- Email: [email protected]
Isolation and description of a few properties of the β2A-globulin of human serum
VOL. 4 ('959)
CLINIC.4 CIIIMICA ACTA
96
ISOLATION
AND
OF THE
DESCRIPTION ,!&A-GLORULIN
J, F. HBRPXANS, Medical Department,
C;niversity
M:TH. Clilzic
OF A FEW OF HUMAN
HEREMASS
St. Pierre,
Marburg/Lahn
PROPERTIES SERUM
AND H. E. SCHl~LTZE
Louvain
(Belgium)
and the Behringwevke,
(Germalzy)
Conventional electrophoretic graphs obtained with normal human sera show, between the well-individualized /3- and y-peaks, a more or less pronounced depression, which is used as a landmark for planimetric work. This also holds for those techniques’ that produce a split P-band. In some pathological conditions, for instance in cirrhosis, the separation of /3 and y may become completely indistinct. For such sera at least, the existence of proteins with intermediate mobility may be assumed. Moreover, many cases of myeloma and macroglobulinemia Waldenstrom are characterized by the appearance of large abnormal protein peaks in this region. The substances responsible for these peaks have been given various designations, such as fi2, yl, 111 or 5‘. Recently our knowledge of the nature of these pZ- (or yl-) proteins has been increased by some more precise information. Thus, GRABAR et ak2, using the immunoelectrophoretic technique, has established beyond doubt that the y-globulins really consist of a whole family of components with identical antigenic specificity. The most rapid members reach the slower (w2region, and the slowest molecules extend to the cathodic end of the graph. From this it follows that the so-called /32(or ‘12, or y,, or 5) (using the word in its immunological region at least contains some “y-globulin” meaning). Eext, KUNIWL'S group3 succeeded in preparing a high-molecular weight component, whose mobility slightly exceeds that corresponding to the classical y-peak. Although the final proof has not yet been given, it may be assumed with reasonable certainty that KWKEL'S heavy y,-globulin at least partially coincides with an antigenie constituent, which BURTIN and co-workers4 called “/32M”. The latter protein appears to be appreciably increased in cirrhotic sera, as is KUXKEL'S heavy y,-globulin, and it has been shown to be specifically involved in macroglobulinemia Waldenstrom 4. Finally, a third globulin with characteristic b2 mobility has been known since the very first immuno-electrophoretic investigations of human sera, performed by GRABAI~ AND WILLIAMS in Igj3. It was merely called “fi,-globulin” by them, until the discovery of the “/3,-macroglobulin” precipitation line prompted them to change the name to “P,,+-globulin”, in order to avoid confusion. This is the protein that will be considered in the present publication. ISOLATIOS
All our attempts to isolate the /3,,-globulin from human plasma or serum have been based upon the hypothesis that it is a substance with high resistance to precipitation with zinc ions. This idea originated from a consideration of the following wellRefeuerlct~s p.IO2
4 (Igjg)
VOL.
known
P,,-GLORULIN
facts.
First,
it has been shown
loma sera show an enormously this protein majority
seems
“paraprotein” tested
by
all of
B-type,
postulated fraction
#?-myelomas
KUNKEL’S
zinc
show
We
values
a fact which the supernatant
still
contained
also remained action
around
peak”.
The
seem to derive
their
towards
flocculation
a few
this
of the y-myereactions,
myelomas,
zinc
and the bZA nature
when
all or almost
ion precipitation”. of the abnormal
by the analysis
We protein
of a large number
by increasing XII.
from
obtained
all of
(after
a conventional but
to 50 mJ!l; concentration
addition),
were
large
found
of
the selectivity
this was obtained of
by
100 mmolesl,
low ionic strength to result
test on
quantities
was made to increase
a zinc sulfate
+29”,
KUSKEL
that
its concentration zinc sulfate
to
from
/3,,-globulin,
An attempt
Finally
+25”
its
PH
and tempera-
in practically
complete
of /3*,,- and y-globulins.
The
zinc ions were
Preliminary supernatant
of normal
p,-globulins,
removed
by adding
that 1.8 flf ammonium
serum.
was submitted
of a2- and its
subsequently
trials showed
the p,,-globulin
method
but
was confirmed
in solution.
neutrality
ranging
into
disturbed test,
stability
this stability
that
separation
“paraprotein
however,
P-mye-
in some cases
sera (to be published).
