THE JOURNAL OF BIOLOGICAL CHEMISTRV

Vol. 234, No. 10, October 1959

Printedin U.S.A.

Human Gamma Globulin Fractionation on

Anion Exchange Cellulose Columns*
JOHN L. FAHEY AND ANN P. HORBETT

From the Metabolism Service, General Medicine Branch, National Cancer Institute, National Institutes of Health, Public
Health Service, Department of Health, Education and Welfare, Bethesda, Maryland

(Received for publication, May 25, 1959)

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
The serum proteins designated as y-globulins include all no history or evidence of hepatic disease, contributed the sera
proteins of the slowest moving of the several major protein utilized in these studies. Each serum was normal by analytical
groups seen on serum electrophoresis conducted at alkaline pH electrophoresis and ultracentrifugation.
(2). Among the y-globulins of normal serum are proteins of Protein and hexose determinations and analytical paper
differing electrophoretic mobility (3,4), proteins with sedimenta- electrophoresis were performed on whole serum and protein
tion coefficients of 6.6 S and 18 S in the ultracentrifuge (2, 3, 5, fractions as described previously (21). When necessary, dilute
6)) many individual antibodies (5, 7)) and the isohemagglutinins y-globulin solutions were dialyzed at 5 against phosphate-
(8). In disease states, within the -y-globulins are also found the buffered saline (1 volume of 0.15 M sodium phosphate pH 8.0:9
relatively specific serum properties identified in rheumatoid volumes of 0.14 M NaCl) .
arthritis (9), Hashimotos thyroiditis (10, II), lupus erythema- Analytical ultracentrifugation was performed in a Spinco
tosus (12)) cold hemagglutination syndrome (13, 14)) multiple model E ultracentrifuge at 59,780 r.p.m. Samples were dialyzed
myeloma (15, 16)) macroglobulinemia (17) and cryoglobulinemia thoroughly against 0.14 M NaCl before analysis. All experi-
(1%. mentally observed sedimentation coefficients were corrected to
Study of the -y-globulins has been handicapped by the poor a water basis at 20 (SZO.w) and are expressed as Svedberg units
resolution of fractionation procedures and the amount of material of sedimentation, S = lo-l3 cm per second per unit field of force.
or effort required to carry out each separation. The introduc- Antisera to normal y-globulin were produced by subcutaneous
tion by Peterson and Sober (19) of substituted cellulose ion or intramuscular injection of rabbits with 5 mg of whole y-
exchangers with a high protein-binding capacity offered an globulin obtained by preparative electrophoresis, as described
opportunity to fractionate and characterize the y-globulins. below, plus Freunds adjuvant and a similar injection 2 weeks
Previous studies utilizing whole serum have demonstrated the later. After another 2 weeks, removal of 40 ml of blood by
capacity of anion exchange cellulose chromatography to separate cardiac puncture was followed by intramuscular injection of 5
the y-globulins into a number of regions of differing electro- mg of y-globulin. Similar blood removals and booster immuni-
phoretic mobility (20, 21). However, with whole serum as a zations were performed at weekly intervals until each rabbit
starting material, many of the chromatogram fractions contain- was killed. All antisera samples were tested by gel-diffusion
ing y-globulins also included large amounts of other serum techniques (22) against y-globulin and whole serum to deter-
proteins. mine qualitative antibody composition and by precipitin proce-
In the present work the y-globulins were first separated from dures (23) to determine potency.
whole serum by electrophoretic techniques and subsequently Gel-diffusion (Ouchterlony) plates were prepared as described
subfractionated by anion exchange cellulose chromatography. by OConnor (24). Samples, 0.1 ml, were applied to each cup
The y-globulin subfractions thus obtained have been charac- and diffusion allowed to proceed at room temperature or at 4.
terized electrophoretically, ultracentrifugally, immunochemically, When the cups were almost empty, another 0.1 ml of solution
and by measurement of the hexose content. The findings was added. Reaction was usually complete in 48 to 72 hours,
demonstrate that the y-globulins are a heterogeneous group of at which time photographs were obtained.
proteins with differing physicochemical and immunological Preparative Electrophoresis-The y-globulins (also referred to
properties, and that anion exchange cellulose chromatography as whole y-globulins) were separated from serum by electro-
is a useful means of subdividing the y-globulins. phoretic techniques. Zone electrophoresis on polyvinyl blocks
was conducted by procedures modified from those suggested by
EXPERIMENTAL
Miiller-Eberhard and Kunkel (4). A trough, 20 x 6 x 2.5 cm,
Materials and Methods was lined with parafilm and moistened wicks of Telfal placed at
its ends. A mixture of 65 g of polyvinyl particles? and 65 ml
Blood samples were obtained in the morning from fasting
of sodium diethylbarbiturate buffer, pH 8.6 and ionic strength
subjects and the serum separated after several hours at room 0.075, was poured into a trough. Excess fluid was removed from
temperature. Serum was used immediately or frozen for future the block by touching the wicks to a towel. A 5.0 X 0.2-cm
use. Individual normal sera, as well as pools of three normal
sera, were utilized. Ten donors, in current good health and with 1 Telfa gauze, Bauer and Black Company, Englewood, New
Jersey.
* A preliminary report of this study has been published (1). 2 Geon, Goodrich Chemical Company, Akron, Ohio.

