analytical MethodB............ 50
Free boundary electrophoresis ............. 50
Paper electrophoresis..................... 50
Starch gel electrophoresis ................. 51
EXPERIMENTAL RESULTS.................................. 53
Expression of Results ........................... 54
Protein units ............................. 54
Antibody units ............................. 55
Antibody enrichment ....................... 57
Significant figures ....................... 58
Negative results ......................... 58
Electrophoresis.............. 59
Protein distribution ....................... 59
Antibody distribution .......... . 62
Antibody enrichment............ 82
Antibody recovery ................. .... 86
Evaluation................................. 90
Starch gel electrophoresis................. 91
Free boundary electrophoresis . ........... 92
PAGE
Sedimentation ................................... 92
Protein distribution ..... ........ . . 93
Antibody distribution ..................... 93
Antibody enrlohment ....................... 94
Antibody recovery ......................... 94
Evaluation ......................... 95
AnIon-Exchange Cellulose Chromatography ..... 99
Protein distribution.......... 99
Antibody distribution.....................100
Antibody enrichment ........................ 101
Antibody recovery .......................... 101
Evaluation................................. 101
Precipitation Methods ........................... 102
Protein distribution ........................ 103
Antibody distribution.....................103
Antibody enrichment.......................104
Antibody recovery ................ 104
Evaluation................................. 104
Stroma-Resln Adsorption ......................... 105
Protein distribution ........................ 105
Antibody distribution ...................105
Antibody enrichment ..... 105
Antibody recovery .......................... 105
Evaluation.................................105
V
PAGE
Combined Methods ............................... 106
Antibody Stability ............................. 107
DISCUSSION.......................................... 110
Expression of Results ......................... 110
Antibody Distribution ......................... 111
Electrophoresis ........................... 111
Sedimentation.......... 112
Modified cellulose ohromatography ........ 113
Precipitation methods..................... 114-
Antibody Enrichment........................... 114
Antibody Recovery ............................. 115
Other Antibodies............................... 117
General Statement ............................. 118
SUMMARY AND CONCLUSIONS ............................. 120
BIBLIOGRAPHY........................................ 122
APPENDIX - METHODS................................. 128
Serological Titrations ................. 128
Anticoagulant Solutions....................... 130
Starch Block Electrophoresis .. ............... 131
Anion-Exchange Cellulose Chromatography . . . 133
DEAE-Cellulose Adsorption, Batch Method .... 133
Solvent Fractionation of Proteins ....... 134
Stroma-Resin Adsorption ............... 135
Pervaporation................................ . 136
vl
PAGE
Concentration "by Dialysis..........................137
Lyophilization.................................... 137
Salting Out of Proteins............................138
Protein Determination, Lowry ................... 138
Nitrogen Determination, Mlcro-KJeldahl ........ 139
Nitrogen Determination, Nessler ................. 140
Protein Determination, Biuret ................... 141
Free Boundary Electrophoresis ................... 142
Starch Gel Electrophoresis ..................... 143
LIST OF TABLES
TABLE PAGE
I. Antibodies in Sera Studied................... 33
II. Correspondence of Standard Starch Block
Fractions with Serum Protein Fractions .... 61
III. Approximate Electrophoretic Mobilities of
Serum Fractions found in Starch Block .... 63
IV. Antibodies Fractionated by Starch Block
Electrophoresis ............................. 64
V. Summary Table of Starch Block Results on
Complete Anti-D ............................. 66
VI. Summary Table of Starch Block Results on
Incomplete Anti-D ........................... 67
VII. Summary Table of Starch Block Results on
Complete Anti-D ..................... 68
VIII. Summary Table of Starch Block Results on
Complete Anti-D ............................. 69
IX. Summary Table of Starch Block Results on
Incomplete Anti-D ........................... 70
X. Summary Table of Starch Block Results on
Incomplete Anti-D ................... 71
XI. Summary Table of Starch Block Results on
Complete Anti-E ............................. 72
XII. Summary Table of Starch Block Results on
Incomplete Anti-E . ......................... 73
XIII. Summary Table of Starch Block Results on
Incomplete Anti-E.................. 74
XIV. Summary Table of Starch Block Results on
Anti-A....................................... 75
XV. Summary Table of Starch Block Results on
Anti-B ............................. ..... 76
vii
▼Ill
TABLE PAGE
XVI. Summary Table of Starch Block Results on
Antl-B........................................ 77
XVII. Summary Table of Starch Block Results on
Antl-B........................................ 78
XVIII. Relation of Antibody Activity to Starch
Block Fraction . ........................... 79
XIX. Proportion of Antibody Activity in Albumin
Area on Starch Blocks....................... 80
XX. Antibody Enrichment ......................... 84
XXI. Complete/incomplete Ratio ................... 85
XXII. Summary of Sedimentation of Incomplete Anti-D 96
XXIII. Summary of Sedimentation of Anti-D ......... 97
XXIV. Summary of Sedimentation of Anti-A ...... 98
INTRODUCTION
The human blood groups, and the antibodies associated
with them, play a vital role in the welfare of the
individual and of the society in which he lives.
Knowledge of blood groups is applied in the selection of
blood for transfusions, in the diagnosis and treatment of
hemolytic disease of the newborn, in cases of disputed
parentage and other medicolegal matters, and in theoretical
studies in human genetics and physical anthropology.
Moses said, "The -life of the flesh Is In the blood"
(1), and this concept underlies many of the religious and
secular customs, of both primitive and more advanced
societies, which have been observed throughout the history
of mankind. Some societies have prohibited the drinking of
blood, while others have demanded it as a means of
acquiring the qualities of another person or animal. Blood
has been the symbol of relationships and has been used to
exact vengeance, to confirm kinship, or to initiate
outsiders Into a group. It has been thought of as giving
strength to the old, courage to the warrior, fertility to
the barren, power to the prophet, and life to the dead (2).
Belief in the peculiar powers of the blood led to an
early Interest in transfusion (3,4). In 1492, the dying
Pope Innocent VIII was given blood from three young boys in
an unsuccessful attempt at rejuvenation. In 1615, Andreas
Llbavius, the Chemist of Halle, described a fictitious
transfusion, and in 1654, Francesco Folli, a Florentine
physician, claimed to have transfused blood from one
animal to another. The first well documented accounts were
those published by English workers. In 1656, Sir
Christopher Wren gave an intravenous injection of blood to
a dog, and in 1665 Richard Lower transfused blood from one
dog to another. Two years later, Jean Denys of Montpellier
administered lamb’s blood to a fifteen year old boy. A
German physician proposed transfusion as a means of
reconciling the partners in unhappy marriages. In 1669, a
third transfusion of sheep blood to an insane man proved
fatal, and transfusions were prohibited thereafter in
France. Similar experiences in other countries caused the
practice to be abandoned as a therapeutic measure. Toward
the end of the century Pierre Bayle, the French skeptic,
inquired whether a transfusion would change a man’s
temperament and whether sheep blood would change a dog
into a sheep.
In 1818, the physiologist James Blundell first
transfused human blood into a human being, but his
experiments and those of subsequent workers often had
unfavorable consequences (5). Ho further progress was made
in the art of transfusion during the nineteenth century.
Blood Groups
Landsteiner1s discovery, in 1901, of the human blood
groups revealed the major source of difficulty in these
early transfusions (6). He found that the serum of some
individuals would agglutinate the red blood cells of
others. On the basis of his observations, Landsteiner
could distinguish three mutually incompatible blood groups.
Later work revealed a fourth group. A few years later
Jansky and Moss applied these findings in blood
transfusion. They designated the four groups by Roman
numerals, but today they are more commonly known as the
AB, A, B, and 0 of the familiar Landsteiner, or ABO, blood
group system.
The hereditary character of blood groups, which follow
Mendel’s laws of segregation and independent assortment,
was first recognized by Epstein and Ottenberg in 1908 (7).
According to Bernstein's theory (8,9), a locus on the
chromosome may be occupied by either (or neither) of two
allelomorphic genes, A and B. The absence of a gene is
denoted by 0. Pairing of the chromosomes can give rise to
six different genotypes, AB, AA, AO, BB, BO, and 00; but
only four phenotypes can be distinguished: AB, A, B, and 0.
Later work has shown that 0 is actually a third allelomorph
( 10,11 ).
Somatic expression of these genes consists in the
presence of specific blood group substances, A, B, and 0,
in the cellular membrane of the erythrocytes. These blood
group substances are mucopolysaccharides which will react
with specific plasma proteins, antibodies, to cause
agglutination of the cells. Agglutination tests with sera
containing anti-A and anti-B provide a simple and sensitive
means of determining the blood groups of individuals.
Antisera to 0 substance are very rare, so that genotypes AA
and AO, or 3B and BO, cannot be distinguished routinely.
On detailed study it was found that individuals
belonging to groups A and AB can be divided into several
mutually compatible subgroups according to the occurrence
of different kinds of A substance in their cells.
Substances Aj , A2, A^, A^, A^, and 0 show progressively
weaker reactions with anti-A sera and progressively
stronger reactions with anti-0 or antl-H sera. Hlrszfeld
(12) suggested that blood group A-j , the most common
subgroup, has been evolved from group 0, or vice versa.
Corresponding subgroups of B have not been found.
Anti-0 and anti-H sera, while they both react with 0
cells and weakly with A2 cells, can be differentiated by
virtue of the fact that antl-H is inhibited by saliva from
so-called secretors and anti-0 is not. Anti-H also occurs
in some eel sera and is produced by goats Immunized to
Shigella dysenterlae (13).
The gastrointestinal tract and other soft tissues also
contain varying amounts of materials with the same
specificities as the blood group substances. In
individuals known as secretors, these specific materials
are water soluble, and they appear in the saliva and other
secretions. Similar specific materials can be extracted
with alcohol from the tissues of non-secretors. The
secretor phenomenon is inherited as a dominant trait (14).
Nearly all individuals have in their plasma antibodies
to the blood group substances not present in their own
erythrocytes, with the exception that antibody to group 0
substance is rarely found. For example, individuals with
type A substance in their cells will have anti-B in their
plasma, but not anti-A.
These are natural, or normal, antibodies, which are
not known to be produced in response to an antigen.
Several theories have been proposed to explain the
existence of natural antibodies. It is quite possible that
their production has been stimulated by minute amounts of
materials which have entered the body from intestinal
bacteria, foods, clothing, airborne dust, or other sources.
Materials with specificities similar to those of A and B
blood group substances are widely distributed in both the
plant and the animal worlds. Another possibility is that
the ability to produce antibodies of this kind is
inherited. Filitti-Wurmser and coworkers (15) have shown
that anti-B differs in persons of different genotypes (AjO,
A-|Ai, 00, AgO, AjO), and they have suggested that the
formation of anti-B is determined hy the same gene which
determines the production of blood group substance A.
As opposed to natural antibodies, immune antibodies
are produced in response to sensitization to a foreign
blood group substance which has entered the circulation,
usually through a pregnancy or a transfusion with
incompatible blood. Yftiile there are no known methods for
qualitatively distinguishing between natural and Immune
antibodies, the latter are generally present in much
greater quantities, so that the immune state is easily
recognized. Details of the immune response will be
discussed in a later section.
Specific plant agglutinins have been found for blood
group substances A (A1+A2), A1, H, and N. Some of these
lectins, as Boyd calls them (16), have a low degree of
specificity and are capable of agglutinating all human red
cells. It would seem that the specificities of lectins are
accidental, and Boyd suggested that this may also be the
case with natural antibodies in human plasma, which often
have less avidity than immune antibodies of the same
specificities.
In 1927, Landsteiner and Levine (17) discovered a
second blood group system, the MN system, which is
genetically Independent of the ABO system. Two
allelomorphic genes (M and N) are analogous to those of the
ABO system, the phenotypes being M, N, and MN. The
antigens belonging to these groups are found only in the
erythrocytes of man and other primates. Normal antibodies
do not occur for these antigens, and immune antibodies in
man are rare. Antisera are obtained from rabbits injected
with appropriate red cells. Later work has uncovered a
second pair of allelomorphs, S and s, which seem to be very
closely linked to the MN genes (18,19). These are not to
be confused with the S and s genes determining the secretor
type.
Studies of blood- groups in different races have
yielded provocative results (13,20). On the basis of
blood groups and other genetic traits, at least six
different races have been recognized: early Europeans,
Europeans, Africans, Asians, American Indians, and
Australian aborigines. It has been observed that there is
a close resemblance between the blood groups of Australian
aborigines and American whites, that Asiatic Mongoloids
have a high frequency of B while it is virtually absent
from American Indians, and that Bushmen and Hottentots
have quite different frequencies of B in spite of their
geographical proximity. Insufficient data are available
8
for a systematic analysis, but there can be no doubt of the
value of blood grouping as an anthropological tool. The
blood groups are easily and objectively determined, their
mode of inheritance Is well understood, and they remain
constant throughout the individual’s lifetime and are
unaffected by environment.
The anthropoid apes (gorillas, chimpanzees, orangs,
and gibbons) have ABO groups with substances very similar
to those found in man, though group 0 substance is
relatively rare in these animals. Substances similar to M
and N are also found in apes and monkeys.
While engaged in studies on group M substance in
primates, Landsteiner and Wiener (21) found that rabbits,
when injected with the blood of the rhesus monkey, will
elaborate antibodies which agglutinate about 85 per cent of
human erythrocytes. It was later shown that this
anti-rhesus (anti-Rh) had the same specificity as the
antibody found by Levine and Stetson in a case of hemolytic
disease of the newborn (22). Subsequent work has shown
that more than 90 per cent of the cases of this disease are
due to Rh incompatibility.
Rh Blood Groups
The Rh blood groups are genetically much more complex
than the ABO and MNS groups. Statistical evaluation of
accumulated data enabled Fisher to formulate his theory of
the genetic transmission of the Rh groups (23). He
postulated three closely linked loci on the chromosome
determining the Rh blood group, each locus occupied by one
of a pair of allelic genes (C or c, D or d, E or e). There
are eight possible combinations of three genes, and, with
pairing of homologous chromosomes, 36 possible genotypes.
Each gene effects the production of a corresponding blood
group substance. Using Fisher's system of nomenclature,
six letters are necessary to describe an Rh genotype,
usually written as CDe/cde, cDe/cDE, and so on. Another
system of nomenclature, developed by Wiener (24), has not
been generally adopted in practical work because it is
harder to understand and more cumbersome than Fisher's
system. V/iener proposed a single locus on the chromosome,
occupied by one of eight allelomorphs. These allelomorphs,
r, r’, r", r^, R°, R°' (or R1), R°* (or R2), and Rz,
correspond to the gene combinations cde, Gde, cdE, CdE,
cDe, CDe, cDE, and CDE, respectively. As in the ABO
system, a number of subgroups have been found.
Normal antibodies to Rh antigens do not occur, and
there is a distinct variation in the ability of individuals
to produce immune antibodies to these antigens. Whether
this ability is determined by the sum total of the
individual’s blood groups is not clear at this time. It
seems likely, however, that there is a greater or lesser
10
similarity in molecular structure among the different blood
group substances and that an individual would not be
capable of producing antibodies which would react even
weakly with substances in his own erythrocytes. Such an
explanation would presumably apply, for example, to the
apparent non-existence of antl-d and the rarity of anti-M,
anti-N, and anti-0.
Other genetically independent blood group systems
which have been discovered recently are the Lutheran, Kell,
Lewis, Duffy, Kidd, Jay, Wright, and Vel systems (13,25).
Antibodies to these occur only rarely in human sera. This
accounts for their relatively late discovery and the fact
that they seldom give rise to clinical problems. Also a
number of blood group substances, each apparently
restricted to a single family, have been Identified. The
Lewis blood groups are of interest because they are linked
to the secretor phenomenon and because one of the genes
(Lea) seems to be dominant in infants and recessive in
adults.
Correlation has been reported between blood groups and
the incidence of certain diseases. Peptic ulceration and
toxemia of pregnancy appear to be most common in
individuals of group 0, and gastric carcinoma and
bronchopneumonia in individuals of group A. Various
reasons for these correlations have been suggested, but not
11
substantiated. In some cases of acquired hemolytic
disease, the patient's serum has been found to contain
autoantibodies, which react with his own erythrocytes.
Rh antibodies are the most frequent cause of
erythroblastosis fetalis, or hemolytic disease of the
newborn. During pregnancy the fetus will sometimes
sensitize the mother to a blood group substance which she
lacks but which has been inherited by the fetus from the
father. In a later pregnancy, if the fetus possesses this
same blood group substance; the mother may produce large
amounts of antibody, which can cross the placenta and
hemolyze the red cells of the fetus. The resulting
disorder is known under various names according to the
clinical manifestations. Hydrops fetalis indicates a
stillbirth with extreme anemia and edema. In other
stillbirths the fetus may be macerated, with a cirrhotic
liver. The name icterus gravis is applied when the newborn
infant develops an elevated serum bilirubin level as a
result of the hemolysis. This may lead to kernicterus, or
staining of the brain cells with bilirubin, and the
associated nervous disorders. Erythroblastosis fetalis is
a more inclusive term, simply denoting the presence of
immature erythrocytes in the blood. Prognosis Is good if
exchange transfusions are given soon after birth (26). If
the infant Is type D and the antibodies anti-D, as Is
12
usually the case, type d blood is given. Delay in
treatment may result in permanent damage to the nervous
system.
All bloods entering the typical blood bank are
routinely typed for the ABO group and for the Rh factor
(D). Before a blood is used for transfusion, its serum and
cells are crossmatched against those of the prospective
recipient to preclude any transfusion reaction due to
incompatibility of blood groups. In emergency cases, a
group 0 blood ("universal donor") may be used; in such a
case, A and B substances are usually added to absorb out
the normal antibodies. Even when crossmatching has
established compatibility, the recipient still may become
sensitized to some unsuspected blood group substance
present in the donor's cells. A more thorough routine
typing of both the donor's and the recipient's blood would
be desirable, but the rarity of many antisera and the
amount of time which would be required make it
impracticable. Also, difficulties arising from antibodies
other than anti-A, anti-B, and antl-D are quite rare.
Assuming that crossmatching is always possible, the only
danger would be in the sensitization of a woman to some
antigen which would be carried by a future child. Such
cases must be extremely rare.
13
Blood Group Antibodies
Antibodies are not uniformly capable of agglutinating
red cells. Those antibodies which can agglutinate cells
suspended in isotonic saline solution are commonly known as
saline agglutinins or complete antibodies. Those which
cannot agglutinate red cells under these conditions are
called incomplete antibodies.
There are several kinds of incomplete antibodies. The
blocking antibody inhibits agglutination by complete
antibodies. The albumin antibody is capable of clumping
cells suspended in concentrated colloid solutions, such as
20 per cent albumin (27,28). Other colloids commonly used
are gum acacia, dextran, polyvinylpyrrolidone, and gelatin.
Other kinds of incomplete antibodies are capable of
clumping red cells which have been altered by treatment
with proteolytic enzymes, such as trypsin, papain, or ficin
(25,29). Still other kinds of incomplete antibodies can be
detected only with rabbit anti-human globulin serum, which
will clump cells which have been coated with the incomplete
antibody (30). A refinement of this last method consists
in reacting the coated and anti-human globulin treated
cells with anti-rabbit globulin serum. This is said to
increase the sensitivity of the method (31).
The kinds of incomplete antibodies described above are
not mutually exclusive. The same antibody will usually
14
react In more than one of the methods.
All these types of Incomplete antibody are not
necessarily found in the same individual. Some individuals
will produce one type, other another. However, the same
Individual can produce several different kinds of antibody,
in various proportions, at the same time; so that the
overall picture can become quite complex.
The strength of the antibody solution is determined by
making serial two-fold dilutions (1:1, 1:2, 1:4, 1:8, etc.).
The highest dilution at which antibody activity is still
detectable is called the titer of the antibody solution.
The procedure is known as a "titration" although it is
quite different, of course, from the acid-base, redox, or
other titrations used in analytical chemistry.
If there are two antibodies present, they may compete
and give rise to a prozone effect. In such a case,
agglutination may become progressively stronger as the
antibody solution Is diluted. Beyond the prozone, i. e.,
beyond a certain dilution, the strength of agglutination
will tend to decrease with increasing dilution In the usual
way.
A variation of the titration method is used for the
detection of blocking antibody. The cells are reacted
first with different dilutions of blocking antibody, then
with complete antibody of known strength. The highest
15
dilution of the blocking antibody serum which will inhibit
agglutination Is taken as the titer of the blocking
antibody. This titer will generally be a function of the
strength of the complete antibody used.
Aside from their clinical effects and serological
