9 and 18).
SUMMARY.
The paroxysm of hematin intoxication in the rabbit undoubtedly
presents many features of striking similarity to the paroxysm of
human malaria; still one must hesitate to apply such results unre-
servedly in an attempt to identify the causative agent of the malarial
paroxysm. When, in addition to the character of the paroxysm,
we consider the sequence of events in the two instances, the analogy
becomes so close that it seems impossible to regard the matter as a
mere coincidence. The injection of hematin, especially in fractional
doses, is in a measure comparable to the liberation of hematin into
the human circulation by the malarial parasite. In these experi-
Digitized by
Google
Wade H. Brown.
691
I
J3
J3
& I
1^
K
00 .a
<^ B
•s
J3
A^UDOJI E
f<)Ooooqoq^MC^
ooooooooo
00 O 00 00
W ^ PO ^
o o o o
[r] t
ei c« M 6
M « O o q CO
fO M ci fo ^ CO
•uopn]os IPS ^^
-iioqj«3{g amipos
ipiM axmwad
-OI»X JO oop«A9{3
00 00 io
M M e«
Mm«J»diiiax
00 CI M o^yQ %n ^ rooo
« « « M
<nrx ^ up»u»H
o q q q
coo ro «
o ro q^ q^oo ^
« M ro r^ t^ d
q io vr» to q q
»o w ti ti lo 6
to 10 q
•D-o 01 noimios ip»s
ai«aoqj«3)g arafpos
9 9 9
6 6 ro
•oopoataijoarKi
i mi
o^o^OkO^OkO^o^ao^
M M M
0« O^ O^ CK
V « « «
fc fc fa fa
!rnaniVJO-OK
•I-
••a .
2
.•a
2
Digitized by
Google
592
Malarial Pigment in Malarial Paroxysm.
1
2to .
il
I
5 .
S 3 a
§11
ill
Us
X
i
<H
'aop^dftii
ajnituddmax
r* ^ r* o w ^
-max JO aoptiA9ia
CO CO
CO CO
CO r* o r* M
CO ci ci CO ^
-aoqjvoig rnnipos
■"*JL P aop»A9ia
•aopaaftxi
JO ainix l»
00 O CO
M « ei
o o o
\n CO too M M
e« M M e«
•8m u\
»q8»M Xpogjo
o q
f- CO
4o
o o o o o
e« ci ci c« ci
•O'DOl
uopnios opvoidH
o o
o o
00 00
o
>6
o o o q o
e« e« e« e« CI
•o'D ai aopnios IP'S
o o o o
00 o >o 6
•uopoafai JO wbq
M
&&&
sas
o^ o^ a o^
M H M M
M ei CO ^
OiOlOi Ot 0>Ol
II
lllll
•toim in iqapM
: psaniY JO 'oj^
•a .
I"
2
•^5 5
<5> «
Digitized by
Google
Wade H. Brovm.
593
1
in
1
1^
Hematin solution s)
months old, same as
No. 7. injection 3. and
No. 8. injecUon 3.
Hematin solution, same as
No. 10.
Ox hematin in all other in-
Ox hematin.
In two doses 30 min-
utes apart. Initial rise
checked by second dose.
Apparently about the
optimum dose.
•uoimfhi
»ini«j»diiiax
in
?
M
000
M M
tno « 00 00 CI CI
<>oq.o^> ^ OM
MMMMM M MM
<nunoHa!9aiix
M M
»«« mt* p<WK»
^4M^4R ^m^^^^^ ^^^ ^^'••^^^
ip}iiain)«i9d
-«»xjooop»A»ia
53
1:?
