After sfding x hr.
1.0 cc
0.7 cc.
0^ cc
0.4 cc.
0.3 cc
0.2 cc.
o.iscc
O.I cc
NoH.*
NoH.
NoH.
NoH.
NoH.
NoH.
C.H.
C.H.
C.H.
S1.H.
Tr.H.
NoH.
C.H.
F.C.H.
Tr.H.
Experiment 11 (Usual SpHUmg).
CeairifugaUMd after z hr. at nom tesiperature.
Guinea pig aenun
z CO. + 0.9 <
Naa9C.c.
Amonnu
takes.
+ oi"c.cK/95NaOH
in wgt NaCl.
Deposit diaaolved in
zocc. ofo.9<NaCl.
ingredienu.
After standing at
for z hr. (control).
total Tohuae aoade «p to z.5 cc Reanlta taken after 90 and <
to each tube and the
Sonin.
30 lain.
6omin.
30 lain.
6o^inin.
jonin.
6omin.
somin.
6omin.
1.0 CC
C.H.
NoH.
NoH.
0.7 CC.
C.H.
NoH.
NoH.
C.H.
CH.
0.5 , CC
F.C.H.
C.H.
C.H.
CH.
0.4 CC
•Mch.H.
CH.
CH.
CM.
0.3 ex.
SLH.
C.H.
CH.
CH.
0.2 CC
NoH.
S1.H.
F.C.H.
CH.
CH.
CH.
0.15 CC
NoH.
NoH.
S1.H.
CH.
CH.
0.1 CC
NoH.
S1.H.
Mch.H.
0.07 cc
1 1
NoH.
^NoH.
"Throughout this work 0.9 per cent sodium chlorid solution was used in-
stead of a.85 per cent usually employed by other workers.
*In this and the following tables, for the sake of brevity, certain symbols
are employed to indicate the degree of hemolysis; thus CH. s= complete hemol-
ysis; F.CH. s= fairly complete hemolysis; S1.H. = slight hemolysis; Tr.H.=
trace of hemolysis ; No H. =: no hemolysis ; Mch.H. = much hemolysis.
Digitized by
Google
604
Biochemical Study of Complement'Splitiing.
TABLE n.
The Effect upon Components of Guinea Pig Serum of Heating for 30 Minutes
at 56'' C,
Experiment /.* Experiment 11.
Guinea pig serum heated at 56° C.
Fresh gninea pig seram i c.c.+8.a c.c. HCl N/aso (in disdlled water).
for 30 minutes + 8.a ccHCl N/as©
(in disUUed water).
Kept at room temperature for 60 minutes, centrifugalized and supernatant
fluid separated, neutnOized, and made isotonic by 0.8 c.c. of N/as
minutes, centrifugalized and su-
pernatant fluid separated, neutral-
iced, and made isotonic by 0.8 c.c.
of N/as NaOH in zo)( NaCl.
NaOH in lofi NaCl.
o.a c.c. of supernatant fluid, a
zo^ susptnsion of corpuscles in eacb tube.
unitt of ^unbooeptor. and o.z c.c-
in each tube.
Resulu taken after :
Resulu taken after :
3omin.
6omin.
9omin.
3omin.
6omin.
gomin.
control.
+0
NoH.
S1.H.
C.H.
NoH.
NoH.
KoH.
+Alanin 0.3 cc of 7.5% solution in
0.9% NaCl.
C.H.
C.H.
C.H.
NoH.
NoH.
NoH.
NoH.
+Deposit from splitting in '0.15 c.c.
C.H.
C.H.
C.H.
NoH.
NoH.
NoH.
NoH.
experiment I dissolvedj
in 10 cc 0.9% salt
solution. L 0.07 cc.
C. H.
C.H.
C.H.
NoH.
NoH.
NoH.
+Depo8it from splitting in C o.a cc.
NoH.
S1.H.
C.H.
NoH.
NoH.
NoH.
NoH.
experiment II dissolvedj
in 10 cc 0.9% salt so- |
ution. Lo.i c.c
C.H.*
C.H.
C.H.
NoH.
No.H
NoH.