reactions, very little is known about blood group
antibodies. This is especially true of the Rh antibodies
because of their rarity and lability. All blood group
antibodies appear to be protein In nature.
The blood group substances, on the other hand, are
mucopolysaccharides (32) which apparently owe most of their
specificity to terminal monosaccharide groups:
galactosamine in A substance, galactose in B substance, and
fucose in 0 substance. This has been Indicated by several
kinds of evidence. Reactions of antibodies with these
antigens are inhibited by the respective monosaccharides,
and enzymatic degradatlonof the blood group substances is
also retarded by the same monosaccharides. Experiments
with proteolytic enzymes have shown that the amino acids
contained In the blood group substances are not essential
to their specificity. All three of these substances will
react with type XIV antipneumococcus serum. This suggests
that their basic structure is similar.
Dunsford and Bowley (25) list conditions which can
affect antibody production in an immunization experiment:
16
antigenic potential of the blood group substance, number of
previous exposures to the antigen, route of injection,
simultaneous introduction of other antigens, ability of the
individual to form antibodies, and intervals between
exposures to antigen. They state that the quantity of
antigen introduced is immaterial.
A minimum of two injections of antigen is required to
initiate production of antibodies, the first injection
acting as a primer. The second injection, given after a
latent period of several weeks, is followed within a few
days by appearance of antibody in the plasma. Some
Individuals, however, do not begin to form antibody until
after several injections of antigen have been given.
The first antibody to appear is often of the complete
type. Later, incomplete antibody predominates, and it may
persist for many years. This indicates a continuous
production of antibody, since its half-life is only 30 to
35 days (33).
Most or all tissues may be capable of producing
antibodies. This fact may account for the heterogeneity of
antibodies found in the plasma. Local injections of
antigenic materials will result in local production of
antibodies, the type of antibody produced depending on the
site of injection. However, most antigens probably enter
the circulation and then are carried to all parts of the
body. In such cases, most of the antibodies seem to be
produced by specialized cells, the Immature plasma cells of
the reticuloendothelial system (spleen and lymphoid
tissue). The evidence for this may be summarized as
follows: the rate of antibody synthesis parallels the
number of plasma cells present, treatment which inhibits
the reticuloendothelial system (blockage with India ink,
x-lrradiation, exposure to mustard gas, or splenectomy)
also inhibits antibody production, and persons with
agammaglobulinemia are deficient both in antibodies and in
plasma cells.
Infants do not begin to produce antibodies until
several months after birth, but most of them have
antibodies which they have acquired from the mother's
circulation during pregnancy. Any of these acquired
antibodies which would react with the infant's erythrocytes
are, under normal conditions, immediately removed from the
circulation by specific substances in the tissues, as shown
by Zuelzer's transfusion experiments (34-). Extensive
hemolysis occurs only when the mother produces overwhelming
quantities of immune antibodies.
According to Jerne's "natural selection" theory of
antibody production (33), an Individual's blood contains
minute amounts of all antibodies the Individual is capable
of forming. Combination of antibody with an antigen
18
invading the system causes it to deposit in special cells,
where it induces formation of more antibody of the same
kind.
Haurowitz (36) proposed the existence of cytoplasmic
particles which become coated with antigen and act as dies
for the shaping of newly formed gamma globulin into
antibody molecules. It has been noted that once an
individual has been sensitized, he continues to produce
antibody for the remainder of his life (37).
According to Pauling's interpretation, a gamma
globulin molecule is folded, to fit over the specific group
of the antigen, and then held together in that form by
hydrogen bonds (38). This process would have to occur
twice on the same molecule for the formation of a bivalent
antibody. The Haurowitz and Pauling mechanisms do not
necessarily require that antigen molecules be present in
the cell during antibody formation.
Burnet and Fenner (39) suggested that adaptive enzymes
are formed in certain cells in response to the need for the
degradation of foreign molecules which have entered the
system, and that once the cell has been stimulated, it will
continue to produce these enzymes. The enzymes escape into
the circulation, retaining their specificity and losing
their enzyme activity, thus becoming antibodies. This
theory has since been revised and enlarged in the light of
19
new evidence. While Burnet no longer believes that
antibodies are adaptive enzymes, he continues to emphasize
the analogy between the two (40).
All these theories on the mechanism of antibody
formation are fundamentally similar. The conclusion is
inescapable that new, specific molecular structures are
formed in response to the stimulus of a foreign antigen,
and that some sort of permanent record permits continuous
formation of these structures. The details of the
different theories, while they are rather vague, tend to be
complementary rather than contradictory.
The mechanism of action of blood group antibodies is
not known with certainty. According to the unitarlan
hypothesis, which arose about 1900 and was stated
definitively by Zinsser in 1923 (41), all antibody
molecules of a given specificity are identical, their
reactions varying with environmental conditions. Few
workers would accept this hypothesis literally, but it does
appear to contain some truth, particularly as applied to
the varieties of incomplete antibodies. It appears that
both complete and incomplete antibodies are capable of
hemolyzing cells in vivo, probably because of the presence
of complement, which is an essential component In lytic
systems (13).
In 1944, Wiener (42) found a non-agglutinating
20
antibody, blocking antibody, which coated red cells and
prevented their reaction with agglutinating antibody.
Later he found that cells suspended in a colloid solution
would clump in the presence of non-agglutinating antibody
(28). This conglutination, as he called it, proved to be
more sensitive than the blocking test for the detection of
non-agglutinating antibodies. Wiener's use of the term
"conglutination" has been opposed by some workers on the
ground that in the reaction to which it was originally
applied by Bordet and Streng (13) complement was present,
while complement does not participate in Wiener's
conglutination reaction. Wiener originally supposed that
the colloid complexed with these antibodies to make them
capable of clumping cells, but such a'complex could not be
demonstrated electrophoretically (43).
A perhaps more sophisticated concept was advanced by
Ponder, who studied surface charges on red cells by means
of electromigration (44). He found that the cells would
clump spontaneously if the charge were reduced below a
critical value. Isotonic glucose solution or complete
antibody reduced the charge below this value; incomplete
antibody, colloids, and proteolytic enzymes also reduced
the charge, but not enough to cause clumping.
Another approach, exploited by Flick and Villafane
(45), is the analysis of clumps of cells formed under
21
different conditions. They studied clumps formed from
mixtures of bacterial cells and red cells. Complete
antibodies gave specific clumps containing only one kind of
cell; incomplete antibodies gave non-specific clumps
containing a mixture of cells.
These observations help to resolve the old conflict
between the Bordet and Marrack theories of agglutination.
Bordet in 1920 suggested a non-specific agglutination of
antigen once it has become sensitized with antibody (46),
while Marrack in 1934 proposed a specific interaction
between particles of sensitized antigen to form a lattice
similar to a crystal (47). It appears, then, that
agglutination follows Marrack's scheme, while
conglutination follows Bordet's.
Experiments on the analysis of precipitates formed
from different proportions of antigen and antibody point to
the conclusion that most or all antibodies are bivalent,
that is, each antibody molecule has two specific groups
and can combine with two antigen molecules (1>). Though
this work was done with precipltins, blood group antibodies
are probably similarly constituted. In the older
literature, complete and incomplete antibodies were often
referred to as bivalent and univalent, respectively. The
terms bivalent and univalent no longer seem appropriate,
and their use should probably be discouraged.
With regard to affinity for cells, Sturgeon observed
that cells selectively remove incomplete antibody from a
serum, leaving most of the complete antibody (4-8). With a
given antibody, some cells will react better than others;
for instance, E/E cells react better than E/e cells with
anti-E serum. Also, presence of D antigen seems to Inhibit
reaction of E antigen with anti-E serum. This dosage
effect is attributed to epistasis, or competition, of the
C, D, and E genes for the limited amount of raw material
from which antigens are made (^9). In this connection,
Belkin and Wiener (50) suggested that there may be many
more sites on the red cell to react with ABO antibodies
than with Rh antibodies, thus explaining why blocking is
observed more frequently with Rh than with ABO antibodies.
It should also be observed that Rh antibodies, unlike
most other blood group antibodies, react better at 37° than
at lower temperatures. The explanation for this difference
may lie in the relative weakness of Rh reactions, or
perhaps in the need for a particular molecular conformation
which can be attained only at body temperature.
Several workers (30,48,51,52) are in agreement that
heating at about 60° will destroy complete antibody
activity, with relatively little effect on incomplete
antibody activity. Pickles (53) found that a serum
developed a prozone after storage for several months in
the refrigerator; this is indicative of a relative decline
in complete antibody activity. Boyd (54) destroyed
antibodies by subjecting them to pressures of several
thousand atmospheres or by photo-oxidation; he found
incomplete antibodies more resistant to this treatment than
complete. Although the serum proteins were found to
polymerize at high pressures, no appearance of complete
antibodies was observed. Boyd concluded that conversion of
incomplete to complete antibodies through polymerization
would be still less likely at atmospheric pressure. Coombs
and coworkers (30) suggested that complete antibodies might
be converted to incomplete by heating, but such conversion
has not been reported by other workers. On the whole,
there is little evidence to support the idea that one kind
of antibody may be converted to another.
Separation of Blood Group Antibodies
Relatively little attention has been given to the
problem of separating complete and incomplete antibodies.
Both antl-A and antl-B have been said to migrate in
the gamma-beta region in free boundary electrophoresis.
Cann and coworkers used a process called convection
electrophoresis to achieve partial separation of Rh
antibodies (55,56). Their results indicated that blocking
antibody and albumin-reacting antibody was in a gamma
globulin fraction of low mobility and that the
24
antiglobulin-reacting antibody was distributed broadly
through the globulins, with maximal activities in different
parts of the globulins in different sera.
Wiener wrote that complete antibodies are alpha and
beta globulins and incomplete antibodies are gamma
globulins (57). Tullis stated that Rh antibodies are found
predominantly in Cohn fraction III, which is mostly beta
globulin, and that they may be partly alpha globulins (59).
Laurell found that starch gel cannot be used as a
supporting medium for electrophoresis of antibodies,
because the gel is impermeable to the large molecules (58).
Witebsky and coworkers attempted to fractionate Rh
antibodies by dialysis of serum against distilled water
(60,61). They observed that nearly all agglutinins
precipitated out upon dialysis, while incomplete antibody
remained in solution. In agreement with these findings,
Wiener reported that complete antibody is a euglobulin
precipitated by sodium sulfate solutions of 15.5 to 17.4
per cent concentration, and that incomplete antibody is a
pseudoglobulin precipitated by 17.4 to 21.5 per cent sodium
sulfate (57).
In the same publication, Wiener gave the sedimentation -
constants of complete and Incomplete antibodies as 17 S and
7 S, and the molecular weights as 930,000 and 155,000,
respectively. These are probably in the nature of
25
estimates rather than experimentally determined values.
The intact placenta is impermeable to complete and
permeable to incomplete antibodies, probably because of the
difference in molecular size, but this has been questioned.
In more recent experiments Campbell and coworkers
centrifuged serum at 105,000 times gravity, thereby causing
complete Rh antibodies to sediment more rapidly than
incomplete antibodies (62).
It has long been known that antigen-antibody reactions
can be reversed, with release of free antibody Into
solution (63). However, only minute quantities of blood
group antibodies can be recovered In this way from large
volumes of red cells, and the cells are too fragile for
convenient manipulation.
In experiments on the purification of antibodies,
Isliker (64) took advantage of immunological specificity by
linking red cell stromata to an ion exchange resin,
reacting the stroma-resin with serum containing antibody,
and then eluting the antibody from the stroma-resin in a
relatively pure state. Isliker claimed that anti-B could
be concentrated as much as 350-fold in this way. His
results were erratic, however, and the procedure was not
regarded as the method of choice for purification of Rh
antibodies (65).
Because of the broad field covered in this
introduction, and because of the confusion which still
surrounds many fundamental problems of antibody function,
it would be difficult to make any generalizations which
would do justice to the subject. It is not out of place,
however, to observe that the study of blood group
antibodies as chemical compounds has only just begun.
Further experimental evidence is needed to confirm or
refute the various theories which have been advanced and
perhaps to establish the basis for another approach to
fundamental problems. The present investigation represents
one attempt to add to the store of basic chemical
information on antibodies.
STATEMENT OF THE PROBLEM AND PLAN OF ATTACK
In addition to the scientific interest in studying the
nature of antibodies, there is strong clinical motivation
for proceeding with such investigations. Perhaps the most
important effect of blood group antibodies is the damage
they cause when Introduced into the blood stream of certain
Individuals, particularly in cases of erythroblastosis
fetalis, in which the fetus is exposed for long periods to
antibodies produced by the mother. Treatment of this
condition by transfusion is not only inherently dangerous,
but it is frequently ineffectual in advanced cases.
Consequently, there has been an interest in the possibility
of prenatal preventive treatment in incompatible
pregnancies. Any such treatment would have to be founded
on a more extensive knowledge of the nature and behavior of
antibodies than exists at present. In view of the need for
basic Information, it would be premature to advance
hypotheses of treatment. One can only assist in the
process of gathering the needed information.
Statement of the Problem
The general purpose of the present investigation has
been to explore the possibility of fractionating and
purifying the blood group antibodies so that their
28
properties can be determined. Even partially purified
fractions could yield valuable information. The primary
goal in this work has been the separation of complete and
incomplete antibodies. A practical application of the
results to be gained would be in the commercial preparation
of useful blood typing sera by the removal of blocking
antibodies from sera otherwise of little use.
In these studies attention has been focused on the Rh
antibodies for several reasons. These are the antibodies
which cause more than 90 per cent of the cases of
erythroblastosis, and they are therefore of great clinical
significance. From the theoretical standpoint, these
antibodies are interesting because, unlike most other blood
group antibodies, they exist in both the complete and
Incomplete forms. The comparative properties of these
forms, if known, could shed considerable light on the
general problem of the nature of the antigen-antibody
reaction. The third reason, a practical one, is that,
aside from the ABO antibodies, Rh antibodies are by far the
most easily obtainable blood group antibodies.
Studies on both natural and immune ABO antibodies have
been considered concurrently with the work on Rh, but these
studies have been incidental to the main problem. The ABO
antibodies seldom give rise to clinical problems, they are
most often of the complete type only, and relatively more
29
ia known about them.
Plan of Attack
Antibodies have not usually been amenable to
manipulation by the conventional physical and chemical
methods because of their molecular size and complexity and
their instability. Only within the past few years have
suitable methods been developed, and even these are
generally in want of refinement. Another problem in
immunochemical studies is the unavailability of reliable
sources of uniform materials (sera and cells) with which to
work.
In the present work several different sera, containing
antibodies of different specificities' and of both the
complete and Incomplete types, have been studied. By
comparing results from different sera it should be possible
to find which properties are characteristic of Rh
antibodies in general, and which properties are peculiar to
individual sera.
Several different fractionation methods have been used
in attempting to obtain preparations of high activity in
good yield. Included were several forms of electrophoresis,
ultracentrifugation, chromatography, salt and solvent
methods, and adsorption on and elution from specific red
cell stromata. Within experimental limitations, conditions
in each method were adjusted to insure the best possible
30
separations.
Fractions obtained by these methods were analyzed by
physical, chemical, and serological methods, and in some
cases the fractions were subjected to further fractionation
by additional methods.
Sufficient data were accumulated to insure validity of
the findings. Interpretations based on experimental
results were supplemented by the results of other workers,
as reported in the literature.
MATERIALS
Two kinds of biological materials were used: selected
high titer, immune sera from human donors, and human
erythrocytes for the detection of antibodies in these sera.
Six different sera were provided by Dr. Phillip
Sturgeon. These sera; which will be designated by the
abbreviations And, Bea, Pic, Wat, Cas, and Ste; contained
antibodies belonging to both the Rh and ABO blood group
systems, and one serum contained a Lewis antibody (anti-Le^).
The donors of sera And, Bea, and Wat had been immunized by
pregnancies several years earlier, the donor of Pic by
blood transfusion about 1950, and the donors of Gas and Ste
by transfusions also. The sera were prepared by clotting,
with topical thrombin, blood which had been mixed with an
anticoagulant solution. Sera Bea, Pic, and Cas were
freshly prepared; and And, Wat, and Ste had been stored at
-20° for more than one year, without apparent change in
antibody titers. Characteristics of these sera are given
in Table I.
Erythrocytes used in titrating anti-D, anti-A, and
antl-B were taken from Red Gross blood, which contained
A.C.D. (acid-citrate-dextrose) anticoagulant (66). For
other titrations, rarer types of cells were required, and
these were provided by laboratory workers or patients of
31
Dr. Sturgeon. In such cases, cells were collected and
stored in Alsever's solution (13). All cells were stored
in the refrigerator and were discarded after they were
three weeks old.
The standard antiglobulin serum used in all titrations
of incomplete antibodies, except where otherwise noted, was
supplied by Dr. Sturgeon. This serum had been prepared by
the immunization of rabbits against whole human serum. It
was diluted to a titer of 32.
33
TAELE I
ANTIBODIES IN SERA STUDIED
Donor Antibodies
ABO
Symbol Type
And 0
Rh
, f ype
Gde/cde
Titer*
Antibody Complete Incomplete
Anti-A 64
Anti-B 4
Antl-D 0 512
Anti-E 4 128
Bea A CDe/CDe Anti-B
Anti-E
2
256 256
Die
Wat
Cas
Ste
A
0 cde/cde
CDe/CDe
Anti-B
Antl-D
Anti-c
Anti-Lea
256
256
0
0
51 2
4
32
* The titer is defined as the greatest dilution at which
activity can be detected. Complete titer by saline
agglutination; Incomplete by antiglobulin reaction,
except for Cas serum, which reacted only in albumin.
METHODS
The methods used may he conveniently grouped under
five headings: serological methods, fractionation methods,
concentration methods, protein methods, and analytical
methods. Detailed descriptions of certain of the methods
are given in the Appendix.
Serological Methods
The procedures used for titrating the antibodies
present in various fractions are adapted from the methods
described by Coombs and coworkers (30). In essence they
consist in determining the greatest dilutions, in series of
two-fold dilutions, at which antibody activity is
sufficiently great to cause observable clumping of red
blood cells under the conditions of the titrations. For
the purposes of the present work, complete antibodies will
be considered as those which are measured in the saline
agglutination titration and incomplete antibodies as those
measured in the antiglobulin and albumin titrations. In
the saline titration, the serum or antibody solution is
mixed with an equal volume of a two per cent suspension of
erythrocytes in isotonic buffered saline solution. The
albumin titration is identical in all respects, except that
a 20 per cent solution of bovine serum albumin is used
instead of saline solution in making the dilutions of serum
34
35
and the suspension of cells. In the antiglobulin
titration, the cells from the saline titration are washed
three times and then mixed with antiglobulin serum before
inspection for clumping.
The use of a clean pipette for each dilution is
important. When the same pipette is used for all
dilutions, a minute but variable amount of antibody is
carried over from tube to tube. The result is an unknown
increase in the observed titer.
The amount of rocking or shaking of the tube during
reading can affect the appearance of any clumps of cells.
Some agitation is necessary to dislodge the packed cells
from the bottom of the tube, but excessive shaking can
reduce the apparent strength of agglutination, especially
In the case of larger clumps. It seems that the tube
should be rocked just enough to spread some of the cells
along the side. Gentle rocking should continue during
observation of the clumps, as it is easier to distinguish
between the different degrees of agglutination while the
cells or clumps are In motion.
The manner of recording the degree of agglutination is
to some extent a subjective matter, and it is necessary
only to obtain titers which are comparable from one
experiment to another. The titers of the fractions which
were determined in the present work are not considered to
"be absolute values. In fact, results show that the titer
found for any given fraction may differ by two-fold from
one time to the next. One reason for this variability is
the difficulty of deciding whether to rate a weak reaction
as + or as ±. As a rule, all of the fractions from a
single fractionation experiment were titrated at one time,
and the reactions compared with one another as well as with
a serum control, so that within any one set of titrations