0^ tn c«
0' c< CO CO ^ ro
i{){A aniimjod
M CI
00 in
M M
o o
M M
SS
M M
>o>o nn t^ »noo
sssssssi
wfoqt^in ^ tn^
MMMMei M MM
00000 00
MMMMM M MM
'9m xn
«II21 j»d a!9«ai»H
o o
ci 00
M M
6 d
tn 00000
K. c{ 06 6
M M CI M
qqqoo q qq
moo 0^0 00^
M M CI M CI M M
•0*0 nf
f% 00
M M
M M
00000
in ^ e« 06 6 06
M M M
M M M M
•ODnniopii|osii»s
CI 06
M M
nH>!1MftlI JO 9i«a
M M
M M
O^ Oi
M M
ii
M M
M M
M M
ini^
MMMMMMMM
o^o^o^o^o^o^aa
MMMMMMMM
ti co^oo 6 t*i^9>
iiiiiiil
Nov. 10. 191 1
Nov. 13. 1911
Nov. IS. 19"
Nov. 17, 191 1
Nov. 30. 191 1
Nov. 33. 191 1
Nov. 34,1911
Nov. 37, 191 1
.'ITOinivjo-OK
6> ^
No. 10.
Weight.
i.8s.
No. II.
Weight.
3.0.
No. 13.
Weight.
3.0.
Digitized by
Google
694
Malarial Pigment in Malarial Paroxysm.
iiiiii
m m tt a a ^
9 9
...AH I.
III III I
I 1 .1 .1*
*>0>p>0^0>s8ff8
GO O O O O
A^txnoH £
anqmodmax
so CI i^i^oo^^ae o in^
ooooooooooo
<n smoH Of •ia]x
â– M ii H i< H i<<i< H i<>
*a|i«uia|{
-a»X JO oopiM»|a
in fo
M to
O M C^yq to CO
CO « « « CO ^
fO fO «
•aopnios Ji»s »»
i{){Aajm«x9d
-n»x JO aop»Aaia
â– loatuixw
cioei«ooor»v>MV>^
S S o o o
•9ta Of
o o
6 6
o o o o o o
9 ^'^ 9
•O'DUj
iio|)ii{os opitnnaH
ro M q O O O
•^ n 6 6 ci ^
M ^ M M
« CO ^
O
^
9l«iio<pK>ia omipos
O ro ro
4 c*
*oopa»{tal >o aivQ
MMMMMMMMMMM
« ^ti»^O0d MfO»000
mi
O^ O^ O^ a a ov
M M M M M M
O 00 M CO tn 00
:i»mnivjo*oM
.1.
fO'g' in
Z
Digitized by
Google
Wade E. Brown.
596
<5
aumiiuadaiax
>0 i/> rooo ^ ^
O t^ ro O
ci « CO ^
O O O O
ro ^ ^ 00 f*5 O
O O O O O O
O O
'IUUL8l4flVJ
-n»x JO oopvAai^
« ^ o in ^
fO fo fO w ct
in e« t^ CI
M M « fO
q^oo •-• o^ c«
fO W fJ W fj
ro CI ei
-noqnoig ranipos
-a»x JO aop«A»ia
*iiop39ftxi
ft wmx vt
ajnnMadai9X
O 00 M o c« o
S S S e o o
<) in »noo
M M M M
O O O O
MO 00 M o tn
M p M e« ci c«
o o o o o o
i«i«PM ^poa JO
oira
o o o q o
6 6 6 6 6
o o o o
6 6 6 6
o e o o o
6 6 6 6 6
o o o
6 6 6
oopmog a;i«uiaf{
c« e« ei e« e«
CO fO CO CO CO
C« C« C« CI
CO CO CO CO
CO CO CO CO CO
CO CO CO CO CO
CO CO CO
fO CO CO
•0*0 Of aoptqos i(»s
9tTrao(|iB3]g mnipog
'OOIlOOQil y^9fQ,
CI «« c« c< c« e« CI
M M M M M M M
a a a Ov Ok a a
lis nil
CI CI CI CI
ei CO ^ »n
CI CI CI c«
iiii
e« n CI CI CI CI CI
& & St & & Si SI
CI « CI
M M M
Oi Ok a
eikOt^oddkOM ei" CO ^
MMMMHC1C1 eieici
CB (9 CB w w ^^ (9 (9 ^^ CB
llTOiniVJOO|»i
Digitized by
Google
696
Malarial Pigment in Malarial Paroxysm,
06
Ox hematin.
Temperature reached 105^
F. in I hour.
Injection here with nc
record.
Qz hematin.