2645
 2646 y-Globulin Fractionation Vol. 234, ?So. 10

ZONE (BLOCK) ELECTROPHORESIS the block sections by displacement filtration, with centrifugation
at 300 X g for 4 minutes in small sintered glass funnels after the
Z I I I I I I I I I I I I I I I I I I addition of Verona1 buffer or saline solution. This procedure
0
was repeated once to rinse the polyvinyl medium. The eluents
NORMAL SERUM #I2 from each section were combined, adjusted to a constant volume,
and the protein content estimated by biuret procedures (Fig. 1)
or by determining the optical density at 284 mn in a Beckman
DU spectrophotometer. This wave length was suggested by
Brattsen (3) in order to reduce the optical density due to barbi-
turate buffer while retaining as much as possible of the protein
optical density. Electrophoretic separation of the y-globulins
from serum was also carried out at 5 in a Spinco model CP
continuous flow paper electrophoresis cell utilizing normal serum
diluted in the ratio of 1 volume of serum to 2 volumes of sodium

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
diethylbarbiturate buffer of pH 8.6 and 0.03 ionic strength. The
diluted serum was applied to the curtain at a rate of 0.8 ml per
hour and exposed to a constant current of 35 ma. The effluent
from the curtain was collected from 32 delivery tips into in-
dividual tubes which were changed every 12 or 24 hours by means
z of an automatic fraction collector. The protein content of each
i tube was estimated by determining the optical density at 284 mn.
k O 5 IO I5 20 The y-globulin electrophoretic region was considered to be
z t-1 BLOCK SECTION NUMBER (+I composed of all the proteins migrating less rapidly than the
a
P-lipoproteins, which appeared to be the slowest migrating of the
FIG. 1. Zone (block) electrophoresis of normal human serum.
@-globulins. This protein location was determined by qualita-
Electrophoresis of 1.5 ml of serum was carried out on polyvinyl
particle blocks in barbiturate buffer, pH 8.6, as described in tively ascertaining the lipid distribution in the electrophoretic
Methods. Cuts of the block, 1 cm, were eluted and the protein fractions by applying aliquots of the fractions to paper strips
content determined by biuret procedures. and staining with Oil Red 0 dye (25). y-Globulin pools were
prepared by combining the content of all tubes in the y region
GAMMA GLOBULIN CHROMATOGRAPHIC that did not contain lipid (Fig. 1, Sections 2 to 7), thus most of
FRACTIONATION the region intermediate between the fl- and y-protein peaks was
included within the y-globulin pool. On ultracentrifugal anal-
DEAE-cellulose: 2gm.
ysis, 90 to 95 ye of these pools were proteins with a sedimentation
d Columns: 1.0 X 30.0 cm.
O - ii! coefficient of 6.6 S, and the remainder of the proteins had a
2.0 - Protein: IO 30 mg. 0.30 sedimentation coefficient of 18 S (cf. Fig. 5). No components
E Initial Buffer: 0.02 M Phos., pH 8 /7 -0
a of 9.5 S were seen.
F Final Buffer: 0.30 M Phos.,pH 8 / E
a Anion Exchange Chromatography-Electrophoretically pre-
E / / ;;1 pared y-globulins equilibrated by dialysis with the initial buffer
El / - 0.20
/ 0
were applied to 30 X l.O-cm (internal diameter) columns pre-
g l.O- / pared as described previously (21). The columns contained 2 g
0 / / E of diethylaminoethyl (DEAE) cellulose3 (19) in equilibrium with
z / / - 0.10 z the initial buffer. For most of the work the initial buffer was
0.02 M in respect to phosphate and pH 8. Phosphate buffer
solutions were prepared as the potassium, sodium, or tris(hy-
I I 1 I droxymethyl)aminomethane salts. An elution gradient of
0 20 40 60 80 100 increasing molarity was established, maintaining pH 8, with a
PERCENT OF EFFLUENT total elution volume of 150 ml. Starting buffer, 90 ml, was
FIG. 2. Anion exchange chromatography of normal human placed in a mixing chamber prepared from a flat bottomed
r-globulin. The total volume of effluent (150 ml) was collected cylinder with 4.7-cm internal diameter. This was connected
in approximately 40 equal fractions. by means of a siphon to a 50-ml Erlenmyer flask reservoir
containing 55 ml of 0.30 M phosphate buffer pH 8. The bottoms
slit was made across the 0.6 cm deep block and 1.0 to 2.0 ml of the two containers were level and mixing was assured by a
of serum was delivered into this slit with a syringe and 26 gauge magnetic stirrer (21). The eluting solution was pumped through
needle. The block was allowed to equilibrate at 5 for + hour the column at a rate of 15 I& per hour and effluent fractions of
with the wick ends immersed in electrode chambers filled with 3.3 ml were collected. The optical density of each fraction was
the above buffer. Electrophoretic separation was carried out measured at 280 mn. Recovery of the y-globulins applied to
in a cold room at 5 with a constant current of 20 ma applied for the column was 90 to 100% complete, as determined by optical
about 18 hours, i.e. until the color bands were well separated. density measurements at 280 mn.
After the trough was carefully removed, the excess fluid was
3 DEAE-cellulose, passing between 230 to 325 mesh and con-
evaporated from the block with a warm air stream, and the taining 0.90 meq/g was generously supplied by Drs. E. A. Peterson
bloek was cut into l.O-cm sections. Proteins were eluted from and H. A. Sober.
 October 1959 J. L. Fahey and A. P. Horbett 2647

RESULTS ELUTION DIAGRAM OF

A representative normal y-globulin chromatogram is illustrated GAMMA GLOBULINS
in Fig. 2. y-Globulins were eluted continuously throughout
WHOLE GAMMA GLOBULIN
most of the chromatogram. For purposes of analysis the eluted
y-globulins were grouped into five fractions composing, respec-
tively: 5 to 15% of the elution volume, Fraction 1; 15 to 25%,
Fraction 2; 25 to 45%, Fraction 3; 45 to 58%, Fraction 4; and 58
to 75+$, Fraction 5. However, it should be remembered that
the elution of y-globulins composing the first four groups and
part of the fifth was continuous, and the division into five frac-
tions was a convenient expedient.
On rechromatography, each fraction was eluted in the same
characteristic region of the chromatogram (Fig. 3). Fraction 2
contained a small amount of the preceding fraction, presumably

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
due to spreading of a portion of the first fraction as it passed
through the column. 0 50 100
The chromatographic y-globulin distribution has been found PERCENT OF EFFLUENT
to be dependent on the molarity of the initial buffer. In Fig. 4 FIG. 3. Anion exchange chromatography of normal r-globulins
are illustrated the chromatograms obtained with the same y- separated by preparative electrophoresis and rechromatography
of the fractions obtained. The amount of whole r-globulin ap-
globulin preparation chromatographed with starting (and equili- plied to the column was equal to that present in 1 ml of normal
brating) buffers of 0.020 M, 0.010 M and 0.005 M phosphate serum. The elution system was the same as that used in Fig. 2.
concentration. A marked and progressive shift of -y-globulin For rechromatography, fractions were concentrated to about 2
from the Fraction 1 to Fraction 3 region is seen as the molarity ml and then prepared for chromatography in the usual manner.
is reduced. This change occurred to the same degree whether
ELUTION DIAGRAM FOR
the potassium, sodium, or tris(hydroxymethyl)aminomethane
forms of phosphate buffers, pH 8, were used. Confirmatorv WHOLE GAMMA GLOBULINS
evidence that Fraction
the Fraction
1 contributes
3 of a lower starting
the y-globulin
molarity
obtained in
was obtained by
r
rechromatography. When Fraction 1 was obtained from a
column with an 0.020 M phosphate initial buffer and was subse-
quently rechromatographed with the use of 0.005 M phosphate
initial buffer, much of the protein was eluted in Region 3 (Fig. ci
4, lowermost graph). On comparison of the whole y-globulin 6 0
2 2.0
chromatograms illustrated in Fig. 4, it is seen that with buffers 0
of lower phosphate concentrations the quantity of protein eluted F
in Region 2 was little altered; the protein in Region 3 showed the 2 1.0
greatest increase; whereas Region 4 increased moderately, and 5
the quantity of protein in Region 5 was not altered.
The change of y-globulin distribution from one region of the
chromatogram to another upon lowering the initial buffer
molarity was considered to be due to an increased capacity of the
z 1.0
anion exchange cellulose to bind y-globulins and/or an alteration
in certain of the y-globulins under these conditions. It is if
probable that a specific group of y-globulins adheres to the 4 0
adsorbent at the lower buffer molarity, but will not adhere or 2.0 .RECHROMATOGRAPHY
remain adherent in solutions of higher ionic strength. The shift 0.020 FRACTION I
does not seem to be due to an association of y-globulin molecules 1.0 AT 0.005 M
into larger aggregates. Ultracentrifugal analyses of Region 3,
obtained after chromatography starting with 0.005 M phosphate
0
buffers, were made in the eluting solution, and after dialysis
0
against 0.005 M phosphate buffer pH 8 and, also, after dialysis PERCENT OF EFFLUENT
against 0.14 M NaCl solutions. These analyses revealed only
FIG. 4. Comparison of the effect of variation in the molarity of
molecules with sedimentation coefficients of approximately 6.6 8. the initial buffer on the anion exchange chromatography of normal
The progressive increase in the size of Fraction 3 with serial r-globulins. Buffers of pH 8 and phosphate concentrations of
lowering of initial buffer molarity indicated that a spectrum of 0.020 M, 0.010 M, and 0.005 M were used as the initial buffers. The
remainder of the elution system was the same as in Fig. 2. Ali-
proteins was contained in Fraction 1. A stepwise lowering of quots of the same electrophoretic r-globulin preparation were
buffer molarity could be utilized for further fractionation of this utilized.
group of y-globulins.
Another consideration in selecting the phosphate molarity of of normal y-globulins with decreasing solvent molarity. In
the initial chromatography buffer is the increasing insolubility 0.020 M phosphate buffers, pH 8, approximately 1 to 4% of the
2648 r-Globulin Fractionation Vol. 234, No. 10

y-globulin mobility were seen. These findings are in accord with

ELECTROPHORESIS
ULTRACENTRIFUGATION anion exchange cellulose chromatographic observations utilizing
elution systems that produced a falling pH coincident with
WHOLE SERUM WHOLE increasing ionic strength (20, 21).
--+ GAMMJ GLOBULIN Ultracentrifugal analyses demonstrated that the first 4 frac-
tions were composed entirely of proteins with sedimentation
coefficients of 6.6 S (Fig. 5). The 18 S macroglobulins were
II n present only in Fraction 5. This fraction also contained a small
amount (up to 10%) of 6.6 S protein.
The hexose content of the y-globulin fractions was found to
increase from 1.1 to 2.3% hexose with progression of elution from
2 * A the columns (Table I). Fraction 5, containing the y-macro-
globulins, was found to have a much higher carbohydrate content
than the fractions composed of 6.6 S proteins. The value of

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
5.0% hexose is less than the 6.0% hexose content reported by
Miiller-Eberhard and Kunkel (4) for normal y-macroglobulin
\ prepared by repeated ultracentrifugal precipitation, and probably
6.6s 18s is due to the lower hexose content of the 6.6 S y-globulins also
FIG. 5. Electrophoretic and ultracentrifugal analysis of normal eluted in this chromatogram region. Possible hexose contribu-
r-globulin anion exchange chromatogram fractions. For refer- tion from the DEAE-cellulose was unlikely. Chromatography
ence purposes, the electrophoretic distribution of normal serum conducted in the usual manner except for the omission of protein
components and the ultracentrifugal distribution of normal whole application revealed no detectable hexose in the effluent.
r-globulins are illustrated. The ultracentrifuge patterns were Immunochemical-Electrophoretically prepared y-globulins
obtained from photographs taken 26 minutes after reaching full
speed. when tested by the gel-diffusion technique against rabbit anti-
human y-globulin serum consistently showed 2 precipitin lines
TABLE I (Fig. 6, tipper). One of these lines was straight and the other
Characteristics of chromatographic y-globulin fractions concave toward the y-globulin cup. Korngold and Van Leeuwen
(26) have postulated that a straight precipitin line between
-y-Globulin fractions antigen and antibody cups is indicative of an antigen of about
- -
the same molecular weight as the antibody. Similarly, a precipi-