the + reactions are all alike and the ± reactions are all
alike.
Fractionation Methods
Five principal fractionation methods were
investigated: electrophoresis in starch blocks,
ultracentrifugation, anion-exchange chromatography on
DEAE-cellulose, salt and solvent fractionation, and
specific adsorption on stroma-resin preparations. Of these,
only three methods were considered sufficiently useful for
the separation of antibodies in serum: these were starch
block electrophoresis, ultracentrifugation, and
anion-exchange chromatography.
Starch block electrophoresis. The method of starch
block electrophoresis used in these experiments was similar
to that described by Kunkel and Slater (67). The starch
block measured 50 by 20 by 1 centimeter. Use was made of
an efficient refrigerator (or a cold room in the later
37
stages of the work) to provide the needed cooling. The
flatness of the block permitted efficient dissipation of
the heat produced by the passage of current. Such a block
will accommodate as much as ten milliliters of serum and
will resolve the serum into at least five distinguishable
fractions. The built up mold used in these experiments is
preferable to the one piece cemented lucite molds which are
sometimes used because of the ease with which parts of the
mold can be stripped off the starch block. Manipulation of
the block itself, which may lead to running, breaking, or
crumbling, is thus reduced to a minimum.
If the block is not adequately dried before insertion
of the sample, the space in the block will fill with
buffer. There is no danger, within reasonable limits, of
drying the block too much. Usually the current averages
about 90 milllamperes, depending on the amount of buffer in
the block, for a potential of 350 volts. For routine
operation, a voltagp x hours product of 7000 was most
satisfactory. If desired, the run can be stopped when the
albumin band, visible because of its yellow color, reaches
a certain point. A second colored band, more pinkish than
the albumin band and visible only when the block is viewed
by transmitted light, migrates with the beta globulin.
After cutting of the block, the proteins can be easily
eluted from the starch by suction filtration on a sintered
38
glass filter.
Ultracentrifugation. The rate of sedimentation of a
particular protein under a large centrifugal force depends
on the size and shape of the protein molecules. Since this
method is not directly dependent on the charge on the
protein molecules, it might be expected to supplement
electrophoresis as a fractionation method. The usefulness
of ultracentrifugation, however, is limited by the fact
that all sedimenting molecules collect in a mixture at the
bottom of the centrifuge tube, so that only the uppermost
fraction can be obtained in pure form. It is also rather
difficult to remove well defined fractions from the tube
without some mixing of the contents. A device designed by
Dr. John Mehl (68) permitted removal of fractions from the
top of the tube while the contents were gradually displaced
upward by addition of a dense salt solution or carbon
tetrachloride at the bottom of the tube. The centrifuge -
used in the present work was a Spinco Model L preparative
ultracentrifuge with No. 40 rotor. At 40,000 rpm the
average centrifugal force was about 105,000 times gravity.
Anion-exchange cellulose chromatography. The
technique of chromatography of proteins on columns of
diethylaminoethyl-cellulose ("DEAE-cellulose") was first
introduced by Peterson and coworkers (69»70) and later
described in more detail by Fahey and coworkers (71).
39
A protein mixture is adsorbed on a column of
DEAE-cellulose, and various fractions are progressively-
eluted by means of a pH and ionic strength gradient.
Because of the newness of the method, ideal experimental
conditions for the separation of serum proteins have yet to
be defined. The rate of appearance of the various protein
components of serum from the column can be altered to some
extent by changing the elution gradient. A pH gradient
with constant ionic strength, or an ionic strength gradient
with constant pH, can also be used.
In later experiments in the present work, a batch
operation replaced the column method. The conditions of
the batch method were based on results obtained in the
column method. In these experiments four fractions were
eluted successively from the DEAE-cellulose by buffers of
increasing ionic strength and increasing acidity.
Solvent fractionation. This method is based on the
early methods of protein fractionation, which took
advantage of differences in solubilities of different
proteins at various ionic strengths. Cohn and coworkers
developed a method by which protein fractions could be
obtained from plasma at much lower ionic strengths than had
been used previously (66). Cohn's system made use of an
alcohol-water mixture of reduced dielectric constant, and
also of specific interactions between proteins and other
40
ions. The method used in the present work has been adapted
from Cohn's well known "Method 10."
Stroma-resin adsorption. This method takes advantage
of specific Interaction between antibodies and their
corresponding antigens, the blood group substances in red
cell stromata (64). Stroma from hemolyzed red cells is
bound to an ion-exchange resin to improve its physical
properties. Free stromata cannot be used because of their
tendency to form colloidal solutions, especially'in the
presence of salts. Serum is added to the stroma-resin
preparation, which selectively adsorbs the desired
antibodies from the serum. The adsorbed antibodies can
then be released into solution with heating and treatment
9
with certain sugars, which reverse the antigen-antibody
reaction.
In spite of the specificity of the antigen-antibody
reaction, the isolated antibody is always contaminated with
other serum proteins which are non-speclfically adsorbed on
the resin. Also, the recovery may be quite low because of
incomplete reaction between antigen and antibody and
because of losses in washing of the stroma-resin prior to
elution. Apparently, much depends on the nature of the
serum and experimental conditions.
Concentration Methods
Four principal methods of concentrating protein
solutions were tried: pervaporation, dialysis against gum
acacia, lyophilization, and salting out. All methods gave
about the same results, but pervaporation proved to be the
most convenient and economical.
Pervaporation. The principle of pervaporation is the
evaporation of water from one side of a semi-permeable
membrane while protein molecules are retained on the other
side (72). Heat removed by the process of evaporation
keeps the protein solution cooler than the surroundings.
When pervaporation is conducted at room temperature, the
solution remains at about 10°.
When pervaporation was used to concentrate the eluates
from starch blocks, no preliminary dialysis was necessary,
as dialysis and pervaporation could be carried out
simultaneously. This was accomplished by having the lower
end of the dialysis bag Immersed in a tray of saline
solution. Simultaneous dialysis also served as a safety
feature by preventing complete drying of the fractions
during pervaporation. Drying destroys most of the antibody
activity. Slightly lower recoveries of antibody were
obtained from pervaporation at room temperature than from
pervaporation in the cold, but pervaporation in the cold
has the disadvantage of taking several times longer, often
as long as a week, to achieve the same degree of
concentration.
42
Concentration by dialysis. Concentration by dialysis
makes use of the difference In osmotic pressure across a
membrane separating dilute and concentrated solutions of
non-diffusing molecules. Water diffuses toward the side of
the membrane on which water molecules are relatively more
scarce. Because of Its similarity to pervaporation,
dialysis should be expected to give similar results.
A 25 per cent solution of gum acacia provided the
non-diffusing molecules in the present work (73). When the
gum acacia solution was being prepared, it was kept
constantly stirred during heating to prevent charring, and
the resulting large volume of foam was scraped off the
surface after cooling. Since the gum acacia solution was
used repeatedly, merthiolate was added to retard bacterial
growth. According to the manufacturer (Ell Lilly & Co.),
merthiolate has no harmful effect on serum proteins, and
experience in the present work has borne this out. During
the dialysis a marble tied in the bottom of the dialysis
bag kept it from floating to the surface of the gum acacia
solution. The marble did not interfere with subsequent
manipulations, as it was separated by a knot from the
protein solution.
The rate of concentration for a completely immersed
bag was about the same as for pervaporation in the
refrigerator, the volume being approximately halved in 24
43
hours. Because of the high concentration of gum acacia, it
removed virtually all of the water from the proteins in the
bags if the process was continued too long; this resulted
in loss of antibody activity. A stirring mechanism kept
the bags in constant motion through the dlalysate during
dialysis (74).
Lyophlllzatlon. The process of lyophilization, or
freeze-drying, involves evaporation of water from the solid
state, which, once achieved, is maintained by removal of
the heats of fusion and of vaporization during evaporation.
The rate of evaporation, which would otherwise be very low
at freezing temperature, is accelerated by reducing the
pressure within the system with a vacuum pump (75).
Lyophilization has the advantage over other methods of
protein concentration of reducing thermal denaturation to
a minimum. Another advantage--that proteins can be
completely dried— was of less importance in the present
work. The Dewar flask used in this work had a capacity of
some ten liters, and 25 pounds of dry ice sufficed for one
or two days' operation. The apparatus used was similar to
that of Flosdorf and Kudd (76), with lyophile flasks of
about 800 ml. capacity. Protein solutions shell-frozen in
these flasks were dried at a rate of about 15 ml. per hour.
Aside from the slowness of drying, the volume of solution
which could be dried at one time in the four lyophile
44
flasks was limited by the size of the water trap, which
also had a capacity of 800 milliliters.
If several samples were to be dried simultaneously,
they were all shell-frozen just before placement on the
manifold, and the flasks were not allowed to touch any warm
surface after that. Thawing at this time or during the
drying process reduces the surface available for
evaporation and Increases the risk of denaturation. If
thawing does occur, refreezing can be repeated at any time
during the drying.
Salting out. The classical method of salting out of
proteins was of special value in the present work because a
rough separation of albumin and globulins could be obtained
(77). The method used in the present work was designed to
remove globulins from the very dilute eluates from starch
blocks. Recovery of active material was slightly lower
than for other concentration methods.
Protein Determination
Yftiile several different methods of protein
determination were studied, the method which offered the
best combination of sensitivity and convenience was the
Lowry method. This was standardized against the other
methods of protein determination.
Lowry method. The method of protein determination
used for all practical work was a modification of the
method published by Lowry and coworkers (78,79). It is
much more sensitive than the biuret method, so that only
very small samples are required for protein determination.
The Folin-Clocalteu reagent (80), which was obtained from a
commercial source, had to be diluted to about 1 N in acid,
or about two-fold, to make the color developing reagent. A
Beckman Model C colorimeter was used. The readings were
generally made on the day following the addition of color
reagent, because it was found that the color intensity
increases slowly for some time after the initial color
development but stabilizes after several hours. The
increase in Intensity is proportional to the protein
concentration, however, so that any set of readings made at
the same time are comparable. Moreover, there is not more
than a 10 per cent increase in optical density after the
first 30 minutes.
In an experiment with various dilutions of serum, the
optical density of the developed color was found to be
proportional to protein concentration in the working range.
The optical density bears approximately the same relation
to protein concentration, as determined by the
micro-Kjeldahl method, for different fractions of serum
proteins.
In the section Experimental Results, protein
concentrations are expressed as milligrams of bovine serum
46
albumin per milliliter of solution. This is done in order
to provide a common denominator for comparison of results
on different fractions. Calibration work shows that the
optical density is about twenty per cent greater, for the
same amount of nitrogen, for bovine serum albumin than for
human serum proteins. The protein concentrations given in
the tables are consistent in themselves even though they
have all been converted to bovine serum albumin equivalents.
Three other methods of protein determination were used
in calibrating the Lowry method. These were the
micro-Kjeldahl, Nessler, and biuret methods.
Kjeldahl method. The KJeldahl method is a widely
recognized method for nitrogen determination. The
micro-KJeldahl procedure used in these experiments was an
adaptation of the procedure of Scales and Harrison (81).
The proteins in a sample were digested with sulfuric acid,
the acid was neutralized, and the nitrogen was distilled as
ammonia into a known volume of boric acid, which was then
back-titrated with standard acid. The nitrogen values
determined by this method were a close approximation to
absolute values, since they were obtained by comparison
with an ammonium sulfate standard. However, the factor
6.25, which is generally used to convert nitrogen to
protein values, has never been completely experimentally
verified in the case of serum proteins.
In the digestion step, failure to add a boiling chip
results in dangerous spattering of the sulfuric acid and it
causes loss of sample. A catalyst, such as copper sulfate
or selenium is often added to the digestion mixture, but
control experiments showed that neither of these materials
had any effect on completeness of digestion in the present
work. However, the amount of protein found was about three
per cent greater after two hours of digestion than after
one hour. It was considered that any increase with longer
digestion times would be insignificant, because of the
variability of the method. It was found necessary to use
30 per cent hydrogen peroxide in the digestion; without it,
the digest did not decolorize.
The composition and concentration of the indicator dye
in the boric acid trapping solution was not critical, and
no attempt was made to reproduce It exactly. All that was
necessary was that the dye be a blue-green color in the
end-point standard and that it be dark enough to be easily
seen. The colors of the sample and the end-point standard
were compared by placing the flasks on white paper in a
well illuminated spot and looking down through the necks of
the flasks. The most convenient concentration of
hydrochloric acid used for titrations depends on the size
of the burette which is used. The 0.0025 N acid was used
in a five milliliter burette. In earlier experiments,
48
0.001 N acid was used in a 50 milliliter burette. The more
concentrated acid is to be preferred, because the color of
the indicator dye depends to some extent on the volume of
the solution. When a more dilute acid is used, the volume
of the titrated samples varies over a wider range, so that
a large number of end-point standards of different volumes
are required.
Two methods were used for comparing the unknown
samples with the standards. In general, the standards were
found to contain more than the theoretical amount of
nitrogen, as determined from the volume of standard acid
required to neutralize the ammonia. The difference was
assumed to be due to traces of nitrogen in the reagents.
The average amount of reagent nitrogen per sample was then
subtracted from the observed nitrogen values of the
unknowns, calculated from the volume of 0.0025 N acid used
in the titrations. The reagent nitrogen varied slightly
from one time to another, but it was usually less than five
micrograms per sample. The other method of determining the
nitrogen contents of unknown samples was to read the values
from a line connecting the points representing standards on
a graph with milliliters of 0.0025 N acid as the ordinate
and milligrams of nitrogen as the abscissa. This second
method has one advantage over the first. Since the
apparent color of a sample at the end of the titration is a
function of the volume of the sample (that is, it Is a
function of the volume of acid added to the sample during
the titration), a small error is introduced when this
volume is not approximately the same as that of the
end-point standard. In the graphical method, these errors
due to volume differences are the same for both the
standard and unknown samples, so that the errors have no
effect on the final result. That these errors do indeed
occur is borne out by the fact that the empirical line on
the graph is never quite parallel to the theoretical line.
On the other hand, it was found that the final results
obtained by the two different methods seldom disagree by a
significant amount.
Nessler method. The hessler method of nitrogen
determination (82) is based on the measurement of the
colored complex produced when an alkaline solution
containing mercuric ions is added to protein digested in
the same manner as for the Kjeldahl determination. The
acid in the digestion mixture must be exactly neutralized
by the mercuric solution; otherwise, an orange precipitate
forms and no color develops in the solution. While this
method has about the same sensitivity as the Lowry method,
it was considered too long for routine use, and the
necessary digestion adds another possible source of error.
Biuret method. The biuret method (82) is generally
50
considered an excellent method for protein determination,
because the intensity of the violet color which results
when an alkaline copper sulfate solution is added to
protein is proportional to the number of peptide bonds
present, the peptide linkage being a fundamental
characteristic of all protein molecules. The biuret method
suffers from the disadvantage of being about a hundred
times less sensitive than the methods discussed above.
Analytical Methods
Although the serological and protein determination
methods are, in a sense, analytical methods, the discussion
in this section will be limited to three physical methods:
free boundary electrophoresis, paper electrophoresis, and
starch gel electrophoresis. In the present work, starch
gel electrophoresis has been of interest primarily as a
preparative fractionation method.
Free boundary electrophoresis. The principles of free
boundary electrophoresis (83,84) are so well known as not
to require detailed description here. The Perkin-Elmer
apparatus, Model 38-A, equipped with Polaroid camera, was
used, and the Instructions supplied with the Instrument
were followed. The photographs of separated proteins were
used both for qualitative analysis and for the measurement
of mobilities.
Paper electrophoresis. For paper electrophoresis, the
51
conventional Spinco hanging strip system was used (85)*
The principles, conveniences, and shortcomings of this
method are now so well known that little comment need be
made here. While paper electrophoresis is convenient and
generally satisfactory for qualitative analyses of protein
mixtures, it suffers from several disadvantages. In the
hanging strip method, the migration is not proportional to
mobility, proteins become adsorbed on the paper, and
quantitation is difficult and unreliable. Other methods of
paper electrophoresis, such as the closed glass plate
method, can partially overcome such disadvantages, but they
lack the convenience of the Spinco method.
Starch gel electrophoresis. Starch gel
electrophoresis (86,87) is a method similar to starch block
electrophoresis, the main differences being that an elastic
starch gel and a very dilute borate buffer are used in the
former, so that relatively more of the total current
passing through the supporting medium is carried by the
proteins themselves. In the present work, the measured pH
values of the buffer and bridge solution were 9.0 and 8.2,
respectively. The hardness of the gel depends on the
nature of the starch and on the duration of the hydrolysis
step, as well as the starch concentration. Several batches
of starch from the United States and Denmark were tried,
and all gave satisfactory gels with appropriate hydrolysis
52
times. Other factors being equal, lengthening of the
hydrolysis time gives a softer gel.
The protein patterns obtained generally resemble those
in other types of electrophoresis, except that additional
minor components appear, often ahead of the albumin band.
The patterns are not easily reproducible, although Smithies
and coworkers have been able to distinguish different types
of sera, and they have devoted much attention to the
characterization of minor components (87). There is some
indication that the structure of the starch gel is
impermeable to certain large molecules, including
antibodies (58).
Another related technique, electrophoresis in 1 or 2
per cent agar gel, yielded somewhat similar results; but
the gel was harder, less elastic, and almost transparent,
while the protein bands appeared more diffuse and less
uniform.
EXPERIMENTAL RESULTS
Because of the complex nature of antibodies and the
relatively wide array of methods which have been used,
problems are encountered in the presentation and
interpretation of experimental results. First, the
conventional language of immunology lsJinadequate for a
full description of the results. Second, while there is
considerable variability in results, titration data cannot
be subjected directly to statistical analysis. Also, the
laboriousness of the methods and scarcity of materials
preclude extensive replication. Third, the methods of
fractionation and analysis have evolved somewhat in the
course of the work, so that often experimental details have
to be stated when a general description of a method would
otherwise suffice. And fourth, uncertainties in
Interpretation arise from the fact that all fractions
obtained have been mixtures of proteins rather than pure
compounds.
The presentation of results will be divided into
several sections, each treating a particular fractionation
method with respect to protein and antibody distributions,
enrichment and recovery of antibodies, and other points of
interest. First of all, however, it is important to
describe the terms in which the results will be expressed.
53
54
Expression of Results
In the expression of results there Is need for units
of protein concentration, of antibody concentration, of the
ratio between the two, and of total amounts of protein and
antibody in a given fraction. Protein units may be defined
with little difficulty, but certain reasonable assumptions
must be made in setting up antibody units. While these
assumptions may be open to question, the theoretical
uncertainties about the units need not detract from their
practical value.
Protein units. The Lowry method has been used almost
exclusively for protein determinations on serum fractions,
because of its sensitivity, rapidity, and reproducibility.
For convenience, the optical density measurements obtained
in this method have been converted to milligrams of protein
per milliliter. The factor used for this conversion is
based on a comparison of the Lowry method with other
methods of protein determination, particularly with
nitrogen methods. Bovine serum albumin, a readily
available protein of reasonable purity, has been used in
this comparative work, and for this reason, all of the
protein concentrations given in this section must be
understood to be "BSA equivalents," Comparisons have been
made between bovine serum albumin and total human serum
proteins by the mlcro-KJeldahl, Nessler, biuret, and Lowry