While the phenomenon suggests a double constitution of the com-
plement, yet it does not exclude the possibility of its being caused
by an excess of acid or alkali. Guinea pig serum, when mixed with
hydrcxrhloric acid in the proportion prescribed in this procedure,
becomes almost neutral or but faintly alkaline. This causes the
precipitation of certain proteids, mainly the globulins. Thus the
acid introduced is quickly taken up by certain alkaline salts of the
serum and there is none left free. The precipitated globulins also
bind a certain amount of the acid. The supernatant fluid which is
almost neutral and contains chiefly the albumin fraction of the
serum, is then mixed with sufficient free alkali to neutralize exactly
• The total volume was brought up to i c.c. by 0.9 per cent salt solution in
both experiments.
•This phenomenon docs not occur constantly. With some specimens of
complement (very few) the mid-piece is undoubtedly destroyed by heating the
whole serum previous to the splitting, but it is easier to destroy the mid-piece
after its separation.
Digitized by
Google
Jacob Bronfenbrenner and Hideyo Noguchi.
605
the amount of acid added, but this amount of alkali is in excess
when added to the serum mixture from which the globulin fraction
was removed after its precipitation by hydrochloric acid. Thus it
happens that in the serum treated with the acid and then neutralized
with the alkali in the manner prescribed, in the presence of the
globulin fraction, the complement action is restored. On the other
hand, if the globulin fraction is first removed and the same amount
of the alkali is added to the remaining supernatant fluid, the com-
plement activity of the supernatant fluid is paralyzed. The exami-
TABLE III.
Demonstration of Complement Inactivation by Sodium Hydroxid in the
Process of Complement-Splitting,
Tubes.
X
3
3
4
5
6
Guinea pig serum
N/250 HCl
0.9%NaCl
I.OC.C.
8.2 c.c.
I.OC.C.
8.2 C.C.
I.OC.C.
8.2 C.C.
I.OC.C.
8.2 c.c.
I.O C.C.
1.0 C.C.
9.0 C.C.
DiatiUed water
8.2 C.C
After I hour at room temperature all the tubes were centrifugalized. From tubes i
and 3 only supernatant fluid was used ; from the others the entire contents of the tubes were
used.
N/25 NaOH in 10%
VftPl
0.8 cc
0.8 cc
10% NaC)
[
0.8 cc.
0.8 cc
8 c c
Supernal
ant fluid
Theei
titire contei
Its of the tubes
Reaction of resulting
5.6 cc
2.5 cc
2.5 cc
5.6 cc.
6.2 cc
6.2 cc.
digo' (calculated for
Ha
NaOH
NaOH
HCl
HCl
HCl
the whole volume).
N/250
N/250
N/250
N/250
N/250
N/250
Complement activity of
the resulting mixtures
(with 0.1 c.c. of 10%
suspension of corpus-
cles and 2 units of
amboce]
)tor.
1.0 c.c.
C.H.
NoH.
1
0.7 c.c.
SLH.
*«
0.5 c.c.
Tr.H.
C.H.
0.3 cc.
NoH.
C.H.
C.H.
F.C.H.
2 '
0.2 c.c.
F.C.H.
F.C.H.
Mch.H.
C.H.
0.15 cc.
Mch.H.
Sl.H.
Tr.H.
F.C.H.
0.1 cc.
NoH.
NoH.
NoH.
S1.H.
0.07 cc
NoH.
'The use of this indicator, first introduced by Bernhardt, was suggested to
us by Prof. W. J. Gies. The indicator is not noticeably influenced by the
presence of proteids or carbon dioxid.
Digitized by
Google
606 Bioohemical Study of Complement-Splitting.
nation of the so<alled neutralized supernatant fluid for alkalinity
indicates a slight excess of alkali over that present in the same
dilution of the untreated serum.
That this slight excess of alkalinity is not solely responsible for
the loss of action of the untreated complement, but that the nature
of the alkali employed has more to do in paralyzing the complement
seems very probable. Thus, the free alkali in the form of -sodium
hydroxid fails to restore the original alkaline salts, and exerts an
injurious effect upon the active complement constituents. That
this is the case is easily shown by the experiments in table III.