COHS’S Method
tures
myelomas,
turbidity
found
of the zinc
many
5. On the other hand the vast majority
serum
y-globulin
that
line. Indeed,
of the so-called
show extremely
sulfate
that
corresponded,
of myelomatous normal
the whole
an unusual
therefore
,5,,-precipitation
of the p-type
from the y-antigen
and some
using
to constitute
97
by immuno-electrophoresis
increased
of the y-, and many
lomas,
OF HU>IA?; SERUN
Therefore
the above
to this salting-out
of which
about
electrophoretic
components
macromolecular
/3ZM-globulin of normal
completely
described
procedure,
7596 was
by
I to 2 gy6 glycine.
sulfate
which
pZA. The
KUNKEL‘S
precipitated
glycine-containing yielded
mixture
starch
block
a mixture
was
resolved
electrophoresis
‘.
The
contaminant
of our preparations,
of it. Dialysis
against
distilled
the purest water
serum
of which
was consistently
found
were still not completely
at pH 6 to 7 should,
however,
as a
devoid
remove
most
of
the impurity. IDENTIFICATION Since
/Y,,-globulin
by any other
tion has to be based graphs,
which
respond
of normal
procedure were
than
human
serum
upon this technique. made
to the normal
has hitherto
immuno-electrophoresis,
with
Fig.
instance,
such serum both
serum
ideally
complete
anti-y
(IA),
from
shows
anti-p,,
absorption,
of serum
a ,!-myeloma,
from
of BZA antigen.
that
lt is obvious
that
pZhI line may
IZefmttccsp. 102
cirrhosis
the PZM precipitation
The
ever,
a faint
with
antiserum,
and with
(I B) or, at least partially,
last-mentioned consist
a patient
anti-human
the antiserum where other
our preparation
any identifica-
our preparations
really
cor-
protein.
I. Testing of /12_Aagainst various antisera is illustrated each
that
I shows some immuno-electrophoretic
the aim of proving
/?,, serum
not been individualized
it follows
in Fig.
was
line.
used
Fach
similar
the “paraprotein“ consists
be distinguished
with
performed
exclusively
on some
which
with
For the
been found pure
to
antigens.
of PZA antigen.
preparations
with either
1/400 of its volume
had previously
were
almost
from
(I C) was removed.
had been incubated
exhaustions
since
slide was developed
antiserum
anti-pzA
I A, B and C. In
as a reference,
HOW-
corresponding
to
J. F. HEREMANS,
XI.-TH.
HEREMASS,
H. E. SCHLKTZE
VOL.
4
(1959)
Fig. I. Immuno-electrophoretic anaiysis of ~~~-globulin, isolated from normal human plasma. A: developed with complete antiserum (left) and with antiserum absorbed with normal y-globulin (right). The latter still develops the BzA line. R: developed with complete antiserum (left) and with antiserum absorbed with /?,-macroglobulin (right). The latter still develops the pZAline. C: developed with complete antiserum (left) and with antiserum absorbed with a trace of a myeloma serum of the Baa variety. The PpA line is not developed. D: normal serum (upper tracing) and serum enriched with pure bZA (lower tracing). 15: a cirrhosis serum developed with complete antiserum (below) and with antiserum absorbed with isolated Bti (above). The latter faik to develop @*Aglobulin. F: a cirrhosis serum developed with complete antiserum, against which a @211solution was allowed to diffuse from the upper reservoir. A long straight precipitation line prolongs the fiSA bow generated by the corresponding serum protein. G: a combination of E and F.
VOL.
4
(1959)
/!&,-GLOBULIN
OF
HUMAN
SERUM
99
the previously mentioned impurity. In addition, the main BZA precipitation line is always accompanied by a faint secondary curve, which lies in its concavity. This will be discussed later. 2. Addition 0-f @%Ato serum,, as shown in Fig. I n, considerably increases the /!I*,, precipitation line of the serum, and causes its displacement towards the antibod! reservoir, which indicates excess antigen. 3. Absorption of a complete nntiserzbm, by means of our preparation, specifically removed the anti-pZA antibodies without affecting the various other antibodies, as is shown in Fig. I E. We consider this a most rigorous test, since even traces of y-antigen considerably weaken, or even abolish the y-precipitation line. It is interesting to notice, however, that with preparations showing no visible y-line upon direct testing with antiserum, the absorption test sometimes weakens the y-line to a considerable extent. The significance of this phenomenon will be discussed later. 4. Fusion of the precipitation lines of serum PZA and of our preparation of jjeA constitutes a further proof of their identity. This effect, which is shown in Fig. I F;, was obtained in the following way. After completion of the electrophoresis of a cirrhotic serum, a long slit running parallel with the migration direction was cut out and filled with a 0.27; PzA solution. Upon diffusing towards the antibody reservoir, antigen formed a long straight precipitation line, which clearly developed the BZA completely into the usual BZA line of serum in the region where the added antigen joined the corresponding protein from the cirrhotic serum. This experiment was of course inspired by the classical Ouchterlony technique which it combined with the electrophoretic separation of the serum antigens. A somewhat similar experiment is illustrated in Fig. I G where cirrhotic serum was developed both by complete antiserum and by antiserum to which a large excess of prepared ,LIZA had been added. Not only was the anti-Bza antibody wholly absorbed, but enough antigen remained in excess to form a new precipitation line, which, as in the previous experiment, joined the PZA line produced by the unabsorbed antiserum. PHYSICAL I.