methods. Results of the biuret and Lowry methods parallel
each other closely-,- but there is a disparity between the
nitrogen methods and the Lowry method. On BSA and serum
protein solutions of the same nitrogen content, the optical
density In the Lowry method is greater for BSA by a factor
of 1.22. Among the various electrophoretic fractions (from
starch block electrophoresis) of human serum proteins, the
optical density/nitrogen ratios have been found to be about
the same (within 5 per cent). This result would be
expected, because such fractions contain mixtures of
proteins, and variations in the optical density/nitrogen
ratio, among the various proteins present in a given
fraction, would tend to nullify one another. The
similarity of this ratio for different electrophoretic
fractions lends further support to the validity of the
Lowry method for protein determinations on fractions in the
present work.
Antibody units. The only available method for the
detection of blood group antibodies is the observation of
the clumping of erythrocytes. The extent of the clumping
seems related to the amount of antibody present, as well as
its avidity, but there is no assurance of proportionality.
The serial dilution method of titration gives a better
semblance of quantitation. The titer, which is the
greatest dilution at which antibody activity can be
detected by a given method, is clearly proportional to the
amount of antibody in the serum, provided the assumption is
made that the reactivity of the antibody is not changed in
the process of dilution. Experience has shown that this is
a generally good assumption, though it is contradieted\±n
some cases, for example by the prozone phenomenon and by
the fact that total protein concentration has a marked
effect on the reactivity of albumin antibodies.
Unfortunately, we have no clue to the absolute amount
of antibody present, except through titration as an
expression of activity. The activity may or may not be
proportional to the antibody concentration, and there is no
direct way of determining whether a serum contains factors
which will enhance or Inhibit antibody activity. However,
in the present case the assumption of proportionality seems
the most reasonable basis for the expression of antibody
concentration.
The titer of a given serum or protein fraction will
henceforth be accepted as the best measure of antibody
concentration, with the understanding that it is and
empirical and somewhat arbitrary quantity. In order to
emphasize the assumed proportionality to concentration, and
to facilitate calculations, the titer will be expressed as
an Integer, rather than as a reciprocal, as. is often done
in the literature. The total amount of antibody in a given
57
fraction will be defined as the product of the titer and
the volume of the fraction in milliliters, designated
herein as "antibody units."
The need for some expression of the variation and
reproducibility of titration results is another source of
difficulty. The tube chosen as the end point of an
antibody titration may vary according to minor variations
in technique, and according to the judgment of the person
reading the titration; there Is no definitive end point.
But if the same person titrates the same serum with the
same cells on different days, the individual end points
will rarely differ by more than one tube from the mean end
point. Statistical analysis of a series of such titrations
has shown that the 95 per cent confidence limits would
equal approximately plus or minus one tube. Because of the
logarithmic relation (log2) between the titer and the
number of the end point tube, it is impracticable to give
confidence limits for individual titers. The tube numbers
themselves cannot be used in expressing results because
they are not proportional to the antibody concentration.
Expression of a titer of 16, for example, would be awkward
and confusing if it were stated as antllog2 4±1 or as
8 ^ x 02.
Antibody enrichment. As a measure of antibody
enrichment with respect to protein, the "specific antibody
58
activity" will be defined as the ratio of the titer of a
given fraction to the protein concentration, in milligrams
per milliliter (BSA equivalent). The specific antibody
activity is analogous to the specific activities of enzyme
or vitamin preparations or radioactive materials.
Enrichment of complete with respect to incomplete
antibody will also be of interest, and this will be
expressed simply as the complete/incomplete ratio, the
ratio of the respective titers.
Significant figures. In conformity with conventional
practice, all titers will be presented as the observed
dilutions, without rounding. Other data, including protein
measurements, specific antibody activities, antibody units,
and recoveries, will be rounded to two significant figures.
It should be emphasized, however, that in view of the
variability of titration results, the second figure must be
regarded with caution, even in the averaged data of the
summary tables.
Negative results. A negative result on an antibody
titration, expressed as a titer of zero and zero antibody
units in the tables, does not constitute proof that no
antibody is present. It Is rather an indication that no
antibody was detected with the method used. It is entirely
possible that methods of greater sensitivity could detect
antibody in such cases.
59
Electrophoresis
Starch block electrophoresis was the most extensively
used fractionation method In the present investigation.
Fifty-five blocks were run, most of them for preparative
purposes. Ten milliliters of serum was generally
fractionated on a single block. Because of the large
accumulation of data from starch blocks run with different
sera and under different conditions, a consideration of
starch block electrophoresis provides an excellent
introduction to the presentation of experimental results.
Protein distribution. In the methods used for
fractionating antibodies, the general nature of the
proteins in each fraction is of interest. Determination of
protein distribution is not only a convenient check on the
efficiency of a method and a help in arriving at the
optimal conditions for separation, but it also is a helpful
indication of the mobility class of proteins to which the
antibodies belong. The presence of antibody in a given
fraction cannot, of course, be proof that it is nearly
identical with the other proteins in that fraction.
Longitudinal strips, one centimeter wide, were cut
from several of the starch blocks, and the strips were
divided into fifty segments, each one cubic centimeter in
size. Direct protein determinations on these segments
permitted curves to be made of protein concentration versus
60
distance of migration. Such a curve bears a close
resemblance to the familiar protein curves obtained in -
paper electrophoresis and also to the refractive index
*
gradient curves obtained in free boundary electrophoresis.
This resemblance suggests that corresponding peaks in these
three methods represent much the same classes of proteins.
The earlier starch blocks (no. 1-19) were cut into
four segments, corresponding to the four principal protein
peaks. Later blocks were cut into ten standard segments,
each five centimeters long; this method of cutting was
adopted in:order that the segments might still coincide
with the principal protein peaks. Cutting the block into
ten segments increased the effective resolution in the
components of greatest interest, gamma globulin and
albumin. The serum proteins contained in the ten standard
fractions eluted from these sections were tentatively
identified as indicated in Table II.
Further evidence of the identity of these fractions
was provided by the subjection of the fractions to paper
electrophoresis and free boundary electrophoresis. Results
with paper electrophoresis were disappointing because of
adsorption of proteins on the paper, but the proteins in
the different fractions did appear to have the same
relative mobilities as in starch block electrophoresis.
In free boundary electrophoresis (barbital buffer, 0.1 M,
61
CORRESPONDENCE
WITH
TABLE II
OF STANDARD STARCH BLOCK
SERUM PROTEIN FRACTIONS
FRACTIONS
Block Protein Per Cent
Fraction Fraction of Total
1 Gamma1 globulin
0.3Q_
2 Gamma globulin 3.0
3
1 1 1 1
10.2
A Beta1 globulin 5.8
5
Beta globulin 6.A
6
I t I t
8.2
7 Alpha globulin 8.1
8 Albumin
37
9
1 1
20
10 Albumin1 1 .00
62
pH 8.6), fraction 4 gave a single broad peak with an
approximate mobility of -2.9*10“5 cm.2/volt/sec.,
corresponding to that of beta globulin; and fraction 8 gave
two peaks, with mobilities of -5.2 and -6.4*10"^
cm.2/volt/sec., corresponding to those of alpha globulin
and albumin, respectively.
The two colored bands which may be observed in the
starch block after electrophoresis, yellow in fractions 8
and 9 and pink in fraction 6, may be considered to
correspond to albumin and transferrin, respectively (70).
There Is good evidence, therefore, from several sources,
that the major peaks of protein, at least, in the present
method of starch block electrophoresis are substantially
the same as those of free boundary electrophoresis.
Approximate electrophoretic mobilities, in starch, of
the major protein components were calculated from observed
rates of migration and from values given by Kunkel and
Slater for electroosmotic flow and albumin mobility (67).
These are shown in Table III.
Antibody distribution. The antibodies fractionated by
starch block electrophoresis are shown in Table IV. The
results of several starch blocks have been averaged for
each of these antibodies, and the averaged data are
presented in Tables V through XVII. No tables are given
for And complete antl-E and Ste anti-Lea, because no
TABLE III
APPROXIMATE ELECTROPHORETIC MOBILITIES
OP SERUM FRACTIONS FOUND IN STARCH BLOCK
(Barbital Buffer, 0.05 M, pH 8.6)
Alpha globulin -4.9
Beta globulin -3*6
Beta’ globulin -2.3
Gamma globulin -0.8
Gamma1 globulin +1.1
Protein Mobility
-7.8*10~5 cm.^/volt/sec.
-6.5
Albumin1
Albumin
6b
TAELE IV
ANTIBODIES FRACTIONATED BY STARCH BLOCK ELECTROPHORESIS
Serum Antibody
And Anti-D
Complete
Incomplete
Anti-E
Complete
Incomplete
Anti-A
Anti-B
Complete
Incomplete
Complete
Incomplete
W at Antl-D
Be a Anti-E
Die Anti-B
Ste Antl-Lea
65
antibody activity could be detected in any of the fractions
obtained. The column labeled "Antibody Units" gives the
antibody distribution among the ten standard fractions.
The major peak of antibody activity appears, in nearly
every case, in fractions 2-4 (Table XVIII). Such a peak
probably existed also for Bea antl-B, but was too weak to
be detected; results obtained with low titer antibodies
cannot be accepted as conclusive. The complete antibodies,
in general, have a slightly greater mobility than the
Incomplete antibodies in this region. This is readily
apparent in Tables V through XVII. On individual blocks,
the separation of the two types of antibody was
sufficiently great so that fractions were frequently
obtained which contained complete or incomplete antibody
exclusively.
The minor peaks of activity in fractions 8-10 are
rather surprising, because there antibodies are generally
considered to be gamma globulins, while fractions 8-10 are
predominantly albumin. In some cases a relatively large
proportion of the total antibody activity appears in this
forward peak (Table XIX). It is Interesting, and perhaps
significant, that these cases Include both the natural
anti-B's and four of the five antibodies in And serum.
Two starch blocks were run with buffers other than pH
8.6 barbital. With pH 4.6 acetate buffer, the albumin peak
66
TABLE V
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON COMPLETE ANTI-D
(And Serum)
Titers Specific Antibody
Fraction (Range) Activity Units
1 0-8 5.8 6.7
2 0-4 0.96
10.3
3
0-1 0.0 60
2.5
4 0-1
0.069 1 .43
5
0 0 0
6 0 0 0
7
0 0 0
8 0-2 0.026
3.3
9
0-4 0.084
6.3
10 0-16
4.5 16.0
Total -
47
Serum 0 0
Recovery {%) - -
(Data from Blocks 22-24, 27-29, 37)
67
TABLE VI
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON INCOMPLETE ANTI-D
(And Serum)
Fraction
Titers
(Ran^e)
Specific
Activity
Antibody
Units
1 0-64 106
93
2 4-256 102 850
3
8-1 28 1 7.6 730
4 2-32 1 1 .8 260
3
0-16
2.5 45
6 0-8 1 .76 50
7
0-32 1 .56 46
8 0-8 0.26 34
9
0-1 0.057
4.1
10 0-2 1 .39
11 .0
Total -
2100
Serum 1 1 .4 4800
Recovery {%) - 44
(Data from Blocks 22-24, 27-29, 37)
68
TABLE VII
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON COMPLETE ANTI-D
(Wat Serum)
Titers Specific Antibody
Fraction (Range) Activity Units
1 1
33
14
2 0 0 0
3
0 0 0
4 0-32 1 2 192
5
0-2
1 -1
14
6 0 0 0
7
0 0 0
8 0 0 0
9
0 0 0
10 1
0.25 13
Total - 230
Serum 4.4 3700
Recovery {%) -
6
(Data from Blocks 49-50)
69
TABLE VIII
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON COMPLETE ANTI-D
(Wat Serum)
Titers Specific* Antibody
Fraction (Range) Activity Units
1 0 - 0
2 0-32 - 180
3
16-64 - 530
4 16-32 - 330
5
0-2 -
9
6 0 - 0
7
0 - 0
8 0 - 0
9
0 - 0
10 - - -
Total - 1050
Serum 4.4 3700
Recovery {%) ~
28
(Data from Blocks 51-52)
* Protein concentrations of fractions were not determined,
as buffer contained albumin.
70
TABLE IX
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON INCOMPLETE ANTI-D
(Wat Serum)
Fraction
Titers
(Range)
Specific
Activity
Ant ibody
Units
1 0 0 0
2 0 0 0
3
4-64 90 37
4 16-128 69
420
3
2-16 6.2 140
6 1-4 3.0 26
7
0-1 0.46
7
8 1
0.045 8
9
0-1 0.028 7
10 0 0 0
Total - 650
Serum 8.8 7400
Recovery {%) -
9
(Data from Blocks 49-50)
71
TABLE X
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON INCOMPLETE ANTI-D
(Wat Serum)
Titers Specific* Antibody
Fraction (Range) Activity Units
1 8-32 -
1 90
2 16-32 - 240
3
— 16-256 - 1800
4 16-128
-
750
5
0-2 -
9
6 0
-
0
7
0
-
0
8 0 -
0
9
0 - 0
10
- - -
Total -
3000
Serum 8.8 7400
Recovery {%) - 41
(Data from Blocks 51-52)
* Protein concentrations of fractions were not determined,
as buffer contained albumin.
72
TABLE XI
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON COMPLETE ANTI-E
(Bea Serum)
Fraction
Titers
(Range)
Specific
Activity
Antibody
Units
1 0-4 8.4 1 0.0
2 0-32 7.4 51
3
32-128 44 1060
4 1-64
25
210
5
0-16 4.0 45
6 0-1 0.173 2.5
7
0 0 0
8 0-1 0.0140
2.5
9
0-2 0.022
1 .6
1 0 0-1 0.22 0.8
Tutal - 1380
Serum
5.7 2560
Recovery (%) - 54
(Data from Blocks 26 and 38-41)
73
TABLE XII
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON INCOMPLETE ANTI-E
(Bea Serum)
Titers Specific Antibody
Fraction (Range) Activity Units
1 0-16 24
25
2 0-64 20 176
3 8-1 28 25 750
4 2-16
7.5 68
5
0-4 0.90 10
6 0-1
0.173 2.5
7
0 0 0
8 0 0 0
n
u
0-1 0.0112 0.80
10 0 0 0
Total -
1030
Serum
5.7 2560
Recovery {%) -
40
(Data from Blocks 26 and 38-41)
74
TABLE XIII
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON INCOMPLETE ANTI-E
(And Serum)
Fraction
Titers
(Rangel
Specific
Activity
Antibody
Units
1 0 0 0
2 0-4
2.5
27
5
2-4
1.19 43
4 0-2 2.1 46
5
0 0 0
o 0 0 0
7
0-1
0.075 2.3
8 0 0 0
9
0 0 0
10 0 0 0
Total -
118
Serum
2.9 1 280
Recovery {%)
-
9.2
(Data from Blocks 23-24 and 28-29)
75
TABLE XIV
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON ANTI-A
(And Serum)
Fraction
Titers
(Ranp;e)
Specific
Activity
Antibody
Units
1 0-4
3.5
4.0
2 0-8 1 .32 14.6
3
2-4 1 .60 56
4 2-4
2.3 77
5 0-4 0.60 13.2
6 0-4 0.30 8.8
7 0 0 0
8 0 0 0
o
0-2 0.058 4.4
10 0-4
1 .34 4.8
Total -
183
Serum 1 .42 640
Recovery {%) -
29
(Data from Blocks 23-24 and 27-29)
76
TABLE XV
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON ANTI-B
(And Serum)
Fraction
Titers
(Range)
Specific
Activity
Antibody
Units
1 0-4
7.1 8.0
2 0-1 0.42
4.5
3
0-1
0.105 3.8
4 0-2
0.57 12.3
5
0 0 0
6 0 0 0
7
0 0 0
8 0 0 0
9
0-4 0.148 1 1 .0
10 0-4 1 .68 6.0
Total -
46
Serum
0.089 40
Recovery {%) -
115
(Data from Blocks 23-24 and 28-29)
77
TABLE XVI
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON ANTI-B
(Die Serum)
Fraction
Titers
(Range)
Specific
Activity
Antibody
Units
1 0 0 0
2 0-16
0.67
24
3
4-256 13.7
104
4 16-128 36 510
5 8-32 16.4 290
6 1 -8 1 .49 65
7
0-1 0.38 7
8 0 0 0
9 0 0 0
10 0-1 0.28 5
Total - 1000
Serum
5.69 2560
Recovery {%) -
41
(Data from Blocks 42 and 44-46)
78
TABLE XVII
SUMMARY TABLE OF STARCH BLOCK RESULTS
ON ANTI-B
(Bea Serum)
--------
Titers Specific Antibody
Fraction (Ran^e) Activity Units
1 0 0 0
2 0 0 0
3
0 0 0
4 0 0 0
5
0 0 0
6 0 0 0
7
0 0 0
8 0 0 0
9
2 0.1 1 1 8
10 2 2.2 4
Total - 1 2
Serum 0.045 20
Recovery {%) -
60
(Data from Block 26)
79
TABLE XVIII
RELATION OF ANTIBODY ACTIVITY TO STARCH BLOCK FRACTION
Serum
And
Nat
Bea
Die
Fractions with
Antibody Activity
Antl-D
Complete 1 -2 and 9-10
Incomplete
2-3
and 10
Anti-E
Complete -
Incomplete 3-4 and
7
Anti-A 3-4 and 9-10
Anti-B 4 and 9-10
Anti-D
Complete 3-4 and 1 0
Incomplete 3-4
Anti-E
Complete
3
and
8-9
Incomplete
2-3
and
9
Anti-B 9-10
Anti-E
4-5
and 10
TABLE XIX
PROPORTION OF ANTIBODY ACTIVITY IN
ALBUMIN AREA ON STARCH BLOCKS
Proportion In
Antibody Fractions 8-10
Bea anti-B 100 %
And complete anti-D 54
And antl-B 37
And antl-A 5
And Incomplete anti-E 2
remained at the origin, while the antibodies migrated
toward the cathode. With pH 7.1 phosphate buffer, the
antibodies had approximately the same distribution among
the fractions as with pH 8.6 barbital buffer, but the
albumin peak migrated a short distance toward the cathode.
This rather anomalous result is difficult to reconcile with
theoretical considerations or with other observations;
further experiments would have to be done to clarify the
situation. It is interesting that even at pH 4.6 recovery
of antibody activity was not significantly reduced. In
neither experiment was the separation of complete and
incomplete antibodies Improved.
On two starch blocks, selected fractions from earlier
blocks were remigrated. On block 33, pooled gamma globulin
fractions reappeared in fractions 2-4. Substantially all
of the original activity was recovered in these fractions.
Apparently this gamma globulin fraction has about the same
mobility either in the presence or in the absence of other
serum proteins. This was not true of pooled fractions 10,
which were remigrated on block 34. The activity, which was
recovered quantitatively, appeared not In fraction 10, but
in fraction 4; while the peak of protein concentration,
representing albumin contamination, appeared in fraction 7.
Evidently, this antibody does not migrate to fraction 10
except in the presence of other serum proteins. The
82
antibody appearing in fraction 10 may be identical to
certain of the antibodies which usually appear in the gamma
globulin fractions. Serologically, the only difference
between these and the antibodies in fraction 10 is the
relatively high complete/incomplete ratio of the latter;
this may be merely an indication that complete antibodies
are more inclined to appear in fraction 10.
As described under "Antibody recovery" below, a number
of experiments were done in which human serum albumin was
added to the electrophoresis buffer. These were based
partly on the idea that the rapid migration of these
fraction 10 antibodies might be due to complexing with
albumin, or in some other way related to albumin or total
protein concentration. It would be desirable to Induce
more of the antibody to appear in fraction 10, because of
the difference in complete/incomplete ratio in that
fraction. It was found, however, that albumin did not have
the desired effect. The presence of some more specific
protein may be required.
Antibody enrichment. Although partial separation of
complete and incomplete antibodies has been demonstrated, a
less satisfactory situation was found with respect to
purification. The greatest enrichment of antibody with
respect to protein, as reflected in the specific antibody
activity, does not necessarily occur in the fractions with
83
the highest titers. In fact, the peaks of antibody
activity coincide fairly closely with the peaks of protein
concentration. As a consequence, those fractions which
show the greatest enrichment of antibody usually contain
only a small proportion of the total antibody units, a
circumstance not conducive to purification of the
antibodies with good yield.
In a large number of experiments, the specific
antibody activity of fractions has seldom exceeded ten or
twenty times that of the original serum.
In Table XX are given the fractions which, on the
average, show the best gain in specific activity over the
original serum. In most cases, this is in fraction 1 ; less
often it is in fraction 3 or 4.
Another kind of enrichment is enrichment of complete
antibody with respect to incomplete, and this is indicated
by the ratio of the titers or antibody units for the two
kinds of antibody. As will be seen from the data in Table
XXI, this ratio tends to be highest in fractions 9 and 10.
For Bea serum, the ratio is relatively large in the gamma
globulin region. A possible explanation for this
difference would be the presence in the other two sera of
blocking antibodies, which would reduce the apparent titers
of complete antibodies, therefore reducing the apparent
complete/incomplete ratio. This explanation seems
84
TABLE XX
ANTIBODY ENRICHMENT
Fraction with
Antibody Best Enrichment
And Anti-D
Complete 1 and 10
Incomplete 1-2
Anti-E
Incomplete 2-4
Anti-A 1 and 4
Anti-B — 1 and 10
Bea Anti-E
Complete 3 and 10
Incomplete 1 -3
Anti-B 10
Wat Antl-D
Complete 1
Incomplete 3
Die Anti-B 4
85
TABLE XXI
COMPLETE/INCOMPLETE RATIO*
Fraction And Anti-D
Antibody
Bea Anti-E Wat Anti-D
1 3.2 0.40 0.144
2 0.54
0.29 1 .46
3 0.153
1 .41 0.50
4 0.25 3.1
0.82
5
0
4.5 0.32
6 0 1 .0 0
7
0 0 0
8 4.5
00 0
9 69
2.C 0
10 65
00 oo
* Adjusted to correspond to a ratio of 1.00 for the
original serum.
86
justified for And serum, at least, because complete anti-D
cannot be detected at all in the original serum. It can be
detected only after electrophoresis, which presumably
separates the complete from the blocking antibody. The
infinite ratios given for three of the fractions are merely
a reflection of inadequacies In the serological methods;
they very likely represent large finite numbers.
Antibody recovery. Considerable variation is found in
the completeness with which antibody activity is recovered
in different experiments, and this is no doubt the result
of small differences in experimental conditions. It is not
easy in a method such as starch block electrophoresis to
control, or even to recognize, all of the variables. The
excellent recoveries achieved from 3tarch block no. 23
indicate that good results are possible with the right
combination of conditions; there is nothing in the records,
however, to distinguish this block experimentally from the
less successful ones which preceded or followed it.
Under standardized conditions the average recovery of
antibody activity, ai. shown in Tables V through VII, ranges
from 9 to 115 per cent for different antibodies and
averages about 43 per cent for all antibodies. Recovery of
only 9 per cent of And incomplete antl-E is probably an
Indication of unusual lability in this antibody, especially
when it is remembered that the recovery of And complete
anti-E was always zero. In contrast, good recoveries have
been achieved of both types of anti-E in Bea serum.
Recovery of 115 per cent of And anti-B may indicate that
this antibody was Inhibited by blocking antibody in the
original serum. The somewhat lower recovery of Pic antl-B
is perhaps a consequence of the immune nature of this
antibody. Next to And complete antl-E, And complete antl-D
seems to be the most labile of the antibodies studied;
while the recovery of the latter could not be calculated,
because it was not evident in the original serum, recovery
was very often zero. Taken together, all these
observations seem to suggest that antibodies of the same
specificity taken from different sources may have different
degrees of stability.
Because of its importance to the separation and
purification of antibodies, the question of antibody
stability and the factors affecting it have received much
attention in this work.
Total protein recovery is usually between 80 and 90
per cent, and it is reasonable to assume that only a part
of the protein loss is mechanical. Therefore, mechanical
losses of antibody, resulting in the manipulations of the
starch block, the eluates, and the concentrates, may be
estimated at not more than 10 per cent. Consequently, the
greater portion of the 50 per cent or greater loss of
88
antibody activity which is usually experienced is assumed
to be due to denaturatlon of the antibody, at least with
respect to the part of the molecule that confers the