Injection here with no
5
I
4
CI CO t"-^ CO CO
St iiss
MM M M M M
CI ro ro M
ssts
M M M M
CI q too 00 c^
M M M M M M
sist
M M M M
oi tjnoH ui aoiix
M CO M M M M
M M M M
ct fo'^Hi'^rtr
M M M M
-u»x JO aoji»A»ia
CIOOO
^ M fO CI CI
00 aooo
CI CI CI CI
^00 too ^
CO ci ci c« CO
00 00<> fO
cow ro CO
-awiCTOia uraipos
-n»X JO oop»A»ia
<>
CI
•oopD»ftii
K) araix l»
(•J) aji««J9dn»x
c« CO ^^00
MM M M ci ci
00 0000
MM M M M M
ci CI CI CI
0000
M M M M
*nyq q CK CO to
M M ci M ci M
000000
M M M M M M
A c« in r«
ssss
M M M M
'8m n\
'»l»PM^Poajo
0000
lo to «o lA in
M M M H M
to to to to
M M M M
q q
10 to to to to
M H H M M
M M H M
•O'O UJ
«0 10 10 to V)
:^u:^
to to to 10 in
in in in in 10
tntninm
in to to 10
to
in
to
•oopDaJtaijowa
WW n CI CI CI c«
MM M M M M M
MM M M M M M
CI t^OO di d M
MM M M M C« CI
ii illll
Jan. 22,1912
Jan. 23,1912
Jan. 24,1912
Jan. 25, 191 2
CI « CI CI CI CI CI
M M M M M M M
0^ a a a a a a
H M M M M M M
CI t^oo 6^6 *S
M M M M M CI CI
iiiiiii
« « CI CI
M M M M
M M M M
CI CO ^ in
M C« CI CI
4444
rpraiHivjoo^
No. 17.
Weight,
1.79.
No. 18.
Weight.
1.84.
Digitized by
Google
Wade E. Braum. 697
ments, both solution and finely divided suspensions of hematin have
been found equally effective in eliciting the phenomena of the
paroxysm, and while it seems possible that a portion of the malarial
pigment might be dissolved in the alkaline human senun, such an
assumption is probably not essential.
It might be objected that the toxic action of foreign hematin thus
injected into the circulation would probably be greater than that of
hematin derived from an animal's own blood, but as far as I have
been able to determine, this objection does not seem valid, as rabbit
hematin, dog hematin, and ox hematin produce in the rabbit effects
that are alike in both character and degree.
The dose of hematin remains as the one factor to which it is
possible to attach some degree of uncertainty, but even here the
author feels that the range of experimental conditions has been
kept within the bounds of legitimate analogy with conditions exist-
ing in the human subject of malarial infection.
Finally, the most conservative estimate of the value of such
experiments points strongly to the fact that we have at least a
potentially toxic substance in the pigment hematin as liberated by
the malarial parasite into the circulation of the human host.
There is also abundant evidence to show that the action of
hematin is not confined to the paroxysmal phenomena of malaria,
but that other features of the disease may fold their explanation in
the action of this pigment. For the present^ however, it seems
advisable to confine the discussion to this one phase of the question.
CONCLUSIONS.
1. Alkaline hematin in doses commensurate with the amounts of
hematin liberated in the human circulation by the segmentation of
the malarial parasite, produces, when injected intravenously into the
rabbit, a paroxysm which is characterized by a short prodromal
stage, a stage of chill and rising temperature, and a hot stage. In
their details the phases of this paroxysm are practically identical
with the corresponding ones in the paroxysm of human malaria.
2. The phenomena in human beings infected with malaria are, at
least in part, directly referable to the toxic action of this malarial
pigment.
Digitized by
Google
Digitized by
Google
A BIOCHEMICAL STUDY OF THE PHENOMENA
KNOWN AS COMPLEMENT-SPLITTING
Bt JACOB BRONFEN6RENNBR and HIDEYO NOGUCHI
Digitized by
Google
[Reprinted from the Journal of Experimental Medicine, Vol. XV, No. 6, 1912.I
A BIOCHEMICAL STUDY OF THE PHENOMENA
KNOWN AS COMPLEMENT-SPLITTING.
First Paper : Splitting of the Complement Associated with
Globulin Precipitation.*
By JACOB BRONFENBRENNER and HIDEYO NOGUCHI.
{From the Laboratories of The Rockefeller Institute for Medical Research,
and the Laboratory of Biological Chemistry of Columbia University,
New York.)
INTRODUCTION.
During the last few years much attention has been directed
toward a phenomenon known as the splitting of complement. The
conception is that complement, when treated according to certain
procedures, can be decomposed into two definite portions each of
which is by itself inactive but can act in combination with the other
as complement. The two components of the complement differ in
their behavior in hemolytic processes; one has the property of being
bound by the sensitized erythrocytes, while the other possesses a
lytic action upon sensitized cells that have been acted upon by the
first component. Thus, according to certain investigators, comple-
ment is composed of an intermediary portion, called the mid-piece,
and a l3rtic portion, called the end-piece. The terms "mid" and
" end ** are employed because the conception is that the mid-piece
stands between the sensitized cells and the lytic portion of the com-
plement, while the lytic portion when viewed graphically, becomes
the end-piece.
In view of the high lability of complement it is natural to ask
if the phenomenon of splitting brought about by certain procedures
may not be due to some modifications of the complement, such as
a reversible inactivation through changes in the reaction of the
medium in which the complement is present. It appeared to us of
interest to approach this question from a biochemical standpoint.
In our first paper we deal with the splitting phenomena which are
* Received for publication, April i, 1912.
•598
Digitized by
Google
Jacoh Bronfenbrenner and Hideyo Noguchi. 599
associated with the precipitation of the globulin fraction, and in the
second with those which occur without visible change in the physical
conditions of the serum proteids.
HISTORICAL.
In fresh blood serum Landois found a hemolytic action upon alien corpuscles,
and Fodor, Nuttall, and others, a bactericidal property, but none of the early
Investigators made exhaustive studies upon the nature of the active principles.
It was Hans Buchner who first pursued the study in a systematic manner and
threw much light on this interesting phenomenon. Buchner (1890) established
the fact that the active principle causing bacteriolysis or hemolysis is very labile
and can be made inactive by a temperature of 55® C, by dialysis, or by dilution
with distilled water. The active principle was designated by him alexin.
Bordet (1899) subsequently found that the alexin of Buchner is composed
of two distinct principles, one a sensitizing substance and the other a thermo-
labile alexin. Somewhat later Ehrlich and Morgenroth (1899) made a similar
observation, but by Ehrlich the sensitizing portion was called the amboceptor,
and the alexin the complement In this study Ehrlich and Morgenroth mention
that the complement becomes inactive when mixed with water. Markl (1902)
observed that sodium chlorid in concentration above 3 per cent suppresses the
activity of the complement.
While the disappearance of the complement activity after heating the serum
to 55® C. was recognized by all as due to the destruction of the active principle
at this temperature, no one was able to explain why the complement should lose
its action in distilled water. This question was difficult to approach inasmuch
as in a hypotonic salt solution a hemolytic experiment could not be carried out.
Ferrata (1907), however, utilized the well known fact (noted by Hamburger,
Madsen, and Arrhenius), that the corpuscles can be kept intact when suspended
in an isotonic solution of sugar, and he studied the action of the salt-free comple-
ment (dialysed) in a salt-free isotonic medium, and found the complement to be
inactive in this medium. Ferrata found also that during dialysis the serum
separates into two portions, one the serum globulin (precipitate) and the other
the albumin fraction (in solution). When tested separately in an isotonic salt
solution neither of them caused hemolysis, while when united the action of com-
plement was restored. He affirms that the component of complement remaining
in the albumin portion is inactivated at 55® C.
The finding of Ferrata was confirmed by Brand, who showed further that
the component carried down with the globulin fraction attaches itself to the
sensitized corpuscles and that the component in the albumin fraction will after-
wards act upon the cells so altered, He introduced, therefore, the terms mid-
piece (Mittelstuck) for the former and end-piece (Endsttick) for the latter.
According to Brand the mid-piece of complement remains active when kept
as precipitate (globulin) in distilled water, but loses its property on standing in
salt solution. Brand and Tsurusald (1908) foimd the mid-piece as well as the
end-piece inactivated at 55* C.
Hecker (1907) and Sachs and Bolkowska state that the binding of the mid-
Digitized by
Google
600 Biochemical Study of Complement'Splitting.
piece takes place at o^ C. as well as at higher temperatures, while that of the
end-piece occurs only at higher temperatures. Thus they claim that the selective
absorption of the natural hemolysin by means of corpuscles at o** C. effects not
only the fixation of the amboceptor, but also the mid-piece of complement.
Sachs and Teruuchi (1907) observed that the complement loses its activity
in distilled water or isotonic sugar solution when the degree of dilution with dis-
tilled water is about i : 10, but not in much higher dilution. They were inclined
to believe that the inactivation of the complement was caused by a certain fer-
ment of the serum capable of exerting an injurious action only in a salt-free
medium. They state that the inactivation of complement in a salt-free medium
does not take place at a low temperature or in the presence of an excess of
amboceptor. Complement a few days old does not undergo similar alteration
(inactivation with water, etc.).
Tsurusaki found that the so-called splitting of complement is quite incon-
stant and seems to depend upon the quality of the dialysing membranes, upon
the variations in different individuals of the same species of animal, and upon
the temperature. For instance, the unaltered complement of rabbit was found
in the supernatant fluid when dialysed at o^ C.
Sachs, and Sachs and Altmann (1908) then introduced a new method of
splitting the complement. They used a weak solution of hydrochloric add by
which a certain portion of serum is precipitated (globulin). The precipitate
corresponds with the mid-piece, and the supernatant fluid with the end-piece
obtained by dialysis.
Later Michaelis and Skwirsky found that when corpuscles, amboceptor, and
complement were put together in a medium containing a certain quantity of
NaHiPO* and NaJiPO* in proportions resulting in isotonicity, there was no
hemolysis. But on further analysis, they found that in this mixture the cor-
puscles had fixed not only the amboceptor, but also the mid-piece of comple-
ment The corpuscles so modified were called persensidzed, because they
readily underwent hemolysis when the neutralized supernatant fluid (containing
the end-piece) was added. In other words, persensitization can take place in a
medium containing an excess of NaHjPO* which suppresses the action of the
end-piece, while the activity of the end-piece can be restored by neutralizing the
excess of the acid phosphate with NasHP04.
Liefmann and Cohn introduced carbon dioxid for splitting the complement^
whereupon the mid-piece is precipitated and the end-piece remains in solution.
This phenomenon occurs only in a serum diluted with a sufficient quantity of
water. They studied the influence of certain acids and alkalis and concluded
that carbon dioxid exerts a far more inactivating effect than the acids or
alkalis studied by them. Cholesterin inactivates the mid-piece in a salt-free
medium. From the fact that the persensitization of corpuscles is affected with
a minute amount of the mid-piece, they were inclined to think that this com-
ponent acted like a ferment, an assumption held also by Ruszny4k.
Jacoby and Schutze state that the complement inactivated by shaking and
also the precipitate that is formed may be reactivated either by the mid-piece
or the end-piece. On the other hand, the supernatant fluid of the shaken com-
plement becomes active only with the end-piece.
Complement is known to be carried down by certain colloidal substances
Digitized by
Google
Jacob Bronfenbrenner and Hideyo Nogttchi. 601
through an adsorpdon (Landsteiner). Skwirsky confirmed this and points out a
very interesting difference existing between the colloidal adsorption and the specific
complement fixation as it occurs in the Bordet-Gengou or the Wassermann
reaction. According to this author, later confirmed by Amako, in the former
instance the end-piece as well as the mid-piece is carried down from the solu-
tion, while in the complement fixation the mid-piece alone is absorbed, leaving
the end-piece free in the supernatant fluid.
Landsteiner, Liefmann and Cohn, and Frankel state that the mid-piece pre-
pared by the hydrochloric acid or carbon dioxid method from sera of different
animals can persensitize the corpuscles for the end-piece of guinea pig serum,
but not vicg versa. This phenomenon is ascribed by them to the presence of
natural amboceptor in the fraction containing the mid-piece. Frankel points
out certain differences in the behaviors of mid-pieces derived from the sera
of different animals when these are acted upon by cholesterin, or sodium
chlorid, or inactivated by the temperature of 57® C.
Marks observed that the reactivation of the end-piece by the mid-piece occurs
when the latter is employed in a smaller quantity than the ratio of i : i. He
found that the heated as well as the fresh serum can furnish an active mid-
piece, but a prolonged heating of the mid-piece at 55® C. finally destroys its
function.
Liefmann studied the relation of certain lipoids and soaps to the mid- and
end-pieces of complement and found that they have no effect upon the end-piece,
but that they destroy the mid-piece. He brought out also the fact that per-
sensitization can not take place unless the amount of amboceptor reaches nearly
fifty units.
That the active principle of complement undergoes an inactive modification
in the presence of various acids, alkalis^ and salts has been carefully studied by
Noguchi, von Liebermann, Hektoen, and others. According to the experiments
of Noguchi, the complement is permanently injured by a strong solution of
these substances, especially by the acids and alkalis, but when used within a
certain Hmit, the activity can be restored with but slight loss of power.
EXPERIMENTAL PART.
There are at least five different procedures by which it is claimed
that complement may be split* into two components. Sachs employs
hydrochloric acid; Liefmann, carbon dioxid gas; Ferrata, dialyzes
against distilled water; Michaelis and Skwirsky use a mixture of
acid and alkaline phosphates; the fifth method is that of splitting
by complement fixation.
'"Complement-splitting" is the term generally adopted by English speaking
bacteriologists. This is the reason for the employment of the term in this
paper, although a better word might be substituted, and the following work
almost necessitates its replacement by the term " inactivation," since this is what
has really been proved to take place in almost all cases of so-called splitting.
Digitized by
Google
602 BiochenUcal Study of ComplemenUSplitting.
In our first paper we shall consider only the first three methods.
In these the globulin fraction of the serum is precipitated during
the process of splitting.
COMPLEMENT-SPLITTING WITH HYDROCHLORIC ACID.
The procedure introduced by Sachs is as follows: One cubic
centimeter of guinea pig serum (complement) is mixed with 8.2
cubic centimeters of N/2S0 hydrochloric acid solution (in distilled
water), and after standing at room temperature for one hour the
mixture is centrifugalized, whereupon the precipitate settles down
to the bottom leaving a clear supernatant fluid above. The super-
natant fluid is carefully separated from the deposit^ and is mixed
with 0.8 of a cubic centimeter of N/25 sodium hydroxid solution (in
ID per cent, sodium chlorid). The addition of alkali is intended to
neutralize the acid introduced into the serum for splitting, while the
resultant solution is made isotonic at the same time (about 0.9 per
cent, sodium chlorid). The deposit is washed with water and
finally collected by centrifugalization, and then dissolved in 10 cubic
centimeters of 0.9 per cent, sodium chlorid solution. The super-
natant fluid and the precipitate, both now in clear solution, repre-
sent the end- and mid-piece, respectively. Thus, the supernatant
fluid has no action upon the sensitized corpuscles unless the precipi-
tate portion is added, and vice versa. Apparently the complement
is separated into two components.
Experimental evidence was also brought forward to show that
the end-piece or supernatant portion carries the labile zymotoxic
component, and the mid-piece, or precipitate portion, possesses the
property of enabling the end-piece to act upon corpuscles sensitized
by amboceptor, and that it is not destroyed by the temperature of
56" C.
The above experiments have been confirmed by various workers,
but we have been able to corroborate the results in part only, inas-
much as the supernatant fluid was found to be active upon sensitized
corpuscles in most of our experiments (table I, experiment II) and
was very seldom inactive.
' Finding that the supernatant fluid was often active without the addition of
mid-piece, we took the precaution of filtering the end-piece through a hardened
filter paper in order to be certain of the absence of the mid-piece.
Digitized by
Google
Jacob Branfenbrenner and Hideyo Noguchi.
TABLE I.
Cmnplement^SpliUing by Hydrochloric Acid.
Experiment I {Ideal Case)}
j808
GviMa piff teram z cc. + N/aso HCl 8.a cc. (in dbdlled water).
Gninea pic serum
zccH-ol^KliIaCl
9C.C
Sopwnatant fluid
+ o.8c.cN/s5
NaOHinzoperctfit.
Depodt disMhred
in M cc of
0.9 per Mat. N«a.
both ingrtdiems.