Electropkoretic
AXD
CHEMICAL
PROPERTIES
mobility
In the paper electrophoretogram shown in Fig. z the relative mobilities of prepared p,,-globulin, commercial y-globulin (K.A.B.I.), and normal serum proteins can be compared. The /12* clearly lies exactly in the depression between the conventional ,!L and y-peaks, and the remarkable obliteration of the normal inter-/+ depression which characterizes cirrhotic sera must at least partially be ascribed to increased serum fiZA levels. 2.
~Tltracentrif~~gation
The analytical Spinco ultracentrifuge was used for the determination of the sedimentation constant of isolated ,!J,&obulin. The picture obtained after 27 minutes is shown in Fig. 3. A main component with a sedimentation constant of S,, = 7 probably corresponds to the normal /3ZAof human serum. The similarity of this value to the sedimentation constant of normal intact y-globulin (also 7) is striking, and probably accounts in part for the fact that /3,,-globulin so long escaped detection. References 0. 102
J. 1:.HENEX4SS,
100
M,-TH. HEREM_4NS,
H. E. SCHVLTZE
VOL. 4
(1959)
However, it was by no means unexpected to us, since it is known that most of the even those from p-myelomas, have the same sedimentation myeloma “paraproteins”, constant ( = 7) as normal ~-globuIin 8. A second and much less important component has a sedimentation constant of S,, = 10.5. It is possible, though not proved, that this component corresponds to the accessory precipitation line which usually borders the main curve of immunoelectrophoretic tracings obtained with our preparations. The exact antigenic composition of this line is not yet clear, but it possibly represents some artefact derived from pziz during the fractionation process, such as a polymerized form of the latter protein. A comparable problem also exists for most commercial y-globulin preparations $.
Fig. 2. X comparison of electrophoric mobilities of y-globulin, isolated &-globulin and serum proteins. 3,
Fig. 3. L7tracentrifuge diagram of isolated &,-globulin. The major component has a sedimentation constant of about 7 S. An impurity of S = 17.9 corresponds to fi2-macroglobulin.
Composition Only a few preliminary data are given here. Lipid staining of paper electrophoretograms of pzA failed to reveal any protein-
bound fat. On the other hand, the protein is intensely stained by the periodic-acid-Schiff reagent, which indicates a large carbohydrate content. The following figures should be regarded as first estimates only (on the basis of IOO mg of polypeptide, as determined by the biuret method} : 16.20 g4’i, Xitrogen : Hexoses (aldoses, as determined by IIOLT'S anthrone method) : 4-p g% Hexosamines (as determined by ELSON AND MORGAN'S method) : 3.74 gT/o Refemces
JI, IOZ
VOL. 4 (‘959)
&,-GLOBULIN
OF HUMAN SERUM
101
On the whole, ,82A-globulin appears to be rather rich in carbohydrates, as compared to y-globulin. This again, could have been expected, as it is well known that /3-myeloma proteins usually show markedly higher carbohydrate levels than do proteins from y-myelomas lo, Il. 4. Sonze other firoperties p,,-Globulin is precipitated by ammonium sulfate in concentrations between and 1.8 M. In the presence of distilled water it behaves as a pseudoglobulin, thus again showing contrast with y-globulin. Finally, it is resistant towards precipitation with zinc ions. This has already been mentioned, and has been used as a means of fractionation. 1.2
Fig. 4. Antigenic relationships between BzA- and y-globulins. A : developed with horse antiserum against human serum. B: developed with rabbit antiserum against human serum. The tracings on the left side were obtained with I gojOprotein solutions; those on the right side with 0.5 ~7; protein solutions. In each tracing the left reservoir contains a ,& preparation, and the right reservoir a y-globulin solution of corresponding concentration.