activity. That this loss is Irreversible is indicated by
the fact that recombination of fractions does not restore
the original activity.
It is difficult to determine whether the loss in
activity occurs during electrophoresis, elution, or
pervaporatlon. The eluates from the starch are too dilute
for reliable titrations, antibody being undetectable in
most fractions. Titrations of eluates do suggest, however,
that relatively small losses in activity occur during the
concentration of the eluates. If whole serum Is diluted
with barbital buffer, mixed with starch, eluted in the
usual manner, and concentrated to its original volume,
there is no loss in antibody activity. Experiments on
different methods of concentration, all carried out on the
same pools of eluates from starch blocks (no. 25, 29,
31-32, and 35-36) suggest that substantially the same
recovery is achieved with pervaporatlon, dialysis against
gum acacia, or salting out. If whole serum is diluted with
nine volumes of Isotonic saline solution and then
concentrated to Its original volume, complete recovery of
activity is achieved, regardless of whether the
concentration is effected by pervaporatlon in the cold or
at room temperature, dialysis against gum acacia, or
lyophilization. These results seem to point to the
conclusion that, while the antibodies are quite stable in
whole serum, they are less stable when partially isolated
from other serum proteins. The general impression one gets
is that the greatest losses occur during electrophoresis,
and that these are the direct result of the separation of
antibodies from other serum proteins. Further observations
on the stability of antibodies will be described in a later
section.
Several attempts were made to minimize loss of
activity by adding human serum albumin to the serum applied
to the starch block (no, 49 and 50), by using buffer
containing 1 per cent human serum albumin (no, 51-52) or
3.3 per cent sucrose (no. 48), or by using AE plasma (no.
43) or 0.4 per cent human albumin as the eluting solution
(no. 46). Recovery was poor on blocks 43, because the
plasma clotted during pervaporatlon, and 48, because the
fractions were Inadvertently subjected to elevated
temperatures- overnight. On the remainder of these blocks,
recovery was good, but probably no better than would have
been achieved without addition of albumin to the serum,
buffer, or eluting solution. Since albumin does not seem
to improve the stability of the antibodies, it is possible
that some specific factor or group of factors found only in
90
whole serum Is essential to maximal stability of
antibodies.
One experiment (block no, 30) was performed with Ste
serum, which contained anti-Lea with a titer of 32. A
special antiglobulin serum (titer, 512) was used to titrate
this serum and the fractions obtained from it. Wo activity
was found in any of the fractions, even after concentration
to one-fifth the volume of the original serum. There was
not sufficient serum available for further experiments.
Evaluation. Starch block electrophoresis combines
high capacity and unusual adaptability with the excellent
protein resolution obtainable in electrophoresis methods.
It is an almost indispensable tool in exploratory work on
serum proteins which requires the availability of fractions
corresponding to the conventional components found in free
boundary electrophoresis. Reasonably good separation of
complete and incomplete antibodies has been achieved by
starch block electrophoresis, and, in addition, antibodies
have been separated into two peaks of activity at opposite
ends of the block. In some fractions, antibodies have been
enriched ten- to twenty-fold, with fairly good yield.
Among the disadvantages of starch block
electrophoresis are the difficulty of closely controlling
experimental conditions and the length of time required for
a single separation, usually more than a week. Location of
the protein "bands and cutting of segments from the "block is
necessarily an imprecise process, so that averaging of
results from several starch blocks causes an apparent loss
of resolution. Also, the number of segments which can be
cut is limited not only by the amount of labor Involved,
but also by the requirement of a workable volume of final
concentrate in each fraction; spreading the antibody over a
larger number of fractions would Increase the difficulty in
detecting it.
The interesting results obtained from the large number
of starch blocks run in the present study attest to the
general usefulness of the method.
Starch gel electrophoresis. Although it yielded
striking serum protein patterns, starch gel electrophoresis
proved disappointing as a method for the separation of
antibodies. Pic anti-A was held up at the origin, there
was no enrichment of antibody, and recovery was only 5 per
cent.
Because of the localization of antibody at the origin,
gel electrophoresis was considered as a possible method of
separating antibodies from other serum proteins. Ten
milliliters of Pic serum was applied at the center of a
short, thick gel. After electrophoresis, the starch at the
origin was cut out and eluted. Only 10 per cent of the
antibody was recovered, and this had a specific activity
92
only twice as great as that of the original serum. No
further work was done on this method.
Free boundary electrophoresis. Free boundary
electrophoresis was used as a fractionation method in one
experiment. After separation of Wat serum, small fractions
were taken from the tops of the ascending and descending
limbs of the cell. Photographs taken before and after the
removal of these fractions indicate that they contained
mostly albumin and gamma globulins, respectively. Because
of the small volumes, these fractions could not be
concentrated. No antibody activity could be detected in
the albumin fraction. The gamma globulin fraction had a
titer of 16, and a specific antibody activity nearly ten
times as great as that of the original serum. The findings
suggest that at least a part of the antibodies are, by
definition, gamma globulins.
Sedimentation
Studies on sedimentation complement those on
electrophoresis, because these two methods of fractionation
are based on different properties of the protein molecules.
There is no necessary correlation between molecular charge
and molecular size and shape. The Splnco No. 40 rotor,
used in the present work, accepts tubes of ten milliliter
capacity. Four fractions, three of three milliliters and
one of one milliliter, were collected for analysis, and the
93
fractions were numbered U-1 through U-4, beginning at the
top. When plasma was centrifuged, a gelatinous plaque
usually formed at the bottom of the tube. This probably
consisted mostly of fibrin; it was suspended in isotonic
saline solution and called fraction U-5.
Protein distribution. After centrifugation of serum,
the part in the top third of the tube was colorless and
cloudy. The remainder of the serum was yellow, with the
intensity of the color increasing gradually toward the
bottom of the tube. Protein determinations on the
fractions averaged 21, 39, 54, and 72 milligrams of protein
per milliliter for U-1 to U-4, respectively. No studies
were made on the electrophoretic components present in the
various fractions.
Antibody distribution. Fifteen sedimentation
experiments were performed on a Red Cross 0 plasma (anti-A,
64), And serum, Wat serum, Cas serum, and three starch
block fractions. In general, the antibodies sediment at a
greater rate than most other serum proteins, complete
antibodies sedimenting more rapidly than incomplete. A
summary of representative data is given in Tables XXII
through XXIV. The specific antibody activities and the
complete/incomplete ratio become progressively larger from
the top to the bottom fractions, while for the original
serum these quantities have intermediate values. From the
94
observed rates of sedimentation, the minimum molecular
weights of complete and incomplete antibodies were
estimated as 1,000,000 and 120,000, respectively (88).
Antibody enrichment. The greatest antibody enrichment
was achieved in fraction U-4 or U-5. The specific activity
in this bottom fraction was from two to eight times as
great as that of the original serum.
Antlbody recovery. Recovery of protein averaged about
99 per cent. Recovery of antibodies was approximately
twice as great as would be expected from starch block
electrophoresis. Recovery of anti-D from Wat serum
averaged 38 and 65 per cent, respectively, for complete and
Incomplete types, as compared to 17 and 27 per cent,
respectively, from starch blocks. Recovery of anti-A from
Red Gross 0 plasma averaged about 41 per cent, as compared
to 29 per cent recovery of anti-A (from And serum) from
starch blocks. Recovery of incomplete antl-D from And
serum by ultracentrifugation was about 230 per cent. This
result was obtained consistently and is significantly above
100 per cent. Recovery of this same antibody from starch
block electrophoresis averaged only 44 per cent. The
reason for the large difference between results obtained
with the two methods is not clear. No complete anti-D
could be detected in the ultracentrifugal fractions from
And serum. The albumin antibody in Cas serum showed a
95
recovery of 120 per cent.
Pooled gamma globulin fractions from starch blocks,
when subjected to ultracentrifugation, lost about 23 per
cent of their total activity and gained but little In
specific activity. Pooled gamma1 fractions lost 95 per
cent of their total activity, and specific activity was
diminished.
For evaluation of ultracentrifugation as a method for
concentration of dilute antibody solutions, fraction 4 of
starch block no. 50 was diluted ten times with isotonic
saline solution, and a 10 ml. aliquot of this dilution was
ultracentrifuged. The original dilution had complete and
incomplete anti-D titers of 3.2 and 1.6, respectively; no
activity was detectable in the fractions after
centrifugation.
Evaluation. Ultracentrifugation is a high-capacity
method which can supplement electrophoresis methods. In
some ultracentrifugal fractions incomplete antibodies can
be obtained relatively free of complete antibodies, and
recovery of antibody activity is often better than that
achieved by starch block electrophoresis.
On the other hand, resolution of proteins into
discrete fractions is relatively poor with
ultracentrifugation. It seems that in most cases, more
efficient methods of fractionation are available.
96
TABLE XXII
SUMMARY OF SEDIMENTATION
OF INCOMPLETE ANTI-D
(And Serum)
Fraction
Specific
Activity
Antibody
Units
U-1 8.2 510
U-2 21 2600
U-3
30 6200
U-4 30 2400
Total - 11700
Serum 11 .4 51 20
Recovery
(%) - 230
(Data from Experiments
7-9)
97
TABLE XXIII
SUMMARY OP SEDIMENTATION
OF ANTI-D
(Wat Serum) -
Specific Activity Antibody Units
Fraction Complete Incomplete Complete Incomplete
U-1 0 1 2.0 0 640
U-2 0 10.0 0 1280
U-3 1 .65 13.3 320 2560
U-4 9.8 18.6 960 1 920
Ppt#
19.9 9.8 640 320
Total - -
1 920 6700
Serum 1 1 .4
23 51 20 1 0240
Recovery{%) - -
38 65
(Data from Experiments 10-11)
# "Ppt" indicates protein which precipitated as a result
of contact with the salt displacing solution (saturated
NaCl). This precipitate was redissolved and dialyzed
against Isotonic saline solution for titrations.
98
TABLE XXIV
SUMMARY OF SEDIMENTATION
OF ANTI-A
(Red Cross 0 Plasma)
Fraction
Specific
Activity
Antibody
Units
U-1 0.056
3
U-2
0.155
20
U-3
0.161 26
U-4 0.172 1 28
U-5
11 .00
83
Total - 260
Plasma 1 .42 640
Recovery
{%) -
41
(Data from Experiments 1-5)
99
Anlon-Exchange Cellulose Chromatography
Two DEAE-cellulose columns were run on Red Cross serum
and And serum. The 200 tubes of eluate collected from each
column were divided into seven fractions, on the basis of
the curve obtained by plotting protein concentration versus
tube number. These fractions were numbered C-1 through
C-7, beginning with the first to be eluted. In later work
the procedure was simplified to a batch operation, from
which four fractions were obtained, 8.0, 6.5, 5.6, and 4.5,
numbered according to the respective pH values of the
buffers used to elute them from the cellulose. These four
fractions were intended to correspond to the column
fractions approximately as follows:
Column Batch
C-1 and C-2 8.0
C-3 6.5
C-4 and C-5 5.6
C-6 and C-7 4.5
The batch operation was used in the fractionation of
And, Wat, and Cas sera and certain ultracentrifugal
fractions.
Protein distribution. Fractions C-1 and C-2 contained
a sharp peak of protein which appeared very early in the
elution process, and fractions C-4 through C-7 contained a
broad peak which represented most of the serum protein.
100
These two peaks were reported by Fahey and coworkers (71)
as consisting primarily of gamma globulins and albumin,
respectively, while C-3 contained mostly beta globulins.
Not enough experiments have been done in the present work
to ascertain the percentage of total protein in each of the
fractions. Each of the four fractions from the batch
operation contained roughly equal quantities of total
protein.
Antibody distribution. Both anti-A and anti-B from
the 0 serum showed large peaks of antibody activity in
fractions C-1 and C-5, with considerable activity in the
intervening fractions. With And serum, most of the
activity of anti-A and incomplete antl-D was localized in
fractions C-1 and C-2, with some Incomplete anti-D activity
also in fractions C-3 through C-5. With the exception of
And antl-A, therefore, the antibodies were broadly
distributed throughout the fractions.
More striking results have been achieved with the
batch procedure. For And antl-D, the peaks of complete and
incomplete activity were found in fractions 4,5 and 8.0,
respectively; and for Wat anti-D, the peaks of complete and
incomplete activity tended to appear also in fractions 4.5
and 8.0, respectively, with Wat anti-B in fraction 4.5.
The same pattern of distribution was observed when Wat
anti-D, from pooled ultracentrifugal fractions U-4, was
101
subjected to the batch process.
Antibody enrichment. In the column experiments, the
largest increase in specific activity of fractions over the
serum was usually observed in the earlier fractions, C-1
through C-3, which contained mostly gamma globulins. The
greatest activities were 2 to 15 times the specific
activity of the serum.
In the batch experiments, specific activity was
increased 2 to 5 times in some fractions.
Antibody recovery. In both column experiments,
recovery of anti-A was about 25 per cent, of antl-B about
19 per cent. Recovery of And incomplete anti-D was 133 per
cent, which is significantly greater than the recovery from
starch block electrophoresis. No And complete anti-D was
detected in the column fractions.
In the batch experiments, And complete anti-D could be
demonstrated. Recovery of And incomplete anti-D was about
57 per cent. This compares with average recoveries of 25
and 78 per cent, respectively, for complete and Incomplete
Wat anti-D. Recovery of Wat antl-B was 72 per cent.
Cas serum was fractionated by the batch procedure, but
no activity was detected in the fractions, probably because
they were too dilute to show the low-tlter antibody.
Evaluation. Although this method, like
electrophoresis, depends on molecular charge for its power
102
of separating proteins, it is not limited to the use of a
single buffer solution per experiment. Variation of pH and
ionic strength of the buffer allow the molecular charge to
vary, and at the same time there are opportunities for
transient interactions between proteins and the
DEAE-cellulose and among the proteins themselves. As a
result, the cellulose fractions do not exactly parallel
those obtained from electrophoresis, even though there is a
general correspondence; and an extra dimension can be added
to fractionation experiments through combination of the two
methods.
Results obtained with the laborious column method have
not beaiparticularly good, although the real value of this
method could not be decided on the basis of two trials.
The batch operation, contrary to expectations,
permitted much better resolution of peaks of antibody
activity. Separation of complete from incomplete
antibodies is better than with any other method tried in
this work. Recovery of the various antibodies seems to be
at least as good as with starch block electrophoresis.
Advantages of this method include simplicity of equipment
and rapidity, and there is usually no need for fufther
concentration of the fractions.
Precipitation Methods.
Under this heading are grouped experiments based on
103
solubility-ionic strength relations. Included are Cohn
fractionation, salting out, and dialysis against distilled
water.
Protein distribution. Ho attempt was made to
characterize the protein fractions obtained in these
methods. The Cohn method, an abridgement of his Method 10,
yielded three fractions, A, E, and G, which were intended
to contain albumin, alpha and beta, and gamma globulins,
respectively. No doubt, there was considerable
overlapping. In the salting out and dialysis methods, the
precipitate was assumed to contain principally gamma
globulins. About 50 per cent of the total proteins of And
serum were salted out by 1.5 inolar ammonium sulfate, and
about 5 per cent of the total by dialysis against distilled
water.
Antibody distribution. And and Nat sera were
subjected to Cohn fractionation. In both experiments,
fraction G contained the major portions of complete and
incomplete anti-D, and small amounts of Incomplete anti-D
were found in fractions A and B. It is interesting that
antibody distribution in this method superficially
resembles that found in ultracentrifugation, a quite
different method.
At 1.5 molar ammonium sulfate, about 95 per cent of
the incomplete anti-D of And serum was salted out.
104
Dialysis of And serum against water resulted in
precipitation of about 80 per cent of both anti-A and
incomplete anti-D.
Antibody enrichment. The specific antibody activity
of fraction G was in all cases about 4 times as great as
that of the original serum. Salting out yielded a
precipitate with 16 times the original specific activity
of the serum. Dialysis against distilled water jrielded a
precipitate with 4 times the specific activity of the
original serum.
Antibody recovery. In Cohn fractionation an average
recovery of about 50 per cent was achieved with And
incomplete anti-D and Rat complete and incomplete anti-D.
Recovery was practically quantitative from salting out and
dialysis experiments.
Evaluation. The Cohn fractionation method, as used
here, possesses no special merit for the separation of
complete from incomplete antibodies, and neither enrichment
nor recovery figures are especially noteworthy. The same
may be said of dialysis against distilled water. Salting
out results in substantial Increase in specific activity,
and this method has been investigated with respect to the
concentration of eluates from starch blocks. Details are
given in another section.
105
Stroma-Resin Adsorption
Specific adsorption of antibodies on red cell stroma
linked to a resin, followed by elution of the antibodies,
holds promise of being a powerful method for the
preparation of purified antibodies. The theoretical
advantages of an imiaunologically specific method are
obvious. Only ABO stromata and antibodies were used in the
present work, because adaptation of the method to the
preparation of Rh antibodies was not feasible at the time.
Protein distribution. Antibody solutions eluted from
stroma-resin were subjected to free boundary
electrophoresis. The pattern obtained resembled that of
whole serum. Apparently, there had been considerable
non-specific adsorption of serum proteins on the resin.
Antibody distribution. Regardless of the titer of the
original serum, only a small fraction of the total
antibodies was adsorbed on the stroma-resin.
Antibody enrichment. In none of the experiments was
the specific antibody activity of the eluate from the
stroma-resin more than double that of the original serum.
Antibody recovery. Recovery of antibody in the eluate
ranged from zero to about 8 per cent.
Evaluation. Several batches of stroma-resin were
prepared, several different plasmas and sera were tried,
and the conditions of adsorption and elution were
106
systematically varied. Despite this, satisfactory results
were not obtained with this method. Further investigations
along these lines could well be rewarding, however.
Combined Methods
Application of different fractionation methods in
sequence has theoretical advantages, and it was considered
as a possibility in the present worlc on the separation of
antibodies. Descriptions of several examples of coupling
of methods have been given in previous sections. Starch
block fractions have been taken to ultracentrifugation,
DEAE-cellulose fractions to starch block electrophoresis
and salting out, ultracentrifuge fractions to
DEAE-cellulose fractionation, and starch block fractions to
salting out or remlgration under altered conditions.
As an example, 20 ml. of Y/at serum was
ultracentrifuged. Fraction UC-4 was separated by the
DEAE-cellulose batch method. Fraction 4.5 was salted out.
In the final precipitate from salting out, which contained
less than one per cent of the total original activity, the
complete and incomplete antibodies had specific activities
of only 23 and 13, respectively, as compared to 11 and 23
in the original serum. An increase of about four-fold was
gained in the complete/incomplete ratio, but at the expense
of about 99 per cent of the activity.
While such coupling of methods yielded interesting
107
results, It did not appear economical of time and
materials. The additional losses of antibody activity in
the second separation more than compensated for any
Increase in resolution or specific activity. With
increases in the number of separations, the antibody
recovery diminishes In a geometrical manner, and with the
low recovery per separation which has been observed In this
work, the total amount of antibody activity soon reaches
the vanishing point.
This prospect could best be remedied by finding a way
of stabilizing the antibodies or otherwise increasing
overall recovery.
Antibody Stability
Aside from the experiments described in previous
sections, several observations have been made on the
subject of antibod3 ' - stability.
Three aliquots were taken from an antibody solution,
prepared by electrophoresis on a starch gel, and stored at
4°, 23°, and 37° for 16 days. Antibodies were known to be
stable for months or years in the frozen state at -20°, but
it was suspected that they might deteriorate at higher
temperatures. In each of the three aliquots, however, the
titer was found unchanged on the sixteenth day.
As noted earlier, antibodies in serum were not
inactivated by dilution with saline solution,mixing with
108
starch, elution from starch, vigorous shaking for several
minutes, or by concentration by various methods.
The starch block electrophoresis experiment at pH 4.6
indicated that the antibodies were as stable at this pH as
at 7.1 or 8.6. In another experiment, antibodies (diluted
serum) were left at room temperature in solutions with pH
ranging from 2 to 12. Within the pH range 3 to 9, the
titer appeared undiminished at the end of two weeks, even
though some of the proteins had begun to precipitate out.
Ho activity was detected in the tubes below pH 3 or above
pH 9. In view of the fact that these solutions were not
adjusted to pH 7 before titration, the antibody-antigen
reaction is apparently not affected to any great extent by
variations in pH.
When antibodies can survive such drastic conditions,
one wonders why the recovery is so poor in fractionation
methods employing milder conditions, or in some cases
almost physiological conditions. The present studies do
not provide a direct answer to this question, but they do
suggest that the difficulty may not lie in the conditions
of the experiment. Except in special cases, where blocked
antibody is apparently uncovered, fractionation of a serum
results in loss of antibody activity. This loss seems to
be of the same order of magnitude regardless of the method
used. Although some antibodies seem to be Inherently
109
more stable than others, the recovery figures for most are
not far apart.
The rather unusual behavior of certain Isolated
fractions, noted here and there in the description of
results, provides further evidence that antibodies are