%% %
1 2 3 4 5
_- -- -- tin line concave toward the antigen cup is indicative of a greater
% % molecular weight for the antigen. Support for this interpreta-
Distribution of the -y-globu- tion is illustrated in Fig. 6 where two purified y-globulin prepara-
lins (7 determinations) tions, containing only normal 6.6 S components in the first
Mean....................
Range .
71.0
(63.2-
79.6:
7.7
(5.1
9.6 i
(kz ,Is(5
13.4 1 5.1 )
-
13.2)
instance, and in the second only 18 S macroglobulins
from the sera of a patient with Waldenstroms
reacted with rabbit antinormal human y-globulin
macroglobulinemia,
prepared

serum on each
Ultracentrifugal analysis side of the normal whole y-globulin preparation. It is seen that
(szo,3 the concave precipitin line of the whole y-globulin demonstrates
6.6 s.. .. . 100 100 100 100 10
a reaction of identity (22) with the similar concave line of the
18 s.................... 90
18 S macroglobulin. Also, the straight precipitin line reacts with
Hexose content (% of pro-
tein content). . .. . . 1.1 1.6 1.9 2.3 5.0 the corresponding line formed by the known 6.6 S y-globulins.
- - - These findings demonstrate that y-globulins possess at least two
antigenic properties, one characteristic of the 18 S macroglobulins,
proteins became insoluble and were precipitated by centrifuga- the other characteristic of the 6.6 S globulins. This is in agree-
tion before application to the column. However, with 0.005 M ment with the observations of Franklin and Kunkel (27).
phosphate buffers, the loss was 5 to 10%. In sera obtained in The immunochemical findings when multiple chromatographic
disease, y-globulin losses on equilibration with 0.005 M phosphate y-globulin subdivisions were tested against rabbit antinormal
solutions, pH 8, may be much greater. For this reason the human y-globulin sera are illustrated in Fig. 6, lower. The
chromatographic fractionation of y-globulins was conducted chromatographic y-globulin Fractions 1 to 4 seemed to have a
with initial buffers of pH 8 and 0.02 M phosphate concentration single antigenic component which was shared in common by these
as the best compromise. fractions. Although not illustrated here, when these fractions
were tested in hexagonally arranged cups about a central anti-
Characteristics of y-Globulin Fractions sera cup, a single straight precipitin line was seen with each
Physicochemical-Electrophoretic characterization of the y- fraction which showed a reaction of identity with the chroma-
globulin fractions revealed a progressive increase in the electro- tographically adjacent -y-globulins. The proteins in Fraction 5,
phoretic mobility of the proteins comprising Fractions 1 to 4 in addition, demonstrated a strong precipitin line that was
(Fig. 5).4 In Fraction 5, however, only proteins of intermediate
somewhat different from those used here, the sequence of y-
4 Moving boundary electrophoresis of normal serum chromato- globulin elution is believed to be similar. The boundary electro-
gram fractions was performed by Sober, Gutter, Wyckoff and phoresis observations are in agreement with the relative mobility
Peterson (20). Although the conditions of chromatography are differences noted on paper electrophoretic analyses here.
October 1959 J. L. Fahey and A. P. Horbett 2649

concave toward the antigen cup. This finding is consistent with

previous ultracentrifugal demonstration of 18 S y-macroglobulins
in this fraction.
These observations clearly demonstrate that normal y-
globulins of differing chromatographic behavior, electrophoretic
mobility, and hexose content share a common antigenic com-
ponent. However, the y-globulins with sedimentation coeffi-
cients of 18 S, which can be separated from almost all of the 6.6 S
y-globulins by anion exchange cellulose chromatography, possess
distinctive antigenic properties which are retained after chroma-
tographic fractionation.

DISCUSSION

This work emphasizes the heterogeneity of the normal y-

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
globulins. The heterogeneity is of two types. The first is the
division between the large and small y-globulins, i.e. those with
sedimentation coefficients of 18 S in the ultracentrifuge and those
of 6.6 S. The second heterogeneity is within the 6.6 S group of
y-globulins, and probably also within the 18 S group.
Relevant to a discussion of y-globulin heterogeneity are
preliminary findings on the chromatographic distribution of a
number of -y-globulin physiological activities (I). In Fraction
1 from appropriate sera were found antibodies to mumps virus,
Histoplasma capsdatum, Salmonella H antigens and thyroglobu-
lin. No activities have been located in Fractions 2, 3, or 4. In
Fraction 5 the rheumatoid factor and the major portion of the
isohemagglutinins A and B, of several Rh antibodies and of
lupus erythematosus antinucleoprotein activity were found.
The large and the small y-globulins are distinguishable in the FIG. 6. Upper: Gel diffusion characterization of r-globulins.
ultracentrifuge, by hexose content, immunochemically, and by Whole normal r-globulins, normal 6.6 S r-globulins (Fraction 1)
both normal and pathological physiological activities. On anion and an 18 S r-macroglobulin preparation (from a patient with
exchange cellulose chromatography, the y-globulins with sedi- Waldenstroms macroglobulinemia) reacted with rabbit antiwhole
normal human r-globulin serum in an agar gel diffusion cup for
mentation coefficients of 18 S can be separated from at least 95 $& 72 hours at room temperature. Lower: Immunochemical charac-
of the 6.6 S y-globulins. Although Deutsch and Morton (28) terization of chromatogram fractions obtained from whole normal
have postulated that the 18 S -y-globulins arc polymers of 6.6 S r-globulin. The effluent from a normal r-globulin chromatogram
-y-globulin units, the behavior on anion exchange cellulose was divided into II pools which were concentrated and added to
consecutive diffusion cups. The opposite cups were filled with
chromatography and the higher hesose content indicate that the rabbit antihuman y-globulin sera, and diffusion was allowed to
18 S y-globulins differ appreciably from the bulk of the 6.6 S proceed 48 hours at 4 before the picture was obtained. A draw-
y-globulins. The extent of heterogeneity within the 18 S y- ing of the precipitin lines is included for clarification.
globulins is uncertain. Several differing physiological properties
are found in this group normally and in disease. Also, the more subunits. One of these subunits, that possessing anti-
homogeneous 18 S y-globulins found in patients with Walden- genicity for the rabbit, might very well be identical in all of the
Stroms macroglobulinemia differ with each instance of the 6.6 S y-globulin molecules. The other subunit or subunits
disease (29). would be of various composition to account for the many differ-
The heterogeneity among the normal 6.6 S y-globulins, the ences observed in hesose content, clectrophorctic mobility, etc.
second type of y-globulin heterogeneity, may bc considered to The concept that the 6.6 S y-globulin molecules in man are
be an example of microheterogeneity (30). Many studies have composed of subunits, some of which have variable composition,
indicated that a spectrum of these y-globulin molecules exists is influenced and supported by recent work of Porter (34) who
in human serum (4, 5, 7, 20, 21, 31, 32). Each molecule or has enzymatically divided rabbit y-globulin into subunits of
group of molecules must have distinctive and differing charac- differing properties. Pauling (35) had earlier postulated a com-
teristics to account for the variations found in electrophoretic mon portion and a variable portion for y-globulin molecules to
and chromatographic behavior, hexose content, and antibody account for the many specific antibodies within the y-globulins.
distribution. Despite these differences, all of the 6.6 S y-globu- Currently available evidence would indicate that the 6.6 S
lin molecules apparently have similar antigenic components. y-globulin molecules vary in physicochemical properties as well
The observations reported here, and studies by most investi- as in antibody activity.
gators, indicate that the 6.6 S y-globulins represent a single The concept that the normal y-globulins comprise a large
antigenic immunochemical entity (33) when tested with rabbit number of differing molecules is particularly relevant to any
antisera. consideration of y-globulin metabolism and of the changes
The finding of both similarities and differences among the 6.6 S occurring in disease. It is conceivable that these proteins are
human -y-globulins would seem to be best understood by regard- formed in individual and differing plasma cells. This has been
ing the 6.6 S y-globulin molecule as being composed of two or suggested by Askonas et al. (36) who found that amino acids
2850 -y-Globulin Fractionation Vol. 234, No. 10

were incorporated into several rabbit -y-globulin fractions at comprising about 5 to 10% of the total y-globulins, were readily
different rates by the spleen, lymph nodes, and bone marrow. distinguished from the smaller (6.6 S) y-globulins by their
It is further possible that different y-globulins, once out of the distinctive ultracentrifugal and immunochemical properties, a
plasma cell, may not have the same metabolic status. If the relatively high hesose content, and characteristic physiological
turnover times of the various y-globulins are found to differ, the activities.
use in metabolic studies of y-globulin pools without subdivision 4. The y-globulins with sedimentation coefficients of 6.6 S,
might miss significant variations, either normally or in disease. that normally comprise 90 to 95yo of the y-globulins, were
It will be of particular interest to determine if the chroma- distinguished by an anion exchange chromatographic distribu-
tographically discernible y-globulin subgroups arc similarly tion indicating the existence of a spectrum of molecules. The
or differently affected by pathological states such as hepatic differing electrophoretic mobility, chromatographic behavior,
cirrhosis, chronic infections, the collagen diseases, leukcmias hexose content, and distribution of physiological activities of the
and lymphomas, multiple myeloma and macroglobulincmia, and chromatogram fractions support the view that these are a spec-
other disorders in which marked alteration in the total amount trum of y-globulin molecules. The ultracentrifugal and im-
of y-globulin may occur. munochemical findings indicate that these molecules share,

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
Considerable attention has been devoted to the question of however, similar size and antigenic properties. The implications
whether the large amounts of homogeneous serum protein seen of these observations are discussed.
in multiple myeloma are qualitatively or quantitatively ab-
normal. However, the isolated, apparently homogeneous, serum Acknowledgments-The authors wish to thank Drs. H. A.
myeloma proteins have been compared with and shown to differ Sober and E. A. Peterson for consultation and encouragement
from the heterogeneous group of normal y-globulins which may and Miss Patricia Duggan for assistance in the ultracentrifugal
be composed of hundreds or thousands of individual y-globulins. analyses.
Thus the evidence currently available can not be regarded as REFERENCES
conclusively demonstrating that the myeloma proteins are 1. FAI-IEY, J. L., Federation Proc., 18, 43 (1959).
abnormal. It would be unrealistic to insist that the homogene- 2. TISELIUS, A., Biochem. J.. 31, 1464 (1937).
ous y-globulins found in certain diseases should be compared 3. BRATTSE~, I:, A&iv Kern;, 8; 347 (i955):
with isolated individual normal -y-globulins. Nevertheless, the 4. MILLER-EBERHARD, H. J., AND KUNKEL, H. G., J. Exptl.
implications of extensive heterogeneity within the normal y- Med., 104, 253 (1956).
5. DEUTSCH, H. F., ALBERTY, R. A., GOSTISG, 1,. J., AND
globulins does bear upon the problem of relating a single protein WILLIAMS, J. W., J. Immunol., 56, 183 (1947).
found in a disease to the proteins normally present. 6. WALLENIUS, G., TRAUTMAN, R., KUXKEL, H. G., ANU FRANK-
Further subdivision of the normal y-globulins beyond that LIN, E. C., J. Biol. Chem., 225, 253 (1957).
obtained in the present report seems to be feasible. Fraction 1, 7. ENDERS, J. F., J. Clin. Znvest., 23, 510 (1944).
8. PEDERSEN, K. O., Ultracentrifugal studies on serum and serum
for instance, can be further fractionated by rechromatography fractions, Almqvist and Wiksells Boktrgckeri-A.-B.,
on anion exchange ccllulosc columns utilizing a starting buffer Uppsala, 1945.
of lower ionic strength. Refractionation on columns of the 9. FRANKLIN, E. C., HOLMAN, H. R., MUELLER-EBERHARD, H. J.,
cation exchanger carboxymethyl cellulose can be used (37). AND KUNKEI,, H. G., J. Exptl. Med., 105, 425 (1957).
The use of such additional procedures together with anion 10. ROITT, I. M., DONIACH, D., CAMPBELL, P. N., ANU HUNDSON,
R. V., Lancet, 271, 820 (1956).
exchange chromatography should facilitate the detailed study of 11. WITEBSKY, E., ROSE, N. R., TERPLAN, K., PAINE, J. R., AND
y-globulins in normal and disease states. KGAN, R. W., J. Am. Med. Assoc., 164, 1939 (1957).
12. HASERICK, J. It., LEWIS, L. A., AND BORTZ, D. W., Am. J.
SUMMARY Med. Sci., 219, 660 (1950).
13. GORINN, R. S., JR., J. Immunol., 71, 220 (1953).
1. y-Globulin fractionation by means of anion exchange 14. FUDENBERG, H. H., ASD KUNKEL, H. G., J. Exptl. Med., 106,
diethylaminoethyl (DEAE) cellulose chromatography has been 689 (1957).
performed on electrophoretically prepared normal y-globulin 15. KEKW~CK, n. A., Biochem. J., 34, 1248 (1940).
pools. Chromatographic procedures have been modified so that 16. PUTXAM. F. W.. AND UDIN. B.. J. Biol. Chem.. 202. 727 (1953).
17. WALDEN~TRRM,J., Acta Med. &and., 117, 216 (1944).
the y-globulins of as little as 1 mL of serum can be satisfactorily 18. LERNER, A. B., AND WATSON, C. I., Am. J. Med. Sci., 214,
subdivided. 410 (1947).
2. The fractions of y-globulin obtained by anion exchange 19. PETERSON, E. A., ANU SOBER, H. A., J. Am. Chem. Sot., 78.
cellulose chromatography have been considered in 5 groups for 751 (1956).
purposes of analysis and comparison. Fractions 1 to 4, which 20. SOBER, H. A., GUTTER, F. J., WYCKOFF, M. M., AND PETERSON,
E. A., J. Am. Chem. Sot., 78, 756 (1956).
contained only proteins with sedimentation coefficients (~~0, W) of 21. FAHEY, J. L., MCCOY, P. F., AND GOULIAX, M., J. Clin. Invest.,
6.6 S on ultracentrifugal analysis, were composed of y-globulins 37, 272 (1958).
with progressively increasing electrophoretic mobility and in- 22. OUCHTERLONY, o., Acta Pathol. Microbial. Stand., 26, 507
creasing carbohydrate content (from 1.1 to 2.3% hexose). (1949).
23. KABAT, E. A., AND MAYER, M. M., Experimental immuno-
Immunochemical testing revealed Fractions 1 to 4 to have chemistry, Charles C Thomas, Springfield, Illinois, 1948.
antigenic properties in common. Fraction 5, containing all of 24. OCONNOR, G. R., A. M. A. Arch. Ophthalmol., 57, 52 (1957).
the 18 S y-macroglobulins, had the highest hexose content (5.0%). 25. JENCKS, W. P., AND DURRUM, E. L., J. Clin. Invest., 34, 1437
Fraction 5 also possessed distinctive antigenic characteristics, (1955).
compatible with the presence of the 18 S proteins in this fraction. 26. KORNGOLD, L., AND VAN LEEUWEN, G., J. Immunol., 78, 172
(1957).
3. The heterogeneity of the y-globulins was emphasized. The 27. FRANKLIN, E. C., AND KUNKEL, H. G., J. Immunol., 78, 11
-y-globulins with sedimentation coefficients of 18 S, normally (1957).
October 1959 J. L. Fa.hey and A. P. Horbett 2651

28. DEUTSCH, H. F., AND MORTON, J. I., J. Biol. Chem., 231, 1107 32. WILLIAMS, C. A., JR., AND GRABAR, P., J. Immunol., 74, 158
(1958). (1955).
29. PUTNAM, F. W., J. Biol. Chem., 233, 1448 (1958). 33. KABAT, E. A., Am. J. Med., 23, 299 (1957).
34. PORTER, R. R., Nature, 182, 670 (1958).
30. STEINBERG, D., AND MIHAYLI, E., Annual review of biochemis-
35. PAULING, L., J. Am. Chem. Xoc., 62, 2643 (1940).
try, Vol. 26, Annual Reviews, Inc., Palo Alto, California, 36. ASKONAS, B. A., HUMPHREY, J. H., AND PORTER, R. R., Bio-
1957, p. 373. them. J., 63, 412 (1956).
31. MCFADDEN, M. L., AND SMITH, E. L., J. Am. Chem. SOL, 76, 37. SOBER, H. A., AND PETERSON, E. A., Federation Proc., 17,
2784 (1953). 1116 (1958).

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
Human Gamma Globulin Fractionation on Anion Exchange Cellulose Columns
John L. Fahey and Ann P. Horbett
J. Biol. Chem. 1959, 234:2645-2651.

Access the most updated version of this article at

http://www.jbc.org/content/234/10/2645.citation

Alerts:

Downloaded from http://www.jbc.org/ at Universidad Nacional Autonoma de Mexico on August 15, 2017
When this article is cited
When a correction for this article is posted

Click here to choose from all of JBC's e-mail alerts

This article cites 0 references, 0 of which can be accessed free at

http://www.jbc.org/content/234/10/2645.citation.full.html#ref-list-1