ANTIGENIC PROPERTIES
The study of the antigenic properties of our isolated p,,-globulin was considerably hindered by the fact that none of our preparations were absolutely pure, as judged by immunological standards. Therefore no attempt was made to obtain a specific antiserum by immunization of animals with our preparations, as it could be expected that several antibodies would be produced. Nevertheless complete anti-human antisera provided some useful information concerning the relationship between y- and p,,-globulins which will be briefly summarized here (Fig. 4). Fig. 4 A illustrates an experiment where a preparation of b2* and a solution of commercial y-globulin (K.A.B.I.) diffused against a polyvalent horse anti-human antiserum. The bZA preparation contained a rather large amount of pZM, whose precipitation line is visible near the antigen reservoir. Two different concentrations (I and 0.5 g:b resp.) were used for each antigen. At the point where the main precipitation lines intersect, some fusion apparently occurs. This is clear in the case of the I l;Oantigen, and is shown by a slight bridging bow in that of the weaker concentration. Each line seems to continue ind.ependently in the original direction, although considerably weakened, at least for the I:/; antigen experiment. No such weakening is visible for the 0.50; antigen. Both phenomena are again seen when a rabbit antiReferems
9. 102
J. I;. HEREMANS,
102
M.-TH.
HEREMANS,
H. E. SCHULTZE
VOL.
4 (1959)
human serum is employed for similar experiments (Fig. 4 R) ; here the fusion seems to dominate over the independence. Tt is tempting to interpret the above findings as indicative of ~~r~~~Lavttigcnir correspondenct?. Finally it may be of interest to mention that in Fig. I F a similar weakening of the @,,-precipitation line can be seen at the point where it crosses the strong gamma curve. ACKNOWLEDGEXENTS
The authors are greatly indebted to Miss C. to Dr. PH. DUSART for their skilful assistance.
GILLANT,
Miss M. TH.
DELCOX
and
By a combination of zinc sulfate precipitation, ammonium sulfate precipitation and starch block preparative electrophoresis, ~~~-globulin could be obtained from normal human serum and plasma. Immuno-electrophoretic tests showed the preparation to be identical with native &, serum protein. Its sedimentation constant was found to be S,, = 7, the same as for y-globulin. The carbohydrate content of @aA-globulin was determined and found to be much higher than that of y-globulin ; the new protein is therefore classified as a glycoprotein. Some other properties of this protein are mentioned, amongst them a certain immunological relationship with y-globulin. REFERESCES 1 C. B. LAURELL, 2 P. GRABAR AKD 3 C. WALLENIUS,
S. LAURELL AND X. SKOOG, Cl&. Chess., L (1956) 99. C. A. WILLIAMS JR., ISiochim. Biophys. *4cta, TO (1953) 193. R. TRAUTMAX, H. G. KUNKEL AND E. C. FRANKLIN,J. BioE.Ckem.,
22.5 (1957)
Zjj.
* P.
BURTIN,
R.WIEME, 161. 5 I’.
GRABAR,
175.
L. HARTMAW, CH.WUNDERLY, R.
6 L. L. GRIFFITHS 7 1-I. G. KUNKEL, 8 9
10 I1
FAUVERT,
J. J. SCHEIDEGGER, J. HEREMANS, Ii.FAUVERT AND 1'.GRABAR, Heu.fraq. P.
BURTIN
AND
L. HARTMANN,
Rev.frm~.
F. WESTENDORP-BOERMA,
e?tudes rlin. biol., L (1957) e’tudes clitz. hiol., I (1956)