subject to changes in their characteristics when Isolated
from whole serum.
As stated in a previous section, attempts to stabilize
antibodies during fractionation procedures by the addition
of albumin to the solutions have not been highly
successful. The stabilizing nature of whole serum may be
due to some more specific factor. If purified antigen were
available, it might have possibilities as a stabilizing
agent, much as substrates tend to stabilize enzymes.
DISCUSSION
In the present work the primary objective, the
separation of complete and incomplete blood group
antibodies, has been achieved. In the course of the
studies, observations have been made on the properties and
behavior of these antibodies. Also, the present
investigation has delineated some of the practical
obstacles to be overcome as studies in this general field
continue.
Expression of Results
Methods of organizing and presenting experimental data
have been studied at some length, and it is felt that the
present system of presentation, while it represents some
degree of departure from established usage, is sound and
well suited to the requirements of clarity and thoroughness
in the treatment of the large body of data collected in
these experiments.
The desirability of a clear and thorough presentation
is emphasized by the difficulty often experienced in
comparing and evaluating published results which do not
include sufficient data. This difficulty is all too common
in the field of blood group antibodies.
110
111
Antibody Distribution
As noted in the Introduction, several reports on the
separation of blood group antibodies have appeared in the
literature,
Electrophoresis. Deutsch and coworkers (89), using
free boundary electrophoresis, and Payne and Teming (90),
using paper electrophoresis, both found anti-A and anti-B
to migrate in the gamma-beta region. Cann and coworkers
(55,56), using convection electrophoresis to fractionate
several Rh sera, were able to classify the sera into
several types on the basis of differences in antibody
distributions. In all sera they found albumin-reacting
antibodies and blocking antibodies in low-mobility
fractions, the latter occasionally in two peaks of
activity. Antiglobulln-reactlng antibodies occurred in one
or two peaks in various globulin fractions.
These results are in general agreement with those of
the present investigation. Most antibodies have been found
to migrate in the fast gamma globulin region (fractions
3-4), and different kinds of antibodies appear in different
peaks of activity In different sera. The results of Cann
and coworkers seem more variable, but this may be
characteristic of the convection electrophoresis method,
which was not used in the present studies.
Migration of antibodies in the albumin region, as
112
observed In starch block electrophoresis, has not been
previously reported, although there have been occasional
unsubstantiated claims that antibodies are alpha globulins.
No evidence has been found in the present studies which
would support such claims, since even the antibodies in
fractions 8-10 appear to be gamma globulins. Association
of the antibodies with albumin, or perhaps with some more
specific proteins in the serum, is suggested as an
explanation of this rapid migration. Wiener (43) suggested
such an association to explain his conglutination reaction.
Sedimentation. A single paper, by Campbell and
coworkers (62), has appeared on ultracentrifugation of
blood group antibodies. These workers studied the
sedimentation behavior of Hh antibodies in several sera,
and their results were substantially the same as those
obtained in the present studies: complete antibodies
sedimented more rapidly than incomplete.
Estimates of the minimum molecular weights of complete
and incomplete antibodies were given as 1,000,000 and
120,000, respectively, in the Results section. Wallenlus
and coworkers (91) have observed gamma globulin components,
from starch block electrophoresis, with sedimentation
constants of 7, 19, 28, and 44 S. It seems likely that
incomplete antibodies occur in the first component, and
that complete antibodies occur in one or more of the faster
113
sedimenting components. A sedimentation constant of 44 S
would correspond to a molecular weight of several million.
Since "both kinds of antibodies have the same specificity,
the differences in serological behavior could be a result
of the difference in molecular size. Ho evidence has been
found in the present studies on the question of whether
complete antibody is a polymer of incomplete.
Modified cellulose chromatograph.y. Since completion
of the present work on fractionation with DEAE-cellulose,
a paper has appeared by Abelson and Rawson (92) which
generally confirms the results obtained with the batch
method. These workers used both DEAE-cellulose and
carboxymethyl-cellulose with a number of sera containing Rh
and ABO antibodies. They also obtained good separation of
complete and incomplete antibodies with DEAE-cellulose;
results with the CM-cellulose were less successful. Speer
and coworkers (93) achieved some separation of antibodies
by chromatography on DEAE-cellulose columns.
Proteins in the fractions obtained from DEAE-cellulose
do not correspond closely to electrophoresis mobility
classes (70,71). The fact that complete, and some
Incomplete, antibodies occur in a relatively minor gamma
globulin component in a fraction which is predominantly
albumin has interesting theoretical and practical
implications. This phenomenon tends to strengthen the
114
suggestion that association with albumin also occurs in
starch block electrophoresis and is responsible for the
appearance of antibodies in fractions 8-10. Occurrence of
antibody in the albumin fraction from DEAE-cellulose is
also of practical convenience because globulins are easily
separated from albumin by simple salting out. Combination
of these two methods is described in the Results section.
Precipitation methods. No specific data have been
found in the literature on separation of blood group
antibodies by Cohn fractionation, although Tullls (59)
indicates that Rh antibodies are found in Cohn fraction
III, which consists mostly of beta globulins. This is not
at variance with results of the present studies.
Witebslcy and coworkers (60,61 ) found some separation
of complete and incomplete antibodies upon dialysis of the
serum against distilled water. The present studies,
however, Indicate that this is a relatively Inefficient
method of separation.
Antibody Enrichment
The literature contains little reference to the
subject of enrichment of blood group antibodies with
respect to protein. Pillemer and coworkers (94) achieved
an 8-fold enrichment by fractional precipitation with
methanol. Campbell and coworkers (62) found that recycling
of ultracentrifuge fractions, because of large losses in
activity, did not result in any enrichment of antibody.
Isliker claimed an enrichment of anti-B of 350-fold by
stroma-resin adsorption; 75 to 90 per cent of the protein
in his purified solution was identified as specific
antibody. If this finding is generally applicable, then
the maximum possible enrichment of such antibodies would be
about 400-fold. In the present work, enrichment has not
usually exceeded 20-fold.
Some enrichment of complete antibody with respect to
incomplete, or vice versa, is almost inevitable with any
fractionation method, but the results obtained in this work
by starch block electrophoresis, ultracentrifugation, and
the DEAE-cellulose batch method have been striking. The
nearest approximation to these results were those of
Campbell and coworkers (62), who noted an Increase in the
complete/incomplete ratio upon recycling the bottom
fraction in ultracentrifugation.
Antibody Recovery
Many workers mention losses of antibody activity, but
actual recovery figures are not always given. Recovery of
Rh antibodies from convection electrophoresis (56) is given
as 40 to 70 per cent, and recovery from ultracentrifugation
by Campbell and coworkers (62) was comparable to recovery
116
from ultracentrifugation in the present investigation.
Isliker found up to 50 per cent recovery of ABO antibodies
from stroma-resin experiments.
Indications in the present work suggest that most
losses in activity occur in the initial separation of
antibodies from other serum proteins, and that this loss is
irreversible. Simple dilution and concentration of
antibody solutions seems to have little effect on their
stability. Considering the delicacy of their specificity,
antibodies exhibit remarkable stability under a variety of
conditions. Apparent recovery may exceed actual recovery
if antibodies blocked in the original serum are separated
from the blocking antibodies. Sturgeon (48) states that a
complete/incomplete ratio between 0.5 and 8 can result in a
prozone in titrations of complete antibody, and It is
possible that smaller ratios could result in total blocking
of complete antibody. In such a case, greater than 100 per
cent recovery of complete antibody could be realized in a
fractionation experiment which would effectively increase
the complete/incomplete ratio.
Since there are several kinds of incomplete
antibodies, one Incomplete could conceivably block another,
as has been suggested by Witebsky (60). Recoveries of more
than 100 per cent might therefore occur with incomplete as
well as with complete antibodies. In several experiments
117
in the present work, recovery of incomplete antibodies has
exceeded 100 per cent.
Other Antibodies
Assumption of relations which do not in fact exist is
a frequent source of error. For this reason, conclusions
about blood group antibodies based on reports of the
properties of other kinds of antibodies must be viewed with
caution. It may be observed, however, that all kinds of
antibodies do seem to have generally similar properties.
Stelos and Taliaferro (95), for example, in studying
hemolysins produced by rabbits against sheep cells,
obtained results with starch block electrophoresis and
sedimentation which resemble some of the results obtained
in the present work.
Kuhns (96), in studying diphtheria antitoxin, found
that the precipitating form migrated as s . slow gamma
globulin and that the non-precipitating, skln-sensltlzlng
form migrated as a fast gamma globulin in starch block
electrophoresis. He was able to separate mixtures of the
two forms with this method.
Christenson and coworkers (97) found that cold
hemagglutinins migrated as fast gamma globulins in
electrophoresis on cellulose powder columns.
Various other properties of antibodies in general are
118
also similar to those of blood group antibodies. Examples
are molecular weight, reaction temperature optima, lysis of
cells in the presence of complement, and the existence of
complete and incomplete forms (13).
General Statement
Full advantage has been taken of every available means
to arrive at a satisfactory solution to the problem of
antibody separation. Within the limitations Imposed by
circumstances and the nature of the antibodies themselves,
the major objective of this investigation has been reached.
There has been opportunity to observe the behavior of
antibodies under different conditions, and these
observations are offered as a contribution to the growing
knowledge of blood group antibodies.
The principal drawbacks which have been experienced
have been in the limited quantities of materials with which
to work (high titer sera must be used as starting materials
in all of the methods available at present), the relative
non-speclficlty of available physical chemical methods of
protein fractionation, the low yields of antibodies from
fractionation experiments, and problems in the titration of
antibody preparations and in the interpretation of
titration results. Combination of several methods in
sequence has generally yielded disappointing results
because of the factors stated above.
119
Future investigations in this field will have to face
these same problems. Opportunities may open up, however,
with respect to improvement of the resolution of existing
fractionation methods, development of new methods, methods
for stabilizing antibodies during fractionation, or more
sensitive and reproducible methods of detecting and
quantitating antibody activity (such as a precipitin
method).
The real key to both the problem of isolating
antibodies and the problem of preserving their activity
lies in the development of specific reagents. Such
reagents could be prepared either by extracting specific
biological materials or by synthetic methods. The
stroma-resln method is the closest approach to a practical
specific method, and further work on the development of
such a method should be rewarding. The successfulness of
the DEAE-cellulose method suggests that it may eventually
be possible to synthesize materials which would
specifically adsorb antibodies from solution. Synthetic
materials would probably be more stable and reliable than
crude extracts from red cell stromata or other biological
sources. Further work on the nature of the specific groups
of antigens and antibodies would be helpful in determining
what characteristics such synthetic materials should have.
SUMMARY AND CONCLUSIONS
Separation of complete and incomplete blood group
antibodies has been achieved, and observations have been
made on the properties of these antibodies.
Complete antibodies in several sera, both ABO and Rh,
were found to migrate faster than incomplete antibodies in
starch block electrophoresis at pH 8.6 and to sediment
faster in ultracentrifugation at 40,000 rpm. Excellent
separations and interesting distributions of antibody
activity, not previously described in the literature, were
obtained in starch block electrophoresis and DEAE-cellulose
adsorption. In both methods appearance of antibody
activity in albumin fractions was suggestive of specific
protein-protein interactions, as the antibodies were shown
to be gamma globulins. Several other fractionation methods
were studied. Minor differences were observed in the
properties of antibodies from different sera.
Units of antibody activity were defined, and
experimental results were described in terms of these
units. Antibody activity was enriched ten- to twenty-fold
by several of the methods. Enhancement of activity was
observed in some cases, apparently as a result of
separation from blocking antibodies, but overall recovery
was usually less than 50 per cent. Attempts at further
1 20
121
purification of crude fractions were attended by further
losses of activity.
Because of the low yield in these separations,
attention was concentrated on the problem of antibody
stability. On the basis of balance studies and the effects
of adding albumin, it was suggested that the loss of
antibody activity was a direct result of separation from
stabilizing substances in the serum.
Every protein fractionation method available to this
laboratory has been studied with respect to the separation
of antibodies. The few published reports on antibody
separation have been confirmed and considerably extended.
The results, while encouraging, have been limited by the
relatively small amounts of materials available, the
non-specificity of the available fractionation methods, the
heterogeneity and instability of the antibodies, and the
lack of sensitivity and precision in the titration methods.
Y/ith the present methods and limited supplies of suitable
sera, it is not likely that better separation or purity of
antibodies could be achieved.
For future work, the development of more specific
methods of fractionation and more precise methods of
quantitation has been suggested.
BIBLIOGRAPHY
1. Leviticus,17:11.
2. Frazer, J. G., The golden bough, London (1922).
3. Mettler, G. C,, and Mettler, F. A., History of
Medicine, Philadelphia (1947).
4. Sturgis, C. C., Blood Transfusion, history, in Oxford
Medicine, Oxford (1947).
3. Anonymous, Blood transfusion, In Encyclopaedia
Britannica, Chicago (1951).
6. Landsteiner, K., Wien. klin. Wochschr., 14, 1132
(1901).
7. Epstein, A. A., and Ottenberg, R., Proc. N. Y. Fathol.
Soc., 8, 117 (1903).
8. Bernstein, F., Klin. Wochschr., 3, 1495 (1924).
9. Eernstein, F., Abstammungs. u. Vererbungsl., 37, 237
(1925).
10. Boorman, K. S., Dodd, B. E., and Gilbey, B. E., Ann.
Eugenics, 14, 201 (1943).
11. Morgan, W. T. J., and Watkins, W. M., Brit. J. Exptl.
Pathol., 29, 159 (1948).
12. Hirszfeld, L., and Amzel, R., Ann. Inst. Pasteur, 65,
251 (1940).
13* Boyd, W, C., Fundamentals of immunology, 3rd edition,
Hew York (1956).
14. Schiff, F., and Sasaki, H., Klin. Wochschr., 11, 1426
(!932).
15. Filittl-Wurmser, S. et al., Ann. Eugenics, 18, 183
(1954).
16. Boyd, W. C., and Reguera, R. M., J. Immunol., 62, 333
(1949). “
122
17.
18.
19.
20.
21 .
22.
23.
24.
25.
26.
27.
28.
29.
30.
31 .
32.
33.
123
Landsteiner, K., and Levine, P., J. Exptl. Med., 47
757 (1928).
Race, R. R., and Sanger, R., Blood groups in man, 2nd
edition, Oxford (1954).
Sanger, R., and Race, R. R., Nature, 160, 505 (1947).
Ashley Montagu, M. F., An introduction to physical
anthropology, 2nd edition, Springfield (1951).
Landsteiner, K., and Wiener, A. S., Proc. Soc. Exptl.
Biol. Med., 43, 223 (1940).
Levine, P., and Stetson, R. E., J. Am. Med. Assn.,
113, 1 26 (1 939).
Race, R. R., Nature, 153, 771 (1944).
Wiener, A. S., Lancet, 1, 343 (1948).
Dunsford, I., and Bowley, G. C., Techniques in blood
grouping, Edinburgh (1955).
Diamond, L. K., Smith, N. J., and Vaughan, V. C., Ill,
Diseases of the blood, in 'Textbook of pediatrics,
Nelson, W. E., editor, 6th edition, Philadelphia
(1954).
Diamond, L. K., and Denton, R. L., J. Lab. Clin. Med.,
' 30, 821 (1945).
Wiener, A. S., J. Lab. Clin. Med., 30, 662 (1945).
Morton, J. A., and Pickles, M. M., Nature, 159, 779
(1947).
Coombs, R. R. A., Mourant, A. E., and Race. R. R.,
Brit. J. Exptl. Pathol.. 26, 255 (1945).
Jones, A. R., M & R Pediatric Research Conference,
Columbus, Ohio (1954).
Kabat, E. A,, Blood group substances, New York (1956).
Wiener, A. S.. Nappl, R., and Gordon, E. B., Blood, 6,
522 (1951). -----
1 24
34. Zuelzer, W. W., M & R Pediatric Research. Conference,
Columbus, Ohio (1954).
35. Jerne, N. K., Proc. Natl. Acad. Sol., 41, 849 (1955).
36. Haurowltz, F., Chemistry and biology of proteins, New
York (1950).
37. Wiener, A. S., Nappi, R., and Gordon, E. B., Blood, 6,
799 (1951 ).
38. Pauling, L., J. Am. Chem. Soc., 62, 2643 (1940).
39. Burnet, F. M., and Fenner, F., The production of
antibodies, Melbourne (1949).
40. Burnet, F. Ii., Enzyme, antigen and virus, Cambridge
(1956 ).
41. Zinsser, H., Infection and resistance, Hew York (1923).
42. Wiener, A. S., Proc. Soc. Exptl. Biol. Med., 36, 173
(1944).
43. Wiener, A. S., Ann. Allergy, 10, 535 (1952).
44. Ponder, E., Blood, 6, 350 (1951).
45. Flick, J. A., and Villafane, A. L., J. Immunol., 68,
41 (1952).
46. Bordet, J., Traite de l’Immunite, Paris (1920).
47. Marrack, J. R., The chemistry of antigens and
antibodies, London (1934).
48. Sturgeon, P., J. Immunol., 73, 212 (1954).
49. Malone, R. H., and Dunsford, I., Blood, 6, 1135 (1951).
50. Belkin, R. B., and Wiener, A. S., Proc. Soc. Exptl.
Biol. Med., 56, 214 (1944).
51. Diamond, L. K., and Abelson, N. M., J. Clin. Invest.,
24, 122 (1945).
52. Wiener, A. S., and Handman, Ij , Exptl. Med. Surg., 6,
416 (1948). B
53.
54.
55.
56.
57.
58.
59.
60.
61 .
62.
63.
64.
65.
66.
67.
68.
69.
125
Pickles, M. M., Haemolytic disease of the newborn,
Oxford (1949).
Boyd, W. C., J. Exptl. Med., 83, 401 (1946).
Cann, J. R., Brown, R. A., Gajdusek, D. C., Kirkwood,
J. G., and Sturgeon, P., J. Immunol., 66, 137
(1951)
Cann, J. R., Brown, R. A., Kirkwood, J. G., Sturgeon,
P., Clarke, D. W., J. Immunol., 68, 243 (1952).
Wiener, A. S., and Katz, L., J. Immunol., 66, 51
(1951).
Laurell, C.-B., Unpublished observations.
Tullis, J. L., editor, Blood cells and plasma proteins,
New York (1953).
Witebsky, £., J. Lab. Clin. Med., 33, 1353 (1948).
Wltebsky, E., Mohn, J. F., Howies, D. J., and Ward, H.
Pr°c. Soc. Exptl. Biol. Med., 61, 1 (1946).
Campbell, D. H., Sturgeon, P., and Vinograd, J. R.,
Science, 122, 1091 (1955).
Landstelner, K., and Miller, C. P., Jr., J. Exptl.
Med., 42, 853 (1925).
Isllker, H. C., Ann. N. Y. Acad. Sci., 57, 225 (1953).
Isliker, H. C., Personal communication.
Cohn, E. J., Gurd, F. R. N., Surgenor, D. M., Barnes,
B. A., Brown, R. K., Derouaux, G., Gillespie, J.
M., Kahnt, F. W., Lever, W. F., Liu, C. H.,
Mlttelman, D., Mouton, R. F., Schmid, K., and
Uroma, E., J. Am. Chem. Soc., 72, 469 (1950).
Kunkel, H. G., and Slater, R. J., Proc. Soc. Exp.
Biol. Med., 80, 42 (1952).
Mehl, J. W., Personal communication.
Peterson, E. A., and Sober, H. A., J. Am. Chem. Soc.,
78, 751 (1956).
70.
71 .
72.
73.
74.
75.
76.
77.
78.
79.
80.
81 .
82.
83.
84.
85.
86.
126
Sober, H. A., Gutter, F. J., Wyckoff, M. M., and
Peterson, E. A., J. Am. Chem. Soc., 78, 756 (1956).
Fahey, J. L., McCoy, P. F.. and Goulian, M., J. Clin.
Invest., 37, 272 (1958).
Thalhimer, V/,, Proc. Soc. Exptl. Biol. Med., 37, 639
(1938).
Enriquez de Salamanca, F., Jr., Zuazo de Leon, C.,
Gonzalez Alvarez, J., and Otero de la Gandara, J.,
Arch. Med. Exptl. (Spain), 16, 359 (1953).
Oxford Instruments, San Francisco.
Greaves, R. I. K., The preservation of proteins by
drying, London (1946).
Flosdorf, E. W., and Mudd, S., J. Immunol., 29, 389
(1935).
Fruton, J. S., and Simmonds, S., General biochemistry,
New York (1953).
Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and
Randall, R. J., J. Biol. Chem., 193, 265 (1951).
Daughaday, 'i. K., Lowry, 0. H., Rosebrough, N. J., and
Fields, Y , r. S., J. Lab. Clin. Med., 39, 663 (1952).
Folin, 0., and Ciocalteu, V., J. Biol. Chem., 73, 627
(1927).
Scales, F. M., and Harrison, A. P., Ind. Eng. Chem.,
12, 350 (1920).
Natelson, S., Microtechniques of clinical chemistry,
Springfield (1957).
Longsworth, L. G., and Maclnnes, D. A., Chem. Rev.,
24, 271 (1939).
Longsworth, L. G., and Maclnnes, D. A., Chem. Rev.,
30, 323 (1942).
Williams, F. G., Jr., Pickels, E. G., and Durrum, E.
L., Science. 121, 829 (1955).
Smithies, 0., Biochem. J., 61, 629 (1955).
87. Poulik, M. D., and Smithies, 0., Blochem. J., 68, 636
(1958).
88. Technical manual, ultracentrifuge model L, Beckman
Instruments, Inc., Palo Alto, California (1952).
89. Deutsch, H. F., Alherty, R. A., Gosting, L. J., and
Williams, J. W., J. Immunol.. 56, 183 (1947).
90. Payne, R., and Deming, Q. B., J. Immunol., 73, 81
(1954).
91. Wallenius, G., Trautman, xR., Kunkel, H. G., and
Franklin, E. C., J. Biol. Chem., 225, 253 (1957).
92. Abelson, 11. M., and Rawson, A. J., J. Immunol., 83,
49 (1959).
93. Speer, R. J., Prager, K. D., Kelley, T. F., and Kill,
J. k., J. Lab. Clin. Red., 54, 685 (1959).
94. Plllemer, L., Oncley, J. L., Melin, M., Elliott, J.,
and Hutchinson, M. C., J. Clin. Invest., 23, 550
(1944). "
95. Stelos, P., and Taliaferro, K., Anal. Chem., 31,
845 (1959).
96. Christenson, W. H., Dacie, J. V., and Croucher, B. E.
Brit. J. Haematol., 3, 262 (1957).
97. Kuhns, W. J., J. Exptl. bed., 99, 577 (1954).
98. Wall, R. L., Jayne, k., and Beauchamp, J., J. Lab.
Clin. Med., 36, 378 (1950).
APPENDIX - METHODS
A thorough understanding of the methods employed in
the present investigation is essential to the appreciation
of the results obtained. Detailed descriptions may be
helpful to those who are unacquainted with the methods.
Descriptions of the methods are also Indispensable to
anyone who wishes to repeat or extend aspects of this work.
Because of the large number of methods used, however, only
general descriptions could be given in the text. It is for
these reasons that method descriptions are collected in
this appendix. The sequence followed is the same as in the
Methods section.
serological Titrations
Reagents.
Buffered isotonic saline solution. 0.68 g.
NaH2P04.H20, 3.16 g. Na2HP04 (or 7.96 g. NaqKPO^•12HpO),
and 32.4 g. NaCl per 3.6 liters.
Red blood cells of appropriate type.
Anti-human globulin serum ("Coombs serum").
Procedure.
1. Prepare a series of tubes, 10 x 75 mm., numbered 1,
2, 3, etc.
2. Pipette 0.3 ml. of serum into each of tubes 1 and
2, and pipette 0.3 ml. of saline into every tube except 1.
3. Mix the contents of tube 2 and transfer 0.3 ml. of
this mixture to tube 3.
128
1 29
4. Repeat step 3 for tubes 4, 5, 6, etc., using a
clean pipette each time. Discard the 0.3 ml. from the last
tube.
5. Wash a sufficient volume of red cells by mixing
with saline solution, centrifuging at 2300 rpm for 2