These experiments show that the supernatant fluid from which
the native alkalinity has been almost completely removed by
hydrochloric acid still has nearly two thirds of the original com-
plement power of the serum. But by adding gradually increasing
amounts of sodium hydroxid, the activity is steadily reduced (table
V) until in some cases it disappears completely (table VI), when
the amount of sodium hydroxid reaches the point supposed to be
necessary to restore the original alkalinity of the treated serum.
The completeness of the inactivation of the complement action of
the supernatant fluid is also enhanced by a slight excess of the free
sodium hydroxid toward which the complement shows a great sus-
ceptibility. It may be assumed that the salts that cause the native
alkalinity, principally the carbonates and phosphates, are split by
hydrochloric acid and form sodium chlorid and the apids, but the
greater part of carbon dioxid leaves the fluid, thus causing the free
alkali introduced to act directly upon the complement.
It is interesting, however, that, when the serum was acted on by
hydrochloric acid and sodium hydroxid was subsequently added,
without removing the globulin fraction, much if not all of the
complement activity was restored (table IV, A).
The reason for sodium hydroxid not causii^ the same injurious
effect upon the complement in this instance (table IV, A) must be
due to the presence of the globulin fraction which seems to have a
greater affinity for the free sodium hydroxid than the albumin
fraction in which the active principle of the complement seems to
reside.
That the globulin fraction has a greater affinity for sodium
Digitized by
Google
Jacob Bronfenbrenner and Eideyo Noguchi.
607
<
«
•ft
3
•ft
2;
I
^
f
3
9^
555-;
Is
I
I
J1
JS
33 ji
8.-1
-56
II
asSo
uSa5Z
»
.US a
x4x^
.IB .3;
• as
uuuuudOO
oi^tAfoetMHio
Moodoodo
Digitized by
Google
608
Biochemical Study of Complement-Splitting.
s
^ s,
s
o
Co
«
o
^
Co
Ta
^5^
%
^
^
Q
1
2
1
a
1^^
5 d o
au o
uta2:
V
ja -g
K
e
5d
.J ►-• 35 55 ffi
*. *. o o o
-a o.
s
c^
o ixiiz;
mmzzz
a o
«
•»- 1^
1^
5
g
1
»
c5
6
. ffiffi
xxxxx^
S|
1
d
ffi uu
8 S
s
^
U fedi
cS o5 CO UiU^Z
c "^
£
1
« «
<
Z
. 5d ffi K
i «
1 1
1
?
ffi ucJ U
* g
1
1
d fe fe fe
m <ji c^Hiz
M •^
. *J
V^ •*«
t
•:s
o
M
^.«i
S"
S'S,
d
CJ to
w c^2:
1:
III
^X
ii
g 1 ?
ffi cJ U
3!o
g -U
6
'&&i
d ix; ix;
w2;
2 *5
13 '^
m
X 3 0-
0"o
c« ^
*— \ X-N
ru ru
A-
n S
m
N— ' >^
N-*
Is.
1
1
.8
c«
6
fci ^ '^
G
.go
+ .
1
S
&
^
"c^«^
u,^ m i
•S d XS
IB*?
S"
S iii ucJ
c^hI
S ^
S
E B
'o
S
4> ^
6
ft K
3
1
a a: K 5d ^
. . »: c o
O
uuudOuuuuu
o
'sajnixno
uuOuuuuuuu
o
JO sionooiy
r^ to ro CI H Q
tOfOetM O O O OOO
1
6
o
o o
o
b
6
o o ol
"S
c8
Digitized by
Google
Jiicob Bronfenbrenner and Hideyo Noguchi. 609
hydroxid than the albumin fraction is easily demonstrated by ex-
periments dcfscribed later.
In order to ascertain whether the loss of complement action in
the albumin fraction, the so-called end-piece, is really due to a
reversible inactivation through the action of sodium hydroxid, such
as had been observed by Noguchi in his earlier experiments, and
not due to the removal of one of the interdependent components,
the mid-piece, it was necessary to make a series of reverse experi-
ments with an acid.
The following experiments show that the complement activity of
the inactive end-piece can be restored to a great extent by means of
neutralization of the alkali with adequate amounts of hydrochloric
acid (table VII).
We now come to consider the mechanism of the restoration of
the complement activity of the end-piece by the globulin fraction
known as the mid-piece. From the reactivation experiments with
hydrochloric acid, it is not improbable that the combination of the
complement and sodium hydroxid can also be split by means of an
amphoteric compound. Among the amphoteric substances we find
various proteids, proteoses, and many non-coagulable s)mithetic
amino acids. The serum globulins are undoubtedly amphoteric.
In the present experiments we have also used egg-white, alanin,
deuteroalbumose, and peptone.®
These experiments (table VII) demonstrate that the com-
plement action of the so-called end-piece can be readily restored
by adding adequate quantities of the amphoteric substances herein
employed. One may wonder how such a weak acid as the amino
acid, which does not form a salt at ordinary temperature, can
reverse the combination of the alkali and the complement fraction.
But, it must be remembered that the affinity between the comple-
ment and alkali cannot be stronger than that which exists between
any other amphoteric bodies. Moreover, the direction of the rever-
sion i& dependent upon the total number of acid radicals possessed
by a given substance. Thus, a greater amount of the substance is
•We arc indebted to Dr. F. J. Birchard, of The Rockefeller Institute for
Medical Research, for the albumose and peptone preparations. The albumose
was prepared from fibrin according to Pick's method, and the peptone from
Witte's peptone according to Sigfried's method.
Digitized by
Google
610
Biochemical Study of Complement'SpUtting.
Z
r«5
IK
Co
t
9.
©•a
^6
r
I
I
o*o o o ^ o o
»j4
Nrt »j M ^ KC *^ w w S S {U {U s s
»JCW3:oq*.*S o o o o o o
mvmuzt^^i^zz z z z z
u u cJu
X X
X
b Ixi U
X .33 .
2u2u
X X
cj cj
U o z z
fO fO ^0 fO f) fO fO fO fO PO fO fO fO f) fO f) fO fO
666666666666666666
o
6
W5
M ft f*5
O O O
6
o
55j
r
*^ ^ JC- a 33 S
•8-S Z
ffl . SIS ffi
ffi Ox o
6 txiu fa fa
ZO u
. X
W 6
U Z
t^ fO to fO f*5 fj fO fO fO f) fi fO to to tO to to CO
666666666666666666
o
6
6
^'- 8 s i
J-
en ^
'3 «
SI'S.
ss
6
"•a
8.2
•CO
GO O
I
•-•a _
Digitized by
Google
Jaodb Brcnfenbrenner and Hideyo Noguohi.
611
^
to
IK
> t
z
5^S rw
I
s a
O
'" o
0.66
o*^ 6 6
• X
X
s
X
X
o g3
X
j5
u m
X
3i * 33
X
6
to
a:
X
55
X
o
X
u
fa
•8
55 f5|
X ^X
K XX
4 aw
5 ^2
X ss*
W 55|
•K teX
53 S*
I I
>8 I
I ^
B •§
^ -M V
_. * *
pa
^ §.2
5l u *>
•a •«
. 5 ««
^ B 6
^ S « o
Si o
— £■28
<^ .1: o
_S -2 ^
'§•2 3 «
lag's
«^ G
a 4^
« 5 'Si's
^ ill
K S3 2 I
^3
I 2 .? S"
a ga Jz;
Digitized by
Google
•^12 Biochemical Study of ComplemenUSplitting.
required if it possesses fewer acid radicals. That the globulins
iiave a greater affinity for sodium hydroxid than the albumins may
be accounted for by the larger size of the molecules of the former
-class of proteids. The solubility of the substances seems also to be
important, as certain insoluble amino acids, such as asparagin,
asparagin acid, and leucin failed to reactivate the end-piece.
COMPLEMENT-SPLITTING WITH CARBON DIOXID.
Liefmann splits the complement in the following manner. One
half of a cubic centimeter of guinea pig serum (complement) is
mixed with two cubic centimeters of distilled water, and carbon
<iioxid gas is passed through until no more precipitate is formed.
The treated serum is then centrifugalized to separate the precipitate
from the supernatant fluid. The clear supernatant fluid is carefully
decanted and filtered through hardened filter paper into another
receptacle and is used as the end-piece after being made isotonic
with sodium chlorid. The deposit, which represents the globulin
fraction of the serum, is then dissolved in 2.5 cubic centimeters of
•0.85 per cent, salt solution. This portion is used as the mid-piece.
In a certain number of instances the phenomenon of splitting is
ideally shown, the end- and mid-pieces separately having no action
upon the sensitized corpuscles (table X). The two components
obtained by the carbon dioxid method behave in much the same way
as those prepared by the hydrochloric acid method of Sachs. It
may be stated, however, that the end-piece portion remains quite
active in many instances, as has been observed also by Liefmann
and others (tables IX and X).
In considering the processes that lead to the inactivation of the
end-piece by carbon dioxid and the subsequent reactivation of this
portion by the mid-piece of the globulin fraction, one can conceive
that an essential difference exists between the inactivation by carbon
dioxid and that by the hydrochloric acid method. In case of the
hydrochloric acid-splitting, the carbonates as well as other alkaline
salts of the serum are completely converted into chlorids. Small
amounts of free carbon dioxid or other weak acids remaining in the
fluid cause a partial inactivation of the complement.
On the other hand, the passage of carbon dioxid through a
Digitized by
Google
Jacob Bronfenbrenner and Hideyo Noguchi.
618
IS
IS
5
s
u
^
(X.
J
o
s
?
u
»
S
s
r
i
2
X
s
X
r
s
1
i
^"^
1
X
J
u
ffi
rr;
ffi
s
r
(3
1
ffi
K
^
s
X
»
S
u
CJ
^
i
1
,
,
X
»
d
o
X
•g
2
CJ
fa
ffi
?
ci
Digitized by
Google
614
Biochemicdl Study of CamplemenUSpUtting.
"n
•?
o
00
XX X X
,M o
o o
Z Z
^x w w»2k
^^ ^_J NN ^b{ ^^ ^b4
u
X
XX
X
U o c
to ^55
•gas
d
e
e
9
.a
V
bl
ti
1
I
5
■3
CO <J
§■?•
S 6
B +
Digitized by
Google
Jacob Bronfembrenner and Hideyo Nogwihi.
615
s.
I
2
5 ?
a
X
t5
si xs w
9 -
^ o
ll§
5355
^
o
«
,
^«
3
X
^
M
6
V
5:
>
t
c5
,
-S.
(5
X
^
s
o
*K
^
^
js"
8
J
u
.?.
^
d "P
^
n
^
a ,
«M
•O.S
2211
Ji •4" *3 '♦!
Ilia
^i£^
1
o5
5
5
a;
K
t
I
5
t
I
u
u
ffl 1
1
i
3
ai a
a X
M
O Ob
°
2; 2^ 2
H
» Sjij a
o
S ^a ^
(5
tsai
ff
» -^-S »
6
S! SS ss
S
a a . w
?
•S da -S
S d^u S
:!
a -a
r
dad
lb U (tl
6
M
a a'
6
d d
;i
:l :
:| :
*§ i
:*** :
3
• c !
a r
•o c
' *3
2 8
H O
s'l •
M
^+ :
•il;
'cd :
11 J
•i0ssii\
id
5? 1
Digitized by
Google
616
Biochemical Study of Complement-SpUtting.
•I
•«
5
Q
^
<2
X X X X
i i i ^
S ^ S jijQ S
^ m ^ m i
ffi ffl sd 2 ^
m m ^ h
X
to
2 ffl J«!
^ '^ ^
X
X
u
ef^i^
;!
X X
*. X X X
fa 2; H w
X 3: SB ffi
u S c^
X X SB
^ H i^
X X
m
X
•S
I
X X
X
t
8
s
p
Digitized by
Google
Jacob Bronfenbrenner and Hideyo Noguchi.
617
TABLE XII.
Reactivation of Complement Inactivated by Carbon Dioxid.
Amoonts of reacdvadng
substances.
AlamD3)(.
Mid-piece
Egg-white
i:zo.
Sheep
serum. 30
min. at
56'' C. 1:5.
Protoalbu-
mosess^.
Deuteroal-
bumoses a ](..
^.
I.O c.c.
0.7 c.c.
a|
0.5 c.c.
as
0.4 C.C.
C.H.
11
0.3 C.C
C.H.
Sl.H.
0.2 C.C.
C.H.
Mch.H.
9-c'
o.is c.c.
F.C.H.
Mch.H.
No. H.
n
0.1 c.c
Mch.H.
C.H.
Sl.H.
S1.H.
No. H.
0.0s cc
Mch.H.
Sl.H.
Mch.H.
Tr.H.
0.03 cc.
S1.H.
C.H.
F.C.H.
S1.H.
Mch.H.
0.02 cc.
C.H.
S1.H.
o •
o.oi cc
C.H.
6
, 0.005 cc
C.H.
mixture of guinea pig serum and water precipitates the globulin
fraction just as in the case of the hydrochloric acid treatment, but
the carbonates as well as phosphates remain unaffected by this treat-
ment. The inactivation of the complement in this instance is doubt-
less due to the overcharging of the serum water with carbon dioxid
gas, which by virtue of its acid nature combines with the ampho-
teric molecules of the serum albumin fraction in which the activity
of the complement is residing. In a medium, whether in the undi-
luted native serum or in a serum diluted with sodium chlorid solu-
tion of isotonic concentration, in which the sodium chlorid content
reaches a sufficient concentration (table XI), a weak acid like car-
bon dioxid fails to enter into even a loose combination with the
albumin molecules, and hence leaves the complement comparatively
uninjured. The reason that carbon dioxid acts upon the comple-
ment molecules in serum sufficiently diluted with water is that the
strong electrolyte, sodium chlorid, can in that dilution no longer
interfere with the action of a weak acid. In diluted serum carbon
dioxid causes an acid inactivation of complement and the activity of
complement may be restored by certain amphoteric substances (table
XII) and also by an alkali (table XV, C), but not by an acid.
The phenomenon of reactivation by means of adequate quanti-
ties of sodium hydroxid may raise a question as to why the result-
Digitized by
Google
618
BiochemiodU Study of Complement-'SpUtting.
9S
^
2 "? o
►5 ^ r
X
•5: rt
t E
I -I
>
H ^
SI
0:1 -co
^ 8
Si ^
8 -
-4 g
'SI
1-2
e S
«9 CO
V> U
II
«0 (A
J* 5»
•s
•s
1
ffl
J
ffl
i
r^
z
z
^"^
^
6
X
rf
CJ
.
+
H
a»
!0
+
(J
SB
1
1
^1
%
f
<S|
ffi
;l
S
'
^
f
1
i
§
2
(J
r
4
z
u
f
*^
UZ
d
u
u*
K
u
X
3
.2
1
a
J
^
u
u
+
sx
.
+
6
ffiffl
»n
?
M
s
^;2:
M
s
H2:
V
'3
3
Jj
,
^
(5
8*
u
u
M
6
1
^
?
i
h4^
1
^
^x
J?
W5
^
d
ulz:
§
02;
1 :
1;
•3 •
•
*» .
a -
^ ;
^ :
^ :
3 :
:
:
^
0/-V
*i
?i
d g
a
J'S
Q s^
rl
B|
2I
§^
i'S
1^
•0 1;
c
d
U2
Qz
Digitized by
Google
Jacob Bronfenbrenner and Hideyo Noguchi. 619
ant NajCOs or NaHCOs ^^ longer exerts the same effect as the free
sodium hydroxid and carbon dioxid, since these are readily dis-
sociable in the fluid. But one must not forget that in the stage of
reactivation the fluid is now made isotonic with sodium chlorid, and
this restrains the dissociation of carbonates or phosphates suffi-
ciently to prevent the free action of the carbon dioxid or sodium
ions upon the complement fraction. It is important to notice that
an acid has no reactivating property for the end-piece prepared with
carbon dioxid, while exactly the opposite is the case with the end-
piece made with hydrochloric acid, which is an inactivation by
alkali.
In the experiments in table XIII we show that amphoteric
substances capable of reactivating complement inactivated by car-
bon dioxid lose this reactivating property if treated with carbon
dioxid before being added to the inactive complement. The way