J. Clin. Pathol., 6 (1953) ~87. Electvo;bhoresis, in D. GLICK, Mcfhods of Biochemical Inc., New York, 1954. p. 141.
AND
V. A. L. BREWS,
d~mlysis, vol. I, Tnterscience Publ., F,. L. SMITH, D. M. BROWN, M. L. MCFADDEX, V. RUETTNER-JANUSCH AND B. V. JAGER, ,/.RiOE.ChW?.., 216 (1955) 601. H. E. SCIIULTZE, Behringwerke-Mitt., 26 (195~) 3. C. K. LAURELL, H. LAURELL ANI) J. WALDENSTRGM, .4?%. J. Afed., 22 (1957) 24. J. SONNET, L. LOUIS AND J. HEREMANS, .4&x Haematol., 14 (1955) 193. Zor&e
CLINIC.4 CIIIMICA ACTA
96
ISOLATION
AND
OF THE
DESCRIPTION ,!&A-GLORULIN
J, F. HBRPXANS, Medical Department,
C;niversity
M:TH. Clilzic
OF A FEW OF HUMAN
HEREMASS
St. Pierre,
Marburg/Lahn
PROPERTIES SERUM
AND H. E. SCHl~LTZE
Louvain
(Belgium)
and the Behringwevke,
(Germalzy)
Conventional electrophoretic graphs obtained with normal human sera show, between the well-individualized /3- and y-peaks, a more or less pronounced depression, which is used as a landmark for planimetric work. This also holds for those techniques’ that produce a split P-band. In some pathological conditions, for instance in cirrhosis, the separation of /3 and y may become completely indistinct. For such sera at least, the existence of proteins with intermediate mobility may be assumed. Moreover, many cases of myeloma and macroglobulinemia Waldenstrom are characterized by the appearance of large abnormal protein peaks in this region. The substances responsible for these peaks have been given various designations, such as fi2, yl, 111 or 5‘. Recently our knowledge of the nature of these pZ- (or yl-) proteins has been increased by some more precise information. Thus, GRABAR et ak2, using the immunoelectrophoretic technique, has established beyond doubt that the y-globulins really consist of a whole family of components with identical antigenic specificity. The most rapid members reach the slower (w2region, and the slowest molecules extend to the cathodic end of the graph. From this it follows that the so-called /32(or ‘12, or y,, or 5) (using the word in its immunological region at least contains some “y-globulin” meaning). Eext, KUNIWL'S group3 succeeded in preparing a high-molecular weight component, whose mobility slightly exceeds that corresponding to the classical y-peak. Although the final proof has not yet been given, it may be assumed with reasonable certainty that KWKEL'S heavy y,-globulin at least partially coincides with an antigenie constituent, which BURTIN and co-workers4 called “/32M”. The latter protein appears to be appreciably increased in cirrhotic sera, as is KUXKEL'S heavy y,-globulin, and it has been shown to be specifically involved in macroglobulinemia Waldenstrom 4. Finally, a third globulin with characteristic b2 mobility has been known since the very first immuno-electrophoretic investigations of human sera, performed by GRABAI~ AND WILLIAMS in Igj3. It was merely called “fi,-globulin” by them, until the discovery of the “/3,-macroglobulin” precipitation line prompted them to change the name to “P,,+-globulin”, in order to avoid confusion. This is the protein that will be considered in the present publication. ISOLATIOS
All our attempts to isolate the /3,,-globulin from human plasma or serum have been based upon the hypothesis that it is a substance with high resistance to precipitation with zinc ions. This idea originated from a consideration of the following wellRefeuerlct~s p.IO2
4 (Igjg)
VOL.
known
P,,-GLORULIN
facts.
First,
it has been shown
loma sera show an enormously this protein majority
seems
“paraprotein” tested
by
all of
B-type,
postulated fraction
#?-myelomas
KUNKEL’S
zinc
show
We
values
a fact which the supernatant
still
contained
also remained action
around
peak”.
The
seem to derive
their
towards
flocculation
a few
this
of the y-myereactions,
myelomas,
zinc
and the bZA nature
when
all or almost
ion precipitation”. of the abnormal
by the analysis
We protein
of a large number
by increasing XII.
from
obtained
all of
(after
a conventional but
to 50 mJ!l; concentration
addition),
were
large
found
of
the selectivity
this was obtained of
by
100 mmolesl,
low ionic strength to result
test on
quantities
was made to increase
a zinc sulfate
+29”,
KUSKEL
that
its concentration zinc sulfate
to
from
/3,,-globulin,
An attempt
Finally
+25”
its
PH
and tempera-
in practically
complete
of /3*,,- and y-globulins.
The
zinc ions were
Preliminary supernatant
of normal
p,-globulins,
removed
by adding
that 1.8 flf ammonium
serum.
was submitted
of a2- and its
subsequently
trials showed
the p,,-globulin
method
but
was confirmed
in solution.
neutrality
ranging
into
disturbed test,
stability
this stability
that
separation
“paraprotein
however,
P-mye-
in some cases
sera (to be published).