minutes, and discarding the saline. Wash three times.
6. Prepare a 2% cell suspension by adding 5 ml. of
saline solution for each 0.1 ml. of packed cells.
7. Add 0.3 ml. of this cell suspension to each serum
dilution.
8. Mix the contents of the tubes and incubate in a
water bath at 37° (20° for ABO antibodies) for two hours.
9. Centrifuge for one minute at 1000 rpm.
10. Read each tube for agglutination. Hold the tube
horizontally and rock it back and forth until the cells run
along the side of the tube. Viewing the tube by strong
light reflected from a concave mirror, record the degree of
agglutination as follows;
++++ Complete agglutination; very large clumps,
few free cells
+++ Large clumps, some free cells
++ Marked granular appearance
+ Slight granularity (microscopically, a
fair number of clumps of 10-20 cells each
± Almost Imperceptible granularity
(microscopically, occasional clumps of
about 5 cells each)
- No agglutination; cloudy appearance, with
swirls as the tube is rocked.
11._Fill the tubes with saline solution, mix well by
inversion, using the index finger to stop the tubes and
wiping it each time, and centrifuge at 2500 rpm for two
minutes. Decant and discard the supernatant. Wash the
cells three times in this manner.
12. For each tube, place a drop of antiglobulin serum
130
on a glass plate.
13. Mix the contents of the tube by shaking and pour
the drop of suspended cells Into the drop of antiglobulin
serum. Mix and spread the drop to 1 cm. diameter by
stirring with the lip of the tube.
14. After 5 minutes, scrape the cells from the plate
with a glass rod, wiping the rod on a towel after each
time.
15. Rock the plate slowly 20 times around an imaginary
pivot at its center.
16. Viewing the plate by transmitted light, record the
degree of clumping, using the notation given in step 10.
17. The results obtained in step 10 indicate the titer
of complete antibody. The results in step 16 indicate the
titer of Incomplete antibody. The titer is given by the
last tube which shows a + or stronger agglutination:
C D
O
•
Titer
1 1
2 2
5
4
4
8
5
16
6 32
7
64
8 128, ■
Anticoagulant Solutions
Alsever1s solution. (2.05 g. glucose, 0.8 g.
trisodium citrate, 0.42 g. sodium chloride, made up to 100
ml. with water and adjusted to pH 6.1 with citric acid).
Cells stored in Alsever1s solution (30 ml. blood per
20 ml. Alsever’s) will last 4 to 6 weeks (98).
A._C.D. (Acld-cltrate-dextrose) solution. (26.7 g.
trisolium citrate*5H2O, 8.0 g. citric acid, 22 g. dextrose,
made up to 1 liter with water).
Cells stored in A.C.D. solution (450 ml. blood per 75
ml. A.C.D.) will last 4 weeks (25).
131
Starch Block Electrophoresis
Reagents.
Barbital buffer, pH 8.6, 0.05 M. 1.84 g. barbital and
10.3 g. sodium barbital per liter.
Dye solution. 0.1-1$ bromphenol blue in ethanol.
Fixing solution. 5.4 g. sodium acetate and 100 ml.
glacial acetic acid per liter.
Procedure.
1. Fill a gallon jar about half full of potato starch,
add enough buffer to cover the starch, and stir. Add more
buffer, as required, to make a thick paste.
2. Allow the starch to settle, then decant the excess
buffer.
3. Prepare a mold by laying a plastic sheet (65 x 50
cm.) over a glass plate (25 x 50 cm.), placing wooden
blocks (50 x 2 x 1 cm.) along the edges and shorter blocks
at the ends, and clamping the edge blocks to the bottom
plate. Place paper towels between the ends of the plate
and the end blocks.
4. Stir the starch paste to a homogeneous mass and
pour it onto the plate. Lumps of starch remaining in the
jar may be scraped out with the fingers.
5. Spread the starch evenly over the plate, smoothing
out lumps with the fingers, and let it dry until the gloss
disappears from the surface.
6. Blot the starch with paper towels until it no
longer tends to flow.
7. With a spatula, scoop out a space 1 cm. wide across
the block, 20 cm. from the cathode end.
8. Kix 10 ml. of serum with enough dry starch to make
a paste of about the same consistency as the starch block.
9. Carefully put this paste into the space in the
block, then press it firmly so that it becomes continuous
with the block.
132
10. Remove the wooden blocks from the ends of the plate
and unclamp the edge blocks, Wrap the loose ends of the
plastic sheet over the starch block, taking care not to
trap bubbles of air, and again clamp the edge blocks to the
plate.
11. Fill the electrode vessels with buffer, and space
them about 50 cm. apart in the cold room.
12. Rest the plate with the starch block on the edges
of the electrode vessels, and connect the starch block with
the buffer in the vessels by means of cloth towels soaked
in buffer. Press the towels into the ends of the starch
block to insure good contact.
13. Turn on the power supply, and set it at 300-400
volts for overnight operation.
14. After about 20 hours, turn off the power supply.
The buffer in the electrode vessels may be saved and used
again.
15. Peel the plastic sheet from the top of the starch
block and immediately press a sheet of filter paper (50 x
20 cm.) onto the block.
16. After the paper is moistened by buffer from the
block, run the paper through the dye solution for about 5
minutes, wash several times in 5^ acetic acid, immerse in
fixing solution for a few minutes, blot with paper towels,
and dry in the oven for 30 minutes at 100-125°.
17. After the starch block has dried for two or three
hours in the cold room, transfer it to a wooden cutting
block.
18. Using the dyed filter paper as a guide, cut out
with a knife the desired parts of the starch block (e. g.,
ten 5 cm. sections in the present work).
19. Put the starch in a sintered glass filter, add
one-tenth volume of isotonic saline solution, and mix well
with a stirring rod.
20. Apply suction until the starch is packed on the
filter, then slowly add a volume of saline solution equal
to the volume of the starch.
21. Release suction when the flow of eluate stops.
133
Anion-Exchange Cellulose Chromatography, Column Method
Apparatus. After Fahey and coworkers (71).
Reagents.
Phosphate buffer, pH 8.0, 0.01 M. 0.068 g, KH2P04 and
3.4 6* NagHPO^j. per liter.
Procedure.
1. Wash 16 g. DEAE-cellulose with the following
sequence of solutions: 0.3 M Nal-IgPC^., 0.5 H NaOH, water,
acidified ethanol, 0.5 M IT a OH, water, and 0.01 M pH 8
phosphate buffer.
2. Pour a suspension of the cellulose in the buffer
into a chromatographic column 2.5 cm. in diameter, and let
it settle by gravity or under nitrogen pressure.
3* Add 20 ml. of serum to the top of the column and
wash it into the cellulose with buffer. For best results,
the serum should be dialyzed against buffer before it is
added to the column.
4. Fill the reservoir with 1 liter of 0.3 M HaH2P04,
and the mixing chamber with 1 liter of buffer. Place a
magnetic stirrer in the mixing chamber. Then connect these
flasks to the column.
5. Start the fraction collector. It should change at
intervals of about 15 minutes for a column of this size,
although the exact interval should be adjusted to the flow
rate.
6. After all the fractions have been collected, repeat
the washing in step 1 before reusing the cellulose.
7. The protein concentrations of the fractions may be
determined either by the Lowry method or by measurement of
light absorption at 280 millimicrons.
DEAE-Cellulose Adsorption. Batch Method
Reagents.
Buffer, pH 8.0, 0.1 M. 0.68 g. KH0PO4 and 34.0 g.
Na2HP04 (or 85.7 g. NagHPO^.* 12H2O) per liter. (Dilute 10 X
134
to make 0.01 M working buffer).
Buffer, pH 6.5, 0.04M. 3.81 g. KH2PO4 and 4.30 g.
NagHP04 per liter.
Buffer, pH 5.8, 0.1 M. 12.2 g. KH2PO4, 3.58 g.
Na2HP04 per liter.
Buffer, pH 4.5, 0.3 M. 40.8 g. KH2PO4 per liter.
Procedure.
1. Dialvze 1 ml. serum overnight against 500 ml. 0.01
M pH 8.0 buffer.
2. Mix serum with 1 ml. packed DEAE-cellulose, which
has been previously washed with 0.01 l - I pH 8.0 phosphate
buffer.
3. Add 0.01 K pH 8.0 buffer to make a total volume of
11 ml. and mix.
4. Centrifuge. Decant supernatant and label "8.0".
5. Repeat steps 3 and 4 with the following buffers:
0.04 K pH 6.5, 0.1 i - i pH 5.5, and 0.3 H pH 4.5. Label the
supernatant fractions "6.5", "5.8", and "4.5", respectively.
6. Regenerate the cellulose as described in the
section Anion-Exchange Cellulose Chromatography.
Solvent Fractionation of Proteins (Cohn Method 10)
Reagents.
Acetate buffer, 0.1 M, pH 4. 8.2 g. sodium acetate,
32 ml. glacial acetic acid, diluted to Too ml.
Solution A. 25 ml. 95/£ ethanol, 2 ml. acetate buffer,
diluted to 100 ml.
Solution B. 15 ml. 95^ ethanol, 0.2 ml. 0.1 M sodium
acetate, 0.14 ml. 1 M glacial acetic acid, 4.5 g. glycine,
diluted to 100 ml.
Cooling mixture, -5°. 16# glycerol by volume.
135
Procedure.
1. Cool 1 ml. serum to 0° In a centrifuge tube
immersed in the cooling mixture.
2. Add 4 ml. Solution A, one drop per second, with
stirring. Continue stirring for several minutes.
3. Centrifuge 10 minutes at 2000 rpm. Decant
supernatant and label "A".
4. Stir precipitate into a smooth paste, then stir
with 2 ml. Solution B.
5. Centrifuge 10 minutes as before. Decant supernatant
and label , ! B".
6. Dissolve remaining precipitate in 2 ml. isotonic
saline solution and label "G-".
Stroma-Resln Adsorption
Procedure.
1. Lyse 100 ml. packed red blood cells of appropriate
type with water.
2. Centrifuge, and wash the sedimented stroma (about
3 ml.) several times with water.
3. Suspend the stroma in 20 ml. of 0.05/o sodium
tetraborate solution, and add 0.05 ml. of formalin.
4. I-lix with 30 g. of oven-dried Ainberllte IRA-401
(XS-75) resin.
5. Wash the resulting stroma-resin with saline
solution. It may be stored until use either in saline
solution or dry. Wash again in fresh isotonic saline
solution before use.
6. I-lix equal volumes of stroma-resin and serum, and
leave in the refrigerator for 1 hour.
7. Centrifuge, decant the serum, and wash the
stroma-resin three times with 5 ml. portions of cold saline
solution.
136
8. Add to the stroma-resin 5 ml. of a solution 0,5 M
in galactose and 1.5 II in NaCI, and incubate at 43° for 60
minutes.
9. The supernatant solution contains the antibodies
and may be dialyzed to remove the galactose and salt. The
stroma-resin may be regenerated by washing with 0.1 N
acetic acid.
Pervaporatlon
Procedure.
1. Gut a length of dialysis tubing. Lengths of 19 mm.
diameter tubing are required as follows:
Volume of protein solution Length of tubing
100 ml. 19 in.
50 ml. 1 2 in.
10 ml. 7 in.
2. ioak the tubing for a few seconds in distilled
water.
3. Tie two knots at one end of the tubing, with a
marble inserted between knots, being careful not to stretch
the tubing (excessive stretching causes leaks).
4. With a funnel, fill the bag with protein solution.
5. Tie a knot in the other end, and attach a hook and
identification tag.
6. Hang the bag on a rack in front of a large fan in
the cold room. Pervaporatlon may also be carried out at
room temperature.
7. Place a tray or beaker beneath the bag, and add
saline solution until the lower end of the bag is Immersed.
Replace with fresh saline every day or two.
8. When the volume of the protein solution has
decreased sufficiently, after about a week of continuous
pervaporatlon, mix the remaining protein solution by gently
squeezing the bag, cut off the bag just above the surface
of the solution, and empty the contents into a test tube.
137
Concentration by Dialysis
Procedure.
1. Prepare a 25^ solution of gum acacia In saline
solution, heating and stirring to dissolve the gum, and add
0.1 g. merthiolate per liter. Cool overnight In the cold
room, then pour Into the dialysis vessel.
2. Insert a marble in the end of a 20-inch length of
wet dialysis tubing, and tie knots in the tubing on both
sides of the marbles, taking care not to stretch the
tubing.
3. Pour the protein solution into the other end of
the tubing, tie a knot, and hang the bag in the gum acacia
solution. Agitate continuously or at intervals.
4. V.'hen the protein solution has been concentrated
sufficiently, usually after two or three days, take the bag
from the gum acacia solution, wash it with water, and
transfer the protein solution to a test tube.
L.y ophxl 1 zat ion
Apparatus. Gampbell-Pressman type.
Procedure.
1. Fill a 10 liter Dewar flask with dry ice-isopropvl
alcohol mixture (about 1:1 to 1:2).
2. net up the lyophile apparatus so that the water
trap is completely immersed in the mixture In the Dewar
flask. G-rease all ground-glass connections with vacuum
grease.
3. Place the protein solution in a lyophile flask,
and rotate and shake the flask in a freezing mixture (dry
ice-isopropanol in a large beaker) until the__ solution is"
shell-frozen on the walls of the flask. "
4. Turn on the vacuum pump, refreeze the flask in the
freezing mixture, and Immediately attach it to the lyophile
manifold.
5. Let the pump run until the proteins are dry, slowly
break the vacuum, then turn off the pump.
138
Saltlnp; Out of Proteins
Procedure.
1. Add 40 grams of ammonium sulfate to 100 ml. of
eluate from a starch block or other dilute protein solution
(to make 3 M in ammonium sulfate).
2. Mix until the salt is dissolved.
3. Centrifuge at 2000 rpm for 10 minutes.
4. Decant and discard the supernatant solution
(contains albumin).
5. Suspend the precipitate in 5 ml. water and dialyze
overnight against isotonic saline solution.
6. Remove, by centrifugation, and discard any
precipitate which remains undissolved.
Protein Determination, Lowry
Reagents.
Reagent I. 20 g. RapCCjjj, 4 g. VaGH, and 0.5 g. sodium
potassium tartrate per liter.
Reagent II. 1 g. Cu;J04»5Rp0 per liter.
Reagent III. 9 parts of Reagent I to 1 part of
Reagent II, mixed just before use.
Reagent IV. Folln-Clocalteu phenol reagent, diluted
with water to make a solution 1 N In acid.
Procedure.
1. Use an aliquot containing about 100 micrograms of
protein.
2. Add 5 ml. Reagent III, let stand 10 minutes or
longer. A purple color indicates too much protein is
present, and a smaller aliquot must be taken.
3. Mix quickly with 0.5 ml. Reagent IV.
4. Read with green filter after 30 minutes.
139
Nitrogen Determination, K1 cro-K. 1 e 1 dahl
Apparatus. Standard, commercially available
equipment, with separate steam generator, a jacketed still
unit, and a silver tube condenser.
Reagents.
Nitrogen standard. 0.2 mg. N per ml.
Boric acid trapping solution, M-%, with bromcresol
green and methyl red. mixed indicator.
Standard hydrochloric acid, 0.0025 N.
Procedure.
1. Set heating block on medium (500-400° C.) at least
50 minutes before beginning digestion.
2. Place an alundum chip in a 150 x 18 mm. Pyrex tube
(rim of tube should be silicon-coated for easy pouring).
Add a sample containing about 0.2 rug. of nitrogen, then add
1 ml. concentrated IlgSOA.
5. Place the tube in the heating block.
4. After 50 minutes, remove the tube from the block
and let it cool for 15 minutes.
5. Add 2 drops of J>0% H2°2> rnix, and replace the tube
in the block.
6. After an additional 50 minutes remove the tube from
the block and let it cool for at least 15 minutes.
7. On the distillation apparatus, apply clamps at the
filling funnel and steam trap drain, and place a 125 ml.
Erlenmeyerfrask under the condenser.
8. Light the burner and place it under the steam
generating flask. Turn on the water supply to the
condenser.
9. Allow steam to pass through the apparatus for about
50 minutes before the first sample is added.
10. Remove the burner from under the steam generating
flask.
140
11. After the water in the distilling flask has been
sucked into the steam trap, remove both clamps.
12. After the steam trap has emptied, place a clamp on
the tubing between the steam trap and the distilling flask,
and replace the burner under the steam generating flask.
13. Position a 125 ml. Erlenmeye flask, containing 20
ml. of boric acid trapping solution, so that the tip of the
condenser dips below the surface of the acid.
14. Pour the digest into the filling funnel.
15. Rinse the digestion tube with four 2 ml. portions
of water, and pour the rinsings into the funnel.
16. Slowly add 5 ml. of 50$ NaGH to the funnel.
17. Rinse the sides of the funnel with about 0.5 ml. of
water.
18. Place a clamp at the filling funnel, and move the
second clamp to the steam trap drain.
19. Collect 50 ml. of distillate in the receiving
flask. Then lower the flask and collect 1 ml. more.
20. Replace the receiving flask with an empty flask.
21. Repeat steps 10 through 20 for each sample.
22. Prepare an end-point standard consisting of 20 ml.
of the boric acid trapping solution diluted to a volume
equal to that of the sample after titration.
23. Titrate the sample with 0.0025 N HC1 to a color
which matches that of the end-point standard.
24. One ml. of 0.0025 N acid is equivalent to 0.035 mg.
of nitrogen.
Nitrogen Determination, Nessler
Reagents.
Stock standard. 4.719 g. ammonium sulfate per
liter, plus 0.5 ml. conc. K2SO4.
141
Working standard. (20 micrograms N/ml.) Dilute stock
standard 50 X.
Digestion mix. 25 mg. HgO, 25 g. K2S04» and ^00 ml.
conc. H2SO4 per liter.
Nessler's reagent. Dissolve 51 g. KI and 16.2 g. HgO
in 200 ml. water, then add 160 ml. 50$ NaOH (500 g./l.) and
dilute to 1 liter with water.
Procedure.
1 . Use an aliquot of protein solution containing about
25 micrograms of nitrogen, or 150 micrograms of protein.
2. Add 0.5 ml. of digestion mix and a boiling chip.
5. Digest 30 minutes at 300-400° C., cool 5 minutes.
4. Add 2 drops 30$ H2C2» digest 30 minutes, cool 5
minutes. Add 4 ml. water.
5. I-iix quickly with 3 ml. Nessler's reagent.
6. Dilute to 10 ml. with water and read with the blue
filter.
Protein Determination, Biuret
Reagent.
Biuret reagent. 60 g. NaOH, 1.5 g. CuSO^ySHgO, and
6.0 g. sodium potassium tartrate per liter.
Procedure.
1 . Use an aliquot containing about 10 milligrams of
protein.
2. Dilute to 3 ml. with water.
3. Add 3 ml. biuret reagent.
4. Read with green filter aft3r 30 minutes.
142
Free Boundary Electrophoresis
Apparatus. Perkin-Elmer, Model 38-A.
Buffer. Barbital buffer, pH 8.6, 0.1 M. 2.797 g.
barbital and 20.6 g. sodium barbital per liter.
Procedure.
1. Adjust the protein concentration of the sample to
about 2 to 5 per cent, and dlalyze it overnight against the
electrophoresis buffer.
2. Add ice cubes to the water bath and turn on the
stirrer. Insert the conductivity cell, filled with 0.01 N
KC1 solution, in the bath; and measure the resistance with
the resistance indicator. Then fill the cell with
electrophoresis buffer, and determine the conductivity of
the buffer.
K = g R
where K is the conductivity cell constant, g is the
specific conductance of 0.01 N KC1 solution at the
temperature of the ice bath (0.000844 at 3°), and R is the
measured resistance with the cell filled with KC1 solution.
3. Lightly coat the flanges of the electrophoresis
cells with silicone grease, assemble the cell, and place it
in the rack, taking care not to touch the optical surfaces.
Attach the electrode vessels containing the electrodes.
4. Fill the bottom section of the cell and one limb
(the descending limb) of the cell with protein solution.
Shift the center section to close off the bottom section,
then rinse the ascending limb with three changes of buffer
and finally fill it with buffer. Then shift the top
section of the cell to close off the center section, and
fill both sides of the top section and the electrode
vessels with buffer. Formation of bubbles on the walls of
the cell must be avoided in this step.
5. Add 10 ml. of 1/3 saturated IiCl solution to each
electrode vessel, letting it flow down Inside the glass
tube.
6. Place the rack, cell, and electrode vessels in the
water bath, and screw down the rack. Allow about 15
minutes for temperature equilibration, then close the gate
in the top section of the cell and connect the electrodes
to the terminals.
143
7. Shift the center section of the cell so that it
communicates with the top and bottom sections, and very
slowly and gently add buffer to the electrode vessel on the
descending side until the boundaries in both limbs are
visible on the ground glass screen when the lamp is turned
on.
8. Set the polarity switch, turn on the current, and
set it at about 10 milllamperes. The positions of the
protein boundaries can be Inspected with the ground glass
screen and photographed at any time during the run.
9. On the photograph, measure the distance of
migration with a centimeter scale. The mobility is
calculated as follows:
u = d K A
I R m t
where u is the mobility (in cm.2/volt/sec.), d is the
apparent distance of migration in cm., K is the conductivity
cell constant, A is the cross-section area of the
electrophoresis cell, I is the current, R is the measured
resistance of the buffer, m is the magnification factor of
the optical system (m=1.06), and t is the time in seconds.
Starch Gel Blectroohoresis
Reagents.
Borate buffer, pH 9*0, 0.2 I - I total borate. 24.7 g.
K3BO3, 6.4 g. HaOII per 2 liters. Dilute 8 X to make 0.025
M working buffer.
Bridge solution. 37.2 g. IlgBOg and 4.8 g. NaOH per 2
liters.
Procedure.
1. Suspend 100 grams of starch in 200 ml. acetone
containing 2 ml. concentrated HC1.
2. Incubate the mixture at 37.5° for 60 minutes
(adjust time for different batches of starch).
3. Add 50 ml. 1 M sodium acetate.
4. Pour the mixture into a sintered glass filter and
1 44
wash with 500 ml. water, applying vacuum to hasten
filtration.
5. Suspend the starch in 1 liter of water and leave
overnight.
6. Wash the starch with 500 ml. water as in step 4.
7. Wash with 100 ml. acetone.
3. Leave the vacuum on until the starch dries on the
filter. This starch is sufficient for several gels.
9. Clamp fiber strips along the edges of a glass plate
(52 x S cm.) and smaller strips at the ends, leaving a
space in the center 50 x 4 cm., and place the plate on a
level support.
10. Llix 50 grams of hydrolyzed starch, obtained in step
8, with 200 ml. of 0.025 M buffer.
11. neat the mixture in a beaker on a hot plate, with
constant stirring, until it boils, thickens, and then thins
slightly.
12. Immediately pour this material into the form
prepared in step 9. It should flow sufficiently to spread
Itself evenly on the plate.
15. Clamp a second glass plate on the top of the form
to flatten the surface of the liquid and to squeeze out any
excess.
14. Leave the entire assembly in the cold until the gel
hardens, about 30 minutes.
15. Unclamp the top plate and gently slide it off so as
not to disturb the gel. At this stage the gel should be
soft, resilient, and slightly cloudy, but not lumpy.
16. With a razor blade cut two 2 cm. slits 0.5 cm.
apart in the center of the gel and perpendicular to its
length. Also, make short cuts to join the ends of the
slits.
17. With a spatula, lift out the part of the gel
between the slits, leaving a slot 2 x 0.5 x 0.5 cm.
18. Lay strips of thin plastic sheet on the surface of
145
the gel all around the slot.
19. Mix equal volumes of serum and dry starch, and
transfer this mixture into the slot, then dust with dry
starch to harden.
20. Pull off the plastic strips to remove any overflow
and gently push the starch-serum mixture to insure good
contact between it and the gel and to make it level with
the surface of the gel.
21. Remove the fiber strips from the end, and cover
the entire surface of the gel, except for 1 cm. at the
ends, with plastic sheet.
22. Fill the electrode vessels with bridge solution,
rest the ends of the gel plate on the edges of the vessels,
and connect the gel and bridge solution by means of cloth
towels soaked in bridge solution.
25. Apply a current of about 10 milliamperes (200
volts) for about 24 hours.
24. Cut away the edge of the gel. Cut a narrow strip,
about 0.5 cm. wide, from the center of the gel and transfer
it to a glass plate.
25. Cut the strip in half lengthwise with a razor
blade held horizontally.
26. Immerse both halves of the strip in solvent
(methanol:water:glacial acetic acid::50:50:10, by volume)
saturated with naphthalene black, and leave overnight. The
remainder of the gel may be discarded, or it may be cut
into sections and frozen for later use.
27. - ‘ /ash the strip in several changes of solvent.
About one week is required to remove background dye from
the gel, which becomes harder, opaque, and white. The
protein bands retain a dark blue color. The dyed gel lasts
indefinitely if kept immersed in solvent.
28. Frozen sections of the gel, after thawing, may be
squeezed out like sponges for recovery of the solutions
they contain.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Galactose metabolism in human blood cells