COHS’S Method
tures
myelomas,
turbidity
found
of the zinc
many
5. On the other hand the vast majority
serum
y-globulin
that
line. Indeed,
of the so-called
show extremely
sulfate
that
corresponded,
of myelomatous normal
the whole
an unusual
therefore
,5,,-precipitation
of the p-type
from the y-antigen
and some
using
to constitute
97
by immuno-electrophoresis
increased
of the y-, and many
lomas,
OF HU>IA?; SERUN
Therefore
the above
to this salting-out
of which
about
electrophoretic
components
macromolecular
/3ZM-globulin of normal
completely
described
procedure,
7596 was
by
I to 2 gy6 glycine.
sulfate
which
pZA. The
KUNKEL‘S
precipitated
glycine-containing yielded
mixture
starch
block
a mixture
was
resolved
electrophoresis
‘.
The
contaminant
of our preparations,
of it. Dialysis
against
distilled
the purest water
serum
of which
was consistently
found
were still not completely
at pH 6 to 7 should,
however,
as a
devoid
remove
most
of
the impurity. IDENTIFICATION Since
/Y,,-globulin
by any other
tion has to be based graphs,
which
respond
of normal
procedure were
than
human
serum
upon this technique. made
to the normal
has hitherto
immuno-electrophoresis,
with
Fig.
instance,
such serum both
serum
ideally
complete
anti-y
(IA),
from
shows
anti-p,,
absorption,
of serum
a ,!-myeloma,
from
of BZA antigen.
that
lt is obvious
that
pZhI line may
IZefmttccsp. 102
cirrhosis
the PZM precipitation
The
ever,
a faint
with
antiserum,
and with
(I B) or, at least partially,
last-mentioned consist
a patient
anti-human
the antiserum where other
our preparation
any identifica-
our preparations
really
cor-
protein.
I. Testing of /12_Aagainst various antisera is illustrated each
that
I shows some immuno-electrophoretic
the aim of proving
/?,, serum
not been individualized
it follows
in Fig.
was
line.
used
Fach
similar
the “paraprotein“ consists
be distinguished
with
performed
exclusively
on some
which
with
For the
been found pure
to
antigens.
of PZA antigen.
preparations
with either
1/400 of its volume
had previously
were
almost
from
(I C) was removed.
had been incubated
exhaustions
since
slide was developed
antiserum
anti-pzA
I A, B and C. In
as a reference,
HOW-
corresponding
to
J. F. HEREMANS,
XI.-TH.
HEREMASS,
H. E. SCHLKTZE
VOL.
4
(1959)
Fig. I. Immuno-electrophoretic anaiysis of ~~~-globulin, isolated from normal human plasma. A: developed with complete antiserum (left) and with antiserum absorbed with normal y-globulin (right). The latter still develops the BzA line. R: developed with complete antiserum (left) and with antiserum absorbed with /?,-macroglobulin (right). The latter still develops the pZAline. C: developed with complete antiserum (left) and with antiserum absorbed with a trace of a myeloma serum of the Baa variety. The PpA line is not developed. D: normal serum (upper tracing) and serum enriched with pure bZA (lower tracing). 15: a cirrhosis serum developed with complete antiserum (below) and with antiserum absorbed with isolated Bti (above). The latter faik to develop @*Aglobulin. F: a cirrhosis serum developed with complete antiserum, against which a @211solution was allowed to diffuse from the upper reservoir. A long straight precipitation line prolongs the fiSA bow generated by the corresponding serum protein. G: a combination of E and F.
VOL.
4
(1959)
/!&,-GLOBULIN
OF
HUMAN
SERUM
99
the previously mentioned impurity. In addition, the main BZA precipitation line is always accompanied by a faint secondary curve, which lies in its concavity. This will be discussed later. 2. Addition 0-f @%Ato serum,, as shown in Fig. I n, considerably increases the /!I*,, precipitation line of the serum, and causes its displacement towards the antibod! reservoir, which indicates excess antigen. 3. Absorption of a complete nntiserzbm, by means of our preparation, specifically removed the anti-pZA antibodies without affecting the various other antibodies, as is shown in Fig. I E. We consider this a most rigorous test, since even traces of y-antigen considerably weaken, or even abolish the y-precipitation line. It is interesting to notice, however, that with preparations showing no visible y-line upon direct testing with antiserum, the absorption test sometimes weakens the y-line to a considerable extent. The significance of this phenomenon will be discussed later. 4. Fusion of the precipitation lines of serum PZA and of our preparation of jjeA constitutes a further proof of their identity. This effect, which is shown in Fig. I F;, was obtained in the following way. After completion of the electrophoresis of a cirrhotic serum, a long slit running parallel with the migration direction was cut out and filled with a 0.27; PzA solution. Upon diffusing towards the antibody reservoir, antigen formed a long straight precipitation line, which clearly developed the BZA completely into the usual BZA line of serum in the region where the added antigen joined the corresponding protein from the cirrhotic serum. This experiment was of course inspired by the classical Ouchterlony technique which it combined with the electrophoretic separation of the serum antigens. A somewhat similar experiment is illustrated in Fig. I G where cirrhotic serum was developed both by complete antiserum and by antiserum to which a large excess of prepared ,LIZA had been added. Not only was the anti-Bza antibody wholly absorbed, but enough antigen remained in excess to form a new precipitation line, which, as in the previous experiment, joined the PZA line produced by the unabsorbed antiserum. PHYSICAL I.
Electropkoretic
AXD
CHEMICAL
PROPERTIES
mobility
In the paper electrophoretogram shown in Fig. z the relative mobilities of prepared p,,-globulin, commercial y-globulin (K.A.B.I.), and normal serum proteins can be compared. The /12* clearly lies exactly in the depression between the conventional ,!L and y-peaks, and the remarkable obliteration of the normal inter-/+ depression which characterizes cirrhotic sera must at least partially be ascribed to increased serum fiZA levels. 2.
~Tltracentrif~~gation
The analytical Spinco ultracentrifuge was used for the determination of the sedimentation constant of isolated ,!J,&obulin. The picture obtained after 27 minutes is shown in Fig. 3. A main component with a sedimentation constant of S,, = 7 probably corresponds to the normal /3ZAof human serum. The similarity of this value to the sedimentation constant of normal intact y-globulin (also 7) is striking, and probably accounts in part for the fact that /3,,-globulin so long escaped detection. References 0. 102
J. 1:.HENEX4SS,
100
M,-TH. HEREM_4NS,
H. E. SCHVLTZE
VOL. 4
(1959)
However, it was by no means unexpected to us, since it is known that most of the even those from p-myelomas, have the same sedimentation myeloma “paraproteins”, constant ( = 7) as normal ~-globuIin 8. A second and much less important component has a sedimentation constant of S,, = 10.5. It is possible, though not proved, that this component corresponds to the accessory precipitation line which usually borders the main curve of immunoelectrophoretic tracings obtained with our preparations. The exact antigenic composition of this line is not yet clear, but it possibly represents some artefact derived from pziz during the fractionation process, such as a polymerized form of the latter protein. A comparable problem also exists for most commercial y-globulin preparations $.
Fig. 2. X comparison of electrophoric mobilities of y-globulin, isolated &-globulin and serum proteins. 3,
Fig. 3. L7tracentrifuge diagram of isolated &,-globulin. The major component has a sedimentation constant of about 7 S. An impurity of S = 17.9 corresponds to fi2-macroglobulin.
Composition Only a few preliminary data are given here. Lipid staining of paper electrophoretograms of pzA failed to reveal any protein-
bound fat. On the other hand, the protein is intensely stained by the periodic-acid-Schiff reagent, which indicates a large carbohydrate content. The following figures should be regarded as first estimates only (on the basis of IOO mg of polypeptide, as determined by the biuret method} : 16.20 g4’i, Xitrogen : Hexoses (aldoses, as determined by IIOLT'S anthrone method) : 4-p g% Hexosamines (as determined by ELSON AND MORGAN'S method) : 3.74 gT/o Refemces
JI, IOZ
VOL. 4 (‘959)
&,-GLOBULIN
OF HUMAN SERUM
101
On the whole, ,82A-globulin appears to be rather rich in carbohydrates, as compared to y-globulin. This again, could have been expected, as it is well known that /3-myeloma proteins usually show markedly higher carbohydrate levels than do proteins from y-myelomas lo, Il. 4. Sonze other firoperties p,,-Globulin is precipitated by ammonium sulfate in concentrations between and 1.8 M. In the presence of distilled water it behaves as a pseudoglobulin, thus again showing contrast with y-globulin. Finally, it is resistant towards precipitation with zinc ions. This has already been mentioned, and has been used as a means of fractionation. 1.2
Fig. 4. Antigenic relationships between BzA- and y-globulins. A : developed with horse antiserum against human serum. B: developed with rabbit antiserum against human serum. The tracings on the left side were obtained with I gojOprotein solutions; those on the right side with 0.5 ~7; protein solutions. In each tracing the left reservoir contains a ,& preparation, and the right reservoir a y-globulin solution of corresponding concentration.
ANTIGENIC PROPERTIES
The study of the antigenic properties of our isolated p,,-globulin was considerably hindered by the fact that none of our preparations were absolutely pure, as judged by immunological standards. Therefore no attempt was made to obtain a specific antiserum by immunization of animals with our preparations, as it could be expected that several antibodies would be produced. Nevertheless complete anti-human antisera provided some useful information concerning the relationship between y- and p,,-globulins which will be briefly summarized here (Fig. 4). Fig. 4 A illustrates an experiment where a preparation of b2* and a solution of commercial y-globulin (K.A.B.I.) diffused against a polyvalent horse anti-human antiserum. The bZA preparation contained a rather large amount of pZM, whose precipitation line is visible near the antigen reservoir. Two different concentrations (I and 0.5 g:b resp.) were used for each antigen. At the point where the main precipitation lines intersect, some fusion apparently occurs. This is clear in the case of the I l;Oantigen, and is shown by a slight bridging bow in that of the weaker concentration. Each line seems to continue ind.ependently in the original direction, although considerably weakened, at least for the I:/; antigen experiment. No such weakening is visible for the 0.50; antigen. Both phenomena are again seen when a rabbit antiReferems
9. 102
J. I;. HEREMANS,
102
M.-TH.
HEREMANS,
H. E. SCHULTZE
VOL.
4 (1959)
human serum is employed for similar experiments (Fig. 4 R) ; here the fusion seems to dominate over the independence. Tt is tempting to interpret the above findings as indicative of ~~r~~~Lavttigcnir correspondenct?. Finally it may be of interest to mention that in Fig. I F a similar weakening of the @,,-precipitation line can be seen at the point where it crosses the strong gamma curve. ACKNOWLEDGEXENTS
The authors are greatly indebted to Miss C. to Dr. PH. DUSART for their skilful assistance.
GILLANT,
Miss M. TH.
DELCOX
and
By a combination of zinc sulfate precipitation, ammonium sulfate precipitation and starch block preparative electrophoresis, ~~~-globulin could be obtained from normal human serum and plasma. Immuno-electrophoretic tests showed the preparation to be identical with native &, serum protein. Its sedimentation constant was found to be S,, = 7, the same as for y-globulin. The carbohydrate content of @aA-globulin was determined and found to be much higher than that of y-globulin ; the new protein is therefore classified as a glycoprotein. Some other properties of this protein are mentioned, amongst them a certain immunological relationship with y-globulin. REFERESCES 1 C. B. LAURELL, 2 P. GRABAR AKD 3 C. WALLENIUS,
S. LAURELL AND X. SKOOG, Cl&. Chess., L (1956) 99. C. A. WILLIAMS JR., ISiochim. Biophys. *4cta, TO (1953) 193. R. TRAUTMAX, H. G. KUNKEL AND E. C. FRANKLIN,J. BioE.Ckem.,
22.5 (1957)
Zjj.
* P.
BURTIN,
R.WIEME, 161. 5 I’.
GRABAR,
175.
L. HARTMAW, CH.WUNDERLY, R.
6 L. L. GRIFFITHS 7 1-I. G. KUNKEL, 8 9
10 I1
FAUVERT,
J. J. SCHEIDEGGER, J. HEREMANS, Ii.FAUVERT AND 1'.GRABAR, Heu.fraq. P.
BURTIN
AND
L. HARTMANN,
Rev.frm~.
F. WESTENDORP-BOERMA,
e?tudes rlin. biol., L (1957) e’tudes clitz. hiol., I (1956)
J. Clin. Pathol., 6 (1953) ~87. Electvo;bhoresis, in D. GLICK, Mcfhods of Biochemical Inc., New York, 1954. p. 141.
AND
V. A. L. BREWS,
d~mlysis, vol. I, Tnterscience Publ., F,. L. SMITH, D. M. BROWN, M. L. MCFADDEX, V. RUETTNER-JANUSCH AND B. V. JAGER, ,/.RiOE.ChW?.., 216 (1955) 601. H. E. SCIIULTZE, Behringwerke-Mitt., 26 (195~) 3. C. K. LAURELL, H. LAURELL ANI) J. WALDENSTRGM, .4?%. J. Afed., 22 (1957) 24. J. SONNET, L. LOUIS AND J. HEREMANS, .4&x Haematol., 14 (1955) 193. Zor&e