PDF

Microheterogeneity of fetuin

PDF

Purification and identification of trypsin inhibitors in human serum

PDF

Effects of sex hormones on cholesterol metabolism

PDF

Studies on the transport, metabolism and chemistry of iron-sugar chelates

PDF

Studies of trypsin-binding ɑ₂ macroglobulin of human plasma

PDF

Studies of steroid 11β-hydroxylation and cholesterol side chain cleavage in adrenal cortex mitochondria

PDF

Studies of phosphorylated metabolic intermediates

PDF

Interaction between 11β-hydroxylation and respiration In bovine adrenocortical mitochondria

PDF

Steric effects in imidazole and ɑ-chymotrypsin catalyzed ester hydrolysis

PDF

Studies on the biosynthesis of glycoprotein in reticulocyte stroma

PDF

Kinetic investigation of the hydrolysis of aryl β-D-glucopyranosiduronic acids by β-glucuronidase

PDF

Metabolism of 5-hydroxyuracil compounds

PDF

The Biosynthesis Of Ergothioneine

PDF

Studies on acyl transfer reactions on n-acylimidazoles

PDF

Structure and biochemical function of the photosynthetic apparatus

PDF

Steric effects in the acid, amine, and glyceraldehyde-3-phosphate dehydrogenase catalyzed hydrolysis of acyl phosphates

PDF

Substrate and inhibitor interactions of glyceraldehyde 3-phosphate dehydrogenase and intramolecular general base-catalyzed alcoholysis of amides and esters

PDF

Studies On The Metabolism Of Iron

PDF

Energy linked functions and steroid hydroxylation of bovine adrenocortical mitochondria

Asset Metadata

Creator Meacham, Edwin James (author)

Core Title The separation of complete and incomplete blood group antibodies

School Graduate School

Degree Doctor of Philosophy

Degree Program Biochemistry

Degree Conferral Date 1960-06

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag chemistry, biochemistry,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Bergren, William R. (committee chair), Mehl, John W. (committee member), Sturgeon, Phillip (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c18-81558

Unique identifier UC11357996

Identifier 6005483.pdf (filename),usctheses-c18-81558 (legacy record id)

Legacy Identifier 6005483.pdf

Dmrecord 81558

Document Type Dissertation

Rights Meacham, Edwin James

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA