y< . : » if 4 d f
Aki tc L
* .) * n° i
Foe ay.
« LAN are
4
se
mt
aia ag
ee"
"OK
.
*
4
it
’
‘
s
i
'
LPP wx
m7
a
*
.
'
.
'
.
ian’.
>,
i
T a
.
at
bat
:
‘fon.
» #
SEEM teat
cee oF Wy ao ae |
‘
oe e/a,
= oe @AGerid yt
~~
: 7
4 - ‘
n
y ty
‘ j
yn
ia
% fs :
t ¥ = iy ,
y ;
A <1 “4 é
“ \ ns
i ; ee?
walt 6] See “ i .
ay 2 o *
i. Ad i" 4
~ 4 E ,
OF
IMMUNOLOGY
VOLUME V
rretty af Torvaty
BALTIMORE, MD,
1920
A\BRAR p
JUN 15 1964
WA of
% &
E5i7y 9 WA
906198
CONTENTS
NuMBER 1, JANUARY, 1920
On the So-Called Neisser-Wechsberg Inhibiting Phenomenon in Bactericidal
RRR aE mem ee TIO UGA boi cons aye mt PANES Siigini® mo. 48 +, +s Wine 22 wis aap 1
The Relation of the Rate of Absorption of Antigen to the Production of
eeemrreu ome VEAEIGTIC WV.) COOK. «<<. os. vt acca cis tie ce voce vee sens 39
Studies on the Meningococcidal Activity of Blood. Toitsu Matsunami.... 51
NuMBER 2, Marca, 1920
Natural Antihuman Hemolysins and Hemagglutinins in Horse Sera in Rela-
tion to Serum Therapy. John A. Kolmer and Motomatsu Matsumoto... 75
An Attempt to Produce Specific Immune Agglutinins and Hemolysins for the
Four Groups of Human Erythrocytes. John A. Kolmer and Mary E.
SO aie Re aso, msgs ope Ee od x8 9,4 eb me Re Se 89
A Comparative Study of Methods for the Preparation of Typhoid Agglutin-
TELE. USGL DE IS C1006 | SE ee See 97
A Study of Different Methods for the Preparation of B. typhosus Antigen.
MUMoe hears cous UM Car SUIT OU Oe eh scysiy ta tishe cons << o!> ses, 0 wienehebet eee otetole ® Sys eels BEN s 111
An Experimental Study of the Effect of Autogenous B. coli Vaccines on the
Intestinal Colon Bacilli of Dogs. John C. Torrey and Alfred H. Rahe.. 133
Experimental study of the Sensitized Cholera Antigen. Y.Miura.......... 145
A Dropping Bottle as an Aid in Macroscopic Slide Agglutination. Charles
oo TSHERE Ee. 2 2 Se 0 er 155
The Complement Fixation Test for Tuberculosis. Hassow O. von Wedel.. 159
Numper 3, May, 1920
On the Transfer of the So-called Normal-Antibodies from Mother to Off-
Seen APH GIning.«. Gr. OC. ROYMAM - . 6 sce <6 cow oe ee sec ce ect aes 227
Studies in Anaphylaxis. The Relation of Certain Drugs to the Anaphylactic
Reaction, and the Bearing Thereof on the Mechanism of Anaphylactic
SIRT, INEST ES RRS nH 8 pee RS ss. on) SA ae a 230
The Antigenic Properties of Hemocyanin. Carl L. A. Schmidt............ 259
On the Nature of Bacterial Toxaemia. Hans Zinsser. .................... 265
Studies in Anaphylaxis. I. On the Quantitative Reaction of Partially
Neutralized Precipitin in Vitro and in Vivo. Arthur F. Coca and Mitsuji ©
Ie ee... Ss, Cerne hae civ te aE ae cas os 4 = 297
NuMBER 4, Juty, 1920
A Study of the Precipitin Test in Cases of Pneumococcus Empyema. Cleve-
De ONES Hepat d tC ke dies Eire dA Stee ee 321
iv CONTENTS
A Nephelometrie Method of Estimating the Number of Organisms in a Vac-
ame. Gearre 6. Dunham. oe. ee eee eae in os a Gee eee ee 337
Effect of Ultraviolet Rays on Antigenic Properties. I. Studies on Meningo-
euceus. Hrederick Hberson:: .'. i.cc2etecaes oes cet coe ee ee 345
Hypersensitiveness: Anaphylaxis and Allergy. Arthur F. Coca........... 363
The Relation of Sputum Bacteria to Asthma. Francis M. Rackemann.... 373
Some Observations on the Constitution of the Complements of Different
Animals: “Pe J! Miaekic -0;.; <25 eres ee vere areal tonls ee 379
On the Placental Transmission of So-Called Normal Antibodies. II. Anti-
iryptic-Acting: Bodies. G:C: \Reymanin..0 6 oe oc. cs. see ess cone 391
NuMBER 5, SEPTEMBER, 1920
Simplification and Partial Revision of the Factors Involved in the Comple-
ment Fixation Test for Infectious Abortion in Cattle. Charles 8. Gibbs
and ‘Leo H' Retiger.. ....: oscpeeeen cor a < Ain en noe one Oe 399
The Antigenic Properties of Globin, with a Note on the Independence of the
Properties of Serum and Tissue Proteins, as Exemplified by the Absence
of Antibody from the Globin of an Immunised Animal. C. H. Browning
and Go HaswelloWilson )jemerces ches cnr eee One eT ee 417
The Protective Value of Pneumococcus Vaccination in Mice and Rabbits.
ATIPUSHIS, "b.- WAGSWOFbH ceeite sc. cs cele oreo nee ecg Rene eee 429
Serological Relationships of Liver and Kidney. Moyer S. Fleisher, T. G.
Halland Natalie-Arnistein Sens. citer reece 0 a a oe ORE 437
On the Placental Transmission of So-called Normal Antibodies. III. Anti-
lysing. “GC. Reyman . 23: Sie eetcles os os bc ote ee eee 455
A Serological Study of Cholera Immunity. I. Agglutinin. Rokuro Ume-
PIU Bed referee: nice so S a ayes ore ease oeroe eee RE MERD Clinse ote, eve a0 soe snore) cero ee eee Rete 465
The Value of the Intra-Palpebral Mallein in the Diagnosis of Glanders.
Edward-H. Mason and’ R. Vi Beimmons. 6). 5s ees oe eee 489
NuMBER 6, NOVEMBER, 1920
Comparison of Smear, Culture and Complement Fixation in Chronic Gonor-
rhoea in Women. A Preliminary Report. James D. Smith and M.A.
RY MON oes as hic ole «ens, syahs «6 se oR He ce > esos ets ce ge 499
Experiments Upon the Production of Antihuman Hemolysin with Special
Reference to Immunization with Erythrocytes Sensitized with Heated
Serum. Motomatsu,Matsumoto. : yteeecs oie a -- ccna ele eee 507
A Note on the Non-Specific Production of bodies. Motomatsu Matsu-
ER OUO 2s oA ioe Sates 3 eee 8505.4 2 > 3 oc5 0 ORME ear oie Sn nee eee 517
The ene oaee Precipitation Test for Syphilis. Thomas G. Hull and
Hive. Baighte oo. oss co.cc:- + > 21.) a Rene Sole ae ee 521
A Study of the Mechanism of Human Isohemagglutination. Herbert L.
Kooeckert 5.030.600 23 S36 4aes,0 2 wre oe ole renee 529
Immunization with Blackleg Aggressin. Thos. P. Haslam................. 539
A Study of the Specificity of the Absorption of Anti-Bacterial Precipitins.
Charles Krumwiede and Georgia M. Cooper....................0.00085 547
ON THE SO-CALLED NEISSER-WECHSBERG INHIBIT-
ING PHENOMENON IN BACTERICIDAL
IMMUNE SERA
TH. THJ@TTA!
From Dr. F. G. Gades Pathological Institute, Bergen, Norway ;
Director, Dr. M. Haaland
Received for publication December 15, 1919
A. HISTORICAL
In the year 1901 Neisser and Wechsberg described a peculiar function
of immune sera, which they called ‘‘Complementablenkung,” and
which later has been named the ‘‘ Phenomenon of Neisser and Wechs-
berg.””’ The phenomenon consisted in the observation that, whereas
small or medium doses of bactericidal sera exhibited the bactericidal
function with the homologous strain, the larger doses were without
effect. Thus an inactivated immune serum against Vibrio Metschnikoff
with added complement, showed bactericidal action in doses of 0.05 to
0.0025 cc., but in larger doses it had no such effect. Likewise, Neisser
and Wechsberg showed that the dose of complement-bearing serum,
which was sufficient to activate a certain dose of an inactivated serum,
was not capa»le of doing so when larger doses of immune serum were
added.
The explanation of this paradoxical serum function was sought, by
the authors, in the great richness of the immune serum in bactericidal
amboceptors. These antibodies, which, in the immune serum, must be
present in far greater quantities than the complement in the activating
serum, were supposed to unite with the complement and form thereby
a lysin that dissolves the bacillus after becoming attached to it; but on
account of the predominance in the number of the amboceptors, there
being insufficient complement to satisfy all of the amboceptors, some
of the latter were assumed to remain as unaltered amboceptors, free
of complement, and thus devoid of bactericidal function.
1 Director of the Bacteriological Laboratory of the Norwegian Army.
1
THE JOURNAL OF IMMUNOLOGY, VOL. v, No. 1
2 TH. THJOTTA
When under these conditions, fixation between the serum bodies
and the bacilli takes place, and if we assume that the bacilli have no
greater tendency to attach themselves to the lysin than to the free
amboceptor, the possibility may be considered that some of the bacilli
will combine with the lysin and will be dissolved, while others will
unite with the free amboceptors and will not be affected. The latter
bacilli will survive, and the test will not show bactericidal function.
This result would be the more likely under the assumption that effective
lysin has a lesser avidity for the bacilli than the free amboceptor. If
this supposition is true, the greater number of bacilli will be attached
to the ineffective, free amboceptor, while the effective lysins will remain
unused, and these will have no opportunity of bringing about the
bactericidal effect. When, on the other hand, the amount of ambo-
ceptors is small, as in a diluted serum, all of the amboceptors will be
supplied with complement. Consequently, there will be only effective
lysin in the test and the bactericidal effect will occur.
Thus, according to Neisser and Wechsberg, the surplus of ambo-
ceptors in an immune serum brings about the inhibition of the bacteri-
cidal action, because this surplus makes the effective utilization of
complement impossible (‘‘Complementablenkung’’).
Contrary to Neisser and Wechsberg, Gruber supposes that the
immunization calls forth antibodies that act antibactericidally and
antihemolytically. He, therefore, denies the significance of the surplus
amboceptors.
Lipstein, however, entertains the same opinion of the phenomenon
as Neisser and Wechsberg, and shows that the phenomenon is of a
strictly specific character, that it is not called forth by normal serum-
bodies, and that the agglutination does not play any part in the origin
of the phenomenon.
By absorbing the bactericidal serum with the homologous bacillus,
Lipstein succeeded in robbing the serum of its inhibiting action. He
therefore supposes that the serum bodies that bring the phenomenon
about, are of the nature of an amboceptor, and he simply identifies them
with the bactericidal amboceptors.
Levaditi rejects the theory of the lysin, showing that serum, wherein
the homologous bacteria have been sensitized, -loses all bactericidal
effect, although, according to the theory of Neisser and Wechsberg,
such serum ought to contain numerous unattached lysins, after the
bacteria are removed in the centrifuge. To explain the phenomenon
Levaditi supposes that ineffective amboceptors (“‘ambocepteur inac-
NEISSER-WECHSBERG INHIBITING PHENOMENON 3
tive’) are produced in the course of the immunization. Furthermore,
he supposes that these amboceptors have a greater tendency to attach
themselves to the bacilli, than the effective ones.
Gay believes that precipitating antibodies arise during the immuni-
zation and that these antibodies, with the homologous antigen, form a
precipitate, that absorbs the complement. He further shows that
the phenomenon of Neisser and Weehsberg also can be found in hemol-
ysis tests.
Sormani likewise observed the phenomenon in hemolysis tests, when
very strong solutions of serum are used for the sensitizing of the red
blood corpuscles, and furthermore, he found that the blood corpuscles,
under these circumstances, could be dissolved mechanically by shaking,
without the influence of complement. He calls this phenomenon
specific fragility (‘“‘specifische Sproedigkeit”’).
Microscopically Sormani could show that the corpuscles, after the
treatment in serum, were shrunken and jagged. He therefore surmises
that something happens to the surface of the corpuscles during the
treatment and that this ‘“‘something” tends to make the corpuscles
more resistant to the influence of the hemolytic antibodies. Hesupposes
that the cause is to be found in a precipitation of the albumin of the
surface and that this precipitation forms a covering around the corpuscle.
Through this supposition Sormani also explained for himself the specific
fragility since it was reasonable enough to assume that the hardened
corpuscles could be broken by vigorous shaking. Likewise, it was a
natural assumption that the hemolytic antibodies could be hindered
in their action, when the corpuscle was protected by a more or less
thick covering of coagulated albumin.
The theory of Sormani, however, has a drawback in a fact that cannot
be easily explained. Sormani himself stated that corpuscles that had
been treated with concentrated serum, showed neither the phenomenon
of Neisser-Wechsberg nor that of specific fragility. He tried to explain
this fact through the theory that the covering of albumin in this case
is so thick, that it cannot be broken mechanically, and that the surface
gets so shrunken that small pores are formed, through which comple-
ment forces itself into the corpuscles and causes hemolysis.
That the precipitating faculty of the serum is of importance for its
inhibiting action, Sormani shows in his statement that the phenomenon
of specific fragility and that of Neisser and Wechsberg, stand in direct
relation to the precipitating power of the serum. Contrary to Gay,
however, he does not believe that the specific precipitates absorb
complement. Thus, even large quantities of a specific precipitate
a TH. THJGTTA
(sheep serum in rabbit immune serum) added to his hemolytic mixtures
did not diminish the force of his complements. Only if a mixture of the
emulsified precipitate and complement was kept in the incubator for
one hour, was the complement inactivated, but not if the mixture
remained for the same time in the room. Consequently the absorption
of complement by the precipitate was not strong enough to explain
the lacking hemolysis.
The phenomenon of the inhibition of the bactericidal action has
also been studied earlier in this institute. Brekke studied the occur-
rence of the phenomenon in sera from typhoid patients. His results
are somewhat contrary to those of Neisser and Wechsberg, as Brekke
finds the phenomenon so seldom (in 22.2 per cent of the sera) that he
does not consider it a specific phenomenon. He also calls attention to
the fact that it is not always possible, as was supposed by Neisser and
Wechsberg, to suppress the bactericidal function of an active im-
mune serum by adding more immune serum to the test, and that a large
dose of a strong immune serum does not always inhibit the action of
complement. é
Brekke, furthermore, showed that there was no direct relation
between the bactericidal and the inhibiting titers of his sera. The
inhibition of the bactericidal function was found in weak bactericidal
sera as well as in strong ones. In several sera the inhibiting function
could not be found, although they showed very high bactericidal titers,
such as 0.0000001.
Since he was dealing with sera, which undoubtedly contained a con-
siderable amount of bactericidal antibodies, Brekke had reason to
claim that all of these sera should exhibit the phenomenon of Neisser
and Wechsberg, if the theory of these authors was correct. As this was
not the case, Brekke rejected that theory and sought another explana-
tion. This he thought to find in the theory of the complementoids.
These bodies are considered to be partly destroyed complements that
possess the haptophore group unaltered, while their zymophore
group is destroyed. They can unite with the amboceptor like a com-
plement, but they cannot act asone. Consequently, the bacilli attached
to an amboceptor-complementoid will not be dissolved but will remain
alive. Now Brekke never saw the phenomenon in active sera, but
only in such that had been inactivated through heating. He therefore
supposed that the inhibiting phenomenon was due to the complemen-
toids that had been transformed from the complements during the
heating. Consequently, Brekke had to consider the phenomenon as
an un-specific one throughout.
NEISSER-WECHSBERG INHIBITING PHENOMENON 5
B. OWN INVESTIGATION
1. Introduction
During the work on the classification of dysentery-bacilli
cultivated in Bergen and vicinity, the bactericidal test tube
reaction was used as a method of separating the bacillary groups.
It then turned out that the dysentery bacilli were well suited for
the bactericidal test and good objects for the study of the phe-
nomenon of Neisser and Wechsberg. This phenomenon occurred
so frequently in the sera employed, that it must be considered
as a regular faculty of the dysentery-immune sera from animals.
It was, therefore, natural to put together the results of these
tests with a special bearing upon the inhibition of the bacteri-
cidal action and to try to find out fixed rules for the appearance
of the phenomenon, and perhaps bring forth facts toward its
explanation. :
Further special experiments were carried out in order to ascer-
tain the dependence of the phenomenon on amboceptors and
complements, its importance for the total bactericidal action of
an immune serum and its variation during the immunization.
Finally, the question was taken up whether the phenomenon is
due to already known antibodies, or is brought about by un-
known ones.
2. Technic
The sera employed were the inactivated sera of rabbits that had
been given repeated intravenous injections of bacillary emulsions in
normal saline solutions. The injections were begun with + to 4 eose of
an agar slant culture and this amount was increased up to several
whole cultures. The complement-bearing serum was obtained usually
from guinea-pigs; in some of the tests, however, fresh human sera have
been used on account of the scarcity of animals.
Before the reaction itself, a preliminary test was always carried out
to determine the bactericidal action of the normal serum as well as its
power of activating the immune serum.
An example of the preliminary test of the normal, complement-
bearing serum is presented in table 1.
TH. THJGTTA
= Complement
hactemerht aantoCrphor
> Qui gen
In the first six tubes of the series are mixed broth, the unit
dose of bacteria (1/8000 of an oese) and the normal serum jin
quantity diminishing by the usual geometrical progression. The
aie
Ww
a“
NEISSER-WECHSBERG INHIBITING PHENOMENON 7
TABLE 1
Protocol of a preliminary test of the direct bactericidal power and of the ‘‘activating”’
power of normal complement-bearing serum
TUBE AMBOCEPTOR | COMPLEMENT ANTIGEN BROTH COLONIES
ce cc. oese drops
1 0 0.05 1/8000 2 0
2 0 00.025 1/8000 2 0
3 0 0.0125 1/8000 2 0
aa 0 0.0063 1/8000 2 ©
5 0 0.0032 1/8000 2 oo
6 0 0.0016 1/8000 2 oo
fi 0.001 0.05 1/8000 2 0
8 0.001 0.025 1/8000 2 0
9 0.001 0.0125 1/8000 2 0
10 0.001 0.0063 1/8000 2 0
11 0.001 0.0032 1/8000 2 About 1000
12 0.001 0.0016 1/8000 2 oo
13 control 0 0 1/8000 2 ©
TABLE 2
Protocol of the titration of a bactericidal immune serum (ambocepter)
TUBE AMBOCEPTOR COMPLEMENT ANTIGEN BROTH COLONIES
ce cc oese drops
1 Out 0.005 1/8000 2 co
2 0.05 0.005 1/8000 2 o
3 0.025 0.005 1/8000 2 o
4 0.0125 0.005 1/8000 2 oo
5 0.0063 0.005 1/8000 2 About 100
6 0.00382 0.005 1/8000 2 0
7 0.0016 0.005 1/8000 2 0
8 0.0008 0.005 1/8000 2 0
9 0.0004 0.005 1/8000 2 0
10 0.0002 0.005 1/8000 2 0
il 0.0001 0.005 1/8000 2, 0
12 0.00005 0.005 1/8000 2 0)
13 0.000025 0.005 1/8000 2 0
14 0.0000125. 0.005 1/8000 2 About 100
15 0.0000063 0.005 1/8000 0 Many 1000
16 0.0000032 0.005 1/8000 2 ©
17 0.0000016 0.005 1/8000 2 ©
18 0.0000008 0.005 1/8000 2 co
19 0.0000004 0.005 1/8000 2 co
20 Complement control 0.005 1/8000 2, ©
21 Saline control 0.005 1/8000 2 ey
8 TH. THIGTTA
mixture in the tubes 7 to 12 duplicate those of the first six,
excepting that an effective quantity of the bactericidal immune
serum has been added to each mixture. The bactericidal titer
of the normal serum (“‘complement’’) is seen to be 0.0125 (tube
3), and this represents the total bactericidal titer of this serum.
Its activating titer lies between 0.0063 and 0.0032.
In table 2 is presented the protocol of the main test of the
bactericidal power of the immune serum, in which the same
reagents that were employed in the preliminary test were used.
The bactericidal titer of this serum is seen to be 0.0000125,
the titer of inhibition is 0.0063.
This is the technic that was employed where the bactericidal
action of a serum against a certain bacillus was tested. If it is
desired to determine only the minimal inhibiting dose of the
immune serum, it is sufficient to use an amount of complement
that we know to be large enough to activate the immune serum.
In this case it does not matter if this dose is so large that it has
a bactericidal action of its own, because, even so, the degree of
inhibition can be determined. In the following tests advantage
was now and then taken of this fact.
3. Specificity of the inhibition
Numerous tests have been carried out according to the de-
scribed method, and the results of some of them are shown in
table 3. ‘The bacteria used were:
(Group I. Bacillus of Shiga
hey f Bacillus Flexner and Strong,
B. dysenteriae } Group II. || jand bacillus ¥.
| Group III.
B. typhosus
Vibrio cholerae
B. metacoli (Bacillus no. 1, of Morgan)
These strains of bacteria have been tested in their homologous
sera and for the greater part, also, in heterologous sera. Further-
more, all, with the exception of the last two, have been tested
in normal rabbit sera.
NEISSER-WECHSBERG INHIBITING PHENOMENON 9
The results of these tests may be stated as follows: Normal
serum shows a considerable bactericidal action against the
dysentery bacilli of group I and IJ, but no inhibiting action can
be discovered in the doses smaller than 0.1 cc. Against dysen-
tery bacilli of group III normal serum has no bactericidal action.
Consequently, the question as to the existence, here, of the
function of inhibition must be left open. Against typhoid
bacilli, normal serum shows some bactericidal power, but no
inhibiting action.
The serum against dysentery bacilli of group I (Shiga) has a
considerable inhibiting action and a strong bactericidal power
against the homologous organism. Against the strains of group
II, it has some bactericidal power and with some of these strains
it exhibits the phenomenon of inhibition.
Against dysentery III and typhoid bacillino bactericidal action
whatever is seen in the serum.
The serum against dysentery bacilli of group II has no inhib-
iting action against the Shiga strain. It will be seen that the
sera showing the best bactericidal action also show the best
inhibition.
Against the strains from group II the group II sera show a
different action. The highest titers of inhibition are always to
be found in tests between a serum and its homologous strain, or
a closely related strain.
As group II is composed of strains that show some individual
differences, it is only natural that there should be some variance
in the results of the inhibiting reaction in this group. Neither
of the sera of group II shows any bactericidal action against the
strains of group III, nor typhoid bacilli.
The group III serum had the same bactericidal action, without
inhibition against the strains of groups I and II as had a normal
serum, but it showed a strong inhibiting and bactericidal action
against the homologous strain from this group.
A typhoid serum showed a normal bactericidal effect without
inhibition against the strains of groups I and II, and no effect
against the group III strains, but it exhibited a good bactericidal
effect without inhibition, against its homologous strain. Another
TABLE 3
B metacoli
A[SNOUSABIZUT SON}
-[NO JUB[S-Ivde FG ‘OT6T
‘IZ IequIsAON 03 ‘TT 19q
-wajydeg wo1j peztunul
~UlT "TTOOGZOW “STL AISA
A[SNOUDABIY
eo
EF | -UL saingino fg ‘OTGT ‘3s
Se | Fouwy OF ‘9T6T “9Z Arwen
Sg | -uep wow pezrunmut
© | ‘(eBrgfOYo) “TOL HAGVY
3 A]SNoueaBIzUL
S | Sainj[no iBBe cyt ‘OTE
& | ‘er Azenuee 03 ‘corer ‘z &3 S S = =
> | 1aq0}0Q Woy pazlunu = s S S S
- | -ur ‘y odAy ‘ 1qqe : - : ° é
Q _ + 16 WAISH (S=>) oo oo oo i=)
A[snousa ss}
ke “UT Saing{no J ‘QTL ‘ZI a 0 © )
z [dy 0} ‘ore ‘zt Azen S =] S
a -use MEO By Sagi S S S
o “ur ‘ 013 ‘66 31qQqB Z ; d Z
S Tar ‘TI dn 663999F | OS oo Ss 6. oS
= N Re) Re)
5 A[SNOUGABIZUI 89IN4 ae) bond —
-[NO > ‘QT6I ‘ZZ Wore Ss =e
az 01 ‘O16, ‘g AsBnigeT S NX se)
wolj peztunmwMt ‘jy SS S =
dnoiz ‘jg ° 1qqu é Sagan Sapa 6 6
3 TE “A FOT F998 oo oo coo oo
B Aj[sno Re)
= -UW9AVI}UL S9IN}[NO 1VBB S Ne) a
2 | Of YI ‘ST6I “LT tequreo N = S
> | -9d 0} ‘STG ‘% Jequrez S Yo} S 19 S
TS | -deg wor} peztunuut —) — a) — a) > >
M | -VIl dn0ip 96. 41498H.| Oo S'S SS) 16 So S'S
A[SNOuUsAGIzUT 8 a te)
sinsdoo} 21 WM ‘OTT | op S = = aS 068
Il [dy 07 ‘9T6I ‘92 se S Sy AS A
Arenuve woly peztunur Ss = i
-UIl ‘(BSI ‘ 1qqz ei ate ° ° Sine G
eG e Glee ee eee SB Se SS
(e@) ie) (o4)
UINIOS 41GQqUI [BULION S S 3 a =
St eee
oo KS) =) oo QS
2 a ‘a a 3 a
|
Be 2 Bee (SESS a Ee |e ae
& OL ors One ORS ore ors
OER - agai 1 RS Ba eS Ege
fe fe 82 28 £2 BS
Ef gs: | Shes. erties | celmesty Wome
2 SOQ AMR FAA AA BQ
ae —S———— se Sa ae
a H tH AQ N
asod INAWA1dNO9 o i) o —) o
i=) o i=) i=) i)
a —
: Bi) Z
yet 5 aml
E 3 Bae Ey yest sper eines
E wn
zg a _ SSS eee aa
a iB _ _
2) = =
: os)
a ———————————————— ae
8
co)
5 S
5 :
a S ~
2 =I
%
in
S
ea)
10
¢20'0
0
[Bplolieyoeg
Aroyiqrqauy
[Bplol1o}OV_
ALOVIGIYUT
eZI0'0 GO'0| [eprotsojoeg
0 QO} Aroqzrqryay
CZI0'O OQ) [BplolsoyoBVg
0 ALOVqry ay
0] Teprorsoyoeg
ATOpquyuy
eZ10'0 OL ere eee
A10}1qIYO]
i=)
10°
10°
¥0
c0
10°
TO"
TO"
TO"
0
0
0
0
0
0
0
Se ed ce OOO uBZ10JA]) T[OOBJOUL “|
JAG EAS Gf
OF “A
OF “A
Wd
Il
**9BIB[OYO OLIGIA
Me Oe snsoydAy “g
eelioyuesXp “g
11
12 TH. THJOTTA
typhoid serum; however, showed both inhibiting and bactericidal
action against this very same strain.
The cholera and metacolon sera showed strong inhibition and
a moderate bactericidal effect against their homologous strains.
On the whole these tests show that the inhibiting effect is
chiefly exhibited toward the homologous strains, but that is also
seen, though less frequently, with the closely related strains.
4. Factors influencing the appearance of the phenomenon
On the basis of the results presented in table 3, we can examine
the importance of the various factors involved in the bactericidal
test with reference to the origin and growth of the inhibiting
phenomenon. These factors are the antigen, the bactericidal
immune serum; i.e., (the amboceptor) and the complement.
As to the antigen, it seems that the very easily dissolved strains
have a special faculty of giving rise to the inhibiting antibodies,
while the strains that are more resistant to the serum bodies
possess that faculty in only slight degree.
In the immunization of the animals it does not seem to matter
much whether very large or smaller doses of antigen has been used.
It must, however, be mentioned that all of these immunizations,
in reality, have been carried out with considerable doses and,
as a rule, by intravenous injection. Between 1 and 10 agar
slant cultures have been used at a time. The bacteria, previous
to their injection, were always suspended in physiological saline
solution and heated for one hour at 60°C.
Rabbits, as a rule, weigh one twenty-fifth as much as a full-
grown man; hence the proportional amounts of the bacteria for
the average man, would be 25 to 250 agar slant cultures. It is
clear that these doses represent many more bacteria than ever
occur in man during a natural infection. It is possible that this
is one of the reasons why Brekke found the phenomenon of
inhibition so seldom; that is, in only 22 per cent of the sera of
typhoid patients.
The dependence of the phenomenon of inhibition on the specific
relationship between the antigen and the antibodies of the
NEISSER-WECHSBERG INHIBITING PHENOMENON 13
immune serum has been pointed out. Furthermore, it has been
shown that normal sera lack the inhibiting function entirely.
Hence it is not reasonable to suppose that the inhibition could
be due to anything in the treatment of the serum after it has
been obtained from the animal. However, the experiment was
carried out to see whether the inhibiting function is influenced in
any way by difference in the treatment of the serum.
A Shiga serum was tested in the fresh condition, then dried
in a desiccator. The serum powder was then dissolved, partly in
distilled water, partly in normal saline. All the tests showed the
same inhibiting action of the serum.
Further, the effect of heating of the serum (one-half hour at
56°C.) on the inhibiting phenomenon was, also, examined and
it was found to be nil; the inhibition was the same in the active
and in the inactive immune serum, provided that the dose of
complement added in the two tests was the same. The phenom-
enon is consequently not due to nor affected by any alterations
in the serum after the bleeding of the animal.
5. Appearance of the inhibition during the immunization
The tests recorded in table 3 showed that the phenomenon of
inhibition is just as specific as the bactericidal action. It seemed
probable, therefore, that the first appearance and the growth of
the phenomenon could be traced during the immunization.
This was done in experiments presented in table 4.
It is seen that the inhibiting phenomenon, which is absent in
the normal sera, appears after the first injection, and becomes
more pronounced during the immunization. In the last test
made four weeks after the last injection, the inhibiting as well as
the bactericidal action is a little reduced.
6. Relation of the phenomenon to the bactericidal power of the serum
It is apparent that the inhibiting action as well as the bacteri-
cidal one is a result of the immunization, but this conclusion
does not justify the supposition entertained by Neisser and
Wechsberg and by Lipstein that both of these functions are
exercised by the same antibody—the bactericidal amboceptor.
14 TH. THJGTTA
TABLE 4
Antigen B. dysenteriae of group II (two different strains employed)
|
SERUM OF RABBIT 109| SERUM OF RABBIT 117
Before treatment............... Jenene ees = Eom a
One week after first injection............ Saas Be nan
One week after second injection......... en gaan
One week after third injection........... alii a
Four weeks after fourth injection........ : iat ein
*| = Inhibiting action; B. = bactericidal action.
If, namely, the bactericidal titer (B) is taken as an indication
of the amount of amboceptors present, this and the inhibiting
titer (1) must stand in a direct proportion to each other, if these
two functions be carried by the same antibody. However, such
is not the case, as an analysis of the results presented in table 4
shows. With serum 109 the proportions were:
After the first injection..........1: B = 0.025 : 0.000025 = 1:0.001
After the third injection........ I: B = 0.0032: 0.0000063 = 1: 0.002
With serum 117 the proportions were:
After the first injection.........1: B = 0.0125: 0.0000008 = 1: 0.000064
After the second injection. .......1: B = 0.0032: 0.0000008 = 1: 0.00025
After the fourth injection... ....1: B = 0.0063: 0.0000032 = 1: 0.00052
Thus, it is obvious that the proportion between I and B is not
a fixed factor. This is demonstrated even better if we put
together all the results of the bactericidal reactions and reduce
the I function to 1 while the B function is proportionally reckoned
out. This has been done and the results are presented in table 5.
NEISSER-WECHSBERG INHIBITING PHENOMENON 15
TABLE 5
STRAIN INHIBITION BACTERICIDAL ACTION I ° B
>0.1 0.0032 ?
Shi 0.0002 ?
Shiga >0.1 0.0004 fe
oe es Se ee 0.0125 0 0000032 1:.0.00025
0.0063 0.0000032 1: 0.00052
0.0032 0.000000025 1: 0.0000078
>0.1 0.0008 15
>0.1 0.0001 ?
: >0.1 0.0000063 ?
IDE Ae eee ee 0.05 00000016 4:0.000032
0.0125 0.0000032 1: 0.00025
0.0063 0.0000025 1: 0.0038
>0.1 0.0063 ?
> OF 0.0008 ?
1D GS hh >0.1 0.0001 ?
0.05 0.0000063 1: 0.000126
0.0125 0.0004 1: 0.032
>0.1 0.0016 r¢
>0.1 0.00005 16
LV O40) oo Bs ae ato wee ei 1: 0.008
0.0032 0.000025 1:0.0078
0.0004 0.0000016 1: 0.004
0.0001 0.0000016 1: 0.016
ike hs>0.1 0.0016 ?
>0.1 0.0008 ?
SU 0.000025 ?
1) noon Rata ae 0.0125 0.0001 1: 0.008
0.0125 0.000025 1: 0.002
0.0032 0.0000016 1: 0.0005
0.0008 0.0000025 1: 0.03
F 41 0.0063 0.0002 1: 0.03
1) 410) U6 Re ae 0.0032 0.0002 1:0.06
oS Le >0.1 0.000025 r¢
0.0032 0.00005 1:0.015
Maer cholerae... 2... csi. .... 0.0032 0.0001 1:0.03
16 TH. THIGTTA
It is seen that the bactericidal titers vary in their propor-
tion to the inhibiting titer (taken as 1) from 0.06 to 0.0000078
This would be quite impossible if these two functions were due to
the same antibody, because, in that case, there would have to
be a constant ratio between the two functions.
The variations in the ratio 1: B confirms Brekke’s statement
that a high bactericidal titer does not eo ipso convey a high
inhibiting titer, and vice versa. By these findings we are forced
to reject the theory of the significance of the bactericidal ambo-
ceptor for the inhibition. Since, however, the inhibition arises
simultaneously with the bactericidal function and is quite specific,
TABLE 6
COMPLEMENT INHIBITION COMPLEMENT INHIBITION
0.05 0.05 0.02 0.0016
0.025 0.0001
0.04 0.05 0.0125 0.0016
0.025 0.01 0.0125
0.03 0.0125 0.0002
0.00382 0.009 0.2
0.025 0.0125 0.025
0.0032 0.007 0.0032
0.0008 0.0032
0.02 0.0125 0.005 0.0125
0.0032 0.0063—0. 0032
the former must owe its origin to specific antibodies other than
the amboceptor. How these antibodies must be supposed to
act and how they originate, will be presently dealt with.
7. Relation of inhibition to complement
If we examine the relation between inhibition and the dose of
complement, we find that this seems to be of a more stable
character. From the material of the bactericidal tests made in
this institute, it is obvious that the very high inhibiting titers
are to be found in tests where small doses of complement have
been employed, while the tests in which large doses of com-
plement have been used, usually show low titers. This will be
seen quite clearly from table 6.
NEISSER-WECHSBERG INHIBITING PHENOMENON 17
Still better we will see the relation between the dose of com-
_ plement and the inhibition in some tests made upon the same
serum with different doses of complement. The tests were made
at different times and not for the purpose of showing this fact.
Test A. Antigen: dysentery Shiga.
Immune serum: produced against dysentery Shiga, dose 0.1=0.0008.
Complement: guinea-pig serum (doses: 0.005 and 0.02).
C = 0.005 C= 0.02
1 = 0.0032 = 00125
1:¢ = 1.: 1.56 Le G2: 1.60
Test B. Antigen: dysentery, (group iii) (F. 79).
Immune serum: produced against F.79 (doses: 0.2 — 0.0032).
Complement: guinea-pig serum (doses: 0.009 and 0.08).
C = 0.009 C = 0.08
1 = 0.025 1 = 0.02
1-C = 0.36 1: C =a20240
Test C. The following test is made directly to demonstrate how 1
varies with C. The protocol of this test is presented in table 7.
Antigen: B. typhosus.
Immune serum: Produced against B. typhosus. Doses: 0.0063,
0.0125, 0.025, and 0.05.
Complement: Guinea-pig serum. Doses: 0.4 — 0.0125.
C = amount of the normal guinea-pig serum used.
= the “‘inhibiting titer.”
These three tests very clearly show how the titer of the in-
hibition varies according to the dose of complement. This re-
lation between inhibition and complement, indeed, is so close
that, in the same serum, we find an absolutely stable ratio be-
tween the smallest dose of the serum that gives the inhibition
and the employed dose of complement. It is clear, however,
that we can not expect in all cases to find the same ratio between
I and C, as a fixed amount of a complement-bearing serum does
not always have the same functional value.
The establishment of a fixed ratio between I and C is further
prevented by the fact that the inhibiting power of each serum
THE JOURNAL OF IMMUNOLOGY, VOL. V, No. l
18 TH. THIGTTA
TABLE 7
IMMUND SERUM
COMPLEMENT 0.0063 0.0125 | 0.025 0.05
Number of colonies
0.4 5 10 10 20
0.2 20 10 10 20
0.1 20 5 5 2000
0.05 20 20 Many 1000
0.025 1000
0.0125
1:C 1, Iho Leo
TABLE 8
GROUE EELN SS OF Sonate outta aie RATIO I: Cc
0.0032 0.03 I: 8.1
0.0005 0.005 I: 10.0
ae er Dys. Shiga 0.0032 0.005 1: 1.56
0.0125 0.02 I: 1.60
0.0125 0.025 I: 2.0
Oe 0.05 0.05 I: 7.0
0.0125 0.03 I: 2.4
Bolt 0.025 0.04 I: 2.0
4 F 52 0.05 0.04 1: 0.8
Danish I 0.025 0.04 I: 1.6
0.0001 0.02 I: 200
0.0002 0.01 I: 50
we 0.0016 0.0125 I: 7.8
0.0032 0.03 To
0.0125 0.01 I: 0.8
PES ee F 41 0.0032 0.007 I: 2.2
0.0032 0.006 I: 1.9
Lyphoid...'..2..|Lypa ik 0.0032 0.02 I: 6.25
NEISSER-WECHSBERG INHIBITING PHENOMENON 19
is different and varies from time to time. It is only possible to
show the relation between I and Cineach serum. The ratio I: C
will then be the same in all tests made with the same immune
serum and the same complement.
As this ratio is of a quite stable character, it looks as if a
certain dose of immune serum inhibits the action of a certain
dose of complement and that the larger doses of serum can
inhibit the action of correspondingly larger doses of complement.
In table 8 some inhibition titers are put together with the em-
ployed dose of complement. The tests were made at different
times and with different complements; consequently, we cannot
expect to find the ratio I: C so constant as in tests with the same
serum and complement.
The table shows that the value of the different immune sera
reckoned according to their faculty of inhibiting the action of
the complement is very different. Thus, while one unit of the
strongest inhibiting serum checked the action of 200 units of
complement, the same amount of the weakest serum could pre-
vent the function of only 0.8 units of complement. It is seen,
furthermore, that in three instances I: C is I: 10 or more, four
times it is between I:10 and I:5, eight times it is about I: 2
and three times it is I:I or less. Thus, in the majority of the
cases, we find that one dose of immune serum is capable of pre-
venting the function of two doses of complement as one of the
demonstrated tests shows.
The foregoing results make it obviously incorrect to compare
the inhibition titers of various sera, unless due regard is paid to
the employed dose of complement.
8. Inhibition in normal sera
Since the inhibition had been found to be dependent, to such
an extent, upon the employed dose of complement, it was natural
to think that normal sera also might show the phenomenon, if
only the dose of complement was small enough or the dose of
serum large enough.
20 TH. THJ@TTA
This consideration applies especially to the tests where normal
rabbit sera was found to show a quite high bactericidal action
against the dysentery strains of group I and II. It seems
possible that these sera would have shown the inhibition phe-
nomenon in higher doses.
In table 9 are presented the protocols of some tests with three:
normal rabbit sera in high doses.
Rabbits I and II in these tests show a strong inhibition down
to the dose 0.25 cc., while rabbit III shows a weak inhibition
only in the dose 0.75 cc. Thus, it seems that even wholly nor-
mal inactivated serum can inhibit the function of complement,
when large doses are employed. Since this non-specific inhi-
TABLE 9
RABBIT I RABBIT II RABBIT III
“Gy Sue | Daa Dn. | Bou
Sc C* = 0.008 che 00d. 2 see PYG Ene by c ye ue
Colonies Colonies
1L-(0) co roe) 0.75 1000 1000
0.5 co co OR25 400 100
0.25 Many 1000 | Many 1000 0.125 100 100
0.125 About 1000 50 0.063 50 50
0.063 10 5 0.032 30 50
0.032 10 5 0.016 30 20
* C = Complement.
bition is found only with much larger doses of serum than are
used for the bactericidal tests, there is no occasion on account
of this phenomenon to doubt the specificity of the corresponding
phenomenon in the immune sera.
Whether this normal inhibition is of the same nature as the
specific one, is an open question. It seems likely that this need
not be the case. In fact the doses of normal serum that give
inhibition are so large, that it is possible that the imactivated
serum taken as a solution of albumin absorbs the complement
in the manner of colloids. It may also be supposed that a
concentrated, inactivated serum contains so many complemen-
toids that these become more numerous than the effective com-
plement, and so inhibit the function of the latter.
NEISSER-WECHSBERG INHIBITING PHENOMENON 21
A study of the preceding experiments makes it clear that we
can expect to find a zone in the dilutions of serum between the
concentrated and the I: 10 dilution, where the inhibition phe-
nomenon can occur without its being possible to decide whether
it is specific or non-specific. In sera diluted more than I: 10 we
should most likely be out of this zone of inhibition. When we
begin the bactericidal tests with the doses 0.1 or 0.05 cc., we
will, therefore, usually not be troubled by the non-specific
inhibition. However, it has often happened that some hundreds
(200-300-500) of colonies have grown up in the largest dose of
serum used. This has not been taken as an inhibition, but has
been looked upon as an accidental occurrence. It seems possible,
now, that this was something like the tip of the tail of the in-
hibition, and that this phenomenon might have been found in
full strength had only higher doses of serum been employed.
We would then, however, have been incapable of deciding
whether this was a specific or a non-specific inhibition.
- 9, Inhibition in active, immune sera
The ratio between I and C, referred to above, unveils another
and very peculiar action of immune sera. It must be supposed
that the immune serum that is able to inhibit the function of a
foreign complement is able, also, to bring about the same action
against the complement of the immune serum itself. If, now,
we suppose that the test of an immune serum shows the most
frequent ratio between I and C; namely 1:2, one unit of the
serum taken as an immune serum should be able to check the
function of two units of the same serum taken as a complement.
Consequently, 0.5 cc. of the immune serum will be able to inhibit
the function of the complement contained in 1 cc. of the same
serum. The conclusion must be, that such a serum never can
show any bactericidal action against its homologous strain of
bacilli, whatever amounts of the active serum are employed
without added foreign complement. This will further lead to
the assumption that all the sera with a high titer of inhibition
must have lost the faculty of bactericidal action that they had
22 TH. THJGTTA
before the immunization. If this assumption proves to be true,
we will see that sera with some normal bactericidal effect will
not grow more effective with regard to total bactericidal power
as the immunization proceeds, but they will lose all the effect
that they had before. And if this is so in the tests in vitro, it is
most likely the same in vivo, because here the serum or plasma
never is present in diluted condition or with added foreign com-
plement, but it is always concentrated. In fact, the results pre-
sented in tables 10 and 11 show that the assumption is correct.
In these tables are demonstrated the results of tests carried out
with the serum of a rabbit that had been immunized first against
B. typhosus and then against B. dysenteriae, Shiga. Before the
treatment the serum of the rabbit very rapidly and easily dis-
solved typhoid as well as dysentery bacilli, while this faculty
was completely lost after some injections of antigen.
During the immunization with typhoid bacilli, tests with
dysentery bacilli were carried out to demonstrate that the
action of the serum against these latter bacilli was unaltered
and although this latter function remained undiminished, the
bactericidal effect against the typhoid bacilli was completely
lost.
During the immunization with the bacillus of Shiga, the same
control tests were carried out with a strain belonging to dysen-
tery group III. This group III stain remained very sensitive
to the serum, while the serum lost all its former effect against
the bacillus of Shiga.
The strong specificity of the phenomenon is demonstrated in
these tests.
As regards the typhoid bacilli, tests were carried out one
year after the last injection. The serum then had again ac-
quired its former faculty of total bactericidal effect. The in-
hibiting antibodies had diminished so much that they only
could act in the dose of 0.1 cc. with 0.02 cc. of complement.
According to the proportion I:C = 0.1: 0.02 = 1: 4, it is seen
that one unit of the serum as an inhibiting serum inhibits the
action of only 4 of the complement at hand. Consequently,
there is left + of the complement to bring about the bactericidal
action.
ee — — ee ee ee aa en i) Se aan (ane ye nee
co €Z0000'0
co 00000
a 1000°0
co co 0 © © QOOT AUB | © co co co © co Z000°0
Ooor Aue co I co oo OoOT AuvyT | © oo co © | 0001] © $0000
QooT AUBIN co I co co OOOT AUB | © © co © | 00g co 8000°0
OoT AUBIN co 0g co co ooor Auvpy | > [O0OT AUB] © co 02 |000T Aue 91000
oor Aue co 0g © co ooot Aus | © | OOTAURIN] - | © | 0 - [000T AULIY Z£00'0
oor Aus co 000T co OOOT AUBIN | 02 OOOT 0 o | 0 OOT £9000
oot looor Aueww} OOOT joodT AUBIN} O00T 002 Or OO0T 0 0 0 0 &Z10°0
OOT oor ooor Avy OOT OO0T OOT ZE jo00T Aue | 0 0 0 0 ¢z0'0
OOT og Oo0OT AUR 91 Ooot Aux | OOT AULT | OF jO0OT AUB} O 0 0 0 c0'0
Ss o¢ OOOT Aue 9 OOOT AUR OOOT 8 co 0 0 0 0 10
og 6 © OO00T Aueyy 0 0 0 0 z'0
000T Aue] 0 0 0 0 b'0
proydsy, | proydsy, | proydse | gropdoseq | PIOUUSL | Proadé a proyday, roe Pee rm proyd4y,
oIgr ee $ ore hha z Arenue Uu0148z sasoda wonuas
smricen Shee sre} on is eetuusieiie OT YOUR SAL 6 TOBIN 389.L Ge Avenger ae 3 4897, f -1UNUIWIT sxoyed 48a],
“ur UL 4saJ, -Ul UT 48aT,
See SS SS — — a ne,
SJINOTIOON JO UAAWON !‘WOUAS GAAILOV NI SLSTL
Se Se
s[njdooy g ‘ST6T “¢ Gorey
s[njdoot Z ‘8I6T “1% Areniqe,y
sinjdooy Z ‘8161 ‘8 Arenaqayy
[njdoot ¢ ‘gT6r ‘og Arenuee
[njdoo] ¢ ‘ST6T ‘ST Avenues
{njdooy 2 ‘gi6T ‘6 Axenuve :suoryoofuy
yyog proydhy ysurvbo woynzvunuuy
OST #498
Ol AIAVL
eR
23
TH. THJGTTA
a a a an
G9) 0
© £900000 0
Q00T Aue GZ10000 0
OOT GZ0000 0
OS ¢G0000 0
OOT 10000
ros) foe) co roe) co OOOT AUBIN © foe) co foe) co Z000'0
© © foe) © fo} QOOT AUBIT (oe) fos) © © © F000'0
foe) co foo) oo foo) ? foe} co foo) foo) (oe) foe) 8000 ‘ 0)
© foe) eo) foe) co foo) fee) feo) foo) co foe) CLOO x (0)
co foo) © co roe) © © co OOOT AUBIT foe) foe) Z200'0
oo co co OOT Aueyy © co 00¢ | © | 000T NOGY co 000T Aue €900°0
© co co 0g OoOT Aueyy © o¢ © | QOOT Luv | OOOT Aue | OOT LUV GZ10'0
OOOT co) od OoT OOOT © 9 © OO0OT AUBIN OOT OS Gc0'0
oor © © OS OOOT © 0 ee) OOOT AUB oT 0€ ¢0'0
0G oo oo G 0006 0 © OOOT AUBIN 9 0€ 10
© © 0 0008 0) cs) ee) 6 G0
0 od oo F'0
ae =S sf of JII Ateyuashq | yj A1oyueshq saiceene a pees i ane pAsozuesfq | [J Aroqueshq | J Ar1oyUasAG
Boed| se | 22 ‘so0=9 [sa
wos] os ~ 4 : ee tay 5 sasoa waugs
aA eg, E BIBI ‘T 1940990 489,], 14897, g sung 4sa,Z, ZI ABW SAL WOTZVZTUNULUIT ALOJaq $4SaJ,
SAINO100 JO UAANWON ‘WOUAS GAILOV NI SLsaL
a a ee ee ee ee
SoIny[No Z Noge ‘IGT ‘6 YUBA {oanyno [ ‘TE oquieseq ‘{njdooy T ‘gTé6T ‘T Ane
soinq[no Z ynoge ‘IGT ‘ET Youvy feanypno J ‘g¢ soquieseq ‘{nydooy $ ‘eT6T “g ounr
sainyno z ynoge ‘“GIGET ‘Sz Arwnaqoy foungpno J ‘6 LoquIoAON ‘[njJdooy ¢ ‘ST6T “eT AVIA
soiny[no Z ynoge ‘“GIGL ‘IT Aveniqoag foanqino T ‘gz 10q0}790Q ‘[nJdooT 2 “STET ‘g ABI
soiny[no Z ynNoge ‘6I6T ‘6z Aenuve foing[no T ‘g I9aqoyoO ‘[nJdooy] %* ‘STET ‘e ABI
soinq[no Z% ynoge ‘EIET ‘Gl Arenuve foinyjno [ “p 10qQ0300 ‘[nJdooy °Y* ‘QTET “FT Yous] :suorqooluy
(vBiyg) J dnowb) 17]100q fuaquashp ysuwbo woynzvunmur SOT 12990T
Il Have,
‘
NEISSER-WECHSBERG INHIBITING PHENOMENON 25
If we expect to find inhibition in active sera, I:C must be
I: I or exceed this ratio.
As now the greater part of our immune sera show the propor-
tion I: C = 1:2 or more, we must suppose that all these sera
have lost their total bactericidal action during the immuniza-
tion, and this is not only the case with the sera from rabbits.
On the contrary, 34 out of 64 sera from typhoid patients (Brekke)
showed so high inhibiting titers that they no doubt must have
been robbed of all their total bactericidal action. The same
has been found by the author as regards sera from dysentery
patients.
If we wish to transfer the results of these experiments in vitro
to the conditions in vivo, we must conclude that blood fluid from
immune organisms (if it is supposed that this is identical in its
action with the serum) so far from being more bactericidal than
the same in the normal state, loses its bactericidal effect
during the immunization or the disease. Consequently, the
bactericidal action of the serum can not play any part in the
stable immunity after a disease, or after an artificial infection.
The experiments show. clearly that the immune animals, whose
serum cannot kill the bacteria in question, tolerate the inocu-
lation much better than the normal animals, whose serum is
able to kill large amounts of bacteria. And we have no reason
to believe that the patients that show a high degree of inhibition
are not quite as immune as those, whose serum shows a high de-
gree of bactericidal effect. If the results of the test tube experi-
ments are applicable to the conditions in the living organism, it
would appear probable that the animals with the strong in-
hibition tolerate the inoculations better than the normal animals
just on account of the acquired faculty of inhibition. The effect
of an injection of bacilli into the organism of a normal animal
with bactericidal action, may be that the bacilli will be very
rapidly dissolved and their endotoxins made free to intoxicate
the organisms, if the dose of bacilli was large enough. This
will, perhaps, not be the case in an immune organism with a
well developed inhibiting action. Here the bacilli will either not
be dissolved in the blood stream or their solution will go on
26 TH. THJGTTA
more slowly and other immune forces can come into play and
deal with the bacilli. Consequently, the endotoxins are not
made free or at any rate they are liberated very slowly, and the
animal is saved from the intoxication that kills the normal
animal.
Since most normal sera in active, concentrated condition are
able to dissolve or kill the bacteria here dealt with, but lose their
faculty (or acquire the faculty of inhibition) during the disease
or the immunization, it is probable that the specific inhibiton
plays a far more dominant part in the protection of the organism
during the infection than is the case with the bactericidal action
of the serum. On the other hand we cannot consider the in-
hibiting action as any protection against becoming infected, and
the phenomenon, therefore, can hardly be of any importance to
the lasting immunity after the disease or inoculation.
We have considered the relation between inhibition and com-
plement without setting up any definite theory as to its ex-
planation. We have also expressed the idea that the inhibiting
serum interferes with or checks the function of the complement.
The question now to be answered is this:
What has happened to the complement?
That it is not fixed to a surplus of amboceptors, as held by
Neisser and Wechsberg, must be considered beyond doubt. Con-
sequently, we must consider the following possibilities:
1. The complement may for some reason be hindered in
attaching itself to the bacteria and amboceptors and remains
free in the fluid.
2. The complement may be fixed to the amboceptor in the
normal manner, but unknown antibodies render the bacteria
invulnerable.
3. The complement may be fixed to other antibodies than the
bactericidal ones, and cannot, therefore, be available for the
bactericidal reaction.
1. To test the first possibility the following experiment was
carried out. To an active anti-Shiga serum showing inhibition
homologous bacilli were added and the mixture was incubated
for 3 hours at 37°C. The bacilli were separated from the fluid
NEISSER-WECHSBERG INHIBITING PHENOMENON 24
by centrifugation and the absorbed serum was tested as to its
total bactericidal action against a strain of dysentery ITI, which
is very sensitive to the normal serum.
Simultaneously, a control test was made in active, not ab-
sorbed serum from the same animal.
Result: In the absorbed serum, no bactericidal effect.
Result: In the not absorbed serum: B = 0.0063.
It is seen that complement does not remain free in immune
serum during its contact with the related bacteria in the zone of
inhibition.
2. The second possibility is considered in the theory of Sor-
mani. It supposes that the surface of the antigen is altered by
the inhibiting antibodies in such a way that it is rendered invul-
nerable. If this be the case, it should be possible to show this
acquired invulnerability, when the bacteria that have been in
contact with an inhibiting serum, are brought into another, not
inhibiting, solution.
The following experiment was carried out to test this suppo-
sition.
Bacteria were kept in contact with the specific antiserum—
F.79—(I = 0.025; C = 0.009) in the ratio 1/8000 oese to 0.1
ec. of serum in a volume of 1 cc. After the mixtures had stood |
for different periods of time at different temperatures, the bac-
teria were separated from the fluid by centrifugation and by
washing in sterile saline solution and compared with the untreated
bacteria of the same strain as to their susceptibility to the normal
bactericidal action of fresh guinea-pig’s serum. As some growth
of the bacteria took place during their contact with the inhib-
iting serum corresponding dilution of the bacterial sediment had
to be made before the rest in the guinea-pig’s serum..
The results of the experiment are presented in table 12.
It is seen that the treated bacteria were killed as easily as the
untreated ones, consequently, during their contact with the
inhibiting serum they had not acquired any faculty that rendered
them invulnerable to the antibodies.
3. The complement may be supposed to be absorbed by anti-
bodies of another nature than the bactericidal ones. The con-
28 TH. THIGTTA
nection between these antibodies and the complement may be
thought to take place with the production of a precipitate that
ean be brought down as a sediment. Also, a union may be
assumed to occur between antigen and antibodies that remains
in solution, and for this reason cannot be removed from the
fluid by mechanical means. In the first case it must be possible
to remove the inhibiting antibodies after addition of antigen to
the serum and later centrifugation of filtration of the fluid.
The antigen then must be supposed to absorb the inhibiting
antibodies from the serum and thus remove them from the
solution. In the other case, the removal of the antigen cannot
be thought to rob the serum of its inhibiting faculties, as the
TABLE 12
A. FOUR HOURS B. FOUR HOURS Cc. TWENTY.FOUR
CONTACT CONTACT HOURS CONTACT IN| D. NOT TREATED
ee ee aigarou|) iN INCUBATOR IN THE COLD INCUBATOR
Colonies
0.5 3 0) 0 0
0.25 2 0 0 2,
0.125 5 2 50 0)
0.063 10 1 5 4
0.032 Many 100 1 4 5
0.016 Many 1000 100 50 lo)
Control © ©0 oo oS
soluble connection between antigen and antibodies will remain
in the fluid after centrifugalization or filtration and be capable
of absorbing complement when this is added.
To test these possibilities, the following experiments were
carried out:
1. To an anti-Shiga serum was added an emulsion of living
Shiga bacilli in great excess. After two or three hours in the
incubator at 37°C., the bacteria were removed in the centrifuge
and new bacteria were added. This procedure was repeated
eight times during thirty hours. No agglutination could be
seen after the last addition of bacteria.
In the serum thus absorbed, bactericidal tests were carried
out as well with a dose of complement suitable for the activation
NEISSER-WECHSBERG INHIBITING PHENOMENON 29
(0.005) of the serum as with a dose that itself had a bactericidal
action (0.02). Simultaneously, control tests were carried out
with untreated serum. The results of this experiment are pre-
sented in table 13.
It is seen that the prolonged contact of the immune serum
with the antigen caused diminution of its inhibiting action
although all of its bactericidal property had been removed.
Another anti-Shiga serum was treated in the following manner.
2. One part of the serum was diluted 1:5 and mixed with a
great excess of Shiga bacilli; this mixture was shaken at room
temperature. After four hours the bacilli were removed in the
TABLE 13
A. COMPLEMENT = 0.02 B. COMPLEMENT = 0.005
IMMUNE SERUM paar (= ete eee | Ae
Colonies
0. 1 joe) © c foe)
0.05 foe) foe) © foo)
0.025 ro) co co co
0.0125 co © © ea
0.0063 About 100 About 100 © oo
0.0032 0 0 About 100 co
0.0016 0 0 0 co
0.0008 0 0 0 oe)
Complement control 0 0 © co
centrifuge and the supernatant fluid was mixed with fresh
bacteria. This was repeated every fourth hour during forty-
eight hours (nights excepted). After the last centrifugation the
fluid was divided in two halves. One portion was heated at
56°C. for one hour and again centrifugalized. The serum was
then quite clear and showed no growth in a control test. The
other half was filtered through a Berkefeld filter. Control
culture showed no growth. As a control test some non-absorbed
serum was filtered.
With these three sera bactericidal reactions were carried out.
As complement was used a dose of fresh serum that was capable
30 TH. THJOTTA
of bringing about bactericidal action by itself. The results of
this experiment are presented in table 14.
It is seen that here again prologned contact of the inhibiting
serum with the homologous bacteria resulted in no diminution of
its inhibiting power whether the bacteria are removed with
centrifugation or by means of filtration.
Another portion of the same serum was treated in the follow-
ing manner:
Three cultures (surface 7 by 15 em.) of Shiga bacilli emulsified
in normal saline solution (80 ec.) were heated at 60°C. for one
TABLE 14
A. ABSORBED, B. ABSORBED, C. UNHEATED,
HEATED SERUM FILTERED SERUM FILTERED SERUM
SERUM
Colonies
0.1 (oe) ee) fo)
0.05 roo) co co
0.025 About 2000 Many 1000 About 1000
0.0125 1000 1000 200
0.0063 200 50 3
0.0032 50 20 5
0.0016 10 10 0
0.0008 5 2 10
0.0004 10 10 10
0.0002 5 50 10
0.0001 50 20 10
Complement control 5
Serum control 0 0 0
(0.1 Serum without bacilli)
hour, the emulsion was then centrifugalized, the bacilli washed
in normal saline and afterwards mixed with the serum diluted
1:10; after being shaken at room temperature for 24 hours, the
mixture was centrifugalized and a new bacterial emulsion was
added to the supernatent fluid. This procedure was repeated
four times.
* With this treated serum tests are made in the ordinary
manner. Complement was employed in amount (0.025 ec.) that
was itself capable of producing bactericidal effect. The results
of this experiment are presented in table 15.
NEISSER-WECHSBERG INHIBITING PHENOMENON 31
It is seen that the prolonged contact of the inhibiting serum
with the killed and washed bacteria failed to reduce its inhibiting
power.
Similar tests were made with the same serum absorbed at a
temperature of 5°C. with living bacilli during six days. The
serum thus treated showed a little more inhibition than the
untreated serum.
i» Likewise the same tests were carried out with a serum produced
against B. dysenteriae group III (twenty-four hours contact with
living bacilli). Also in this treated serum, a rather stronger
inhibition was noted than in the untreated serum.
TABLE 15
A. ABSORBED SERUM B. UNTREATED SERUM
SERUM
Colonies
0.05 o co
0.025 ey ©
0.0125 oo ea
0.0063 Many 1000 Many 1000
0.0032 About 200 About 500
0.0016 50 50
0.0008 10 10
0.0004 10 3
0.0002 5 6
Complement control 10
Cerum control (0.1 serum) 0
Control on bacilli used ©
Thus it has not been possible to remove the inhibiting effect
from the sera examined through absorption with the homologous
antigen.
Objection may possibly be made to the first experiment, that
living bacilli may have been left in the serum, as no control test
was made to exclude this possibility. The dose of active serum,
however, was so large (0.02 cc.), that it should easily have been
able to kill the small number of bacilli that might have been left
after the centrifugation, as it killed all the bacilli in the control
test. This objection cannot be made against the last two experi-
ments, as the serum here was found to be sterile before and
during the test.
32 TH. THJOTTA
It has thus been impossible to free the sera of the inhibiting
action through absorption with homologous antigen, as carried
out in this investigation. From the above mentioned experi-
ments we must conclude that the inhibiting antibodies cannot
be fixed to the bacillary bodies or to sedimenting antigen and
be removed from the serum with these; also, that the inhibiting
antibodies are quite different from the bactericidal amboceptors,
as the inhibiting action may be found in sera that have been
deprived of all bactericidal power.
DISCUSSION
If, on the basis of the foregoing experiments and results, we -
wish to try to form an opinion as to the nature of the inhibiting
antibodies and the value of the different theories concerning
them, it may be useful to consider the various possibilities in the
light of Ehrlich’s side-chain theory and to compare these possi-
bilities with the facts brought forth by these experiments.
1. Neisser and Wechsberg did not assume any special inhib-
iting antibody, but thought that the great surplus of bacteri-
cidal antibodies made the action of the complement impossible.
This conception is graphically represented in figure 1.
This theory is made impossible by the following facts:
a. There is no relation between the bactericidal titer and the
inhibiting titer.
b. There may be found immune sera with a high bactericidal
titer, but without any inhibiting action.
c. The inhibition can still be found in sera from which the
bactericidal amboceptors have been removed by absorption with
homologous antigen.
2. The idea that during the active imunization of an animal
an antibody is produced that is capable of rendering the bacteria
invulnerable to the action of the bactericidal amboceptor-com-
plement complex although the latter is not prevented from
combining with the bacteria. This idea is contained in the
theory of Sormani. It is represented graphically in figure 2.
This assumption is rendered untenable by two facts brought out
NEISSER-WECHSBERG INHIBITING PHENOMENON 33
in the present investigation, namely, (1) the hypothetical anti-
body cannot be absorbed from the inhibiting serum and (2) the
bacteria that have been in contact with an inhibiting serum are
not less vulnerable to the bactericidal amboceptor-complement
complex than are untreated bacteria.
3. The hypothesis that the inhibiting antibodies act by
attaching themselves to the antigen, thereby hindering the union
of the latter with the antigenophil group of the bactericidal
amboceptor, is represented graphically in figure 3.
This hypothesis is excluded by the fact, demonstrated above,
that the bactericidal amboceptors can be absorbed from the
serum by the bacteria in the presence of the inhibiting anti-
bodies, and that the latter actually do not become attached to
the bacteria at all but remain free in the fluid.
4, During the immunization it is conceivable that an antibody
of the nature of an antiamboceptor is produced. By fixing itself
to the antigenophil group of the amboceptor; this antiambo-
ceptor might hinder the connection between amboceptor and
antigen. The latter would, in this case, remain free and the
bactericidal action could not take place. This conception is
represented graphically in figure 4. This theory necessitates that
the combination amboceptor-antiamboceptor absorbs the com-
plement with greater avidity than the combination amboceptor-
antigen. The bactericidal amboceptors namely, are at hand in a
far greater amount than the supposed antiamboceptors (I. gives
always a lower titer than B.) and the amboceptor-antigen com-
pound therefore would absorb the greater amount of complement,
if the avidity in both cases were the same. But if an antiambo-
ceptor connected with its homologous amboceptor could absorb
all of the complement, the nature of the antigen would play no
part in the inhibition, and the inhibition would thus take place
against any bacteria tested in an inhibiting serum. As this is
not the case, the theory cannot be accepted.
5. Conceivably, the production of antiamboceptors against the
complementophil group of the amboceptor (complementoids)
may take place during the immunization. Such antibodies
would not prevent the union of the bactericidal amboceptor
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 1
34 TH. THJOTTA
with the bacteria, but they would prevent the subsequent
cdoperation of complement. According to this view, which is
graphically represented in figure 5, bacteria after contact with
an inhibiting concentration of immune serum should be invul-
nerable to the action of fresh active serum, since all of the
receptors are supposed to be satisfied, moreover, an absorbed
inhibiting immune serum should possess bactericidal power since,
according to this view, its complement should remain unaltered
in the fluid after the absorption. Since neither of these two
conditions are met in actual experiment, the theory cannot be
sustained.
The theory of Brekke must be dealt with under this last possi-
bility. This theory must be rejected, as it is demonstrated that
the inhibition takes place, also, in quite fresh and active sera.
6. Finally, it is possible that during the immunization specific
antibodies are produced, which in connection with dissolved
antigen, absorb complement with a greater avidity, so that the
complement cannot effect its bactericidal action. Figure 6
represents this idea graphically.
This supposition necessitates that the titer of inhibition varies
with the dose of complement, since a certain dose of inhibiting
antibody must absorb a certain dose of complement. If the dose
of complement is larger than that required by the inhibiting
antibodies, we must suppose, that the rest of complement is
made use of by the bactericidal antibodies. If, on the other
hand, the dose of complement is very small it should be possible
to show the existence of very small doses of inhibiting antibodies.
Further, the bacteria that have been in contact with an inhib-
iting serum can be entirely vulnerable in a non-inhibiting serum,
since nothing has happened that should diminish their vulner-
ability. On the other hand, the absorbed serum, according to
this view, should lose all its former bactericidal action, as no
free complement will be left in the solution. f
All the above described experiments satisfy these claims, and
this theory may therefore be accepted.
The specific inhibiting antibody thus acts like an anticom-
plement inasmuch as it hinders the function of the complement.
NEISSER-WECHSBERG INHIBITING PHENOMENON ao
Since, however, it can act only in the presence of the homologous
antigen, we must suppose that a combination of this antibody
and the antigen absorbs the complement. We must, therefore,
look upon the inhibiting antibody as being of the nature of an
amboceptor, but not as an amboceptor that enters into union
with the solid sedimented bacterial protein, since it has not been
possible to remove the inhibiting antibody from the serum by
absorption and centrifugalization or filtration. We must rather
suppose that the inhibiting antibody in connection with dissolved
antigen forms larger colloidal molecular complexes that absorb
the complement at hand with great rapidity without the forma-
tion of a precipitate, which could be removed from the fluid.
This explanation is the same as that used by Gengou and by
Moreschi in their work on anticomplementary action. We
have here, also, an explanation of the great similarity between the
specific action of the mentioned inhibiting antibodies and the
non-specific action of such complement inactivating bodies as
bile, peptone, albumose, citrate, and oxalate solutions. The
action of the latter bodies is non-specific; the complement is
inactivated in any combinations whatsoever. The action of the
inhibiting antibody, however, is dependent on the presence of the
homologous antigen. But with this exception the action and
the result of the action is the same in both cases; namely, the
inhibiting of the function of the complement, and hereby the
hindering of the bactericidal action.
It is possible that the examination of the physical colloidal
conditions of the serum absorbed with its homologous antigen
would bring us closer to the solution of the problem of the inhi-
bition than this investigation has been able to do.
If we wish a working hypothesis as to the origin of the inhib-
iting antibodies, we may assume the following:
When antigen gets into the veins of man or animals, it will,
partly at any rate, go into solution. The dissolved antigen will
bring about the production of immune bodies against itself.
Among those bodies will be found some that precipitate the
dissolved antigen, and others that enter into combination with
the antigen without producing a precipitate. This last combi-
36 TH. THIOTTA
nation is capable of binding complement and the antibodies
concerned will, for this reason, appear as inhibiting antibodies.
Thus, the production of inhibition in a serum must be looked
upon as a reaction of immunity just as agglutination, precipi-
tation and the bactericidal action. It is possible that the inhib-
iting reaction might be employed as a diagnostic aid in infections
with easily soluble bacilli.
The phenomenon of specific inhibition has hitherto been
looked upon as a peculiar and paradoxical function of immunesera.
Practically, however, it has not yet attracted any interest. It
is nevertheless probable that the phenomenon of inhibition may
play a part in the action of the therapeutic immune sera. This
will, however, take place only in sera possessing bactericidal or
bacteriolytic action which, in practice, means the antimenin-
gococcic serum. The antimeningococeus serum is widely used
and is considered to be of use in the treatment of meningitis.
However, its therapeutic effect must be due to antibodies that
do not need the coéperation of complement, because this function
is absent in the stored serum, and the normal spinal fluid lacks
complement. Under these circumstances the bactericidal function
of the serum cannot be brought into action.
If the attempt should be made to improve the effectiveness of
the antimeningococeus serum by the addition of active
serum—say from the patient himself—it would be important,
first to ascertain the titer of inhibition of the immune serum, if
that property exists in such sera—a question which has not been
investigated. If, for example, the ratio I: C were found to be
1:2, this would mean that the ordinary dose of the immune
serum (25 cc.) would nullify the complementary action of 50 ce.
of active serum. Hence, it would be necessary, under these
circumstances, in order to secure the codperation of complement,
to inject, with the immune serum, more than 50 cc. of the normal
serum. However, as the spinal column is not able to accom-
modate this large volume of fluid, it would be necessary to adopt
the other alternative of injecting less of the immune serum than
usual. For example if we choose to inject the usual volume of
25 cc., and if we wish to inject a mixture of immune serum and
ee
NEISSER-WECHSBERG INHIBITING PHENOMENON 37
complement-bearing serum in which the ratio of the latter to
the former is more than 2:1, say 5:1, we can use the following
mixture:
HOTTER CR SE LUTTE NOR Ea fase. cus (Live a ald ee eb oe ole cee ecseay: 1.0
Horgein mairmaal Serr eere ene Ge ee an ee et Tb 8d tS 5.0
SOMES CISO NT DLOM fc rds tasetcte «sta oe sie cc cee Ue Bab oe see aracve ss —19.0
It is likely that such a diluted mixture would be more efficacious
than an undiluted one, as it might penetrate more easily into the
meninges.
The necessity, in such procedure, of determining the ratio I: C
for each specimen of immune serum, must be pointed out.
The possibility that the use of a strongly bactericidal mixture
may cause an intoxication as the result of a sudden liberation of
endotoxins in large amounts, should be borne in mind.
Where, on the other hand, sera are dealt with that are to be
injected intravenously, the inhibiting action cannot, as a rule,
play any part in the therapeutic effect. In such case there will
always be a far greater amount of active serum at hand than the
dose of inhibiting serum. Only if the serum has a very high titer
of inhibition, may there be a slight possibility of inhibition.
Thus if I: C were 1: 200, as in one of our dysentery sera, 10 ce.
of inhibiting serum would nullify the complement in 2000 ce. of
active serum, and in this case 20 cc. of inhibiting serum injected
intravenously would produce an inhibiting mixture, as a grown
~ up man has about 2500 ce. of blood fluid.
CONCLUSION
1. The inhibiting phenomenon of Neisser and Wechsberg is of
a specific nature. It is to be found in active as well as in inactive
sera; it develops during the immunization and can be found in a
very high degree in dysentery immune sera. In active sera
from immunized animals examined without the addition of
foreign complement, the phenomenon presents itself as a com-
plete abolition of the normal bactericidal action.
2. The inhibition is due to antibodies that arise during the
immunization or during the natural disease. These antibodies
38 TH. THJGTTA
are not identical with the agglutinins, the bacteriolysins or the
precipitins. They must be considered as specific antibodies,
which combine with dissolved antigen to form molecular com-
plexes, that have a marked tendency to absorb complement and
to withdraw it from the bactericidal antibodies.
3. The titer of inhibition is directly proportional to the em-
ployed dose of complement. With a small dose of the latter,
smaller doses of inhibiting antibodies can be demonstrated than
with a larger dose of complement.
4, The inhibiting antibodies do not effect the bacteria them-
selves, nor can they be removed from the serum by absorption
with an emulsion of the homologous bacilli. They can be demon-
strated in sera that lack any bactericidal action.
REFERENCES
BREKKE, ALEXANDER: Om tyfusvakcinations og seroologiske methoder til bedgm-
melse av dens effect. (Tillegshefte til Med. Revue, Bergen 1916.)
Gay: Citation by Sormani.
GerNGou: Sur les sensibilisatrices des sérums actifs contre les substances albumi-
noides. Ann. de l’Inst. Pasteur, 1902, 16, 734.
GruseEr, M.: Theoretisches tiber die Antikérper im Blute. Wiener klin. Woch.
1901, 14, 1244.
LevapITI: Citation by Sormani.
Lirstern: Die Complementablenkung bei bactericiden Reagenzglasversuchen
und ihre Ursache. Centralbl. f. Bakter., Orig., 1902, 31, 460.
Morescui, C.: Zur Lehre von den Antikomplementen. Berl. klin. Woch., 1905,
42, 1225.
Netsser, M., anp Wecuspera, F.: Ueber die Wirkungsart bactericider Sera.
Miinch. med. Woch., 1901, 48, 697.
Sormant, B. P.: Eine neue Erklarung des Neisser- und Wechsbergschen Phano-
mens vermittels des Phinomens der spezifischen Sprédigkeit. Zeits.
f. Immunitatsforsch., Orig., 1916, 24, 336.
Tuugrra, TH.: On the bacteriology of dysentery in Norway. The Jour. of Bac-
teriology, 1919, 4, 355.
= ~~
|
|
:
THE RELATION OF THE RATE OF ABSORPTION OF
ANTIGEN TO THE PRODUCTION OF IMMUNITY
MARJORIE W. COOK
From the Bacteriological Laboratory of Brown University, Providence, Rhode Island
Received for publication January 14, 1920
Studies of the rate of antigen absorption in sensitized and
immunized animals, as contrasted with the rate of absorption
in normal animals, have shown that previous treatment with
an antigen confers upon an animal an increased power of absorp-
tion for that antigen. This fact was established with respect to
sensitized rabbits and guinea-pigs by Doerr and Pick (1), Fried-
berger and Lura (2), Romer and Viereck (3), and others. Smith
and Cook (4) confirmed the work of Doerr and Pick and further
(5) demonstrated that absorption of antigen in immunized ani-
mals proceeds much more rapidly than in sensitized or normal
animals.
As it is evident that upon sensitization and particularly upon
immunization, an animal acquires increased powers of absorption
for the specific antigen, the question arises as to whether the
process of immunity production is in any way dependent upon
the changed absorptive powers of the organism. In other words,
if a high degree of immunity is always accompanied by a marked
increase in the rate of absorption of the antigen, can any cause
and effect relationship between the two phenomena be established?
Or, approaching the subject from another viewpoint, is it possible
to demonstrate that conditions which furnish possibilities for
increased powers of absorption of antigen also result in the pro-
duction of a high degree of immunity?
A possible method of investigating this question is suggested
by certain facts established by a number of workers with respect
to the production of changes in the permeability of the cell
39
40 MARJORIE W. COOK
membrane through the use of varying concentrations of electro-
lytes in the surrounding media. Loeb (6), Lillie (7), Osterhout
(8), and McClendon (9), have shown that whereas certain marine
organisms may be considered to be in equilibrium when the
surrounding medium contains a definite proportion of Ca and
Na ions, well recognized changes in the cell activities, such as
simple stimulation, initiation of cell division, and in extreme
cases toxicity and death, may be set up by varying the proportion
of these ions. These investigators have demonstrated that the
above results are due to changes in the permeability of cell
membranes, an increased permeability being caused by an excess
of Na ions, a decreased permeability by a slight excess of Ca ions.
Clowes (10) in a similar work upon the action of antagonistic
electrolytes upon living cells has shown that in higher organisms
a disturbance of the equilibrium of the cells can be effectively
produced by the use of calcium chloride on the one hand and
by sodium citrate upon the other.
With these facts as a basis for further procedure, it seemed
possible that by introducing an excess of Na ions or of Ca ions
into the blood of experimental animals, changes might be effected
in their powers of absorption for an antigen simultaneously
injected. To obtain the maximum effect of these electrolytes
upon antigen absorption, it seemed advisable to administer them
in connection with the usual immunizing procedure, that is in
a series of five or more injections. By means of such a procedure
a series of observations could be made during the entire course
of immunization, and any cumulative effects, as well as any
immediate changes due to the action of the electrolytes, could
be observed.
In carrying out the experiment, as outlined above, the electro-
lytes used were calcium chloride and sodium citrate. These
were employed in oa concentration in amounts of 1 cc., toxicity
tests upon both electrolytes having shown that this amount
usually caused no marked disturbance when given intravenously
to rabbits of 1500 to 2000 grams weight. As antigen, a simple
protein, egg albumen, was chosen, since the course of absorption
RELATION OF RATE OF ABSORPTION TO IMMUNITY 41
of a simple protein can be more readily followed by means of
the precipitin reaction than can be accomplished by any of the
usual in vitro tests upon a more complex antigen, as for instance
a bacterial emulsion.
The rate of absorption of antigen into the blood was measured
by bleeding the animals before, and at intervals following each
injection and titering these bleedings with an anti-egg precipi-
tating serum for their content in egg albumen. In view of the
importance ascribed by various workers (Hektoen (11), Melni-
kowa and Wersilowa (12)) to the leucocyte reaction in connection
with any attempts to change the reactivity of an organism to an
antigen, leucocyte counts were made before and at intervals
after each injection. The rapidity of antibody formation under
the conditions of the experiment was measured by titering the
bleedings taken before each injection and a bleeding taken ten
to fourteen days after the last injection for precipitins for
egg albumen. Further details of the procedure are given in
experiment 1.
Experiment 1
Three groups of animals were treated as follows:
Group 1. Rabbits 1, 2, and 3 received for five successive times at
m
five-day intervals intravenous injections of 1 ce. of 10 sodium citrate
and immediately following, intraperitoneal injections of 1 cc. of a 2
per cent solution of powdered egg albumen.
Group 2. Rabbits 4, 5, and 6 received for five successive times at
m
five-day intervals intravenous injections of 1 ce. of i0 calcium chloride
and immediately following, intraperitoneal injections of 1 cc. of a 2
per cent solution of powdered egg albumen.
Group 3. Rabbits 7, 8, and 9, controls, received at five-day intervals
five intraperitoneal injections of 1 cc. of a 2 per cent solution of powdered
egg albumen.
Bleedings were taken from all animals immediately preceding each
injection and one, two, four, six, twenty-four, and forty-eight hours
following. The anti-egg precipitating serum used for the titration of
the content of each bleeding in egg albumen was 1: 12800 in titer. The
42 MARJORIE W. COOK
usual precipitin procedure was employed, 0.1 cc. of this serum being
added to 2 ce. of dilutions of 1:50, 1: 100, 1: 200, 1: 400, and 1: 800
of the serum obtained from the bleedings. In recording the time re-
quired for the absorption of antigen, the highest dilution of the serum
in which a positive reaction was obtained is given to indicate the relative
concentration of albumen in the serum as well as the time interval.
These values are given in table 1. In no case was egg albumen present
in the bleedings taken immediately preceding the injections.
Leucocyte counts made before each injection and for periods of
four to eight hours following are recorded in table 2.
Data in regard to the course of antibody production of each rabbit
are given in table 3.
Examination of the data given above shows that there are
considerable differences between rabbits receiving electrolytes
and the control animals. It is evident that the administration
of electrolytes influences very markedly the rate of absorption
of antigen, for while antigen was first detected in the blood of the
control animals at an interval of four hours after the intraperi-
toneal injection, the animals receiving sodium citrate gave evi-
dences of absorption in every case after an interval of one hour.
Of the animals receiving calcium chloride, on the other hand,
no. 5 showed no evidence of absorption at any time, no. 4 gave
positive tests for antigen in the blood after an interval of twenty-
four hours in four out of five injections, and in no. 6 antigen was
absorbed into the blood after an interval of six to eight hours.
The most significant feature of these results, however, is the
fact that variations in antibody content of these animals were
parallel to the differences found in the rates of absorption of
antigen. Precipitins appeared more rapidly and in very much
higher concentrations in the rabbits of the sodium citrate series
than in the control animals. The reverse was true of the animals
receiving calcium chloride, one rabbit of this group giving no
positive reactions.
With respect to the leucocyte count there was throughout a
considerable variation in the individual rabbits. In the control
animals there were no marked reactions. In the rabbits receiving
sodium citrate there was a rather pronounced fall in the count
43
002 + T P OOT = T v OOT = T v OOT = T v 0¢ :T v OOST | 6
OOT = T i OOT + T v 0G :T v OOT + T v OOT = T p OPgT | co) ite. ‘*Touyuoo “ou0N
002 = T P 006 = T P OOT + T P OOT + T 9 OOL = T uy 0006 | 2
OOT + T 9 OOT :T 8 0¢ :T 8 0G :T 8 0S :T 8 Oost | 9
uotdiosqe oN | uordsosqe on | uordiosqe oN | Uoldiosqe ON | UoTydIOsqe ON | 000Z
0S :T VG 0S :T VG 0S :T VG uorydiosqe ON | OOT‘T ¥G OO9T | ¥
{Ye}
pines eplwoyyo unt1aley
004 = T I 006 + T T 002 :T T OOT -T I OOT :T i OOST | &
006 - T i 006 + T i OOT :T i 0¢ :T T 0g :T I OO9T | & "res" "9F@IZ1D UINIPOS
00 + T I OOT :T i OOL + T I 002 + T i! OOT + T \! OcoT | T
sunoy sunoy sunoy sunoy sunoy | sunub
STL any eee an ee ek PTE eR ihe on leene © hue easae
UoNnid | Swty, | WONT | wry, voyniqg | eur | wound | omy, | uoynq | om | F43!AK| on N
¢ uonoofuy 7 uorjoofuy ¢@ worqoofuy Z uorqyooluy J uorooluy @LATOULOATA
Liaava
UqLIV GOOId AHL NI NADILNVY JO GONVUVAddyV LSuId
Se ee ee eee ee ee ee ee eee
uabyun fo uoydsosqn fo sans aaynjay
T @TAVL
b
d
:
2)
B
Z
o)
Bi
Ay
(om
2)
Dn
a
ic
o)
2
<
a
&
°
Z
E
I
pS
COOK
MARJORIE W.
0086 ‘iy G
000‘8 BunG
009‘L Me |
OOF ‘OL ILO Jo
000‘9T | “4 ¢
OOO'TT | “14 z
008‘9T | “AUT
000‘ST o10Jod
OoO'TT | “24 ¢
00F‘8 Gs
0089 iT
000‘0L | 210J0q
00s‘OT | “24 F
000‘OT | “4 z
004.2 | AUT
00¢‘OT oLO fod
000‘CT | “AU F
00s‘OT | “AU
o00‘Or | “24 T
OOS‘IL | e10Joq
000‘Or | “YL
000'9 IG
00F‘S Aq T
000°6 | 90Jog
0089 “449
009‘F AY ZG
009 F “aq T
008‘6 | e20jJog
yunoy oul y,
gG NOILLOGLNI
p NOILOGLNI
009‘TT
008 ‘TT
000°0T
000‘ZI
000°ZT
000‘TZ
00¢*L
008‘TT
00001
008‘2
000‘L
008‘6
yunog
IY F
a
“Ty
olOjogq
ES:
“IU ¢
PI eay
alo jog
“14 9
“Tye
TUE T
alOjog
€ NOILOALNI
“IY 9
"Iu, Z
“IY T
a0 jog
cLuay
‘IWS
“Ty T
a10 Jog
‘IU F
a BITE
“IY 1
alojog
eB)
Tyg
aT
a10Joq
a10 Jog
YF
"IY Z
as
OOF TT |e10joq
009‘or | “4 F | 009°6
o0s‘6 | “az | 00F‘Or
0006 ‘Iy T 009‘6
OOS‘OL | e40Joq | O00‘EST
o0s‘6 | “48 | 008'6
000‘0r | “4 F | 000‘8
000'8 | “4 T | 000'9
0096 | 240J0g | OOO‘OT
000°OT | “AX 8 | OOL'6
000‘FZ “IY Pp 000‘0T
00002 | “4 T | 009'F
009'6 | e40joq | 0006
009‘6 “IU Pf 000‘6
000°2 | “4% | 00F'9
009°¢ ‘IY T | 009°¢
000‘6 | et0J9q | 00Z‘OT
0006 | “4%
000°9 | “1%
OO re aT
000‘0L | e40J0q | 000'6
00¢"9
000.9 | ‘24% | 00F6
OOF ‘OT ILO JO
qunog oul], rane.
Z NOILOMLNI
squnod ajhiv0onaT
¢ WIAVL
oun,
T NOLLOGLNI
oud “*9plloyyo WINTD]BD
Patt 8781}10 UINIPOg
GQLATOULAI
45
RELATION OF RATE OF ABSORPTION TO IMMUNITY
*9} BIPIO
UINIPOS Jo UoTZIof[UT OY} SUIMOT[OJ UOTJOVOI OIOAOS B JO DDUBIINIDO OY} JO OSNvOIG oye} JOU SvAL 4UNOD 0}AV0ONIT OUT, »
000‘FT
000‘2T
000‘Ez
008‘ET
00r‘6
00F‘Z
008‘2
008‘ST
“IY F
“ag
Ty T
a10jog
“IY F
“TY Z
aq T
a1ojog
000‘TT
00F‘S
0086
00021
0086
008‘2
006‘¢
006 ‘OT
000‘0T
006‘6
0086
009‘TT
LEY.
ITC
ae!
d10 Jog
“I >
“ay €
aq 1
d10JO
‘IY F
‘a Z
“aq
a10jog
002 ‘TT
000‘0T
000‘¢T
000‘¢T
00¢‘ET
OOF ‘TT
OOT‘2T
006 ‘ET
00F ‘ZT
00¢‘9T
009‘ZI
OOT‘ST
“TY 9
“TY Z
TY T
a1ojog
“AY F
“TY g
‘ay T
a1OJOT
ortTay,
a0
|
a10jog
00¢‘ST
006‘OT
006‘9T
00F‘8
0086
0086
00¢‘8
009 ‘FT
002 ‘FT
002 ‘ZT
002 ‘ZT
“IY F
“qq 1
a10jog
ay
st
ea!
alo jog
cE
iG
ete
a1ojog
00¢‘9
00s‘2
009‘6
008‘T
008‘TT
00F‘6
000‘ST
006
008 ‘ZT
009‘2T
000‘TT
002L‘TT
‘IU P
“Ca ce
eee
d10JO
“I >
“TY Z
ay T
a10Jog
eA
TUG
gta
a10Joq
oe
se eree *]0.17 009 ‘QUON
46 MARJORIE W. COOK
one to two hours after injection, and this fall was not followed
by a leucocytosis. In the calcium chloride series there was
usually a drop in the count the first hour after injection. This
was occasionally followed by a leucocytosis with a return to
normal in four to five hours. With the exception of the leuco-
penia observed in the animals receiving sodium citrate, however,
the reaction in none of the rabbits was sufficiently distinct to be
of any particular significance. The leucocyte reaction cannot,
therefore, in this series of experiments be considered as an impor-
tant factor in determining or reflecting changes in immunity
production.
TABLE 3
Precipitin content of bleedings
RABBIT TITER OF BLEEDINGS TAKEN
ELECTROLYTE NUM- TIME OF FIRST APPEARANCE 14 DAYS AFTER
BER THE LAST INJECTION
1 | Preceding the 5th injection Positive in 1: 51200
Sodium citrate... 2 | Preceding the 5th injection Positive in 1: 25600
3 | Preceding the 5th injection Positive in 1: 51200
4 | 14 days after 5th injection Positive in 1: 400
Calcium chloride. 5 0 Negative in 1: 100
6 | 10 days after 5th injection Positive in 1: 1600
7 | 10 days after 5th injection Positive in 1: 3200
None, control.... 8 | 10 days after 5th injection Positive in 1: 3200
9 | 10 days after 5th injection Positive in 1: 1600
As the results of the above series indicate that the production
of precipitins is distinctly affected by the administration of
sodium citrate and calcium chloride during immunization and
that the rate of antibody production follows the rate of antigen
absorption it is of interest to ascertain whether the same relations
hold true with respect to other antibodies. A similar experiment
was therefore carried out, in which a series of animals were im-
munized to typhoid bacilli and the effect of sodium citrate and
calcium chloride was determined upon the production of agglutin-
ins and opsonins, as well as precipitins. Actual determinations
of the rate of absorption of the antigen were omitted in this
series, as a bacterial antigen when killed before injection cannot
RELATION OF RATE OF ABSORPTION TO IMMUNITY 47
be readily detected in the animal body. In this case, it was
assumed that the electrolytes would act upon the animal in the
same way as when actual determinations were made, as in
experiment 1.
Experiment 2
Rabbits 10, 11, 12, 13, 14, and 15 received seven intraperitoneal
injections of suspensions of heat-killed typhoid bacilli in physiological
saline, one hundred million bacteria being given at each injection. The
first three injections were given at five-day intervals. A period of ten
days with no injections followed, after which the four remaining injec-
tions were given at five-day intervals.
Immediately preceding each intraperitoneal injection of typhoid
m
antigen, rabbits 10 and 11 received 1 ce. 10
sodium citrate intravenously,
m
rabbits 12 and 13 received 1 cc. 10 calcium chloride intravenously,
while rabbits 14 and 15 served as controls with no intravenous injections.
Bleedings were taken before the first and fourth injections and seven
and seventeen days following the last injection. These bleedings are
designated as 1, 2, 3, and 4 respectively. Bleedings 1, 2, and 3 were
titered for agglutinins only. With bleeding 4 the titers of agglutinins,
opsonins, and precipitins were determined. Table 4 is a comparative
study of the development of agglutinins in the six animals. Table 5
gives the concentration of agglutinins, opsonins and precipitins of
bleeding 4.
It is evident from table 4 that the production of agglutinins
is noticeably influenced by the use of electrolytes during immuni-
zation. Animals receiving sodium citrate showed throughout a
much higher titer than the controls, while animals receiving
calcium chloride were lower in titer than the controls. Table
5 shows, moreover, that agglutinins, opsonins, and precipitins
followed parallel courses. The results of this experiment, there-
fore, confirm those of experiment 1, and while direct evidence
cannot be brought that in this case the rate of absorption of the
antigen has a direct influence upon the degree of immunity
produced, it is nevertheless evident that the use of those agents,
which caused such marked changes in the rate of absorption of
48 MARJORIE W. COOK
egg albumen, is followed by changes in antibody production to
the typhoid bacillus, which are exactly similar to the changes
effected in the production of precipitins for egg albumen.
While the influence of electrolytes upon the reactivity of an
organism to an antigen is of considerable interest, the fact which
is of chief importance for the present discussion is that a rapid
TABLE 4
Agglutinin titer of bleedings
BLEEDINGS
ELECTROLYTE RABBIT
1 2 3 4
Sodi itt 10 0 1: 640 1: 2560 | 1: 10240
OGluMeMTaAte: /--7..o: ooo 0 1: 640 1:2560 | 1: 10240
: ; 12 0 0 1:160 | 1:320
Calciumuchloride.4..-. a2 2 13 0 1:30 1:160 |1:320
N oi 14 0 1: 160 1: 640 | 1: 1280
NONE CONLLOL sc: ¢ tec ce ar 15 0 1: 80 1:320 | 1:1280
TABLE 5
Antibody content of bleeding
ELECTROLYTE | RABBIT ig a | eee PRECIPITINS
Sadiuin citrate 10 1: 10240 1.91 1: 6400
26 016101 6. 6) 6)'8) 88,6 6 66 6:6 11 1: 10240 1.80 1: 3200
- 4 12 1: 320 0.93 1: 100
Calcium chloride..............- 13 1: 320 0.91 | Negative in 1:50
None. control 14 1: 1280 |Z! 1: 1600
PRC OULLO eens ciciors tele ekeles ace 15 1: 1280 1.14 1: 300
rate of absorption of antigen was followed by an increased pro-
duction of antibody. The reverse was also true—a much
retarded rate of absorption of antigen was accompanied by a
marked decrease in the production of antibody. These results,
when taken into consideration with the statement made previ-
ously—namely, that the condition of immunity is accompanied
by increased powers of absorption for the specific antigen—may
RELATION OF RATE OF ABSORPTION TO IMMUNITY 49
be of significance in throwing some light upon the mechanism
of immunity production. Certainly support would seem to be
given to the idea that the condition of permeability of the cell,
with respect to its absorptive powers for the introduced antigen,
is a factor of considerable importance in the production of
immunity.
REFERENCES
(1) Doerr, R., anp Pick, R.: Centralbl. f. Bakteriol., 1912, 62, 146.
(2) FRIEDBERGER, E., anp LurA, A.: Ztschr. f. Immunititsforsch u. exper.
Therap., 1913, 18, 272.
(3) Romer, P. H., anp Vierecx, H.: Ztschr. f. Immunititsforsch. u. exper.
Therap., 1914, 21, 32.
(4) Smirx, G. H., anp Cook, M. W.: J. Immunol., 1917, 2, 421.
(5) Smiru, G. H., anp Cook, M. W.: J. Immunol., 1917, 2, 269.
(6) Loxrs, J.: J. Biol. Chem., 1917, 32, 147.
(7) Linu, R. S.: Am. J. Physiol., 1910-1911, 27, 289.
(8) OstERHOUT, W. J. V.: Science, 1917, n.s., 45, 97.
(9) McCuiEenpon, J. F.: Am. J. Physiol., 1911-1912, 20, 302.
(10) Crowes, G. H. A.: J. Physical Chem., 1916, 20, 407.
(11) Hextosn, L.: J. Infect. Dis., 1916, 19, 69 and 737.
(12) Metnixowa F. J. anp Wersitowa, M. A.: Centralbl. f. Bakteriol., 1912,
66, 525.
THE JOURNAL OF IMMUNOLOGY, VOL. Y, NO. l
STUDIES ON THE MENINGOCOCCIDAL ACTIVITY OF
BLOOD
TOITSU MATSUNAMI
From the Government Institute for Infectious Diseases of the Imperial University
of Tokyo (Director, Prof. Dr. M. Nagayo)
Received for publication January 20, 1920
A method for measuring the bactericidal activity of blood in
vitro by employing the many stemmed capillary pipets of Wright,
has been recently devised by Heist (1). Heist has shown that
this method—which having been in part suggested by Prof. B.
F, Lacy, has been termed the Lacy-Heist method—may be used
for determining resistance or immunity to pneumococcus infec-
tion. By employing the same technic, Dr. Kolmer and I (2)
have shown the existence of a relation between the meningo-
coccidal action of the blood of normal animals and the resistance
of the animal to infection with virulent meningococci, and also
that the high natural immunity or resistance of certain of the
lower animals to the meningococcus is to be partly ascribed to
a higher. meningococcidal activity of their blood, and that the
bactericidal blood test as described by Dr. Heist, possesses defi-
nite value as a test or measure of bactericidal activity of the blood
for meningococci in vitro.
The object of the present investigation was to determine,
whether or not active immunization with virulent meningococci
in rabbits will be accompanied by an increase of the meningo-
coccidal activity of the blood, and also to study the nature of
the test for the measuring of antimeningococcal activity of the
blood in vitro.
5l
or
bo
TOITSU MATSUNAMI
EXPERIMENTAL
The experiments were conducted with a single strain of normal
meningococcus B!:2, the virulence of this strain in mice is shown
in the results presented in table 1.
The virulence test was conducted according to the method of
Hitchens and Robinson (3) with the exception that active guinea-
pig serum and serum-water-dextrose broth were used in prepar-
ing the meningococcus emulsion, instead of guinea-pig serum
alone.
TABLE 1
Results of virulence tests at varying pertods in mice with meningococcus strain B
suspended in serum-water-dextrose-broth*
APRIL 23, 1919 APRIL 26, 1919 MAY 19, 1919 JUNE 4, 1919 JUNE 20, 1919
=e |e) 7) er ei ae
2 Result 2 Result 2 Result 3 Result 2 Result
al ce Pave Ale i i nd
mgm. gms gms gms gms gms
AZO 12| D.24hrs. | 13} D.17 hrs. | 10} D.48 hrs. | 11) D.17 hrs .| 11) D.17 hrs.
0.5 11} D.17 hrs. | 12) D:17hrs. | 10) D.17 hrs: | 14 S. 11 S.
0.25 11} D.24hrs. | 11) D.17hrs. | 10} D.24hrs. | 11) D.20 hrs. | 11) D.17 hrs.
0.125 | 10 S. 10} D.17 hrs. | 10) D.24hrs. | 10 S. 10} D.24hrs.
0.06 | 10 S. 10} D.17 hrs. | 10 S. 10 S. 9} D.24hrs.
0.03 | 10 S. 10 S. 10 S. 10 S. 9 S.
Control | 10 S. 10 S. 10 S. 10 S. 9 S.
Control | 10 S. 9 S. 10 S.
* These tests were conducted with strain B transplanted every three days on
serum-dextrose-agar.
Tt Doses in 0.4 ec. of serum-water-dextrose-broth.
D. = died.
S. = survived seven days or longer.
Control = injected 0.4 cc. of serum-water-dextrose-broth.
Bactericidal test. The bactericidal tests were conducted prin-
cipally after the method described by Heist, a brief account of
which is as follows:
Several dilutions of the culture are arranged in sterile tubes and
allowed to run by capillary attraction into the many stemmed sterile
capillary pipets of Wright, numbered respectively, and measuring about
1] am indebted to Dr. K. Iyehara for conducting the identification test for
the type of this strain.
2 This strain was kindly furnished by Department of Serum Therapy of the
Institute.
———
_——_—_
MENINGOCOCCIDAL ACTIVITY OF BLOOD 53
9 cm. in length and about 1 mm. or less in thickness. The emulsion
in the pipets is now removed by touching the tip to moist sterile
gauze (which attracts the fluid but leaves a film of microdrganisms
sticking to the wall of the tube) and each is loaded to the same level
with blood, secured by pricking the cleaned skin surface. The pipets
are now sealed by dipping the tips in melted paraffin and they are incu-
bated for twenty-four hours, when a smear is made from each pipet
and stained for meningococci.
In order to measure the bactericidal activity both of normal
and immune rabbit blood, I have employed several modified
methods besides the original method of Heist, of which a further
description will be given in subsequent tests.
MENINGOCOCCIDAL ACTIVITY OF THE BLOOD OF THE NORMAL RABBIT
That the blood of the normal rabbit possesses high bacteri-
cidal activity for normal meningococcus, and that these animals
are highly resistant to infection with the same strain injected
intraperitoneally in doses of culture, according to body weight,
comparable with those given to the mice and guinea-pigs has
been reported by Dr. Kolmer and myself (2). In the present
investigation with normal strain B, it was found, also, that the
blood of the normal rabbit possesses about the same high menin-
gococcidal activity, as shown in table 2.
In conducting the test I have employed suspensions of 10
loops of eighteen hour serum-agar culture of meningococcus in
1 ce. of broth undiluted and in four dilutions prepared with
serum-water-dextrose-broth, namely 1:5, 1:25, 1:125 and 1:625.
For the method of cloring the distal end of the pipet, in-
stead of dipping it in melted paraffin, a “‘peep’’ flame was used,
after the method of sealing a Wright’s blood capsule; after the
blood has been drawn up, the empty end of the pipet is sealed
with a flame and then cooled: The blood is drawn toward
this sealed end. The distal end, whichis now left emptied of
blood, is sealed with a “‘peep”’ flame, and the first sealed end is
opened by filing, and then incubated.
As is to be seen from table 2 the results of the meningococcidal
blood tests with rabbit bloods conducted by the Lacy-Heist
method as described indicate that the growth of meningococci
54 TOITSU MATSUNAMI
in the capillary tubes varies in degree, and also that the blood of
the normal rabbit possesses high bactericidal activity for men-
ingococci. I have conducted a series of experiments in search of
a method which yields suitable results for the purpose of com-
paring the bactericidal power of the normal and immune rabbit
blood as follows:
TABLE 2
Meningococcidal activity of the blood of normal rabbits and rabbits immune to
bacteria other than meningococci
MENINGOCOCCIDAL TESTS
RABBIT WEIGHT
Undi, | 1:5 | 1:25 | 12125 | 1:625
grams
INGrEAP EBs. LO Se _ - _ _
Nr oo. 2 es cidteleie islet Se al ROOO asta t= +--+ + 4 —
INonmallisigsrent... acme bices 2,850) -+-+ -+- _ — —
Normal 4..................-}2,240|. ++ _ -- - _
INOEMABI Ss cists sec os 98S DOOM. steak +--+ -L — —
Normal Gis.) oj. cies cece eiets| 2e4OO (sec — — — —
IVORMAAST Oi). co ss ieeleieitis sereietents (Gt OO) stat ++ + — =
INiforree yey bos Es Beem Olean Hee 1,785 | ++ + = —
Norma GO teveveuctsi «coe seela ele Poe AO Ny toot — = — —
INGrmA VAD z as. 62,0 a) b's ate wee] OROOO en, set ++ a _ —
ING ee eh he ee RO Stat _ _ _ —
Mommas ID ht 552 2h) ct ee OM) oat yer eee ~ = a
Immune B. typhosus....... 3,100; ++ _ - — —
Immune gonococcus......... 3,000; ++ 4--++ + a —
Immune B. dysenteriae
(Shiga) Reet. Ah AEE 2,700; ++ = = - —
Immune sheep blood cells .. .| 3,050} ++ + _ _ —
Immune sheep blood cells... .| 3,100} ++ — = = —
Wonitrolderc tec cts cto neues ++ $+ +--+ ++ ++
PIAteSSaeniy reece ee Une.{ | Une. Une. 3,000 1,500
to 5,000} to 500 | to 100
* Prepared by suspending 10 loops of eighteen hour serum-dextrose-agar cul-
ture in 1 cc. serum-water-dextrose-broth.
7 ++ = heavy growth; + = light growth; — = sterile.
t Serum-water-dextrose-broth substituted for blood.
§ Plates were prepared by drawing the culture into the capillary tubes em-
ployed, expelling the culture, and then washing the cocci, adhering to the inner
wall, into Petri dishes with twelve changes of serum-water-dextrose-broth; the
number of the colonies gives an idea of the number of cocci adhering to the inner
wall of a pipet and subject to the bactericidal activity of the blood.
{ Colonies per plate; Unc. = too many colonies for counting.
i MENINGOCOCCIDAL ACTIVITY OF BLOOD 55
Experiment 1
In order to measure more accurately the growth of meningo-
cocci in capillary tubes after the incubation, three sets of experi-
ments were conducted on the same day, with the same blood and
culture; namely, the contents of one set of capillary tubes were
plated in Petri dishes and the number of colonies of cocci which
grew were counted; also the contents of the second set of tubes
was cultured on slants of serum-dextrose-agar to determine
whether any cocci survived in the tubes; and, furthermore, the
stained smears of the contents of the third set of tubes was
examined under the microscope for cocci, as described in the
original method.
Table 3 shows the results of these experiments: while the results
obtained by examining the stained smear proved quite compar-
able to the others, for the purpose of obtaining an accurate
result of the test, the plate method or slope culture method was
found to be preferable.
Experiment 2
In this experiment I have studied the influence of the length
of the incubation time upon the meningococcidal action of the
blood. In Table 4 are shown the results of the meningococcidal
blood test, conducted at various lengths of time of incubation;
namely, directly after the test, after thirty minutes, after three
hours and after twenty-four hours. It will be seen, by this
experiment, that the meningococcidal action of the blood is
almost completed within three hours after incubation of the
tubes has begun.
Experiment 8
In this experiment I have studied the influence of the num-
ber of cocci and the dilution of blood upon the bactericidal
action of the blood of rabbits. In order to enumerate the num-
ber of cocci for the test, I have employed the bacterial emul-
sion, instead of using the cocci adhering to the inner wall of the
pipet: One volume of this bacterial emulsion was mixed with an
equal volume of blood in the pipet as follows:
pe ee Oe ee
fe
a gee
TOITSU MATSUNAMI
56
ay} OJUL IvBv-ssoi}xop Butunod usyy pus ‘yjo1q-
‘19}VM WINIIS BULUIe}UOD SoYystp I1yaq OFUT Soqny Arey[ides oy} Jo s}uezU0d oy}
dIN{[Ne IvBv-d801}Xo9p 1o}VM-
. * soqsIp
WINIOS YIM souty [erodes yodrd oy} Jo [[BA JouUr ot]} Suryseam Aq pus
Burmolq Aq poredoid o19M soyerd $380} osoyy UT
‘SUIJUNOD IO SoTuo[oo AUBUT 00} = “OUL) fayetd 19d soruopop ft
‘g[ldays = — {ypMoIZ 4YySI] = + ‘q}MoI3 favoy = ++ I
“YJOIq-o801}X9p-19}BM-WUNIIS JO 990 T UL
uIn1as noy wsee7Yy3tIe Jo sdooy OT Surpuedsns Aq poredaid uors[nte estep Y}IM popONpUoY ,
SE eS eee
‘aug | ‘oUQ | “oUQ | “Ou
— — — | o0r'z
— — — ‘ou
_ _ Moj —_
— — — 09
«}89} [Bp (o0000Zulua
INQ000 ALVId
-ougy |++[+4+)+4+]4+4]4+4]4+4+]4+4[44|4+4+| ++ Sue ket [OUIUOG)
rouge) = We lee leer | — = le |e | te O02! gi) o> (amy) eutseyuesED el amma
‘ou £5 ase EE (seat niet ee oe Seas) ee 000‘ shee helieliabal olenersisiie * + sno9000U08 aUNUIWT
‘ou, be = 2 = + rel 2s os +4 oor‘ sejeisieci> Hoe TNO UCLA ‘gq ouNUIWT
‘ou Ss a tae Ss ++ a rave aS = ++ OOr‘s shedanias waste) nkebure ie) einssnene\e cess emelns Il [BUIION
‘ou a =— a + t+ eee as £8, + +4 OS Se a Se OD LEEON
foug oa ae a = ++ x5 a= oe es {++ OFS‘'Z sip farie}ce. elre/elel-e.el ene) .eieie,e. (6:66 amen m eae 6 [BULLION
sub
Bea se| olga ap beelless| i \aweleneag
+p bo bo o +5 nr bo o +p
ae on on ae on on ae
DTHSIaM LIGdvu
-8}89} [Bplovo00suTUaTY
@YoOL1T090 7AdO18
UVAWS GANIVLIS
«3899 [Bp loo000au UOT
Ee ek ae Se ie ee SSS ee
junoo 940)d pun aunzqna adojps ‘anaus pours fig pauyutajap sD S7rqqns YILM $789}? P0OT4 popr0090buruau fo sznsoy
€ WIaAve
ae
MENINGOCOCCIDAL ACTIVITY OF BLOOD 57
Each pipet is marked at two levels. After the culture was
allowed to run into the pipet by capillary attraction and to
reach the first mark, the tip of the tube was withdrawn from the
culture tube and introduced into the drop of blood, care being
taken to avoid air bubbles between blood and culture; the blood
flows up by capillary attraction. When the ascending column of
the mixture of blood and culture has reached the second mark,
the tube is moved aside and mixed well by keeping the tube con-
TABLE 4
Results of meningococcidal blood tests with rabbits after various lengths of incubation
time
LENGTH OF MENINGOCOCCIDAL TESTS*
RABBIT WEIGHT INCUBATION
ae Undiluted| 1:5 | 1:25 1: 125
grams i eee er a en
At once Une: iadne: 5,400 | 2,700
, 30 minutes | Une. | 5,400 240 0
MNGTMANS....-..-.6...| 1,785 B hates Une 1 0 0
24 hours Une. 0 0 0
At once Une. Une. 5,400 | 2,480
30 minutes | Une. 8,100 800 30
NGPA eo. oldies o 2,540 Bh ae 15 0 0
24 hours Une. 0 0 0
At once Une. Une. 6,950 | 2,700
30 minutes | Une. | 10,800 780 60
MarmialiO............| 3,050 re Tel 12 0 0
24 hours Uhe. 0 0 0
* These tests were conducted with a culture prepared by suspending 10 loops
of eighteen hour serum-water-dextrose-agar culture in 1 ee. of broth.
{ Colonies per plate; Unc. = too many colonies for counting.
stantly rotating with a lateral movement and bringing the
column of fluid in the pipet up and down to insure a uniform
mixing.
By employing this method the number of coeci was enumerated
on the one hand, and the blood was diluted one to two on the
other hand. It is seen from table 5, in which the result of this
experiment is given, that the meningococcidal activity of rab-
bit’s blood is reduced by this method as compared with that
obtained by the original method.
58 TOITSU MATSUNAMI
It was found in these experiments that this meningococcidal
blood test, conducted by mixing equal volumes of blood and
culture in capillary tubes, incubating for three hours and then
plating for the purpose of counting the surviving cocci, may be
used for comparing the meningococcidal activity of normal and
immune rabbit’s blood.
TABLE 5
Meningococcidal activity of rabbit blood, equal volumes of blood and culture being
used
MENINGOCOCCIDAL TESTS*
RABBIT WEIGHT In slope culture In plate count
Undi- | 4:10 | 1:100 | 1:1000) UR | 1:10 | 1:100 | 1:1000
grams iio a Ones: |
Normal 8.........| 1,785 |+++|+++| ++ | ++ | Unc.{| Unc. | 8,100; 600
Normal 9........./ 2,540 /+++/+++] ++ | + | Une. |10,800| 2,700} 360
Normal 10........ 3,050 |+++/+++/+++| + | Une. /13,500/10,800) 800
ImmuneB.typhosus.| 3,100 |+++/+++/+++] + | Une. | Une. | Une. | 5,400
Immune gonococcus| 3,000 |+--+-+)/+++|++-+]| ++ | Une. | Une. 13,500) 5,700
Approximate num- 10 1 aed 10 10 1 100 | 10
ber of cocci em- mil- | mil- | thou-| thou-} mil- | mil- | thou-} thou-
ployed for the lion | lion | sand | sand | lion | lion | sand | sand
UGEWs aig cats ome
* These tests were conducted with a culture prepared by suspensing 5 loops of
eighteen hour serum-water-dextrose-agar culture in 1 cc. of serum-water-dex-
trose-broth.
+ +++ = heavy growth; +-+ = moderate growth; + = light growth.
t Number of colonies per plate; Une. = too many colonies for counting.
MENINGOCOCCIDAL ACTIVITY OF THE IMMUNE RABBIT BLOOD FOR
MENINGOCOCCI
1. Method of «immunization employed by the author
A number of rabbits were immunized with meningococci principally
after the method of Amoss and Wollstein (4), which was devised for
preparation of antimeningococcic serum in the horse. It consisted in
inoculating, alternately, living and autolyzed products of meningococci
into the vein. Living cultures of meningococci grown on serum-dex-
trose-agar slant for eighteen hours were given to each rabbit in doses
as follows: on the first day 0.02 cc., on the second 0.05 cc. and on the
MENINGOCOCCIDAL ACTIVITY OF BLOOD 59
third 0.1 cc. of one loop of cocci suspended in 2 cc. of normal salt solu-
tion. After the lapse of seven days a series of three injections of auto-
lyzed meningococci, consisting of the filtrate of a suspension of meningo-
cocci in normal salt solution, which had been incubated for twenty-
four hours at 37°C., in doses of 2 cc. containing autolysate of 0.1 cc. of
one loop of meningococcus culture in 2 cc. of normal salt solution was
injected. After the lapse of seven days, for the second time, two series
of three injections each of living and autolyzed cocci were given in the
same manner. ‘This process of immunization was repeated for the third
and the fourth time.
2. The bactericidal blood test
The bactericidal blood test was conducted by the method
described above. Equal volumes of blood and culture were
mixed in a capillary tube, incubated for three hours, and plated
for counting the surviving cocci.
3. The agglutination test
The agglutination test with the serum was conducted before
the beginning of each immunization.
The results of a number of these tests are summarized in
table 6.
As shown in table 6 the meningococcidal activity of immun-
ized rabbit blood has generally been found to be stronger than
that of normal blood or of that drawn before immunization.
This increase of bactericidal activity was marked in the blood of
rabbits that had received only the first series of the injections of
cocci, and distinctly more with rabbits that had received the
injections for the second time. But, as may be seen from the
results recorded in table 6, the blood of rabbits which had re-
ceived more than two series of injections have been found to be
quite irregular as far. as bactericidal activity in vitro, is con-
cerned. The bactericidal activity of the blood does not accom-
pany the increase of the agglutination titer of the serum. The
bactericidal activity of the blood of rabbits 3 and 7 as regards
the tubes which contained the smaller number of cocci, and of
60 TOITSU MATSUNAMI
TABLE 6
Results of meningococcidal blood and agglutination tests with immune rabbits for
meningococcus strain B
RAB- AGGLUTI- MENINGOCOCCIDAL TESTS*
BIT BLOOD WHIGHT| ‘NATION. |-o= Use te iy EEE eee
No. TITERT | Undiluted 1:10 1:100 | 1:1000
grams
mie 2,710)| 1 = 20 Une.t{ Une. 3,780 180
eee 4 | 2,0 eD Une. 16,200 | 4,700 | 240
2,710 | 1:20 Une. 13,500 | 3,780 | 300
1
ae Ij 2,500 | 1 : 40 13,500 2,700 120 12
Reenter eee 5,400 120 120 0
IIT) 2,450 | 1 : 640 18,900 3,780 600 8
Before 2,595 | 1 : 20 Une. 13,500 5,400 300
immunization Une. 3,780 420
kb
or
te}
or
_
to
oO
(S|
i=}
fe)
—
tS
So
—
oO
iw)
(=)
oO
5,400 180 10
F PREY II| 2,550 | 1 : 640 3,240 360 60 0
immunization J
Reler I} 2,450
IIT} 2,500 3,280 800 12
—_
i)
—_
(or)
So
oO
Before
immunization 2,300 | 1:10 Une. 8,100 5,400 360
Ne J) 2,250} 1 : 80 13,500 4,700 240 180
He ae II} 2,170 | 1 : 640 8,100 10,800 80 60
III} 2,400 | 1 : 1280 140 0 | 10,800 | 5,400
Before
immunization 2,240 | 1: 20 Une. 13,500 7,800 300
“After Ij 2,100 | 1 : 80 10,800 3,200 180 8
eT eae II} 1,750 | 1 : 320 16 5,400 240 0
III} Died
Before
immunization 2,500 | 1: 10 Une. 7,560 8,100 320
5
After ]| 2,350; 1 : 160 8,100 5,400 360 | 730
immunization |II| Died
Before
immunization 2,450} 1:10 Une. Unc. 8,100 | 600
O Wegewe 1/2,3001:80 | Une. 3,780 360 | 0
zt a II} 2,150 | 1 : 320 5,400 600 180 | 320
III} 2,400 | 1 : 640 8,100 0 120 | 600
MENINGOCOCCIDAL ACTIVITY OF BLOOD 61
TABLE 6—Concluded
RAB- AGGLUTI- MENINGOCOCCIDAL TESTS*
BIT BLOOD WEIGHT NATION
No. TITER} | Undiluted 1:10 1:100 | 1:1000
grams
Before
immunization 1,750 | 1: 10 27,000 10,800 5,400 360
7 Ij 1,750 | 1 : 80 18,900 3,780 120 60
After II} 1,950 | 1 : 640 3,780 600 680 11
immunization] III) 2,000 | 1 : 640 600 8 5,400 |3,240
IV| 2,100! 1 :1280} *18,900 180 800 360
Approximate number of cocci employed gee alice ce
to to to to
HOTMRUIVORLES Gena bays cielcbesch leis boeiei sues 10,000,000 | 1,000,000 | 100,000 | 10,000
* Conducted with an equal volume of blood and culture, prepared by suspend -
ing 5 loops of eighteen hour serum-water-dextrose-agar culture in 1 cc. of broth.
t Agglutination tests were conducted at 55° C. for twenty hours.
t Colonies per plate; Unc. = too many colonies for counting. Each immuni-
zation consisted of three successive injections of living and, after seven days
interval, autolyzed product of meningococcus respectively.
rabbits 1 and 2 as regards nearly all the tubes employed in these
tests, has been found to be much less after the third series of
injections than that of blood taken after the first or second series.
In regard to this phenomenon I have found a similar result in a
horse highly immunized with meningococci. As shown in table
7 the bactericidal activity of this horse conducted by the method
described, has been found surprisingly low compared with that of
normal horse blood.
TABLE 7
Showing that highly immunized horse blood is less bactericidal for meningococct
than normal horse blood
In
HORSE AGGLUTINATION MENINGOCOCCIDAL
TITER* TESTST
STINET CRS AEP elo oho eraicc oes c cigha, 6.6 odie o/s) aislarepeme a 1 : 1280 Uncountable f
WIGITBTGN Sc Be SNe eee 1 : 160 600
* Agglutination tests were conducted at 55° C. for twenty hours.
+ Conducted with equal volumes of blood and culture, prepared by suspend-
ing 10 loops of eighteen hour serum-dextrose agar culture in 1 cc. of serum-water-
dextrose-broth.
t Colonies per plate; Uncountable = too many colonies for counting.
62 TOITSU MATSUNAMI
This phenomenon may possibly fall in a class with phenomena
of asomewhat similar nature. For instance, in the case of agglu-
tination, clumping of the bacteria may not occur in low dilu-
tions of an immune serum, while it may be complete in high dilu-
tions. As another instance, Neufeld (5) and others state that
too much agglutinin in a serum inhibits opsonie activity, result-
ing in irregular and low bactericidal activity for meningococci.
It may be that in the process of immunization the blood acquires
the power of inhibiting autolysis of meningococci. Whether
one or the other of these factors, or a summation of them all is
responsible for the facts observed, or whether some other factor,
(or factors) is concerned is not determined by the present inves-
tigation.
The results of my tests with normal and immune rabbit blood
may be summarized as follows:
1. The blood of normal rabbits and rabbits immunized against
B. dysenteriae (Shiga), B. typhosus and gonococci has been found
to possess marked bactericidal activity for a virulent normal
strain of meningococcus within three hours 77 vitro.
2. The blood of a rabbit immunized against meningococcus by
intravenous injection of a living and autolyzed culture of men-
ingococcus has been found to be distinctly more bactericidal
than that of normal rabbits or rabbits immunized against B.
typhosus, B. dysenteriae (Shiga) and gonococcus.
3. The increase of the meningococcidal activity of the blood of
rabbits after immunization has been found to be parallel up to a
certain limit with the process of immunization, but further im-
munization did not appear to show increase of bactericidal
activity in vitro.
STUDIES OF THE NATURE OF MENINGOCOCCIDAL ACTIVITY OF
RABBIT BLOOD IN VITRO
Discovery of the bactericidal activity of the blood for
various microorganisms (6) (7) and that of the phenomenon of
phagocytosis have thrown a light on the explanation of the re-
sistance of the organism to certain microorganisms. By further
ae
a SS
ee a
a eee
_
MENINGOCOCCIDAL ACTIVITY OF BLOOD 63
investigations it has been proved that the bactericidal activity of
the blood is to be ascribed to the action of the various immune
bodies or antibacterial substances contained in blood serum or
plasma, in phagocytes and in blood platelets. In this respect I
have conducted a series of experiments to study the nature of
the meningococcidal activity of the blood in vitro as follows:
1. Bactericidal activity of the serum for meningococer
Potent antimeningococcus serum is generally regarded as
possessing a certain meningococcidal activity in addition to
specific opsonin, toxin neutralizing antibodies, agglutinins, pre-
cipitins and the complement fixing antibodies, and upon these its
curative powers probably depend. But the meningococcidal
activity of antimeningococcus serum does not appear to be
ascribable to the presence of specific bacteriolysin requiring the
presence of complement for its lytic activity: While Davis (8)
and McKennzy and Martin (9) were able to demonstrate in
vitro the complemental bacteriolysis with the serum of meningitis
patients, Flexner (10) proved that heated serum also possesses
bactericidal activity, and he advanced the hypothesis that it is
only necessary that the fresh or heated serum should injure the
cocci in order that their intracellular enzymes should be rendered
active, and thus destroy the microorganism. Jobling (11) and
others deny the action of complemental bacteriolysis with the
antimeningococcus serum. According to the results obtained
by Drs. Kolmer, Toyama and myself (12) (13) while studying
the influence of active normal serum (complement) upon mening-
gococci, the bactericidal activity in vitro of horse antimeningo-
coccus sera is quite low, although some bactericidal activity is
generally apparent as compared with the control, and largely
independent of complemental bacteriolysis; it was also found
by our experiments, that the higher dilutions of serum not
infrequently are more bactericidal than the lower. Jochmann
(14) also found this to be the case. In the present investigation
I have conducted a comparative study of the meningococcidal
activity of rabbit blood and serum and controls, the results of
64 TOITSU MATSUNAMI
which are given in tables 8 and 9. As shown in table 8 normal
rabbit’s sera have been found to posses a certain meningococcidal
activity, quite comparable with that of the sera of immune
TABLE 8
Showing that whole blood is more bactericidal for meningococci than serum*
AGGLUTINATION RESULTS WITH | RESULTS WITH
per man | SCO? A>) | Sanaa
IUCeo eT LES aa aie ee por NN IE Rm Thee a0) 8, 100+ 18,900
AQ ETE EO State cee caine ais Sarikei 1520 13,500 Une.
WormalelOQi t icguaoes do cess oteaee en 13420 10,800 16,200
Normal 11... 1:20 13,500 Une.
PMMIMUTIO Mees. oes wos w+ slow easeise RPE e Oars 1 : 640 600 16,200
framaunnera te ttt. o.oo cate cone oon ee 1 : 640 60 Une.
minum erstpaeces). teint eee ee ae 1:80 600 16,200
lieci vin, Oe pees Sos. oe oe soca Sloman 1 : 320 300 8,100
ARENA apg o'= «ico A nares ee ik eee en 1 : 1280 600 Une.
* These tests were conducted each with an equal volume of blood or serum with .
twenty-four hour egg-yolk-dextrose-broth culture; one volume of this culture
contained approximately 50,000 meningococci.
+ Colonies per plate; Une. = too many colonies for counting.
TABLE 9
Antimeningococcidal action of normal salt solution, sodium citrate in normal salt
solution, cerebrospinal fluid and normal rabbit serum
MENINGOCOCCIDAL TESTS*
APPROXIMATE
NUMBER OF
COCCI EXPOSED 1.5 per cent Normal rabbit
serum Control
sodium citrate .
to ee, ee ermal cel solution in ere Dt es spinal (inactivated at | (egg yolk dex-
normal salt 55°C for thirty trose broth)
solution minutes
80,000 OF 0 Uncount.t 18,900 Uncount.§
* These tests were conducted with meningococci adhering to the inner wall
of a capillary tube; culture of which is prepared by suspending 5 loops of eighteen
hour serum-dextrose agar culture in 1 ee. of broth.
t+ Number of colonies per plate: Uncount.t = Colonies in these plates ap-
peared to be comparable to the plate containing the cocci exposed to germicidal
action of fluids; Uncount.§ = Colonies in these plates appeared to be more than
in the plate containing the cocci exposed to germicidal action of fluids.
rabbits. While the normal rabbit blood is more bactericidal
than serum, the meningococcidal activity of the immune rabbit
blood has been found to be distinctly stronger than that of
MENINGOCOCCIDAL ACTIVITY OF BLOOD 65
immune serum. ‘The results of the experiment indicate, there-
fore, that the bactericidal activity of the rabbit blood can not be
ascribed totally to the serum. As found by Flexner (10) and
others, normal salt solution and 1.5 per cent sodium citrate in
normal salt solution were very toxic for meningococci, while the
cerebrospinal fluid did not appear to be toxic for them within
three hours in vitro, as shown in table 9.
2. Bactericidal activity of the defibrinated blood for meningococct
The meningococcidal activity of the defibrinated rabbit’s
blood, prepared by shaking with glass beads, has been found to
be quite similar to that of the serum and to lack the high bac-
tericidal activity of the whole blood as shown in table 10.
TABLE 10
Showing that whole blood is more bactericidal for meningococci than defibrinated
blood*
RESULT WITH
=~ RESULT WITH =
RABBIT AGGLUTINATION WHOLE BLOOD DEFIBRINATED
TITER BLOOD
AND CULTURE | ,xp CULTURE
: 640 6007 18,900
rane W PEE ere acc tie oe. cyerdicleteiaiseede «| OL
Ane eee oe keke cycinecescses| 1 2640 120 Une.
Mera Men er Pet fy. e.8 ia cinietslsee'ecle | Lt 820 360 16,200
NCNM. al encosciisene.sc.c)|, 2 21280 600 Une.
* In these tests a twenty-four hour egg-yolk-dextrose-broth culture was em-
ployed; the number of meningococci in one volume of culture being approximately
50,000.
Agglutination tests were conducted at 55° C. or twenty hours.
{t Number of colonies per plate; Unc. = too many colonies for counting.
8. Bactericidal activity of rabbit's blood, consisting of blood cells
and serum
The bactericidal test with the blood, which is prepared by a
method shown in table 11, consisting of whole formed constitu-
ents of the blood and serum, has been conducted with the men-
ingococci. The result of this experiment, shown in table 11,
indicates that the meningococcidal activity of the untreated
blood is still distinctly higher than that of the blood cells plus
serum, although the bactericidal activity of the latter has been
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 1
66 TOITSU MATSUNAMI
found to be generally slightly stronger than that of serum alone
or of defibrinated blood. It will be seen by this experiment
that the difference in bactericidal activity between the two
kinds of blood is not to be ascribed to the difference in the num-
ber of leucocytes acting as phagocytes; as the leucocytes con-
tained in both samples of blood may.be regarded as about the
same in number.
TABLE 11
Showing that whole blood is more bactericidal for meningococci than blood cells
plus serum*
RESULT WITH
RABBIT AGGLUTINATION Beep eeu BLOOD CELLS,
. TITER SERUM
AND CULTURE | ,Np CULTURE
1G TaoTTOUAVS) 1] UP Aen erre Mauve mee ones dict ecole dh oGz00 3007 10,800
Lica) rr ena eae Aloe ang rail TES (2270) 600 16,200
Immune 6.............. 2.0602 ee sees eee! 1 : 320 300 8,100
Immune 7....-.6. 6.0 seer eee eee eee] 1 : 1280 60 18,900
* These tests were conducted with a twenty-four hour egg-yolk- dextrose-
broth culture in equal volume and incubated for three hours.
Number of cocci in one volume of the culture being approximately 50,000.
+ Number of colonies per plate.
The blood used for this experiment, consisting of blood cells and serum, was
prepared as follows:
One cubic centimeter of immune rabbit blood was taken from an ear vein by
puncture into a syringe containing 4 cc. of sterile 1.5 per cent solution of sodium
citrate in normal salt solution. This blood-citrate mixture was transferred to
a sterile centrifuge tube and centrifuged. The supernatant fluid was transferred
into a second sterile centrifuge tube and centrifuged again at a high speed. The
supernatant fluid was then drawn off. The blood cells left in the first and sec-
ond centrifuge tubes were mixed with the active serum, which was previously
separated by coagulation of 1 cc. of the same rabbit blood, thus making the mix-
ture of blood cells and serum correspond to a certain amount of blood. This
mixture of blood cells and serum was then used for comparative bactericidal
blood tests for meningococci with whole blood.
4. Bactericidal activity of citrated blood
For a study of the bactericidal activity of the blood pre-
vented from coagulating I have conducted the test with blood,
the coagulation of which was prevented by the use of sodium
citrate. The result is given in table 12. Since sodium citrate
in normal salt solution is very toxic for meningococci, as shown
MENINGOCOCCIDAL ACTIVITY OF BLOOD 67
in table 9, and inasmuch as Otani (25) found that spontaneous
phagocytosis in a mixture of citrated blood of certain micro-
organisms occur quite freely the results of experiments with
citrated blood, may, strictly speaking, not be comparable with
those with untreated blood, but the bactericidal activity of both
bloods being concerned, the untreated whole blood was found
more bactericidal for the meningococcus than blood which was
prevented from coagulating by the use of sodium citrate,
TABLE 12
Showing that whole blood is more bactericidal for meningococci than citrated blood*
RESULT WITH A EMEY ANE
RABBIT AGGLUTINATION} WHOLE BLOOD CLERAED
TITER AND CULTURE BLOOD AND
> CULTURE
PT CRORE ee i) coc te is te ne ctsean-o| 1 21280 54007 Uncount.
MPTP e eh pein ch cwce es oeivcs eens .| 2 2.1280 8100 Uncount.
* These tests were conducted with a twenty-four hour egg-yolk-dextrose-
broth culture in equal volume and incubated for three hours.
Number of cocci in one volume of the culture counting approximately 80,000.
{7 Number of cocci per plate; Uncount. = too many colonies for counting.
Citrated blood was prepared by adding 0.02 gram of sodium citrate to 1 cc.
of blood.
5. Influence of coagulation of the blood upon the meningococcidal
activity of the blood
Since the results of experiment 1, 2, 3 and 4 indicate that the
bactericidal activity of whole blood is greater than that of serum
alone or defibrinated blood or blood cells plus serum or citrated
blood, the reason for the difference of the bactericidal activity in
vitro between whole blood and others appears to be in the process
of coagulation of the blood. Although the direct influence of
coagulation of blood upon meningococci may be negligible for
the comparative study of the bactericidal blood test in vitro by .
reason of the fact, that the meningococci will not be destroyed by
the process of coagulation of blood alone if the blood does not
possess bactericidal activity as indicated in the results of ex-
per:ments with mice and other animals (2), coagulation taking
place in the test in both immune and normal blood. It is, how-
68 TOITSU MATSUNAMI
ever, highly probable, that the indirect influence of the process
of coagulation plays a considerable and important réle in the
bactericidal blood test in vitro with respect to the meningococci.
As pointed out by Wright and Dr. Heist (1), when the blood is
allowed to clot in the capillary tube a ‘‘semi-solid” is formed,
and in this condition and furthermore by the possible influence
of chemical changes of blood which occur in the clotting, phago-
eytosis, by which meningococci undergo intracellular digestion,
may be more active than that of the blood ina fluid state.
Therefore the influence of coagulation of blood, at least as one
factor of difference between the whole blood and defibrinated
blood and blood cells plus serum or citrated blood may not be
a negligible one with respect to the bactericidal activity for
the meningococcus in vitro. But it appears difficult to ascribe
this difference solely to the influence of coagulation of the blood,
because the degree of that influence upon the meningococcidal
activity of blood can not be determined, inasmuch as a method
to measure the bactericidal activity of whole blood which can
be prevented from coagulating without making any change in
the original nature of blood, is unknown. According to Dr.
Heist (1), the blood of pigeon, which is highly resistant against
pneumococcus infection, has distinct antipneumococcic factors
in vitro, while this activity is not to be found in serum or in
defibrinated blood or in blood of the pigeon influenced by
coagulation.
The results of the preceding study on the significance of bac-
tericidal activity of the blood for meningococci may be summa-
rized as follows:
1. The meningococcidal activity of whole blood of normal
rabbit in vitro was found to be stronger than that of serum.
2. The meningococcidal activity of serum and of defibrinated
blood of the immune rabbit for meningococci was found com-
parable with that of the normal rabbit and slightly less strong
than that of blood cells plus serum of the immune rabbit.
3. It was found that the meningococcidal activity of whole
blood of the immune rabbit is strikingly stronger than that of
serum, defibrinated blood, blood cells plus serum or citrated blood
_of the same rabbit.
MENINGOCOCCIDAL ACTIVITY OF BLOOD 69
4, It is suggested, that at least one factor explaining the differ-
ence in meningococcidal activity in vitro of whole blood, defibrin-
ated blood, citrated blood and serum is because coagulation in
the bactericidal blood test with whole blood, favors phagocytosis
of the meningococci.
SUMMARY
1. It has been found by the pipet method that normal rabbit
blood and serum are capable of killing considerable numbers of
virulent normal meningococci 77 vitro within three hours.
2. The meningococcidal activity in vitro of normal rabbit blood
was found to be increased up to a certain limit by the intravenous
injection of the living and autolyzed meningococci. After that,
further immunization did not appear to increase bactericidal
activity, was generally rather irregular and not infrequently
even decreased meningococcidal activity of the blood. The
more highly immunized rabbit’s blood was found sometimes less
bactericidal than that of slightly immunized rabbit’s blood.
3. The meningococcidal activity of normal rabbit’s serum has
been found not to be increased by artificial immunization and
also to be comparable with that of defibrinated blood of an
immune rabbit.
4. The meningococcidal activity 7 vitro of immune rabbit’s
blood was found by the pipet method to be distinctly stronger
than that of the serum, of defibrinated blood or of blood con-
sisting of blood cells and serum or of citrated blood. It was
suspected that at least one factor in explaining the higher men-
ingococcidal activity in vitro of immune rabbit’s blood compared
with defibrinated blood, citrated blood and serum lies in the
influence of coagulation of the blood, which is permitted in the
regular blood test as described, favoring the phagocytosis of
meningococci.
5. The meningococcidal blood test can not be accepted on the
basis of the present investigation for the purpose of measuring or
determining the artificially induced immunity against meningo-
cocci.
However, as pointed out by Dr. Heist (1) the bactericidal
blood test described possesses the advantage of employing whole
70 TOITSU MATSUNAMI
blood; hence when this method is used any mechanism of im-
munity existing in the blood, may be brought into direct relation
with the microorganism. Moreover, in this test any anti-
bacterial factors existing in the blood, and the influence of indi-
vidual variation of fluid constituent as well as phagocytes in blood,
upon the bactericidal activity of the blood, have equal chance
to come into play. Furthermore, it was found that a parallelism
between the bactericidal activity of the blood and resistance to
certain bacteria including meningococci, exists under normal
conditions (1) (2). Therefore, the bactericidal blood test de-
scribed may be regarded as a method possessing definite value
for the measuring the natural resistance of the organism for cer-
tain microorganisms especially for meningococci, inasmuch as
with respect to meningococci no accurate method sufficiently
reliable to serve as a definite measure of antibody content has
yet been devised for measuring immunity to this microor-
ganism (16).
I wish to express my thanks to Dr. Yutaka Nakamura for
advice and asistance in carrying out this work.
REFERENCES
(1) Hetst, G. D., Sotts-CoueEn, S., anp Sonis-Conen, M.: The bactericidal
action of whole blood, with a new technique for its determination.
Jour. of Immunology, 1918, 3, 261.
(2) Marsunami, T., AND Koutmer, J. A.: The relation of the meningococcidal
activity of the blood to resistance to virulent meningococci. Jour. of
Immunology, 1918, 3, 201.
(3) Hircuens, A. P., AnD Ropinson, G. H.: Standardization of antimeningitis
serum. Jour. of Immunology, 1916, 2, 345.
(4) Amoss, H., anp Wotusre1n, M.: A method for the rapid preparation of
antimeningitis serum. Jour. Exper. Med., 1916, 23, 403.
(5) Neureip, F.: Bakteriotropine und Opsonine. Kolle und Wassermann’s
Handbuch d. Patholog. Mikro., 1913, 2, I, 416.
(6) FrrEDBERGER, E.: Die Bakteriziden Sera. Jolle und Wassermann’s Hand-
buch d. Patholog. Mikro., 1913, Bd. 2, I, 298.
(7) Zinsser, H.: Infection and Resistance. McMillian, 1918, Ed. II, 134.
(8) Davis, D. J.: Studies in meningococcus infections. Jour. Infect. Diseases,
1905, 2, 602-619.
(9) McKenziz, D., ann Martin, W. B. M.: Serum therapy in cerebrospinal
meningitis. Jour. Path. and Bact., 1908, 12, 539-549.
Bene
a en
MS, a
sl a ae
MENINGOCOCCIDAL ACTIVITY OF BLOOD 71
(10) Fiexner, §.: Contribution to the biology of diplococcus intracellularis.
Jour. Exper. Med., 1907, 9, 105-141.
(11) Jopuine, J. W.: Standardization of the antimeningitis serum. Jour. Exper.
Med., 1909, 11, 614-621.
(12) Koutmer, J. A., Toyama, I., anp Matsunamt, T.: The influence of active
normal serum (complement) upon meningococci. I. Jour. of Immu-
nology, 1918, 3, 157.
(13) Matsunamt, T:, AND Koutmemr, J. A.: The influence of active normal serum
(complement) upon meningococci. II. Jour. of Immunology, 1918,
3, 177.
ea) JOCHMANN, G.: Versuche zur Serodiagnostik und Serumtherapie der epidem-
ischen Genickstarre. Deutsch. Med. Wochenschr., 1906, 32, 788-793.
(15) Oranr, M.: On the acceleration of phagocytosis in the Feomaied blood. The
Kitasato Archives of Experim. Med., 1918, 2, 2, 147.
(16) Kotmer, J. A.: Infection, Immunity and Specific Therapy. Saunders,
1917, Ed. II, 792.
DESCRIPTION OF FIGURES
Showing colonies of meningococci grown in Petri dishes, survived of the
meningococcidal blood test in the capillary tube; C = Coagula of blood. (The
author’s modified method for estimating the result of the meningococcidal blood.
test in the capillary tube.)
I. Plate of strong meningococcidal blood
II. Plate of marked meningococcidal blood
Ill. Plate of weak meningococcidal blood
“i
bo
4
PLATE 1
MENINGOCOCCIDAL ACTIVITY OF BLOOD
T. MATSCUNAMI
r
ae
ere a
tee - SS
— : ~ 2
aes
73
Sy ret ict aE. ACM oi si .
eal pet Sara, Care re A Eoin a eS mies
¥ + *
: 7 » Vier
: . ; ’ F : ) aye
: t '
>A.
- ‘
——| TP rs
. & -
® ‘es E
A a
é
-
aa 7 < £
: 5
*
; - te
&
’
7
f
'
=
s -
- 7 va
. <
4
: , ‘
p x
tnd, i
_*.
ite
i
‘
20
7
« ‘
’
“
-
.
. Mi ; .
=
o
NATURAL ANTIHUMAN HEMOLYSINS AND HEMAG-
GLUTININS IN HORSE SERA IN RELATION
TO SERUM THERAPY
JOHN A. KOLMER anp MOTOMATSU MATSUMOTO
From the McManus Laboratory of Experimental Pathology of the University
of Pennsylvania
Received for publication February 9, 1920
Inasmuch as the intravenous injection of persons with large
amounts of immune horse serum as in the serum treatment of
pneumonia, is sometimes followed by symptoms not referable to
anaphylaxis, the purpose of this investigation was to study nor-
mal and immune horse sera for hemagglutinins and hemolysins
for human erythrocytes to ascertain whether intravascular agglu-
tination or hemolysis are responsible in part for these symptoms.
This study was conducted with agglutination and hemolysin
tests in vitro in which normal and immune horse sera and the
erythrocytes of different persons were employed; horse sera were
found to contain relatively large amounts of agglutinins for the
erythrocytes of some rabbits and were injected intravenously
into these animals to determine the effects of agglutination and
hemolysis in vivo.
Williams and Patterson (1) tested 19 horse serums and found
12 to contain agglutinins for human erythrocytes, and suggest
that in fatalities following large injections of serum, the possi-
bility of agglutination of red corpuscles is to be considered; they
also suggest the advisability of testing the sera of horses against
human corpuscles and of rejecting any horse whose serum had
distinct agglutinating power.
In this investigation 30 horse sera have been used, 28 being
obtained from immunized and 2 from normal horses; all sera
were kindly furnished by Dr. John Reichel of the Mulford
Biological Laboratories. The majority of immune sera con-
tained a preservative as prepared for distribution and adminis-
tration to persons.
75
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 2
76 JOHN A. KOLMER AND MOTOMATSU MATSUMOTO
RESULTS OF AGGLUTINATION TESTS
Macroscopical tests. Macroscopical agglutination tests were
conducted by mixing in small test tubes 0.1 ec. of each horse
serum unheated with 1 cc. of 1 per cent suspension of washed
erythrocytes from 9 to 12 different persons, the final dilution
being about 1:11; results based upon tests with the erythro-
cytes of any one person were found only approximately correct
since the agglutinins in horse sera for human erythrocytes occur
in groups, similar to the groups of agglutinins and hemolysins in
human sera for the erythrocytes of the lower animals (2).
TABLE 1
Agglutinins in unheated horse sera for human erythrocytes
HORSE SERUM * HUMAN ERYTHROCYTES, 1 cc. oF 1 PER CENT SUSPENSION
UNHEATED 5.1 cc.
Antistreptococcus..| — = = = = = = = =
Antistreptococcus..}| — = _ = = = = = =
Antistreptococcus..| — —= - = = = = a =
Antistreptococcus..| — _ - - = = = = ote
Antistreptococcus..}| — - — — = = = SF =
Mormal ncaa... sails _ = - = = = = = =
Normal, 2a5). cocdie — = = =
Antipneumococcus..| + + — + _ ais
Antistreptococcus..| — — + =
Antipneumococcus..| — _ _ _
Antistreptococcus..| — - + —
Antistreptococcus..| + = ap =f
alee
eel
* Incubation in water bath at 38°C. for one hour. Readings made after
standing overnight in refrigerator; —, no agglutination; +, = agglutination.
The mixtures and the controls were incubated in a water-bath
for one hour at 38°C. and the results were read after the mixture
had stood in a refrigerator over night.
The results observed with 12 sera tested with the erythrocytes
of 9 different persons, are given in table 1; the results with 28
sera tested with the corpuscles of each of 12 different persons
are shown in table 2.
As indicated in these tables about 50 per cent of horse sera
showed the presence of hemagglutinins with this technic in a
HEMOLYSINS AND HEMAGGLUTININS IN HORSE SERA re
final dilution of about 1:11, for the erythrocytes of certain
persons; no single serum contained hemagglutinins for the
erythrocytes of all persons tested, indicating the presence of
TABLE 2
Agglutinins in unheated horse sera for human erythrocytes
HUMAN ERYTHROCYTES 1 Cc. OF k PER CENT SUSPENSION
UNHEATED HORSE SERUM
0.1cc.
Antistreptococcus....... Ssh tac |i eat itt Ao pol (i Ne a | a
Antistreptococcus....... —fo—-fo}omly mle ol om l clot ol odo
Antistreptococcus....... —}o}oroy} oly ol omy omy ml aml cml odo
Antipneumococcus.......-—|—|—]|—]|—]/—}]—/—/+]—-]-]-
Antistreptococcus....... i AP (| ar i (ce dm) Ne ieee eee (ee
Antipneumococcus......;— | —}|—-|]—-|—-/|-|-—|]-
Antistreptococcus....... lial eral (iin Cire (eat (ical | Se:
Antistreptococcus....... S| oe SE | a)
Antipneumococcus...... = lsat a ae el We (Beret 9 (Se Se
Antipneumococcus...... =i | et il
Antipneumococcus...... mar] Samy he | ES
Antipneumococcus......| — | — | —|— |] —
Antipneumococcus......] — | — | — | —
Anti-influenzal.......... |S Pak Wi | |
Anti-influenzal..........)} —|—]|—-]—|—-
Anti-influenzal.......... SS ia
Anti-influenzal..........) -—|—-|—|—|+
Anti-influenzal.......... eo SN | Ee ||
ae
|
|
he pea Jt eas pad aa
eel
ler |
seen leet ie |
[=e |
lal
Ieal
|]
|
ied |
(eal
Antimeningococcus...... —{—}—}|—
Antimeningococcus...... —-{|-|-|- eae Shs) lade] Se Siecle
Antistreptococcus....... oa) a cecal ian Mitel (|i | a ei facie (lore
Antistreptococcus....... Se RY RS at ad i ft |
Antistreptococcus...... oh erty eG | Ul lai Pies | |) 29) | SS ia hael
Antistreptococcus....... mac |= Pd Wim (tee |e oh al Neen ileal A hers
*—, No agglutination; +, agglutination.
group hemagglutinins in varying amounts for the different
groups of human erythrocytes.
Microscopical tests, however, in which lower dilutions of sera
were employed, showed the presence of these hemagglutinins in
a larger percentage of sera.
78 JOHN A. KOLMER AND MOTOMATSU MATSUMOTO
Microscopical tests.
mixing on cover-glasses one loopful of serum and one loopful of
a 1 per cent suspension of washed cells, the results being read
ten to fifteen minutes later; the final dilutions with this technic
TABLE 3
Microscopical tests were conducted by
Microscopical agglutination tests with horse sera and human erythrocytes
HORSE SERUM
NO TIAR I Wes e Sac ss
1 (ry git 0 ee ae
Antipneumococcus......
Antistreptococcus..
Antipneumococcus.
Antistreptococcus..
Antistreptococcus.
Antipneumococcus.
cee eee
Antipneumococcus......
Antipneumococcus......
Antipneumococcus.
Antipneumococcus.
Anti-influenzal..........
Anti-influenzal..........
Anti-influenzal..........
Anti-influenzal..........
Antimeningococcus
Antimeningococcus
Antitetamus..+.5.-2 2. 4c.
Antitetanus........
Antistreptococcus..
Antistreptococcus..
Antistreptococcus.......
Antistreptococcus..
see ee
[+H 1h 1 +Ett+HR I +H+i
+H +1
HHH ++ Ltt t++t+Htttsei ti
++hi +t +tt+++h ++tt+et++sest
+I bi ++ ++H
I++ b+++h i+
b+) ++H+4+
+H +L tt+++t+++itsti
+ttt++tt++tteet¢+e yr t+ t+stig¢
Hos [pefocatiy
HUMAN ERYTHROCYTES
|
DHE LEE I++ I
l+++H RH
I+l! Holttocoo
+++H++H+I1++Hh#+4+4+4+4+4+4+444+4+4+
+++ +++ HEHE HH I
coolottococl |
l+++++t
I++ TRA +H+4
tH |
*—, No agglutination; +, partial agglutination; +, strong agglutination.
were 1:2 and showed the presence of small amounts of hem-
agglutinins in practically all horse sera examined.
The results of one series of such tests with 24 sera and the
erythrocytes of each of 12 persons, are shown in table 3; the
results observed with 3 sera and the corpuscles of 24 persons are
given in table 4.
HEMOLYSINS AND HEMAGGLUTININS IN HORSE SERA 79
As shown in these tables all normal and immune horse sera
contain hemagglutinins for human erythrocytes when tested
microscopically with a technic similar to that proposed for the
detection of human isohemagglutinins; when tested with the
corpuscles of a number of different persons, however, only a
TABLE 4
Microscopical agglutination test with horse sera and human erythrocytes
SERUM
1]2/3|4{5]|6| 7] 8 | 9 | 10] 11] 19] 13 | 14| 15 | 16| 17 18 | 19 20] 21) 2223 | 24
Normal I...J—J | |4 |-— |-— |— J+ |+ J4/- |-|43)+ |- |-|- Fe =:
Normal II. .|—|+3|+3|+3/+2/+2}+2/+3|-+3|+2/+|+2/ |42|—|42)494/4/4!-l+3 49
Normal III..|-+/+2|-+3|-+2|+ |+3/+3/+ |+ |4/— |—|+ |+2|-+2!-|+2|- be +
—, Negative no agglutination; +, very doubtful agglutination; +, very weak
agglutination; +2, marked positive agglutination; +3, very strong agglutination.
TABLE 5
Quantitative agglutination tests with preserved horse sera and human erythrocytes
CORPUSCLES CORPUSCLES CORPUSCLES CORPUSCLES
I II Ilr IV
SERA
Macro-| Micro- |Macro- | Micro-| Macro-} Micro-| Macro-| Micro-
scopic | scopic | scopic | scopic | scopic | scopic | scopic | scopic
Antistreptococcus....... 1:4
Pea) VSS LSS RSZN eS ees 16+ bs2
Antipneumococcus...... ESS) hs4et hs 16 1 158) Tete BsS) 4s 16) 18
Antipneumococcus...... Te Zieipeis4e | Us ab 2 a e225 es? tts 2
PANT GILCCANUS.....< . 640.0 1:2 | None} None} None| 1:4 | None] 1:2 | None
Anti-inivenzal..........| 1216] 1:8 | 1:16) 1:8 |] 1:16) 128 | 1:82) 1:8
Anti-influenzal.......... ESS ay LS | LAG | VUSS BSR GS ease Ts Solis
ING TaD Oe Oe HAIGH e416 1855), 0 0 0 0
iT CTE ED] | i ee SZ E38) | 1264 | £21640 0 0 0
ING ETAT] | A opener eee 1:16 | 47 |) 1264 | Ls32e 0 0 0
few sera show the presence of hemagglutinins for the cells of all
persons, so that a single serum may not agglutinate the corpuscle
of a certain individual.
The amount of hemagglutinins in preserved horse sera for
human erythrocytes is relatively small; when tested macro-
scopically in final dilution of 1:11 only about 50 per cent of sera
contain hemagglutinins for the corpuscles of some persons
80 JOHN A. KOLMER AND MOTOMATSU MATSUMOTO
(tables 1 and 2); in higher dilutions fewer and fewer sera are
found to contain hemagglutinins and only about 20 per cent
were found to yield positive macroscopical reactions in dilutions
higher than 1:20 and none with dilutions higher than 1:64. Second Test.
106 JOSEPH E. SANDS
SUMMARY
1. Agglutinogens of broth cultures vs. saline suspensions. Of
primary importance are the results in relation to agglutinogens
prepared by cultivation of the microérganism in broth as com-
pared with those prepared by cultivating on plain solid media
and suspending the bacteria in saline solution: a general review
of the results observed in this study with a single strain of
typhoid bacilli, indicate that suspensions in saline solution so
prepared as to break up clumps and diluted to proper density,
are superior to broth cultures.
2. Saline solutions vs. distilled water. Asis well known, sodium
chloride exerts an important réle in the physicochemical phe-
nomenon of agglutination; of agglutinogens prepared by sus-
pending typhoid bacilli in strengths of sodium chloride in sterile
distilled water varying from 0.85 to 5 per cent (nos. 17, 18, 19
and 20), best results were observed with the 0.85 and 1 per cent
solutions (nos. 17 and 18); agglutinogens prepared with 2 per
cent solutions of sodium chloride were less susceptible to agglu-
tination and 5 per cent solutions were decidedly less susceptible.
Agglutinogens prepared with distilled water alone (no. 21) were
least susceptible to agglutination and yielded some of the lowest
titers.
3. Influence of heat. Of primary importance is the effect of
heating an agglutinogen upon its susceptibility to agglutina-
tion; in the majority of laboratories cultures or suspensions are
usually heated at 56° to 60°C. for one-half to two hours. In
order to be able to observe the most marked influence of heat,
if any, antigens were heated in a water bath at 60°C. for two
hours; a general survey of the results of this study indicates
that heated antigens are somewhat more susceptible to agglu-
tination than unheated antigens.
4. Influence of chemical germicides and preservatives. In this
investigation phenol, tricresol, formalin, mercurophen and gly-
cerin were employed for chemical sterilization and preservation
(antiseptic activity); with the exception of formalin in 0.1 to 5
per cent; agglutinogens prepared without preservatives were
i
PREPARATION OF TYPHOID AGGLUTINOGENS 107
somewhat superior to those containing phenol, tricresol, mer-
curophen and glycerin.
5. Kind of chemical germicide and preservative. Of the chem-
icals employed in the preparation of agglutinogens of B. typho-
sus, best results were observed with formalin; in fact the addition
of 1 to 2 per cent formalin? to suspensions in isotonic saline solu-
tion yielded the best agglutinogens of the series included in this
study (nos. 9 and 10). Agglutinogens were prepared with 0.1,
0.5, 1, 2 and 5 per cent neutral formalin; comparative tests have
usually shown that those containing 1 and 2 per cent were least
likely to show spontaneous agglutination, were most susceptible
to specific serum agglutination and were never contaminated.
Agglutinogens prepared with phenol in 0.5 per cent and
tricresol in 0.5 per cent (nos. 4 and 5) yielded similar results,
but they were somewhat inferior to preservative free agglu-
tinogens and decidedly inferior to those containing 0.1 to 2 per
cent formalin. Agglutinogens prepared with 1 per cent phenol
and tricresol (nos. 12 and 14) were decidedly inferior to those
containing 0.5 per cent of these substances and an antigen
containing 5 per cent phenol (no. 13) was almost insusceptible
to agglutination and proved most unsatisfactory of all.
Mercurophen was employed because of its high germicidal
activity and freedom of precipitating and coagulating influence
upon proteins including bacterial proteins; agglutinogens pre-
pared with 1: 1000 mercurophen in physiological saline solution
(no. 15) and 1: 2000 (no. 6) yielded results similar to those con-
taining 0.5 per cent phenol and tricresol. An agglutinogen
containing 1: 5000 mercurophen (no. 16) was generally satisfac-
tory and from the standpoints of freedom from spontaneous
agglutination, susceptibility to specific agglutination and freedom
from contamination ranked next to plain and formalized agglu-
tinogens.
The addition of 10 cc. of the best grade neutral glycerin to
each 100 cc. of heated agglutinogen containing 0.5 per cent
phenol (no. 3) reduced susceptibility to specific agglutination
? The formalin used contained 39.2 per cent formaldehyde gas; the 1 per cent
solution, therefore, contained 0.39 per cent of formaldehyde.
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 2
108 JOSEPH E. SANDS
and proved inferior to the same agglutinogen prepared without
the addition of glycerin (no. 1).
6. Spontaneous agglutination. Of agglutinogens 1 to 16 pre-
served over a period of four to eight weeks in a refrigerator,
no. 15 containing 1: 1000 mercurophen, showed most tendency
to spontaneous agglutination and nos. 7, 8, 9 and 10 containing
0.1 to 2 per cent formalin, least spontaneous agglutination; of
the freshly prepared antigens, nos. 19 and 20, containing 2 and 5
per cent sodium chloride frequently showed spontaneous agglu-
tination. Macroscopical tests for spontaneous agglutination
were conducted with each antigen whenever employed by di-
luting 1 cc. with 1 cc. of 0.85 per cent saline solution and incu-
bating at 55°C. for twenty-four hours; at the same time micro-
scopical tests were made and occasionally antigens showed small
clump of bacilli microscopically, which appeared perfectly hom-
ogenous and satisfactory to the closest scrunity with the naked
eye.
CONCLUSIONS
1. A comparative study of agglutinogens prepared from a
single strain of typhoid bacilli which had been used for agglu-
tination tests for several years, was made by comparing their
susceptibility to specific agglutination by rabbit immune sera,
tendency to spontaneous agglutination, keeping qualities and
susceptibility to contamination.
2. The density of the agglutinogen was found to have an
important influence, regardless of the method of preparation;
thick suspensions obscured results and reactions while very thin
suspensions were difficult to read with the naked eye. The
density of a particular agglutinogen should be adjusted accord-
ing to the diameter of the test tubes employed and total volume
of fluid.
3. Suspensions in saline solution of microérganisms washed
from solid media, were generally superior to broth cultures.
4. The best saline solutions for the preparation of agglu-
tinogens were found to be those containing 0.85 to 1 per cent
chemically pure sodium chlorid in distilled water.
PREPARATION OF TYPHOID AGGLUTINOGENS 109
5. Distilled water alone was found unsatisfactory for the
preparation of typhoid agglutinogen.
6. Heating an agglutinogen at 60°C. for two hours generally
increased its susceptibility to specific agglutinins.
7. Agglutinogens prepared without preservatives with the ex-
ception of those preserved with formalin were generally superior
to those containing phenol, tricresol, mercurophen and glycerin.
8. The best agglutinogens were found to be those containing 1
to 2 per cent formalin.
9. The addition of more than 0.5 per cent phenol and tricresol
to an agglutinogen reduced its susceptibility to specific serum
agglutinins; the addition of glycerin also reduced the suscepti-
bility to specific agglutination.
10. An agglutinogen of typhoid bacilli is best prepared by
cultivating on solid media for forty-eight hours, removing the
growths with 0.85 to 1 per cent chemically pure sodium chlorid
in distilled water, shaking with beads until a perfectly homo-
genous emulsion is secured, diluting with saline solution to
proper density (about 2,000,000,000 per cubic centimeter), and
adding neutral formalin to 1 per cent.
The writer wishes to express his sincere gratitude to Prof.
John A. Kolmer for his very kind assistance throughout the
entire course of this study; he is also indebted to Dr. M. Matsu-
mato and Dr. Yosiho Saeki for their aid in the preparation of
the agglutinogens.
‘
thy
9 .
i
S i
'
%
‘ %, 1
\ 4
= iat
n
z, 4
‘ t + {
7 i pe
4
“
‘ —
» “
.
j
fi
A STUDY OF DIFFERENT METHODS FOR THE
PREPARATION OF B. TYPHOSUS ANTIGEN
MOTOMATSU MATSUMOTO
From the McManes Laboratory of Experimental Pathology of the University of
Pennsylvania
Received for publication February 9, 1920
One of the reasons generally assigned for the unsatisfactory
status of complement-fixation tests in the diagnosis of bacterial
infections, is the difficulty of preparing efficient and stable anti-
gens; among the diseases of bacterial origin complement-fixation
is probably most widely employed in the diagnosis of glanders,
tuberculosis and gonococcus infections, but in the last men-
tioned the test is well known as lacking in the sufficient delicacy
and it probably can be rendered more sensitive by further im-
provement of the antigen, and the same may be true of the
tuberculosis complement-fixation test.
Bacterial antigens for complement-fixation tests may be di-
vided into four main groups, namely (a) those composed of whole
bacteria and their soluble products in the fluid medium in which
they have been cultivated; (b) those in which the bacterial cells
alone are utilized suspended in sterile salt solution; (¢c) those in
which the bacterial cells are disrupted but not filtered and (d)
those in which the cells are disrupted and the insoluble portions
removed by filtration, the filtrate being employed as antigen.
The antigen commonly employed for the gonococcus comple-
ment-fixation test may be classified under the last mentioned
in which the soluble intracellular substances are utilized as anti-
gen; in a study of these gonococcus antigens Kolmer and Brown
(1) found that whole suspensions of gonococci in saline solution
classified under (b) above, proved superior in antigenic sensitive-
ness to filtrates, and similar results were observed in complement
fixation tests in typhoid fever (2), diphtheria (8) and canine dis-
temper (4).
Hit
1t3 MOTOMATSU MATSUMOTO
PURPOSES OF INVESTIGATION
In view of the practical importance of increasing the sensitive-
ness and delicacy of complement fixation in bacterial infections,
Professor Kolmer suggested a further systematic and compara-
tive study of prevailmg methods for the preparation of bacteria
antigens taking a single pure culture of B. typhosus as the test
microérganism and preparing antigens after the four main varie-
ties described above; a secondary object was the study of com-
plement fixation in typhoid fever and after active immunization
of persons with typhoid vaccine, as an additional means for com-
paring the antigenic sensitiveness of the various antigens.
Each antigen has been studied for its anticomplementary,
hemolytic and antigenic values; the antigenic titrations were
conducted with the sera of rabbits immunized with the same
strain as used in the preparation of the antigens and also with
the sera of persons containing typhoid antibodies. The results
are summarized in this communication.
PREPARATION OF ANTIGENS
Antigen 1. This antigen was composed of living bacilli sus-
pended in physiological saline solution and freshly prepared as
required by removing twenty-four hour cultures on plain neutral
agar with saline and shaking with sterile glass beads until a
homogenous suspension was secured.
Antigen 2. This antigen was a forty-eight hour growth in
plain beef extract broth neutral to phenolphthaleim, shaken me-
chanically for an hour to secure a homogenous suspension fol-
lowed by heating in a water-bath at 60°C. for one one hour and
preservation in a refrigerator with 0.5 per cent phenol.
Antigen 3. The same as antigen II except that a fourteen-
day broth culture was employed.
The remaining six antigens were prepared from mass cultures
of the strain of B. typhosus removed from a large series of agar
cultures in Blake bottles, due care being utilized against remov-
ing bits of culture medium. In order to make sure that agar
PREPARATION OF B. TYPHOSUS ANTIGEN 113
and other bacteria were not included, the bacterial mass was
briefly centrifuged and cultured before use.
Antigen 4. Five hundred cubic centimeters of a heavy sus-
pension of bacilli in sterile distilled water was heated in a water-
bath at 56°C. for one hour, then at 80°C. for an hour. The
heated suspension was shaken mechanically for twenty-four
hours after which treatment it was centrifuged and the super-
natant fluid was passed through sterile neutral porcelain filters.
The filtrate was then heated at 56°C. for one hour on three suc-
cessive days and preserved in a refrigerator with 0.5 per cent.
phenol.
Antigen 5. A saline suspension of bacilli was thoroughly cen-
trifuged and the sediment was dried over calcium chlorid; each
0.05 gram of dried bacterial mass was ground into a very fine
powder and gradually suspended in 25 ce. of physiological saline
solution. This emulsion was shaken mechanically for twenty-
four hours and passed through a porcelain filter and the filtrate
was preserved in a refrigerator for antigen.
Antigen 6. Five hundred cubic centimeters of a heavy saline
suspension of bacilli was precipitated with an equal quantity of
absolute ethyl alcohol and thoroughly centrifuged; the sediment
was dried over calcium chlorid, ground into a fine powder, weighed
and suspended in sufficient saline solution to make a 2 per cent.
emulsion. The resulting product was quite thick and required
further dilution with saline prior to use.
Antigen 7. This antigen was prepared after the method de-
scribed by Hitchens and Hansen (5) for the preparation of men-
ingococcus antigen; briefly the technic consisted of precipitation
of 500 cc. of a heavy suspension of bacilli in distilled water with
an equal amount of 95 per cent. ethyl alcohol and thoroughly
centrifuged at once for the sediment; the sediment was resus-
pended in alcohol and again centrifuged, this process being re-
peated several times with alcohol and finally several times with
ethyl ether. The final sediment was freed of ether, dried over
calcium chlorid and ground to a very fine powder and 0.02 gram
was suspended in 20 ce. of sterile saline solution for antigen as
required.
114 MOTOMATSU MATSUMOTO
Antigen 8. This antigen was prepared after the method de-
scribed by Small (6); briefly the technic consisted in thoroughly
centrifuging a heavy saline suspension of bacilli and drying the
bacterial sediment at 56°C.; 0.5 gram of this powder was mois-
tened with chloroform and thoroughly ground with the addition
of a small amount of ether from time to time until a very fine dry
powder was obtained. This powder was now suspended in a
mixture of equal parts of chloroform and ether and shaken me-
chanically for six hours followed by several washings of the sedi-
ment with ether and drying of the ether moist residue at 56°C. ;
when used about 0.5 gram of the powder was suspended in 25 ce.
of saline solution and further diluted with saline solution.
Antigen 9. This antigen was prepared after the method de-
scribed by Miss Wilson (7) for the preparation of antigen of
tubercle bacilli for the complement fixation test; 2000 cc. of a
five day culture of the strain of B. typhosus in plain neutral broth
was heated in an Arnold sterilizer for one hour and thoroughly
centrifuged for the bacilli; the sediment was then treated a num-
ber of times with ten volumes of absolute ethyl alcohol and fin-
ally with ether, the sediment being secured each time by centri-
fuging. After the final treatment with ether the bacterial sedi-
ment was dried at room temperature, ground into a fine powder,
weighed as required and prepared in a 0.5 per cent suspension
in sterile saline solution.
TECHNIC
Titrations for anticomplementary power. Each antigen was ti-
trated at intervals and just prior to complement-fixation tests
with immune sera, for its anticomplementary or antilytic value,
the smallest amount producing the slightest inhibition of hemoly-
sis being taken as the anticomplementay unit. All antigens were
titrated at the same time and with the same hemolytic system
in order to render the results strictly comparative.
In conducting these titrations the antigens were used undiluted
or diluted with saline solution as required and placed in a series
of 12 test tubes in amounts ranging from 0.02 to 2 cc.; complements
PREPARATION OF B. TYPHOSUS ANTIGEN 115
were furnished by the mixed sera of guinea-pigs in a dose of 0.5
ec. of 1:20 dilution. After salt solution had been added to each
tube carrying antigen and complement to bring the total volume
to 2.5 ec., incubation was conducted in a water-bath at 38°C.
for one hour and this was followed by the addition of two units
of antisheep hemolysin and 0.5 ce. of a 2.5 per cent suspension
of sheep corpuscles; the tubes were then reincubated for an hour
and placed in a refrigerator over night, the results being read
the next morning. The usual hemolytic, complement and cor-
puscle controls were included.
Hemolytic titrations. By including relatively large doses of
each antigen in the anticomplementary titrations the direct hem-
olytic dose of each in the presence of complement was generally
obtained and served as a means for comparing the hemolytic
activity of the various preparations.
Antigenic titrations. ‘These were conducted with sera of im-
munized rabbits and of persons having had typhoid fever or ac-
tive immunization with typhoid vaccine. Rabbits were im-
munized with the same culture of B. typhosus employed in the
preparation of the antigens, increasing doses of heated killed and
finally living bacilli being injected intravenously and the sera
obtained from these animals were employed in the complement
fixation tests when the agglutinins had reached a titer somewhat
comparable to their concentration in the sera of persons during
typhoid fever or after a course of injections of typhoid vaccine.
In order to avoid the non-specific fixation of complement by
rabbit sera described by Kolmer and his associates (8), each
serum was used unheated or after heating at 62°C. in a water
bath for thirty minutes and in small amounts ranging from 0.0001
to 0.02 cc. Human sera were used unheated and after heating
at 56°C. for thirty minutes in amounts ranging from 0.01 to 0.1
ee.
In conducting the antigen titrations with immune sera to
bring out the differences in antigenic sensitiveness of the various
preparations the following method was employed; each antigen
was used in a constant dose equal to one-third its anticomple-
mentary unit in a series of twelve test tubes and unheated rabbit
116 MOTOMATSU MATSUMOTO
sera added in amounts ranging from 0.0001 cc. to 0.02 cc.; a
serum control on each serum was included in which 0.02 ee.
serum alone was used. A control on each antigen was included
as likewise hemolytic, complement and corpuscle controls. Com-
plement was furnished by the mixed sera of guinea-pigs in a con-
stant dose of 0.5 cc. of 1:20 dilutions and saline solution was
added to bring the total volume in each tube to 1 cc. Primary
incubation was conducted at 38°C. in a water-bath for one hour
and this was followed by the addition of two units of hemolysin
and 0.5 cc. of 2.5 per cent sheep cells; after reincubation for an
hour the results were read after the tubes had been placed in a
refrigerator overnight.
Agglutination tests. The agglutinin content of each human and
rabbit serum for the strain of B. typhosus employed in this work
was determined in a macroscopic test with the employment of
living suspensions and an incubation at 38°C. for one hour; the
results were read by the naked eye after the tubes had been
placed in a refrigerator over night.
RESULTS
a. Anticomplementary and keeping qualities of the different
antigens. The results of anticomplementary titrations with
freshly prepared antigens are shown in table 1; these titrations
were repeated at subsequent periods and the results are summar-
ized in table 2, the smallest amount of each antigen producing
slight-inhibition of hemolysis being registered as the anticomple-
mentary unit.
None of the antigens were markedly anticomplementary but
a strict comparison could not be made in-as-much as this would
have required the preparation of antigens according to numerical
numbers of bacilli entering into a uniform given volume of each
preparation.
All of the antigens were preserved in a refrigerator near the
freezing point and the series of titrations of anticomplementary
power over a period of six weeks following their preparation
PREPARATION OF B. TYPHOSUS ANTIGEN 117
TABLE 1
The first anticomplementary titration of antigens
ANTIGENS
EY a ree
ANTIGEN No.1 No. 2 No.3 No.4 No.5 No.6 No.7 No.8 N
: : K - , : : No.9
Undilu- Cashes Undilu- ceo hae 1:10 1:10 1:10 1:50 L:
* C.H., Complete hemolysis; M.H., marked hemolysis, the anticomplementary
unit; §.H., slight hemolysis; N.H., no hemolysis.
TABLE 2
Summary showing anticomplementary units of each antigen as determined at varying
intervals
*FIRST SECOND THIRD FOURTH
TITRATION TITRATION TITRATION TITRATION
CTU a a 0.2t 0.6 0.4 0.3
LOU 7 a 0.6 0.6 0.6 0.5
RATE AE STAN Ue Ios Wises eee. oars 0.15 0.2 0.2 0.3
SRNR ERE NTN S20 0)ei 0d) 52) 2 aie\s eras’ « 0.2 0.2 0.4 0.2
PREDIZEURO Neier: eras sin 6 ssiefas 0.15 0.3
JCC) De 0.04 0.04 0.06 0.05
2 Cu Gr 0.15 0.4 0.1 0.06
PRELIM Stas raha le a cies 0) v bye\s 0.03 0.03 0.06 0.07
Loy ee 0.01 0.04 0.02 0.01
* First titration conducted with freshly prepared antigens; second titration
conducted ten days later; third titration conducted one week after the second and
fourth titration about six weeks after the third.
t+ Amount in cubic centimeters of undiluted antigen producing slight inhi-
bition of hemolysis.
118 MOTOMATSU MATSUMOTO
TABLE 3
Comparative antigenic sensitiveness of antigens with active serum of rabbit 1
(agglutination titer 1: 100)
DOSE | ANTIGENS USED IN ONE-THIRD ANTICOMPLEMENTARY UNITS
OF rn TT ee
BS eNon! No. 2 No.3 No. 4 No. 5 No. 6 No.7 No. 8 No.9
0.0001; — - _ - - - — Se -
0.0002 -- _ _ _ _ -- — -- —
0.0004 — = =FSF = = = = = =
0.0006; + = ates = = = = = =
0.0008) + = AFSF aa = = = ae ae
0.0015) -E=- aR Steetaats = = + ap Space | arceae
002s -i-atecialGataaieats | ak aie ta = = 2PaF Sele |paraeas| Sea PF
OORie-aleatal aivaieste at | iat ate ke 7 Spa eps eagac mrersrerparica aesrseo
0.006 |++++]++++/t+++4) + [4++4)++4+4]44+4+4/+4+4+4]44+4++
0.008 |++++/++++|t+++4+) +++ [t++4+4]+4+4+4)4++44]4+4+4]44+4+4+
0.01 (++++]++++)t+4++4+) 444 [44+44)4+4+4+4/44+44]4+4+44]44+4+
0.02 ++++|4+++/4++4/t++4]t4++4]t+4+4]4++4+4]4+4+4+4]44+4+4+
— Negative; + doubtfully positive; + very weakly positive; ++ weakly posi-
tive; +++ , moderately positive; ++-+-+ strongly positive.
TABLE 4
Comparative antigenic sensitiveness of antigens with active serum of rabbit 2
(agglutination titer 1: 80)
DOSE ANTIGEN USED IN ONE-THIRD ANTICOMPLEMENTARY UNITS
0.0002) — |; = ae = Bs = - =
0.0004, — = = = = = = = =
0.0006} — _ + _ _ = - _ —
0.0008} + - - - - - - — -
0.001} + + | ++ | - ~ - ~ ~
0.002; +++] ++ | +++] - - = ~ + | +++
0.004 |+++4+] +++ |++4+4] - - + | tt | +44 [+444
0.006 |++++\+++4+/t++4+4+) + | +++] +++ | +44 [+4+4+4)+4+44+
0.008 |++++)++++/++4+4) + |+++4)t+++4]4+4+4+4+/44+44]+4+4+4+
0.01 |++++)++++/t4++4+) +++ |+++4/+4+4+4]4+4+4+4/44+44]4+4+4+4+
0.02 |++++]+t++i/++t++]t+++/4+t+[t4++4]+t+4i4+t4}t+4++
— negative; + doubtfully positive; + very weakly positive; ++ weakly positive;
+++ moderately positive; ++-+-+ strongly positive.
PREPARATION OF B. TYPHOSUS ANTIGEN 119
TABLE 5
Comparative antigenic sensitiveness of antigens with active serum of rabbit 3
(agglutination titer 1: 60)
ee ANTIGEN USED IN ONE-THIRD ANTICOMPLEMENTARY UNITS
SERUM
No.1 No.2 No. 3 No. 4 No.6 No.7 No. 8 No.9
cc.
0.0001 = = = = = = = =
0.0002 | — = = = = = = =
0.0008 | — = = = = - =
0.000 | — = = = = = - =
0.0008 | — — = =: = - = =
0.001 = = = = = = = =
0.002 + ~ — - + _ - -
0.004 | ++ | + [4444+] - + - + -
0.006 |+++] + [44+4++) -— | +++] - | +44) 4+
0.008 | +++] ++ [44+4++) - | +++] 4 [+++] 44+
0.01 |++++)4+4+4+ 4444) - [+444] + | 444+) 444+
0.02 |+4+4/44+4+4+/44+4+4+) + [44+4+4/44+44)44+44| +44
— negative; + doubtfully positive; + very weakly positive; ++ weakly posi-
tive; +++ moderately positive; ++-+-+ strongly positive.
TABLE 6
Comparative antigenic sensitiveness of antigens with heated serum of rabbit 3
(agglutination titer 1: 60)
ANTIGENS USED IN ONE-THIRD ANTICOMPLEMENTARY UNITS
SERUM
No. 1 No. 2 No.3 No. 4 No. 6 No.7 No.8 No.9
cc.
0.0001 - = = = _ = = =
0.0002 - _ _ — - - _ _
0.0004 = = - — - - - -
0.0006 _ = _ - — _ _ _
0.0008 - _ _ — - - _ -
0.001 — — + - - _ - -
0.002 + - ~ - - - - -
0.004 | ++ | — |4+4++)] - - = + =
0.006 |+++4+/++++/++++) 0 | +++] 0 0 0
0.008 |++++/t++++/+4+4+4+) - | 44+] - | ++ | +
0.01 |+++4+]44++4l44+44) -— [+444] 4 [444+] +
0.02 |++++l+++4i++4+4) t+ [4444] +44 [4444/4444
ae negative; + doubtfully positive; + very weakly positive; ++ weakly posi-
tive; +++, moderately positive; ++++ strongly positive; 0 lost by accident.
120
MOTOMATSU MATSUMOTO
TABLE 7
Comparative antigenic sensitiveness of antigens with active serum of rabbit 4
(agglutination titer 1: 240)
DOSE OF
SERUM
~-
ae
aa
aaa
tae
++4++
+4+4++
oe
ANTIGENS USED IN ONE-THIRD ANTICOMPLEMENTARY UNITS
aa
++++
++++
wae
mee
wae
No. 3 No. 4 No. 6 No.7 No. 8 No.9
a — — — — _
+++] - ~ ~ - -
+444) - ~ - ~ -
++4+4+) -— |++4+4+) 4+ [+444] +
+444) 4 |44+4+4/4+4+44/44+44) +44
++4++) + [44+4+4)44+4+4/+4+4+4]4+4+4+4+
+++4+|) $4 |+4+4+4]4+4+4+/44+4+4+/44+44+
+4+4++) 444+ |+4+4+4]44+44)44+4+4+/44+4+4+
— negative; + doubtfully positive; + very weakly positive; ++ weakly posi-
tive; +++ moderately positive; ++-+-+ strongly positive.
TABLE 8
Comparative antigenic sensitiveness of antigens with heated serum of rabbit 4
(agglutination titer 1: 240)
DOSE OF
SERUM
ANTIGENS USED IN ONE-THIRD ANTICOMPLEMENTARY UNITS
No.1 No.2 No.3 No. 4 No.6 | No.7 No.8 No.9
= — — — — _ _ —
- - + _ = - - ~
++] - | ++] - - - ~ -
+++] + |444+4) - | + - - ~
++44/44+4+4]44+4+4] -— [44+4+4+) 4+ | 444+] -
+4+44l[4t44i4444] — |44+4+4]44+4+4]4+4+4+4) +44
+444it+t+i +444) -— [tt+t]t++t4+]4+4+4)/+++4+
+4+44)t+4+4++/4+4++
+++4)t++4)44+4++
+4+44|+++4|4+4+4|4+4++
++t+)/4++4+|44+4+4+|4+4++
+ |
— negative; + doubtfully positive; + very weakly positive; ++ weakly
positive; +++ moderately positive; +++-+ strongly positive.
Lag a ae
PREPARATION OF B. TYPHOSUS ANTIGEN
TABLE 9
Summary showing smallest amount of each un
121
heated and heated rabbit immune
serum giving +++-+ with one-third the anticomplementary unit of each antigen
UNHEATED RABBIT SERA
ANTIGENS
Rabbit 1 Rabbit 2 Rabbit 3
itt 0.002 0.004 0.01
1s ae 0.004 0.006 0.02
1.03 Uae nae 0.002 0.005 0.004
1. ole 0.02 0.02 0
No. 5 0.006 0.008 0
Wor Gren... > - 0.004 0.008 0.01
Ls UY i ae aS 0.004 0.008 0.02
Lo 0.002 0.006 0.02
ai s'<'ss <3 0.004 0.004 0
HEATED RABBIT SERA
Rabbit 4 Rabbit 3 Rabbit 4
0.004 0.006 0.004
0.004 0.006 0.004
0.002 0.006 0.002
0 0 0
0 0 0
0.004 0.01 0.004
0.006 0 0.006
0.004 0.02 0.006
0.008 0.02 0.008
Agglutination titer of rabbit 1 was 1: 100, rabbit 2,1: 80; rabbit 3, 1: 60; rabbit
4, 1: 240.
TABLE 10
Results with the unheated serum of a person in
the fourth week of typhoid fever;
agglutination titer 1: 40
RESULTS WITH DIFFERENT ANTIGENS
SERUM
No.1 No. 2 No. 3 No.4] No
ce. ee |
0.01 |++++) + |++4+4) - | -
0.02 j+++4+) + |4+4+4+] - | -
0.04 |++++] +4 |44+44/ - | -
0.06 |++++\t++lt++4+| - | -
0.08 | +++ |++4+]+4++4) - | -
01 | +++ |+4++/4+++4+) - | -
TABLE lil
— | + [44+4+4+/4++++
— | + |+++4+/++++
— | ++ |44+4+4+/4+4++
— | ++ 444+4+/4+4++
— |44++]) +++ | +++
— | + | +++) +++
Results with the unheated serum of a person in the sixth week of typhoid fever;
agglutination titer 1
ee
0.01 |++++/+++4+)++++)t+4+++4+4++4+
0.02 |++++/t+++4+\++++/4++4+4/+4+4++4+
0.04 |++++/+++4+)+++4|+4+4+4/+4+4++4+
0.06 |++++/+++4+\++++/t+4+++/4+4+4+4+
0.08 |++++/t++++/+4+4+4+44+4+4+ 4444
O.1 [t+t+i++t+i++t+i+t+t+it+tt+
: 820
RESULTS WITH DIFFERENT ANTIGENS
+++4/+4+++
+++4+)4++++
+4+44]+4++4+)4+4+4+/4+4+4+4+
+4+4++i¢t+4++/+t+4++/t+++
+Ht+l[ttt+lttt+itt++
+++) t+++4+++4+ 4444+
+++4+/4+4++4+
++++\t+++ .
122 MOTOMATSU MATSUMOTO
TABLE 12
Results with the heated serum of a person on the thirteenth day of typhoid fever;
agglutination titer 1: 40
RESULTS WITH DIFFERENT ANTIGENS
0.01 |+++ | - - };-;-]-] - |-] +
0.02 ++++}/ + | +4) - | - | - | - | = | +4
0.04 0 | ++ |+4+4+4+) - | - | - | - | = |44+4+
0.06 0 |+++4+)+4+4+4]) - | - | + | + | - [4444+
0.08 |++++)+++4]44+4+4+) - | -— |44+4+)44+4+4/44+4)+4+4++
0.1 [+4++4+)t+4++4]4+4+4+4) -— | - [44+4/}44+4+4/+4+4+]}44+4+4+
TABLE 13
Results with the heated serum of a patient convalescent from typhoid fever;
agglutination titer 1: 40
RESULTS WITH DIFFERENT ANTIGENS
DOBE OF
SERUM Ae STP SS | RS Eis so ea SE
No.1 No.2 | No.3 | No.4 | No.5 | No.6 No.7 No.8 No.9
0.01 j++++) - | +] -/}/ -|-] = — |44+4++
0.02 |44+4++) - |/4+4+] - | - | - |] + + |++4++
0.04 |++4++) - |444+) - | - | — | ++ | +44 /44+4+4+
0.066 |++++) + [t+4+4+) - | - | -— | 444+) 444 |44+44
0.08 |++++) + [44+4+) - | -— | - |44+4+4)44+4+4+/44+4++
0.1 j++++)4+4+ |t+4+4+) - | -
++4++)44+4+4|4+4++
TABLE 14
Results with the heated serum of a patient on the seventeenth day of typhoid fever;
agglutination titer 1: 40
DOSE RESULTS WITH DIFFERENT ANTIGENS
OF pan a ne
SEE OM Noel No.2 No.3 | No.4] No.5 | No.6 No.7 No. 8 No. 9
cc.
0.01 |4+4+4+4+/4+4+4+4+/4+4+4+4|] —- = = = = a5
0:02 |--+ +--+ )4+-4+4+4+|4+-+-+4| = = == oF Sera |) oncr ne
0.04 |+4+4++4+/4+4+4+4+/4+4+4++4+] - = ee) ee OEE la ete
0.06 |+-++4++|/4+4+4++4+|/4++++|] — = |+-+-5-+| +--+ | 4 ae
Oe tele se arararadeeesspaeleeeecses acces
0.1 Paste teta tat tt, SPoRarSE| = — j+++4+)44+4+4+/44+4+4+/4+4+4+4+
at .
ee
PREPARATION OF B. TYPHOSUS ANTIGEN 123
TABLE 15
Results with the heated serum of a patient convalescent from typhoid fever;
agglutination titer 1: 40
DOSE ; RESULTS WITH DIFFERENT ANTIGENS
0.01) +++ |+++4+) +4 | =) - | - + OP t+}
0.02 J++++|+4++4]4+4+4+4+) — | — | 44 F444) $44 4444)
0.04 |+4++4/4++4+4+/+4+4+4) — | — 44+4+4)44+4+4)44+44/44+4+4
0.06 FF 4+44/44+44/t4+44) — | — 44 t4iett+l]t444it+4+
0.08 |++++/++++|4+4++) — | + 44+4++4+44)44+4+4/4444)
O.1 J++++i+t++]+t4+4+) — | + |++4+4)44+4+4+/44+44/444+4)
TABLE 16
The results with the heated serum of a person who had typhoid fever one year ano:
agglutination. titer, 1.40.
RESULTS WITH DIFFERENT ANTIGENS
DOSE OF _
SERUM
= egal: ++) +44
0.04 [44+4+4+/4444/t+444] + | 44+ [4444144444444
0.06 J+t+4+4+i+++4[++t4][t+4+4) 4+ [t+4+4+4)t+444]4+4+4+4+
(Le Se eesti al etl olf ral
O.1 |t+t+it+++]t+++t)++4+4+) +44 44+4+4+)44+4+4)44+44+
TABLE 17
Rewulis with the unhedted serum of a person who. had tr vohora ae and hee
- vaccine ey years ago; agglutination titer 1: 10
DOSE = = = = RESULTS WITH DIFFERENT ANTIGENS
0.01 }+4++4+4+) + f+t++4i+4+44) -— | +. f+t4+4l++4+4) ++:
0.02 /++++| + [+444] 0 [+444] + [+4+44]444+4/4+4+44+
0.04 )-++++] ++ |++++)t++++)44+4+4)+4+4|/++4+4]4++4+4/44++4+
0.06 |++++) ++ |++t++/+++4+)4+4+4+4) +4 [4+4+4]44+4+4/+4+4++
0.08 |+++-+) ++ |++++)++4+4+/+4+4+4]. +) +++ |44+4+4/+4+4+H:
O.1 | +++ [t+tit+t+itt+t+tts+) + | +44 4444+ 4444+
THE JOURNAL OF IMMUNOLOGY, VOL. Vv, NO. 2
124 MOTOMATSU MATSUMOTO
TABLE 18
Results with the unheated serum of a person who had vaccine two years ago;
agglutination titer 1: 20
RESULTS WITH DIFFERENT ANTIGENS
DOSE OF
SERUM
No. 1 No. 2 No. 3 No.4 No. 5 No.6 No.7 No. 8 No.9
O01 |t+++) — |+tt++)t++4+)+4+4+4] +++ [4+t+4+4)+4+44)44+44+
0.02 |+t++it++ti+tt+4[Ft++l|t+4+t)++t4]tt+4+i[ +444] 444+
0.04 |+++)t+++/+++4/+4++4]+4+4+4]44+4+4/+4+4+4]4+4+4+4+]44+4+4+
0.06 |Jt+++[++++)+++4/$++4)/44+4+4]44+44]4t4+4[+4+44)44+4-+
0.08 | +++ | +++ |++++/44++4+)/+4+4+4) +44 | 444+ | +44] 444+
OL | ttt | +++ [+t+4)t++4+4i+4+4+4) +44 | +44 | +44 ] 444+
TABLE 19
Results with the unheated serum of a person who had typhoid fever eighteen years
and vaccine six years ago; agglutination titer 1: 40
°
DOSE RESULTS WITH DIFFERENT ANTIGENS
OF
BEEUMI Noel No. 2 No.3 No. 4 No.5 |No.6| No.7 No. 8 No.9
cc.
0.01 )/++++) — - - — | = |44+4+4)4+4+4+4) 444+
0.02/++++) + j+t+) + | ++ | — |44+4+4)44+4+4/+4+4++4+
0.04 )++++)+++4]+++4]4++4+4]44+44) — |+4+4+4)4+4+4+4)4+4+4++
0.06 |++++)+4++4)+++4/++4+4)t++4+4]44]44+4+4/44+44)4+4+44+
0.08 |++++)+++4+)4+++4+)4++4+4+/44+44) — [++4+4]4+4+4+4)4+4+4++4+
O.1 | +++ [+t+t]t+++4]++4+4)4+44] — [44+4+4)4+4+4+4)+4+4+4+
TABLE 20
Results with the unheated serum of a person who had vaccine one year ago;
agglutination titer 1: 40
RESULTS WITH DIFFERENT ANTIGENS
DOSE OF
8ERUM
cc.
0.01 |++++)+++4+)+4+4++/+4+4++/44+44] -— +4+4+4+/44+4+4)4+4+4++
0.02 |++++|t+++4]+++4+/4+4+4/t+4+4+4] +4 |44+4+4/44+44]+4+4++4+
0.04 |++++l++++lt++++i/t++4]4+44)t+4+4)+4+44)44+4/4+4+44+
0.06 |++++/+++t]++++it+++)4+44]44+4+4/+4+4+4)44+44/4+4+44+
0.08 |j++++++++)++++|t+++/4+4+4+) + [44+44/444+4/4+4+44+
OL [+++ti++t+]+t+4+[t+4+4+4+/ 4444) -— 44+4+4/44+4+4/4+4+4+4+
PREPARATION OF B. TYPHOSUS ANTIGEN 125
TABLE 21
Results with the heated serum of a person who had vaccine two years ago;
agglutination titer 1: 20
DOSE OF RESULTS WITH DIFFERENT ANTIGENS
—————— ————— —————
No. 1 No. 2 No. 3 No. 4 No. 6 No.7 No. 8 No. 9
cc. Nine
0.01 = — - + — ms = wk
0.02 = - — + — = aE =
0.04 = — — ++ _ = at it
0.06 - + - ++ _ + ++ ~
ee ce) fT | ee | ee pet
0.1 = + — |4+++] - + |+++] ++
TABLE 22
Results with the heated serum of a person during immunization with typhoid vaccine;
agglutination titer 1: 160
RESULTS WITH DIFFERENT ANTIGENS
DOSE OF
el No.1 No. 2 No. 3 No. 4| No.5 | No.6] No.7 | No.8 No.9
0.01 = 4 = a 2), = ge =
0.02 zsh is = a a Sen ee 8 =
oo | ++] - - |-—-}|-];-]+]- | ++
0.066 |++++)/ + —- | - | - | = J++] 4+ |4+4++
0.08 |t+++) ++ | - | — | — | — | +4] 4+ [4444
01 |+t+++}4+4++}] - | - | -— | - [44+] 4+ 44+
TABLE 23
Results with the heated serum of a person who had vaccine one year ago;
agglutination titer 1: 20
RESULTS WITH DIFFERENT ANTIGENS
DOSE OF
SERUM
cc.
0.01 - - - — ~ - ~ -
0.02 - ~ - ~ - - - -
0.04 - ~ _ - - - - -
0.06 - ~ - - - - - -
0.08 - - - ~ - - - -
0.1 = = = - - _ _ -
426 “ADTTVAMOROMATSD . MARSUMOTO.2.A 7S 44
TABLE. 24
Resets with the heated serum of a person who had vaccine about two years' ago;
“agglutination titer t;.80
Se“ 30tty >. SRESULTSiWIFH: DIFFERENT ANTIGENS ees
DOSE OF 420¢
No. 2 No.6 | Noo?! No.8 |
TABLE, 25
Resulis with the heated serum, of a person, who.had typhoid fever eighteen years ago
and vaccine two years ago; agglutination titer 1: 40
exacts tp) RESULTS WITHDIFFERENT ANTIGENS
DOSE OF
—_SERUM = =
8.07% No 3 +] < No.4¢
cn
0.01 - |_-
0.02 cS (iS
40.04 = a a
~ 0-06. = |-7
+ 0.08. = il 5
+ 0.1. =) =a ae
LABIA 26
Results, with. the heated. serum. of..a, person who, -had. vaccine, four years ago;
agglutination, titer, 1:10
C30 auaorre. +x aRESYMTS, WIPE, PIFEERENT ANTIGENS
PREPARATION OF +B. TYPHOSUS:;ANTIGEN F27
47 ‘TABLE 2
ae rr
r
Lith
gmszitas, svitogg Spys sy
Restilts with the MALE serum ‘of a person never Rivthg had fo id: ate or bacciye?
= ateeedie test showed Bessie result AT stsdo
as ca | ePa9
r73 tp t9vet
showed that all kept en well a
increased anticomplens@mtagy, activiti
b. Ferkolytic ptewe Me the dif :.
gens Vee es emolytigz ‘ag Cpe in
tables Hd antigens, / were nk Hy hemablytie-and .@s-
peciallantigen Gewhith wash suspergion @prottad dried bacilli
in salifige a4 Ameren was a fonrt@em-daag broth culture
heatedeat 6G). dal prescryad with Og. henol} antigen
4) was se suspend d water, pre-
| e prdpatiton
}
pared : P nf x sf only e
ff free of
The anti-
of gon
There
hemol¥ es. co &
c. Agbiger 4 te di nig eAS previously
stated gies@ were determine 9! - barative tests
in which each ‘antigen \ was s employed in an amount equal to ¢ one-
human and rabbit immune sera; all antigens ‘were ed at one
time/ with each,serumjand with,the same -hemolytic system; in;
order to render the,.tests strietly gomparativee 2t{yes1 oft eovig
ailhe results observed with ;fqurrabbit immune; sera,of, yaryng
agglutinin. content; are shown. in. tables 3, 4,-5,.6, 7 and. 85 chart J,
shows graphically the.-yariation, im, antigenic, sensitiveness of, the;
different er ae Tables 9 and 10 also summarize the re-
sults, Vvintg the smallest’ amount’ of éach® ‘impjune: seruit Hiéated,
128 MOTOMATSU MATSUMOTO
and unheated, yielding reactions with the respective antigens
(chart 1).
Tables 10 to 26 give the results observed with the various
antigens and the sera of persons with typhoid fever or convales-
cent from this infection; also the sera of persons who had typhoid
fever or typhoid-paratyphoid vaccine at varying periods prior
0.001
0,005
OC: OsS—
0.02
CHART 1. SHOWING THE COMPARATIVE ANTIGENIC SENSITIVENESS OF THE
VARIOUS ANTIGENS WITH RABBIT IMMUNE SERA
to the time when sera were collected for these tests... Table 27
gives the results observed with a normal serum.
As shown in these tables the antigens varied considerably in
antigenic sensitiveness and may be summarized somewhat as
follows in order of delicacy and sensitiveness.
1 T am indebted to Dr. John Eiman and Dr. Stanley P. Reiman for several of
these sera from typhoid fever patients.
if
PREPARATION OF B. TYPHOSUS ANTIGEN 129
1. Antigen 3 was uniformly most antigenic in all tests; this
was one of the simplest, being a fourteen-day broth culture killed
by heating at 60°C. for one hour and preserved with 0.5 per cent
phenol.
2. Antigen 1 was generally second in antigenic value and was
prepared by suspending living bacilli removed from agar slant
culture in sterile saline solution as required.
3. Antigens 2, 6, 7, 8 and 9 were about equal in antigenic ac-
tivity and generally represented antigens prepared by grinding
dried bacillary sediment secured with or without alcoholic pre-
cipitation, and suspending the very fine powder in saline solution.
4, Antigen 4 was uniformly poorest in antigenic sensitiveness
and this result is of considerable significance inasmuch as it was
prepared after a method commonly employed in the preparation
of gonococcus antigen.
In general terms the results of this study with antigens of
B. typhosus have shown that the best products are those in which
are used whole bacilli in suspension with or without their soluble
products elaborated during growth in fluid culture media; the
next best are those antigens composed of thoroughly disrupted
bacillary bodies in suspension and the lowest in antigenic value
was the antigen prepared by autolysing the bacilli for intracel-
lular substances and utilizing the filtrate which probably carries
the soluble products and lacks the insoluble bacillary bodies.
Exactly similar results and conclusions have been recorded by
Kolmer and Brown in a study of gonococcus antigens, previously
referred to.
A further result of this study was to show the marked effect of
heating upon the typhoid antibody concerned in complement-
fixation with human and to lesser extent with rabbit sera; all of
the unheated sera of persons with typhoid fever or convalescing
therefrom and the majority of those who had received typhoid-
paratyphoid vaccine, yielded a positive reaction with the majority
of antigens; after heating at 56°C. for thirty minutes a marked
reduction in the degree of complement-fixation was observed;
this was due, presumably, in part to the thermolability of the
r3Q - «+ MOTOMATSU MATSUMOTO
antibody inasmuch as the unheated sera were used fresh and
found free of demonstrable anticomplementary activity in the
serum control tubes, which always showed complete hemolysis.
ae
ecw bak oullty oitenisce CONCLUSIONS
a” Nine gneieens prepared ie a single strain of B. typhosus af-
ter various methods have shown well defined differences i in anti-
ae sensitiv eness. |
Antigens prepared from living or dead - suspensions e ba-
a in saline solution or culture broth proved most antigenic;
antigens prepared by suspending the powder of dried and ground
bacilli i mn saline solution proved next best in antigenic sensitive-
ness and an antigen prepared of the filtrate of bacilli autolysed in
distilled’ water ‘aided by heating at a high temperature, “proved
least antigenic. The method of preparing typhoid antigen has,
therefore, a marked effect upon the occurrence ‘and degree of
col uplement- -fixation tests and the s same is probably true oO bac-
ee antigens in general.
3. The anticomplementary activity of the various antigens did
not appear to differ to a marked extent; also, several were more
hemolytic than the others; all of the antigens’ appeared to keep
uniformly well. over, a ‘period of six weeks at or near the LUBEES
point. /
as The general result of this study and a review of investiga
ons by’ others indicates that the similar bacterial antigens in
W Fick j is employed the whole microorganism either living or dead
D hysiqlogical saline solution or in culture broth, are superior
iltrates and constitute the antigens of choice for the conduct
of | complement-fixation tests in bacterial infections. I beg to
oe my appreciation to’ Professor Kolmer na directions and
aid in conducting this work.
Uhinopacs ond Alen waits “REFERENCES |
b a [3271
(), Korner, bp 1% AND ‘Brown, C. Pp. Complement fixation, in gonococeus in-
d ectigns. ‘Jour. ‘Infect. Dis., 1914, 15, 6-21.
ahi Koumbn,! geo ) AND ‘ BERGE, Jie ‘The Hélation of the typhoid in skin reac-
tion to sagen in typhoid fever. Jour. Immunology, 1916, 1, 409.
PREPARATION OF B. TYPHOSUS ANTIGEN 131
(3) Kotmer, J. A. The relation of the diphtheria skin reaction to immunity in
diphtheria. Jour. Immunology, 1916, 1, 443.
(4) Koutmer, J. A.. Matsunami, T., anp Harkins, M. J. The relation of the
bronchisepticin skin reaction to immunity in canine distemper includ-
ing the bactericidal action of dog serum for B. brochisepticus. Jour.
Immunology, 1916, 1, 571.
(5) Hitrcuens, A. P., anp Hansen, G. Studies on antibacterial serums. A
stable bacterial antigens with special reference to meningococci.
Jour. Immunology, 1916, 1, 355.
(6) Smatu, J. C. A method of preparing bacterial antigens. Jour. Immunol-
ogy, 1918, 3, 413.
(7) Witson, M. A. A contribution to the study of the complement fixation
reaction in tuberculosis. Jour. Immunology, 1918, 3, 345.
(8) Kotmer, J. A., aND Trist, M.S. Non-specific complement fixation by nor-
mal rabbit serum. Jour. Infect. Dis., 1916, 18, 20-26.
i‘
i
>, a ee
AN EXPERIMENTAL STUDY OF THE EFFECT OF
AUTOGENOUS B. COLI VACCINES ON THE INTES-
TINAL COLON BACILLI OF DOGS
JOHN C. TORREY anp ALFRED H. RAHE
From the Department of Hygiene, Loomis Laboratory, Cornell University Medical
College, New York City
Received for publication February 11, 1920
From time to time, within recent years, the use of autogenous
B. colt vaccines has been advocated as a therapeutic measure in
the treatment of such conditions as chronic intestinal toxaemia
(1) and eczema (2) on the assumption that the toxic substances
giving rise to these conditions are produced through the activi-
ties of certain B. colt vegetating in the intestinal tract, and that
these strains may be suppressed or eliminated through specific
immunization. Apparently, however, this mode of treatment
has not been substantiated by any experimental evidence either
that the B. cola of the intestinal tract may be controlled through
specific therapy or, if particular strains are reduced in numbers
through this procedure, that the effect obtained is more than
transitory. The study reported here was undertaken with the
hope of throwing some light upon these points.
Normal dogs were utilized in these experiments in which an
attempt was made to reduce in numbers or eliminate certain
strains of B. coli naturally vegetating in their intestines.
All organisms belonging in the colon group ferment lactose,
but only certain varieties split sucrose. Advantage was taken
of these distinctions within the group in selecting strains for the
preparation of the vaccine and in estimating the specific effect
of inoculation on the distribution of colon types in the fecal
specimens. In other words, if there exists any rational basis
for attempting to control in a practical way through specific
vaccine therapy the types of B. coli within the intestinal tract,
133
134 JOHN C. TORREY AND A. H. RAHE
then in these experiments inoculation with representatives of
the sucrose-positive B. coli should cause an elimination or, at
least, a marked reduction of these types as revealed in examina-
tions of the fecal specimens. 2) :
The dogs, used int these! &xperiments, Were’ ke tonal donstant
diet of boiled rice 4nd bdifed beet’ hearts’ in'the Patid BY Avéight
of about 2 to 1. ATHY Giet Was found by dnedfi us (3) to be
favorable for the dexelpprent of, an gatestinal;flora dominated
by B. coli. . In fact. often. the enly, colonies. appearing on the
Endo plates were B. colt-like: «;‘Vhe,anmrount of rice and meat
fed was not weighed each day, as there would be no advantage
in such precautions, but the relative proportions were kept
approximatély xconstaint.is9 7 ta9091 aidtiw ,satit of satt croTT
‘The fecal:specimensi were hatural movements eolleeted inthe
morning. | Rather: heavy: emulsions; itepreséntative of theswhole
stool; were made in normali salme-solution, and: froms suitable
dilutions Endo platés weve seededz 1 In the»preparation iofuthese
Endo plates sucrose was substituted: fersthé usualdactoseio On
these plates, of course, théosuerbse-fermenting straims of B3-calt
appeared as red) colonids, whereas! the! varieties of. du colicwhichi
cannot split sucroseogaverise;to: white colomes: After! twenty
four or more hours)incubattom the:ratio:oh “whites)’. fosffreds!’
among: the B. \coli-like: colomies iwas determined: and decorded:
Frequent control. tests werésmaidet td} demonstrate: the sucrose
fetménting properties of the bacilli forming the red colonies and
they were invariably fount ecapablé: ofs splittimg: this !sugar!
Difiérential cultural tests were also carried:out»on large number
of isolated straits to establish:itheinidentity: sAmongsthe’severak
hundred°cultures examined no! representativesat the B. aerogenes:
type was encountered; .Of)onel hundred: sutrosespositive, gela-
tin-negative cultures from one dog ‘thirty were positive-for'salicin!
and seventy were negative. «According to: Levine’s:(4) i¢lassifica-
tion’ of) the sucrose; fermenters:the:salicin+positive types:should;
be) designated » as’ B:>meapolitanws |and bthelisalicin-negative: as
B. communior. or B. eoscoroba depending upon motility «Among
three out of the four dogs) differential fermentation’ tests’ with!
the sucrose-positive B.'t¢oli: showed) the: salicin-negative tiypes
}
EFFECT ‘OF -AUTOGENOUS SB GOLI VACCINES 135
considerably in the majority. Gelatin Hquefying strains among
the sucrose-positive cultures seemed to’ be comparatively rare as
Gut of 150 strains examined only 10 liquefied gelatin; accordingly
about 7 per cent! of these ‘cultures should be placed in the B.
cloacee-groups The main objective‘in these differential cultural
4éstso was; of! course, the selection’ of ‘sucrose-positive *B.: coli
representative ‘of’ all’ present ‘in the intestinal flora. All the
varieties ‘of! sucrose-positive B. coli isolated: were incorpor sical
in-the vaceine-except the gelatin: liquefiers. «
For the purpose of these experiments, then, a definite group
of-Be colt, viz. the sucrose ferméenters| was selected to test the
practicability of controlling: these:and: related intestinal organ-
isms through specific immunization. “These bacilli are constant
and normal imhabitants of-the intestinal tract) and ones’ which
may ‘be recoghized readily through differential: cultural and
serological tests! “In some ways it ‘would have been preferable
if bacterial: species foreign’ to” thé intestinaltract could ‘have
been utilized. © Implantation of such foreigm strains, however,
does aotseenrto be'possible;jand ne experiments? along: that
Vineswere: attémpted. - could tdeis9 o3 doris? 9990 cintt bee
> Preliminary «examinations ~for\‘eachlianinial’ were made at
regular intervals for a period of four to six:week&$ ‘with the pur
péséof determining not) only!the average! ratio'and the degree
of variation ir: théiconiparative' prevblehee ‘of the red’ and white
éoldn ecloniesforithe normal animal, but also for the: selection
of Tepresentatives (ofo the: omam vbulturali variants’ among the
sticrosespositive’ Bircolst Wath these selectedicultures'a vaccine
was préparedoand alsocrabhits were! immunized for the produc:
tion of asspétific anti¥eserumi? (When!the range inthe ratio’ of
the red and'white colonies for theinormal animal had been deter-
rained; iavseries of vactitieinoculations Were ‘given and ‘the cul-
dural léxaminations were (continued! atfrequent intervals: At
each plating during the period of immunization ten well isolated
fed’ colonies were? transferred to! agarslarts'and! agglutination
tests’ wére' carried out with cach, by sing! the serum from the
Fabbit inoculated with tHe vactine cebturéds.)) This anti-serum
had @ titer of 1-5000 to 1210,000 foreach of the vaccine cultures.
136 JOHN C. TORREY AND A. H. RAHE
Tests with the isolated strains were made macroscopically at
dilutions of 1-50 and 1-500.
aa,
at 8
Sa a ee, ee eee .
SENSITIZED CHOLERA ANTIGEN 147
(containing 2 mgm. of bacilli) of ordinary sensitized .vaccine and
in the second group 0.1 gram of the sensitized cholera vaccine
powder (containing 2 mgm. of bacilli) was injected into the vein
of the ear of each rabbit. On the ninth day following the injec-
tion, when the production of antibodies is always highest, the
blood was withdrawn from the heart. The separated serum was
heated for thirty minutes at 56°C. to be made inactive and then
the following experiments were made with it.
1. Experiments on the development of agglutinin and bacteriolysin
In order to know proportionately how much of the antibody
is agglutinin and bacteriolysin, and how much is present in this
serum, the agglutination reaction and Neisser-Wechsberg’s
method were used. Results obtained are shown in table 1.
a. Experiment on agglutination index. One hundred test tubes
were divided into two groups, the first group for the ordinary
cholera sero-vaccine, and the second group for the powder vac-
cine, and then each group was arranged into five series of ten
tubes each, each series for one immune rabbit. Into the first
tube of each series was put 0.90 cc. of the saline solution and in
the others 0.50 cc. and further in each first tube 0.10 cc. non-
diluted immune serum of its rabbit. 0.5 cc. of this was removed
and placed in the next test tube, and the process repeated for
each succeeding tube. Then to each tube was added 0.50 cc.
of living cholera bacilli in suspension (1 ce. of this suspension
contains 1.0 mgm. agar culture of cholera bacilli in eighteen
hours cultivation). The total volume in each tube was, thus,
1 cc. The mixtures were thoroughly shaken. Thereupon these
tubes were incubated for two hours at 37°C., and the result was
read as shown in table 1.
b. Neisser-Wechsberg’s experiment. ‘The serum was diluted as
in the preceding experiment and 0.30 ce. of complement (1: 10)
and 0.50 ec. of bacilli suspension (1/500 mgm. in 1 ce.) were
added to each dilution of serum. ‘Then, after these tubes were
incubated at 37°C. for two hours the agar medium was poured
into each tube. When the agar was completely coagulated it
Y. MIURA
148
IOULNOD
we
+++
0982
O82T
Il do0up
oa
+++ | +++
+++ | +++
+++] +++
+++ | +++
+++ | +++
+++] ++
+++] +
++ | -
+ —
+ =
4- —
- ~
- +
++ | ++
XOpUt UOI}eUIyN [say
0 | ¢
8}IqQqei JO oquINyy
++ 1
1 @TAV.L
++ | +++] 444
I aqanoup
444
a
aoe
ae
“+
op
4
fe
+
+
44+
+++] ++
T000000 °0
2000000 *0
000000 °0
T00000 °O
$%00000 0
00000 0
T0000 ‘0
¢Z0000 °0
¢0000°0
T0000
¢z000 0
¢000°0
T00°0
$2000
G00°0
T0°0
¢20°0
G0°0
0992 O8eT
X@aNI
OILATOINALOVA NI LIOsaH
SENSITIZED CHOLERA ANTIGEN 149
was incubated again for twenty-four hours and the number of
colonies was counted.
According to table 1, the index of agglutination was 1280 to
5120 in rabbits of the first group and 640 to 2560 in those of the
second group. The bacteriolytic result in the first group shows
the index of 0.00025 to 0.000025 and in the second group of
0.0005 to 0.00005, thus proving that the agglutination and
bacteriolytic indices are nearly the same for these two vaccines.
2. Pfeiffer’s experiment
From the preceding experiments it was evident that the
antibody was produced by the intravenous injection of 0.10
gram of this vaccine powder. In order to obtain a more accu-
rate result, however, we tried the following experiment:
TABLE 2
GROUP I GROUP II
20 Alive} Alive} Alive} Died} Alive} Alive} Alive} Alive} Died! Alive
240 Alive} Alive} Alive} Alive} Alive} Alive] Alive] Died| Died! Alive
80 Alive} Alive] Alive! Alive} Died} Alive] Alive} Alive} Alive! Alive
160 Alive! Alive} Alive} Alive} Alive} Alive} Died| Alive} Alive} Died
Alive! Died} Alive} Alive} Alive} Alive] Alive} Alive} Alive! Died
Alive} Alive! Died| Alive} Alive} Died| Alive} Alive} Alive! Alive
: 1280 Died! Alive! Died| Died} Alive}! Died! Died} Died} Alive} Died
: 2560 Died| Died| Died| Died! Alive} Died! Died| Died | Died! Died
: 5320 Died} Died} Died| Died| Died| Died| Died} Died| Died! Died
: 10640 Died} Died! Died| Died! Died| Died| Died} Died} Died} Died
Ph
One hundred healthy guinea-pigs from 150 to 200 grams body
weight were divided into two groups. In the first group an
immune serum of the common sensitized vaccine, and in the
second group that of the sensitized vaccine powder, was injected
intraperitoneally, then immediately 1 cc. of the living cholera
bacilli suspension (that is twice the lethal dose) was introduced
intraperitoneally into each of them, and examination made
after twenty-four hours. The results are shown in table 2.
150 Y. MIURA
TABLE 3
GROUP I GROUP II
BATS Num-| p, a Dose Hone Body Dose ir
Asati weight jee ane fault ber of weight ahr Ween Re
Tt Sy eee grams cc. mgm. grams cc. mgm.
1 175 1 2 | Died 61 200 1 2 | Died
2 181 1 2 | Died 62 | 200 1 2 | Died
First 3 195 1 2 | Died 63 | 200 1 2 | Died
4 190 1 2 | Died 64 195 1 2 | Died
5 170 1 2 | Died 65 198 1 2 | Alive
6 200 2 | Died 66 | 200 2 | Died
Uf 170 1 2 | Died 67 | 160 1 2 | Died
8 185 1 2 | Died 68 182 1 2 | Died
Geeand 9 190 1 2 Died 69 | 195 1 2 Died
10 170 1 2 | Died 70 199 1 27 Died
11 192 1 2 | Died 71 187 1 2 | Died
12 200 2 | Died 72 | 198 2 | Died
13 170 1 2 | Died 73 | 180 1 2 | Died
14 150 1 2 | Died 74 160 1 2 | Died
Third 15 172 1 2 Died 7A Plies 1 2 Died
16 178 i 2 | Died 76 165 1 2 | Died
17 198 1 2 | Died te) G55 1 2 | Died
18 185 2 | Died 78 | 200 2 | Died
19 185 1 2 | Alive) 79 | 195 1 2 | Died
20 195 1 2 | Died 80 | 180 1 2 | Died
Reeth 21 191 1 2 Died 81 180 1 2 Died
22 192 1 2 | Died 82 | 180 1 2 | Alive
23 155 1 2 | Died 83 | 185 1 2 | Died
24 200 2 | Died 84} 200 2 | Died
|
25 190 1 2 | Died 85 | 190 1 2 | Alive
26 195 1 2 | Alive} 86] 155 1 2 | Alive
Fifth 27 155 1 2 | Died 87 | 175 1 2 | Alive
28 160 1 2 | Alive) 88] 175 1 2 | Died
29 175 1 2 | Alive} 89} 160 1 2 | Died
30 200 2 | Died 90 | 200 2 | Died
31 190 1 4 | Alive} 91 190 1 4 | Died
32 170 1 4 | Alive} 92] 195 1 4 | Alive
Sixth 33 160 il 4 | Died 93 | 180 1 4 Alive
34 165 1 4 | Died 94] 170 1 4 | Died
35 155 1 4 | Alive} 951} 170 1 4 | Alive
36 200 4 | Died 96 | 195 4 | Died
ala
SENSITIZED CHOLERA ANTIGEN
TABLE 3—Concluded
GROUP I
eee Num- Dose Dose Num-
Hagel weight Vansiie living es pects weight
day grams cc. mgm, grams
37 200 1 4 | Alive} 97); 170
38 195 1 4 | Alive} 98] 170
Seth 39 175 1 4 Alive 99 | 192
40 180 1 4 | Alive} 100] 190
41 195 1 4 | Alive} 101 185
42 195 4 | Died| 102} 200
43 190 1 4 | Alive} 103 | 200
44 195 1 4 | Alive} 104] 200
; 45 175 1 4 | Alive} 105 | 200
eee 46 | 165| 1 | 4 | Died} 106| 190
47 165 1 4 | Alive} 107} 190
48 195 4 | Died} 108} 200
49 160 ik 4 | Alive} 109} 170
50 190 1 4 | Alive} 110 | 175
Ninth 51 185 1 4 | Died| 111 159
52 183 1 4 | Alive] 112 165
53 180 1 4, | Altve). 113, )al77
54 190 4 | Died] 114] 200
55 180 1 4 | Alive| 115] 155
56 182 1 4 | Died| 116 157
Tenth 57 175 1 4 Died 117 168
58 177 1 4 | Alive] 118 175
59 180 1 4 | Alive} 119} 177
60 200 4 | Died} 120} 200
GROUP II
Dose
of
vaccine
et a el ee
ea
Dose
of
living
bacilli
PEP PR PPP PPP PP PP PP
De ee
151
Comparing the results again with those of the above aggluti-
nation and bactericidal experiment, we have proved that the
activities are nearly the same in both vaccines.
Five control
guinea-pigs used in this experiment all died within twenty-four
hours.
3. Experiments on prophylaxis
We divided 100 healthy guinea-pigs into two groups; | ce. of
the sensitized cholera vaccine was injected subcutaneously into
152 Y. MIURA
each of the first group, and 0.10 ec. of the sensitized cholera
vaccine powder into each of the second group. Then every day
six healthy ones were selected from each group and 1 ce. of bacil-
lary suspension (2 to 4 mgm. of bacilli) was injected intraperi-
toneally. Examination was made every twenty-four hours.
As a control a healthy guinea-pig was taken every day and the
same dose of the bacillary suspension was injected intraperi-
toneally, but all died within twenty-four hours.
In considering the result of the prophylactic experiment we
found that the production of antibodies increases gradually from
the fourth day, in each group, until it is at its highest on the
seventh, eighth and ninth days, and begins to decrease again on
the tenth day.
III. OBSERVATIONS ON THE REACTION IN THE HUMAN BODY,
FOLLOWING THE INJECTION OF SENSITIZED CHOLERA
VACCINE POWDER
The most important means for comparing the grade of the
reaction is the recording of the body temperature after the
injection and the local and general subjective symptoms.
The results observed after injection of the powdered prepara-
tion into the human body are shown in table 4.
There was almost no temperature reaction. In some cases
we saw a slight induration, but no pain, with recovery in one or
two days. The general symptoms were only slight lassitude or
headache in a few patients but none were confined to bed. Here,
also, we obtained a favorable result by using the sensitized
vaccine powder, as in case of the common sensitized vaccine, in
which the reaction is very slight.
We avoided injection into the aged, infants, invalids, drunk-
ards and pregnant women and into cases with heart disease,
pulmonary tuberculosis, nephritis and beri-beri.
153
SENSITIZED CHOLERA ANTIGEN
ODATPVION
OAT] VIIN
OAT}BVION
SuI[[OMS JO Burpaay
dAIZVION
dAIYBVION
ured yy3I1g
dATP VION
ured yyaiTg
DATPVIIN
OAIYBVION
OAT} BION
AAI} BION
Ssoulepudy,
OATZVBON
BuI[OMS JO Surlpoay
DATPVBIN
OAT} VBON
SSOUIOpUd} FYSITS
OATPVBON
NOILOVaUY TVOOT
OAT} VION
DAI} BVIIN
OAT} VION
peoy ul
SSOUTABOTT
pvoy ur
SSOULABOH]
DATPBSIN
OATZVBON
OAIYVION,
apnysse'y]
OATPCION
OAIYVION
DATYBVIIN,
OAT} BION,
psey
ul Ssou
-IABOY
WSIS
DATPCIIN,
opnyisse’yT
OAT}BIIN,
opnyisse'yT
dAIVVSIN
OATPBSON
SNOLIWAS
GALLOara os
IVUGNAD
ba |
SSS SS OS oS ot ot
poo)
IIB
pood A194
pood A194
peo’)
peo.)
pee
pood A190
Peos)
ATV
poosd A104
pood A104
poos Ala,
poos A1aA
[ora
poo
ITB]
poo!)
pood AoA.
Poo,
Iooyyo ATVYITLTAY
UBUIODT[Og
yuepNys
Ia0qjo Arey
10490q
UBUIOM SsouIsng
OFTMPTTAL
10390q
[SoHo
[BOBO
[vyidsoy ut 1adpo'T
[syidsoy ul 1odpo'T
[eyidsoy ul 1edpoT
[eyidsoy ul 1adpoT
[ejtdsoy ut 1aspo'T
[eydsoy ut 103po'T
[syidsoy ut 1ospo'T
[syIdsoy ul 10dpo'T
[eyidsoy ur 1e3poT
[eyIdsoy ul 1espo'T
ioredng
CON IPerAl
roredng
rotedng
iotedng
ioredng
CEnTpeTy
ioredng
Joradng
CORPS
io1edng
CP e TN
CORTP STN:
UmnIpely
UNI pel]
oriedng
CNP
Man PeTN
io1edng
unIpey
N
=x
OF OFIOR ND Ng DOF CE ‘D ‘oD oo
WO mee
AHN swe
G°9€ 9°9€
0°LE P'9€
8°9§ 8°9§
T 2é 1°98
6° 9€ & 9E
¥'9€ PF 9€
€°9€ ¢'9€
9°9€ ¢ 9€
8°9§ Pb 9€
9°9€ G9
G'9§ PF 9€
0°98 €°9€
0'LE 6°98
OLE €°9§
0°98 P9§
0°LE 9°9€
0°9€ T'98
¢'9§ €'9€
0°LE 9°9€
& 9E T'9E
*— saa.bap |‘ saatbap
NoI“oar
“NI SOME | SOILOELNI
Peet gos
auoL
-vagawag |"V2tdWan
ISaHDIH
NOW
-NI WO
asoa
NOILINLON
NOILVYd2000
fipog ununy uodn uonyovay
¥ UTAVL
NOILOLILSNOD
@OvV | Xas
GWYN
154 Y. MIURA
CONCLUSIONS
1. The quantity of the agglutination and bacteriolysin pro-
duced by the treatment with the sensitized vaccine powder is
nearly equal to that by the common sensitized vaccine. .
2. The reaction on the part of the human body is very slight
after the injection of either of these vaccines.
3. The sensitized vaccine powder does not decompose as
rapidly as the common sensitized vaccine. It has been pre-
served for nineteen months without losing its value as an antigen.
REFERENCES
(1) Besrepxa, A.: Ann. Inst. Pasteur, 1902, 16, 918.
(2) Mercunikorr, E., AND BesrepKA, A.: Ann. Inst. Pasteur, 1911, 25, 193.
(3) Suraa, K.: Kitasto’s Arch. Exper. Med. 1918, 11.
A DROPPING BOTTLE AS AN AID IN MACROSCOPIC
SLIDE AGGLUTINATION
CHARLES KRUMWIEDE
From the Bureau of Laboratories, Department of Health, New York City
Received for publication February 25, 1920
The macroscopic slide agglutination method has been more or
less generally known for years. The value of the method, how-
ever, has not been generally appreciated, probably due to the
fact that most of the serums employed for identification purposes
have not been sufficiently active to give an immediate or prompt
agglutination even in low dilutions. Coca (1) utilized the method
in cholera examinations and, to our knowledge, he was the first
to publish a description of the method. We have utilized it
for some years as a routine in examinations of feces for mem-
bers of the typhoid-paratyphoid dysentery types (2) and also as
an aid in the search for meningococcus carriers (8). As very
potent sera become more generally available its use will un-
doubtedly spread to most laboratories.
There is one inconvenient feature in the method which we
believe to have overcome by the device here reported. In plac-
ing on the slide the drops of diluted agglutinating serum and the
saline or diluted normal serum for control it has been the custom
to utilize the ordinary platinum loop. Where many colonies are
to be tested this method is time-consuming and tedious.
The dropping bottle that we have devised to deliver the drops
depends upon the use of a capillary delivery tube with a flat
end and a rubber diaphragm to force the delivery of the drop.
The delivery tube is prepared from a piece of heavy walled
tubing (thermometer type) having an outside diameter of about
4 mm. and a bore of about 1 mm. ‘This is heated thoroughly in
the flame and drawn out slowly. A piece is then cut as shown in
the outline sketch (figure 1) and the larger end ground flat and
155
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 2
156 CHARLES KRUMWIEDE
smooth on a fine stone. The diameter of the bore is reduced to
about the bore of 20 gauge hydpodermic needle.
The diaphragm is made from a so-called ‘‘no-air” stopper,
size No. 23. The cork end is cut off leaving only sufficient to
give stability to the delivery tube. The delivery tube is thrust
through the cork.
The bottle employed is the regulation 4 ounce wide mouth
homeopathic vial. The neck of the bottle must be wider than
the cork portion of the stopper to allow up and down play
otherwise there will be no diaphragm action.
The bottles are partly filled with the diluted serums or saline
and the drops are placed on the slide by holding the bottle ver-
tical and pressing till the appropriate sized drop is delivered.
With very little practice many drops can be delivered in a few
seconds. The appropriate pressure to apply soon requires no
conscious attention.
The amount of air or air tension in the bottle needs readjust-
ment from time to time. Air can be introduced by inverting the
bottle and pulling on the delivery tube. If there is too much air
pressure this can be equalized by holding the bottle upright and
pressing down the diaphragm which empties the tube and results
in equalization of the pressure.
There are other possible applications of this dropping bottle.
It may be utilized to deliver drops of saline or other fluids used
in the preparation of smears of cultures. In the classroom it
could be utilized to give out suspensions of microérganisms for
smear examinations. Other similar uses will probably be found
for it.
REFERENCES.
(1) Coca, Artuur F.: Bulletin Manila Med. Soc., 1910, 2, no. 1.
(2) Krumwiepe, Cuarzes: Journ. Inf. Dis., 1918, 23, 275; 1917, 21, 141.
(3) Krumwiepbe, Cuarues: Journ. A. M. A., 1917, 6-9, 358.
AN AID IN MACROSCOPIC SLIDE AGGLUTINATION 15
delivery Tube... , cs
oo , (1 i?
; Cross-section o}
No-atr” Sto b per
a-Tim turned down
b= Yim Turned a
dotted line indi-
cates where cork
AS eut,
Fria. 1.
ig
e.
adh aR
THE COMPLEMENT FIXATION TEST FOR
TUBERCULOSIS
HASSOW O. von WEDEL
Received for publication February 26, 1920
CONTENTS
oi Te VUEETE ce he's AE Ae Se 159
Mare IOREUIE WRIST OAG oo Soares oo cc's F's cos ce eR PTCA asc pa'enee wa 162
. Technic of the standard complement fixation test for tuberculosis used
STIR AE CAPA PA A SS oc 3) dong «in: 2) 98 6 Seabees aie sein b me * 169
. Determination of the optimum time and temperature of fixation......... 173
map suandaraizavion Of Complement...............cdecccrccaccccecccess 174
. Determination of the thermolability and the thermostability of anti-
Snes eel Momolyeing in human SeLAa. .... .... 5 0:deene dedaas sewers nciens 177
7. Determination of the relative antigenic value of tubercle bacillus anti-
gens prepared and sterilized by twenty different methods............. 184
8. Determination of the relative value of the Hecht-Gradwohl technic
and the technic of natural antisheep amboceptor absorption in com-
parison with the standard complement fixation test................... 191
9. Comparative results obtained by tests with sera one day old and with
these same sterile sera after preservation for one week in the ice box... 195
10. Determination if any cross fixation exists between the tubercle bacillus
antigens used in this study and syphilitic reagins..................... 199
11. The relationship of the von Pirquet reaction to the complement fixation
SPERM IMURERORTITORS es Se clack eh via oo coe de ce sata eee cc ss css 200
12. The classification of tuberculosis patients.................0002 cece sees 201
13. The relationship of the patients’ temperature, pulse, respiration and
ago £0. the complement fixation test..............c0cecccesncdeswecees 204
14. Summary of six series of complement fixation tests....................-- 205
. The value of the complement fixation reaction to the clinician in the
diagnosis and prognosis of tuberculosis..................0000eeeeeeeee 213
si 8 es, Sas v,alaw p Aon apie See Rok ok me cease 220
NE IE REPRE TE 0,2 553. ins ch a. 0 a cip: 0 0 20's oceinig stelg aU Saisie wsinc pee ce 222
1. INTRODUCTION
In spite of the epoch-making discovery of Koch, the early
diagnosis of tuberculosis is still a problem to be solved, even as
it
was one hundred years ago in the days of Laennec.
Although methods for the early diagnosis of tuberculosis are
still wanted, the wide distribution of this disease is well estab-
159
160 HASSOW O. VON WEDEL
lished. In fact, tuberculosis, not unlike syphilis, must be con-
stantly kept in mind in making a diagnosis, whatever may be
the clinical picture of the case.
McCrae and Funk (1) in 1919 stated that, although the recog-
nition of chronic pulmonary tuberculosis is generally regarded
as a simple matter in which there is slight chance of error, there
is a definite percentage of errors made in diagnosing the disease
in this stage. They found that 72 out of a series of 1200 con-
secutive cases admitted to the Jefferson Hospital as advanced
pulmonary tuberculosis were incorrectly diagnosed.
In the quest for a reliable method for this diagnosis, numerous
investigators have repeatedly attempted to apply to the diag-
nosis of tuberculosis methods on which the diagnosis of other
infectious diseases are based, such as the agglutinin, precipitin,
meiostagmin and epiphanin reactions. Their attempts were
crowned with partial success only, as the tests were found to be
of little diagnostic value.
The discovery of the tubercle bacillus by Koch in 1883 placed
the diagnosis of tuberculosis on a substantial foundation, but the
bacillus cannot always be demonstrated early in the discharges,
frequently never appearing, and even if present giving little clue
to the degree of activity or inactivity of the disease. Thus far
biologic methods of diagnosis have been of little practical value
with one exception—complement fixation. Though not ful-
filling the early expectation, this method of diagnosis has been
gradually improved so that there was promise of its becoming as
valuable a diagnostic test as the Wassermann reaction in syphilis.
The remarkable usefulness of the Wassermann reaction espe-
cially stimulated the efforts of numerous investigators along the
lines of application of the Bordet and Gengou reaction to the
diagnosis of tuberculosis. The earlier efforts along this line
were not very encouraging. ‘This could be due to several causes.
It is possible that (due to the walled-off nature of the lesions and
the slow process of the disease in certain cases) there may be no
immune bodies or only a very few of these present in the circu-
lation. Moreover, the concentration of circulating antibodies
is subject to constant and quite marked fluctuation in the same
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 161
patient. Besides, the antibodies in tuberculosis may not be, to
any great extent, of the nature of amboceptor. Again, tuber-
culosis amboceptor, as suggested by Davidowitch, may be more
thermolabile than most others, and since complement deviation
was usually performed with inactivated serum, the amboceptor
may have been largely destroyed in heating; thus the amount
remaining in the serum may not have been large enough to be
detected even by the delicate method of complement fixation.
This complement fixation reaction for tuberculosis has occu-
pied, for nearly twenty years, the attention of many investi-
gators who have studied it mainly from the standpoint of its
possible value as a diagnostic and prognostic aid in clinical
medicine.
Results obtained by different workers show considerable
variation. When, however, it is remembered that the value of
the test is entirely dependent on the specific action of the antigen
and that the antigens used by many investigators have been
prepared by widely divergent methods, it is to be expected that
the results would be somewhat dissimilar. At best, we are but
feeling our way toward a common understanding of the relation
which our laboratory findings bear to the disease itself, and the
test will eventually be established as of practical value only by
a thorough comparison of all careful investigations rather than
by the consideration of the work of one person.
In the diagnosis of tuberculosis it is not to be expected nor
desired that the complement fixation test should replace the
ordinary examination of sputum for tubercle bacilli, but if it is
to be of any practical value to the clinician, a positive reaction
must be specific. With these facts in mind, the writer has
studied this complement fixation reaction to determine its value
as a routine diagnostic test for tuberculosis.
In this work, the following questions have been taken up.
First, what is the best general technic to employ for this test?
Is special complement a necessity? What is the optimum
fixation time and temperature, and is it necessary to keep the
patient’s serum for any period of time before testing? Secondly,
what antigen will give the highest possible percentage of specific
162 HASSOW O. VON WEDEL
positive reactions and none or only an occasional non-specific
reaction? Thirdly, is there any advantage in using the fresh
non-inactivated serum with its native complement and ambo-
ceptor, or in removing the natural amboceptor from the inacti-
vated serum? Fourthly, will the tubercle bacillus antigens used
in this study give any non-specific cross fixation with syphilitic
sera? Fifthly, is there any relationship of the complement
fixation test for tuberculosis to the age, temperature, pulse or
respiration of the patient? Sixthly, what is the best method for
classifying tuberculosis patients from the combined view-point
of the serologist and the clinician? Seventhly, what type of
tuberculosis case gives the highest percentage of positive reac-
tions? And lastly, just what value is this reaction going to be
to the clinician as an aid to diagnosis and prognosis?
In this study the writer has made 6128 complement fixation
tests on 1207 sera taken from 1000 patients. Of these sera 633
were from 484 patients in tuberculosis hospitals. The remain-
ing 574 sera were from 516 patients in general hospitals suffering
from various other diseases. There were made 1167 compara-
tive tests with sera from 60 patients with 20 different tubercle
bacillus antigens. Tests were made with 60 sera by the Hecht-
Gradwohl technic and by the technic of absorbing out their
natural antisheep amboceptor.
2. HISTORICAL REVIEW
In 1901 Bordet and Gengou described a method for detecting the
presence of specific antibodies in the serum by means of complement
fixation. Five years later the principle of this method was success-
fully applied by Wassermann, to the serum diagnosis of syphilis.
Widal and Le Sourd (2) appear to be the first who used the comple-
ment fixation reaction in attempting to arrive at a more certain method
of diagnosing tuberculosis. Bordet and Gengou (8) in 1903 demon-
strated the presence of antibody capable of uniting with tubercle
bacilli and fixing complement in the sera of tuberculous animals.
Wassermann and Bruck (4) in 1906 also demonstrated the presence of
an antibody to tuberculin in patients treated with tuberculin.
|
;
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 163
Caulfield (5) and Beattie (6) in 1911, using bacilli emulsion as anti-
gen, obtained 33 per cent positive reactions in primary tuberculosis, 70
per cent in moderately advanced cases and 62 per cent in far advanced
tuberculosis.
Deilman (7) in 1911, using carbolized emulsion of tubercle and
other acid-fast organisms as antigen, obtained fixation with tubercu-
lous serum, the other acid-fast organisms giving about the same results
as the tubercle bacilli.
Laird (8) in 1912, using a watery emulsion of tubercle bacilli, obtained
fixation in only 4 out of 34 cases.
Mollers (9) in 1912, concluded that the fixation reaction permits no
diagnostic or prognostic conclusions.
Hemmer (10) in 1912, using old tuberculin as antigen, obtained
100 per cent positive reactions on 48 tests of tuberculous cows and only
4 per cent non-specific reaction with non-tuberculous cows. He also
reports (6) 97 per cent positive results with human tuberculous sera.
Zweig (11) in 1912, using a bacillen emulsion as antigen reported
that his fixations were proportional to the severity of the disease.
Calmette and Massol (12) in 1912, using watery and dialyzable
extract antigens reported fixation in 92.5 per cent of their cases.
Much (13) in 1912, using various acid-fast bacteria as antigens,
with sera from tuberculous and healthy persons, obtained fixation in
77 per cent of the healthy cases, in other words, a large number of non-
specific fixations.
Letulle (14) in 1912, using Calmette’s antigens, obtained 89 per cent
of fixation in tuberculous cases.
Fraser (15) in 1913, using various antigens, found that 96.6 per cent
of sera from normal individuals gave no fixation with antigens made
from living bacilli, and that with this antigen 42.3 per cent of sera
- from tuberculous individuals gave positive results. She states that
the most reliable antigen is prepared from living human bacilli, and
thinks that the complement-fixation test made with living bacillary
antigen is of more value in the diagnosis of tuberculosis than any
other reaction thus far discovered.
Dungeon, Meek and Weir (16) in 1913, report 85 per cent positive
reactions with tuberculous patients not given specific treatment.
Tuberculous patients treated with tuberculin 100 per cent positive,
arrested cases 75 per cent positive. They conclude that killed tubercle
bacillus emulsion makes the best antigen.
164 HASSOW O. VON WEDEL
Bank and Anderson (17) in 1918, using emulsions of killed tuber-
cle bacilli as antigen, obtained a marked per cent of strong positive
reactions.
Ammand (18) in 1913, using crude tuberculin, obtained only 4 per
cent definitely positive reactions while with a peptone soluble antigen
they obtained 92 per cent definite reactions.
Wyschellesky (19) in 1913, using an emulsion of tubercle bacilli and a
solution of the bacilli in 2 per cent lactic acid, obtained 18.1 per cent
positive results with tuberculous cattle and also obtained 9.7 per cent
positives with healthy cattle.
Kinghorn and Twitchell (20) in 1918, using a bacillus emulsion as
antigen, reported 37.5 per cent positive in the incipient stage, and 93
per cent in advanced tuberculosis, with no fixations in normal cases.
However, they tested only 33 cases altogether.
Rothe and Bierbaum (21) in 1913, reported that they did not obtain
strong fixation with tuberculous cattle before treatment except in a
few cases.
Harris and Lanford (22) in 1913, attempting to differentiate acid
fast bacilli by means of the complement fixation test concluded that
regardless of the various methods used to produce these sensibilizators
no clean-cut specificity for complement fixation was found for the acid
fast bacilli which they made use of in their experiments. They pro-
duced their anti-substances by injecting rabbits with whole bacilli or
extract of the bacilli.
Momose (23) in 1913, using bodies of tubercle bacilli, after extrac-
tion of the fats, as antigen, obtained 100 per cent positive reactions
with all tuberculosis patients and in nearly all exposed to the disease
as well as 50 per cent in healthy persons.
Besredka (24) in 1914, using his special antigen, reported that all
first stage tuberculosis patients react positive and nearly all second
stage cases, but that in the third stage often partial or negative reac-
tions were obtained.
Wwednesky (25) in 1914, using various tuberculins as antigen, re-
ported 82.8 per cent positive and 16.2 per cent doubtful reactions in
tuberculous cases. Ten non-tuberculous cases were negative.
Debains and Jupille (26) in 1914 using Besredka’s antigen reported
that the reaction is very sensitive in all forms of tuberculosis except
miliary and meningeal and absent in healthy non-tuberculous people.
Kuss, Leredde and Rubinstein (27) in 1914, using Besredka’s anti-
gen, reported 89 per cent positive reactions in well-developed cases of
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 165
pulmonary tuberculosis, about 66 per cent in mild cases and negative
in all normals except those giving positive Wassermann reactions.
Inman (28) in 1914, using Besredka’s antigen, reported 95 per cent
positive reactions in 100 cases of pulmonary tuberculosis and 24 per
cent positive in non-tuberculous patients.
Bierbaum and Berdel (29) in 1914, using bovine old tuberculin as
antigen with the serum of 120 slaughtered cattle stated that the autopsy
findings and serological findings agreed in only 65 per cent of the cases.
McIntosh and Fildes (30) in 1914, using fresh living tubercle bacilli
as antigen, reported 76.7 per cent of positive reactions in pulmonary
tuberculosis, 80.7 per cent in surgical tuberculosis (not glands) and no
positives in the controls.
Radcliffe (31) in 1914, using the same antigen as McIntosh and
Fildes, reported about 85 per cent positives with tuberculosis cases.
Dudgeon, Meek and Weir (32) in 1914, using an alcoholic extract of
tubercle bacilli, reported 89 per cent positive reactions with tuber-
culosis cases when repeated examinations were made.
Meek (83) in 1914, stated that the greatest amount of antibody was
found in severe cases with extensive lesions. Cases similar clinically
may give different reactions.
Bronfenbrenner in 1914, in several communications (34, 35, 36, 37)
on the use of Besredka’s antigen, reported 93.84 per cent positive
reactions in active tuberculosis. He stated that Besredka’s antigen is
specific, and syphilitic and tuberculous antibodies occurring in the
same patient’s serum are distinct and separable. Forty-three per
cent of syphilitics gave positive reactions with Besredka’s antigen,
indicating an undue prevalence of tuberculosis in this class of patients.
Craig (38) in 1915 concluded that complement binding bodies are
present in the blood serum of both active and inactive tuberculous
infections. His polyvalent antigens prepared from several strains of
tubercle bacilli have been found by him to give excellent results in the
complement fixation test for tuberculosis. He obtained positive results
in 96.2 per cent of active tuberculosis and 66.1 per cent in clinically
inactive cases. The tests were negative with all his normal individuals,
and with patients suffering from other diseases, with the exception of
two patients infected with syphilis in whom symptoms of a coincident
tubercular infection were also present.
Stimson (39) in 1915, using a variety of antigens, reported a small
number of cases with but fair results. Corper (40) in 1916, using an
autolysate as antigen and also a bacillary emulsion antigen, concluded
166 HASSOW O. VON WEDEL
that the complement fixation test for tuberculosis is not absolute, being
positive in only about 30 per cent of all clinically definite cases of tuber-
culosis, both active and inactive.
Miller and Zinsser in 1916, in a communication to the New York
Pathological Society, reported 100 per cent positive results in active
eases and 100 per cent negative results in non-tubercular controls.
In a subsequent communication (41) they reported 98.5 per cent posi-
tive reactions in active cases of tuberculosis and 32 positive fixations
with 140 doubtful cases (i.e., patients suffering from diseases clinically
diagnosed as other than tuberculosis). In a still later communication
(42) on the clinical value of complement fixation in tuberculosis, Miller
reported 96.8 per cent of positive results in active cases, 100 per cent
negative results in non-tubercular and normal patients, and about 90
per cent negative results in inactive cases.
Woods, Bushnell and Maddux (48) in 1917 employing partial anti-
gens (i.e., alcoholic extract antigens prepared by disintegrating the
bacilli with 1 per cent lactic acid, filtering and extracting with alcohol)
obtained positive results with 90 per cent of sera from cases classed as
incipient, 87 per cent from active cases and 92 per cent from advanced
tuberculosis cases.
McCaskey (44) in 1917 stated that specific complement-binding
bodies were present in the blood of tuberculous patients, but not
constantly so, even in clinically active cases; on the other hand, they
may be present in cases having no clinical manifestations. These
bodies, when present, may be demonstrated by the usual complement-
fixation technic, and proves the existence of a focus which is patholog-
cally active. A negative fixation test does not absolutely exclude
clinically active tuberculosis. The results of the tests in which tuber-
culins or bacillary suspensions were used as antigens are probably as
dependable as the subcutaneous tuberculin test, and removes that
element of danger to the patient which may be caused by the latter
test. When the blood of the patient gives positive reactions with the
tuberculosis complement-fixation test, and the Wassermann test, both
tuberculous and luetic foci are present.
Brown and Petroff (45) in 1918, in a study correlating clinical and
laboratory experience, found the test of greater value to them as a con-
trol of the therapeutic regimen than as a diagnostic measure. It
parallels the subcutaneous tuberculin reaction in that a negative reac-
tion in the tuberculous individual is of more value than a positive one
in determining which patients need treatment. They have found the
ee, Eee Se Te eee eee ee
’
EE
ae 5g Sao
;
ae a
Se
a
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 167
reaction of value, too, in pre-determining which patient will be bene-
fited and which harmed by exercise and activity.
Lange (46) in 1918 reported fixation of some degree in 5 per cent of
tuberculosis cases with 12.6 per cent in non-tuberculosis cases. He
examined 864 serums with four different antigens, including Miller’s
and Petroff’s potato broth culture antigen.
Stivelman (47) in 1918 reported on a series of 205 cases, 22 of which
were non-tuberculous. Using Miller’s antigen, he found that about
50 per cent of the tuberculous cases, active and inactive, gave positive
fixations. He was unable to corroborate the favorable report of Miller.
Lewis (48) in 1919 has critically studied the reaction and has come
to the conclusion that certain inherent defects will limit its usefulness.
He suggested some modifications of the technic, however, such as
increasing the time of fixation and increasing the quantities of com-
plement and antigen in the effort to overcome the merely transient
binding of complement that has possibly been interpreted by some
workers heretofore as true deviation. It was his impression that it is
unsafe to apply the reaction to the diagnosis of tuberculosis unless as
a matter of confirmation of a clinical decision.
Pritchard and Roderick (49) in 1919, reported 69 per cent of posi-
tive reactions in active moderately advanced cases and 16 per cent of
reactions in cases not proved to be tuberculosis. They thought that
this test was a great aid in differential diagnosis.
Cooke (50) in 1919, concluded that in tuberculosis the serum con-
tains complement binding substances that gave fixation in about 87
per cent of his cases.
Stoll and Neuman (51) in 1919, concluded that, from their experi-
ence, it would seem that with suspicious symptoms and suggestive,
yet with inclusive signs, a negative fixation test, using the method
described in this study, increases to a considerable degree the prob-
ability of the non-tuberculous nature of a given case. With the same
symptoms and signs a persistently positive reaction probably signifies
an active tuberculosis. A positive reaction occurring with neither
symptoms nor signs does not justify a diagnosis of active tuberculosis,
though it is quite probable that there has been an active process re-
cently. In such a case, roentgenoscopy should be employed and the
patient observed for several months.
To show how the opinion of laboratory workers in regard to this
complement fixation test is divided even today, one needs only mention
Mourseend’s (52) conclusions as written in 1920. He stated that the
168 HASSOW O. VON WEDEL
complement fixation test for tuberculosis as described in his article is
of no value as a diagnostic or prognostic aid, that this complement
fixation test with alcoholic extract of tubercle bacilli as antigen is not
specific and that a large percentage of serums giving a positive Wasser-
mann give fixation with tubercle bacillus antigens.
Mourseend used a methyl alcoholic extract of the tubercle bacilli
containing all the alcohol soluble fats along with other alcohol soluble
substances; therefore, it is not to be wondered at that he, like Corper,
should get non-specific cross-fixations with syphilitic sera.
TABLE 1
Results of experiments by different workers
SUSPICIOUS wonseoeoe eae ab: Ne pase NORMALS
SIS vancen | 4DVANCED | pinions
‘nen es oa ee Peo | en ce aka ror yc
g/22| 2 | 83/2) 23/2 | #2] g |£8] g | 82
6G") S| a" / 5 [a* |S [ao] 6 lao] Ss la”
Caulhieldeeesoeteciesce 33.0 70.0 63.0
Radcliffe........ 88.6 89.6 79.0 204! 0
IM BIL seen ee eee 140/22.8] 32 |100.0} 110/98.0 | 83] 98.0} 45] 4.4] 144] 0
MIYINVAT etter: rue cate oe 50/60 .0 100/95 .0 100/24.0
Webaimsageces ie 90.3 81.3 17.3 3.2
Bronfenbrenner...... 50/72 .0 65/93 .8 375} 8.0
Craigie sik eee 30 | 96.7] 61/98.3 | 54 | 96.4} 450} 4.4) 200] 0.5
Mc Cake. ocii5 «+42 8/25 .0 36/77 .7 74/20.2) 9/11.1
TRH AO Gon ROR eee 20/65 .0} 64 | 81.2) 123/91.0 | 2 |100.0} 14] 7.0) 14) 0
(GOreroner esos isc cic 69/43 .6] 28 | 35.7) 61/70.5 | 63 | 61.0} 31/19.3
IME OOT ee oss lave css 61/60.0| 24 | 87.5) 49/85.7 | 83 | 84.3] 23/26.0) 100/12.0
Calmettenversce ss... 134/92 .5*
Dud geonseeeaes. sei: 234)89 .3*
MelIntosh, 26%). .:..:.. Zonda 87] 3.5
* Various stages grouped.
Moon (53) has tabulated the experiments of the more important
studies according to the general broad classification of tuberculosis
patients. His summary is given in table 1.
Bronfenbrenner (54) in 1917 suggested that at least two reasons for
the failure of advanced cases to give fixation can be offered tentatively;
one is that the resistance of the patient having been exhausted, there
is no new antibody formation; and the other, that the circulating anti-
body is taken up as formed by the combination with antigen which
may greatly increase during the last stages of the disease.
Sa ee ee
— - “
ae.
=.
—_=
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 169
Boez and Duhot (55) in 1919 said that if one considers the various
stages of pulmonary tuberculosis, the curve of the antibodies, at first
low, rises during the first and second period; that it is maintained or
increased at the beginning of the third period; and at the ultimate
phase, the antibodies can disappear in a rapid manner, in coincidence
with the progress of the cachetic premonitories of death.
Depending on the antigen and the technic used, the percentage of
tuberculosis cases giving positive fixations vary from about 95 per cent
down to a low figure. While the antigens and technics giving the
highest percentage of positive results are more valuable in confirming
. suspected cases or in detecting unsuspected ones, they tend to approach
such tests as the von Pirquet in failing to give information as to the
activity of the tubercular process and are therefore misleading as
suggesting in any degree an active process.
It seems to be generally conceded that a large percentage of human
individuals have had some tuberculous lesion during life. Many, how-
ever, become quiescent; all traces of the bacilli even disappearing.
Those antigens and technics giving a lower percentage of positive
findings are of more value in that they give positive evidence of the
activity of the disease with practically no non-specific reactions.
The most significant feature of this summary is that, while there are
differences in percentages due evidently to the different methods and
reagents used, there is agreement that complement fixation under
proper conditions gives positive results in the majority of cases of
active tuberculosis. The laboratory technic employed in this test is
the same in principle and in main details as in complement fixation
applied to the diagnosis of syphilis. As in that test there have been
many variations in the technic and in the reagents employed, and as
would be expected, the results have also varied. The widest variation
occurs in the preparation of the tubercle bacillus antigen. The reac-
tion is one of biologic specificity depending on the presence in the pa-
tient’s serum of free antibodies specific to tubercle bacilli. In this
particular, the situation differs from that in syphilitic infection.
3. TECHNIC OF THE STANDARD COMPLEMENT FIXATION TEST FOR
TUBERCULOSIS USED IN THIS STUDY
The technic employed was similar to that originally used by
Wassermann with the following modifications.
At first the tests were carried out in both one-quarter and
one-tenth the original Wassermann volume, but as the writer
17
70 HASSOW O. VON WEDEL
found no difference in the results, he has since continued to use
the one-tenth Wassermann volume only.
Complement. The pooled blood serum from six to ten healthy
guinea-pigs was used as complement; in addition, we used serum
from separate guinea-pigs untested for its complement fixation
value; serum from separate guinea-pigs after having been tested
for complement fixation value and serum from six to eight
guinea-pigs, all of which had been specially tested for comple-
ment fixation value. All these complements were titrated with
a 2.5 per cent sheep cell suspension, sensitized with three units
of antisheep amboceptor; the unit was recorded at the end of
fifteen minutes. Exactly two units were used in the regular
test. Many different balances of the hemolytic system have
been tried but the most constant results have been with the two
hemolytic units of selected complement combined with cells
sensitized with three standard units of amboceptor.
Antigen. The Wilson antigen (56) which was used as the
standard antigen throughout this study is a suspension in 0.9
per cent saline of dried bacilli, from which all constituents soluble
in alcohol and ether have been removed. The bacilli were
grown in glycerin-broth. The monovalent antigen cultures
were grown for three months. The broth cultures were killed
by heating them in the Arnold sterilizer for one hour. The
cultures were then filtered through filter paper. The filtrate
was discarded, and the residue was placed in absolute alcohol, in
the proportion of one volume of residue to ten volumes of alcohol.
This mixture was shaken thoroughly by hand, and was placed
in the ice box for two weeks. It was then filtered through paper
and the filtrate was discarded. The residue was washed in
absolute alcohol and the sediment obtained by centrifugalization
was washed in ether. After further centrifugation the ether
was discarded and the centrifuge tube containing the residue was
plugged and placed in the dark at room temperature over night.
By this simple procedure the residue was dried within twenty-
four hours. The dried powder was emulsified in a large mortar
with 0.9 per cent saline in the proportion of 1 gram of powder
in 200 cc. of saline. This gave a concentrated emulsion con-
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 171
venient for storing as a stock antigen. ‘The emulsion was heated
for one hour at 80°C. The antigen was then ready for use, and
it was standardized to be used in such a dilution that 0.1 cc.
contained two standard fixation units and one-fourth, or less, of
the anticomplementary dose. The unit was determined by
titrating varying amounts of the antigen with 0.01 cc. of a known
positive tuberculosis serum, and two hemolytic units of a com-
plement known to be potent for tuberculosis fixation. The
standard dilution of the antigens employed is usually 1:50.
This makes a final dilution of dried bacilli 1:10,000. These
antigens are not anticomplementary in the amount used in the
test. They have given uniform and constant fixation reactions.
Sheep cells. A 5 per cent suspension of sheep cells, which had
been washed five times in sterile saline was used, after having
been sensitized with an equal volume of diluted amboceptor in
the water bath for half an hour.
Amboceptor. Three units were used in the tests.
Fixation period. After the patient’s serum, complement,
antigen and saline were mixed, the mixtures were incubated in
the water bath at 37.5°C. for one hour. The sensitized cells
were then added and the reading was made in exactly fifteen
minutes.
Results were reported as plus minus if any degree of fixation
was observed; 1 plus if there was marked fixation in the first
antigen tube; 2 plus if there was complete fixation in the first
tube; 3 plus if there was complete fixation in the first tube and
marked fixation in the second tube; and 4 plus if there was com-
plete fixation in both tubes.
During 1918, Cyrus W. Field and the writer carried out a
series of 730 Wassermann reactions in the Bellevue Hospital
Laboratories, using the regular amount of serum prescribed by
Wassermann and also twice, three times, four times and five
times that amount. These amounts of patient’s serum were
tested in all of the 730 cases, our controls being carried out with
double the amounts of serum used with the antigen. Discard-
ing all those cases that were anticomplementary in the regular
Wassermann amounts, and considering only those which ordi-
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 2
172 HASSOW O. VON WEDEL
narily would be considered as not anticomplementary, we found
that we had no anticomplementary and no non-specific reactions
with double the usual amount of serum. Three times the usual
amount of serum gave about 1 per cent of anticomplementary
reactions, four times the usual amount of serum gave about 5
per cent of anticomplementary reactions and five times the usual
amount of serum gave about 25 per cent anticomplementary
reactions.
As these results were so favorable, and as several other inves-
tigators have made favorable reports on the use of larger quan-
tities of patients’s sera, the writer has made all his tests since
January 1, 1918, with the regular Wassermann amount and with
double that amount of patient’s serum; that is, 0.04 ec. of serum
in the first antigen tube with 0.08 cc. of serum in its control
tube; 0.02 cc. of serum in the second antigen tube with 0.04 ee.
of serum in the control tube and 0.01 cc. of serum in the third
antigen tube.
However, it is perfectly possible to obtain approximately the
same results by either of the two following methods:
First, using the regular amounts of patient’s sera and antigen
containing four antigenic units. Second, by using double the
regular amounts of patient’s sera and antigen containing two
antigenic units.
The writer has found that if his antigen has a relatively small
range between its fixation and its anticomplementary dose, it is
best to double the amount of the patient’s serum, and use only
two antigenic units. If, however, one has a wide range with
the antigen, the regular amount of patient’s serum and four
antigenic units can be used.
While at the Walter Reed Army Hospital, the writer experi-
mented with the Noguchi system, using human red blood cells
as the indicator instead of the sheep cells in order to do away
with the troublesome feature of natural anti-sheep amboceptor.
It was soon discarded, however, as the amboceptor was of a.
very low titer. This necessitated the use of a large amount of
complement in the working system. Most of the results were
negative with this system due probably to the use of this excess
ee
re
a
a oe ee
Pepin mal
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 173
amount of complement, as the smallest excess of the regular
amount of complement used very markedly reduces the percent-
age of positive findings. In fact, the complement factor appears
to be the most important point to watch in making the comple-
ment fixation test for tuberculosis. Apparently the, comple-
ment fixation in this test is very much weaker than when one
uses the lipoid antigens for the Wassermann test.
4, DETERMINATION OF THE OPTIMUM TIME AND TEMPERATURE OF
FIXATION
During this study, tests have been made to determine the
optimum time and temperature of the complement fixation
period. In all, 135 comparisons at various fixation periods have
» been made as follows: One hour at 37°C. in the water bath; two
hours at 37°C. in the water bath; two hours at 10°C. in the ice
chest followed by two hours at 37°C. in the water bath; four
hours at 10°C. in the ice chest; and over night at 10°C. in the
ice chest.
The results of a few of these comparisons are given in table 2.
Apparently the one hour water bath fixation period gives the
most uniform results. The ice box fixation for longer periods
of time gave very weak results. Two hours in the water bath
gave marked increased fixation in a considerable number of
cases, but there were more anticomplementary reactions than
with the one hour fixation period. Two hours at 10°C. in the
ice chest followed by two hours at 37°C. in the water bath gave
about the same results as the two hour 37°C. fixation.
The above findings in regard to the weak fixation of tubercle
bacillus antigens at ice box temperature are in variance with the
findings of Ruediger (57) who investigated the optimum time
and temperature for Wassermann fixations. He reported that
fixation at 1°C. for twenty-four hours gave the strongest positive
results when compared with fixations at various temperatures
from 2° to 37°C., and from one hour to twenty-four hours.
He concluded that it is advisable to warm the serum-com-
plement antigen mixture before adding the sensitized blood
corpuscles.
174 HASSOW O. VON WEDEL
_ TABLE 2
Showing variations in the results of a few of the tests obtained by fixation at various
times and temperatures
TWO HOuRSS§ aT 10°C.
o,
AND 2 HOURS AT 37°C, | FOUR HouRs ar 10 Cc.
ONE HOUR* aT 37°C. | TWO HOURS AT 37°C.
serum | Control % 3 Control 3 3 Control 3 i) Control | 8 3
NUM- tubes So DA tubes Be > tubes ere p tubes 3 a >
sa ° ° 25 ca 3 rs) KE Bo 3 3} ee So rs) ° FE cy
é |S |<2/35] 2 | £ )x3/28) 2 |S |e /88] 2 | 2 pala
s sia a — os |N om) So o ia iam So coin ex}
yi fae) (eee comin Mee ree [mer be a | ee ee ee eee ee eee.
Ne) = | — | — | — | =) ee) ee
206 | — | — | 4+) 4+} — | — | 44} 44) — | — | 44) 44} —] — | agt at
28 |—-|-—/3t+/+]—-]-/3+}/+]/-]-]34/+]-]-]4+]-
213 —}—/2+)/ + ]/—] — |] 44+) 2+) — |] — |] 44) 2+) —] —J] 4+] -
SAWP 22 4h ASS ea AMS a el ae ae
216 | — | — | 4+) 24+] — | — | 44+} 3+] — | — | 44) 34] — | — fod}
219 | —|— |] 4+) 44} +] | 44) 44) +] 4 | 44} 44] — | — | 44] 4g
220 —/}—|3+/+)—-—] — | 44+) 2+) -—} — | 34+) 2+) —} —|]+]-—-
221 Sl i Th el een ica ce Pe fe
231 | — | — | 44/34] — | — | 44) 3+) — | — | 4+) 34) — | — |] 34] a+
233 |—|—|}—-{/-|=|/-/4)-1-]-)/4)-]-)=}/-12
35 }—|—|—|-|/—|)-—hele]-)-i44=1-—| ee
246 +} a4] 44+) 2+) +} — | 44+) 44+) + | — | 44+) 2+) +] + | 44) 4+
247 —;}—}]2+) + }]—] — | 4+) 4+) -— | — | 34+) 24+) —|} —J] -—-J] —
4o7 |—|—|24+/+]4]—-—]2+}/4+]4+]—-]44/4+]/-]-]4]4
409 |—|—|3+/4]/4]—]4+)2+/-]-—]34}4+]—]-—]2t}4
* Refers to the fixation time in the water bath at 37°C.
} Refers to the use of double the regular Wassermann amount of patient’s
serum.
t Refers to the use of the regular Wassermann amount of patient’s serum.
§ Refers to the fixation time in the ice box at 10°C.
5. THE STANDARDIZATION OF COMPLEMENT
M. A. Wilson (56) has observed that the serum of a number
of guinea-pigs is unsuitable for use in the complement fixation
test for tuberculosis because the complement of these guinea-pig
sera is not fixed, under the usual conditions of the test with the
sera of tuberculous individuals.
A similar irregularity in the guinea-pigs’ sera, when employed
in the Wassermann test, was reported by Noguchi and Bronfen-
brenner (58) in their article on the ‘‘ Variation of the complement
activity and fixability of guinea-pigs’ sera in Wassermann work.”
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 175
They state that positive patient’s sera will often fix complement
from some guinea-pigs and will not fix complement from other
guinea-pigs, but that there is no definite relationship between
the complementary activity and the fixability of a given speci-
men of guinea-pig’s serum. However, the irregularity observed
by the latter authors was relatively infrequent as compared
with that reported by M. A. Wilson.
With the purpose of further studying this phenomenon, the
writer carried out a number of tests with sera from actively
tuberculous patients; at first, in two series, one with pooled
guinea-pigs’ sera that had not been tested as to fixability, the
other with pooled guinea-pigs’ sera, each of which had been
separately so tested and found satisfactory. The pooled serum
was derived, in each case, from six to eight guinea-pigs. These
parallel tests gave practically the same results. In a few in-
stances, the writer obtained a 2 plus reaction with the tested
complement instead of a 1 plus reaction with the untested com-
plement, or a 3 plus instead of a 2 plus reaction; but this differ-
ence was not regularly encountered and in a few instances, in
fact, better results were obtained with the pooled untested com-
plement than with the specially tested complement.
This experiment being inconclusive, the writer then carried
out a series of tests with sera from seven frank tubercular cases
(six of which he had already examined) testing each guinea-
pigs complement separately against each positive serum. The
results of these tests are shown in table 3.
Serum 1, tested with complement from guinea-pig 1, gave
negative results in the antigen tubes containing 0.02 cc. and
0.04 cc. of patient’s serum on the first and fourth days after tak-
ing the specimen from the patient. On the sixth day and again
on the seventh day, complement from guinea-pig 2 gave a 3
plus reaction with 0.02 ce. of patient’s serum and a 4 plus reac-
tion with double that amount. Tests made on the ninth and
eleventh days with complement from guinea-pig 3 gave nega-
tive results with all amounts of patient’s serum. On the four-
teenth day, the pooled complement, made from sera of six tested
guinea-pigs, gave results which were practically the same as
176 HASSOW O. VON WEDEL
those obtained when complement from guinea-pig 2 was used;
that is, a 2 plus reaction with 0.01 cc. and 0.02 cc. of patient’s
serum and a 4 plus reaction with 0.02 ce. and 0.04 ce. of patient’s
serum. On the sixteenth day, complement from an additional
guinea-pig which may be designated as no. 4 gave a plus-minus
and a 1 plus reaction.
TABLE 3
Showing variations in flexibility of the complement of different guinea-pigs’ sera in
the complement-fixation test in tuberculosis
POOLED
parrenn’s — |.SOMPLE | ‘tmnerg | COMPLE: | ‘armen | COMPLE’ | tmwnt | Counre.
Se pe Prue TEee LATER TEST| 12S? TEST ae SUSE aualas Lee en wanes eee
1 —*| —j|/ + | — | 44) 2+] 44+] 84+ — | — | — | — | 44 2+
2 — f= | — | =) eps 44) 3+) 14+) — | — | — |) eee
3 ele |
4 —|—j} —| — | 44) 44+] 44) 44+) 14) -— 4+) 3+
5 2+| — | 3+] 14+) 44+) 44+] 44+] 44) 44) — | 44] — | 44] 44+
6 =f — | | 4 se EE | |) = ee
7 — | — | 2+] — | 44+) 44+) 44) 44+) -—]|] -—|] -]| - 3+| 1+
* The results in the first column were obtained with 0.04 cc. of patient’s serum.
+ The results in the second column were obtained with 0.02 cc. of patient’s
serum.
t Complement was preserved in the interim with an equal amount of 18 per
cent salt solution and kept in contact with ice.
Patients’ sera 2, 3, 4, 6 and 7 gave almost identical results.
Serum 5 gave a 1 plus and a negative result with complement
1, a 4 plus reaction with complement 2, negative results with
complement 3, a 4 plus reaction with the pooled complement,
and a 1 plus reaction with complement 4, when the regular
amount of patient’s serum was used.
These observations seem to confirm the statements of M. A.
Wilson, Noguchi and Bronfenbrenner on the variation in com-
plement activity and fixability of guinea-pigs sera, but the use
of pooled complement from six to eight guinea-pigs as a rule
apparently overcomes this variability of separate guinea-pig sera.
In studying the difference in the first and seventh day tests
and also in studying the loss of natural anti-sheep hemolysins in
the patients’ serum due to ageing, the writer used a salted com-
a eee eee eee
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 177
plement. A sufficient quantity of this salted complement was
made up to last for all these comparative tests so that no differ-
ences in the complement would complicate the results. After
studying the various methods of preserving complement includ-
ing Ronchese’s (59)—sodium fluoride method, Thompson’s
(60)—sodium chloride method, Rhamy’s (61)—sodium acetate
method, and MeNeill’s (62)—freezing method, the writer con-
cluded from many experiments that simply adding 8.5 per cent
of dry powdered salt to the pure guinea-pig serum, and keeping
the tube (not necessarily sterile) in a thermos bottle packed with
plain cracked ice, preserved the complement for at least one
month without the slightest loss of complementary activity.
6. DETERMINATION OF THE THERMOLABILITY AND THE THERMO-
STABILITY OF ANTI-SHEEP CELL HEMOLYSINS IN
HUMAN SERA
To determine whether or not the natural anti-sheep hemolysin
in human sera is thermolabile or thermostabile, the writer has
made hemolysin titrations on 187 different sera. These have all
been titrated after being heated in the water bath at 56°C. for
various periods of time. All received the initial thirty minutes
inactivation and were then reheated for fifteen, thirty, forty-five,
sixty or one hundred and twenty minute periods in different
lots. On 97 of these sera, these heat tests were repeated after
keeping the sera for seven days under sterile conditions. These
kept sera were inactivated the first day for the regular one-half
hour period at 56°C., and were then stored in the ice box. On
the seventh day the titrations were repeated after reheating the
sera In separate tubes for various periods of time as in the pre-
vious titrations. The volumes of serum used in most of these
titrations were 0.01 cc., 0.02 cc., 0.04 ec., 0.06 ec., 0.08 ec., and
0.1 cc. At first, as high as 0.16 cc. was used but the volumes
above 0.1 ce. were so frequently anticomplementary that later
only from 0.01 to 0.1 cc. were used in the titrations. These
volumes gave a wide range and proved very satisfactory. The
complement used in these titrations was all tested for natural
178 HASSOW O. VON WEDEL
anti-sheep amboceptor and any complement showing the slight-
est trace of hemolysis was discarded. All titrations were set up
with two units of complement. The complement volume neces-
sarily varied as one day the serum volume of the one unit of
complement would be 0.04 cc. and possibly the next week it
would be 0.06 ec. Also the 0.06 ce. might have been just exactly
one unit while the 0.04 cc. was possibly a strong one unit (1.2
units) as the complement titrations were regularly made with
0.01, 0.02, 0.03, 0.04, 0.05 and 0.06 ec. volumes. The unit
therefore would be somewhere between 0.03 and 0.04 ce.
One-tenth of a cubic centimeter of a 5 per cent suspension of
washed sheep cells was added to each tube. Cells from a dif-
ferent sheep were used each week. Consequently, some were
more resistant to hemolysin than others. These variations in
the blood cells and the complement probably account for some of
the apparent discrepancies noted in the results of the first and
the seventh day titrations.
In going over the results of these titrations one at once notes
the marked loss of hemolysin in the heated sera. This loss
seems to be progressive; i.e., the longer the sera are heated the
greater is the loss of hemolysin. In table 4, eight sera show no
hemolysis in the tests after being reheated one hour at 56°C.;
four show a lesser degree of hemolysis and three contained such
an excess of hemolysin that a loss was not indicated by the tests.
In tables 5, 6, and 7 this loss of hemolysin is also noted to a
more or less degree. Even fifteen minutes extra heating caused
‘a slight loss of hemolysin. In table 6 the sera kept for seven
days seemed to show a greater loss of hemolysin due to heating
forty-five minutes than the fresh sera heated the same length of
time. Table 7 does not show any great loss of hemolysin as the
extra period of heating was only fifteen minutes. However,
heating at 56°C. even for two hours does not destroy all the
hemolysin present in human sera. In table 4 we see that sera
nos. 59N, 70N, 71N, 77N, and 90N still show hemolysis in the
tests to a greater or less degree after the two hours heating.
This seems to point to the presence of a thermostable hemolysin
which may cause complete hemolysis if a sufficient quantity of
complement is present.
“UOTPEXY ojoTdur09 “Jo SuOIVexY Buo1ys “Js /UoIZVxY Jo oovs4 ‘Jy ‘Bou 07 siojor BAIWVBAN' : ALON
179
3} Jj} | dou | Jou} 39 yo dou | dou | dou | Jou yu J4\ | NOG
js dou | dou | dou | jy jh dou | Zou | Zou | Jou ju js | NZZ
a ju ju ji ju | JM jM dou | Zou | Zou | Jou ja JM | NIL
; 5 js js | ju] ju] yu | Js ju] ju | |] Bou] 9 | JM | Nos
8 yo zo Jo OER i yo jy Jo 38 jo 9 2 | NZ9
oO ja J} | dou | dou |] #9 yu dou | dou | Zou | Zou | Zou | Zou | NEG
e jo yo go> 90 y J? Jo qd Jo jo jo jy Jo Jo J2 | TOTS
eal yo jo jo | jo yo yo yo jo jo jo J} | dou | Zou | Zou J} | 660°
fe yo yo jo | jo yp jo yo yp 5) 5) Zou | Zou | Zou Bat JM | 960E
2 yo jo jo | jo yo jo yo yo jo jo Zou | Zou | Zou | Zou jt | 260¢
2 ge jo ba j9 Jo yp jo jo Jo jo ye jo id jy | 680¢
je Jo | so jo Jo Jo Jo jo jo yuu 4 yM ee J? | 080¢
jo- yp jo | jo yp yo jo yo yo jy sou | Zou | Zou sat jt | 290¢
BR yo yo jo | jo yo jo jo jo jo jo sou | Zou | Zou | Sou jo | T¢o¢
= jo yo jo | jo yo jo jo jo jo jo | Zou | Zou J} | Sou | Zou 19 || ‘Geoc
fo) sou | Sou | Zou | Zou | Sou J} | dou | Zou | Zou | Zou | Bou | Zou | pEop
a gou | dou | dou | dou | Zou | Zou | Zou | Bou | Zou | Bou | Zou | Bou | QG0F
rd sou | sou | sou yi yh jo | Sou | Zou | Sou | Zou ju jo | 60F
Ee te ieee We Os ica 0 sea ol) S94) S| Pe] gs |-F00g
= jo 8 ~~ yx yu Js Js yaa 4 4 4 JM | 2607
a sou} J9 | Ju] Jw] Ju] Js | dou | ju | JX | 4s | 6807
Ss sou | dou | Sou | Sou | Zou | Zou | Zou | Zou | Zou | Zou | Sou | Zou | gop
= Sou | sou | dou | Zou | Zou | Zou | Zou | Sou | Zou | Zou | Bou | Zou | ggop
By
Ss “09 OT'0 | “99 80'0 | 99 90°0 | “9 $0'0| “99 ZO" | “99 10° | 99 OT'0| “99 800} *99. 90'0| 99 $0'0| “99 Z0'0| °° T0'0| *92 OT'0 "29 80°0 | “99 90°0| *99 400] °99 20°0 | "99 100) aa
a GT aS ee RR ad ee me nes Seman a || SE eas ETS) |} Te MME Op Ra Re See, La a es
. "0.99 LY SYQ0H OME ‘0.9¢ LY YO0H ANO ‘O.9¢ LV SALONIW NAL . KOAES
YIDQ 12]DM 9Y} U2 *O,9g 10 sporsad Buo) sof Buryvay 0} anp urshjoway daaysyun yoinzou fo sso) ay) Buunoyy
* WIGVL
VON WEDEL
HASSOW O.
180
jo
jo
*09 O10 “99 100 | °99 OT'0 | “99 80°0 | .99 90'0 | °° #0'0 | “99 20'0 | “99 T0°0 | “99 OT'O | “99 80°0 |°29 90°0 | “99 F0'0 | “99 20°0| 99 TOO
‘O.9¢ LY SHLONIAW GAIM-ALHOA GALVAILOVNIGU
‘u0178xy a}a[duUIO9 ‘Jo puw ‘uoTVexY BZuosys ‘Js uoTyexY Yeom “JA SUOTZEXY Jo vovI} “Jy SoAtyeZou 09 Siojor Bay f
‘6I6I ‘I ySnSny uo opeul o19M 84804 oseyy, |
“A[snoraoid Avp ay} °O,9¢ 38 ANoY J[VYy UO [eNsn oy} pozVApoVUl Us0q PY V1OS 9S9q} [TV +
q2 38
32 4
Zou | Zou | Zou
32 jo a2
0 seu | Sou
38 0 Be
jy | dou ae
3 34 4
“29 80°0 | *9° 90'0 | °99 400 | “99 200
Sou
Zou
js ju
jm po
js jy} | dou
0 0
yo yy pe
jo gou | Sou
js ju Bo
js } yt
jy pe yy
jy Zou | Zou
yo sou | Zou
jo js yu
jM yy pe
Js } }}
js J} | Sou
pan
}}
ju
Sou
Sou
sou
Zou
Sou
3ou
‘9.99 LV SHLONIW NAGI GALVAILOVNIGDY
oooococoooco
a
js
34
sQGLVAILOVNIGHY LON
"2.99 10 sporsad ayosapowm sof Buryyway 0} anp pias upwny ur 4oj}dav0qup daaysyun pounjgou fo sso] ay} Buwmoyy
¢ ATAVL
8¢9P
LOOP
voor
ESOP
TS9P
OS9F
6F9F
609F
8c9P
SCOP
GC9P
029F
PI9P
181
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS
‘g]80} UOLPBAIPOVUT OYNUIMT OAYG-A4JY OY} Ul 9[qvoorqzou ATUO SBA SIT], “BLS Ot{} JO Suresv oy} 03 onp oov{d Yoo4 ulsAjouoy
JO S80] OFBIOpOUT B ATMO 4YVY} 99S [ITA OUO ‘A 9[GBy UO poyV[Nge, Vsoy} YIM 8759} BS0{} JO s}[Nsor oy} Zuteduios Ag :aLON
*(8}S80} 1078] OSOY} 1OJ OpIse qos SBM
wios JO uomI0eds yowe Jo 41d) GI6T ‘T ISBN UO pozso} UooG PVY YOIYM V1Os oY} Suisn ‘7 ysnsny Uo opvuUl o1OM 8480} osay,T, |
‘6I6L ‘TE A[nE WoO soynuIU AqQI1G} JOJ poyVATpOVUT U9Eq PBY BIOS OSOY TL, »
re a ee SS SS
Je yo yo jo tO be q2 J fu ue J 38 4 J} | dou | dou 4 J+ | 8o9F
fe 38 yt-) A 0 38 bao 4 i a jy | dou | dou | Zou J+ | Lg9P
0 0 y3 | dou | Zou yy gou | Sou | Zou | Zou gou | Sou | Sou | Sou | Poop
yo jo yo 5) jo } jo pat 3 jt yy yy jm | Sou | Zou | Sou | Zou | dou J} | 1¢9F
Sou | Zou | dou | 33 gou | Zou | Zou | sou gou | Sou | Zou | sou | QCor
0 0 0 yy | ja i re) 3 jy} | Zou | dou | Zou | sou 3 }} | 8h9F
js yu ju a jy | ya gou | Zou | Zou | dou | dou jm | Sou | Sou | Sou | dou | dou | Zou | Oop
jo jy yp yo yo | jo 4) yy yy yy 4 ym | Zou |} Zou | dou | Zou | Zou | Zou | HEP
jh yah yi y jy a] Zou | Sou | Zou | Zou | sou j3 | Sou | Zou | dou | Sou | Zou | dou | 8z9P
5a y yy | Bou | sou | Jy gou | Jou | Sou | Sou | Zou | Sou | Zou | dou | Zou | Zou | Zou | Bou | CZOp
js jn yA jh jm | jo yy J} | dou | sou pat jm | Sou | Zou | Zou | dou | Sou | Zou | 2Z9F
gou | dou | Zou | Zou | Sou | 39 gou | Sou | Sou | Sou | Sou | Sou | Sou | Jou | Jou | Sou | dou | dou | OZ9F
ju ju ju ju js | jo jy | Sou | Sou | Zou jy jm | Zou | Jou | Zou | Zou y JM | PLOF
yo yo yp js js. || Js yy | Bou | Zou | dou | Zou jy} | Sou | Sou | Zou | dou | Zou | dou | EI9F
jo yo yp jo jo | jo jo js jM ju ju js | Sou | Sou | Zou | dou | Zou jo | I19F
yp yp yp yp jo | jo js js yu ju ju js | Zou | Sou | Sou | Zou }} JM | OL9PF
js yy | dou i 10 |e at) Zou | Zou | Sou | Sou | Sou | Sou | Sou | Zou | Zou | Zou | Zou | Bou | CQoF
jo yp jo yu jm | jo 3 jj | dou pat yy jm | Sou | Sou | Sou | Zou | Zou | Bau | E09
“99 OT'0| °20 g0°0| *99.90'0 | 99 F0'0 | *99 Z0"0 | 99 T0°0 | “29 OT"0| 99 80° | “99 90°0 | “99 $0°0 | *99 200 | “99 T0'0 | “99 OT'0 | “29 80'0 | "99 90'0| “99 0'0 | “99 Z0'0 | “99 100 =
-WON
SALONIW GAM-ALAA AALVALLOVNIGH SALONIWN GAI-ALNGAML GGLVAILOVNIGH SALONIN NAL AGLVAILOVNIGY
od nega’ ae SE ES et Se Se eee eee
“099 10 Burypoy 07 anp Dias uDWNY Ur Urshjzowoy doaysyuD younjoU fo $80o] OY} Bunoysy
9 ATAVL
_— = - ~— a eS ee Ag om asa ~~ Y
TABLE 7
Showing the loss of natural antisheep hemolysins in human sera due to short periods
of heating and to ageing the sera
FIRST DAY* NOT REINACTIVATED
SERUM NUMBER 5
SoS alaeea Seis
= 3 3 S$ 3
oO o i) o o
31 ef ef ef ef ef
33 neg | neg} neg} neg} neg
34 wf | neg} neg} neg| neg
35 neg | neg} neg} neg
36 sf sf sf sf
37 neg | tf neg | neg
38 neg | neg| neg| neg| neg
39 neg | neg} neg| neg} neg
40 sf tf neg} neg | neg
41 wif | neg} neg} neg} neg
42 neg} neg| neg; neg] neg
43 tf neg} neg| neg] neg
44 neg | neg| neg| neg] neg
45 neg | neg} neg} neg} neg
51 sf | neg] neg] neg] neg
52 Ch Sinch Miser wulach
53 tf neg | neg} neg] neg
54 ef ef ef ef ef
55 ef: “) cf >| ‘ef “cf zen
60 neg | neg} neg| neg} neg
SEVENTH DAY NOT REINACTIVATED
31 CL: C8 yer ere ce
33 neg} neg} neg} neg] neg
34 wi | tf neg | neg | neg
35 neg | neg} neg | neg
36 CE ck .\sek ebemieek
37 tf | neg} neg} neg] neg
38 tf | neg] neg] neg| tf
39 neg | neg} neg| neg] neg
40 ti | neg! ti | tf | tf
41 neg | neg} neg] neg] neg
42 neg} neg} neg] neg
43 tf | neg} neg] neg
44 neg | neg} neg} tf
45 tf neg | neg] neg] neg
51 tf neg | neg | tf
52 Ch cf | eb = etm chee
53 Cis |) neg |tis«|-owd. siiiwd
54 ef | ef cf ef ef
55 0 0 0 0 0
60 neg | neg| neg} neg | tf
wi
ef
tf
FIRST DAY FIFTEEN MINUTES
REINACTIVATED
_——_— |) | |S | | | | |
SEVENTH DAY FIFTEEN MINUTES
REINACTIVATED
neg | neg| neg| neg} neg/ neg
tf | neg} neg| neg| neg} neg
tf neg | neg} neg | tf
Ch ef | ey | seks oh clariger
Waeryoth jneg | tf. bil ieee
efmiliteh. fuck | ets. 4) ehetiees
0 0 0 0 0 0
neg} neg| neg] neg| tf | tf
* The sera had been inactivated the usual thirty-minute period at 56°C. the
day previously.
Note: Negative, neg; trace of fixation, tf; weak fixation, wf; strong fixation,
sf; complete fixation, cf.
182
183
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS
‘uoIyexy oJo]duI09 ‘Jo SuolexYy Zuoszs ‘Js ‘uoyexy Yom “JM SUOIyVexY jo oovI} ‘J4 [OAIZVFOU 04 SIOJOI BON : ALON
*S}]SO} OS9Y} IOJ ‘O,9G 9B SOINUIUL 994 PoyVATZOVUIaL 9IOM Bios OL], f
"S}]S0} SOY} IO} Po}PVATPOVUIO YOU O1OM BIOS OT, |
Zou
*).9¢ 7B soynurur AZI1Yy [BNsN oy} IO} poywaryzour o19M ABP YsIY OY} UO pasn BIG ,
Sou | Zou | Zou | Zou | Zou js js Bat yu jm“ ju sou j} | dou y yi yu are
gou | Sou | Zou | Zou | Zou | Zou js ju yu js jo yo ja jm“ jh yu js yp dze
jo jp jo jo jo yp jo jo js js ja ju js | js 8PSP
js ja jy jo js ju | ju ju 4 jt i} ju ju | JM | LecP
jy jo jo jy jo jo jy jy jo jo jo | jo CPSP
jp jo jo jo jo jo jo jo jy jo jy jo 6ESP
jo jp jo jo jo jo jo jo yo jo jo jo 9ESP
jp yp jo jo jo jo jo jo jo J jo jo Tecp
Zou sou jo js ya js | ju | js 3 3 ba 3 y+ | 9 | SzcP
jo jo jo jo jo jo jo jo jo jo jo yo jo jo CCCP
js jt jo jo jo py}; Pp jo yo jy ju ja ju | JM | PCGP
js ju jp jo jo js js js ju jt jM ju ju | JM | ECSP
jo jy jo jo jo jo jo yo jo jy jo jo jo jo TCSP
y} 4 jo js js ju | ja js J} yy yy ne} ju | JS OZSP
jo jo jo jo jy jo jo jy jo jo jo jo jo jo 61SP
sou Sou yA yu jd 30u jt y Zou | Zou | Zou | Zou | Zou sat 9ISP
jy jo jy jo yp yp jo jo jp jp yp jo jo jo ISP
dou Sou yp jo p yp yp yo yp yp jo yp yo 5) ZIGh
“99 0T'0 | °99 800 |" 99. 90'0 | *99 £0'0 | *99 Z0'0 | “99 [00 | *99 OT'0 | “99 80'0 | *99 90'0 | “90 #0'0| “99 Z0'0 | °99 T0'0 | “99 O10 | °29 80°0 | “99 90°0 | *99 £00 | “99 20'0 |" 99 100 ae
java HINGILUNOd ‘agisaL vaas Java HINGAGS ‘aaLsaL vuas ieee
eA Vd Isuld ‘aGagisaL vaas
‘Burabp 03 anp urishjpoway daaysyup yoinjou fo sso) 247 UO
8 AIAVL
184 HASSOW O. VON WEDEL
Besides making these titrations to determine the loss of hemol-
ysin due to heat, the writer made titrations on 97 sera before
and after they had been kept for seven days in the ice box to
determine the stability of anti-sheep hemolysin in human sera.
Some were kept for fourteen days and again retested. Results
tabulated in tables 7 and 8 seem to indicate that there is appar-
ently no marked loss of hemolysin during this period of time.
The apparent discrepancies in the titrations and the apparent
increase of the hemolysin in sera nos. 4512, 4516, 4520, 4528,
4528, 4544, 32B, and 34B, after being kept for fourteen days as
shown in table 8 is probably due to weak blood cells, or strong
complement units, or both.
Kolmer (63) divides thermolabile and thermostable hemolysins in
human sera as follows: ‘‘thermolabile hemolysins”’ those found in fresh
unheated serum; ‘“‘thermostable hemolysins’”’ those found after the
serum has been heated at 56°C. for one-half hour—he states that all
these natural hemolysins with the occasional exception of anti-sheep
hemolysin are completely destroyed by heating at 62°C. for one-half
hour, and that immune hemolysins suffer only partial deterioration at
this exposure.
Thiele and Embleton (64) state that a portion of the natural anti-
sheep hemolysins present in active human sera are thermolabile, being
destroyed by heating at 56°C.
Sherman (65) claims that all hemolysins are thermostable and that
a reduction in hemolytic activity of a serum as a result of heating is
due to ‘‘masking” of the hemolysin rather than actual destruction.
7. DETERMINATION OF THE RELATIVE ANTIGENIC VALUE OF
TUBERCLE BACILLUS ANTIGENS PREPARED AND STERI-
LIZED BY TWENTY DIFFERENT METHODS
The literature on this subject has been recently abstracted by
H. R. Miller (66), and the various antigens used by the several
workers tabulated in four general groups. Group 1. This group
comprises those antigens composed of the whole bacillus, includ-
ing not only antigens prepared from the tubercle bacilli, but also
antigens prepared from other allied acid-fast bacteria. The
results reported vary both in constancy and the percentage of
_
Ee ES ee ee ee ew ee
|
|
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 185
positive fixations in positive cases. Group 2 are antigens which
consist of various tuberculins. In the main the results reported
by these workers have been encouraging, the chief difficulty
being non-specific fixation, especially with the sera of non-
tuberculous syphilitic patients. Group 3 are those antigens
derived from the tubercle bacillus by means of chemical diges-
tion and extraction. The chief workers in this field have been
Deyke, Much, Leschke and Altstaedt. It is to this group that
the ‘‘partial antigens’ belong. Group 4 comprises antigens
derived from normal and tuberculous tissue. Such antigens
have not given uniform results. In going over the results of the
various workers, the antigens in group 1 seemed to be the only
ones that would give practical results. In the preparation of a
tubercle bacillus antigen of this group, the first question which
arises is whether to use a suspension of living bacilli or of killed
bacilli. Corper (67) states that, while the most reliable investi-
gators concede that a suspension of living tubercle bacilli is the
only one of the many antigens used, that is of specific value, the
objections to the living bacillary emulsion are the small leeway
between the antigenic and the anticomplementary dose, the
turbidity produced in the tubes and the fairly high percentage
of non-specific reactions.
The writer has attempted on several occasions to use suspen-
sions of living bacilli but found that they became anticomple-
mentary very quickly and for this reason could not be used
satisfactorily. Besides, the danger attending the handling of
antigen made with the living bacilli precludes its general use in
diagnostic laboratories. He has not, therefore, attempted to
use suspensions of living bacilli in this comparative study. All
his suspensions have been made from killed bacilli.
The next question which arises is the best method of killing
the bacilli without destroying the antigenic value of the finished
antigen.
M. A. Wilson advocated killing the bacilli in the broth culture
by sterilizing in the Arnold sterilizer for one hour and then
washing the bacilli from the filter paper with 95 per cent alcohol.
186 HASSOW O. VON WEDEL
White of the Otisville Sanitarium advised the writer to make
the antigens from the living bacilli withou™ the use of heat,
washing the living bacilli directly with alcohol and preserving all
the sediment from the alcoholic mixture.
To definitely settle the above questions, the writer made up
antigens in 18 different ways from one strain of tubercle bacilli
grown on broth for four months and has made comparative tests
with these 18 antigens, Wilson antigen no. 330 and Wilson
antigen no. 400, on sera from 60 patients. Forty-five of these
were tuberculosis patients and 15 were normal controls. In all
1167 comparative tests were made. The writer was unable to
make comparative tests with the entire 20 antigens on all the
sera as the quantity of serum obtained from some of the patients
was insufficient. Duplicate comparisons were made on as many
as possible. The antigens were made up as follows: The broth
flask containing the tubercle bacillus growth was thoroughly
shaken so that an even distribution of the tubercle bacillus pel-
licle was obtained. This was divided into three bottles. Bottle
1 was placed in the steam sterilizer and kept at 100°C. for one
hour. It was then taken out, thoroughly shaken, and the mix-
ture of broth and bacillus pellicle was poured on a large filter
paper and allowed to drain until no more broth dripped from the
lower orifice of the funnel. The filter was then broken and the
sediment washed off with 95 per cent alcohol. This was left to
macerate in the alcohol for ten days.
Bottle 2 was also placed in the Arnold sterilizer and heated
for one hour at 100°C. This mixture was poured on a filter
paper, and at once washed with cold normal saline solution until
the wash water ran through perfectly clear. The filter was then
broken, and the sediment was washed off the filter paper with
95 per cent alcohol and left to macerate for ten days.
Bottle 3 was not heated at all. The cold live bacillus culture
was poured on a filter paper and washed with cold saline until
the wash water was clear. The filter was then broken and the
live bacilli were washed off the filter paper with cold 95 per cent
alcohol, and left to macerate in this alcohol for ten days. At
the end of ten days, the supernatant alcohol was poured off
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 187
from all three bottles. The sediments of bacilli and precipitated
proteins were washed in ether three or four times for a period of
twenty-four hours; they were thoroughly shaken each time.
The ether was poured off and the sediments were washed with
ether into centrifuge tubes. These were centrifuged and the
supernatant ether was poured off. In this way all constituents
soluble in alcohol and ether were removed. The tubes were
then loosely plugged with sterile cotton and placed in an incu-
bator over night to dry by evaporating off the ether.
The dried powders were weighed and suspended in 0.9 per
cent saline in the proportion of 1 gram of powder to 200 ce. of
saline. These were thoroughly emulsified in a mortar. Sus-
pension 1 was then divided into six bottles labelled 1A, 1B, 1C,
1D, 1E and 1F. Every precaution was observed to prevent
bacterial contamination of the bacillus suspensions.
Bottle 1A was left without heating, or the addition of any
preservative.
Bottle 1B was sterilized in the Arnold sterilizer at 80°C. for
one hour.
Bottle 1C was sterilized in the Arnold sterilizer at 100°C. for
one hour on two consecutive days.
Bottle 1D—an equal part of 95 per cent alcohol was added to
the bacillus suspension, shaken and stored without heating.
Bottle 1E—one part of alcohol was added to three parts of
the bacillus suspension; the mixture was shaken and stored
without heating.
Bottle 1F—0.5 per cent carbolic was added to the bacillus
suspension, which was stored without heating.
Suspensions 2 and 3 were similarly divided and treated as
suspension 1. In this way, the writer had antigens made of
the same culture, but treated in 18 different ways.
Dilutions of 1-10, 1-25 and 1-50 were made of all the A, B,
C, and F antigens, and were titrated to determine their anti-
complementary dose. Antigens D were made up in 2-10, 2-25,
and 2-50 dilutions because these antigens were mixed with equal
parts of alcohol and therefore contained only one-half as much
powdered sediment as the A, B, C, and F antigens. Antigens
THE JOURNAL OF IMMUNOLOGY, VOL. Y, NO. 2
HASSOW O. VON WEDEL
188
S| a Ul Sa ca a ar eel Sf
tN Ne | i ec | re aU
Pigment sl ea tal Oe ae | ee eon =
Se fea HT i | ee a S| Pee cl ge eg OP
Se Ae fe ees NP tac cl Ml Fn alc | se Pt
=e, =| pees Re | cal MN 8a oe mer el A Hk fe
Bi pee | sees |e Sea | she ll aa Sl ah | ate
PER es Nara Pee Pate ae Set es PF
Bbseee | eet ea sak Bat ose reeal oh Des | one Coceaakee ete |e
So famed fae elated tr ees) Perl asl pe ef = pee ee
se Great cle Gog | say | eae wale Te | ea eel ah P| aetna [eee
se Re aml Per ees ral US
me | ee ae te ee ee ee
1} = = a fet ee eeepc ele =
game teen EY eet | Sh dal ts og ee | Man a ah el ee
peace al See ca | eel eg | git onl eee > ip al ee |)
1 | ae a ied Becta ol We 1 fet Za Pee fe
a en |e sags cs Eo ol OP ate Ih ed Pe
eee | ierree | am e | en atl e | te Vg | eee el teal Ae De [ es
ee oe I lee a eae ml cia MY ie ft] eb 2
+6 66 £9 Tg SP OF & T€ ¥I I
uIndeg | UINdag | UINIeg | UNdeg | UINJeg | UINJOg | UINdeg | UNdIeg | WNnIeg UInIeg | UINJeg | UNIeg | UINnIEg
snopnoi9qn4-uo Ny
SISONDVIG TVOINITO
spoyjau quasaf{up fiywany hq pasodasd suabywn ay} YUM appu DLAs JDULOU PUD snoj}NI.1aqQn} WO 8789} aarynLvdwod ay} fo maf v Buunoyy
6 ATAVL
*(9AIJBSOU SB poyIOdal) UOIY’XY Jo 9dV1} 0} SIOJOI -14
|
ay
IHHI HHI
—
ss
as II
snoynoseqn J,
8
+2 [+p | +e | +1 | +2
SEO S|) eee] sie osl oie al ees
+2 {+e | +1 | +1 | +2
+ {41 | -— | = | 41
+ |+e | + | 41 | +2
+2 | +e | +2 | +e | +2
+ |+r | +r | 41 | 41
UIndeg | UINIeg | UINIeg | UINJeg | UNJeg
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 189
E were made up in 1.5-10, 1.5-25, 1.5-50 dilutions because in
these antigens one part of alcohol was mixed with three parts
of bacillus suspension and therefore contained only three-fourths
as much powdered sediment as the A, B, C, and F antigens.
The titrations were set up in the following volumes: 0.05 cc.,
0.1 ce., 0.2 cc., 0.3 cc., 0.4 cc., 0.5 ec., 0.6 cc., 0.7 ec. and 0.8 ce.
TABLE 10
Titration results of antigens prepared by twenty different methods for their
anticomplementary activity
QUANTITY OF ANTIGEN
ANTIGEN
SS SS nD ED
0.1 cc. 0.2 ce. 0.3 cc. 0.4cc. 0.5 cc. 0.6 cc. 0.7 ce. 0.8 cc.
1A — — — _ = = — tr
2A = _ _ _ = = = tr
3A - _ _ _ = — tr +
1B = = _ _ = = = tr
2B — _ - _ = = = tr
3B _ _ _ _ = = = tr
1C = = = = = = tr tr
2C = = = = = = = tr
3C = = = = = = = tr.
1D - = + + ~ -
2D - - - = + | + -
3D - - + ~ + | ++ | ++
1E ~ - - - = = ~
2E _ - _ — + + +
3E - — == = = = + tr.
1F — — = _ _ tr. tr.
oF = =: = = = = tr
3F — = = = == = tr.
All antigens were made up in 1-50 dilutions, except the D and the E antigens.
The D antigens were preserved by the addition of equal parts of 95 per cent
alcohol. The final dilution was therefore made by adding 0.2 cc. of this alcohol
antigen mixture to 48 ec. of 0.9 per cent saline solution.
The E antigens were preserved by the addition of 1 part of 95 per cent alcohol
to 3 parts of stock antigen. The final dilution was therefore made by adding 1.5
ec. of this alcohol antigen mixture to 48.5 ec. of 0.9 per cent saline solution.
The 1-10 and the 1-25 dilutions were found to be too anticom-
plementary. The 1-50 dilutions gave satisfactory results as the
anticomplementary dose for the A, B, and C antigens was from
0.7 to 0.8 cc. Antigens D, E, and F were anticomplementary
in much lower doses, as shown in table 10.
190 HASSOW O. VON WEDEL
In going over the results of the 1167 comparative tests made
with these twenty antigens on the 60 sera, one notes that prac-
tically all the tests made with the no. 2 antigens gave weaker
results than either the no. 1 or no. 3 antigens. Antigens 1A,
3A, 1B, 3B, 1C and 8C all gave approximately the same results,
and gave practically no non-specific results with the normal
control sera.
Referring to table 9 one notes that antigens D, E, and F gave
stronger results on an average, but also gave a moderate num-
ber of non-specific results with the normal sera. Lower dilutions
of these were tried in a few instances and more consistent results
were thereby obtained. The alcohol and the carbolic acid used
as preservatives apparently gave a marked anticomplementary
tendency to these antigens. This was borne out by the titra-
tions tabulated in table 10.
These observations disprove the assertion that heat destroys
the antigenic value of bacillary antigens, also that antigens
made from bacilli without being killed by heat are better than
antigens made from heated cultures. Apparently washing the
hot broth sediments on the filter paper with saline removed
some of the antigenic properties of the antigens.
Bronfenbrenner (69) states that it is necessary to free each
sample of tuberculin (tubercle bacillus antigen) of all its lipin
fraction before using such tuberculin for the complement fixation
test.
One hundred sera were tested by a worker in this laboratory
with an antigen made from bovine tubercle bacilli. There was
no distinction with the bovine and human antigens. The varia-
tions were in about the same proportion as occurred in tests
with some of our different kinds of human strain antigens. One
hundred sera were tested with the glycerine extract antigen
sent to the laboratory by Petroff (68). This antigen gave the
same reaction as M. A. Wilson’s in 95 per cent of the cases, and
in 5 per cent the reaction was weaker with Petrofi’s antigen.
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 191
8. DETERMINATION OF THE RELATIVE VALUE OF THE HECHT-
GRADWOHL TECHNIC AND THE TECHNIC OF NATURAL ANTI-
SHEEP AMBOCEPTOR ABSORPTION IN COMPARISON WITH THE
STANDARD COMPLEMENT FIXATION TEST
The Hecht-Gradwohl tests were performed with the simplified
technic advocated by W. J. Bruce (70).
According to Lewis and Newcomer (71) every one of 70 con-
secutive fresh, sera contained enough native complement for a
Hecht-Gradwohl test if susceptible corpuscles were used.
Gradwohl (72) concluded that the Hecht-Gradwohl test in his
hands has been a far better check or control of the Wassermann
reaction than the use of any of the controls now in vogue. The
so-called border-line Wassermanns if truly positive will show a
strong Hecht-Gradwohl reaction. In other words, the Hecht-
Gradwohl is a far more sensitive test than the original Wasser-
mann reaction.
In his mind, no negative Wassermann was worth anything
unless backed up by a Hecht test. He stated that even with
this control test 25 per cent of syphilitic sera will give negative
reactions.
L. E. Schmidt of Chicago, Carl C. Warden of Ann Arbor,
M. L. Heidengspeld of Cincinnati and others concluded that this
test was a very good control to use with the Wassermann reac-
tion but that it should not be used alone as it may be too sensitive.
Rubinstein (73) and Radossavlievitch (74) stated that the
results of the Hecht system did not offer as much guarantee of
specificity as the standard technic.
Eschbach and Duhot (75) guarded against the inevitable
danger of doubtful reactions, which are provoked with the Hecht
method, by the use of a very dilute antigen.
In this study, the natural antisheep amboceptor was removed
from the sera by the addition of equal parts of a 10 per cent
sheep cell suspension. This mixture was incubated for one
hour, centrifuged and the clear serum pipetted off from the sedi-
mented blood cells. As the serum was diluted 1 to 2, two times
the regular volume was used in setting up the tests. This is
192 HASSOW O. VON WEDEL
the method advocated by Bauer (76-77). The procedure some-
times causes the sera to become anticomplementary, which is
known as the Sachs-Friedberger phenomenon.
Rossi (78) advised absorbing the heated serum at a low tem-
perature—keeping the serum-blood-cell mixture packed in ice
and chilled during the centrifuging.
Wechselmann (80) advocates absorption of hemolysin from
sera with barium sulphate. Noguchi and Bronfenbrenner (81)
found that this procedure removed the natural antisheep hemol-
ysin and also syphilitic antibody.
Simon (79) has also described a method for removing the
natural antisheep amboceptor. Yor ten minutes 0.4 cc. of sera
is inactivated at 56°C., mixed with 1.6 cc. of 2.5 per cent sheep
cell suspension, extracted for ten minutes at 37°C. in a water
bath and centrifuged free from corpuscles; each cubic centimeter
then contains 0.2 cc. of serum.
Jacobaens (82) found in a study of 257 cases, 10 per cent more
positive Wassermann reactions after absorbing the sera with
sheep corpuscles. Olmstead (83), Van Saun (84), Ottenberg
(85), Sielman (86), Stern (87), Kaliski and others have advocated
various methods for the closer adjustment of the antisheep
hemolytic system to avoid the effect of excessive amounts of
natural hemolysin.
Stimson (88) stated that the criticism of the use of sheep
cells in complement fixation tests on the ground of their being
subject to the action of native amboceptor in human sera is not
altogether well founded. It is granted that a fair percentage
of human sera have this amboceptor, but these same sera may
give excellent fixation, because if the complement has been fixed
in the first phase, the mere increment of amboceptor will not
cause hemolysis in the second phase. However, the removal of
natural antisheep amboceptor will somewhat increase the
percentage of positive reactions in a series of tests.
Neill (89) found in a series of experiments that while the
natural antisheep hemolysin does reduce fixation when present
in syphilitic sera, the amount of hemolysin must be very large
and the amount of syphilitic antibody very small in order to
ee eee
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 193
produce a significant effect. He concludes that the presence of
antisheep hemolysin in sera is not an objection to the use of
the sheep cell hemolytic system, if the sera are employed in
amounts corresponding to not less than 0.1 ce. to a total of 5 ce.
for the test.
Kolmer found that with sera containing large amounts of
syphilitic antibodies excessive amounts of hemolysin have prac-
tically no influence upon the results or, at the most, reduce a 4
plus to a 3 plus reaction; but weak positive results are often
reduced to negatives when a large excess of hemolysin is present
if the readings are made after the tubes havestood in the refrig-
erator over night. Readings made immediately after taking
from the water bath did not show such marked differences.
However, all methods of removing the natural antisheep
amboceptor are probably too time consuming for a routine
diagnostic laboratory to perform when large numbers of sera are
examined each day. |
The writer has made comparative tests on sera from 60 pa-
tients, forty tuberculous individuals and twenty normal controls.
In these comparisons the same antigen (C3—1-50 dilution, 0.2
cc.) and the same sheep cells were used for all the tests.
In table 11 thirty-five of these comparative tests have been
tabulated. The tests were run in four sections: first, the stand-
ard complement fixation test was made on the serum one day
after its removal from the patient; second, the Hecht-Gradwohl
modification was performed on the serum two days after its
removal from the patient; third, the serum was tested after the
natural antisheep amboceptor had been absorbed from it; fourth,
the standard complement fixation test was again made on the
serum after keeping it for seven days in the ice chest under sterile
precautions.
Upon examining the chart, one notes that the tests agreed in
the majority of cases. However, there are a moderate number
of variations. By absorbing out the natural antisheep ambo-
ceptor, 5 normal sera gave some degree of fixation; 5 tuberculous
sera gave stronger fixation results than the first day standard
test; 3 tuberculous sera, negative with the standard test, gave
TABLE 11
Showing a few comparative tests made with the regular complement fixation test on
the first and seventh days, the Hecht-Gradwohl test and after absorption
of the natural antisheep amboceptor
& &
“hl (Pama Rat JSP ed a
2 & PS See | & 4
2 a ¢ ORI fy CLINICAL DIAGNOSIS
ea] ee | SE]
a i) | < &
1 + 1+ | neg 1+ | Normal
2 neg | neg | neg + | Tuberculosis type IV
3 neg | neg | neg | neg | Normal
4 neg | neg | neg | neg | Moderately advanced, tuberculosis type I
5 neg | neg | neg | neg | Normal
6 4+ | 4+ | 4+ | 4+ | Moderately advanced, tuberculosis type III
7 neg a neg | neg | Suspicious
8 1+} 1+] 1+ | 1+ | Moderately advanced, tuberculosis type III
9 neg | neg | neg + | Moderately advanced, tuberculosis type IV
10 1+ | neg | neg 1+ | Far advanced, tuberculosis type IV
il neg | neg | neg | neg | Normal
12 neg | + | neg | 1+ | Moderately advanced, tuberculosis type IV
14 2+) 2+ 1+ | 2+ | Moderately advanced, tuberculosis type III
15 neg = = neg | neg | Normal
16 4+ |} 2+ | 4+) 4+ | Tuberculosis type III
ile neg + + + | Tuberculosis type II
18 neg | neg | neg | neg | Tuberculosis type I
19 3+ | 44+ | 38+ | 4+ | Tuberculosis type III
20 4+) 41+ | 4+ | 4+ | Tuberculosis type III
21 tr | neg — + | Tuberculosis type III
22 1+ | neg | 2+ | + | Tuberculosis type III
23 1+} 44+) + + | Tuberculosis type II
31 4+ | 44+ | 44+ | 44+ | Incipient, tuberculosis type II
32 neg | neg | 2+} + | Incipient, tuberculosis type II
33 neg | 4+ | neg | neg | Incipient, tuberculosis type II
35 neg + 4+| + | Incipient, tuberculosis type II
37 1+ | 4+] 3+] 2+ | Incipient, tuberculosis type II
40 + ac + + | Incipient, tuberculosis type I
4l neg | 2+] + neg | Incipient, tuberculosis type I
42 4+ a. 4+ | 4+ | Incipient, tuberculosis type II
44 neg + 2+ | neg | Incipient, tuberculosis type II
52 neg | neg + tr | Normal
53 neg | neg + neg | Normal
58 neg + 2+ tr | Normal
60 neg | neg 2+ + | Normal
*Refers to the day the reaction was made after the specimen was removed
from the patient.
+ These tests were made two days after the specimens were removed from the
patients.
Note: neg = negative; tr = trace; ac = anticomplementary.
194 |
ht ee
'
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 195
positive results with the Hecht-Gradwohl modification; 6 tuber-
culous sera, negative with the standard test, gave positive results
and 2 tuberculous sera, positive with the standard test, gave
negative results with the amboceptor absorption method.
In reviewing the results, one is struck by the fact that some
sera from tuberculous patients will not fix complement with any
of the modifications tried out, even when there is a certain
amount of non-specific fixation going on in the normal control
sera due to the weak system used. Also, as soon as one attempts
to make the complement fixation test more sensitive, normal
sera begin to give some degree of fixation. While both the
Hecht and the amboceptor absorption methods gave more posi-
tive results with the sera from tuberculous patients, they also
gave some non-specific results with normal sera.
9. COMPARATIVE RESULTS OBTAINED BY TESTS WITH SERA ONE
DAY OLD AND WITH THESE SAME STERILE SERA AFTER PRE-
SERVING FOR ONE WEEK IN THE ICE BOX
These comparative tests have been made on sera from 900
patients (500 tuberculous and 400 non-tuberculous). In a
previous publication by the writer, he reported that certain sera
which gave negative or weak positive reactions when tested on
the day after the specimens were taken from the patient, gave
strong positive reactions when tested after the specimens had
been kept for one week under sterile precautions in the ice
chest. At that time, he stated that this occurred in a
large percentage of cases and was due, he thought, to the
presence of some inhibitory substance in the serum which disap-
peared after the week’s interval. He also stated-that he did not
observe this phenomenon in any of the sera from non-tuber-
culous cases.
While we have continued to observe this difference in fresh
and kept sera (note table 12) it has been in a much lower percent-
age of tuberculous cases and has also appeared in a moderate
number of non-tuberculous cases, showing that it is not a specific
phenomenon and that it is probably caused by certain anti-
196
BERUM
SCO OND OF Wt Fe
_
NUMBER
HASSOW O. VON WEDEL
TABLE 12
DIAGNOSIS
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Non-tuberculous
Tuberculous
Tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Perse ee
SEVENTH
Results of complement fixation tests made on the first and seventh days
DIAGNOSIS
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Non-tuberculous
Tuberculous
Tuberculous
Tuberculous
Non-tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
Tuberculous
——_
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 197
TABLE 12—Concluded
BS y a -
a2 7 2 DIAGNOSIS aa & z DIAGNOSIS
mz Fe > R 7, Fe >
z = Fy Fy i Fs
139 _ - Non-tuberculous 164 = 4+ Tuberculous
140 — — Non-tuberculous 165 — — Tuberculous
141 -- — Non-tuberculous 166 -- -- Tuberculous
142. — — Non-tuberculous 167 |} 2+ | 4+ Tuberculous
143 — - Non-tuberculous 168 | + + | Tuberculous
144 - — Non-tuberculous 169 =< -- Tuberculous
171 -- - Non-tuberculous
145 _ = Non-tuberculous 172 — — Tuberculous
173 _ _ Tuberculous
146 -_ 2+ Non-tuberculous 174 — Non-tuberculous
175 - = Non-tuberculous
147 a — Tuberculous 176 - - Non-tuberculous
* Day reaction was made after the specimen was taken from the patient.
7 tr = trace of fixation.
complementary changes, loss of natural antisheep amboceptor
or by the unavoidable differences in the complement, blood
cells, and antigen suspensions used in these comparative tests.
The fact that in my previous publication a high percentage
of negative or weak positive results was reported with tuber-
culous sera on the first day tests which on the seventh day gave
markedly stronger reactions, and that no non-tuberculous sera
showed this change, was probably due to one or more of the
following reasons: First, the antigen used in this series had a
very wide range between its anticomplementary unit and its
antigenic unit, and only two antigenic units were used: ie.,
0.1 cc. of a 1-50 dilution. Second, only the classical Wasser-
mann volumes of sera were used while in all our later tests twice
the regular volumes of sera had been used as a basis for our
reports. Third, nearly all the tuberculous cases in this series
showing this marked change were active type 2 cases which
should have given strong positive reactions on the first day
tests. The antigen dose was probably too weak to bring this
out. The additional help of slight anticomplementary changes,
the loss of some of the natural antisheep amboceptor due to the
extra fifteen minute inactivation, and other unknown and uncon-
198 HASSOW O. VON WEDEL
trollable factors when added to the specific tubercular reaction
were sufficient to give specific positive results but not sufficient
to give positive reactions with the normal sera.
In the last series of 101 cases every precaution was taken to
avoid or to note any difference between the antigen, complement
and blood cells used on the first and the seventh day tests.
The complement used was carefully titrated against sheep
cells and then salted (by the addition of 9 per cent dry powdered
salt to the concentrated complement). This salted comple-
ment and a portion of the unwashed sheep cells was then packed
in ice in a thermos bottle and kept in this way for the seven-day
period. Even after these precautions, differences were found in
the antigenic units on the first day and the seventh day tests on
several occasions.
In summing up the results of the first and seventh day tests
on sera from a group of 101 patients (61 tuberculous individuals
and 40 normal controls), 15 tubercular sera gave stronger reac-
tions on the seventh day than on the first day tests, 4 of these
changed from complete negative to positive. At the same
time, 4 of the 40 normal control sera changed from negative or
doubtful reactions to some degree of positive ranging from plus-
minus to positive 3 plus. These results are shown in table 12.
It, therefore, seems probable that no change in the specific
fixability of the tuberculous sera occurs, but that the apparent
change is due to anticomplementary changes and non-specific
variations in the reagents employed in making the tests.
Rapoport in 1919 (90) made a similar observation during his
study of the fixation test in influenzal pneumonia, using strains
of B. influenzae as antigen, stating that the fixation test on sera
made during the time the patient was still acutely ill were in
many cases negative or slightly positive; but when these sera
were kept in the refrigerator for from six days to three weeks,
many gave strong positive reactions.
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 199
10. DETERMINING WHETHER ANY CROSS FIXATION EXISTS BE-
TWEEN THE TUBERCLE BACILLUS ANTIGENS USED IN THIS
STUDY AND SYPHILITIC ANTIBODIES
During this study the writer has made Wassermann reactions,
as well as complement fixation reactions for tuberculosis on
the sera from 192 patients to determine the possibilities of
non-specific cross fixation with sera giving strong Wassermann
reactions.
Eighty-five of these sera gave positive Wassermann reactions
and absolutely negative tuberculosis complement fixation reac-
tions; 104 gave negative reactions with both the Wassermann
and the tubercle bacillus antigens; 2 gave positive reactions
with both antigens and had clinical histories and symptoms of
both diseases; and 1 gave positive results with both antigens
whose history could not be obtained. This test was made at a
general army hospital, and as the writer was shortly afterwards
transferred to another camp, he was unable to follow up the case
and determine whether or not the patient had tuberculous
complications.
At any rate, 85 out of 85 positive Wassermann sera gave nega-
tive tuberculosis complement fixation tests.
Burns (91) of the Boston Board of Health reported that he
made 912 complement fixation tests on sera which they received
for routine Wassermann tests. Of these 221 were positive and
691 were negative for the Wassermann reaction.
Of the 221 positive Wassermann tests, complement fixation
tests for tuberculosis gave 21 positive, 29 doubtful, and 171
straight negative. Of the 691 negative Wassermann tests,
55 reacted positively for tuberculosis, 35 moderately positive.
There were 21 delayed negative and 580 negative reactions.
They were unable to obtain clinical histories in many of these
cases to confirm the specificity of the tuberculous reactions; but
the evidence seemed conclusive that there was no cross-fixation
with syphilitic serum.
The percentage of positives, 15, seems higher than should
be expected; but when we consider that tuberculosis causes the
200 HASSOW O. VON WEDEL
death of about one person in every ten and that careful observers
calculate at least three living active cases to each death, perhaps
this percentage is not excessive, especially among that class
of patients who appear for free Wassermann tests.
From their observations, it appeared that the test was specific
for tuberculosis.
Bronfenbrenner (92) stated that the complement fixation
test for tuberculosis, using the Besredka antigen, appeared to be
specific for tuberculosis.
Craig stated that in his experience the complement fixation
test with his tubercle bacillus antigen does not give positive
results with syphilitic sera when no coincident tubercular
infection is present. He also concluded that a positive reaction
is specific and that it apparently indicates active lesions.
H. J. Corper (93) is one of the few serologists who do not agree
with the above statements of the high percentage of specific
results with this test. He stated that in his series of cases, he
used 92 sera with positive Wassermann reactions and of these,
65 gave cross fixation with all tubercle bacterial antigens. He,
therefore, concluded that in the presence of a positive Wasser-
mann reaction, the presence of a positive complement fixation
test for tuberculosis is of no practical value.
1l. THE RELATIONSHIP OF THE VON PIRQUET REACTION TO THE
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS
In this study the writer has been able to compare the results
of the von Pirquet reaction with the complement fixation test
on only a small number of cases, as both the clinicians and the
patients objected to the von Pirquet test.
However, when the results of this small number of compari-
sons are examined on table 13 we at once see that no parallel
relationship exists between these two reactions.
Boez (94) states that the reaction of fixation has no necessary
relation of coexistence or intensity with the cutaneous reaction.
During the first stage of pulmonary tuberculosis the cutaneous
reaction and reaction of fixation generally exist together. The
i — a fo _
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 201
dissociation following the type where the cutaneous reaction is
negative, and the fixation reaction is positive indicates a step
advanced and an evolution unfavorable to the patient. Finally,
the formula is modified to the terminal phase; failure of all of
the reactions, even the fixation reactions disappearing.
TABLE 13
Table showing the comparison between the von Pirquet reaction and the
complement fixation test
SPECIMENS FROM PATIENTS WITH POSITIVE SPECIMENS FROM PATIENTS WITH NEGATIVE
VON PIRQUET REACTIONS VON PIRQUET REACTIONS
Positive fixation Negative fixation Positive fixation Negative fixation
23* 74 per cent 8 26 per cent 8 61 per cent 5 39 per cent
* Refers to the number of patients.
12. THE CLASSIFICATION OF TUBERCULOSIS PATIENTS
The classification of pulmonary tuberculosis has been under-
taken by many investigators since ancient times. Noted classi-
fications have been made by Williams of the Brompton Hospital,
Cornet, L. Bard, Koeniger, Turban (95-96), Meissen (97) and
Walter Rathbun (98). Rathbun classifies tuberculosis as incip-
ient A, B, and C; moderately advanced A, B, and C; far advanced
ABC.
While Rathbun’s classification, accepted by the American
Sanatorium Association, undoubtedly is a very broad classifica-
tion and takes in every possible stage of the disease, it is divided
into so many divisions (nine) that no two clinicians can ever agree
on the types in a group of 100 cases. Also, this classification,
if condensed into its three main divisions only; i.e., incipient,
moderately advanced and far advanced, disregarding the activity
of the disease manifested by the patient at any given time, does
not run parallel with the degree of antibody production in the
patient’s blood serum. As the fundamental basis of the com-
plement fixation test is the degree of antibody production in
the patient’s blood serum, and as this antibody production is
primarily based on the activity of the disease regardless of
whether it is incipient or far advanced, no parallel relation-
202 HASSOW O. VON WEDEL
ship can exist between this classification and the test under
investigation.
In this study the complete clinical data has been obtained on
practically all of the cases, and X-ray findings, von Pirquet tests
and Wassermann reactions have been made on a large number.
The clinical data consists of age, past and present temperature,
pulse and respiration records, sputum reports and _ physical
symptoms with the clinical diagnoses and classifications. The
classifications were made by various diagnosticians, who used
the National Association’s, Stoll’s and a new classification by
the writer. Many of the recent investigators agree that their
highest percentage of positive fixations are in the moderately
advanced cases, while the far advanced cases give a rather low
percentage of positive results. The cases with a very poor
prognosis, especially the laryngeal cases frequently give abso-
lutely negative results.
In checking up the findings of all the series of tests by means
of the National Association’s classification, the results were
rather indefinite. One particularly noticed that the positive
findings were most numerous and most definite in the cases
showing active constitutional symptoms regardless of whether
they were incipient, moderately advanced or far advanced; and,
vice versa, the positive results were least numerous and fre-
quently doubtful in the cases where there were slight or no
constitutional symptoms again regardless of the degree or stage
of the lesions. In fact, the strength and percentage of the
positive findings ran a fairly parallel course with the degree
of activity manifested by the patient’s constitutional symptoms
rather than by the stage of the disease or the extent of the lesion.
The early or incipient cases even with very small lesions but
with active constitutional symptoms gave quite a high percent-
age of negative results.
The writer, has, therefore, arranged a classification based on
the antibody production in the patient’s blood serum as indi-
cated by the patient’s constitutional symptoms. Incipient
A, moderately advanced A, and far advanced A cases produce
relatively few antibodies. Incipient B, and incipient C pro-
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 203
duce more, moderately advanced B and far advanced B appar-
ently produce the most antibodies in the greatest percentage of
cases. Moderately advanced C and far advanced C frequently
are overwhelmed by the disease and therefore produce only few
or no detectable antibodies.
To be of any value to the clinician, a classification of this kind
must be interchangeable with those most commonly used by the
clinicians. The following simple classification, while by no
means perfect, fits in fairly well with the complement fixation
findings and can easily be transcribed by the clinician in terms
of the National Association’s or Stoll’s classifications.
THE WRITER'S CLASSIFICATION act Deciterictos Tare STOLL’S CLASSIFICATION
Non-tubercular patients | Negative controls We
Suspects Suspects 5A and 5B
Inactive Inactive cases Inactive
Type I Incipient A, moderately
advanced A, and far
advanced A 1A, 1B, 2A, and 3A
Type II Incipient B, incipient C | 1C and 2B
Type III Moderately advanced B
and far advanced B 3B
Type IV Moderately advanced C,
and far advanced C 1V
The following classification was published by the writer in a
previous communication (99).
Type I. Primary cases; very few physical symptoms present; no
tubercle bacilli found in the sputum or found only after the exami-
nation of many specimens by the antiformin method.
Type II. Active cases; patient expectorating tubercle bacilli, the
diseased area being walled off incompletely or not at all.
Type III. Active cases in the last stage; patient in a dying condition.
Type IV. Partially inactive cases; that is, cases expectorating
tubercle bacilli but having very marked fibrous formation with the
consequent complete walling off of the diseased area from the body
proper.
Type V. Inactive cases.
Type VI. Cases reported as suspicious but expectorating no tubercle
bacilli and having no symptoms of tuberculosis.
Type VII. Non-tubercular cases.
THE JOURNAL OF IMMUNOLOGY, VOL. VY, NO. 2
204 HASSOW O. VON WEDEL
As this classification was not interchangeable with those
commonly used by the clinicians it was dropped.
The writer has tried definitely to answer the question, ‘‘ Which
type of tuberculosis gives the most and strongest positive
results?”
With this idea in view, he attempted to obtain a series of cases in
which there would be no question as to the accuracyof the classifi-
cation of the patients. He personally took all the specimens from
the patients in two duplicate series, three weeks apart. The physi-
cian in charge assured him that the classifications (made by the
National Association’s grouping) would be as accurate as they
could be made. When the tests were all completed and the classi-
fications of the first set were compared with the second set on the
same patients made three weeks later, a wide discrepancy in the
duplicate classifications made by the hospital physicians was
found. This was especially noticeable on the female side where
a difference of 50 per cent in the comparison of the first and the
second classifications was found. The male side showed a
difference of 21 per cent in this comparison. This shows the
extreme difficulty of correctly classifying the various types of
tuberculosis patients by a classification as broad as the National
Association’s, and also how difficult it is for the serologist cor-
rectly to tabulate his data according to the types of cases.
Mourseend (100) said in his recent paper on the complement
fixation test for tuberculosis that no effort would be made to
give a detailed classification of the cases as it was felt that in the
hands of different clinicians the same set of cases would receive
different classifications. Cases considered as incipient by some
observers would be considered as moderately advanced by others,
and vice versa.
13. THE RELATIONSHIP OF THE PATIENT'S TEMPERATURE, PULSE,
RESPIRATION AND AGE TO THE COMPLEMENT FIXATION TEST
In attempting to determine whether the temperature, pulse
and respiration records of the patients bore any direct relation-
ship to the complement fixation reactions, the writer has com-
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 205
pared his laboratory findings with the record charts, and he
has found that the cases giving a 4 plus reaction had an
average temperature of 99.6; pulse, 104; and respiration, 29.
The patients giving a 2 plus reaction had an average tempera-
ture of 99; pulse, 94; respiration, 25. Those giving a doubtful
or negative reaction had an average temperature of 99; pulse, 88;
and respiration 23. Fifty-five per cent of the 4 plus cases and
only 10 per cent of the doubtful and negative cases had a tem-
perature of over a hundred.
These averages are somewhat misleading as there were enough
wide discrepancies in all the groups to bring the averages rather
close together. However, a large percentage of cases having
high temperature, pulse and respiration records gave strong
positive reactions; the percentage of positive results running
about parallel with the temperature, pulse or respiration record;
i1e., the lower the temperature, pulse or respiration, the fewer
were the positive results.
Sixty-five per cent of the 4 plus, 12 per cent of the 2 plus and
only 5 per cent of the doubtful and negative cases had a respira-
tion record of 30 or over. The writer also attempted to see
whether, possibly, the age of the patient had any effect on his
reactions and found that the average age of the patients giving
a 4 plus reaction was 38; the average age of those giving a 2 plus
reaction was 35; the average age of those giving a 1 plus reaction
was 40; the average age of those giving a plus-minus reaction
was 34. As all of the groups contained patients both young
and old, no conclusions could be drawn.
14, SUMMARY OF SIX SERIES OF COMPLEMENT FIXATION TESTS
These tests were made in six separate series. The first during
the winter of 1917-18 on 200 specimens of blood serum collected
from the Westchester County Hospital and the New Rochelle
Hospital. One hundred and one of these were from tuberculous
cases and 99 from patients suffering from various ailments in
the general wards. The second series was performed in the
spring of 1918 on 154 specimens of blood serum obtained from
206 HASSOW O. VON WEDEL
patients in three general hospitals, two tuberculosis hospitals
and a tuberculosis sanitarium in Westchester County, New York.
The third series was performed at the Walter Reed General
Army Hospital in Washington, D. C., during the summer of 1918
on 187 specimens of blood serum from patients in the tubercu-
losis wards and the syphilis wards of this hospital.
The fourth series was performed during the winter of 1918,
on 168 specimens of blood serum from patients in Bellevue
Hospital in New York City and three general hospitals in West-
chester County, New York.
TABLE 14
Report of complete study embracing all six series
Number of complement fixation tests made..................00000 6128
Number of sera, examined .... ee sce erie aie ais lelow & 0 oer eee 1207
Number’of patients examined ta jac cuics ees cis 6 «bees cisions cinernelone 1000
Number of clinically tuberculous patients......................00 484
Number of non-tuberculous patients. .................cceeeeeceeee 516
Number of all types of active cases giving + reaction...331 or 69.8 per cent
Number of all types of active cases giving + or —
FOAGUION «55's a cve coc 2c RET RAE a ctece rise cre 157 or 30.2 per cent
NUMBER TYPE* POSITIVE DOUBTFUL OR NEGATIVE
539 Non-tuberculous 9or 1.7 percent | 530 or 98.3 per cent
71 Suspects 11 or 15.5 per cent | 71 or 84.5 per cent
47 Inactives 6 or 12.7 per cent | 41 or 87.3 per cent
135 1 55 or 40.8 per cent | 80 or 59.2 per cent
52 2 36 or 69.2 per cent | 16 or 30.8 per cent
243 3 207 or 85.2 per cent | 36 or 14.8 per cent
58 4 33 or 57.0 per cent | 25 or 43.0 per cent
* Refers to the writer’s classification.
The fifth series was a special series of 318 specimens of serum;
196 of these were from the New York City Tuberculosis Sani-
tarium at Otisville, New York; the remaining 122 were negative
and positive controls from the general hospitals in Westchester
County. In this series, the writer tried to obtain serum from
100 tuberculous patients and 75 non-tuberculous patients on
two different occasions three weeks apart. The duplicate speci-
mens were taken in order to check the accuracy of the first
run of tests. He failed to obtain the duplicate specimens from
ae A Ne,
—— aa
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 207
a moderate number of the patients as they either refused or had
left the hospitals.
The first lot of 106 specimens from the tuberculous cases of
this fifth series was tested on April 24, and again on May 1.
At this time, only the names of the patients and the numbers
of the specimens were recorded, as we did not wish to have the
histories or clinical data influence the results. The second lot
of 93 specimens was taken on May 15 from most of the tubercu-
losis patients mentioned in the first group. In this second
group, only the numbers running from one to 93 were known,
but not the names, and of course, no clinical data; so again the
TABLE 15
Showing the percentages of positive results obtained in the different types
of cases in the six series of complement fixation tests
TYPE OF CASE ara SERIES aap cy ian ea Sa
per cent|per cent\per cent) per cent per cent |per cent
Normals (non-tuberculous)...... 0 2 3 3 3 hea
SUG}? BES) SESE ee Se eee 0 12 20 0 25 33.0
INTENTS 6 at Chicos eee EEE aoe 25 0 0 0 No cases} 21.0
Tuberculosis type l............. 31 28 30 20 60 23.0
Tuberculosis type II............ 100 | 100 60 | Nocases| 60 73.0
Tuberculosis type III........... 98 84 92 87 84 90.0
hupercwosis type 1V............) 27 0 0 59 80 60.0
laboratory findings could not be influenced by the personal
equation. The sera from the non-tuberculous cases were tested
with both groups. The results of these two groups agreed in
the majority of cases.
The sixth series is the combined work of all the special antigen
tests, Hecht-Gradwohl comparisons and odd tests not included
in the other series. The results of the tests in these six series
are given in table 14. .
The percentages tabulated on table 15 are the positive results
of all the tests made on the sera in each group and from each
type of case.
Table 16, series 1, gave 100 per cent positive results with the
type II patients (these are the incipient cases with some con-
208 HASSOW O. VON WEDEL
TABLE 16
Report of the First series
Number of complement fixation tests made...................0000- 1127
Num ber Of ‘Sera examMIN ed.) He ays eee litecis siete ale crohene Riera 200
Numberjof patients examined ./eeee. ope. hcticti ss «nth whe Se eee 160
Number of clinically tuberculous patients..................000000- 70
Number of clinically non-tuberculous patients..................... 90
Number of Type 1, 2, 3, and 4 cases giving + reactions. .61 or 73 per cent
Number of Type 1, 2, 3, and 4 cases giving + ornegative
TOACTION SS oh deccie iors Saspeeie hae SI ee ee etek ee ee 22 or 27 percent ~
PERCENT-
NUMBER | NUMBER hion Penteuk veut cee
fiotn vo to[y PEFRsghasenrere gl RODEEEEE |, wommtere | Boor tere
3+or4+
Non-tuberculous*... 99 0 0 0 0 0
SUSPELLS Shs Sunes: 4 0 0 0 0 0
Inactlves..-.. eaten 2! 12 3 0 25 0 0
| NE a oe ete ces 19 6 5 31 26 0
1S ees rae ER 6 6 0 100 0 83
110 ES Ee eae cee 47 46 1 98 2 90
Vinee td eons 11 3 3 27 27 100
* Refers to patients in the general wards of hospitals suffering from various
diseases and diagnosed as clinically non-tuberculous.
{ Two sera anticomplementary.
TABLE 17
Report of second series
Number of complement fixation tests made..................e cece 1458
INUMbBEr OF SERA CXAMIN EG. lees rs seein cere etone antes esas sh cree 154
Number of patients examined t) 2368. 2.28)... s Warten sean Stele a sce ee 154
Number of clinically tuberculous patients......................05- 70
Number of clinically non-tuberculous patients..................... 84
Number of Type l, 2, 3 and 4 cases giving + reactions. . .39 or 68 per cent
Number of Type 1, 2, 3 and 4 cases giving + or negative
TEACTIONG 31. ae deh he dic dete RTT LS oe Ie SE oe 18 or 32 per cent
POSITIVE DOUBTFUL NEGATIVE
Number | Percent | Number |} Percent | Number | Per cent
60 Non-tuberculous 1 2 3 5 56 93
23 Suspects 2 12 5 20 16 68
9 Inactives 0 0 3 33 6 66
14 I 4 28 5 36 5 36
4 II 4 100 0 0 0 0
37 III 31 84 4 10 2 6
2 IV 0 0 if 50 1 50
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 209
stitutional symptoms) and 98 per cent positive results with the
type III patients (these are the moderately and far advanced
cases in good condition but with constitutional symptoms).
Table 17, series 2, gave 100 per cent fixations in the type IJ,
and 84 per cent in the type III cases.
TABLE 18
Report of third series
Number of complement fixation tests made...................02005. 634
ECR Ee ERAMIINOU wn «o's as ne v bclob oka tae Oadesveessen 187
Spee CIE TICHGS CXATHITICN. ... 5. 5 20 onic Cle clue dtina cis ccteosece 160
Number of clinically tuberculous patients.....................2005. 77
Number of clinically non-tuberculous patients....................-. 83
-+Wassermann sera giving + T. B. results.....................0000: 1
-Wassermann sera giving — T. B. results... ...<.-6.eeseccesve secs 40
— Wassermann sera giving + T. B. results....................00000- 1
— Wassermann sera giving — T. B. results.....................20005 37
Number of Type 1, 2, 3 and 4 cases giving + reactions.. 33 or 67 per cent
Number of Type 1, 2, 3 and 4 cases giving + or — re-
Siem ME Se YER. cS ides 2 ssa s coving cet deans 16 or 33 per cent
DOUBTFUL AND
NUMBER = DORSEY NEGATIVE
OF CASES i. i aS ee
Number | Percent Number | Percent
—————————— | eee J __ |__|...
84 Non-tuberculous 2 3 82 97
30 Suspects 6 20 24 80
11 Inactives 0 0 11 100
13 I 4 30 9 70
10 II 6 60 4 20
25 III 23 92 2 8
1 IV 0 0 1 100
13 Unclassified and contaminated
Table 18, series 3, gave 60 per cent fixations in the type II,
and 92 per cent in the type III cases.
Table 19, series 4, included no type II cases, and gave 87 per
cent positive results with the type III cases.
Table 20, series 5, gave 60 per cent positive results in type I,
60 per cent in type II, 84 per cent in type III and 80 per cent in
type IV cases.
Table 21, series 6, gave 73 per cent positive results in the
type II, 90 per cent in type III and 60 per cent in type IV cases.
210 HASSOW O. VON WEDEL
Table 14. All the types of tuberculosis patients in these six
series including incipient, moderately advanced and far advanced
cases regardless of whether they were showing constitutional
symptoms or not, gave 69.8 per cent positive results; the sera
from the non-tuberculous controls gave 1.7 per cent positive
results on separate single examinations. Repeated examinations
of some of the sera from the non-tuberculous cases which gave
TABLE 19
Report of the fourth series
Number of complement fixation tests made.................20eee00 814
Num ber. Of SETS CXAMINE .) eee ENR (526 ol pol a nica onl Rn 168
Number Of Patients CXaMINE sei eieee mie ral lc icin «45 oroleracaasee cestanete 140
Number of clinically tuberculous'patients:)..... 2.0.0... 22. #2568 80
Number of clinically non-tuberculous patients...................004 60
Number of Type 1, 2, 3 and 4 cases giving + reactions.. 32 or 62 per cent
Number of Type 1, 2, 3, and 4 cases giving + or — re-
ACUIONS wise erro eels eed ELL Ee Creare i mosiaie 20 or 38 per cent
DOUBTFUL AND
NUMBER Poets NEGATIVE
OF CASES ee eee
Number Per cent Number Per cent
a es |, fe | | ee
72 Non-tuberculous 2 3 70 97
4 Suspects 0 0 4 100
1 Inactives 0 0 1 100
14 I 3 20 ul: 80
0 1 0 0 0 0
24 Ill 21 87 2 13
14 IV 8 59 6 41
39 Unclassified* 24 62 15 38
* Refers to clinically tuberculous patients whose history and classification
the writer was unable to obtain.
positive reactions on one or more of the tests were made. These
sera on repeated examinations did not continue to give positive
fixation, showing that the positive fixations were probably non-
specific due possibly to anticomplementary bodies combining
with the antigen and complement.
In the report of the first series, 73 per cent of all types of active
tuberculous cases gave positive reactions. The second series
gave 68 per cent positive reactions, the third series 67 per cent,
the fourth series 62 per cent, the fifth series 73 per cent, and the
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 211
sixth series 73.2 per cent. The average percentage of positive
results on complement fixation tests, embracing all the work up
to date, and from all types of active tuberculosis patients is
69.8 per cent.
TABLE 20A
Report of fifth series (Otisville series)
Number of complement fixation tests made.......................0.. 579
MEER E ONCE EWATIINICGS |... 5... - + 5 «2-00 ois slcia tn edlede neiycsccccceee 268
immer or patients €XAMINGd...............0.ccccceceses eet ae 185
Number of clinically tuberculous patients.........-..........0+ee00 126
Number of clinically non-tuberculous patients...................00 59
FIRST SPECIMENS| SECOND SPECI- |FIRST SPECIMENS| SECOND SPECI- |FIRST SPECIMENS
WITH NUMBER | MENS WITH NO.| WITH NO. 319 | MENS WITH NO.| WITH NO. 125
TYPE 170 ANTIGEN 170 ANTIGEN ANTIGEN 319 ANTIGEN ANTIGEN
+ = + - + - + - + -
Non-tuber-{| 1 | 46 1 | 22 0 | 47 2 | 21 0 | 40
culous.... 2%| 98%| 4%] 96% 100%} 9%] 91% 100%
eee 1 3 0 0 0 4 0 0 2 2
USPECES---\| 951 75% 100% 50%| 50%
16 18 20 12 16 16 13 19 | 20 13
eens 47%| 53%! 63%| 37%] 50%| 50%] 40%| 60%! 60% 40%
a 6 5 6 4 3 8 3 7 6 5
“seat ae 54%| 46%] 60%! 40%] 28%| 72%] 30%] 70%] 54%| 46%
tt 30 14 | 28 7 24 19 11 24 | 30 12
eat 49% 70%} 30%| 80%} 20%] 56%! 44%] 30%! 70% 71%] 29%
IV 11 2 8 4 8 5 6 2 10 3
° aay 85%| 15% 66% 34%| 61%] 39%] 75%] 25%] 80% 20%
All types of
tubercu- || 63 | 39 | 62 | 27 | 51 A Nasr 5201 66 | 33
lar pa- 61%| 39%] 70%] 30%] 52%] 48%] 39%] 61%] 66%] 34%
tients....
In the first series, 98 per cent positive results in the type III
cases was reported. ‘The results in the five recent series of type
III cases show 84 per cent in the second series, 92 per cent in the
third series, 87 per cent in the fourth series, 84 per cent in the
fifth series and 90 per cent in the sixth series. The difference
212 HASSOW O. VON WEDEL
in the percentage of positive results which have been obtained in
this type of tuberculosis patients is probably due to the differ-
ences in the classifications made by the various diagnosticians.
The results of all tests in these six series tabulated on Table 15
demonstrate the fact that the moderately and far advanced cases
in good condition showing constitutional symptoms (type III)
TABLE 20B
Report of all fixations test on the Otisville cases (fifth series) using the
No. 170 antigen
Number of type I, II, III and IV cases giving + re-
BELIOHS cos eee ne ene Salinas 88 or 73 per cent
Number of type I, II, III and IV cases giving + or —
MO ACELONS yi oidin teens aeRO eCPM oF aac esos ehovote 33 or 27 per cent
RESULTS
TYPES
+ —or=
Sera from non-tuberculous controls................. - es
3% 97%
Suspects: (doubtfiulicases):5....5 basse erect ae : 3
En as SA 25% 75%
- 25 17
£495) NAAR FO OE, SOLAR UIE TER axl 13 AR eR Ly Ad 60% 40%
7 5
1h Gis LR 2 Ren SN as Pa IE ECS Fv acy tae Oi IN SA A Ae
{) coe | ate
43 8
1) 0 BUSS) Se AE Ins iene BS eo SA ee Bea e Siler cs 84%, 16%
TEARS NG as Mircea Vermin apn ASC OE! RRR AN {| 38 e
\ 80% 20%
gave uniformly the highest percentage of positive reactions,
and that both the incipient cases with constitutional symptoms
(type II) and the advanced cases with very poor prognoses
(type IV) gave irregular results. The patients exhibiting few
if any symptoms regardless of the stage of the disease (type I)
gave uniformly the lowest precentage of positive reactions. The
clinically non-tuberculous cases gave only 1.7 per cent positive
fixations.
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 213
TABLE 21
Report of the sixth series
Number of complement fixation tests made..................+-.05- 1516
MEG MMERELATCKAMIIMNOG.. 0... .... cc ee ec ccaheccctaedeneuncascees 230
MUMNCrO! PALICHtS eXAMINGG -....... 0. oobc ode da me adeces cutee. 201
Number of clinically tuberculous patients......................04. 61
Number of clinically non-tuberculous patients..................... 140
Number of Type 1, 2, 3 and 4 cases giving + re-
LOGS Ms 2) os Be On Sr 41 or 73.2 per cent
Number of Type 1, 2,3 and 4 cases giving + or —
TECHS Oey AS Cane ae ee pene! Bee See cy 15 or 26.8 per cent
DOUBTFUL AND
NUMBER TYPE pian ts cil atte
Number Per cent Number Per cent
154 Non-tuberculous 2 1.4 152 98.6
6 Suspects 2 33.0 4 66.0
14 Inactives 3 2120 11 79.0
9 I 2 23.0 Ul 77.0
11 II 8 73.0 3 27.0
31 TT 28 90.0 3 10.0
5 IV 3 60.0 2 40.0
15. THE VALUE OF THE COMPLEMENT FIXATION REACTION TO THE
CLINICIAN IN THE DIAGNOSIS AND PROGNOSIS OF TUBERCULOSIS
When all is said and done, the final answer to the question
of the value of this reaction to the clinician will be given by the
clinicians themselves in either using or discarding this aid to
diagnosis.
The tuberculosis specialist will probably say, ‘‘We can make
our own diagnoses with the aid of auscultation and the signs and
symptoms of the patient in practically all active tuberculosis
cases except the very early ones with lesions too small to detect
and with practically no symptoms.” They will then ask, ‘Will
this test give positive reactions in a large percentage of our very
early cases with practically no symptoms?” The answer is,
‘“No.”? In such cases the fixation reaction will be positive in
only a small percentage, possibly from 25 to 40 per cent. The
specialist will also ask, ‘‘Can you be sure of your prognoses in
all cases by reading this test?” Again the answer will be,
“No.” It will be of value only as a prognostic aid when it is
214 HASSOW O. VON WEDEL
added to the clinical data already available. The above two
answers will cause many of the specialists to reject the test as
of no value to them except as confirmatory evidence.
However, this fixation tests will be of the utmost value to the
general practitioner who very frequently calls a case incipient,
when the specialist would classify it as moderately advanced.
Many cases, with small lesions but with moderate symptoms—
hemorrhage, night sweats, afternoon temperature, etc., come to
the practitioner. No tubercle bacilli can be demonstrated in
the sputum and he thinks, but is not at all certain, that his
case has tuberculosis. This is the type of case that the comple-
ment fixation test will pick up and by repeated positive findings
prove the presence of an active tuberculous focus.
Again, as an aid in prognosis this test will be of marked value
to the average clinician if he uses it intelligently along with his
clinical findings; for instance, if he follows with the fixation
test, at rather frequent intervals, a moderately advanced case
in apparently good clinical condition, a weakening of the reaction
from a strong positive to a weak positive and finally a negative
reaction is a very good prognostic sign for the time being. Then
as everyone knows some outside factor may later complicate
the case and light up the healed lesion into fresh activity. The
complement fixation reaction under such a condition would at
once become positive again.
Or if he follows a far advanced case in a very poor clinical
condition; the sudden weakening of the reaction from a strong
positive to a weak positive or negative reaction would be a very
poor prognostic sign, indicating the loss of resistance; 1.e., the
power of the body to elaborate antibodies is lost, or these anti-
bodies are overwhelmed by the excess of antigen thrown into
the blood stream.
The clinician will ask to what degree in each class of cases
does the complement fixation reaction help in diagnosis and
treatment beyond clinical and sputum examinations? To this
the reply must be—it will be of very great diagnostic aid to the
average practitioner in the incipient cases showing symptoms
but with no or only few definite signs and with negative sputum
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS PAYS
reports. The reaction in such cases will be positive in probably
70 to 90 per cent. :
In the moderately and far advanced cases in good condition
showing symptoms, it will be only of diagnostic value, of course,
as a confirmatory test, being positive in from 80 to 90 per cent.
In these cases, if made frequently and used with the clinical
data, it will be of great aid in prognosis.
In the advanced cases in poor condition it will give very
irregular results and is of most value as a prognostic aid.
In the cases having few or no symptoms, no matter what
stage the disease is in, the reaction is practically worthless, as
the percentage of positive results is only from 25 to 50 per cent.
In the general summing up of its appearance in all types of cases
(active) one finds that it is comparable with the Wassermann
reaction in syphilis.
Gradwohl (101) states that he, like others, has been forced
to conclude that a positive Wassermann test is laboratory
proof of syphilis, but that a negative Wassermann is by no means
a method of proving that syphilis does not exist. He concludes
that of all treated syphilitics examined, only 60 per cent have
given him positive findings.
Snow and Cooper (102) conclude that the sera of non-syphilitic
tuberculous patients may give partial to complete Wassermann
reactions when cholesterinized antigens are used in about 31
per cent of the cases.
Lewis and Newcomer (103) state that the Wassermann
reaction is positive in other conditions, but that it is not so
generally recognized. They cite fresh instances of this in cer-
tain febrile cases. They also conclude that a positive Wasser-
mann reaction fails to appear in a considerable percentage of
syphilitics.
What the future will bring is hard to say. In all probability
no better results can be expected from the complement fixation
test based on the specific tuberculous antibodies in the patient’s
blood serum than have already been reported by numerous
serologists.
OMI om oo ty | NUMBER
Clinical and laboratory report of the Otisville series
NAME
C. Nicolelle
E. Klein
Bessie Horowitz
D. Smith
V. Kreis
C. Zikarsky
R. Burns
H. Shapiro
M. Rosch
M. Berman
L. Gebbardt
C. Sullivan
C. Kircher
R. Anerheim
Mary Cleary
C. O’ Donnell
E. Wagenbach
A. Rifkin
M. Kenny
R. Gachino
N. Donohue
A. Drennan
A. Kiley
J. Frey
A. Wood
J. Healy
M. Bachman
M. Kroin
H. Ferrera
M. Mulhalley
M. Clyner
B. Wolff
C. McGovern
A. Constantine
A. Ficearrota
J. Kreiger
J. Rossling
O. Ogren
HASSOW O. VON WEDEL
SPUTUM REPORT?
FIRST OTISVILLE
CLASS!
TABLE 22
SECOND OTISVILLE
testes aly deste ot) al
Veesteteaal
++i tt+++ 1 t++4+4++4+01
De
We
WRITER’S CLASSIFI-
M
=]
n
ue)
PWR PW PRE RPE WW PRR RW ER Wr wee Bee wee
REPORTS
WITH NO. 170 | wirH No. 319
ANTIGEN
specimen
First
Second
specimen
As:
a
REPORTS
ANTIGEN
specimen
First
Second
specimen
REPORTS WITH NO.
125 ANTIGEN
Ww Le)
ith |
ww
HR b+ ++
He bo
++
He He He He He e em De ee
+P te tot ttrr+
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 217
TABLE 22—Continued
a i REPORTS REPORTS 6
3, a R = WITH NO. 170| wirH NO. 319| Z
5 3 Ss a ANTIGEN ANTIGEN i z
a 5 L 4 =e
a = = $ ze g § 8 8 E I
B) gs |ge| gs | BB| 88| #8| 28) és
Zz a & n Ee & mn n a
39 | J. Kaplan — | 3B 3B 3 _ S| oy ss
40 | A. Higgins ey te a eee’ Tse gts hae
41 | E. Roder + | 3B 3 4+ 4t 4+
42 | R. Antonio +] 2A 2A 1 a -- =: — | 3+
43 | F. Maklary + | sac 3C + 4+ )/3+/, +] 2+) 4+
44 | J. Becker + 1B 1B 2 + - - - _
45 | J. T. Searoni +} 2B 1 2+ + | 4+ +} 2+
46 | H. Hukenson +! 1B 1B "2 4+ /44+ | 44+ | 44+ | 4+
47 | J. Brown + 3A 3A 1 4+ | 44+ | 4+ | 44+ | 4+
48 | V. Olsen + | -1B 1B 2 2+ | 34+ | 2+ — | 2+
49 | W. Blay + | 2B +] 2B 3 Fi +) =] — | 3+
50 | G. DeShay +]°2C |2C 4 4+ |} 4+ | 44+ | 44+ | 44
51 | J. Malcolms — | 2B 2B 3 qe) Sae 4) Se teas) ==
52 | A. Miller —| 2A 2A 1 3+ | 4+ | 2+ | 3+ | 4+
53 | J. Coffin _ 2B 2B 3° 4+ |} 44+ | 44+ } 2+ | 4+
54 | S. Hickey Fel M2A 1 See eee |) toe
55 | W. Trabuci +} 1A 1 4+ | 3+ | 44+ | 3+ | 4+
56 | H. Hageman +} 2B 2B 3 2+ |] 3+ | 2+ = +
57 | H. Kildea +] 3C 3C 4 - — _ — | ac
58 | J. Schneider = 2B 2B 3 -- _ + - -
59 | L. McGann — | 2B 2B 3 —} +] +/ -| +
60 | J. Dunn + | 2B 2B 3 — + - +] 0
61 | J. Atty + 2B 2B 3 4+ | 44+ | 2+ | 2+ |} 2+
62 | A. Glassheim + 1A 1A 1 4+ |} 24+ } 44+ 4+
63 | A. Larghi +} 2B 2B 3 4+ | 3+ — + | 4+
64 | L. Weisberg +] 1A 2A 1 +/] —- — —| +
65 | E. Anderson aa 2B 2B 3 _ + - _ -
66 | H. Swiss +] 3B 3 — = —
67 | A. Steislet +} 1A 1 3+ 4+ 4+
68 | A. Torak +} 4C 4 4+ 3+ 3+
69 | C. Seedyk +] 2A 2A 1 |+tac} +/ —| —| 0
70 | J. Alletsee + | 2B 2A 1 2+/3+/ +] —]| 3+
71 | F. Frascino - 2B 2B 3 - — — — | 2+
72 | H. Slusek + 1B 1B 2 + + — | 2+ a
73 | S. Naroty _ 1B 1B 2 2+ | 2+ o — =
74 | G. Cleaver _ 2B 2B 3 2+ /3+/3+;) +); +
75 | H. Franklin — 2B 2A 1 — + - _ ~
76 | S. Scott —_ 2B 2A 1 + _ - — ==
HASSOW
O. VON WEDEL
TABLE 22—Continued
NUMBER
NAME
H. Lavelle
J. Willis
G. Gould
R. McDonnell
N. Kelly
W. Day
J. Pestrok
J. Patten
J. Como
M. Chipront
H. Robinson
N. Cheturachino
J. Kaminski
E. Murmane
W. Kirk
B. J. Walsh
H. Pearson
A. Bauer
J. Walsh
G. Murganeo
H. Darci
W. Hunt
J. Christ
W. Allen
P. Masterson
F. Bittanto
J. Devlin
J. McDonald
Geo. Early
F. Cahill
Ed. Sweeny
J. Reilly
J. Outney
G. Bowers
W. Liber
Cleary
Wm. Pienery
J. Mahony
1+4+44+14+41'1t4+44 144+ 1 ++ | seorom revorr?
b+t++4+41
FIRST OTISVILLE
% %
‘ ; e |e
cS) a) 3) aS
ay
2B 2B $
2B 2B 3
3B 3 4
2B 2B 3
1C il ¥ 4
1B 1B 2,
4C 4
IC 2B 3
2B 2B 3
2C, 2
1B 1B 2
2B 2B 3
2B 2B 3 4
3B 2B 3
1B 1B Ps
2C 2B 3
2B QA 1
2B 2B 3
3B 3
2B 2
2C 2C 4
4C 4 4
3B 3 2
3B 3 4
3B 3B 3
2B 2B 3 2
4C 4 3
NTY| NT] NT
ING Sale| NL
Ne ea oN ic Leea tc IN pede 2
NaN) aN
INGE oe N Ga aN
ING Ss SINGlea|) Niels
UN Ten IN ele | Na
INGEN ee Ne:
WAL | ISA Nay
NT | NT] NT
IN IN ad aNd
SECOND OTISVILLE
WRITER’S CLASSIFI-
REPORTS
specimen
L+tittit+ht++ur
rRI+H + |
REPORTS
WITH NO. 170 |wiTH No. 319
ANTIGEN
ANTIGEN
ws) rs
+FI +I FREI REET
iw)
Pwo -
Peeled lect al east ote al ae cle steoeteeal iat
Second
specimen
REPORTS WITH NO.
125 ANTIGEN
COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 219
TABLE 22—Concluded
REPORTS REPORTS
es a 3 2 WITH NO. 170 | wiTH No. 319 g
E 4 5 @ ANTIGEN ANTIGEN 3 =
NAME ce | | Smee bg | ee
a a/ 52 |eb| we | 8) e8] 2] 28 | EF
a n & D e i) 1 VL‘98°%) 0} 12 ‘ularyn[a38
0 | I>\eT IFS 0} O| O01 |#'s] 0 | 02 Gt jo'¢ jer let [o's] St lhe jos | Gs} 2% | {poolq asz0xy
OYSre He Sale SML bs) Om) Tz “ulurynys
0 | I |r Jo's 0/61 161 (9'8 16S | 0} 02 LGG|L TZ/€ ET]O SL|O 6T| O'ST|O OGI2Z°Zz] L°Zs] Sse -38 UO[OD
OM ONO OL aDs| On|) 12
0 0| 0|F'S 0} Oo} O|] Of20] 0} 02 8ST 6% |S 1 IST [o's | St jos jes} 2:2 | oF |) “Urugnys3e
€/O1 \&/z |8/s }€/€ €/L1/8/F1/€/018/L j€/¢ |¢/€ B/T |€/82)8/G0]8/Te}e/21| &/F1/€/0tle/2 | €/9 |*8/e |f Pola wade
0] 0j22 19 F} Of] OF “uluTgnys
0 |e} 11>] I>] t>let jes 0} 0|€'%j|9'F | O| ST 6S IPE |F'S\e 91h'S 16'S [0S JOS O'S 16's JOS |O'F IFS 1#'S | O'S jo's log | 0G] 08 “5B UOlO—
bel otal aaa) > el SC Ret a | (Re KU ee Pee | Is ac
0| 0} O |g°0>/9'0 |z°2 |o°Or O21 ier [2% | Of ST 2s O'S JOS GSS [L's |G°s |G°s |G°s 10'S 12's 19's IL's |G's | O'8 o's |9°% | OF | O'F pong ss10 77
€/1e\e/L1\8/Ft | &/otle/z |8/F 18/2 &/F1e/0Te/2 |8/F 18/6 /16|F/90|F/816/F/ F/T |8/8C/€/ S| E/13|8/L1/8/F1|S/0T\S/2 |8/F le8/] Z/FSIZ/2112/9 |11/ze| 6/S%
0 0 {80 j8°0 Be Ls 0% 6% 0 | &I [P= | ml eet ea ee Pel alee Sa _«|) -uangnysae
0j07)| O 0 |FT eT ge 9 F |T O88 | 0} at LGNE CLC [LS 19'S 19°S 19° FS l2-% |S | $s jes |¢°F | Hg | e's pagid -esrogy
€/€ |%/82\Z/Le |€/12|8/L1/8/F 18/11 }8/L |€/S |&/% |%/82|2/22 /L |F/% |€/8Z|€/SS|8/TS/8/L1E/TTE/2L |€/S 18/8 |aB/L2|0/S2|Z/21\Z1/2T| 6/S%
g . 07} OF “Udy Ny[s
0} 0 \¢7< 010T!) 0/6 BG IGS jos o's o's [o's G's [o's |8's [G's | HS Es OT] BS] SS lps, proyday,
£/€ |%/98|2/F% @/86|6/96|6/ FE €/08|6/82|£/ T/S/LT/S/FTE/TLE/L 18/8 1%/831S/9GlaS/FOIG/C11S 9 |11/ 2s] 6/26 :
0z2! 0 PII9TIZT] 0/8 aratni
CATON E Leataeny |O Tenant |) LOnley fe [9s |9'e 9's |9'€ [FF O'S [OS] HF IPO PF / OE) HS] DS ptoydy,
£/12|8/91|8/11/8/8 |%/83|%/Se}2/82|%/ 12 €/02|8/9T|8/11|8/2 18/8 |%/82|%/Sa|0/e|4o/12/Z/211%/9 |e] 6/S% :
ugg
-WON Git
41IW WOUGS §,d1y4
WOYAS SUAHLOW
1 @TaVL
uod
GQaVWN L8aL
1Z
pue
06
61
pus
8
ae
pur
él
230
“UIUIIN [IIB OU SBA JY} JOPULBUIAI ayy UT !UOATa
a 218 $380} JSIY 94} ATUO ‘aouds Yon 003 dn oy¥4 07 JOU JOpsO UT “YIATG 109)" syQUOUT 99147 JNOGB [GUN Poyso} O1OM SY[TUL [|G PUB Spry []V ‘PZ pus gz pus ‘g puw yY Jo uoTydooxe oy} YIM 7970
i. $< 8 8
0] O} GC] O|] O} I>]9 1] 0] 82 “uTayNs
0 |0'T 0'F OF) OF} 0!) “0 40141 0) 190 1% ET Ie OT Ie0 | 8 t lt i¢'t | er | et | [-38 proqdaéy 3
GO| 0) 0} 0 |G8T IN’ It'2 | 07 82 “UTUTgn]aIB 82
0 |¢°0 {29°0]8'8 Cell 207] 200}! ONT OFT: 10%% 1 O.|| 22 UT |O'T [O'L JOE Jet | 27s [eT 12's | fT | O'% | {poolq ossoxy | }pus
XG
0 190 |6°0 |6'& |a'F 19°F [69 I8'0 | 82
0 {|¢°0 oS 10°82 0] 0 |9°0 |9°2 Jo's jot It'S | Of} 22 4°812'8 |9°9 12'S J9°9 | €°8 |e'8 |o6 | 9'9 | e's “uTUIy Ns
/L1 |8/et |¢/otle/2 |8/¢ ¥/o1|t/# |8/szl¢/or\¢/er|e/orle/2 |e/¢ | ° b/IT|P/¥ |€/8e/€/91/8/s1| &/Otle/2 |x8/S} @/21|e1/0z -#B WO[OED
: 0 | $3 uTUgn sae
G'0>/9'0>| I>|F 1 | ¢< Cte OuNed LG 19'S (0S | 8's |o'F IL°% | 8's | S's | (poolq gtaqey
‘2 0 | ¥ uTUTyN] Ise 14
a 0 {9°0>|] 0 |0°% Jo's Pach |me Onl Be 912% |9% | e's jee [e's | 08 | OTT] {poolq osa07{ | }pus
£3
S/LT |€/11 |€/z |8/9 18/6 G/LU18/11/€/2 | &/9 |u8/8)2/¥C] G/ZI|IT/ee|) -8B wojog
232 . G. C. REYMANN
found by Halban and Landsteiner in the mother’s serum; and
the titer of the milk is everywhere higher than that of the
mother’s serum, so that it is possible to speak of an accumulation
of agglutinin.
In this connection, must be noted the researches of Wegelius
(20) on the laws according to which antihemolysins and agglu-
tinins are passively transmitted from the mother animal; he
found that they were always transmitted, but that the anti-
hemolysins were not transmitted with the milk and Morgenroth
Colon agglutinins; units Per, cubic centimeter
So®
o--—-—-———» Kid No.20 (serum)
i Oe ae ewe Kid No.21 (serum)
O-----———@Mother's serum
o-——————o Milk.
CuHaRrT 1
and Braun give asthe result of their critical examination of this
subject that an assimilation of the antibodies of the milk, it is
true, has been proved in new-born young, but that it is not a
regularly occurring phenomenon. It must, however, be borne
in mind that this criticism applies to antibodies, which have
either been produced artificially in the organism through im-
munization or have been directly mixed with the milk.
The figures quoted in table 1 are, in their reciprocal values, of
such a homogeneous character that a graphic representation of
one may serve as a paradigm of all (see chart 1).
In order to determine the transmission of the agglutinin
through the milk, nursing experiments were undertaken; they
TRANSFER OF SO-CALLED NORMAL-ANTIBODIES 233
were undertaken in such a manner that one kid of a pair of
twins was removed from the mother before suckling and placed
on boiled cow’s milk, whereas the other was permitted to suckle;
in another experiment. raw cow’s milk, free from agglutinin was
used. The experiments are given below in table 2. Milk
samples were taken from both the nipples of the mother ani-
TABLE 2
COLI AGGLUTININ UNITS PER CUBIC CENTIMETER IN
- , |Serum of the
Serum of the kid wae : oa
iather's aeenaeid tear kid which Milk from Milk from
seen” | demetirany | Sikes | ‘Sie | “tha
Experiment 1
The kids born 3/19
Samples taken 3/19 18.2 0 0 58.8 SO
3/20 15.4 0 ye <5 igs
3/21 20.0 0 5.9 — 10.0
3/22 15.4 0 5.5 2.5 Ct
3/24 15.4 0 4.4 2.2 ay |
Experiment 2
The kids born 2/18 | Fed only
; with raw
cow’s milk
free of ag-
glutinin |
Samples taken 2/18 pil 0 | Be 15.4
2/19 5.9 2.0 4.0 1.0 10.0
2/20 Bk !
‘poi0y[gy pue soqynurur
MO} 1OJ UdYyeys ‘UOTN]OS 4]["¥s Ur dn paysem oa10M soMy[Nd Iedy
UOTyBy
€|-15 ‘spoyjyow omy, | pus
Voy] UIZz10y
-[eYoley Younp pun
UjeyNyYs usynulyyYy
es1ula =: Ua. [NY
-UdSSB]] OS|B ‘ poi94
-[¥ pues oinjessd
-W19} WOOL 4B NOY
IT puvys 09 pomoy
-[8 pue ouljes ur
| dn poysem oungpno
gop “d
‘Z “JOA ‘S06T
‘1aquia00q
“uinor ‘pay
"WI ‘PPOL
“410 “O07 S100
snely
281
HE JOURNAL OF IMMUNOLOGY, VOL. v, NO. 3
a
0G ‘[o4
‘6061 «= “311Q.
UIxO}1}UB OU seoNnpoid sBId-veuINS sTDY yoy =| “AYBVE "F°yUID
UIXO} OYJ, ‘“UIxO} YIM pojoefur ueyA ATUO 4ynq WOT}OB oIx04 ‘UUBULIOSU()
-1JUG INI} AVY 0} PI DAOT[OG “WNIIS 91X0jIVUB SNBIS OY} PAMoYg | pu IIo
sinoy
Sh 0% FZ UdsAIZ
4oey A[UO OUT} UOT}
-eqnouy += ‘“peleq[y
pus sXep Z poeyeq
sisA[eied =| -nour ‘uorynjos
ou ‘4ySIem | 48S Ul YO poysem
Inoy jO ssoy ‘vaya =| ‘1838 UO UMOIS UIST
T Oy ‘9.06 -ieip ‘+++-+ |-uvd10 ‘wixojopuam a
Surpueys uo A
UOI}BIOTIO} 61 ‘d ‘TE “JOA
: -op ynoqe ‘sfep $0} smnoy | ‘Oz6T “Pe
surqyou Avs SUOISe, =| Mot uoTZeqnoul ‘dxq = ‘uanor
uosiod jo sosop) | s1oyyne f41 ALOU BIBADS =| f pale} yOIg 390 ‘(e31y9)
18y9°T OOOT JO uory pet annoy puv sIsd]Bi OUI[VY]e Ul YIMOAS ‘19ST
-ezyeqynou § oytoedg|][ JO} “O,¢) -ed ‘+444 |Aep G¢ ‘ulojoxry pus AY¥szTIO
said-voumr s}Iqqey
SHUVAGH NOILOOAGOUd NIXOLIINV ALITIAVIS _ SS en ee LENORE OL EE Oo Cope rey A eee
uod ALIOIXOL
panuyuoj—hajuashq
, 2|GBIOABJ SUIOUT,, s[vulUe 944
St yt Aysnoouvypnuig possed ‘ds jr soyst Lg ‘d ‘0% "JOA
$30030.1d 41 uOS -ulmIp AOIxOT, ‘OOBT ‘1ne4
-10d a10jaq pojoofuy Peg ‘d ,,{po ‘sdjoy u01}BU -seg ‘4suy,[ 9p
sesop -wad = uOTy -o3AxKQy) “ABP UIP ‘uuy ‘1orueq
Q sozI[vigneu ‘00 Z -eqnout ON,, | uo sulseq AylorxKo T, pues neig
sasop WINS , 01X04 sdid-vouins
— Ssozijeaynou ‘09 [| -1yUB,, SesOp [eyye, | 10} uBYyy AT
sosop |Z 0} T sezesqyneu | -prder a10ul
€ soezyBiynou OZ/T| O191A ur UOstod oYyy = | 4sO] SzIqqvI4 aAry posn so}Bi4[y
sosop | YYIA poxiur undies | I0j AYWI -ed0U 901 Aep 2 jij sanoy
Z soezyeigneu (G/T | [BULIONy *O1x0O}Yy -x0O} jj] seyn f+++-+ Woy} SIF, |g@ JI0OJ “9,09 48 eer d
?(SUIXIUI 109jJ8 =| -UB SB IJOM SBlBIIAy | -UTUT QZ I0J ‘as10y A[snousaeiy =| poyeoyyT “UINJaS ‘9¢ ‘0A ‘FO6L
2g y¥8 anoy %) wma | -oeqiyue st uINnieg ‘),021 Aq | pueSoposye | -ul GTO} GQ | as10y puUB poo|q “101g "009
-8S eUnWIWIT as10y UOTVEZIPBAyNoU Jo paAo.14sep ff+++ ynq ssid 10; as10oy poeyeuliqy "YC ‘1erueqd
1I9Yy} OAZIA UL poxiz | VoSep oyesIpoOyY ‘sag | fapqeys AIOA ‘99 G°Q | UBYZ SSE] YON | -ep JO uNtpeur poss) pus neig
D19)0Y4,) z,
"D09E 09 ATUO poyvoy .
pus 4S YIM sony
-[ho sty yO poysea
ey souls ‘saIpoq
T1299 9q} OF UOrZIp
-pB UI SAUIYSBA Iv3de
UI peuleyqoO seouRys
-qns eqnjos 2y4
YIM 4 SULMOUY yNO
-YYIM Suryiom useq saIpoq sn]
easy AvUl OFT “you -[loeq e[OYM pol] ty
ieyj3o §=9yy «6 ‘unas oy} Ul pepnyo
snely oy} Aq pez -ur Yyyoq ‘snuse.
-[B1Znou sv UOTZOe suos ++++ -BUl 19Y4O oY ‘SIS
a1yoied oY} YeYY YOR -10d o1jo.red ysurese ! -Ajeied = soonpoid 1z °d ‘1g ‘Joa
ey3 Aq suosiod omy «=| WINIOS sneIyy = 8y4 ++++4 yorum ouo ‘suos ‘TI6L 3=“VIRA
84} PeyeIUeloyIp oFT | Ul = UIxOyWUe snIT ‘sqiqqey |-1od omy spug osty | "J “QO ‘nesseg
SYUVNAU
parery
-UBJSQNS [[OA\ OU ST
I9SUOT 10
iInoy Uv IOF “—).6E
48 pueys 0} poMo][e
pus poxify “sesop
[847] OL Sv YstYy sv
UIxXO} 9} poZzt[ery
-noeu yevos v JO WING
OLjIA UL Ppoxrul
ysoq ‘UIXO4 104Fe
Seynulud QT UlYyyIM
peyoolur oq ysnut
WIN.IEg ‘90199P
aUIOS 04 pozI[B1yNIN
“‘Oo0L
ulle[d yng ‘sak suTIe[D | 98 paAO14seq
sn290909)da.L1,9
(4) 7YSIOM
yqqet
SITY ¢'0 04
WeAIs JON (4) | Se0 ‘++++
sanoy G UI YYvep fsesop
oB1V] YJLM SoyNUTUL YE 0} OT
:OUIT} UOTYBqnoUl 4LOYys AIBA
SdyNnuUIUL YE
Ur “D0L +8
PEM ELM
A[poyreuL SINOY FZ
‘ATMOS | 0} G WO] UL
S9}VALOLI9}OC] | -YIM “00 G04 T
NOILOOdOuUd NIXOLIINV
sdid-vaumr
ALITIGV.IS
ssid
-wouIns 105
UBY} OIXO} Sso'T
suqqvy
YOu ALIOIXOL
panu tuo Q—v1a7]0 UD
pe1ey[y pus sXkep g
UMOId so} A000N9]
sid-vourns yyIM
yjoiq ulone, pues
[[JOo00A4JS JO wn
-Ipsul pozeorpduio0y | uuy ‘yo1ourreyy
po
-yjour proyddy sty
OESse HO se ven
od ['Q Ul (UTxXO}0
69T “d ‘OT
‘TOA *Z06L ISB
LEZ “d ‘OT “JOA “OG8T “48¥q “4SUT,[ oP
‘uUYy ‘HOYIUYo OJ YIM UOqUTI][VG pues XNOY JO YAOM AaI].AvI OSTe vag
COE “Sh [OA
‘9061 “JOE “J
ea
-x0) B[taids punory | “9g ‘ueApeyoeypy “
skep @
se Ajiva se saved
-dBe uU0lyeiy[y Ur
SSO] ‘syoom Z 04
sAep ¢ ur uinutrTydo
‘auoyded yyouq Joog
NOILONGOUd AO GOHLAN
Ost “d
‘T ‘Joa ‘yonq
-pusyy uuBul
-19SsBA, pus
soy ‘snery
SaLVa
aNV SYuadwuomM
ee ae Se ee ee
UIPIloT[90 YAnoIYy
poezA[sip sosoq.
‘uolgonpoid = urxo}
Joy Aressooou = AT
~ON]OSGB UTqO[ZoweyP]! [Vy] 9 07 doUASISOy
19415 Aq
uosiod jo uormdios
-q@v {UOIyBI011030p
e10Joq '480q ouITy
qysois uInUIIxeyyy
*pesn oq ysnur AZ19
“Ud Y}AMOIZ 4¥013 Jo
SUIBIZS SHUTYy} UNeIg
pezdu19448
UOIZVZIUNUIWIT BAIS
-sud ON ‘“padriay
-u0d seing[nd Zur
-Al] ysurede Ayrtunu
“Wy = “UO1yBzrunuUT
~ul 9ATyoB Aq sesop
xoq
-901 9Y} Ur
sABp Moy B
Ul poysruru
-Ip ‘seynuru
O€ 1OJ °D.0¢
48 paAo14s
“op AWOIXO,
-udul
pouoy
JON
SINOY FZ
0} FI ou
moIyeqno
“Ul Peet
WB1ZO[TY
Jed “00 [ 07 ¢'9
, SOOUONPUL [VUNG [NJWIBY 130430 10 sayho
Oona] ‘seyepnxe [BUIIUe LM 4987009 UT ATMO
suostod svonpoid 1000904da,q4q,,
pus uonsimeue sAvMlTy
*BaYLIBIpP
‘s}[NSeL IepnFo1I17
‘sAep 6 0} Z% J09JB posoq[Yy Yorq put oyep
“NX9 YBUOINI[e WO soinj[Nd oIqo1eBUE UT
peonpord syqqes pay[ey Yory urxKoy, “urs
-Ajoulay pUB UIxXO} UdeMyoq drysuoryefer ON Serpoq url suostiod ON] "Jy “Og ‘suourIg
pues uoneypgur [eoory
s}usTUTpag
pide 104
-eqnour ul
woryeIOlIoj0q | UO yoyo ON | -sI
s#id-voumn3
Sinoy QPF Ul UOYy
-B1}[Y 9soon,s yua0
Jed g'Q pus poorq
pozeuriqyep YyLM
UOIZNIOS s,ayvorT
aT peyearyny
"19900 Jo
suosord 94} [[B Jo Ay10rx04 oy4
peysturuttp “H,0L “Apoq snoo0000;da14s oy} ul Aq1oTx04 qqsI[s
AI9A “plo skep F 04 Z% WoIy soqzeI4g YIM SyyBep [BUOIsBI00
u01}8
-10eule ‘vayt
-I8Ip AMOS
IO sInoy FZ
Ul + aB[ndo.r
t+4++
‘S}UNOWV BF1B] UL USA VATYBBOU ATAITyUd
"P9sN B1OM “09 OZ 07 OT UsyM ydooxe oAtyeSou Aporyy | ‘jog ‘uUosuo.y
soyeaqyy anoy OT | J “DO
SPOT ‘TL
‘TOA ‘SI6I “V
‘W ‘Vv at ‘104
-udT pue YARD
Gg JOA
‘F061 —"qoeg
6 “104 ‘ZO6T
“OOM sf
ese “d ‘Tg "Joa
‘ZI6L «oR
‘uneig
a ee ee
286 HANS ZINSSER
older extraction poisons or direct sediment extracts have failed to
cause paralysis, but have given rise to intestinal symptoms and ma-
rasmus. And Arima’s work with typhoid bacilli gave exactly analogous
results with these organisms in that his so-called typhoid exotoxin
gave him paralysis, while the cell body poisons produced intestinal
symptoms and marasmus. ;
As regards the serological work done to prove the true toxin nature and
the specificity of such poisons, there can be little doubt about the fact
that a great many observers, especially those who have worked with
typhoid, dysentery, and cholera poisons, have produced sera which
neutralized the toxic substances to a limited extent. It is worth not-
ing, however, that in a great many instances there has been a very
marked difference in the results when the poison and the serum were
injected separately, though simultaneously, or when the two have been
mixed and kept in the incubator for from thirty minutes to one hour
before injection. Very powerful neutralization has been observed
only in the latter case, and since in most of the poisons the presence
of true bacterial antigen cannot be excluded, it is by no means impos-
sible that many of these neutralizations, at any rate, depended upon
union of dissolved bacterial antigen and its antibody in the incubator
before injection.
It is noticeable, too, that the sera produced with extracts of typhoid
cell bodies by Besredka possessed neutralizing action for the poisons
obtained by Yamanouchi in seven day filtrates and by Kraus in nine
day filtrates, Also Bessau, using as poisons the killed bodies of the
dysentery bacillus, found that Kraus’ dysentery serum produced with
filtrates and filtered agar washings neutralized the paralytic poisons of
his endotoxins, but not the marasmic effects.?
IV
Our own work was undertaken because we felt that there was
much unclearness regarding many of the claims that had been
made by various investigators regarding the so-called exotoxins
of bacteria, and we believed that no immediate progress of im-
portance could result from further investigations of the endotoxin
problem until the existence of true exotoxins and their proper-
ties were definitely settled. Also we were induced to take up
3 For a possible explanation of this see remarks in tabulation under Bessau.
NATURE OF BACTERIAL TOXAEMIA 287
this matter particularly because observations made by us with
toxic filtrates of haemolytic streptococci indicated that there
was a striking qualitative similarity between these and the
B. influenzae filtrates produced in parallel experiments by Mrs.
Parker. We therefore decided to undertake, with the assist-
ance of Mrs. Parker and Miss Kuttner, a detailed study of
poisons of the possible exotoxic variety with a considerable num-
ber of different bacteria.
Throughout our work we have endeavored to avoid as far as
possible any extensive extraction of the bacterial cell. For this
reason we have worked chiefly with filtrates of very young cul-
tures on liquid media and with filtered salt solution washings of
young agar growths.
Our work is far from completion and is mentioned in this
paper only because some of our preliminary results seem to us to
have important direct bearing on the problems discussed above.
In addition to streptococci and influenza bacilli we have, so far,
employed especially the typhoid bacillus and have done a few
isolated experiments of orientation with dysentery bacilli of the
Shiga and Flexner groups, the bacillus prodigiosus, the colon
bacillus, staphylococcus aureus and the meningococcus. The
details of these experiments will be analyzed and presented at
some furture time. For the present we wish to submit the
following data. —
Working with various strains of haemolytic streptococci we
have found that culture filtrates, (better centrifugates, since the
filter detracts from toxicity) can often be shown to possess toxi-
city for rabbits. The best results are obtained by cultivation
under conditions of partial anaerobiosis (20 em. of mercury)
for twenty-two hours. The poisons appear to some extent on
simple hormone broth and upon ascitic broth but they are most
powerful when the streptococci are cultivated on horse chocolate
broth. They are at best never very powerful, 3 to 5 cc. always
make rabbits of 1200 to 1500 grams very sick, but the action is
irregular in intensity. We have killed in anywhere from five
hours to three or four days with these poisons but many rabbits
eventually recover in spite of a degree of illness that would lead
288 HANS ZINSSER
the inexperienced to expect certain death. The poisons are de-
stroyed at 75°C. to 80°C. in thirty minutes, and they seem to
deteriorate with considerable speed on standing, though the
details of this are not yet worked out. In this, as in the sequence
of symptoms these toxic substances bear great similarity to the
influenza poisons found and reported by Mrs. Parker. The in-
cubation time is short. Rabbits that eventually die get sick
within forty-five minutes to an hour and one half. There is weak-
ness, often diarrhea, respiratory difficulty and a curious watering
of the eyes. In other respects the rabbits act much like animals
injected with a toxic foreign serum. Weakness is progressive
and, finally, after three, four or five hours or longer the animals
lie on their sides unable to rise, in general muscular paresis, but
without paralysis and die, usually without convulsions. Autopsy
shows nothing but congestion of the abdominal viscera especially
the mesentery and bowel, and sometimes a little exudation in
the serous cavities. The poisons, like the influenza ones, have
no action on guinea-pigs and little or none on mice. All neces-
sary controls were of course done, and will be reported.
Typhoid filtrates have been prepared with various culture
media, and most of the experiments have been done with cul-
tures five and a half to six hours old. Such filtrates were always
powerfully toxic for rabbits, with a potency midway between the
more powerful influenza poisons and the weaker streptococcus
ones. In all the other properties mentioned above they were
qualitatively identical with the streptococcus and influenza poi-
sons. In order to compare these substances with the cell ex-
tracts we have carried out several comparisons between the six
hour filtrates and similarly prepared filtrates from six day to
ten day typhoid cultures. On several occasions we have found
that the 6 hour filtrates were as toxic for rabbits (in one case
even more so) than the ten day ones, but while the former were
innocuous for guinea-pigs in relatively high dose, the latter
killed guinea-pigs acutely. These young typhoid filtrate poisons
were also destroyed at 75°C. in half an hour, but their deteriora-
tion index has not yet been fully worked out. Other filtrate ex-
periments are as yet so incomplete that they cannot properly
be included.
a
|
|
|
NATURE OF BACTERIAL TOXAEMIA 289
The first source of error in such experiments that come to
one’s mind is, of course, the possibility of toxic substances of
this kind having originated from changes wrought upon the cul-
ture medium by the growing bacteria, a thought which is par-
ticularly obvious in view of the striking similarity in heat resist-
ance and mode of production, and the complete identity of phys-
iological action of such poisons from different bacteria. How
far this can be absolutely excluded, it is hard to say. We can
rule out histamin (shown recently to be produced on seven and
ten day colon cultures) by the inability of our toxic substances to
act on the isolated guinea-pig uterus, by their innocuousness for
guinea-pigs and by heat instability. The last point would tend
to exclude also most of the substances belonging to the ptomain
series, though about some of these we cannot find statements
concerning this property based on experiment. Peptone effects
we can exclude with considerable certainty, we think, since our
original culture media in considerable amounts do not give the
symptoms; because 20 cc. of a 5 per cent solution made with
our peptone gave no immediate symptoms in a rabbit, killing
only after six days with nothing resembling the symptoms we
were obtaining with 3 to 5 ce. of our 1 per cent peptone broth
filtrates; because guinea-pigs were unaffected and because heat
at 75°C. destroyed the toxicity. Cholin derivatives have not
yet been excluded and unknown possible toxic culture ingre-
dients have not yet been fully searched for by us.
In order to get a more complete understanding of a few of these
poisons, however, before we completed all the details concerning
the broth filtrates, we began to work with agar cultures in which,
as in the dysentery work of Kraus and the typhoid work of
Arima, the bacteria were grown on agar surfaces, the growths
washed off with sterile salt solution and filtered with as little
interval as possible between removal and filtration and with
care that no agar was scraped off, usually without any scraping
of the agar whatever. By this method, astonishing as it was
to us, we obtained powerfully toxic filtrates from influenza,
typhoid, colon and prodigiosus cultures; less powerful ones
from streptococcus growths and from growths of dysentery ba-
290 HANS ZINSSER
cilli. In all cases the differences between these various toxic
substances were purely those of potency. Qualitatively they
were similar to each other and to the poisons obtained from
broth filtrates, at least as far as their toxicity for rabbits, the
resulting symptoms and autopsy findings and the incubation
periods were concerned. Quantitative comparisons must be
made in these experiments since the perhaps confusing impression
given by the equal toxicity of substances obtained from the
saprophytic organisms with those obtained from the pathogenic
ones may lose some of its negative significance when we remem-
ber how incomparably more abundant were the six hour growths
on agar of prodigiosus and bacillus coli. We have excluded the
possible codperation of an agar anaphylatoxin in these effects,
we believe, a control which, by the way, was not made either by
Kraus or Arima in their work.
So far, therefore, although we have made only a beginning in
a rather ambitious program, we have found that many different
bacteria will induce the formation of heat unstable toxic sub-
stances in young cultures. The formation of this substance is
roughly proportionate to the growth energy. The toxic products
are essentially similar in the symptoms they elicit in rabbits
and they are similar in their harmlessness for guinea-pigs. They
differ in some essential properties from the classical endotoxins
of the same organisms, and though we cannot yet be absolutely
sure of it, they seem distinct from most of the more usual toxic
substances produced by the cleavage of culture ingredients.
If the broth filtrate substances are identical with the poisons
obtained from the agar washings, a fact that we deem more than
likely at present, we can probably exclude the codperation of
culture substances definitely. We think that there is little
doubt that the substance we have under observation is at any
rate responsible for a good many vaguely comprehended results
obtained by preceding investigators.
The essential questions of identity of these substances from
different bacteria, their antigenic properties, etc., can be an-
swered only by animal experiments. Such immunization and
cross protection experiments are in progress, and they are ren-
NATURE OF BACTERIAL TOXAEMIA 291
dered extremely difficult by the fact that repeated small doses
often lead to marasmus, loss of hair and eventual death of
animals so treated.
The final question as to whether these poisons play any role
in the symptoms accompanying infection of the animal body
will be more difficult of approach and will not be attacked until
some of the more basic problems have been answered.
One noticeable feature that may have’ some bearing on these
substances is the apparent aggressive action of our poisonous
products. Occasionally rabbits that were injected with unques-
tionably sublethal doses of streptococci contained in supernatant
fluid from centrifuged specimens have died in about two or three
days; and with the poison of B. influenzae the small number of
influenza organisms that occasionally slip through the filters
have caused death in the rabbits with invasion of the tissues,
although this organism in its ordinary relationship does not
infect rabbits.
There remains one further possibility of injury to the body
which has nothing to do with toxaemia in the sense of the pro-
duction of free bacterial poisons. There is no conclusive evi-
dence at the present time which would lead us to doubt that in
chronic infection and in repeated infection with the same organ-
ism the body may become sensitized to the antigen of this organ-
ism. Studies on the typhoidin reaction, and to a less definite
extent on the tuberculin reaction, would point in this direction.
Investigations on the active and passive sensitization of guinea-
pigs against typhoid protein which we published some years
ago, tend to indicate that just before, or at the time when anti-
bodies appear in the circulation, there have developed sessile
antibodies of the same nature. The violent reaction of guinea-
pig uteri existing at such times; the partial protection of the
cells in such cases by increased concentrations of the circulating
antibodies; these facts would, at least, suggest that such a mech-
anism can play a distinct rdle in cases of prolonged, chronic
repeated infections. But whether or not such animal experi-
ments can be translated into the conditions prevailing in the
infected human body, must rest upon further study.
292 HANS ZINSSER
\7
It is quite apparent from the preceding considerations that a
simple answer for the complex problems of bacterial toxaemia is
quite out of the question. Indeed, we have perhaps been illogi-
cal in hoping to explain by any simple mechanism a series of
phenomena as intricate as these, in which there is a struggle
between two rival metabolisms, mutual modification and in-
jury. Yet, difficult as the problem may be, there is hope of
ultimate clearness. In the attainment of this we can best
assist by selecting from the work of the past that which may be
accepted as permanent fact, that which may be set aside, and
that which may, for purposes of experimentation, be regarded as
rational possibility.
For the time being this may be done as follows:
1. The body substances of most Gram-negative bacteria are
toxic for the ordinary laboratory animals. These toxic proper-
ties are common to many non-pathogenic, as well as pathogenic
bacteria of this class. It is uncertain, but unlikely, that they
are pharmacologically specific.
These substances can be obtained by a variety of extraction
methods, as well as by prolonged cultivation in broth.
In a large majority of cases these substances have been found
relatively resistant to heat, and do not deteriorate readily on
standing.
They do not induce neutralizing antibodies of any marked
degree of potency, but they do induce specific protein sensitizers
by means of which partial specific neutralization of their effects
may be accomplished.
These are the so-called endotoxins.
The mechanism of the action of these substances is somewhat
uncertain. They certainly do not form the matrix of toxic split
products produced in the circulation by the sensitizer-alexin
complex as conceived by Friedberger, and others; but injury by
their reaction with the fixed tissue cells as conceived by Vaughan
cannot be excluded, and, indeed, there seems to be no experi-
mental method of differentiating such reaction from the direct
NATURE OF BACTERIAL TOXAEMIA 293
poisoning of tissue cells by preformed poisons of such cell
substances.
Injury by their purely physical effects in the circulation as
suggested, among others, by Novy and De Kruiff, is unlikely,
but must be considered.
Similar endotoxic substances have not been consistently pro-
duced with Gram-positive bacteria.
2. Toxic substances which are probably not identical with the
- poisons of the bacterial cell body have been produced with many
pathogenic and some non-pathogenic bacteria, both Gram-
negative and Gram-positive, by the filtration of young cultures,
and by filtering washings of agar growths.
No conclusive evidence has been brought so far to show that
poisons produced by these two methods are not identical.
The poisons produced by these two methods seem to be exotoxic
in the sense that they do not represent extraction products.
These substances are less heat stable than the endotoxins,
usually being destroyed by 75° to 80°C. in thirty minutes.
They have usually possessed relatively low toxicity for guinea-
pigs.
Poisons produced by these two methods in our own laboratory
from a considerable number of bacteria have been found identical
in regard to heat resistance, innocuousness for guinea-pigs, incu-
bation time, physiological action, and autopsy findings in rab-
bits. This, to us, reopens the question of whether these poisons
are any more specific than are the endotoxins described above.
This problem can be settled only by extensive experiments on
specific immunization and active and passive cross immuniza-
tion. So far, attempts to increase the resistance of rabbits
against these poisons, as produced by us, have met with little
success, since repeated injection led to emaciation, loss of hair,
marasmus and death. (Recently, 10 out of 11 rabbits carefully
treated with slightly increasing doses of young typhoid and
streptococcus filtrates have died, and a horse, placed at our dis-
posal for similar treatment with influenza filtrates made in
our laboratory by Mrs. Parker, also died in the course of immuni-
zation.)
294 HANS ZINSSER
It is still necessary to identify the poisons we are working
with, with similar exotoxic products of other investigators. To
us this identity seems likely at the present time.
Specific antigenic effects of poisons of this class produced from
a variety of bacteria have been claimed by a number of reliable
investigators. Reviewing this part of the problem as a whole,
it becomes apparent that nothing further can be said until more
extensive immunization experiments have been completed.
When these preliminaries have been disposed of, we will be
in a position to proceed to a study of the relative importance of
these toxic bacterial products in the infectious diseases of man
and animals, and only then will we be able to see our way clear
to the practical problems of serum therapeutics.
REFERENCES
Preirrer: Jahresbericht f. Immunitat., 1910, 6, 13. (critical summary by Pfeiffer
himself).
VauGuHan, Victor C.: Protein split products in relation to immunity and disease.
Lea and Febiger, New York, 1913.
FRIEDEMANN, Uuricu: Zeitschr. f. Immunitat., orig., 1909, 2, 591, ibid., 1909,
3, 726; Jahresbericht f. Immunitit., 1910, 6, 31.
FRIEDBERGER AND COLLABORATORS: Zeitschr. f. Immunitit., orig., 1910, 6, 179,
299; ibid., 1910, 7, 94, 665, 748.
THIELE (TEALE) AND EmBLeETON: Zeitschr. f. Immunitat., orig., 1913, 19, 648,
666.
KEYSSER AND WASSERMANN: Zeitschr. f. Hyg., 1911, 68, 535.
BorpET: Compt. Rend. de la Soc. de Biol., 1913, 74, 225.
NatuHan: Zeitschr. f. Immunitat., orig., 1913, 17, 478.
TCHERNOROUTZKY: Compt. Rend. de la Soc. de Biol., 1913, 74, 1213.
JOBLING AND Peterson: Jour. of Exp. Med., 1914, 19, 485.
Novy Anp De Krutr: Jour. of Inf. Dis., 1917, 20, 499 et seq.
Koxter: Cited from Novy and De Kruif, loc. cit.
Mo.povan: Deut. Med. Woch., 1910, 36, 2422.
Doerr: Wien. Klin. Woch., 1912, 25, 331, 339.
SLATINEANO AND Ciuca: Compt. Rend de la Soc. de Biol., 1913, 74, 631.
De Krutr: loc. cit. (Novy and De Kruif.)
Mira anv Ito: Zeitschr. f. Immunitat., orig., 1913, 17, 586.
BrsrEepDKA: Ann. de. l’Inst. Pasteur, 1905, 19, 477.
ARIMA: Central. f. Bakt., orig., 1912, 63, 424.
Conrabi: Deutsche med. Wochenschr., 1906, 32, 58.
KRAUS AND STENITZER: Zeitschr. f. Immunitit., orig., 1909, 3, 646.
Aronson: Berl. Klin. Woch., 1907, 44, 572.
YAMANOUCHI: Compt. Rend. de la Soc. de Biol., 1909, 66, 1051.
NATURE OF BACTERIAL TOXAEMIA 295
MacFapyEN AND Row.anp: Central. f. Bakt., orig., 1901, 30, 753.
| MacFapyeEn: Central. f. Bakt., orig., 1906, 41, 266.
| MEYER AND BERGELL: Berl. Klin. Woch., 1907, 44, 568.
) KRAUS AND STENITZER: Wien. Klin. Woch., 1907, 20, 344.
Brav AND DENIER: Compt. Rend. de la Soc. de Biol., 1904, 56, 433.
Brav AND Denier: Ann. de |’Inst. Pasteur., 1906, 20, 578.
. Kraus; Kotte AND WASSERMANN Handbuch, vol. 1, p. 180.
j MacFapyeEn: Central. f. Bakt., orig., 1906, 42, 365.
j METCcHNIKOFF, Roux AND SALIMBENI: Ann. de |’Inst. Pasteur, 1896, 10, 257.
Doerr, Kraus AND LEvapiTI Handbuch, vol. 1, p. 145.
Kravs AnD DoErRR; Krauts AND LevapitT1 Handbuch, vol. 2.
Krauts AND Dorrr; Zeit. f. Hyg., 1906, 55, 1.
Topp: British Med. Journ., 1903, 2, 1456.
OLITSKY AND KiieLEeR: Journ. of Exper. Med., 1920, 31, 19.
PFEIFFER AND UNGERMANN: Central. f. Bakt., Orig., 1909, vol. 50,
Bessav: Central. f. Bakt., 1911, 57, 27.
MarmoreEk: Ann. de |’Inst. Pasteur, 1902, 16, 169.
Brawn: Central f. Bakt., 1912, 62, 383.
Aronson: Berlin. Klin. Woch., 1902, 39, 979, 1006.
Simons: Central. f. Bakt., Orig., 1904, 35, 308, 440.
CLARK AND FELTON: Journ. of the Amer. Med. Assoc., 1918, 71, 1048.
eS
Le
<<"
Se ee en, ee,
}
9 a
STUDIES IN ANAPHYLAXIS
ARTHUR F. COCA anp MITSUJI KOSAKAI
From the New York Orthopaedic Hospital and the Department of Bacteriology in
Cornell University Medical College, New York City
Received for publication April 30, 1920
I. ON THE QUANTITATIVE REACTION OF PARTIALLY NEUTRALIZED
PRECIPITIN IN VITRO AND IN VIVO
In one of the numefous studies (1) with which Richard Weil
enriched the literature of anaphylaxis he came to the conclusion
that the reaction between antigen and antibody that occurs
within or upon the susceptible cells in the anaphylactic guinea-
pig takes place in a manner different from that of the reaction of
these two substances in wtro. This conclusion was drawn from
a quantitative study of the anaphylactic reactivity of passively,
sensitized guinea-pigs that had been partially desensitized.
Weil states that when he passively sensitized guinea-pigs
with different amounts of immune rabbit’s serum and partially
desensitized these different animals with the same quantity of
the antigen he found that the animals were all equally sensitive
to a further injection of the antigen; the ‘‘minimal anaphylactic
dose”’ of the antigen was the same for all of them. The result
was the same when guinea-pigs passively sensitized with the
same amount of the immune serum were partially desensitized
with varying amounts of the antigen; with all of these animals,
also, the ‘‘minimal anaphylactic dose” of the antigen was the
same.
These experiments, which are summarized in Weil’s tables 12
and 17, revealed a second peculiarity which forced Weil to the
conclusion that ‘‘partially desensitized (or neutralized) antibody
reacts to antigen in a manner which is quite different from that
of pure antibody.”’ This peculiarity hes in the fact that the
297
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 3
298 ARTHUR F. COCA AND MITSUJI KOSAKAI
minimal anaphylactic dose of the antigen was much greater for
the partially desensitized animals—0.5 cc.—than it was for the
animals that had been sensitized with any smaller dose of im-
mune serum, but without being partially desensitized—0.005 to
0.05 ec. According to Weil this latter phenomenon made it
seem evident ‘‘that desensitization cannot be explained on the
basis of the neutralization or saturation of a fraction of the cel-
lular antibody.”
Weil wrote that these two phenomena have no counterpart
in the precipitin reaction in vitro nor, indeed, ‘‘in immunological
science.”’ However, since he did not attempt to reproduce the
phenomena in the test tube, we have taken up this question
experimentally in the present study.
The usual plan of our investigation was, on the one hand, to
sensitize guinea-pigs passively with a precipitating immune
rabbit’s serum and, after partial desensitization with varying
amounts of the antigen, to determine the minimal fatal dose of
the latter in the partially desensitized animals; on the other hand
we mixed in test tubes quantities of the immune serum and
antigen that corresponded with those used in these animal
experiments and after removal of the resulting precipitate by
centrifugation we determined the minimal precipitating amount
of the antigen with the supernatant fluid. The supernatant
fluid was also injected into a series of guinea-pigs and the mini-
mal lethal dose of the antigen was determined for the animals
so treated.
By a comparison of the results of these parallel tests, it can be
seen whether the interaction of precipitin and precipitinogen
is different in vivo and in vitro as Weil thought.
Such experiments were carried out with the pseudoglobulin!
of horse-serum and with crystalline albumin prepared from the
white of hen’s egg.
The immune serum used in the first experiment was a mixture
of sera derived from two rabbits (170 and 397) that had received
1 For a generous supply of this material we are indebted to Charles R. Tyler,
Research and Antitoxin Laboratory, Department of Health of the City of New
York, Otisville, N. Y.
STUDIES IN ANAPHYLAXIS 299
numerous intraperitoneal injections of crystallline egg albumin,
as follows: 2 cc., 10 cc. and 5 ce. on the first, fifth and eleventh
days, and 0.5 cc. daily from the thirteenth to the twenty-sixth
days inclusive. The animals were bled four days after the last
injection.
The minimal sensitizing dose? of this serum mixture was found
to be 0.2 ce.
The minimal anaphylactic dose of the egg albumin was deter-
mined for guinea-pigs that had been sensitized twenty-four hours
previously with 0.4 cc. of the serum mixture. The protocol of
this determination is presented in table 1.
TABLE 1
Determination of the minimal lethal dose of egg-albumin after a sensitizing dose
of 0.4 cc. of serum 170+397
GUINEA-PIG EGG-ALBUMIN, INTRAVENOUSLY RESULT
ce.
1 0.0025 Slight symptoms*
2 0.005 Severe symptoms
3 0.0075 Moderate symptoms
4 0.01 Oe
5 0.01 rl
*h = typical anaphylactic death within 23 to 5 minutes.
Severe symptoms = immediate violent convulsions with eventual recovery.
Moderate symptoms = slight to moderate convulsions usually beginning after
4 to 5 minutes.
Slight symptoms = marked dyspnoea; animal lies down on side; no con-
vulsions.
The minimal lethal dose of the albumin was found to be 0.01
cc. 5 cc. of the immune serum (170 and 397) were then mixed
with 0.01 cc. of the crystalline egg albumin solution and after
two hours at 56°C. the clear supernatant fluid, which was sepa-
rated from the precipitate by centrifugation, was placed in the
ice-box. On the following day, no further precipitation having
taken place, the fluid was compared with the untreated serum
as to its precipitin titer and as to its sensitizing function.
2 In all of these experiments guinea-pigs weighing between 250 and 340 grams
were used.
COCA AND MITSUJI KOSAKAI
ARTHUR F.
300
‘sisAjouoy 040[duI09 = "FY “C—O {sIsh]ouey oyerspoul ="pour {sisAjoursey 4YSI[s ="[s {stsAjowey ou=¢,,
oon) oO (eos OE Oa Boa pou 1s woryexy-juoua;dwoy |(800°0) (F'0) 4
= = = = = = - a aorzeytdroerg | 910° 80
‘H’‘O | ‘H’'O | ‘H'O}.‘O'N | ‘pom ‘Is 0 0 uoryexy-quouroydmoy | 9100'0 | — (#0) as
=e = ss = Te ai she steals uoryeyidioorg | 70° 0's 5
HD 4] “HD |S Os) spom ‘Is 0 0 0 uoryexy-juowatduoy | §000°0 | (#0) S
—- |- - —- | + + + spa 2N(4e) eo uoryepdroorg | 10°0 0's
07 HO | "“H'O2 8 0 0 0 «0 WOTPEXY-}UIUI9] CULO /)
= = - | + + | ++ | ++ | 444+ uoryeqdioarg Sea :
929 “99
|. | ee ae Se ar ae ee Se es ee ee ee ee Se Se
£00000°0 10000°0 20000°0 ¢0000°0 T000°0 2000'0 ¢000°0 TOO1O Ios sage eet 00 ‘ueatyUy |-Nq]e-339) uInIeg SOLI9g
uesIjUy
(Ga0L HOVG NI ‘09 [0 WAUASIINV) daxXiIW NADSIINV GNV WOUGS
NOILVXIA INAWNATINOD AGNV NOILVOEIdI0dUd AHL tO UALIL GQNOWNWMI GAL IO NOLLUNOdOUd
fiop Burnojjof ay} Uo pajsaz som pinyf qunzousadns ay? Spabnfriujuaa uay) pun
SINOY ON} payDgnour pup pan paxiu aan (uriungv-66a) uabyun pun WNLAs aY) 017A UL UOYMZYDYNaU JOYWWT *(L6E+0L1)
wunias pazypoijnau fiyyoysod ay) pun yourbiso ayy fo 49] Uorypar-jwauajduos pup 4ajy wornzidrdaid ay) fo wosr.undwo
¢ ATAVL
=e
STUDIES IN ANAPHYLAXIS 301
The results of the comparative precipitin titration are pre-
sented in table 2 (series 1 and 2).
It is seen that the minimal precipitating quantity of the egg
albumin was about twice as great for the partially neutralized
immune serum as it was for the untreated serum.
A dose of 0.4 cc. of the supernatant fluid was injected into each
of a series of guinea-pigs and on the following day the minimal
lethal dose of the egg albumin solution for these animals was
determined. The results of the test are presented in table 3.
TABLE 3
Determination of the minimal anaphylactic dose of egg-albumin after partial
neutralization of precipitin (in vitro); & cc. serum (170+397)+0.01 cc. egg-albu-
min after two hours at 37°C. centrifuged; 0.4 cc. of the supernatant fluid is used
for the sensitization of each animal
1
GUINEA-PIG EGG-ALBUMIN, INTRAVENOUSLY RESULT
cc.
1 0.01 Slight symptoms
2 0.02 Severe symptoms
3 0.02 oe
4 0.02 | ria
5 0.03 | rts
It is seen that, in harmony with the precipitin titrations, the
minimal lethal dose of the egg albumin was about twice as great
for these animals as it was for guinea-pigs sensitized with the
same amount of the untreated immune serum.
In another series of animals the foregoing partial neutraliza-
tion of the immune serum was carried out in vivo in accordance
with Weil’s procedure.
Each animal received 0.4 cc. of the immune serum’ and on the
third day thereafter each received, by intraperitoneal injection,
0.008 cc. of the egg albumin solution. Twenty-four hours later
the minimal lethal dose of the egg albumin was determined as
usual for the animals.
The results of the test are shown in table 4.
3’ Throughout this study the passive sensitization was effected by intraperi-
toneal injection.
302 ARTHUR F. COCA AND MITSUJI KOSAKAI
TABLE 4
Determination of the minimal anaphylactic dose of egg-albumin after partial de-
sensitization of passively sensitized guinea-pigs. Sensitization: with 0.4 cc. of
serum 1704+-397. Partial desensitization on the third day: 0.008 cc. of egg-albu-
min intraperitoneally. Test on the fourth day
GUINEA-PIG EGG-ALBUMIN, INTRAVENOUSLY RESULT
cc.
1 0.04 No symptoms
2 0.1 Severe symptoms
3 0.1 Very severe symptoms
4 0.2 Very severe symptoms
5 0.2
It is seen that the lethal dose of the antigen by the partial
desensitization method of Weil is relatively the same as it is
after the corresponding partial neutralization in vitro.
In a further series of animals the procedure was the same as
that of the preceding partial desensitization experiment except
that 0.8 cc. instead of 0.4 cc. of the immune serum were used for
the passive sensitization.
The results of the test in this series are shown in table 5.
TABLE 5
Determination of the minimal anaphylactic dose of egg-albumin after partial desensi-
tization of passively sensitized guinea-pigs. Sensitization: with 0.8 cc. of serum
170+-397. Desensitization on the third day: 0.008 cc. of egg-albumin intraperi-
toneally. Test on the fourth day
GUINEA-PIG EGG-ALBUMIN, INTRAVENOUSLY RESULT
cc.
1 0.1 Moderate symptoms
2 0.1 Moderate symptoms
3 0.1
4 0.1 Sf
5 0.2 Oe
It is seen that the animals used in the two tests presented in
tables 4 and 5 responded too irregularly to allow a satisfactory
comparison of the results in the two series although, on the
whole, the animals of table 5 appear to have been more sensitive
than those of table 4.
STUDIES IN ANAPHYLAXIS 303
For the reader’s convenience the results of the foregoing ana-
phylaxis experiments are summarized in table 6.
TABLE 6
Summary of the foregoing experiments
fama Dose BERUM NEUTRALIZING DOSE EGG-ALBUMIN “2 rapginapald
In vitro In vivo EGG-ALBUMIN
ce. ce. cc. a
0.4 0.01
0.4 0.0008 0.02-0.03
0.4 0.008 0.2 -0.3
0.8 0.008 0.10.2
In table 2 are summarized the parallel experiments in partial
neutralization of the precipitin in vitro. The results of these
latter tests show that the minimal precipitating quantity of the
antigen increases exactly in proportion to the amount of the
antigen which is used in the partial neutralization.
The animal and test-tube experiments taken together indicate
that, whether the partial neutralization is carried out in the test
tube or in the guinea-pigs, the quantitative relationship between
antigen and partly neutralized precipitin is the same in vivo and
im vitro.
A second series of experiments was carried out with the
pooled sera of two rabbits (425 and 426), both of which had
received three intraperitoneal injections of whole egg white as
follows: On the first day 5 cc.; on the fifth and tenth days 10 ee.
each time. ‘The rabbits were bled on the seventeenth day. In
a preliminary test 0.1 cc. of each of these sera had been found
capable of fully sensitizing a guinea-pig of about 325 grams
weight. Smaller amounts of the sera were not tested.
As in the previous experiments we first determined the mini-
mal anaphylactic dose of the crystalline egg albumin solution
for guinea-pigs highly sensitized with the immune rabbits’
serum; that is, with 0.3 cc. of the pooled serum 425 and 426.
The results of this determination are presented in table 7.
304 ARTHUR F. COCA AND MITSUJI KOSAKAI
The minimal anaphylactic dose is seen to have been a little
more than 0.002 ce. of our erystalline albumin solution.
Three further series of animals were sensitized each with 0.3
ec. of the serum 425 and 426 and, after a partial desensitization
with the egg albumin solution in amounts differing in the three
series, the minimal anaphylactic dose was determined. The
results of this experiment are shown in table 8-a, b and e.
» Here again the minimal anaphylactic dose of the antigen
increases in exact proportion to the increase in the amount of
the antigen used for the partial desensitization (0.00025: 0.0025::
0.035. : 0.35).
TABLE 7
Determination of the minimal anaphylactic dose of antigen after passive sensitiza-
tion with the untreated immune rabbits’ serum. Sensitization: with 0.3 cc. of the
pooled serum (425+ 426)
GUINEA-PIG EGG-ALBUMIN, INTRAVENOUSLY RESULT
cc.
1 0.00075 No symptoms
2 0.0015 Mild symptoms
3 0.002 Severe symptoms
4 0.002 Oe
5 0.005 Oe
It is seen that the injection of 0.01 cc. of our solution of crys-
talline egg albumin completely desensitized the animals. This
result is In quantitative disagreement with those of Weil.4 For
although Weil’s animals were sensitized with approximately
the same amount of precipitating serum as ours (0.2 to 0.4 ec.)
and although they were equally sensitive to the antigen (0.001
cc. of Weil’s 5 per cent solution as compared with 0.002 ce. of
our 3 per cent solution) they were only partially desensitized
with 0.01 cc. and 0.04 ec. of the 5 per cent solution of the anti-
gen. Our five animals were all completely desensitized with
0.01 cc. of a 3 per cent solution of the antigen.
With the serum mixture 425-426 partial neutralization in vitro
was carried out exactly as in the experiments with serum mix-
ture 170-397.
4 See Weil’s table 18.
STUDIES IN ANAPHYLAXIS 305
The results of this experiment are presented in table 9.
It appears from these tests that after the immune serum
425-426 had been partially neutralized with 0.0025 ce. of the
antigen the minimal precipitating dose of the antigen was about
ten times as great as it was after a partial neutralization with
TABLE 8
Deiermination of the minimal anaphylactic dose of antigen after partial desensiti-
zation (in vivo)
GUINEA-PIG EGG-ALBUMIN, INTRAVENOUSLY RESULT
a. Sensitization: with 0.3 cc. of serum (425+426). Partial desensitization on the
third day with crystalline egg-albumin, 0.00025 ec. intraperitoneally.
ce.
1 0.0133 Slight symptoms
Ae 0.02 Severe symptoms
3 0.025 Slight symptoms
4 0.03 Severe symptoms
5 0.035 Oe
b. Sensitization: with 0.3 cc. of serum (425+426). Partial desensitization on the
third day with crystalline egg-albumin, 0.0025 cc. intraperitoneally.
1 0.2 Slight symptoms
2 0.3 Mild symptoms
3 0.3 oe
4 0.35 mh
5 0.35 ix
c. Sensitization: with 0.3 ce. of serum (425+426). Desensitization on the third
day with crystalline egg-albumin; 0.01 ce. intraperitoneally.
No symptoms
No symptoms
No symptoms
No symptoms
No symptoms
or Whe
Now nN ee
(= — i — i —
0.00025 cc. of the antigen. Corresponding with the animal
experiment, 0.01 cc. of the antigen completely neutralized the
precipitin in 0.3 ec. of the immune serum.
The results of the parallel tests with the method of comple-
ment fixation were in satisfactory agreement with those of the
precipitin tests.
@) @) @) @) 2) @) @) @) (OHA) @) (OM) BO Foe ETO) UOTPEXY
- ~yuouto,du10 T'0| ¢0
— - - |- — — _ — — — — — — — | — | uoneytdioig
Loss|
i @) @) @) @) @) @) @) O | 180A) 0 0 0 UOTPBXY
a ~yuoura,duroD | > ¢z00°0 | £°0
cg -}-|-|{- a Gr terrae oriar ar |r iar] ae | = |p wconceme lei
| xa
5 Oe [eta TOR oer O) il foe lar ohere | hel) 0 Woryexy
aI ~juauratduroy | +¢z000°0 | ¢°0
= pee eee (ee eee eb Ge eae uoryeytdroorg
=) 5 :
7 @) @) O | pow) oO 0 0 0 UOTPEXY poqvary
< -juowe;duioy =
SES ee ere ete, aera ewer — | woneydoorg -
5 oe) oe)
.S) a Pee eS | ee |S ae wee Ea RE | a ee ee ee ee | ee ee
. yuoo 10d ¢
a 190000"0 200000" s00000"0 10000'0 | 0000'0 | 0000'0| 1000°0 | 2000°0 | ¢000°0 100°0| 2000 |s00'0| r0'0 | zoo | go | uosiquy | ydresma) | TOS
fa | uostUy
i
E dqaxXIW N@SILINV GNV
fo] NOILVXIM-LNAWATINWOD GNV NOILVLIdIONUd AHL JO WALL WoOuds ANOWAL
. ~— —
-_
a a ee
|
RELATION OF SPUTUM BACTERIA TO ASTHMA ati
the 2 cases giving a positive skin test but treated unsuccessfully
a non-hemolytic streptococcus was used in each.
The permanency of the favorable results obtained is yet to
be discovered, but most of these patients have retained at least
during the seven months of observation the result which followed
the treatment as given.
Aside from being an index as to the probable outcome of
treatment, it was hoped that the skin test might become another
method of differentiating various sub-groups of organisms and
as a matter of fact, in the skin of the same patient, definite
differences of reactions were obtained with different cultures of
not only hemolytic and non-hemolytic streptococci but also of
staphylococci. The data on this point is, however, far from
complete.
Granted that the results outlined above are substantial and
that the importance of a positive skin test as a prerequisite to :
successful treatment is not overestimated, we may discuss
intrinsic asthma as follows:
We know that in horse asthma and ragweed pollen asthma,
the symptoms depend upon an exquisite sensitiveness to the
particular foreign protein. Inasmuch as circulating antibodies
are not found, we assume that this condition of sensitiveness is
cellular. The specific protein will produce a positive skin test
and repeated injections will cause relief of symptoms. This
treatment is specific.
In intrinsic asthma, vaccines likewise produce a positive skin
test and as treatment with them is successful only in case the
test is positive, their action is “specific.”” By analogy, therefore
we may assume that asthma due to bacteria depends probably
on a condition of specific cellular sensitiveness either to the
bacteria themselves or to the products of their action in the
organism.
REFERENCES.
(1) Montgomery AND Sicarp: Treatment of bronchial asthma by vaccination,
with report of cases. Am. J. Med. Sci., 1917, 158, 856.
(2) Wauxker, I. C.: Treatment of bronchial asthma with vaccines. Arch. Int.
Med., 1919, 23, 220.
(3) Stittman, E. G.: A contribution to the epidemiology of lobar pneumonia.
Journ. Exp. Med., 1916, 24, 651.
.
i
Ta
7 ad
Cd
.
)
- .
i
; J
a
Boe &
a
he
[Ss
.
, >
SOME OBSERVATIONS ON THE CONSTITUTION OF
THE COMPLEMENTS OF DIFFERENT ANIMALS
T. J. MACKIE
From the Department of Bacteriology, University of Cape Town, South Africa
‘Received for publication May 26, 1920
In the past a considerable amount of attention has been paid
by serological workers to the complex characters of serum-
complement and many attempts have been made to subject
complement to more detailed biochemical analysis. Valuable
information has thus been obtained regarding the biological
action and properties of this important element of animal serum.
The study of complement has been principally based on its
cytolytic effects towards red blood corpuscles sensitised with the
homologous immune body, and the general tendency of research
on this subject has been to elicit the complexity of constitution
of complement and the numerous factors on which its action
depends.
The fresh serum of the guinea-pig represents with ox corpuscles
+ rabbit versus ox immune body or cobra venom, one of the
most active complements and it has therefore been commonly
used for studies on complement action.
Attention was first drawn by Stephens (1) to the hemolytic
effect of certain snake venoms along with fresh serum. In the
case of guinea-pig’s serum, which has a powerful activating effect,
this characteristic property is annulled by heating the serum at
55°C. and corresponds as regards thermolability to the comple-
menting action of serum with immune body. Observations made
by Browning and Mackie (2) on the complementing action of
serum with immune body in relation to its hemolytic effect
with cobra venom showed that the venom-activating constituent
of serum though similar in many characters to the complement
which acted with immune body was not identical with the latter.
379
380 T. J. MACKIE
Ferrata (3) and later Brand (4) Liefman (5) Sachs (6) demon-
strated that complement could be fractioned into two compo-
nents neither of which exhibited any degree of complementing
effect by itself, although together they reproduced the full
hemolytic activity of the native serum. These two moieties
were designated ‘“‘end-piece” and “mid-piece’”’ the latter combin-
ing directly with red blood corpuscles + immune body.
In Liefman’s method carbon dioxide gas was passed through
serum diluted with distilled water and this led to the precipi-
tation of part of the globulins of the serum which represented
the ‘“‘mid-piece’’ while the ‘“‘end-piece”’ consisted of the albumin
and that portion of the globulin still remaining in solution.
These fractions were carefully studied by Browning and Mackie
in the case of guinea-pig’s serum and it was found that the venom-
activating constituent could also be fractioned into two similar
components.
A “third component”? of complement was described by Ritz
(7) which had no end-piece or mid-piece properties and unlike
these components was invariably stable at 57°C. This constit-
uent was demonstrated by inactivating complement with venom
and by the restoration of its activity on the addition of heated
serum.
It was subsequently found (Browning and Mackie (8)) that
complement could be fractioned into four different components
by Liefman’s method followed by precipitation of the proteins
in different concentrations of ammonium sulphate. These four
components were represented respectively by the albumin,
pseudoglobulin from ‘‘end-piece,’’ pseudoglobulin from ‘“mid-
piece” and euglobulin. None of these fractions corresponded
to any of the previously described complement components
“end-piece,” ‘‘mid-piece”’ or “third component.” These obser-
vations were carried out with guinea-pig’s serum and the general
results showed that the complement constituents were distributed
over all the different proteins of the serum but appeared to be
concentrated chiefly in the pseudoglobulin. Of the four compo-
nents, three, including always the albumin, were generally neces-
sary for full restoration of complement action and the albumin
‘
COMPLEMENTS OF DIFFERENT ANIMALS 381
fraction appeared to represent an essential constituent of the
complement.
While it is doubtful if albumin, pseudoglobulin and euglobulin
separated by ammonium sulphate constitute homogeneous pro-
tein entities (Martin and Chick (9)), there is strong evidence
that they represent different complement constituents, and
Liefman’s method certainly elicits a striking difference between
the two moieties of the pseudoglobulin separated by carbon
dioxide. With a view to throwing further light on the structure
of complement and especially the venom-activating constituent
of serum, further experiments have been carried out with the
sera of certain other animals.
The technique followed was that originally described in the
Journal of Pathology and Bacteriology and the Zeitschrift fiir
Immunititsforschung (8).
The hemolytic systems used were (1) ox red blood corpuscles
- + immune body (rabbit versus ox), and (2) ox red blood cor-
puscles + cobra venom. ‘The sera of man, rabbit, and horse
were selected for comparison with guinea-pig’s complement.
With ox corpuscles + immune body rabbit versus ox the
average minimal hemolytic dose of these sera as shown by Muir
(11) are: Guinea-pig, 0.01 cc.; rabbit, 0.1 cc.; man 0.11 ce.;
horse, ~ cc.; for 1 ce. of a 5 per cent suspension of blood.
EXPERIMENTS WITH RABBIT’S SERUM
Various specimens of rabbit’s serum were fractioned by Lief-
man’s carbon dioxide method into ‘‘end-piece”’ and ‘‘mid-piece”’
and the two moieties were further subdivided by the ammonium
sulphate method.
The different globulins were then tested separately and in
certain combination as regards their complementing action with
immune body.
Table 1 demonstrates the results of one of these experiments;
the pseudoglobulin from end-piece showed distinct activity which
was increased to the standard of the native serum by the addi-
tion of the pseudoglobulin from mid-piece; the latter by itself
382 T. J. MACKIE
displayed no complementing properties. The addition of euglo-
bulin to the mixture of pseudoglobulins did not further add to
the hemolytic value. A mixture of the pseudoglobulin from
end-piece and euglobulin also yielded a fully active complement
but euglobulin along with pseudoglobulin from mid-piece was
quite inactive.
Thus the complementing property of rabbit’s serum is in-
variably resident in the globulins and distributed among them;
but a mixture consisting of only two of these globulin fractions,
TABLE 1
LYSIS OF 0.5 CC. OX BLOOD SUSPENSION
+ 5 DOSES OF IMMUNE BODY
RABBIT’S SERUM Bo eee
0.01 0.025 0.05 0.075 0.1
cc. cc. cc. cc. cc.
iINativercomplementinane ce sees posse coeee trey dist. |e: c. c
AUT EL OA EEN AOS aE. Rok eee 0 0 S730 0
Pseudoglobulin from end-piece............... 0 0 0 | dist. j(c
Pseudoglobulin from mid-piece............... 0 0 0 0 0
Biglobualimg £00.66 ities es a eke foie Mae ee ere 0 0 0 0 0
Pseudoglobulin from end-piece + pseudoglob-
ulinsfrom\anid=piece.2i.2.% 2M DAasi. 2 en oleae 0 tr. c Cc. c
Pseudoglobulin from end-piece + pseudoglob-
ulin from mid-piece + euglobulin......... OP idist.\|\ se: c. c
Pseudoglobulin from mid-piece + euglobulin..| 0 0 0 0 0
All‘four-components'.) <1 22 Soon 5.0 ee oes a0) ea) dist?.| Pie: Cc. c
Pseudoglobulin from end-piece + euglobulin..| 0 | dist. | ce. C. c
Albumin + pseudoglobulin from mid-piece....| 0 0 0 0 0
Albumin = seuclobulinke ene eae eer ee 0 0 0 0 0
In this and in subsequent tables: tr.=trace; f.tr.=faint trace; dist. =distinct;
mk.=marked; c.=complete; j.c.=just complete; al.c.=almost complete.
provided they do not both belong to the mid-piece, is sufficient
to reconstitute the complement. In the case of rabbit’s serum
also certain constituents may be considered as interchangeable.
It is to be noted that there is an actual qualitative differentiation
of the pseudoglobulin of end-piece and that contained. in the
mid-piece fraction. This was also noted in the case of guinea-
pig’s serum. In contrast with guinea-pig’s serum however the
albumin fraction does not appear to contain any complement
constituents.
COMPLEMENTS OF DIFFERENT ANIMALS 383
In general, rabbit’s serum has no activating effect with cobra
venom and ox’s corpuscles, though with immune body it shows
marked complementing action.
TABLE 2a
Lysis OF 0.5 cc. oF 5 PER CENT SUSPENSION OX BLOOD + 0.0065
GRAM COBRA VENOM
RABBIT’S SERUM
0.04 ec. 0.1 ce. 0.16 ce. 0.2 ce. 0.24 ce. | 0.3 ec. | 0.36 ec.
Fresh serum........... 0 0 0 0 0 0 0
Globulin precipitate by
157) 0 0 0 tr. dist. | mk. | al.c.|] ec.
Albumin separated by
‘sl ES) 0 0 0 0 0 0 0 0
; NHy,)SOx. ..| 0.2 cc. | 0.24 cc.| 0.3 ce. | 0.36 cc.
Globul f ‘
co * + S ee
ipoenini fea. .-| 0.05 ec.| 0.06 ec.| 0.075 cc.| 0.09 ee.
0 0 0 0
TABLE 23
Lysis OF 0.5 cc. oF 5 PER CENT OX BLOOD SUSPENSION
-+ 5 DOSES IMMUNE BODY
RABBIT’S SERUM
0.01 | 0.025 | 0.05 | 0.075 | 0.1 0.2 0.3
cc, cc. cc. ec. cc. cc. cc.
IN/UUKVE: SEINE CR a a 0 mk. | al.e. c. c. c. c.
Globulin (NH4)2SO4 method..... mk. c. c. Cc. c. c.
Albumin (NH,)2SO,4 method..... 0 0 0 0 0 0 0
Globulin + albumin (NH,).SO,
TECUIIGL |. 6 does eee mk. c. c. c. c. c.
Lysis OF 0.5 cc. OF ox BLOOD SUSPENSION
(NO IMMUNE BODY)
Selobulina (NE,)eSOu. ............ 0.3 ece.=no lysis
Mipmaepengiin<..4.4.......-...- 0.3 ec.=no lysis
It was found however that the globulins separated by half
saturation with ammonium sulphate showed distinct comple-
menting action with cobra venom.
The albumin was inactive and also inhibited the action of the
globulin even in doses corresponding to one-fourth of the doses
of globulin used.
384 T, J. MACKIE
These experiments clearly demonstrate that rabbit’s serum,
which had no power of producing hemolysis of venomised cor-
puscles, still contained venom activating constituents which
were resident in the globulin fraction. These constituents were,
like complement, thermolabile (at 55°C.). In the case of hemoly-
sis with immune body, as already shown, the albumin exerted
no inhibitory action (table 2 8).
The globulin fraction of rabbit’s serum apparently represents
the whole complement of the serum both for immune body and
venom but in the case of venom the native serum is inactive in
TABLE 3
LYSIS OF 0.5 cc. OF 5 PER CENT OX BLOOD SUSPENSION 0.0005
GRAM COBRA VENOM
Guinea-pig’s complement 0.0075 cce.|0.01 cc./0.02 cc.|0.04 cc.|0.06 ce. | 0.1 ec..
=F aif =r oy = ale cr
Rabbit’s ‘“‘end-piece’’........ 0.015 ce. |0.02 cc.|0.04 cc.|0.06 cc./0.12 ec. | 0.2 ee.
dist. mk. | dist. tr. f.tr. 0
Guinea-pig’s complement..... 0.0075 cc./0.01 ec.|0.02 cc.|0.04 ce.
jece © (CF c.
Guinea-pig’s complement..... 0.0075 cc.|0.01 cc.|0.0 ce. |0.04 cc.|0.06 cc. | 0.1 ce.
ar =i ai ain a aby at
Guinea-pig’s ‘‘end-piece’’..... 0.015 ce. |0.02 ec.|0.04 cc.|0.08 cc.|0.012 ec.| 0.2 ce.
C: ce o GC C. e
Controls
0.2 ce. guinea-pig’s end-piece=0
0.2 ec. rabbit’s end-piece=0
virtue of inhibition by the albumin. It was concluded that the
deficiency of the whole serum in this respect was due to the
albumin antagonizing or ‘‘masking”’ the activity of the globulin.
It was found also that rabbit’s serum-albumin inhibited the
action of guinea-pig’s serum with venom. In the experiment
shown (table 3), varying amounts of guinea-pig’s complement
mixed with quantities of rabbit’s albumin (end-piece) represent-
ing respectively double these amounts of rabbit serum, were
tested in series with 0.5 cc. of ox blood suspension + cobra
venom. There was marked inhibition and a zone phenomenon
was produced.
ee nes
woe
‘
ie
%
é
vf
\
i
COMPLEMENTS OF DIFFERENT ANIMALS 385
Corresponding mixtures of guinea-pig’s complement with
guinea-pig’s end-piece show no inhibition of lysis. ‘This experi-
ment shows an interesting difference in the albumin fraction of
these two sera in regard to cobra venom hemolysis.
EXPERIMENTS WITH HUMAN SERUM
Specimens of fresh human sera were also investigated in the
light of the findings with rabbit’s serum and it was found that
TABLE 4
Lysis of 0.5 cc. OF OX BLOOD SUSPENSION + 0.0005
GRAM COBRA VENOM
HUMAN SERUM
0.01 0.025 0.05 0.075 0.1
0.15 0.5
cc. ec, cc. cc. cc. cc. cc
IRIE SHBSEMIUUTIN sy afere siete cucteseicholee. 3-2. <0. « 0 0 0 0 0 0 0
Globulin precipitated by
GN asO, method. ..:......./ 0 0 tr. | dist. |. mk. «| ice.
Albumin precipitated by
(NHy,)2SO,4 method...........} 0 0 0 0 0 0
Albumin+ Globulin
(NH4)eSO4 method........... 0 0 0 0 0 0
Lysis OF 0.5 cc. OF OX BLOOD SUSPENSION + 5
DOSES IMMUNE BODY
HUMAN SERUM Sas a a a
0.01 0.025 0.05 0.075 0.1 0.15
ce. ce. ce. ec. ce. ce.
IRs) Seo ose ee mk. 7] 5j.¢: c Cc. c. ¢c
OUI eG dhe al dist. | v.mk.| ec c. c. c
Sih 2 0 0 0 0 0 0
moun elobulin 0). 2... se ep ee mk. | al.c. | e c. c. c
Controls
0.5 cc. of suspension (no immune body nor venom)+human serum 0.5 cc.=no
lysis.
0.5 ce. of suspension (no immune body nor venom)+globulin 0.2 ec.=no lysis.
the globulin fraction was actively hemolytic in the presence of
venom even when the native serum had no action. Sera were
fractioned into globulin and albumin by the ammonium sulphate
method and the fractions were tested with immune body and
with venom; it was observed that the globulin displayed practi-
cally the full complementing action of the serum for immune
body. |
THE JOURNAL OF IMMUNOLOGY, VOL. Vv, NO. 4
386 T. J. MACKIE
In the case of venom the action of the globulin was inhibited
by the albumin. ‘Table 4 shows the results. In the experiment
quoted the mimimum dose of human complement with immune
body was relatively small, about 0.025 cc. for 0.5 cc. of the sensi-
tised suspension. Observations were also made as to the power
of the albumin of human serum to inhibit the hemolytic action
of guinea-pig’s serum for venomised corpuscles. For conveni-
ence end-piece of human serum separated by carbon-dioxide
was employed and it was found to exert a marked inhibiting
effect (table 5).
TABLE 5
LYSIS OF 0.5 CC. OF OX BLOOD SUSPENSION + 0.0005 GRAM
COBRA VENOM
0.005 cc. | 0.01 ce. 0.02 ce. 0.04 ce. 0.06 cc. 0.1 cc.
Guinea-pig’s serum........ j.c. c. ce Cc. Cc. c.
Human ‘‘end-piece’’...... 0 0 0 0 0 0
Guinea-pig’s serum........ 0.005 ce.| 0.01 ec. | 0.02 cc. | 0.04 ec. | 0.06 ce.| 0.1 ce.
= ae ote cae “te ak a5
Human “‘end-piece’”’...... 0.005 cc.}0.01 ec. | 0.02 cc. | 0.04 cc. | 0.06 cc. | 0.1 ce.
mk. mk. dist. dist. dist. tr.
Guinea-pig’s serum........ 0.005 ce.| 0.01 ce. | 0.02 cc. | 0.04 ce. | 0.06 cc. } 0.1 ce.
Ag = oe a ote ate he
Human ‘‘end-piece’”’...... 0.01 ce. | 0.02 cc. | 0.04 cc. | 0.08 ec. |0.12 cc. | 0.2 ee.
mk. tr. 0 0 0 0
EXPERIMENTS WITH HORSE’S SERUM
Certain experiments with horse serum yielded very interesting
results. As regards the hemolysis of ox corpuscles in the presence
of venom, horse serum often exhibits a powerful activating effect
and this property is partially retained even when the serum is
heated to a temperature of 56°C.
It is apparent therefore that this activating power is not
entirely due to serum constituents of complement nature.
Kyes (10) assumed that venom activation by serum was
entirely due to the lecithin present in the serum and that the
lipoid existed in combination with the serum proteins. In the
» Oe
+
€
COMPLEMENTS OF DIFFERENT ANIMALS 387
case of horse serum he suggested that the lecithin was very lightly
bound and was therefore available in the fresh serum.
Horse serum has no complementing power for ox’s corpuscles
along with immune body (see Muir (11)).
The globulin and albumin were separated by the ammonium
sulphate method and tested with venom and immune body.
TABLE 6
LYSIS OF 0.5 cc. OF 5 PER CENT OX BLOOD SUSPENSION
+ 0.0005 gram cobra venom + 5 doses of immune body
HORSE SERUM
fe fe lelelelslelelelelelelele
a Pa We a A= og a
o So So So —) Co So oe —) oe o —) —] i]
Paces serum... 22... 2... jc. | cc Jelejetletle|/0{/0}/0)/0;0;0);0
Globulin (NH,4)2SOx,
Tene A tr. |j.c.}e.|e}e}e}e]/0}/0)/0)/0/0/0)0
Albumin (NH4)2SO,
matin wer. .......) dist. | c. | c.}¢.)/¢. |e /¢10)/0]0)/0;010;0
Globulin+ albumin
(NH,)2SO, method....} j.c. | ec. }e.}e./e}e¢}e.}0)/0/0)}0;0)0/0
+ 0.0005 gram cobra venum
(70°C.)
HORSE SERUM
cease CNEL.) 90, Method. ......2..2..0)...00 2k 0 | 0/0] 0 | ftr
Albumin (NH,)2SO, method. iiss. BBL isthe. jie:. |Geui)-c.
Globulin+ albumin (NH,):80, meshed detaya. oh. | TOISEAMGS. | C4 (Gee:
Controls: No immune body nor venom.
0.25 cce.=no lysis.
0.5 cce.=no lysis.
Fresh serum: 0.5 ee.=no lysis.
Globulin
It was found that both the globulin and albumin individually
displayed marked activating powers and the combination of the
two fractions represented the full hemolytic power of the serum,
of course only by a process of summation of effects (table 6).
To ascertain whether the serum lecithin played some part in
producing lysis along with venom, these two fractions both
388 T, J. MACKIE
together and separately were tested with ox corpuscles + cobra
venom which had been heated to 70°C. for one-half hour according
to the method of Morgenroth and Kaya (12), who showed that
with heated venom, lecithin is actively hemolytic while comple-
ment is inactive.
It was found on carrying out these experiments that the
globulin fraction was quite inert with heated venom while the
albumin was as active as with fresh venom (table 6); the lecithin
nature of the activating constituents of the albumin fraction
was thus demonstrated.
While therefore lecithin bodies play some part in the activating
effect of horse’s serum, other constituents probably of comple-
ment nature are equally concerned. It is of interest also to note
that the lecithin substance should be associated with the albumin
fraction. The other activating elements are contained in the
globulin.
DISCUSSION AND CONCLUSIONS
These experiments elicit striking differences in the constitution
of the complements of different animals apart from their relative
activity with hemolytic immune body and venom.
In the case of human and rabbit’s serum acting on ox red
blood corpuscles + immune body or venom, the complement
is entirely associated with the globulins of the serum while in
the case of guinea-pig’s serum which represents with these
hemolytic systems a much more powerful complement, the
albumin fraction is also an essential constituent of the comple-
ment.
Whether the potency of a complement depends on the presence
of constituents associated with the serum albumin is a matter
for further investigation.
In the case of human and rabbit’s sera, however, acting with
venom, the effect of the globulin is ‘‘masked”’ in the whole serum
by the albumin while in the case of guinea-pig’s serum the
albumin also contributes to the full action of the serum along
with the globulin.
COMPLEMENTS OF DIFFERENT ANIMALS 389
It has also been shown how the albumin of human and rabbit’s
serum may inhibit the action of guinea-pig’s serum globulin.
In the case of horse’s serum the activating effect with venom
is due not only to a complement body represented by the globulin
but also to the lecithin contained in the albumin fraction.
REFERENCES
(1) Steruens, J. W. W.: Jour. Path. and Bacter., 1900, 6, 273.
(2) Brownina, C. H., anp Macxiz, T. J.: Jour. Path. and Bacter., 1912, 17,
120; Biochem. Zeitsch., 1912, 48, 229; Zeitsch. f. Immunitits., Orig.,
1913; 17; 1,
(3) Ferrata, A.: Berl. klin. Woch., 1907, 44, 366.
(4) Branp, E.: Berl. klin. Woch., 1907, 44, 1075.
(5) Lizrmann, H.: Miinch. med. Woch., 1909, 56, 2097.
(6) Sacus, H.: Handbuch d. Tech. & Meth. d. Immunitits., 1909, 2, 969.
(7) Omoroxkow, L.: Zeitsch. f. Immunitits., Orig., 1911, 10, 285.
(8) Brownine, C. H., ano Mackin, T. J.: Zeitsch. f. Immunitits., Orig., 1914,
21, 422.
(9) Cuicx, H., anp Martin, C. J.: Biochem. Jour., 1913, 7, 380.
(10) Kyss, P.: Jour. of Infect. Dis., 1910, 7, 181.
Enruicu, P.: Studies on Immunity, 2d ed., 1910, p. 291.
(11) Morr, R.: Jour. Path. & Bacter., 1912, 16, 523.
(12) Morcenrorta, J., AND Kaya, R.: Biochem. Zeitsch., 1908, 8, 378; Biochem.
Zeitsch., 1910, 25, 88.
ON THE PLACENTAL TRANSMISSION OF SO-CALLED
NORMAL ANTIBODIES
II. ANTITRYPTIC-ACTING BODIES
G. C. REYMAN
From the State Serum-Institute, Copenhagen, Denmark. Director, Th. Madsen
Received for publication May 6, 1920
In 1902 Halban and Landsteiner (1) observed that the blood of
child-bearing mothers contained a greater amount of antitryptic-
acting bodies than the funicular blood of their offspring, and in
1909 this question was taken up nearly simultaneously by Graef-
enberg (5), Becker (6), v. Reuz (7) and also Lust (8). They all
determined the antitrypsin content of the blood by means of
v. Bergmann and Bamberg’s (3) and Kurt Meyer’s (4) modifica-
tion of the Gross-Fuld (2) casein method. Graefenberg found
that the antitrypsin titer was nearly doubled during pregnancy,
and then again became normal shortly after parturition. Becker
examined twenty-five new-born children and their mothers with
the result that titers of the new-born were on an average nearly
the same as that ordinarily encountered in adults; on the other
hand they were somewhat lower than those of the mothers, the
titer of the latter being considerably heightened. v. Reuz exam-
ined the blood of diseased as well as healthy children, from six
days to eleven months old, and reported that the values were all
very low before the onset of the sickness. Lust examined the
blood of ten normal children, from fourteen days to more than
two years old, and found the titers almost as high as in adults.
Gammeltoft (9) improved the casein-method by undertaking
his measurements with a casein solution of a special hydrogen-
ion concentration, and by determining the nitrogen which might
be titrated with formol during the process. He corroborated
the presence of the greater antitryptic power of the mother’s
serum as compared with the corresponding funicular serum of
391
392 G. C. REYMAN
the children and also the increase during pregnancy, and he
further found that during the same period no increase took place
in gravid cows and rabbits. Gammeltoft supposes that the
antitryptic action is due to lipoid compounds, and bases this
supposition upon the fact that these bodies, as demonstrated by
Bauer and others, can be shaken off with ether, and also that
the bodies in question are suspended. Gammeltoft gives in
conclusion the observations made by Morgagni, Hunter and
Virchow, that the serum of gravid females is more opalescing
than is serum ordinarily, and that this opalescence, according
to Virchow, is due to phosphorated fat. In my own experi-
ments, also, there was an apparent correlation of opalescence
and enhanced antitryptic power of a serum; for the serum of
kids was generally opalescing, so much so, in fact that on stand-
ing a cream-like layer might form on the surface, whereas the
serum of the mother animal, before as well as after parturition,
was usually clear; and, in concurrence with this difference in
appearance, the serum of kids, as will be shown later, was found
to inhibit tryptic action more strongly than did the maternal
serum.
The antitryptic action of the serum was tested by the writer
in seven goats and their kids, among which were three pairs of
twins; of the latter one kid, though normal, had to be killed
immediately after birth, and as the first sample from the corre-
sponding twin was lost, the sample of the killed kid made a good
supplement. The first blood sample was taken from the jugular
vein, immediately after birth and before the kids had taken nour-
ishment. All test samples were subsequently taken from the
jugular vein and at the same hours.
The measurements of the antitryptic power were determined by
means of the Gross-Fuld method, as modified at the Serum-Institute
for quantitative use. First the entirely digestive dose of trypsin was
tested against 2 ec. casein solution (1 gram casein dissolved in 100 ce.
*, NaOH neutralized with 4 HCl and diluted with a 0.9 per cent
dilution of sodium chloride to 500 ec.). Based upon this preliminary
experiment a suitable excess of trypsin was used in the actual experi-
ment, which was made in the following manner: to a series of test tubes
PLACENTAL TRANSMISSION OF ANTIBODIES 393
decreasing doses of serum were added (with interyals of 20 per cent),
after the suitable quantities had first been determined by means of
preliminary experiments. Following upon the volumetric correction
of the contents of the tubes (total volume 1 cc.) the trypsin-solution
was added, after which the tubes were left for twenty minutes in a
waterbath (37°C.). After they had been cooled with running water
2 cc. casein solution per tube was added, and, after being well shaken,
they were again placed at 37°C. for three quarters of an hour. The
tubes were then cooled once more and 0.1 ce. acetic acid solution (5
ec. acetic acid + 45 cc. absolute alcohol + 50 cc. water) was added.
Thereafter a tube with a definite opalescence was chosen and in each
of the series the tube with corresponding opalescence was found.
The measure was repeated with another tube with a different degree of
opalescence. The mean of the reciprocal values of these two measure-
ments are the figures given in the tables.
A comparison between the figures of the table is only permis-
sible within the same experiment; i.e., a mother animal with
corresponding kids, seeing that all the samples originating from
a group of this kind were measured at the same time. Further-
more, the series with decreasing doses from each serum sample,
including the eventual minima and maxima of the antitryptic
action, extended over such a range that they embraced the same
serum dosage for both mother and kid, and thus the effect of
equal doses of serum against the same quantity of casein solu-
tion and in conjunction with the application of like dosage of
the same trypsin solution was observed.
The method of Gammeltoft (9) is more expedient, in so far
as various experiments might be compared from a quantitative
point of view, but it is difficult to use it when, as in this case,
we are dealing with long series of many serum samples, and the
Gross-Fuld method seemed to me to offer a sufficient founda-
tion for my purpose, which was in isolated experiments to de-
termine the relative proportion of antitryptic-acting bodies in the
blood of the mother animal and the kids, and their fluctuations
during the experimental period. Finally it turned out that
repetitions gave satisfactory results.
As regards the titers found, it appears from the table that the
titers of the kids, as contrasted with the results of the other ex-
S8T Z81
£08°% = |sze"z
19 aL
68 822
gcse |L9T'€
96 Ll
£/oe | ¢/8
08g
06°
T&T
IST
619°%
tL
£/12
FLI®
e/lSs
¢ PIM
108
11%
Or
£/L1
LST
608° n
02%
198° g
eh
gee
$99°Z
eal
S8T
822° s
ete
Orr
8/II
LOI
280°%
08
L8T
999°
el
e/LT
608
208"
F&I
cor
110°%
28
€/91
091
OST
196 'T
LL
0/92
SPT raat
ShS'T I|PI9'T
08 88
P61
902°%
88
&/II &/L
&1Z
9ST
IeI
est LOT
aIZ T= |199'T
LOT LOT
&/L e/8
eh Lgl
gest |sso'T
8 88
&/2 £/8
9ST SLI
6821 09°T
18 IIT
0/82 | &/8%
C93 982
e8t'r |190°T
912 (a6
o/LT a FL
008
€6F'T
802
6/S6
162 682
09°T bP
81 991
G/1Z 2/81
WOUds Say
8ST
oF I
80r
elt
909 °T
Sor
£/8
ALY
PLT
o/II
88
cee 1
gat
GLI
OPT
for
3/82
6L41
(a |
921
202
92'T
POT
6/93
IST
£02 °T
ra
6/83
99T
2/6
POT
80r'T
G26
G/L
602
L8T
G/Toe
C/1S%
SI
G/F
983
G/6«
Cts
G/ Gx
002
G/Se
elo
aQqo elo asQo Ba2o0 alo
esQQo
Or
‘}
|
|
|
|
Ser
*10JOUJ-}YIIOM OY} UT possordxo you sr yout ‘A41pider
[snenun GPA Mold pus os1u] AJOA O1OM Spry osoyT, “YHIG Joyye Avp YQANoy oy} uO FYs OM oY} yurod Furjrvys Aur sb Furye7 Aq JposAur yuW0x109 0} pey J Z puv | spry sper ia al ee ce ca like Ua eh NLS i a *a]0N
€9 29 |89 |22 |Z2 |€8 |s8 |28 |z8 |SoT | SZ 62 6S 501
€/08|8/S2/8/12/8/L1/8/P 18/11 \8/2 |€/8 |0/8z|Z/92|Z/he% | Z/2T | 6/9 LP6
#6 |28 |I2T |ZTI 96 68 |16 JTFT |2et | 16 CIT | 86 8
€/L2|/€/02|8/9T 8/11 8/2 |€/8 |2/82|Z/Ss|%/Es\Z/Te« | Z/2t | 3/9 PL
06 |28 {OTL |Szt |ggr |szt | sor | ¢6 08 9
€/9T/8/TT8/L |8/8 |2/8|e/Sz} Z/8S |Z/Te | 3/21
so lox loz tz les les | es | os dp Ss
G/L1/8/11|8/F |2/8c|%/Sc|0/1| G/9T |Z/tT x eet oD
OOT |FST |Z8T |G2ZT |GLT G2 |Z9T |G2T | Gzt | Szt | OOF de
G/GC\Z/TS|S/6T\Z/91\S/FL|C/GT|G/1T\Z/01] Z/6x | 3/9 |2t/0z
GZ |S8I |SET |OZT |SST |SeT |SeT | Sor | ost | 22 22
G/18|6/816/9T|Z/FTIS/11|6/6 |Z/L | 3/9 | G/S« |21/02 o
9ST |O9T JOLT JOLT |00% JO9T |OFT | OFT | OFT | OOT PI
G/12|6/8T|0/9T/2/FTIS/TT/6 10/2 | 3/9 | G/S« | 1/%
aia
WOUGS SUTHLOW RO LMETON
a a a ea a i, ee
‘7YF1oM OF pozE[or109 10414 ‘Oo f10j0By 4yBIOM ‘q {A]}O0IIP puNoj 104I], “e
‘sAUpyyig, ‘sojdurvs windtes Surye, Jo soyeq
SUUNJAS SPly PUD JOULIIDU UL UrsdhijyuD fo UOYDVUT
T WTaV.L
ae
PLACENTAL TRANSMISSION OF ANTIBODIES 395
perimentators on human beings and other animals than those
used for my experiments, in all cases are higher than those of the
mother animals; whereas the increase in the titer of the mother
animal, which has likewise been demonstrated in all of the cases,
agrees with the findings of other investigators. In some of the
350: Ontitrypsen uniies pet cm serum
o—o Mothet-serum
b----4 serum of kid &. divecl values eis
G9 Weisht correlated values pe
Q-----O setumol kid T.divect values é
=
.
<7
\ .
Oo
x o,” . ea
y & sereeseeee fey NS Y af See Ae 2226
Fn... : “o-" > ee
We iites 4 .. OS erste
& 1 et -
\,
ES cerca Sn seasons BS ms asp eee
"Oeoee..,.
88850. QSuveceeecres ee Ques...
—-O
Q\ 2
me
v
pee]
Airside’ oe ne
CuHartT 1
mother-animals this increase continues after birth, following
which—as in human beings—a decrease takes place; whereas
in others the titer remains constant through the whole of the
experimental period (up to somewhat more than a month). .
From the values which in the table are marked ‘‘a;’”’ namely the
396 G. C. REYMAN
directly found titers, it appears that in the kids there is in most
cases a gradual decrease of the antryptic acting bodies after
birth; if, however, one looks at the weight-correlated values
marked ‘‘c’’? (found by dividing the weight at birth into the
weight of the moment and multiplying the titers with the re-
sulting factor marked “‘b’’) it will be noted that, after some
fluctuation, there is an increase, a constancy or in a single case
a decrease in the titers. This decrease occurred in a kid (no. 3)
which from the first throve badly. It will be noted that the
increase observed in the case of the weight-correlated titers
may be somewhat different, even in the case of twins (7 and 8,
9 and 10), for it appears from the weight factors given that of
7 and 8 (which by the way had very nearly the same weight at
birth, that is, 2850 and 2750 grams respectively) 8 throve least in
the beginning and its titer at the same time sank below that of
7; later on the case was reversed, both as regards titer and
growth. Of kids 9 and 10, the weights of which at birth were
1800 and 2230 respectively, 9 grew proportionally more rapidly
than 10, and its weight-correlated curve also rose more quickly.
If one compares the weight-correlated figures, it appears that
five of the examined kids have an increasing and only one a
decreasing titer, whereas in three it is nearly constant, so that
the total quantity of the antitryptic acting bodies in the cases
examined on an average has increased.
The fact that the increase and decrease of the titer in certain
cases have turned out to follow growth or failure to grow might,
as suggested by other experimentators, indicate that we are
here dealing with a fat- or lipoid-action.
SUMMARY
By examinations of the proportion between the amount of
antitryptic-acting bodies in the blood of goats and their new-
born kids, the titers of the kids were in all cases found to be
higher than those of the mother animals.
The titers of the mother animal as a rule increase before
parturition.
PLACENTAL TRANSMISSION OF ANTIBODIES 397
There seems to be some connection between the growth of the
kids and the antitryptic-acting power of the blood, so that the
titer of the kid decreases when it thrives badly, and increases
when it thrives well. which would agree with the supposition set
forth by previous experimentators, that the antitryptic action is
connected with a fat or lipoid effect.
———— — — —
] REFERENCES
(1) HauBan anp LANDSTEINER: Miinch. med. Wochenschr., 1902, 49, 473-476.
(2) Gross: Archiv f. exper. Pathol., 1907, 58, 157-166.
(83) v. BERGMANN AND BamBera: Berliner klin. Wochenschr., 1908, 45, 1396.
(4) v. BERGMANN AND Meyer, Kurt: Berliner klin. Wochenschr., 1908, 45, 1673-
1677.
(5) GRAEFENBERG: Miinch. med. Wochenschr., 1909, 56, 702-704.
(6) Becker: Berliner klin. Wochenschr., 1909, 46, 1016-1017.
(7) v. Reuss: Wiener klin. Wochenschr., 1909, 22, 1171-1172.
(8) Lust, M.: Miinch. med. Wochenschr., 1909, 56, 2047-2051.
(9) GammeEtTort, S.: Meddelelser fra Rigshospitalet, Copenhagen, 1912.
Any eS runes
Hee) Cite: RAN F
tet i He ]
‘ Ney a”) {
Aedes det :
SIMPLIFICATION AND PARTIAL REVISION OF THE
FACTORS INVOLVED IN THE COMPLEMENT
FIXATION TEST FOR INFECTIOUS
ABORTION IN CATTLE
CHARLES S. GIBBS anv LEO F. RETTGER
From the Sheffield Laboratory of Bacteriology, Yale University, and from the Storrs
Agricultural Experiment Station
Received for publication June 5, 1920
The applicability of the complement fixation test in the
diagnosis of infectious abortion was demonstrated by Holth
(1909) whose observations were soon substantiated by Wall
(1911), Mohler and Traum (1911), Surface (1912), Hadley and
Beach (1912) and others. Holth made a comparative study of
complement fixation and agglutination and found that both of
these tests were specific for infection by the Bang bacillus.
They have since then been regarded by most investigators as
reliable and of much value in the hands of competent technicians.
Considerable criticism has been aimed, however, in recent
years at the complement fixation test because of the alleged
difficulties that were encountered in the carrying out of the
test, which has led in a few known instances at least to its aban-
donment and sole reliance on the agglutination reaction.
Both the agglutination and the fixation reactions are invalu-
able methods in the scientific study of the complicated problem
of infectious abortion. They have been employed in the present
joint investigation of the Sheffield Scientific School Bacterio-
logical Laboratory and the Storrs Agricultural Experiment
Station for almost six years, during which period over 4500
blood samples have been tested. With very few exceptions, the
two tests have served excellently as checks for each other, and
it is for this reason alone that they have been and are still
regarded as equally indispensable.
399
THE JOURNAL OF IMMUNOLOGY, VOL. VY, NO. 5
400 CHARLES S. GIBBS AND LEO F. RETTGER
It cannot be denied that the carrying out of the complement
fixation test as it has been developed in the past is beset with
difficulties which only the most careful and competent operator
is able to surmount. It has been the purpose of this study,
therefore, to attempt clarification of the phases which appar-
ently have given the largest amount of trouble.
The fixation test for infectious abortion in cattle has under-
gone but little, if any, real modification since the time of Holth
and Wall. There has been some variation in the methods of
preparing the antigen; some laboratories have adhered to the
serum broth culture method, while others have substituted agar
medium, with or without the fresh serum, for the bouillon.
The greatest difficulty seems to have been encountered in the
preparation and titration of the bacterial antigen.
In the regular fixation test the complement is obtained from
the guinea-pig, and the hemolytic amboceptor from rabbits that
have been immunized against washed sheep’s corpuscles, which
are employed also as the hemolytic antigen in the test. Four
different amounts of cow’s serum are as a rule used, namely
0.2, 0.1, 0.05 and 0.02 ec. The cow’s serum, complement and
bacterial antigen are mixed with 1.5 ce. of physiological saline
solution and incubated for one hour preliminary to the addition
of the hemolysin and sheep’s corpuscles. The final incubation
is for two hours. For a complete description of the technic of
the complement fixation test in infectious abortion, including
the preliminary standardization of complement, hemolysin and
antigen, the reader is referred to the work of Surface (1912),
Hadley and Beach (1912) and Rettger and White (1918).
The only modifications of any importance which have been
made in connection with the present investigation of infectious
abortion were the adoption of the Wenner (1918) method of
bleeding guinea pigs, the preservation of sheep’s corpuscles by
the Bernstein and Kaliski method of formalinizing (1912),
and the intravenous injection of small but increasing doses of
washed, undiluted sheep’s corpuscles in rabbits for the purpose
of more rapid and certain hemolysin production.
The chief aim in the present investigation has been to simplify
—E—E—EE——
El —
.
INFECTIOUS ABORTION IN CATTLE 401
and standardize the technic of the complement fixation test for
infectious abortion in cattle. The main emphasis has been placed
on the preparation and titration of the antigen, though the other
factors involved in the fixation scheme have received their due
share of attention.
COMPLEMENT
During the past two years the Wenner (1918) method of bleed-
ing guinea-pigs has been employed in this laboratory. During
this period the loss of animals from injury sustained in the
operation has been negligible. In many instances blood was
drawn from the same guinea pigs repeatedly without any appar-
ent untoward effect on them, and without any decreased com-
plement potency of the serum. The saving of guinea-pigs by
this method was a source of much satisfaction. It may be said
without exaggeration that a very small number of large guinea
pigs which are used solely for furnishing complement will supply
enough serum for routine daily fixation tests, and that if facilities
for breeding are reasonably good there will be little, if indeed
any, occasion for replenishing the stock.
The description of the method of drawing blood, as given by
Wenner, was as a matter of necessity somewhat incomplete.
For this reason it has been regarded by some technicians as
perhaps crude and impracticable. An effort has been made in
the present work to refine the technic and to make it so simple
that even the unskilled operator can employ it successfully.
As now conducted in this laboratory the technic should be
acquired easily. The following is a complete description of the
method, including the present refinement.
A large guinea-pig is selected for the operation, preferably
a young healthy male. The animal is suspended from a sup-
port by means of a cord drawn around the hind legs. The
head is immediately grasped with the left hand so that the thumb
comes under the lower jaw, and the body is turned ventral side
up. In this way the animal is easily controlled by very light
pressure. The head is turned back at right angles to the dorsal
aspect, the hair on the neck clipped as close to the skin as pos-
402 CHARLES S. GIBBS AND LEO F. RETTGER
sible, and the bared neck washed with 2 per cent solution of
cresol followed by 50 per cent alcohol. It is well to rub the neck
briskly with absorbent cotton until it is practically dry.
The guinea-pig should be etherized, of course, but only suffi-
cient anaesthesia should be given to make the animal insensible
to pain and the period during which it is under the influence of
the ether must be made as short as possible.
As soon as the vein from which the blood is to be secured is
located a transverse incision 10 to 15 mm. long is made with a
pair of sharp surgical scissors in the skin over the sternomastoid
muscle about half way between the masseter and subscapularis
muscles. The vein is then quickly brought to view and partly
or completely severed with the same instrument. The guinea
pig is lowered instantly to allow the blood to drain from the
wound into a sterile Petri dish or centrifuge tube, without
touching any part of the body.
The vein which is chosen for the operation (see diagram) is
an anterior ventral branch of the external jugular vein, and
corresponds in location to the anterior jugular vein of the higher
vertebrates. It les above the sternomastoid and the sterno-
hyoid muscles, but under the superficial muscles of the neck.
When it is cut while the animal is suspended at full length the
blood backs into it from the large external jugular vein and any
desired amount can be collected, in fact more blood can usually
be obtained in this way than by severing the external jugular
vein and bleeding the guinea-pig to death. The blood runs
slowly but steadily as long as the position of the animal is main-
tained.
As soon as the required amount of blood is obtained the guinea
pig is placed on its back on the table, and several stitches are
sewed through the loose edges of the cut skin. The flow of
blood quickly stops. After washing the wound with dilute
cresol or with alcohol the animal is returned to its cage. In the
course of two or three days the wound is completely healed and
the pig appears as strong as ever.
The preparation of the serum from the blood involves the
usual technic. The blood is allowed to clot thoroughly, prefer-
Pia =...
INFECTIOUS ABORTION IN CATTLE 403
ably in the refrigerator. The clot is broken up and the Petri
dish returned to the cooler for at least four or five hours. A
clear or slightly opalescent straw-colored serum is usually ob-
tained which, with rare exceptions, has good complementary
properties. When prepared for immediate use no attempts are
made to preserve it except by refrigeration in the ordinary ice
box.
_ ff}
HE NON
GaN
fi
Vif
Mf
be
i h WA
i) 1
jj
. Anterior facial veins 6. Internal jugular veins 11. Subscapularis muscles
. Posterior auricular veins 7. Subclavian veins 12. Masseter muscles
. External jugular veins 8. Sternomastoid muscle 13. Pectoral muscles
. Cephalic veins 9. Thyroid glands 14. First pectoral muscle
. Anterior jugular veins 10. Manubrium 15. Trachea
ore WDE
Both the Rhamy (1918) and the Noguchi (1918) methods of
preserving the complement have been employed by the writer,
with considerable success, but mostly the former. The Rhamy
404 CHARLES S. GIBBS AND LEO F. RETTGER
method, as here modified, consists in adding 40 per cent of a 12
per cent solution of sterile sodium acetate to the serum. The
treated serum is allowed to remain in the ice box for two or three
days before using. At the end of this period the titer of the
complement is rather high, but the complement does not very
materially weaken for a period of three or four weeks.
HEMOLYSIN
For the production of hemolysin Coca (1915) advocates intra-
venous injection of three rabbits with 1 cc. of the washed sheep’s
corpuscles. ‘Two injections are made, the intervening period
being five days. The animals are bled five days after the second
injection and the hemolytic properties determined. Coca figures
that at least one of the three rabbits will furnish potent hemolysin.
Stitt (1918) injects intravenously 1 cc. of a 10 per cent erythrocyte
suspension, and five days later 2.5 cc. of the strength suspension.
A third and last injection, this time 5 cc., is given five days after
the second. Seven or eight days later 4 or 5 ce. of blood are
drawn from the ear and the titer determined.
The following method has been used for over two years in this
laboratory with excellent results. One-half to 1 cc. of well-
washed sheep’s corpuscles is injected into the marginal vein of
each of two rabbits; one or two days later 1 cc. of the same mate-
rial is applied in the same manner, and on the third or fourth
day after the first treatment a final dose of 1.5 to 2.0 cc. of washed
red cells is administered. Blood is drawn from the marginal ear
vein four or five days after the last injection, for the potency
test. If the titer is satisfactory one or both rabbits are killed
not later than the tenth day after the third treatment. If a
serum of still greater potency is desired a fourth injection (1.5
to 2 ce. of the corpuscles) is given, and four or five days later the
titer again determined. Formalinized corpuscles can be used
for the production of the hemolytic amboceptor providing they
are not more than a week old and are thoroughly washed with
sterile physiological saline solution just before using (see page
405).
INFECTIOUS ABORTION IN CATTLE 405
Occasionally a rabbit is lost through anaphylaxis. For this
reason, partly, two rabbits are chosen for the hemolysin pro-
duction. The use of more than one rabbit also increases the
certainty of obtaining a hemolytic serum of the desired strength,
though we have never failed to produce a satisfactory serum
by the method just described. With rare exception, the serum
has been so active that a dilution of 1 to 100 has been necessary
in order to determine the titer with any degree of accuracy.
Even with this dilution the exact titer has usually fallen between
0.01 and 0.05.
; SHEEP’S CORPUSCLES
The method of preserving sheep’s erythrocytes described by
Bernstein and Kaliski (1912) and later by Wenner (1918) has
been employed with considerable success for the past two years
in this laboratory.
Immediately after the blood has been drawn from the jugular
vein of the sheep and defibrinated in a sterile bottle containing
shot or glass beads it is strained through absorbent cotton
and thoroughly mixed with a 40 per cent (commercial) solution
of formaldehyde in the proportion of 1 cc. of the undiluted formal-
dehyde to 800 cc. of the defibrinated sheep’s blood. When
preserved in this manner the erythrocytes will keep, if held at a
uniform temperature in a refrigerator, for fully three weeks at
least. It is very important that the bottle, filtermg funnel and
everything else with which the blood comes in contact be sterile.
The corpuscles are washed when needed. ‘The desired amount
of blood is then withdrawn from the bottle with a sterile pipette
and the corpuscles precipitated in the centrifuge and washed
with physiological saline solution in the usual way. For titra-
tion work and for the final complement fixation tests 2 per cent
suspensions of the washed corpuscles are made in saline solution.
BACTERIAL ANTIGENS
The preparation and titration of bacterial antigens have pre-
sented greater difficulties than any other phase of the comple-
ment fixation test for infectious abortion. It is largely because
406 CHARLES S. GIBBS AND LEO F. RETTGER
of these difficulties that the test has been abandoned in many
laboratories. Very little has been done in past years to simplify
this part of the general technic.
The best success in the preparation of B. abortus antigen has
been attained in this laboratory by the use of slant agar cultures.
It has been quite apparent that the choice of peptone for the
agar medium upon which the organism is grown is an impor-
tant factor; also that temperature and period of incubation must
be taken into account seriously.
In the present investigation an attempt was made to devise a
uniform and reliable method of antigen production. This
involved a study of the influence of age of bacterial cultures,
different brands of peptone, different strains of B. abortus, and
of the initial hydrogen ion concentration of the medium, on the
antigenic properties of the bacterial growths.
1. Relation of age of B. abortus culture to antigenic potency
In the following experiments the antigen was prepared from
slant agar growths of B. abortus, Bang strain, which were incu-
bated at 37°C. under aerobic conditions. Fairchild’s peptone
(1.0 per cent) was employed. ‘The growths were washed from the
agar with physiological saline solution, and the bacterial suspen-
sions filtered through absorbent cotton. ‘The suspensions were
shaken vigorously in glass-stoppered bottles for thirty minutes,
after which they were heated in a water bath at 62°C. for one
hour, with the bottles submerged to the neck. Immediately
following thorough cooling the suspensions were carbolyzed
(0.5 per cent). From these stock suspensions the required dilu-
tions were made and the antigen titers determined. The agar
used in the preparation of the different antigens was as nearly
uniform in composition and reaction as possible.
In the first experiment the antigen (a) was prepared from agar
cultures which had been incubated at 37°C. for four days and then
kept in a dark cool closet (15 to 18°F.) for six weeks. The final
suspension was diluted 1:5 with carbolized saline solution.
Table 1 is a record of the results.
INFECTIOUS ABORTION IN CATTLE 407
TABLE 1
ANTIGEN (a)
With immune abortion serum Without immune serum
0.01| 0.03 | 0.05| 0.08| 0.1 | 0.2 | 0.05} 0.1 | 02103 ]04 105 | 06 | 08
cc. cc. ec. ce. cc. cc. cc. cc. ce. cc. cc. ce. ce. cc.
Hemolysis..... 4 Ce ae ee AAD ACA | AL PAS |e
+ indicates complete hemolysis, — absence of hemolysis, and A anticomple-
mentary properties.
The titer of the antigen is 0.05, and from this standpoint quite
satisfactory, but anticomplementary factors are encountered in
as low a dilution as 0.1, which in itself renders this antigen
useless for the complement fixation test which, according to the
technic very generally employed, calls for 4 units of antigen, or
in this case 0.2 cc.
In the next experiment antigen (b) was prepared from cul-
tures which had been incubated for five days at 37°C. and then
kept in the cool dark closet for 10 days. The stock suspension
was diluted 1:10 and titrated, with the results shown in table 2.
TABLE 2
ANTIGEN (b)
With immune abortion serum Without immune serum
0.01 | 0.03 | 0.05 | 0.08; 0.1 | 0.2 | 0.05} 0.1 | 0.2 | 0.3 | 0.4 | 0.5 | 0.6] 08
cc. | cc. | ce. | cc. | cc. | cc. | ce. | ec. | ec. | co. | ce. | cc. | cc. | ce.
Hemolysis..... +ytyi—f}J—-lf—-l|—-{]+/i+/AsJAsSAlJ—][— lees
The titer of this antigen (b) is 0.05, which means that 0.2 ee.
of the antigen would be needed for the regular fixation test, at
the concentration of which anticomplementary action of this
suspension begins to assert itself. This antigen is, therefore,
highly unsatisfactory.
The next antigen (c) was derived from agar cultures that were
incubated for six days at 37°C. At the end of this period the
growths were promptly removed from the medium and con-
verted into the final antigen preparation. The stock suspen-
sions were diluted 1:5.
408 CHARLES S. GIBBS AND LEO F. RETTGER
TABLE 3
ANTIGEN (c)
With immune abortion serum Without immune serum
0.01 | 0.03 | 0.05} 0.08] 0.1 | 0.2 | 0.05} 0.1 | 0.2 | 0.38 | 0.4 | 0.5 | 0.6 | 0.8
cc. | ce. | ce. | ce. || ‘ec |ec: | ee: | cc. |) ce: | ‘cc. ||'"co. || co.) tcosmince:
Hemolysis..... + —;/—};]—-/;]—-;]+;+;)/AsA}7—]}]—-J]-] -
The titer is again 0.05, and anticomplementary factors were
encountered in 0.2 ce. of the suspension. This antigen, also, is
useless on account of its marked anticomplementary properties.
The last antigen of this series (d) was prepared from agar cul-
tures (Fairchild peptone) which had been incubated at 37°C. for
only four days and then immediately washed off with the saline
solution and converted into the stock antigen. The final dilu-
tion was 1: 5.
TABLE 4
ANTIGEN (d)
With immune abortion serum. Without immune serum.
0.2 | 0.3 | 0.4} 0.5 | 0.6 | 0.8
CC. 41 CC. || Ces nCCam| sCOsmimcce
+]/+]}+]+]4+]4
0.2 | 0.05] 0.1
COs ECCen|ecs
b= se | 4
0.01 | 0.03 | 0.05 | 0.08 | 0.1
cee |, (ccs, || ce> | ee.) || "ce:
Hemolysis..... +} te] |= f-—] —
The titer in this instance is 0.05, while the anticomplementary
factors are practically nil. Only a slight inhibition of hemolysis
was observed in the last tube, that is the 0.8 cc. dilution. This
antigen was employed in several complement fixation tests with
very satisfactory results. Subsequent work on antigen produc-
tion has fully corroborated these findings.
The results of experiments thus far indicate that the time
factor is an important one in the preparation of B. abortus
antigen for the complement fixation test in infectious abortion.
The cultures should not be incubated longer than four to five
days, and the growths should be converted at once into the final
carbolized suspension, which must of course be kept at ice box
temperature to preserve its antigenic potency.
INFECTIOUS ABORTION IN CATTLE 409
2. Comparative study of antigens obtained by the use of different
brands of peptone
These experiments involved the use of three well-known
American brands of peptone, Fairchild’s and two others which
will be designated here as A and B. Witte’s peptone was not
included because repeated attempts to develop good antigen on
agar containing this peptone had given very unsatisfactory
results.
The preparation of the different antigens was carried out in
the same way as has already been described (pages 405-406).
Incubation was for four days, at 37°C. In addition, the turbidity
of the bacterial suspensions was controlled by the use of the
McFarland nephelometer (1907), and an effort was made to
have an agar medium with a hydrogen ion concentration as
near P,, 6.8-6.9 as possible. Tables 5, 6 and 7 are self explana-
tory.
TABLE 5
Giving the results (titration figures) for antigen grown on agar containing peptone A
ANTIGEN
With immune abortion serum Without immune serum
0.01} 0.03 | 0.05} 0.08} 0.1 | 0.2 | 0.05] 0.1 | 0.2 | 0.3 | 0.4 | 0.5 | 0.6 | 0.8
GG.) (CC. | cc... | ce. | cc. | cc. | ce: || ec. | ce: | cc. ec. | cc. | cc. |, co.
Hemolysis..... +i4titi}—-;-—-;}—-/]4+/4+/4+/4+/4+/A/sA/A
The antigenic titer is 0.08, and anticomplementary interference begins at
0.5cec. The growths on the agar medium were only moderate.
TABLE 6
Giving the titration figures for antigen prepared by the use of peptone B
ANTIGEN
With immune abortion serum Without immune serum
0.01} 0.03} 0.05| 0.08} 0.1 | 0.2 | 0.05) 0.1 | 0.2 | 0.3 | 04 | 0.5 . 0.6 | 0.8
COMImCC AN CC: 1 (CG: 9|_CC; |) CC. eo. sees |eGens| OG. 11160.
Hemolysis.....,; + | +}]—/—-—/]—]}—/]+]/+]/)4+]/+/]+/A/A]sA
8
2
()
°
3
The antigenic titer is 0.05, with anticomplementary action setting in at 0.5.
The growths on agar were scant, and the amount of suspension was small,
therefore.
410 CHARLES S. GIBBS AND LEO F. RETTGER
TABLE 7
Giving titration figures for antigen grown on agar containing Fairchild’s peptone
ANTIGEN
With immune abortion serum Without immune serum
0.01} 0.03 | 0.05} 0.08} 0.1 | 0.2 | 0.05] 0.1 | 0.2 | 0.8 | 0.4 | 0.5 | 0.6 | 0.8
cc. | cc. | cc. | ce. | cc. | ce. | cc. | cc. | ec. | ce. | ce. | ce. | cc. | co.
Hemolysis..... +1/P}/—-}]—-/}]—-]—-}]+/}/4+/}/4+},4+)/4+)4+)/4+)4+
P indicates partial reaction.
The titer is 0.03, and no anticomplementary action is exerted in any of the
dilutions.
The above results show the importance of selecting the proper
peptone for the production of antigen. The Fairchild peptone
possesses very pronounced advantages over the other two in this
particular instance of antigen production; that is for the com-
plement fixation test for infectious abortion. Other American
brands which have not been used in this comparative study may
be equally satisfactory, however.
The growths obtained on Fairchild peptone agar were quite
luxuriant as compared with those of the other brands employed.
The suspensions prepared directly from the agar slants were so
dense that it was impossible to make turbidity determinations
with the nephelometer without diluting at least five times. A
further dilution of 1:5 was necessary before making the antigen
titration. The titer of this antigen remained constant for at
least seven months. The results obtained with Fairchild pep-
tone have been duplicated again and again, and by following
the method of antigen production as revised or elaborated by
us no difficulty has been experienced in the preparation of
antigen for complement fixation work in connection with infec-
tious abortion.
Solutions of Fairchild peptone are decidedly acid, and much
care must be exercised in properly neutralizing agar that is
prepared with this peptone. A final hydrogen ion concentration
of P,, 6.8 appears to be the most favorable. The composition
of the agar medium as we have been employing it for some time
is as follows:
INFECTIOUS ABORTION IN CATTLE 411
PI tee Ni, S's 5 ale ao Actaha maaan a ee neo 1000 ce.
US and 6 (Cl 0) 2) 4 grams, or 0.04 per cent
CTS 2 DIDS Oren 001 9) ra 10 grams, or 1.0 per cent
IPPC MMSUISUE NN Caaf le lees ose s sa cnaces cone 16 grams, or 1.6 per cent
3. Antigen production by different strains of B. abortus
Seven different strains of B. abortus were employed in these
experiments. One was an old stock culture which was obtained
from the University of Wisconsin, and labeled ‘‘Bang.”’ It had
come originally from Denmark. The others were of different
ages varying from a few months to five years. Two of the
strains still refused to grow by the ordinary aerobic method.
Two of the newer strains had been isolated by the writers from
the intestine of prematurely-born calves of infected dams.
The hydrogen ion concentration of the medium (ordinary pep-
tone agar) was P,, 6.8.
Good growths were obtained on all of the tubes when Fairchild
agar was employed, though there was a decided difference in
favor of the older and of the aerobic strains. On agar contain-
ing the other two American brands of peptone previously men-
tioned (A and B) the growths were less luxuriant in the aerobic
cultures, and scant in the anaerobic. However, all of the sus-
pensions prepared from the readily-visible growths exercised
antigenic properties when subjected to the titration tests. The
most satisfactory results were obtained with the Fairchild agar
antigens. The antigen titer for the seven different strains was
fairly uniform here, varying only between 0.03 and 0.05.
The conclusion may be drawn, therefore, that the different
strains of B. abortus have antigen-producing powers, and that it
matters little whether one or another of the organisms is used
for this purpose, providing sufficient growth is produced within
the desired period of incubation (four to five days). Largely
because of the ease with which abundant antigen may be ob-
tained, the Bang strain has been used almost exclusively by the
writers. The advantages of employing polyvalent antigen may
be such that it will be advisable to make use of at least three or
four representative strains, instead of one. This we are doing
at the present time.
412 CHARLES S. GIBBS AND LEO F. RETTGER
4. Determination of the most favorable hydrogen ton concentration
for B. abortus antigen ‘production
It soon became quite evident, in the course of this work, that
the hydrogen ion concentration of the agar medium upon which
the organism is grown is a very important factor. B. abortus
has the property of changing the initial hydrogen ion concen-
tration of the medium over quite a range and in either direction
from the neutral point, to suit its own need, as Evans (1918) has
shown before us. It seems quite apparent, then, that this organ-
ism requires an optimum concentration for its best development.
By the colorometric method of exact determination of Clark
and Lubs this optimum range can be definitely established.
From the results of preliminary experiments it appeared as if
the most favorable concentration was at or near P,, 6.8. Con-
sequently this point was taken as a mean in subsequent experi-
ments. All of the cultures were grown in tubes of Fairchild
peptone agar of definite but different H ion concentration, and
the antigens carefully titrated. A number of interesting facts
were brought out.
First, it appears that, if the initial hydrogen ion concentra-
tion is P, 6.8, the reaction does not change during the growth
of the culture, and that antigens prepared from such growths
are more satisfactory than those obtained from culture tubes on
which the organisms themselves readjust the hydrogen ion con-
centration during the course of incubation. If the initial H ion
concentration of the agar is at any other point between P,, 6.6
and P,, 7.3 growth is slow until the optimum concentration
P,, 6.8 is reached. Unless this readjustment is very slow, as
we have found it to be when anaerobic strains of B. abortus are
used, normal growth takes place. Some time is consumed
however, at best, and there is some loss of antigen; furthermore,
if the usual incubation period is exceeded, there will be danger of
anticomplementary action when the antigen is employed in the
regular fixation test. Because of their greater ability to adjust
the reaction of the medium within the limits of the P,, 6.6 and
P,, 7.3 range, the aerobic strains are better adapted for antigen
INFECTIOUS ABORTION IN CATTLE 413
production than the anaerobic, irrespective of any other advan-
tages or disadvantages which they might possess. By anaerobic
strains are meant, of course, those strains of B. abortus which
require partial exclusion of oxygen for their development.
ADJUSTMENT OF HYDROGEN ION CONCENTRATION, IN THE PREPA-
RATION OF FAIRCHILD PEPTONE AGAR FOR ANTIGEN
PRODUCTION
Saturated solution of sodium carbonate is employed for the
neutralization of the medium. Instead of adjusting the reaction
at once to P,, 6.8, the optimum for good antigen production, the
point 6.5 is sought. By careful experimentation we have found
that if the initial hydrogen ion concentration, that is just before
sterilization, is reduced to P,, 6.5, the final concentration after
sterilization for fifteen minutes under 15 pounds of extra pressure
is ordinarily P, 6.8—the point desired. These observations
apply to nutrient agar containing Fairchild peptone, which is
quite acid, and when sodium carbonate is used as the neutraliz-
ing agent. Aside from the use of Fairchild peptone, and of the
above procedure for regulating the acidity, the method of pre-
paring the agar is the same as in our daily laboratory routine.
TITRATION OF THE ANTIGEN
By the use of the McFarland nephelometer we have been
enabled greatly to simplify and standardize certain steps which
are preliminary to the actual titration. Heretofore no exact
method of dilution of the bacterial suspension was followed.
As a result titrations frequently had to be repeated with different
dilutions of antigen until the final dilution was reached within
which the limits of the titration scheme fell.
We have found it desirable to prepare bacterial suspensions
which are quite dense; that is by washing off the slant agar
growths of B. abortus with relatively small amounts of the saline
solution. From the concentrated stock suspension thus pre-
pared a dilution is made with carbolized saline solution to match
tube 1.75 of the nephelometer set. This diluted antigen will, asa
rule, furnish the final titration figure in a single antigen titration.
By knowing the exact proportion in which the antigen was diluted,
414 CHARLES 8S. GIBBS AND LEO F. RETTGER
it is only necessary after the first and only titration to prepare the
proper dilution of antigen from the stock suspension, without
further use of the nephelometer set. This method has been
employed in all of the work of the past two years and has proven
itself reliable and time-saving.
The Zinsser (1918) method of preserving antigen has been
employed by the writers, with gratifying results. This differs
from the usual method in that the bacterial growths on slant
agar are washed off with 10 to 17 per cent saline solution, instead
of 0.85, and diluting this stock suspension with distilled water to
0.85 per cent sodium chloride content when needed. ‘This
method has been of particular advantage during the warm
summer months and under conditions of imperfect refrigeration.
We have had little difficulty, however, in preserving antigen by
the usual method over periods of at least five or six months.
OTHER MODIFICATIONS OF TECHNIC INVOLVED IN THE COMPLEMENT
FIXATION TEST
Titration of immune serum
In the titration of the antigen an immune serum of known
strength is required. Instead of conducting several fixation
tests with various dilutions of the cow’s serum, as has been cus-
tomary heretofore, one test can be made to suffice by employing
all of the serum dilutions at one and the same time, as is shown
in the following titration scheme.
Titration of immune serum
Poco 9 | RAB
2 | SP ecb elie TIME Re ‘up| TIME RESULTS
e|S2| £2 | 22 [EE EEE
a Nila Srales nla
cc. cc. cc. cc. cc. cc.
1 | 1.5 | 0.01 |0.045|0.12} Incubate |0.15| 0.5| Incubate | Complete hemolysis
2|1.5 | 0.02 |0.045)0.12) 1 hour |0.15| 0.5} 2hours| Partial hemolysis
3 | 1.5 | 0.03 |0.045/0.12| at37°C.|0.15} 0.5] at37°C.| No hemolysis
411.5 | 0.04 |0.045/0.12 0.15] 0.5 No hemolysis
5 | 1.5 | 0.05 |0.045)0.12 0.15} 0.5 No hemolysis
6 | 1.5 | None |0.045/0.12 0.15} 0.5 Complete hemolysis
The titer of the immune serum is 0.03 ce. This amount of the serum is
added to each of the first six tubes in the antigen titration.
i i ee ee
‘
i
d
INFECTIOUS ABORTION IN CATTLE 415
The use of positive and negative sera as controls
The descriptions of methods for conducting the complement
fixation test in infectious abortion, which have come to our
attention, have not provided for the use of special control sera.
The inclusion of a positive and negative control serum is of
much importance. Control sera may be obtained readily from
animals whose reactions have previously been determined, and,
if carbolized when fresh with 0.5 per cent phenol, they will keep
for at least six months. ‘The positive and the negative serum
are used in the same manner as the test samples, and at the same
time.
SUMMARY
The complement fixation test as applied to infectious abortion
is specific, and serves as a valuable method of diagnosis.
With the partial revision and simplification of technic pre-
sented in this paper the method should be thoroughly practical
and reliable.
The most satisfactory antigen was obtained on nutrient agar
containing Fairchild peptone, and having an initial hydrogen ion
concentration of P,, 6.8.
The incubation period of the B. abortus cultures should not
exceed four to five days, and the stock antigen suspensions should
be prepared immediately following the removal of the culture
tubes from the incubator.
Antigen suspensions with a turbidity of 1.75 in terms of the
McFarland nephelometer lend themselves readily for direct
antigen titration.
The Wenner method of bleeding guinea-pigs, with the present
refinement, is a very practical and economic one, and can be
mastered readily by the ordinary operator.
Complement stabilized with 40 per cent of a 12 per cent
sodium acetate solution retains its complementary properties
for three to four weeks.
Formalinized sheep’s blood may be used for three to four
weeks as the immediate source of hemolytic antigen in the
THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 5
416 CHARLES S. GIBBS AND LEO F. RETTGER
fixation test. Freshly-washed corpuscles from formalinized blood
may be used also as antigen for hemolysin production.
Positive and negative sera should be used as controls in the
final fixation tests.
REFERENCES
Bane, B.: Die Aetiologie des Seuchenhaften (‘‘infectidsen’’) Verwerfens.
Zeitschr. Tiermed., 1897, 1, 241-278.
Bernstein, E. P., anv Kauisxi, D. J.: The use of formalinized sheep cells in
complement fixation tests. Zeitschr. Immun., Orig., 1912, 13, 490-495.
Crark, Wo. M., anv Luss, H. A.: The colorimetric determination of hydrogen
ion concentration and its applications in bacteriology. Jour. Bact.,
1917, 2, 1-34, 109-136, and 191-236.
Coca, A. F.: A rapid and efficient method of producing hemolytic amboceptor
against sheep corpuscles. J. Inf. Dis., 1915, 17, 361-398.
Evans, Avice C.: Further studies on Bacterium abortus and related bacteria.
J. Inf. Dis., 1918, 22, 576-593.
Haptey, F. B., anp Beacu, B. A.: The diagnosis of contagious abortion in cattle
by means of the complement fixation test. Wisc. University Agricul-
tural Experiment Research Bulletin, June, 1912, no. 24.
Ho.ru, H.: Die agglutination und die Komplementbindungs-Methode in der
Diagnose des seuchenhaften Verwerfens der Kiithe. Berlin. tierartzl.
Woch., 1909, 25, 686-687.
McFaruanpD, J.: The nephelometer. J. Am. Med. Assoc., 1907, 49, 1176-1178.
Mouter, J. R., anp Traum, J.: Infectious abortion of cattle. 28th Ann. Rep.
Bur. An. Ind., U. 8. Dept. Agr., 1911, 147-183.
Rertcer, L. F., anp Wuits, G. C.: Infectious abortion in cattle. Storrs Agr.
Exp. Sta. Bulletin 93, Jan. 1918.
Ruamy, B. W.: Further studies on the preservation of complement by sodium
acetate. J. Am. Med. Assoc., 1918, 70, 2000-2001.
Srirt, E. R.: A method of producing rabbit’s hemolytic amboceptor. Practical
Bacteriology, Blood Work and Parasitology, 1918, 200.
Surrace, F. M.: The diagnosis of infectious abortion in cattle. Ky. Agr. Exp.
Sta. Bulletin 166, June, 1912, 303-366.
Watt, Sven: Ueber die Festellung des seuchenhaften Abortus beim Rinde durch
Agglutination und Komplementbindung. Zeitschr. Infektionskr.,
1911, 10, 23-55.
WewneR, J. J.: A note on bleeding guinea-pigs and on preserving sheep’s erythro-
cytes. J. Immun., 1918, 3, 389-393.
ZinssER, H.: Infection and Resistance. 2d edition, 1918, p. 71.
aa >
THE ANTIGENIC PROPERTIES OF GLOBIN, WITH A
NOTE ON THE INDEPENDENCE OF THE PROPERTIES
OF SERUM AND TISSUE PROTEINS, AS EXEMPLIFIED
BY THE ABSENCE OF ANTIBODY FROM THE GLOBIN
OF AN IMMUNISED ANIMAL!
C. H. BROWNING anp G. HASWELL WILSON
From the Pathological Department of the University and Western Infirmary, Glasgow
Received for publication June 20, 1920
In a previous paper (4) we described experiments conducted
with the serum of rabbits which had received repeated injections
of globin, the protein constituent of hemoglobin, derived from
guinea-pig’s blood. The results showed that globin could act as
an antigen and that, together with the corresponding antiserum,
fixation of complement was produced. A high degree of com-
plement-fixation was obtained with seven different specimens of
guinea-pig’s globin along with one or the other of two antisera.
The striking character of the reaction is shown by the fact that
with a suitable specimen of guinea-pig’s complement, after incu-
bation with 0.5 ec. 1 per cent globin solution plus 0.05 ee. of anti-
~ serum (55°C.) for one and one-half hours, more than 60 doses of
complement were required in order to hemolyse the test corpus-
cles, whereas 1 dose of complement gave complete lysis in the
presence of the same amount of globin alone, and 23 doses gave
complete lysis along with the antiserum in saline. (The minimal
hemolytic dose of complement for 1 cc. ox corpuscles plus 5
minimal hemolytic doses of immune body from the rabbit was
0.01 ec.) Globin when mixed with serum causes a precipitate,
but this precipitate per se. does not lead to complement-fixation.
Thus globin failed to fix complement along with normal rabbit
serum or with a variety of heterologous antisera from the rabbit.
1 We have pleasure in acknowledging a grant from the Carnegie Trust toward
the expenses of this work.
417
418 Cc. H. BROWNING AND G. HASWELL WILSON
In view of the suggestion that protein contamination was re-
sponsible for the development of the antisera which we obtained,
the following observations are worthy of note: (1) the solution of
globin gave no fixation of complement with anti-human serum;
(2) the antiglobin serum did not fix complement with guinea-
pig’s serum or acid serum-albumin; (8) anti-guinea-pig globin did
not contain hemolytic immune body for guinea-pig’s red cor-
puscles. Thus the most likely contaminations—from human
protein due to handling, from imperfect removal of serum from
the guinea-pig’s red corpuscles or from receptors of the red cor-
puscles which might have passed through the filter (as Muir and
Ferguson (6) observed)—were all excluded. The present com-
munication contains further work on the subject, which was in-
terrupted owing to the difficulties described later, but which it
appears advisable to publish, especially in view of the fact that
our main conclusion, that globin possesses antigenic properties,
has been questioned and the result, which we obtained, has been
attributed to contamination (Gay and Robertson (5, 7)).
PREPARATION OF GLOBIN
The method employed was based on that of Schulz (11), which
depends on the fact that when hemoglobin has been dissociated
by a suitable concentration of HCl in watery medium, the acid
hematin can be extracted in great part by ether and alcohol. In
regard to our preparations of globin, it is to be noted that (1)
the blood corpuscles were, to begin with, repeatedly washed with
the centrifuge in order to remove any trace of serum, (2) the
laked blood was first centrifuged and then filtered through Berke-
feld or Maassen candles to remove stromata prior to treatment
with acid. After extraction of the pigment, excess of acid was
removed by dialysis and the ether was evaporated off by warm-
ing. The details have already been fully described (4). The
globin solution equivalent to 1 per cent of washed blood sediment
is referred to throughout this work as ‘‘1 per cent” globin.
The solution of guinea-pig globin which has been subjected
to prolonged dialysis against distilled water is highly sensitive
EPR cet he.
ANTIGENIC PROPERTIES OF GLOBIN 419
toward salt, forming a precipitate in 0.85 per cent NaCl solution.
It is also precipitated from distilled water by traces of alkali;
thus N/20,000 NaOH caused clouding of the 1 per cent solution
after warming at 37°C. for several hours. The addition of
N/2,500 HCl to the salt solution prevented precipitation, but
N/5,000 acid did not suffice; on the other hand, when precipi-
tate had formed, the addition of N/2,500 HCl did not cause it
to dissolve completely. Thus the preparation possessed the
characters described by Schulz. As a rule, solutions were em-
ployed which had been dialysed for a shorter time (twenty-four
hours against soft tap water) and which, although reacting
neutral to litmus paper, retained sufficient acid to remain clear
in the presence of 0.85 per cent NaCl.
In addition to the two specimens of antiserum to guinea-pig’s
globin previously described, two antisera, also from rabbits, were
prepared similarly against ox globin.
FACTORS WHICH INFLUENCE COMPLEMENT-FIXATION BY GLOBIN
PLUS ANTIGLOBIN SERUM
It was found that the complement-fixation reaction was in-
fluenced to a marked degree by physico-chemical factors; namely,
(a) the relative proportions of antigen to antibody (zone phenom-
enon), (b) presence of acid or alkali (hydrogen-ion concentra-
tion), as well as by the individual properties of the complement-
containing serum and of the animal receiving the injections of
globin.
1. Quantitative relationships between antigen and antibody
The following is a characteristic example:
Ox globin (dialysed solution which gave no precipitate in 0.85 per
cent NaCl solution); three series were made—A, containing 1 cc. 0.33
per cent globin; B, containing 1 cc. 1.0 per cent globin; C, containing
1 cc. 2.0 per cent globin—each tube contained 0.05 cc. antiglobin serum;
ascending amounts of complement (guinea-pig serum) were then
added and the mixtures incubated for one and a half hours at 37°C.
490 Cc. H. BROWNING AND G. HASWELL WILSON
Finally the test corpuscles were added (consisting of 1 cc. 3 per cent
washed ox blood suspension plus 5 minimal hemolytic doses of immune
body from the rabbit in all the experiments unless otherwise specified)
and there was a further period of incubation for one and a quarter
hours. The readings were made, as in all such experiments, after
the tubes had stood overnight at room-temperature. The results
were as follows:
LYSIS WITH THE FOLLOWING AMOUNTS OF COMPLEMENT
SERIES
0.06 ec. 0.09 ec. 0.13 ec. 0.18 ce.
A None None None Distinct
B None Distinct Marked Marked
C Marked Marked Very marked | Just complete
Controls: Antiserum 0.05 cc. alone in 1 cc. saline plus 0.025 ce. of
complement = complete lysis.
Globin solutions alone plus 0.01 ce. complement = complete lysis.
On the other hand, when the amount of antiserum was varied
a point was reached any diminution below which caused a very
great falling off in the amount of complement fixed, as the fol-
lowing experiment shows:
Guinea-pig globin solution 0.5 cc. plus antiglobin serum in series
(A) 0.05 ec., in (B) 0.025 ec., and in (C) 0.01 ce.
LYSIS WITH THE FOLLOWING AMOUNTS OF COMPLEMENT
SERIES
0.025 ec. 0.035 ce. 0.075 ce. 0.1 ce. 0.15 ce.
A None None None Trace Incomplete
B Almost complete Complete | Complete | Complete | Complete
Cc Almost complete Complete | Complete | Complete | Complete
Controls: Antiserum 0.05 cc. plus 0.5 cc. saline plus 0.015 cc. of com-
plement = complete lysis.
Globin solution 0.5 cc. plus 0.015 ce .of complement = complete
lysis.
ANTIGENIC PROPERTIES OF GLOBIN 421
2, Hydrogen-ion concentration
The results previously published showed clearly that the amount
of complement which was fixed by guinea-pig globin plus anti-
globin serum depended on the hydrogen-ion concentration, and
it appeared as if the optimum lay slightly to the acid side of
neutral, just as Sachs and Altmann (8) had found in the case of
the Wassermann syphilis reaction. Thus in the case of a speci-
men of globin, which had been dialysed for a week against dis-
tilled water and which was highly sensitive to the precipitating
action of 0.85 per cent NaCl, the addition of N / 2,500 HCl more
than doubled the amount of complement which was fixed by
antigen plus antibody. On the other hand, it appeared that
too high a degree of acidity diminished the amount of comple-
ment fixed, as the addition of alkali to such acid solutions in-
creased the complement-fixation. The question of hydrogen-ion
concentration apparently has not been fully taken into consider-
ation by Gay and Robertson and Schmidt (9) as a possible ex-
planation of the failure to obtain complement-fixation in their
experiments.
The hydrogen-ion concentration, or the physical changes
which resulted therefrom, was probably also the determining
factor in the following experiment:
Guinea-pig globin (preparation V)—a 13 per cent stock solution
which when diluted to 0.9 per cent remained clear in 0.85 per cent
saline (= untreated globin): 10 ce. of this stock solution were precipi-
tated by saturation with NaCl; the precipitate was removed by cen-
trifuging and was then dialysed against running water for two days,
and finally diluted to 0.9 per cent (= salted globin). In the usual
complement-fixation tests 0.5 cc. of untreated globin plus 0.05 ce. of
antiserum plus 0.15 cc. of guinea-pig’s complement gave a trace of
lysis of the test corpuscles: 0.5 ce. of salted globin plus 0.05 cc. of anti-
serum plus 0.04 cc. complement gave complete lysis of the test
corpuscles.
Control: Each globin solution alone plus 0.01 ec. complement gave
complete lysis.
422 C. H. BROWNING AND G. HASWELL WILSON
8. The individual property of the complement-containing serum
The importance of individual properties of the complement-
containing serum in determining the amount of complement
which is fixed, has been pointed out by Browning and Mackenzie
(2) and Browning and Kennaway (1) in the case of the Wasser-
mann syphilis reaction and it has been shown that there is no
constant relationship between the hemolytic power and the de-
viability of the complement. It was shown also that by dialys-
ing complement the hemolytic power might remain almost un-
altered, while the deviability was greatly reduced (Browning
and Mackie (3) ). Similar variations in deviability were found
by us (4) in the case of fixation by globin plus antiglobin serum.
In addition, we have tested mixtures of guinea-pig globin plus
antiserum, which actively fixed guinea-pig’s complement, using
(a) rabbit’s complement and testing with the usual sensitised ox
blood suspension and (b) ox complement, with guinea-pig’s cor-
puscles as the indicator, the immune body in the latter case
being that naturally present in the ox serum. In neither instance
could complement-fixation be detected.
4. The individual character of the animal which receives the
injections of antigen
The methods which were used to develop antisera have been
previously described in detail and consisted in repeated intra-
peritoneal injections of globin in suspension or solution, or intra-
venous injections of solutions. The antisera to guinea-pig’s
globin were derived from the second and third animals injected.
The first animal apparently failed to yield demonstrable antibody,
but, as we were at the time unaware of the great importance at-
taching to the relative proportions of antigen and antibody and
of the optimum reaction, the antibody may have been missed.
Similarly, the antisera to ox globin were derived from the second
and third animals tested. But later, repeated attempts to ob-
tain further antisera failed both in the case of ox and of guinea-
pig globin and in spite of the knowledge gained as to the condi-
eee... .
SS aS ees eee
ANTIGENIC PROPERTIES OF GLOBIN 423
tions for eliciting the reaction in vitro. Thus we are forced to
the conclusion that globin is not a potent antigen or, in Ehrlich’s
language, that it does not produce a powerful “ictus immunisator-
ius” except in specially suitable individual animals. Of course,
the factor of variable responsiveness is a commonplace experience
in the practice of immunisation; but this may well explain why
Gay and Robertson (5) and Schmidt (9) failed to obtain an anti-
serum to globin from the three rabbits which they tested. Simi-
larly, in view of our experiences, the development of antisera
which fixed complement in the presence of globin alone, as the
result of injecting a compound of globin with casein into two
rabbits, may have been fortuitous rather than due to any special
effect of the casein, as explained by Gay and Robertson. In
any case the statements of these workers that (1) globin is not
an antigen but that (2) injections of globin-casein cause the de-
velopment of antisera which lead to complement-fixation with a
solution of pure globin, would appear almost to involve a contra-
diction in terms.
THE BEHAVIOR OF HEMOGLOBIN AS AN ANTIGEN
As was shown by us previously, an antiserum which reacted
intensely with globin (faintly acid solution, clear in 0.85 per cent
NaCl) failed to give practically any fixation of complement with
the same specimen of hemoglobin as that from which the globin
was derived, and in the corresponding concentration. The fol-
lowing is a further illustration:
Guinea-pig’s globin solution 0.5 ec. plus 0.05 cc. antiglobin plus
0.3 cc. complement = trace of lysis of test corpuscles. 0.5 cc. cor-
responding solution of hemoglobin plus 0.05 cc. antiglobin plus 0.04
ec. complement = just complete lysis.
Controls: Globin and hemoglobin solutions alone plus 0.01 cc. com-
plement = just complete lysis.
0.05 cc. antiglobin serum in 0.5 cc. saline + 0.03 cc. complement =
just complete lysis.
Thus neutral hemoglobin does not react with a potent anti-
globin serum. The failure of hemoglobin to react with the anti-
424 Cc. H. BROWNING AND G. HASWELL WILSON
globin serum appears to us to afford strong evidence in favor of
the antibody which we demonstrated being in reality developed
by the globin and not by some adventitious constituent, as sug-
gested by Gay and Robertson. In view of what has been stated
already the only possible contamination would appear to be bac-
terial; now it was the solution of hemoglobin which was especially
exposed to infection. During the subsequent procedure, involvy-
ing the addition of acid, alcohol and ether, conditions were much
more unfavorable for contamination and, as has been noted,
the stock solutions of globin remained perfectly clear for many
weeks. Attempts to produce antibodies to hemoglobin appear
to have been generally, but not universally unsuccessful (see
Schmidt and Bennett (10) for an extensive review of the litera-
ture as well as original experiments). Our own attempts to
produce antisera to guinea-pig’s hemoglobin in rabbits were like-
wise mainly unsuccessful and the serum also contained no lytic
immune body for guinea-pig’s red corpuscles, as tested with rab-
bit’s complement. But in one of our experiments marked com-
plement-fixation was obtained with several specimens of guinea-
pig’s hemoglobin along with the heated serum of a rabbit which
had received several injections of guinea-pig’s hemoglobin. In
view of the small number of our observations and the fact that
hemoglobin was necessarily tested in neutral solution, we hesi-
tate to deny that hemoglobin altogether lacks antigenic proper-
ties; but the proportion of animals which react, as shown by the
production of complement-fixing antibody, is small, and it is
evident that any antigenic power which it may possess is very
weak. Schmidt and Bennett’s failure to demonstrate antisub-
stance in the serum of eight rabbits after repeated injections
of hemoglobin is in agreement with this conclusion.
THE TOXICITY OF GLOBIN AND HEMOGLOBIN
Gay and Robertson found that their specimens of globin were
toxic, especially for guinea-pigs which apparently received an
intravenous or intraperitoneal injection of an acid solution. The
combination of globin with casein was stated to be non-toxic,
ANTIGENIC PROPERTIES OF GLOBIN 425
but ‘‘marked symptoms of prostration with polypnoea after each
injection” are mentioned as occurring in one of the two rabbits
which they injected with globin-casein. Our specimens of globin
caused no obvious ill effects in rabbits. Hemoglobin was found
to be non-toxic in our experiments as well as in those of Schmidt
and Bennett and of others.
THE SPECIES-SPECIFICITY OF GLOBIN
In our previous work the action of an antiserum to guinea-pig’s
globin was tested in parallel series with guinea-pig’s globin and
with preparations from ox and rabbit blood. In each case the
specimen of globin was thoroughly dialysed and N/2,000 HCl
added and the same concentration of globin and amount of anti-
serum were employed in each case. The result was that, as
compared with the amount of complement fixed with homologous
globin (taken as 100 per cent), rabbit globin caused the absorp-
tion of 16 per cent of complement and ox globin of less than 6
per cent. Thus a marked degree of species-specificity was shown
to exist. However, when globin from further species was tested
along with antisera both for guinea-pig and for ox globin results
were obtained which are highly complex and difficult to interpret,
as the following examples show:
A. Complement-fixation produced by mixtures of anti-ox globin
serum 0.025 cc. plus 0.5 cc. of different species of globin in 0.2 per
cent solution (tested with 0.5 ec. sensitised ox corpuscles: minimal
hemolytic dose of complement = 0.015 cc.)
LYSIS OF 0.5 CC. TEST CORPUSCLES WITH THE FOLLOWING AMOUNTS
OF GUINEA-PIG’S COMPLEMENT
SPECIES OF GLOBIN
0.025 ce. 0.04 ce. 0.065 ec.
(OD2 Lid pA 5. O See eee Faint trace Faint trace Faint trace
(Gatiny 2 OR eae, Seee ore ee None None None
iD Sate Se None None Faint trace
WORD ES Ai etnie seeks os ec Almost complete Complete Complete
Control: Globin solutions alone plus 0.015 ce. complement = com-
plete lysis in every case.
426 Cc. H. BROWNING AND G. HASWELL WILSON
Antiserum 0.025 cc. in saline plus 0.02 ce. complement = complete
lysis.
B. Guinea-pig and ox globin tested for complement-fixation in
parallel and in crossed series with the respective antisera.
: AMOUNTS OF GUINEA-PIG’S COMPLEMENT
0.5 cc. BEOBIN ANTISERUM
(0.33 PER CENT ae
: : 0.05 cc.
SOLUTIONS) 0.05 cc
0.1 ce. 0.13 ce. 0.18 ce.
Guinea-pig | Anti-guinea - Distinct | Marked | Very
pig marked
Guinea-pig | Anti-ox - Marked | Marked | Almost
com-
plete
Ox Anti-ox — Trace Distinct | Marked
Ox Anti-guinea | Just com-| Complete] Complete] Complete} Complete
pig plete
Controls: Globin solutions plus 0.01 ec. complement = complete
lysis.
Anti-guinea-pig globin in saline plus 0.04 cc. complement = very
marked lysis.
Anti-ox globin in saline plus 0.05 cc. complement = very marked
lysis.
Thus anti-ox globin fixes complement actively with goat and
duck globin. Anti-ox globin also fixes complement along with
guinea-pig’s globin; on the other hand, as we found previously,
anti-guinea-pig globin fixes practically no complement along with
ox globin. Any fallacy would appear to be excluded from the
latter results of ‘‘crossed”’ experiments by the fact that both
specimens of globin were tested with the homologous and the het-
erologous antisera at the same time and with the same comple-
ment. Jn both homologous series marked fixation of comple-
ment was observed, thus showing that the reagents were acting
satisfactorily. Again, anti-ox globin fixes little or no comple-
ment in the presence of rabbit globin. It was hoped to develop
the interesting lines of work suggested by these results, but this
had to be given up owing to the continued failure to obtain further
antisera. ‘Thus while evidence of species-specificity exists in
certain cases, there is also a wide, though not universal, community
——_—- =
ANTIGENIC PROPERTIES OF GLOBIN 427
of antigenic properties shared by the globin of widely separate
animal species. It will be difficult to explain these results on
any basis of hypothetical contamination.
SUMMARY
1. Globin can act as an antigen. In addition to two antisera
for guinea-pig’s globin, two antisera for ox globin have also been
obtained from rabbits. With these antisera and a number of
different specimens of the homologous globins powerful comple-
ment-fixation reactions have been obtained. The obtaining of
the reaction depends on suitable quantitative relationships be-
tween antigen and antibody and also, as was shown previously,
on a suitable hydrogen-ion concentration.
2. Only certain rabbits apparently respond to injections of
globin by the production of complement-fixing antibodies. The
injections of globin caused no obvious toxic effects in rabbits.
Hemoglobin seems to elicit antibody production more rarely.
3. The reactions with antisera show, in certain cases, marked
species-specificity of globin; thus the antiserum to guinea-pig’s
globin does not fix complement with ox globin. On the other
hand, anti-ox globin fixes complement along with goat, duck and
guinea-pig globin but not with rabbit globin; no explanation is
offered of the contradictory behavior of ox and guinea-pig glo-
bins in the crossed experiments. But the results taken together
seem to exclude bacterial contamination as the cause.
4, The evidence points to the phenomena being due to a gen-
uine antibody to globin and not to adventitious protein contami-
nation. Further recorded facts (namely, that anti-guinea-pig
globin does not contain hemolytic immune body for guinea-
pig corpuscles and does not react with guinea-pig’s hemoglobin,
serum or acid serum-albumin, and that guinea-pig globin does
not react with antihuman serum) also exclude the probable con-
taminations which might arise in the course of preparation of the
globin solutions.
428 Cc. H. BROWNING AND G. HASWELL WILSON
NOTE ON THE INDEPENDENCE OF THE PROPERTIES OF SERUM AND
TISSUE PROTEINS, AS EXEMPLIFIED BY THE ABSENCE OF ANTI-
BODY FROM THE GLOBIN OF AN IMMUNISED ANIMAL
An example of the independence of properties of tissue and of
serum proteins was obtained in the following experiment. Glo-
bin was prepared from the hemoglobin of a rabbit which had
been repeatedly injected with washed ox blood and which in con-
sequence contained abundance of the corresponding hemolytic
immune body in its serum (minimal hemolytic dose of immune
body for 1 ce. 3 per cent ox blood suspension plus excess of guinea-
pig’s complement = 0.0025 cc.). Amounts of the globin up to
0.5 ec. of a 6.5 per cent solution (both in saline and in saline plus
N/2,000 HCl, which barely sufficed to keep the globin completely
in solution) led to no hemolysis of 1 ce. of 3 per cent ox blood
suspension in the presence of 0.1 cc. guinea-pig’s complement.
Thus antibody was absent from the globin of an animal which
had reacted to an antigen by the development of powerful serum P
antibodies.
REFERENCES
(1) BRownInG AND Kennaway: Lancet, 1919, 1, 785.
(2) BROWNING AND MackeEnziE: Zeitschr. f. Immunititsforsch., Orig., 1909, 2,
459.
(3) BRowninG AND Mackie: Zeitschr. f. Immunitiatsforsch., Orig., 1914, 21, 422.
(4) BRowNING AND WILSON: Jour. Path. and Bact., 1909, 14, 174.
(5) Gay AND Rospertson: Jour. Exper. Med., 1913, 17, 535.
(6) Murr anp Fercuson: Jour. Path. and Bact., 1906, 11, 84; see also Muir’s
Studies on Immunity, London, 1909.
(7) Ropertson: The Physical Chemistry of the Proteins. New York and
’ London, 1918, p. 155.
(8) Sachs AND ALTMANN: Berl. klin. Woch., 1908, 45, 699.
(9) Scumipt: University of California Publications in Pathology, 1916, 2, 182.
(10) Scumipt AND BENNETT: Jour. Infect. Dis., 1919, 25, 207.
(11) Scuuuz: Ztschr. f. Physiol. Chem., 1898, 24, 449.
lll lr
THE PROTECTIVE VALUE OF PNEUMOCOCCUS
VACCINATION IN MICE AND RABBITS!
AUGUSTUS B. WADSWORTH
From the Division of Laboratories and Research, New York State Department of
Health, Albany, New York
Received for publication July 7, 1920
Although the fact is well known that inoculation of animals
with the pneumococcus develops against subsequent inoculation
with virulent cultures an immunity which is specific for the homol-
ogous type of pneumococcus, there is no record of a quantitative
determination of the degree of immunity that is obtained after
vaccination with standardized vaccine. Accordingly a series of
experiments with the inoculation of mice with pneumococcus
vaccines of types I, II, and III were made, the results of which
it is the purpose of this paper to record.
Miss Thelma L. Franklin, of the laboratory staff, prepared
vaccines of pneumococcus types I, I, and III according to the
standard methods used in the laboratory and vaccinated mice
with these vaccines. In order to test the protection obtained
as a result of the vaccination, living, virulent cultures were
inoculated seven days after the last dose of the vaccine. Stand-
ard cultures of pneumococci of types I, UH, and III are main-
tained in the laboratory at a virulence which fluctuates con-
siderably from time to time but which usually kills mice of
eighteen grams weight in a dose of 0.000001 cc. in less than thirty-
six hours. The variation in virulence in the normal or control
mice is well shown in the tables recording the results of the ex-
periments. On account of this fluctuation in virulence it is not
practical in experimentation with the pneumococcus to determine
1Presented by title at the Twenty-first Annual Meeting of the American
Bacteriologists, Boston, Mass., December, 1919.
429
Sy— T i T0000 63- I T 1000°0 | 6% i! I T000°0
iis I i T0000 '0 68— T T TO000°0 | OF i! I 100000
t= {! I 100000 0 ¥o—- \! I 100000°0 | LP I T 100000 0
ddIU [BUIIO N—S[O1}U0D OOIU [BUILO N—S[O1}U0D OO1UL [BUILO N—S[01} 00D
0&-96 G 0 G T00'0 SE-PE & 0 & 100°0 | $&-8z G 0 G 1000
VE-GS 3G 0 G 1000°0 VE-LG G T € 1000°0 | ZE-8z & 0 § 1000°0
hs 8E-66 € 0 & T0000 0 cg \! G € 10000°0 | 9&-8z & 0 & 100000
= IIT pun ‘TI ‘7 sadhiy, fo saunqynd y70.1g JUAIN.LLA YIUN UOYDINIOUL ISD) Lajfo shvp Waaas PI}sa} £1ajdUWY}UII 92GN9 wad
S swusvupb10 woyjrg 2914) Buxuwyuoo ‘aurvova JT adfiz fo +00 T fo suornpnoour Apyoom 9014) Buraraoas aovpy *(6T-FI-9) IT quawUrwad xq
D
g Se
is geo et, |_| y [tooo te sy |e es ot | to0000 0) ae a eee emma
a] OSNOUL [VULLO N—[01}U0D SNOUT [VULLO N—[O19U0D OSNOUL [VULIO N—]O1}U0D,
mM
P 91 & 0 & T000°0 Ge-L1 & 0 € 1000°0 | T&-F% 6 0 6 TO000°0
5 Go-LI & 0 € T0000 0 8E-ES 6 0 € T0000'0 | 98 G T 6 10000°0
} LY-86 & 0 & 1000000 VP-ES € 0 € TO0000'0 | 9F G i € TO0000° 0
B s4noy ‘90 sunoy “90 sunoy *99
pore] poz] poz¥]
: Pere pease enoour PeIp | PEAT | -noour perp | Peat | -noour
jeu ae a UL a5 rien roied Ayquend eee) et Ti ioaeq roca aise APPUBNY) [8 sie eee road roned oneal APPUBNH
“wn | -UMN | Sony “UMN | “WON | Sony “wn | -WON | Zong
Ill SQ090900OWOD ANd—dyowvto9 Il SQ09000WOANd—AUOLTAD I sn00000WOaANd—aNOALIN9
IIT pve “TI ‘7 sadhy fo saunqyno yj01g quayn.va Ypn UoYDjnI0Ur Lajfo shpp Waaas papsa} §sazaWyUII
auqno sad swsvunbso uoyjrg 0014) Burvumjuoo ‘auroona 7 adhy fo +09 J fo uoynpnoour U0 burardve4 a0rpy*(6I-8]-f) [ Wauruedagy
430.
431
VALUE OF PNEUMOCOCCUS VACCINATION
0&— I i T0000 °0 Le— T I 10000 0
€&— I! if 1000000 8E— I T 1000000
QIU [VUIO N—S]O1}UOD QOTUT [BUIIO N—S[01}U0—) dOTUL [BUIO N—S[O1} U0)
6 0 & 100°0
&o— G 0 G 100°0 6Z-GG G 0 0 € 6 1000°0
TS-O&— | @ 0 G 1000°0 OF-Zo € 0 0 & & 100000
| epee Die 8 0 & TO000' 0 OF-86 € 0 0 € & 100000 0
II] pun ‘TIT ‘y sadfy, fo saunqyno yj0.19 JUAINLLA YPN UOYNpNIOU?L 4s] Jaz fo shipp Uanas pajsa} fLajaULYjWaI 9LQnd
wad susvuvb..o u0yprig aa1y? Buruwjuod ‘aur20na J adhy fo +99 J fo suowyppnoour fipyaan 9a.y7 Buraraoas a0Upy —* (6-6-9) AT auiodxy
Mf Phas I I 100000 wae I if 10000°0 | 62—- T i 10000 0
its I I! 1000000 98— I i 100000°0 | %& i T 100000 °0
QOTUT [VUIIO N—S]O1} UD QOTUI [VULIO N—S[O1} U0) QIU [VUIIO N—S]or} UO
LZ-9B & 0 € 100°0 ZE-BZ G 0 G 1000°0 | S&-8z j 0 G T0000
PS-LZ-61 € 0 & T0000 8Z-96 & 0 & 10000°0 | 82-02 € 0 € 10000°0
OS-LZ & 0 & 10000 °0 Gc-9I & 0 & 100000°0 | S&-re & 0 & 1000000
III pun ‘Ty ‘7 sadhy, fo saunqqyna yj01g quaynia ypim woynjns0Ur Isp] 1azf/0 shop usaas pajsa} f1ajaMIyUaI 92qQnd Lad
swusvupb1o w07]19 92.14) Bururvjuoo ‘aursona pT edhy fo ‘00 J fo suoynpnoour fiyyaam xrs Buraraoas 90.yy “(61-01 -9) [J] Mautwadag
in
fo}
z
r
4
i}
>
-
o
°
a
°
Z
b
2
Ss
i]
fe
3
Z
%
p
i)
id
a
|
tol
AUGUSTUS B. WADSWORTH
432
8o— | Tal | 1 |to0000"o
9SNOUL [BUILO N—[O1}U0D
oo— G
Go— G
Go— G
sinoy
Petp
perp aorul
jeutue suit] Jo 13q
-UINN
I SQO0000WOUNd GANOAWTINID
UOUDINIOUL JSD) ay} Laz{o syaan Lnof pajsa) puv ‘asop
Y9D9 UI 19909 W0214]1q xIs Lo au209DA J adhy fo +99 g°Q fo s]DaLaqU2 fiDp-xIs 7D suO1DINIOUI Lis Buraradas BILTY
"(6I-18-£) [A Mawriaday
Visa I I 60000 0 ce — T T 60000 0 cG— I
o7— I T 1000000 oc I I 100000 0 tetany \!
9dIU [BUIIO N—[01}U0D 9OTUL [VULIO N—]or}U0D,
SAUp FL 0
O0S-GG G 0 G Z0000 0 9E-VG G 0 G 60000 0 LE I
Binoy 20 sunoy 29 sunoy
paz poyv| :
parp | paar | _P2?°1 pap | paar | _Postl parp
eae cet Toned roted aes Ay UEN® ee oat nonee wares ates APVUBNY Gals on LL concn
-wny | -wnN Sith, -uinN | -wnN ane -un N
Ill SQ99000WOENd—duoOLTIOO II SQA0N00DOWOANd—aAYOLTOND
UoyDnNI0Ur 48D] Layfo Shivp wUaaas pajsa? pun ‘asop
yIDa UL 10909 U01)]1q LIS 4 aurz9na J adhy [0°99 g'Q fo s]DasaqUL hop-xis JD SuOYDINIOUL g buradrd90.4 99L
I 10000
i TO0000 0
dOIUI [BUIIO N—]01}U0D
G SG 100°0
if G T0000
29
978
Peat | hoout
. Oorul Aqryquen’dy
FO 194 | jo 10q ‘
SNS earn Ny
I SQ00000WOUNd—auoLtad
(6[-OI-4) A uaumiuadxA
VALUE OF PNEUMOCOCCUS VACCINATION 433
too exactly the minimal lethal dose, but it is practical to main-
tain this standard. The dosage of the test virulent culture,
therefore, was estimated in multiples of this dose.
Preparation of the vaccine. Meat infusion peptone broth was
the medium used with 0.5 per cent glucose. Tubes of dextrose
beef infusion broth were inoculated with 0.2 cc. of the most recent
semi-solid cultures (1) of the standard strains of pneumococcus,
types I (Neufeld), II and III and incubated at 37°C. for twelve hours
to be used as seed cultures; the semi-solid cultures having been plated,
fished, agglutinated and tested for purity. After incubation, flasks
of broth (300 cc. per flask) were inoculated with 1 cc. of the seed cul-
ture, which had been examined and agglutinated against the three
types and tested with bile, and incubated at 37°C. for thirteen hours.
After flasks had been examined for purity they were heated at 53°C.
for one-half hour. The heated culture was then centrifugalized for
one-half hour. The broth was poured off and a small amount of 0.85
per cent saline was added to the sediment. This suspension was trans-
ferred to a sterile bottle, to which a definite amount of saline was then
added and the mixture thoroughly shaken. One cubic centimeter
was removed for bacterial count and the remaining suspension was
again heated to 55°C. for one-half hour. The vaccine was standard-
ized by the Helber counting-chamber and 0.3 per cent tricresol in a 2
per cent solution was added. Cultural and animal tests were made for
sterility and the vaccine was diluted to the desired strength.
The results of the experiments to determine the protective
value of a single dose and a series of doses of the vaccine in mice
are listed in the preceding tables. The vaccine in these tests
was given subcutaneously; the test inoculation of virulent cul-
ture, intraperitoneally.
For purposes of comparison, tests of the protective value of
pneumococcus vaccination were started in rabbits but discon-
tinued when it was found that the protection, which is obtained
in these animals, did not differ essentially from that which was
obtained in mice, the protection being limited to the infection
with the homologous organism. ‘The results of these tests are
recorded in the following tables:
434 AUGUSTUS B. WADSWORTH
Experiment VII (8-26-19). Rabbits receiving six intravenous inoculations of 1 cc.
at six-day intervals of type I vaccine containing six billion organisms per cubic
centimeter; tested eight weeks after last inoculation with virulent broth cultures
of types I and II
CULTURE—PNEUMOCOCCUS I CULTURE—PNEUMOCOCCUS II
5 Number! umber] Number i seve Rurbes Nembex ae i ee
uan- : oO 0 ime animal] Quan- 5 oO er of |Time anima
tit rabbits | pabbits | rabbits died tity pate rabbits | rabbits died
Oe a | lived died ; ad lived | died
lated lated
ce. cc
0.1 2 2 0 4 mos. 0.1 2 1 il
0.1 Control—Normal 41 brs. 0.1 Control—Normal
rabbit, died rabbit, died
Experiment VIII (8-26-19). Rabbits receiving twelve intravenous (9, 1 cc.—3, 2
cc.) inoculations at three-day intervals of type I broth culture heated to 55°C.
for one-half hour and tested eight weeks after last inoculation with virulent
broth cultures of types I and II
0.1
2 | 2 | 0 | 4mos. | 0.1| 2 | 0 | 2 | 22-70 hrs.
Control as above
In these experiments, rabbits were selected weighing between 1400 and 1800
gms. The virulent cultures were given intravenously and the animals were
autopsied and pneumococci recovered from all that died.
CONCLUSIONS
After one dose of three billion pneumococci, type I, there was
only very slight protection against type I and none against
either types II or III. After three doses of type II vaccine
there was slight protection against type II and none against
types I and III. After six doses of type III vaccine there was
no protection against types I, IJ, or III. After three doses of
type I vaccine there was no protection against either type II or
III but definite protection against 0.0001 cc. or 100 times the
standard fatal dose of Type I which the control unvaccinated
mouse received. Finally, vaccination with six weekly doses of
six billion cocci, type I, or a total of thirty-six billion cocci,
protected against a virulent inoculation of 0.001 cc. but not
against 0.01 cc. of the virulent type I culture.
-—
Ta
VALUE OF PNEUMOCOCCUS VACCINATION 435
Definite protection was then obtained against the development
of the homologous types of infection when large doses of vaccine
were used; but the degree of protection that was obtained was
not great considering the quantities of vaccine that were used
to vaccinate the animals. The parasitism of such highly viru-
lent cultures is so great that the pneumococci develop in the
vaccinated animal just as they grow in a test tube of immune
serum. The virulence of pneumococci has been found in pre-
vious studies to be largely dependent upon its growth energy or
vegetative energy which in the animal tissues- constitutes
parasitism (2).
REFERENCES
(1) Wapswortg, A. B.: Proc. N. Y. Path. Soc., 1903, 3, 113.
(2) WapswortTs, A. B.; AND Kirksripe, M. B.: Jour. Exp. Med., 1918, 28, 791.
ee see oD
—— * = = _
SEROLOGICAL RELATIONSHIPS OF LIVER AND
KIDNEY
MOYER 8S. FLEISHER, T. G. HALL, anp NATALIE ARNSTEIN
From the Department of Pathology and Bacteriology of St. Louis University School
of Medicine
Received for publication July 2, 1920
The subject of immune bodies against various tissues has
occupied considerable space in the literature of immunity and
especially has emphasis been laid in the discussions on the ques-
tion of the “‘specificity” of the antibodies produced against
organs. It is particularly with this latter phase of the subject
that our experiments have dealt; the results we wish to report
here concern however only liver and kidney. We do not desire
to go into the literature (1) on this subject at the present time
and will only briefly review some of the work which has ap-
peared bearing most directly upon the antibodies produced
against renal or hepatic tissue.
Bierry (2) and his co-workers, using so-called (3) nucleo-pro-
teins of various organs (liver and kidney) for immunization,
state that following injection of these antisera into animals of
the same species as those whose organs had been used for im-
munization, definite and more marked changes occurred in the
organs corresponding to the antiserum injected. They believed
therefore that a more or less distinctly specific cytotoxic serum
had been developed. Beebe (4) using essentially the same tech-
nic reported even more definite results regarding cytotoxins.
Armand-Delille and Leenhardt (5) agreed in the main with
Bierry but qualified their statements by admitting changes in
organs other than those corresponding with the specific sera.
Pearce (6) and Pearce, Karsner and Eisenbrey (7), on the other
hand, failed to find a distinct selective action of the antisera on
the homologous organ. It is evident that the cytotoxic method
| 437
438 M. S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
of determining in vivo differentiations between various tissue
antisera has not given sharp or clean cut results in all cases and
the reports of various investigators are not conclusive in view of
the contradictory statements.
Two facts, probably of importance, seem to have been brought
out however by various of the above mentioned investigators.
Pearce and his co-workers state that the changes which are
produced in the organs as a result of the injection of the cyto-
toxic sera are in large part due to the agglutinins for the erythro-
cytes which are present in the various anti-tissue sera, and
Armand-Delille and Leenhardt arrived at the conclusion that the
antisera are composed of a complex of antibodies. The anti-
bodies prepared against the cells of a single organ are therefore
apparently complex; really a mixture of several antibodies; it
ought to be possible to separate these by absorption of the
immune ‘sera and to determine what is left after this absorption
by complement fixation tests.
Experiments have previously been carried out in which ab-
sorption has been used to determine the relationship of various
organs, but on the whole they have been inconclusive because
they have not been pushed far enough. Forssner (8) carried out
experiments in which he attempted to differentiate serologically
various organs by absorption and precipitins but, while he could
apparently show certain relationship, he was not able to demon-
strate marked differentiation. Michaelis and Fleischmann (9)
were able by absorption and complement fixation to show differ-
ences between anti-red-blood cell sera and anti-tissue sera in spite
of the fact that the serum prepared against the one or the other
cell contained antibodies reacting with the non-homologous cells.
Rados (10) was not able to determine differences between various
organs by the use of antisera in complement fixation tests alone.
More recently Kahn and McNeil (11) failed to find complement
fixation with antisera prepared against a few tissues. Our
experiments have been directed towards determining what differ-
ences could be shown between liver and kidney antisera by com-
‘plement fixation reactions with various antigens before and after
absorbing the various sera with cell suspensions of the various
organs.
:
bi
a, ee
SERIOLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 439
TECHNIC
We used throughout guinea-pig organs for immunization, and
rabbits for the production of the antisera. The rabbits were injected
four times with an interval of one day between each injection, and
were bled on the ninth to eleventh day after the last injection. In
most cases the animals were injected intravenously with a suspension
of the organ ground in salt solution, and then filtered through sterile
gauze. None of the animals so injected showed any immediate ill
effects as a result of the injection, but many of them showed a tendency
toward loss of weight. A few animals were injected intraperitoneally,
but we found that we lost more animals injected in this manner than
when injected intravenously. The animals injected intravenously
gave sera which were on the whole very slightly more active, but the
difference was not sufficiently marked to permit a statement that the
intravenous method was better. Practically either method of injec-
tion gave satisfactory results. The organs used in immunizing the
rabbits were removed from the guinea-pigs at once after killing them;
they were not washed but were immediately cut up and ground in a
porcelain mortar without attempting to get rid of the serum or of the
red blood cells. These two factors were ignored upon a priori con-
siderations and will be discussed later. The organ particles were
ground without addition of any solution; after they had been reduced
to a rather homogeneous semi-fluid mass, a little 0.85 per cent sodium
chloride was added and the grinding continued until a smooth paste
was obtained. Twenty cubic centimeters of the salt solution were
added to each two kidneys or the approximately equal volume of liver
tissue and from two to three cubic centimeters were injected into each
animal.
The antigens used in the complement fixation tests were prepared
in a number of different manners. At first we used organs which
had been perfused to remove the red blood cells and serum, and which
had become definitely blanched. These were ground in the same
manner as were the organs used for injection and the cell suspension
repeatedly washed in salt solution by centrifugation; this gave a very
satisfactory antigen with kidney but not with liver. The washed
kidney cell residue from two kidneys was diluted with ten cubic centi-
meters of salt solution, and to this 1 cc. of 5 per cent carbolic acid was
added as preservative. This served as stock antigen. Antigens pre-
pared with and without carbolic acid were tested and gave identical
440 M.S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
results. Antigens were prepared in a similar manner excepting that
the organs were not washed by perfusion. In a number of cases
immediately following the grinding the suspension was filtered through
gauze to remove any larger particles and connective tissue which had
not been broken up by the grinding. In some antigens we attempted
to get rid of the red blood cells by first washing the organ paste with
distilled water until the supernatant fluid was no longer discolored,
and finally making up the cell suspension with salt solution as above;
we discarded this method as the antigens were often not satisfactory.
We also used antigens which had been dried rapidly in vacuo after being
ground and washed as above. By all of these methods we were able to
obtain satisfactory kidney antigens.
In the greater part of our work and constantly in the later part, we
have used not the organ cells, but the organ extract as antigen. We
took the unwashed organs, ground them as above and added about 20
ec. of salt solution, centrifuged for twelve minutes at full speed—
approximately 3000 revolutions per minute—and used for the antigen
the supernatant fluid which was slightly cloudy or opalescent. By this
method we obtained excellent antigens with kidney and it was the
only method by which we were able to obtain a satisfactory liver
antigen. In preparing the liver antigen we used a quantity of liver
tissue approximately equal to the mass of the two kidneys. Occasion-
ally antigens were not satisfactory because they were either anticom-
plementary, or not antigenic; such antigens were not used. This
latter method of preparing the antigen had this advantage—that the
erythrocytes were not contained therein; the serum was of course
retained but as will be seen this was a factor of little or no importance.
In absorbing the serum with the various organs we used throughout
the solid residue of the organs as obtained by grinding and repeated
washings (at least four washings were done before the cell suspension
was used as absorbent). We extracted both diluted and undiluted
sera, but found it more satisfactory to work with the undiluted serum.
The serum and organ were left in contact for thirty minutes at 37°C.
and then were separated by centrifuging. In some cases the extraction
was repeated. We used, as a rule, a volume of tissue cells equivalent
to one-fourth the volume of the serum.
At times, and especially after the sera had been in contact with
either kidney or liver, they proved to be anticomplementary after
the absorption. This fact caused us considerable difficulty for some
time but we finally found a method for removing this anticomplemen-
— —
—— Aa. wi
ee eee
Ss
7
i
SS —-s C
. >
ul
;
SEROLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 44]
tary activity. After the serum had been separated from the organ
mass we added a considerable quantity of barium sulphate, about one-
fourth the volume of the serum, shook the mixture and then centri-
fuged, separated the serum and then inactivated. It seems probable
that the barium sulphate carried down fine particles of the absorbent
which had remained suspended in the serum and that the inactivation
played a negligible part in the restoration of the serum.1
The technic of the complement fixation reactions as carried out in
these experiments was relatively simple. As hemolytic system, we
used throughout sheep corpuscles, rabbit anti-sheep red blood cell
serum, and guinea-pig complement. The latter was used in three
unit quantities. The corpuscle suspension used was always a very
dilute one, approximately a 0.5 per cent suspension, as we desired our
end results to be either clearly positive or negative. Every day before
carrying out the actual series of reactions we had in mind, we tested
the anticomplementary and antigenic dose of the antigen; the antigen
was then used in a quantity well under the anticomplementary dose
(less than half of the anticomplementary dose) and in a quantity
representing three to four units of the antigen. Actually we usually
used the kidney antigen in a dose of 0.1 to 0.3 cc. of a one to ten dilu-
tion of the stock cell suspension or in 0.1 to 0.05 ec. of a one to five dilu-
tion of the extract antigen; the liver antigen was usually used in quan-
tities between 0.3 to 0.1 cc. of a one to five dilution of the extract.
Before we turn to the consideration of the results obtained,
certain preliminary considerations may well be taken up which
have an influence upon the interpretation of the more specific
results dealing with the relationships of kidney and liver antisera.
We examined the serum of a number of animals to determine
whether there exist normally in the blood of rabbits, antibodies
giving fixation with liver or kidney cells or extracts of such cells
of the guinea-pig. The sera of ten normal rabbits were examined;
only one serum gave fixation with kidney antigen. It is doubtful
whether this represents a true fixation as this serum which gave
1 This method of removing anticomplementary substances (?) from serum has
been tried out with a small number of anticomplementary sera sent in for the
Wassermann reaction and seems to have some practical value but as yet sufficient
data has not been collected to warrant any definite conclusion as to either the
usefulness or limitations of this method.
442 M.S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
fixation was freshly drawn and had been inactivated; it seems
probable that we were dealing with a nonspecific fixation such as
described by Kolmer and Trist (12) as being present after inac-
tivation of rabbit serum. It appears, therefore, that normal
rabbit serum contains no antibodies reacting in the complement
fixation reaction with guinea-pig cells. This fact, however, in
view of the one exception mentioned will have to be further
investigated.
We have carried out several experiments to determine whether
the immune sera would show complement fixation when inacti-
vated guinea-pig serum was used as antigen. Using several dif-
ferent antisera, both anti-kidney and anti-liver, and using the
guinea-pig serum in quantities of either 0.1 or 0.2 cc., we did
not obtain positive fixation even when 0.1 cc. of the immune sera
were used. It is evident therefore that the serum content of the
extract antigen used in our later experiments (which dose of
course contained the serum content of the organs) cannot be an
interfering factor in the reactions to be described. Furthermore,
the quantity of serum in the diluted antigen used in the com-
plement fixation reaction was even less than the largest quantities
which did not give fixation.
In the case of the guinea-pig erythrocytes the conditions are
not so simple. We examined the anti-organ sera for aggluti-
nins, hemolysis and complement fixing antibodies against guinea-
pig corpuscles and found that such antibodies were present in
both anti-liver and anti-kidney serum; using 0.25 cc. of a 5 per
cent suspension of guinea-pig corpuscles we found that with anti-
kidney serum 0.005 cc. produced agglutination, while with the
anti-liver serum 0.001 cc. gave agglutination. The hemolysins
were, however, not so active as were the agglutinins, as 0.05 cc.
of both sera gave only partial hemolysis of 0.25 cc. of a 5 per cent
suspension of guinea-pig corpuscles. It was evident from these
results that the presence of corpuscles in the antigen might well
be a factor in the complement fixation reaction. In table 1 we
give the results of complement fixation reactions in which various
suspensions of guinea-pig corpuscles were used as antigen. We
do not give here—and will not give in later tables—the details
Sur
SEROLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 443
of the reactions as these points have been covered in the descrip-
tion of the technic.
It will be noted that the anti-liver serum contained a greater
concentration of complement fixing antibodies reacting with
guinea-pig corpuscles than did the anti-kidney serum; this fact
had also been noted in connection with the agglutinins. As to
the possibility of this presence of anti-erythrocyte antibodies
influencing the reactions which we report here, we believe that
in the antigen used in our experiments, that is, in the quantity
used in our work, red blood cells were not present in a quantity
corresponding to 0.5 ce. of a 0.1 per cent suspension; in this quan-
TABLE 1
Complement fixation of anti-liver and anti-kidney serum with varying suspensions
of guinea-pig corpuscles
QUANTITIES OF
GUINEA-PIG
CORPUSCLE SUS-
1 PER CENT SUSPENSION 0.1 PER CENT SUSPENSION
PENSION 0.5 ce. | 0.25 ce. 0.5 ce. | 0.25 ce.
Anti-kidney serum
cc.
0.1 a 0 0 0
0.07 0 0 0 0
Anti-liver serum
0.1 +4+4++ +—-+++- ++++ 0
0.07 +4+4++ ++ wa oat Be 0
tity and even in the lower quantity, which did not give fixation
with either serum, the mixture of serum, complement and cell
suspension showed a definite reddish tinge and was distinctly
clouded by the corpuscles; on the other hand the antigens did
not show this tinge in the dilutions and quantities used. Fur-
thermore, the volume of the ground organ rarely exceeded 2 cc.;
in order to have a 0.1 per cent suspension of erythrocytes in the
final dilution of the antigen one-twentieth of this bulk would
have been composed of red blood cells which certainly was not
the case. Therefore, in those cases in which the cells of the
organs were used as antigen, it is very unlikely that antibodies
444 M. S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
reacting with red blood cells played any part in the reaction.
Of course in those later experiments in which we used the organ
extracts the corpuscles could evidently have had no influence.
CROSS FIXATION WITH KIDNEY AND LIVER ANTIGEN
Considering first the relation of the antisera to the two anti-
gens, we found that in all experiments the anti-kidney serum
gave complete fixation with its homologous antigen in smaller
quantities than when tested against liver antigen; in four cases
out of six the anti-liver serum also fixed in smaller quantities
TABLE 2
Cross fixation of sera and antigens
CUBIC CENTIMETERS OF SERUM
0.03 | 0.01 | 0.008 | 0.096 | 0.004 | 0.002
Kidney antigen
Anti-kidney serum......../++++/++4+4+/4++4++4+] ++ -- 0
Anti-liver serum.......... ++++/)++4+ } +++ — 0 0
Liver antigen
Anti-kidney serum........ we 0 0 0 0 0
Anti-liver serum.......... ++++)/++4+)/+4++4+]/++4++) ++ 0
with its homologous antigen than it did with kidney antigen, in
the two other cases with liver antiserum, fixation was obtained
with identical quantities of the serum with both antigens—in both
of these latter cases we were probably dealing with rather weak
liver antigens as shown by the poor fixation with anti-kidney
serum. When we compared two different antisera in fixation
reactions with a single antigen, we found a tendency towards
fixation in smaller quantities when antiserum and corresponding
antigen were used. This relation did not appear constantly for
in some cases the anti-liver serum showed with kidney antigen
stronger fixing powers than did the anti-kidney serum. This fact
will be referred to again later.
ee
SEROLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 445
From the above results it is evident that there exists a definite
even though not absolutely sharply defined tendency toward a
differentiation between these two organs as shown by cross
fixation.
ABSORPTION EXPERIMENTS WITH LIVER AND KIDNEY
We will consider first only the relationship existing between
similar sera and antigens absorbed with different organs.
TABLE 3
CUBIC CENTIMETERS OF SERUM
=
io)
=
2
0.05 | 0.04 | 0.03 | 0.02 0.01
0.008 | 0.006 | 0.008
o
a. Kidney antigen and anti-kidney serum
Wuapserbed.|4-4-4---|+-+++|++++/++4++|/4++4+4+| + 0 6775.0
Absorbed
with kidney |++++] ++ “ 0 0 0 0 0 10
Absorbed
Mineneen eS tee |+++| ++ 0 0 0 0 10
b. Liver antigen and anti-liver serum
Unabsorbed .|++++/++++4+|/++4+4+/++4+4]/4+4+4+4+|+++4++4+)]/++++4+/]/++4+4+4+] 0
Absorbed
with kidney |++++/++4-4+/4+4+4++4+]/44+44+/4+4+4++4+ 0 0 0 0
Absorbed
with liver. . 0 0 0 | 0 0 0 0 0 0
From table 3, which is characteristic of the results obtained in
a considerable number of experiments, it is evident when we
had the combination of absorption of an immune serum by the
organ against which it had been prepared, and when we tested
the complement fixing power with a homologous antigen, that we
found definitely more of the antibodies had been removed than
when a non-homologous organ had been used for the absorption.
This statement applies equally well in the case of liver as it does
in the case of kidney. We find that a certain amount of anti-
bodies was removed by the non-homologous tissue but that from
446 M. S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
three to five times as much was removed by the homologous
tissue. There was also evident a tendency for liver tissue to
absorb relatively more than kidney tissue. The relationship
as revealed when we tested the various sera against the non-
homologous antigens is shown in table 4.
We found here when we tested the fixation of the antisera after
absorption with an antigen which was not homologous with the
serum, but which was homologous with one of the absorbents,
that in the case of the anti-liver serum the evidence of the def-
inite relationship of the antiserum to the organ against which it
had been prepared and by which it had been extracted, still was
TABLE 4
CUBIC CENTIMETERS OF SERUM
O),o;s alia
Ne) Sal oO a ~/l[olsios/esase
S S S S a a ad I de ec)
Si So Ss — So | Smloulon&)
a. Liver antigen and anti-kidney serum
Absorbed with liver...
ob
Absorbed with liver...
teat 0 0
evident; but that in the case of the anti-kidney serum this re-
lationship did not appear, and quite as much of the antibodies
fixing with liver antigen had been removed from the anti-kidney
serum by liver as by kidney. However, when we carried out
simultaneous experiments in which we tested the various ab-
sorbed antisera of liver or kidney against the two different anti-
gens, we found that relatively less of the antibody content had
been removed by the non-homologous absorbent when tested
with the antigen corresponding to the serum, than when tested
with the non-corresponding antigen, that is, in this latter case
the antigen corresponding to the absorbent. We found, there-
fore, that there exists not only a relationship between the anti-
SEROLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 447
serum and its corresponding antigen which could be shown by
absorption but also a relationship between absorbent and cor-
responding antigen when tested with non-corresponding serum.
It therefore becomes evident that the relationship existing be-
tween antisera against various organs is not simple but that
probably a considerable number of factors must be taken into
consideration.
ABSORPTION EXPERIMENTS WITH OTHER ORGANS
We next tested the fixation of the two antisera against their
homologous antigens after absorption with organs other than
liver or kidney. We first tried absorption with washed guinea-
pig red blood cells, and found that by this means relatively small
quantities of the antibodies were removed. Spleen was also
‘tested and removed very little, and finally brain was used as
absorbent and was found to remove possibly a little more than
either red blood cells or spleen. Probably not as much was
absorbed from the kidney serum by brain as by liver; very nearly
as much was absorbed from the anti-liver serum by brain as by
kidney; there was however a tendency for the kidney to absorb
slightly more than the brain in some experiments, and in no case
did the brain absorb more from the liver serum than did the
kidney.
We carried out a series of experiments in order to determine
whether the differences in absorption might be influenced by
variations in the volume of the absorbent material. When we
used similar volumes of the absorbent we noted that the absorb-
ent corresponding to the antiserum showed, as stated, distinctly
greater action on the serum and even when we used twice as
much of the non-homologous absorbent the homologous absorb-
ent still had removed more reacting substances. In the case of
the comparison of the non-homologous absorbents, however, the
quantities of the absorbent did apparently play a part, and if
the quantities of the two non-homologous absorbents were not
nearly the same the above mentioned relationships were confused.
This fact had no influence on the results given above as they were
THE JOURNAL OF IMMUNOLOGY, VOL. Vv, NO. 5
448 M. S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
carried out with similar quantities of the various organs used for
absorption. This marked influence of the variations of the quan-
tities of the non-homologous absorbents suggests that there was
here a purely physical absorption occurring in which chemical
relationship played little or no part and we tested two inert
absorbent substances not of animal origin, namely, kaolin and
barium sulphate, and found that neither of these two substances
removed any of the antibodies. While, therefore, as a result of
these quantitative experiments, the influence of the use of non-
homologous tissues as absorbents and their inter-relation is not
yet clear, nevertheless the definite relationship between homolo-
gous absorbent and anti-serum is clearly demonstrated.
FIXATION WITH ANTIGENS OTHER THAN LIVER OR KIDNEY
We tested the anti-kidney and anti-liver sera against two other
organ antigens, namely, spleen and brain, and found that both
antisera gave fixation with these antigens. In general anti-liver
serum gave better fixation than did anti-kidney serum with
these non-homologous antigens. When we tested the various
absorbed sera against these two antigens we obtained results
which were in the main confirmatory of those reported above;
however, there was a less marked regularity of results in these
experiments than in those above. On the whole we can say that
the definite relationship between the antiserum and the homol-
ogous absorbent was usually evident, in that these sera (serum
absorbed with its homologous tissue) showed less fixing power
than the other absorbed sera. However, in a few cases the rela-
tionship between the absorbent and the corresponding antigen
became more evident than the relationship between antiserum
and absorbent; in such cases the immune serum absorbed with
the tissues corresponding to the antigen gave least fixation.
Usually there was very little difference between the results on
the one hand with serum and corresponding absorbent, and on
the other with absorbent and corresponding antigen, when the
same serum and antigen was used and only the absorbent varied.
In practically all cases we could note evidence of a definite rela-
SEROLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 449
tionship between the absorbent and corresponding antigen, which
showed itself in the fact that the serum which had been absorbed
with the tissue corresponding to the antigen, showed less fixation
than did the same serum with the same antigen after being ab-
sorbed with other tissues—excepting of course the tissue corre-
sponding to the anti-serum. One exception to this rule did occur
at times, and this was in the case of anti-kidney serum absorbed
with liver tissue; here liver tissue apparently possessed a very
strong absorbent power. We cannot at the present time offer
an explanation for this result but it seems to correspond with the
more active part the liver seems to take throughout these experi-
ments in the production of antibodies.
DISCUSSION
Summarizing now the facts that have been brought out in
these experiments we found that anti-liver and anti-kidney sera
gave positive results in complement fixation reactions, not only
with the corresponding antigens, but also with antigens prepared
from other organs of the same species; however, in cross fixation
experiments there was a tendency for definite relationship be-
tween antiserum and corresponding antigen to become apparent.
Furthermore, it appeared that anti-liver serum tended to give,
in general, complement fixation in smaller quantities than did
the anti-kidney serum. When, however, we tested the comple-_
ment fixation of the antisera after absorption by various tissues
we found marked evidence of a relationship between the anti-
serum and its homologous absorbent especially when tested with
the homologous antigen, which latter relationship was constant
and definite. When we tested the various absorbed antisera
against non-homologous antigens we still noted a distinct ten-
dency for the appearance of this relationship between antiserum
and homologous absorbent, but in some cases this relationship
was not so clear, due to the appearance of a relationship between
the antigen and corresponding absorbent.
We can put aside the possibility that these reactions are due
to the presence of either precipitins or fixing bodies reacting
450 M. S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
with the guinea-pig serum. That the antibodies reacting with
red blood cells play any part is unlikely from what we have noted
regarding the quantitative relationships of the antibodies react-
ing with the guinea-pig red blood cells, and especially since we
have found that after absorption by the various tissues the anti-
- liver and anti-kidney sera no longer contain agglutinins for the
guinea-pig red blood cells. We were therefore apparently deal-
ing in these experiments with reactions to the tissue antigens,
and the erythrocytes as antigens influenced the results little, if at
all.
We can therefore first state that there exist, in the antisera
prepared against liver and kidney, antibodies, which have a rather
wide range of activity, at least as far as other organs of the
guinea-pig are concerned, giving fixation not only with liver
and kidney but also with spleen and brain. It seems, therefore,
that these two organs contain antibody-producing substances
which have common relationships with numerous other organs of
the body and possible with all tissues of the guinea-pig. These
would correspond with so-called species specific antibodies.
As a result of the complement fixation experiments with ab-
sorbed antiserum and the homologous antigens, the definite re-
lationship between antiserum and absorbent is evident. It
certainly appears that each organ contains quantities of anti-
genic and antibody-producing substances which are more char-
acteristic of that organ than of other organs; that the antibody
producing effect of the organ is most active in connection with
these anti-characteristic substances, and that because of the
greater content in the antiserum of bodies reacting with these
substances and the greater content of the tissue in these charac-
teristic substances, absorption of antibodies is greater when we
mix the corresponding antiserum and absorbent. ‘This relation-
ship of the antisera and organs would correspond to the so-called
organ specificity or would signify, at least, a quantitative organ
specificity.
When, however, we study further the experiments in which
we used non-homologous absorbents and carried out comple-
ment fixation reactions with non-homologous antigens, we find
SEROLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 4051
a tendency for a definite relationship to appear between the non-
homologous absorbent and its corresponding antigen. This
tendency is fairly regularly apparent and it seems that we can
interpret it best by assuming that in each organ there exist cer-
tain substances which have a relationship to definite organs
other than the one actually used as the immunizing agent. It
is hardly possible that this relationship is identical with the first
relationship which we have spoken of above, that is, the relation-
ship which apparently is common to all organs of a species, in
view of the very definite limitation of the relationship which
appears in these experiments. However, at the present time, we
offer this only as a possible explanation as we feel that this work
must be extended to include a consideration of the interrelation
of a number of other organs.
It is however evident, as would be expected, that the immune
sera prepared against tissues such as liver and kidney are complex
in their nature, that they are composed of a number of different
antibodies, probably varying in their relationship, some rather
limited in their range of action and others with a wider range of
action. It would seem that the liver contains more of these last
type than does the kidney. Three facts seem to bear out this
conclusion regarding the liver: We find a tendency for liver to
absorb relatively more antibody from anti-kidney serum than
does kidney from anti-liver serum; the liver serum reacts better
with the non-homologous antigens than does anti-kidney serum;
and finally the liver used as an absorbent in the experiments
with non-homologous antigen tends at least in some cases to
approach the absorbing action of the tissue corresponding to
the antigen used.
This suggests the question of the relationship of various organs
to each other; a matter which has been considered by several
previous investigators (Forssner (8), Cesaris Dehmel and Scotti
(13), Fiessinger (14) and Fleischmann and Davidsohn (15)).
We do not think it advisable to go into this matter at the present
time, but desire to push our work further considering various
other organs before we venture upon a discussion of this matter.
452 M. S. FLEISHER, T. G. HALL AND NATALIE ARNSTEIN
We have throughout avoided the use of the term “organ spe-
cific,’ and have done so intentionally, There has come to be
considerable confusion regarding the use of the term specificity
in connection with serological reactions, and we do not care at
this time to enter into a discussion of the use of this word. We
are satisfied with the statement that we have been able to dem-
onstrate a definite relationship between anti-liver serum and its
homologous antigen and between anti-kidney serum and its hom-
ologous antigen. Whether similar definite relationships can be
shown with other organs is now under investigation.
CONCLUSIONS
By means of complement fixation reactions and absorption of
sera prepared against guinea-pig liver and kidney we have been
able to show that there exists a definite relationship between the
anti-organ sera and the homologous antigens.
The antigens and antisera are not simple but are complex in
nature and probably are composed of several different partial
antigens and immune bodies.
Possibly these partial antigens and antibodies can be arranged
in three groups: The first having a very wide range of activity
and having a relationship to all or practically all tissues of the
species; the second having a limited range of activity and having
relationship only with the tissue used in the preparation of the
antiserum; and the third being possibly a group of antibodies,
also rather limited in their range of activity but reacting only
or more strongly with individual tissues other than the one used
as the immunizing substance.
REFERENCES
(1) References to the literature of cytotoxins will be found in the following
articles: Sacus: Biochem. Cent., 1903, 1, 573; RGsste: lLubarsch-
Ostertag. Ergb. d. Path., 1909, 13-2, 124; SoppERNHEIM, KREHL AND
MarcHanp: Handb. d. Allg. Path., 1908, 1, 5385; LANDSTEINER:
Oppenheim’s Handb. d. Bioch., 1910, 2, pt. 1, 542.
(2) Birrry: Compt. Rend. Soc. Biol., 1903, 55,476; Brerry AND Pertit: idem.,
1904, 56, 238; Brrrry AND Mayer: idem., 1904, 56, 1016; Birrry,
PETTIT AND SCHAEFFER: idem., 1907, 68, 496 and 566.
SEROLOGICAL RELATIONSHIPS OF LIVER AND KIDNEY 453
(8) Weuts: Zeitsch. f. Imm., Orig., 1913, 19, 599.
(4) Beese: Journ. Exper. Med., 1905, 7, 733; Brit. Medical Journ., 1906, 2,
1786; Science, 1908, 27, 648.
(5) ARMAND-DELILLE AND LEENHARDT: Compt. Rend. Soc. Biol., 1907, 62, 31.
(6) Pearce anp Jackson: Journ. Inf. Dis., 1906, 3, 742; PeEARcE AND SAWYER:
Journ. Med. Res., 1908, 19, 269.
(7) Pearce, KarsNER AND EISENBREY: Journ. Exper. Med., 1911, 14, 44.
(8) ForssNerR: Muench. med. Woch., 1905, 52, 892.
(9) MicHarLis AND FLEISCHMANN: Zeit. f. Klin. Med., 1906, 58, 463.
(10) Rapos: Zeit. f. Imm., 1913, 19, 579.
(11) Kaun anp McNeiu: Journ. Imm., 1918, 3, 277.
(12) Kotmer anv Trist: Journ. Inf. Dis., 1916, 18, 64.
(13) Cresaris-DEHMEL AND Scorti: Arch. p. le Scienze Mediche, 1907, 31, 135.
(14) Fresstincer: Compt. Rend. Soc. Biol., 1907, 62, 671; idem., 1907, 63, 573.
(15) FLEISCHMANN AND Davinson: Fol. Seriolog., 1908, 1, 173.
ON THE PLACENTAL TRANSMISSION OF SO-CALLED
NORMAL ANTIBODIES
III. ANTILYSINS
G. C. REYMAN
From the State Serum-Institute, Copenhagen, Denmark, Director, Th. Madsen
Received for publication August 22, 1920
In completion of the present series of studies on the placental
transmission of so-called normal antibodies the following obser-
vations as regards antilysins are reported.
A. ANTI-MEGATHERIOLYSIN
The examination for the presence of these bodies in the blood
of the mother animal and the kid, as well as in the milk, was
carried out in accordance with the quantitative method employed
by the Serum-Institute for measuring antihemolytic bodies. It
is to be observed that in the examination of milk only approxi-
mate accuracy can be obtained owing to its opacity With the
purpose of overcoming this obstacle the casein in one experiment
was precipitated with rennet, but this procedure was abandoned
as it was found that the antihemolytic bodies were also almost
entirely carried away.
The samples were inactivated for one-half an hour at 56°C.,
after it had been proved by experiments that the megatherio- |
lysin-neutralizing power was practically unweakened by this
process. The results of the entire series of experiments are given
in the appended table, whereas only a single series is diagrammati-
cally represented in the following curve (chart 1).
It appears that the titer of the blood of the mother animal
follows an average constant; this also applies to the blood of
kids, where the titer in all cases was found to be lower than in
that, of the mother animal, and in the case of twins it was very
455
456 G. C. REYMAN
nearly the same. ‘The directly determined figures remained very
nearly constant during the period of the research, whereas the
weight-correlated figures exhibited a fairly gradual rise.
The titer determinations for the milk were very low.
ANTI-MEGATHERIOLYSIN
209: Units per cc. o-——_—_—-. Mother's serum
© Kid 7 directly found values
O@-——-— @ " " weight correlated "
o—--—--—> 6") 6B hdirectly found u
esceaeaens ---" " weight correlated "
CEKEKREREEEKE KA GY
CuHartT 1
B. ANTI-VIBRIOLYSIN
In a work, of which unfortunately only a brief summary has
been available, Schenk (1) maintains that the anti-vibriolysin
content is the same in the blood of the mother as in that of the
child; for which reason he concludes that these bodies are easily
diffusible. This finding is corroborated in the case of goats and
kids by the researches detailed below. The researches in all
comprised five kids and their mothers. In testing for antilysin
in the case of two of the kids (11 and 22) and the corresponding
mother animal, horse blood as well as goat blood was used ‘as an
indicator; in the case of the others only horse blood was employed.
The measuring technic and the method of inactivation was the
same as in the tests for megatherio-lysin.
3
g
f
=
-
ieee Fe oe Pee.
Pn.
PLACENTAL TRANSMISSION OF ANTIBODIES 457
The results, as appearing from the table, are that the titer
of the blood is very low and nearly the same in mother and kid;
further, that in the course of the period of the research it fluctu-
ates somewhat, but without any particularly marked tendency
in either direction except that the weight-correlated titers show
in some cases a definite rise.
In the two cases in which samples of milk were investigated,
the amount of antilysin was rather great at the parturition, then
it decreased rapidly but later on it increased again in some degree.
C. ANTI-STAPHYLOLYSIN
As in the case of the normal anti-vibriolysin, Schenk found
that the uterus blood and the funicular blood contained the same
quantities of anti-staphylolysin; and further that these normal
antibodies are found in woman’s milk as well as in goat’s and
cow’s milk. Polano, however, found that, although the blood
of both mother and child contained anti-staphylolysin, it was
found in the former in larger quantities than the latter.
The examination of the normal anti-staphylolysin was under-
taken in the same manner as for the above mentioned anti-
hemolysins, namely, goat blood was used as indicator and the
samples were inactivated for one-half hour at 56°C. These
examinations of the serum samples showed that the maternal
sera at the time of parturition contained larger quantities of
normal anti-staphylolysin than did the sera of the kids; in this
respect the serum of kid 31 forms no definite exception in view
of the fact that the first blood sample was taken twelve hours
after birth. The titer of the maternal sera fluctuated rather
strongly, but with no definite tendency in either direction.
As regards the titer for the kids, it increased in most cases
during the first days after birth, but soon afterwards it decreased,
after which it remained more or less constant, although in several
cases one and a half months after birth it showed a secondary
increase; this latter increase possibly bears some relation to the
increase in titer of the milk, which frequently occurred at a
somewhat earlier period. This relationship, however, could not
458 G. C. REYMAN
be established with certainty, partly because the kids at that
period take other nourishment, and partly because they do not
always confine themselves to suckling their own mother. This
antilytie action of the milk can be reduced by skimming it and
so may partially proceed from the cream.
These experiments indicate that the normal antistaphylolysin
is not always, as is claimed by Schenk, transmitted quantitatively
to the young. My results, accordingly agree with those of
Polano (2).
D. ANTI-SAPONIN
As is well known, it was demonstrated by Ranson (3) in 1901
that the constituent in the serum which neutralizes saponin and
hence inhibits its hemolytic action is cholesterin; and further it
was proved by Madsen and Noguchi (4) in 1904 that this binding
admits of a quantitative measurement. Accordingly, as in this
process “antigen” as well as “antibody” are known, it is of
interest through the medium of saponin binding to examine the
occurrence of cholesterin in the sera of goats and their new-born
kids, as well as its presence in the maternal milk. As far as I
know, such examinations have not been undertaken for the life-
period dealt with in the present article. For a description of the
methods used in titrating the samples reference is made to the
process described by Madsen and Noguchi.
It appears from the appended table (I, d) that the titer of the
mother animal remains nearly constant, before as well as after
parturition, whereas that of the young increases directly after
birth until it reaches a higher point than that of the mother,
and then, as a rule, rapidly decreases; only the titer of the kid
15 and to a certain extent kid 13 (1914) remained nearly constant;
later in one or two cases an increase was again observed.
If one looks at the weight correlated titers, it will appear that
after the primary increase, which, as a rule, is followed by a
decrease, the titer on an average increases gradually, so that the
total content in the blood of the saponin-neutralizing bodies
is upon the whole increasing. Only in kid 15, however, is the
increase nearly an uninterrupted and gradual one.
*rNOUIIQWOS OLqno sod Ayn USA[QUY Ajrass syNsor OAYWQUUND oy, “sopdurvs Jo Furyuy or} 4B qUFIOM OY} PUB YIIIG 7B IYFIOM OY} TOOMJO O1VBI OY} YUOsOIdOI BOND OBO], }
i
02 |0% |8t joe |IT 9
TP OAL a Fh a ay q
G'O>/9°0 |¢°0 |S°0 |O'T |O'T JOS +=j0'% aI jet |tr lzt |or ng *P (49 |L49 |*9 #9 Po CSE (S161)
£/22|8/61|8/018/9 |8/% |%/L2|%/8% |%/2ee €/01|8/9 |8/% |0/L2|6/82|0/2e £/22|/S1\8/01/8/9 |8/2 |2/Le |2/eee 9
I8 |€6 |#9 |F2 |99 |8¢ Ize joe 92 |s2 a)
Q°% |9's (E'S |I @ (O'S IST [9'T (9°T jer ITT q
62 ze |82 |Se |s2 |ze |oz |% jos jez |€% BOL
601 |66 |06 |I8 0g CF ILZ (18 0
o's |2's |6'2 |9'2 ats OT HT [OT q
OVO OMIsONO | VOeNiSsh WTR ITS |T8) 18 £2 9% |6r {6L |8% v6 #6 |66 |LIT |Z8 |ZIT |SITE SIT |80T |OOT |88 joOr |OOT |F6
€/S8|8/L1/8/FI/@/2 18/8 8/92 |%/F%e | G/08|8/SG|8/12|8/LT}8/FT|8/1T]8/2L |&/8 |%/8212/92|0/Fo~ £/08)£/92\8/1Z|S/L1|8/P1 18/11 \8/L |8/& |%/82|%/9G|Z/Foe |2/2I 12/9 | OL ¥6
96 |29 zo |9F |ze |e ) P
9°% (8° ST OT IPT ler q
1g |62 62 162 |&% IST |62 vg
gg ige ir |2e joe ze Ist Zt a)
2S |0'S IST 2Z'T |9'L [PT IST ITT q
0 |0 Oesoeigeio0 | 0: ler \t> Gz |6r |e |6r len |€% FT |et |e B 18 |PIT 22 |16 |22 28 |T2 |€9 |£9 19) £9
G/12| 8/97 |G/1T\8/2 |8/& |Z/8z]Z/Sz|%/ES |2/TSe £/1Z|8/9T|G/1TI8/2 |/8 |Z/82|%/S2|/83|0/Tem £/08|8/9T|8/TT8/L |8/8 |Z/82\2/Sc|0/8s|6/ISe 12/21 12/9 8XL %
0g BG
eo joF |9F |9F Ios lee o
Pen Set Osta Pabel Gaal iar q
10/0 |0 {2°0 0 ez «\e@ «(62 |e8 j62 108 cit) BET ISEE |SET |LET |OTT \OZT |SOT |OOT j6Et (F161)
£/91|E/L |Z/82|Z/8S |Z/1Se £/91\8/L 18/8 |Z/82|0/S210/86|2/1S« SLOTIG/TTIE/L |€/& |%/83|2/Ss|0/8% |2/Te» |2/2t 9xg
16 |ZIT |92 |88 |#6 |OOT th o)
#3 |€'°% 10'S |8'T 19 T eT Ler q
0/0 |0 {go> ge |op ise jer |69 |cL |€9 |OP | a ZLT |Z9T |T6T |ZLT |60% |SLT |Zzt |PEt |ZLT
€/1T|8/6 |Z/T2|2/91 |/PTs G/TT8/6 |8/F |Z/83|/0/82|/18|0/81 12/91 |G/F Te €/1118/6 |8/F 10/82\Z/S2|/12|Z/OT |2/PT* |2/20 i
MITIW WoAUds Sain WoUuds SUaHLonw Fook ee
Lc aE ee ote ee S| ee ee ee Ee ee ce ee eee ee ee ee
UISA[OLIOYJVSOUI-1IYUW “VW
qYSIOM 07 poyepars0d 10914 ‘0 fLrO0joRy ysiom ‘q {A[Qoourp punoj s9j1y, “@ “SABpPYWIgdy,
sajdwmps wnias Buryn) fo sajoq
plas Spry pup pprsajou ust surshyuyun fo u0ro.4ey
T ATAVL
I>11> Gt} oT 0 | O $080 ET 22 88 OTIS OOTSOET ET ET ET OT | ® 0°% |9'T \6°0 [60 (0 T OT (O'T I'L [eT let
/228/61) &/9T\ &/9 |Z/Le | /82)G/S1\2/E1/%/S1'/T12/0l« £/01 8/9 re 6/L\6/80, 6/61 G/ST S/Z1'Z/0l« €/S1.8/01 8/9 \8/% |2/L2,2/82)2/61'2/S1 \2/Z1\2/0ls| 68
¢9 9¢'€9 (96 |b FSF 0S €S 9F OE 6S tate B)
| 9 AOq |B 8V q
G'0>9'0>¢'0>|9'0>€ 1 08 4°23 6S O'S OT LT LT IL T T'S ihe (2 1 ILS (9° 8's (62 |b O'S O'S IVT | BOP BITS LT 8'T (OT IPT (oT (8'8 IE's (0's 13's 10'S (10'S (6S (3'2 IOS IL Tul
¥/1Z | $/IT| F/T | €/8%8/Sc8/01/8/2 8/8 &/8« asioy) JO |poopq :107vo|Ipuy 8\s10q) JO |pooyq :1ozwotpuy| 7
BF = 6 OLL IL9'129'8 BLOsS 6909 SF FF 5)
ge fe 08 8% 9% e@°3 (0'S (OL OL (OT ie Lot iT q eee
ie 68 €¢ 168 22 88 68 |0'F las \@'s |2°e L's Bigs 9 0's O'S F's |b [8's [O'S [9°9 FS F's Fs lbs [L's L26'¢ |9°8 les
G/STS/T |$/St! /1CF/S1 F/TLF/8 |B/T |6/808/SZ8/L1/8/F1,8/018/L 8/S |8/es G/ST.G/1|F/SS/1SF/STP/TLP/8 IP/P I6/T \8/82,8/S8/18/L18/F1\8/0le,2 \8/E |8/e«\o/re
4,803 jo poolq 710/} BOL pul 4yeos jo poo [q ool CPC Tpuy GG
Cee 9% \F'S Io 6S 1% 9'T 9'T B)
9 AOQ|8 sy q
UL OLIUT INT OF 0°S OT (9 T 8 L et lor | G'S \L'S |G°S O'S (3's 19'S |S (G's (9's (G's |o-s \e'S Ios
a)s10y] JO |pooyq :1078o\1puy ajsi0oy) JO |pooyq :Loywoipuy} IT
gL 2269 0S QF |3'F 6 T 8 F a)
6% 1310'S 10'S GL PL I3'T o'r q
4% 12°% 0S 2°3 |9°% 10'S 10'S 0S 9 T OF O'S | @ | | L220 Of OF OF OF O'S 220 F O'S OF Le
€/08 8/9G'8/168/LT.6/FTE/T1E/2L |8/F 8/2 |S/LE1C/Sex £/0€ 8/90 8/12 €/LEE/FLE/TLE/L &/b €/% 2/L20/S2% 2/21 2/%
} Boa jo poo, :10}BO) tpuy 4B0 3 jo! poo|q id oyeo Ipuy Il
£3 (8% 0'% 9'2 (A 1 8 2 )
GURL ST et (LT 90 TOT [OT OT q |
GT OC STO IST 2s 27s 0-3 Os O's | 8080 0T (80/80) OT OT ITT ET
| |@/8)%/S0.0/12,0/61 2/91 2/F1/ZTC/TL\2/01|2/ 6% 8/86.2/162/61,2/917/. iG/@1)2/0L |6/6x'0/9 | €
91 PT 10% A) | |
OL IPT let | q |
0'T |O'T |S‘ |F I OT (6T O'S Os | IST | Gailen| Fala Cael ta cal OnGue|canalses
@/81/Z/9T\S/F1IZ/ T1/2/6 |Z/L \0/9 |C/Se | @/81/Z/9T0/F1|Z/11\c/6 \2/L |2/9 S/T» | T
aii
HIIN wouds sala WOUaAS SUAHLOW ean
-WON
UISA[OLIQIA-IJUY “G
panunuwoj—l AIAVL
460
01% 2°81 9°IT oLt a
83 S 0% tT q
L£°6 |2°9 [2°6 [Z°F JOT O'T [60 |I> |F FIO FING IT 6F 92 (I'L |9'L [SL 18'S [2 ITS StF ETI9 FT \z> | B £6 |6°9 \€ Z 69 [82 189 [88 |F OLlP'L |h'G | (ST6T)
£/08/8/92/8/61/8/91/8/ET\/01|8/9 |8/% |L/2|%/92|2/Fe €/83/8/22|8/61|8/91|8/E1|€/01/8/9 |8/% |%/L2|%/9% |Z/tee £/6218/82|8/61\8/91\8/E1|8/01\8/9 |8/% |Z/Lo |2/1ee| 2
FSF bak 1 &% #93 SLT a
6% G3 Ore 61 rT q
. O'L JL°91/0°SZ|2°6 |S FT/S OL16'6 |6'ET]L°ST\L st\6 sl jor | 8
I el
a 01 3F Cae 9° oF 5)
0's G3 [ete 6T i q
FSL [CEh |8Lz [299 |S2z |F6Z [SEF |ScI |0' OFIO'S G'SIIF'S [FT |6'T |9°S |S'T IST |6'T 6'L 10'S |$°c2 |O B (88 |O'T |O'F |8°S |0'F OT |b S |O'F OF OT [OL jOT {(St6r)
¥/& |8/08|8/22|8/61/8/91/8/81|8/01\8/9 8/3 |2/Ses ¥/8 |8/08|8/92|8/20|8/61 8/91 18/S1 18/01 \8/9 |8/% |Z/Le |Z/S%a| SE |t/E |8/08|8/Sc\8/ea|8/61|8/9T\8/ET\8/01 18/9 |8/% |Z/L% |%/Soe E1221
g°¢ 6S |b'9 |6F 5)
8% 13 |6'T PT q
80 $0 |9°S |t> |2°t |I> |I> |t> |t> |e'r je¢ 0'% 89 {F'9 [8's |G [G's V9 | Bge
Hi[ta | pal uuar| ry
¥S 0's 5) =
9'T Tie q F
€& |IT |# {Zt |22 |0'F |$°¢ |O'% |0°% jO's |0'8 Jo's |G°¢ Sat ian oF BLE 88 {Ot |6°8 \ar |r'9 let {et jor | (rt6r)
Y) [Par | peta) uiey|su
G/STIS/L |S/% |¥/82|F/82|F/81|F/ST|F/8 |F/F IF/T |8/83|8/9G/8/Se* G/L |S/Z \t/82|/81|F/8 |F/T (8/9 |8/Se« G/ST\S/L |¥/82|¥/8 |¥/b |B/T \8/L0 |8/SS+|88*LE
O'IE/9°2 |#2°9/8'°9 |Z°S |Z STIO'F IP's je°9 IF F [89 b)
LeGuSie 198e Fe ITao|MacSel (Zeb 9 b ie biter q
€8 [ZEAL 9/2T'9 O'OTIL'S IFS [27S [2's |S'9 2's |2'e IF IFS [29 OL 8 |P'T |L°% |P9 |9°S |9°9 |9°T J4°S |I> 10° |0°S |8°0 (F161)
£/82|8/92)8/E1|8/0t ¥/02|¥/ST\F/11|F/8 |#/F |F/1 |8/82/8/F3|8/02|8/91 E/T |8/01+|8/6 ¥/08\F/11|F/8 |F/F |F/1 |$/82|8/FZ\8/02|8/91|8/EI 8/0l«,8/6 1g
O'9T/S OTIS ET/Z°2 |b S |Z VtlL'9 19°62 5)
8°S G'S [8S |2°s |0'S |8'E I9'T [eT q
0g |0s -WON
bee ursAjo[Aydeys-1yUuy “oO
——— ee a Pa aT eo em my] SS] TERE ee nS
Q a10yMAIOAT
€/08-ZIS
Q Soyep Suruievulol oy} FV
co>e/6l ¢0 &/I
£/20-Z/16
I>
soyep Sutureulol oy} FV
1>¢/6
$/9-8/1
Q soyep Durureulos oy} FV
OT PL
G0 &/&
¥/91-C/L1Z
L'86
ad
18
/91
@ OT
G° GS
GS
06
£16
8%
92
t/6
£/08|8/82
€ TFT 6
GS 8s
G°8T/0 LT
4/9 F/T
e It
Gs
oF \s>
0 02/8 61
SG |6%
£8 |0°6
£/86 8/86
0-6
G3
00
Gar
GZ
0'¢
€/2%
PLE
oS
O°LT
€/8z
6 GT
1G
8g
SSE
06
8g
£/1%
MIIN
£8
Sant
£/61
g¢
£/22
0°92
0%
0&1
£/8
OIL
61
8g
OTT
61
8-¢
8/21
G8
LT
o¢
06
ST
0s
&/F1
OL
61
0S \O'F \L°%
bP
61
GP l€ Ss |L°%
€/€1|8/01|8/9
6¢ Gg
8 T PT
€€ |L°s 19s
€/01\€/9 |&/%
9°GZIT FS
OT |e T
0 9T\S 810° 6T
€/LT\€/F1 18/2
89 |S'F
LY |S t
OF |0'€ \S>
89 |0°9 |T'8
AACN Zed!
OF \0'F% |8°S
€/1T|8/L |8/S
92
rT
rg
£6
Sal
9
6/%
sis
rT
eke
6/12
GG
(aa
o°8T
&/F
col
co) CONN OH SS
NSA OE SH ©
Ge) =
o
q
ocrlo'# | eer
o
q
1110 a oe |o'F lor |z'e love o's |z'e |2'¢ lo'e
G/LE|0/SEx #/& |\¢/0g|¢/sa|¢/ze|€/61|8/91|¢/e1\¢/0r|e/9
a°9 a
VT q
0g e 9% (G's [9's [EZ
G/82|C/TSe €/62|8/S1|8/01|¢/9
ro)
q
¢stio it | 8 SIG IP ICE |S 1tl9 orlt's Iz [F's
£/z |/Ts 6/9 18/1 |€/82|8/Sa|8/12|8/21/8/F1|8/11|8/2
og a)
ii q
og a
79 a)
Tea q
g'¢ |g wzt je’ |9'e@ |2'e |9'e |9's |e's |9'e Jo's [e's |9'e [8°s |o°z
G/8C|G/Les $/9UIP/TLP/L |b/@ |€/8cle/Te|¢/21|¢/F1\e/11\e/2 |8/¢ 8/8
9% |9°% |9°S | (ST6T)
/% \G/LE\6/Se4| STXOT
oD
61 |Z Tt |0°% (S61)
€/% |Z/LZ|Z/1ex| 9
N
S
FOL |£°8 |0'8
&/F |8/% |&/Ts cl
93 |T€ |9§ (F161)
G/LZx|0/GE|Z/2T | STXCT
WOUGS S$ C1
WOUdS S$ WHHLOW
ain
40 Had
-WON
uluodes-1juy ‘q
panuyuoj—|T ATAV.L
PLACENTAL TRANSMISSION OF ANTIBODIES 463
The curves for kids 12 and 13 (1915) are diagrammatically
represented. It appears that their increases and decreases are
concomitant until toward the end of the observations, when
this relationship is disturbed; that is at the time when the kids
take nourishment other than the mother’s milk.
ANTI-SAPONIN
@-—--© Kid 12 (1915) directly found values
IS» Units per cc. ~ Qe-eeeene ® * 13 (1915) " ® 8
——-- hother's serum
lo - i | N
fl ee
i] *y
Bp oN
a] oO
§ 7 ! ak
©! Sy
+ SO
eit
Bement 3) 1S red The A 1S eB Be aS Se gS
Crart 2
Finally the saponin-neutralizing power of the milk proved
slight and offered no support for the supposition that this food
influenced the content of these bodies in the blood of the kids.
REFERENCES
(1) Scnenx, F.: Monatsschrift fur Geburtshulfe und Gynaekologie, 1903.
(2) Potano, O.: Habilitationsschrift Wurzberg, 1904.
(3) Ransom, F.: Deutsche med. Wochenschrift, 1901, 27, 194.
(4) Mapsen, Tu., anv Noaucut, H.: Oversigt over det danske Videnskabernes:
' Sekskabs Forhandlingar, 1904, Nr. 6, 457.
THE JOURNAL OF IMMUNOLOGY, VOL. v, NO. 5
A SEROLOGICAL STUDY OF CHOLERA IMMUNITY
I. AGGLUTININ
ROKURO UMEMURA
From the Serological Laboratory of the Institute for Infectious Diseases, Tokyo
Received for publication August 27, 1920
The relation of serum proteins to antibodies has been the sub-
ject of much investigation in the past, and yet there are a number
of points which we do not fully understand. The earliest re-
search published in the literature on the subject is by Widal
and Sicard (15) in 1897, who concluded that agglutinin is pre-
cipitated together with globulin when the immunized serum is
treated with magnesium sulphate. In 1899 Winterberg (16)
investigated the action of various protein precipitants and found
that agglutinin is almost completely precipitated by sodium
sulphate and less completely by magnesium sulphate, ammonium
sulphate, sodium acetate and sodium nitrate, while only slightly
affected by sodium chloride and potassium chloride. He further
stated that animal membranes are impermeable to agglutinin,
and that when agglutinin is subjected to dialysis lasting more
than a month, its loss seldom exceeds 10 per cent.
Again E. P. Pick (10, 11) came to the conclusion, as may be
seen in the table 1, that the typhoid agglutinin obtained from
the immunized horse is contained in pseudoglobulin, that in
the immune serum obtained from the goat, rabbit and seal
agglutinin exists in euglobulin, and that the cholera agglutinin,
contrary to the typhoid agglutinin, is contained in euglobulin
and never found in fibrinoglobulin or pseudoglobulin.
It was Gibson and Collins in 1907 (3) who stated that their
results disagreed with those of Pick, but it is now generally agreed
that only the typhoid agglutinin from the horse is found in
pseudoglobulin, which is the only exception to the general rule
that it is contained in euglobulin.
465
THE JOURNAL OF IMMUNOLOGY, VOL. v, NO. 5
466 ROKURO UMEMURA
I shall briefly review the results of investigations by the above
workers and others, which I have repeated several times and
carefully examined, in order to determine wherein their results
disagree. E. P. Pick (10), adopting Hofmeister’s method of
classification, regarded euglobulin as being precipitated by the
addition of one-third saturated ammonium sulphate, and pseudo-
globulin by the addition of an equal volume of saturated solution
of the same salt to its filtrate. Gibson and Collins thought
that the precipitate which they obtained by the addition of 3.4
saturation of ammonium sulphate was euglobulin, while its
filtrate was pseudoglobulin. Following the numerous researches
of Weyl (14), Panum (9), Kiihne (7), Burckhardt (1), Pohl (12),
TABLE 1*
AGGLUTININ ANIMAL yap ted EUGLOBULIN yy caecum ALBUMIN
Horse 0 Trace | Nearly all 0
eR PHI s12 Nhe Soe Goat () All 0 @)
Rabbit 0 All 0 0
Seal 0 All 0 0
Wholeras eet ee Horse 0 All 0 0
Goat 0 All 0 0
* Pick: Handb. von. Kraus-Levaditi, Bd. I, p. 331.
Kauder (5) and others, Marcus (8) distinguished in the serum
the water-soluble and water insoluble proteins; while Hofmeister
and Pick (10, 11) discovered that the water-insoluble protein
can be precipitated by 2.8 to 3.6 saturation of ammonium sulphate
and the water-soluble protein by 3.6 to 4.4 saturation of the same
salt, calling the former euglobulin and the latter pseudoglobulin.
Again, Porges and Sapiro (13) classified the serum proteins into
three groups, each group having the precipitation limits by
ammonium sulphate at three distinct saturations, namely, 2.8
to 3.6, 3.3 to 4.2 and 4.0 to 4.6; and each protein body is thought
to contain soluble proteins. Moreover, Freund and Joachim
(2), repeating the experiments of E. P. Pick, found that both
euglobulin and pseudoglobulin contain soluble as well as insoluble
protein bodies.
Ta tt
be
ay.
mi
ie
es
5
SEROLOGICAL STUDY OF CHOLERA IMMUNITY 467
On the basis of the above experimental data, Hammarsten
(4) states, in his recent text-book on physiological chemistry,
that the classification of sera on the basis of water-soluble and
water-insoluble proteins is incomplete, and that the separation
of serum proteins by means of ammonium sulphate is far from
satisfactory. Accordingly, I diluted the serum five times with
distilled water in order to minimize the above interaction of
serum protein bodies, and yet within the limits of non-interfer-
ence with the examination of possible agglutination reaction, and
drew a distinction between the first, the second and the third
euglobulins, all of which were free from potential pseudoglobulin
and were precipitated at various precipitation points. Thus
the first euglobulin was precipitated at the highest point, namely,
3.3 saturation, the second and the third at 3.4 and 3.6 saturations
respectively. Likewise pseudoglobulin is divided into three
groups as follows: the precipitate which is obtained by half
saturating the filtrate which remains after the complete separa-
tion of euglobulin by means of 3.6 saturation of ammonium
sulphate is called the first pseudoglobulin, while the precipitate
obtained by half saturating the filtrates after the separation of
the first and the second euglobulin by means of 3.3 and 3.4 satu-
rations are regarded as the third and the second pseudoglobulin
respectively.
EXPERIMENTAL DATA
1. Preparation of saturated aqueous solution of ammonium sulphate
Ammonium sulphate is at first dissolved in appropriate quan-
tity of distilled water until fully saturated, and any iron present
is precipitated by passing hydrogen sulphide gas. It is evapo-
rated and made into a heat saturated condition, then cooled to
recrystallization by placing in running water. This recrystalli-
zation method is repeated six times. ‘The crystals of ammonium
sulphate thus purified are used in making the saturated solution
for the present experiments, by taking about 20 per cent in excess
of 770 grams of the purified crystals and dissolving in 1 liter of
heated distilled water. This is left at room temperature until
no more crystal is seen to go into the solution. The clear super-
natant portion is used as saturated solution.
468 ROKURO UMEMURA
2. Separation of the serum
The immune serum described in the present experiments has
been obtained from the blood of the jugular vein by letting the
corpuscular elements settle at the bottom of the containing
vessels after standing for twenty-four hours in a cool place.
3. Separation of serum protein bodies
The serum, which has previously been diluted five times its
volume with distilled water, is placed in three Becher glasses
(50 ee. each), and while stirring the saturated aqueous solution
of ammonium sulphate is added gradually in different quantities,
namely, 24.62 cc., 24.76 cc., and 28.12 cc. respectively. The
precipitates formed in the above treatment are the three varieties
of euglobulin; 3.3 saturation being the first, 3.4 saturation the
second and the 3.6 saturation the third euglobulin. These are
filtered after three hours, and the process of filtration is repeated
until the filtrates are perfectly clear. Each of the above pre-
cipitates is washed with 3.3, 3.4 and 3.6 saturation of ammonium
sulphate solution repeatedly until the solution becomes com-
pletely negative when treated with Spiegler’s solution. Each
filtrate is then centrifuged and 37.31 cc., 37.88 cc., and 39.06
ce. of its clear supernatant liquid is treated with 12.49 cc., 12.12
ec., and 10.94 ec., of saturated aqueous solution of ammonium
sulphate by adding the solution gradually while stirring, and the
precipitates (the third, the second and the first pseudoglobulins)
are filtered repeatedly until perfectly clear. The precipitates
thus obtained are dried thoroughly by pressing between the
filter papers, and then redissolved in distilled water, making up
to the original volume.
4. Examination for agglutination reactions
Each of the immune sera and the separated protein bodies
was diluted to twice the volume with sterile saline. The bacillary
emulsion was made by adding 10 ce. of sterile saline to one slant
of eighteen-hour culture in an agglutination tube. Two drops
SEROLOGICAL STUDY OF CHOLERA IMMUNITY 469
of each emulsion and serum protein were put in a tube and incu-
bated first at 37°C. for two hours and then at room temperature
for twenty-four hours and examined.
HORSE SERUM IMMUNIZED AGAINST CHOLERA
K. P. Pick precipitated euglobulin by the direct one-third
saturation with ammonium sulphate without diluting the strongly
positive horse serum immunized against cholera having the
agglutination value of 10,000. He also precipitated pseudo-
globulin by half saturating the filtrate with ammonium sulphate.
TABLE 2*
SERUM DILUTION
20 200 2000 4000 6000 10,000
NPSIGMUIN eas. 0.» 4. + +. = = =
Pseudoglobulin............. a + == = = x
*E. P. Pick: Hofmeister’s Beitrige, Bd. I, p. 378.
TABLE 3°
SERUM DILUTION
50 100 200 500 1000
PTT MOUMDUIN E.'s cas « so os — + = = =
Pseudoglobulin........... ++++ ]/++44+] 4+4+4+ +44 ze
* Gibson and Collins: Journal of Biolog. Chemistry, vol. III, p. 246.
His results are summarized in table 2, in which it is clearly
brought out that while euglobulin retains the original aggluti-
nation value of 10,000, pseudoglobulin shows a titer of only 200,
and yet he explains that even this small value may probably be
due to some substance which escaped through the filter paper and
thus it may be said that pseudoglobulin contains no agglutinin.
But the results obtained by Gibson and Collins are entirely
opposite. They precipitated euglobulin at 3.4 saturation with
ammonium sulphate from the serum having the agglutination
value of 1000, and its filtrate they regarded as being pseudo-
globulin. They found that while euglobulin was positive only
ROKURO UMEMURA
470
— | + | ++ (44e4eieta]e4slessalettietstiettittap cc II wyNqo[sopnesg
— | 4 | AE ETHIE HEH tHE tlt ltt titty I WyNqo|sopnosg
ahs -- ++ +4+4/4+4+4)/+4+4]4+4+4+]+4+4+/44+4+]/4+4++4+ che fekelio lee..e. es: 6) sile «..ef¢ 19 (aust s a0 slams mie lane)s aiakaup J Ulpnqo[sopnesg
= = pees ++ tH4tjt44)t44)+4-4 Nig piew elie \@hiwice! e's) (a,ere. 6,le)(a) jaqelvariaiia) (ea (6:,8 ) 008 (ONS ibl ewe) e) IIT Uyngosng
— == are ++ ++ t44 +44 +44 Are) 6) elie © le ©) pbs ve) 0) wile) =e) is) e/a se! 0; ols) ere (sje) eels lel ene Il uljnqo[sng
a aE tale SLT fish HEENE Ey aE oT Oe Polo. 8 o-cecrny C) Or of CLO ab reo" Dano Oe OsiclO.O.G ] WyNgo[sng
0 Toate = faiMfivecarl pest 06 fugit ean ore irelliera gee |vuetiara nave PR SOR AION, OCIS TER OO 2 REM SM umnsag
NOILOTMG
PE es a tp eR > Sn IRSA a Pee AU es oe SR a eS er ee ie
g asioy fo wnwas ay} ur sarpog urajoid fo sanjoa uorynurnpbby
¢ HITaVL
ee ie ae is ee ne es ae Sa ee ee
= Stealth tetas tetc lata steak late staat ctestotc testeste teat ct atest inl atest alee ee tyes wees III UtNqo[sopnosg
— |p | bb [EEE AHI E+ Hl ttt tt] ttt IL UNqoZopnosq
— | - | HE FEET ++I +4444 ]4 +444 titty [ UNqo[sopnosq
— == J ST +4 44+4)++4)4+4/4+-+4 Aer oh Cees) 5 UOMO, Wah CHC snd GLO TO chaicen III Wypngoysng
a at 4 +44 4++4 +44 +4 Alle (ecailw! e)'w) aie. ome €)n ej) e064 che) et age)sien)isas sale Il ulpnqoysng
== + + +44 +44 +4-+ ey b).e latte) 6 eceiw) oe ells }jchleehate cane elese 000 8 I ulpnqo[sny
2 rere tale ta lett t ete it ett et itt ttt et SBR GSee amen oeon Annee eC oaces wmnsag
NOILO TG
Se ee ee ee 8 SE eS
V asuoy fo wnwas ay) ur sarpog urajzoud fo sanjoa uoryourynp bby
b ATAVL
|
|
|
SEROLOGICAL STUDY OF CHOLERA IMMUNITY 471
to 100, pseudoglobulin was positive up to 1000. ‘The results of
my own study of the agglutination values of the protein bodies,
which have been separated according to the method already
described, may be summarized as shown in tables 4 and 5.
From these two tables it is clear that the agglutinin in the
serum of the horse immunized against cholera not only exists
in the euglobulin fraction, as brought out by Pick, but is present
for the greater part in the pseudoglobulin fraction as Gibson
and Collins have shown.
The reason why these investigators obtained results so much
at variance with one another may be inferred from the fact
that, while Pick used the serum without diluting, Gibson and
Collins diluted it twice its volume. The two questions which
we must at once seek to answer are: (1) Whether or not the
presence of ammonium sulphate hindered the agglutination
reaction, and (2) whether or not the dilution of the serum had
anything to do with the results. In order to test these points,
the following experiments were made:
1. Does the presence of ammonium sulphate hinder the agglutination
reaction?
The serum was diluted five times (100 ee. of serum and 400
ce. of distilled water) and 250 cc. of saturated ammonium sulphate
were added, making the degree of saturation of the entire mixture
one-third. The filtrate was again half-saturated with the satu-
rated solution of ammonium sulphate. The precipitates thus
obtained in both cases were dialyzed for three weeks. After
proving the absence of sulphuric acid and ammonia by means
of barium chloride and Nessler’s reagent respectively, the
agglutination reactions were tested and the results compared
with those obtained with the undialyzed fluid, as shown in
table 6.
This table shows that ammonium sulphate, which may be
present after the manipulation as stated in the introductory
paragraphs, hinders in no way the agglutination reactions.
\
472 ROKURO UMEMURA
TABLE 6
DILUTION
Horse serum... .|-+}-+-+/-+++/-+++)4++4+/4+4+/+$+|444+/4++4+144/4]=
Dialyzed —eu-
globulin....... +4+4+/+4+4+] 4+) + = =
Dialyzed
pseudoglobulin|+-++|+++/+++/++4+/++4+/+++/+++] ++ | + /-
Uudialyzed eu-
globulin....... +4++)/4+4++) ++) 4+ +- =
Undialyzed
pseudoglobulin|+++/+++|+++/+++|++4+|++4+]++4+] ++ | + |-
2. Do the differences in results depend upon whether or not the
serum was diluted?
Ten cubic centimeters of the serum were made 3.6 saturated
without diluting by adding 5.62 cc. of saturated aqueous solution
of ammonium sulphate. Again, as before, 3.6 saturation was
made from the serum diluted five and ten times respectively.
After three hours the precipitates in the above mixtures were
filtered, and the agglutination values of both the precipitates
and the filtrates were compared as shown in the next table (7).
The filtrates were in each case made transparent by repeated
filtration and the precipitates were washed with 3.6 saturated
solution of the salt, and repeated until the latter gave no reaction
with the Spiegel’s solution. The precipitate was then dried
by pressing between the filter papers and the solution was made
by adding distilled water, restoring the original volume of the
serum.
According to table 7, although the amount of precipitate is
greater in the undiluted serum than in the diluted, yet in no
case did the agglutinin come down exclusively in one fraction
as reported by E. P. Pick.
Thus it appears that the above stated questions are in both
cases negatively answered, and hence the above process and
results may be considered normal. I further endeavored to
SEROLOGICAL STUDY OF CHOLERA IMMUNITY 473
determine at what saturation the agglutinin began to come down
and at what point is it completely precipitated. When this is
determined we shall be able to say whether or not it is contained
in fibrinoglobulin and albumin as reported by E. P. Pick. The
following experiments were carried out.
TABLE 7
DILUTION
e|zgi|#e#/2]¢/18s|]3]2 {iz
Ses iss For Gee ty eels Hse Wie bes
Horse serum....... +2555) SPS eoe Gages Qrcece aoe tan crac Poaer sd rece Poe ee
Undiluted
Preempioate...-}---+-+|4-++\/+++i/+++| ++] + +
Piltrate......'. 3 3 3 5° 3 2 1
5 volume dilution
Precipitate..... 25555) 2=S-5ehegrgn, Geambcra= rae +
Pultrate. =<... :-- 3 3 3 3 3 3 74 1
10 volume
dilution
Precipitate.....|/+-++/+++|/+++] ++ /4+/ + +
Hiltrates. 22... 3 3 3 3 3 3 2, 1
Note: In the table, +++, ++, +, — indicate the degree of agglutination
with the precipitate as in table 4. The figures, 3, 2, 1 designate the agglutina-
tion reaction with the filtrate, — strong, medium and weak.
THE RELATION OF AGGLUTININ TO THE PRECIPITATION POINT OF
SERUM PROTEINS
The serum was treated with ammonium sulphate beginning
with 2.4 saturation and gradually increasing the amount of salt
up to half saturation. The precipitate and the filtrate were
examined as to their agglutination values, which may be tab-
ulated as shown in tables 8 and 9.
We were not able to examine the precipitate obtained at 2.6
saturation with ammonium sulphate because of its small amount,
but it may be said to contain some agglutinin, judged by the
fact that the precipitate obtained at 2.8 saturation gave a positive
reaction. In other words, the agglutinin begins to come down
with the precipitate by ammonium sulphate and the precipita-
ROKURO UMEMURA
ozz0¢/T |
OSF0z/T |
09est/T |
474
T G € € € € € € € € ic 0'Flo'Fo'zl 6
+ | ++) ++ saree teeter et ee tee et + eet et et
I G € € € € € € € € € € sez Flozl 8
i. e ‘ G E|8 F/O Z| ¢
eee ci ar ier ra ord Fates Rag Cte
€ € € & €
€ € &
fer ee rer eee era re rea PORTE COCO OC
+ | — | ++) ++ saree eter echt beet eee + ee tt et
I G € tS € € ts € € € € ts € gelpFle-zl 2
+ | + | ++] ++ pire rari werd fs lee fate utes Pneedh ova
I SG S € € € ts € § € € €
So Tee [ace (es leash ete a per cle a
SG € € € € € € € €
Soe
fe
€ € € € € § €
°
Be
€
|e | + feel at plete eels) 8429104 8
Prqiny, 9 SIF SIOZ) S
IBITD ¥c9 Si0'S| T
> Sie Seo] ae te | ee |S SR Sl ee ey elle | |e
Sg) 8) 2 so 8 |S |e Seles ae eS eee es
D F
S
NOILOTIG
8 ATAVL
475
=|) +] 4) EEE THI Ht tlt tft tye ttl + ttl ttt tttit+ti]t+ti]t44] umes joruop
=|) ++)tt4]444lt tlt ttt] t+tyt tlt] 4+t4+i4+t4+i4+t4i+t4+it+t]t++ti++4 ove lovejo-z
— |] 4] Ft] ttt ele tet] + tt] tyt tt] tttittt [+ t4+i+t+i+ttit++4]t++4] 8p lz elonz
ial
€1
Lemeletcesahevtea Ricca cae ere: eama
T I Sj § § § € € +¥ lo‘eloczl IT
SEROLOGICAL STUDY OF CHOLERA IMMUNITY
| I z g g g ¢ g ¢ g g g z+ lg-elo-zl Ot
I Z g g e g g g g g g g lor Salida a
Be eco teiee \ie eele else Pearle la eee re
feetleee elec eee lee) oc: wlcp lec | cele «oe alee ape cl ae alae
Si |p telt || ete es TE i Fee Te tees Ts A re fee ie |e re itera
Teese cot eeeedliee lee log cle fluo 1-6 gee. eee ag eee leg aca
S| SETS St ees repress erie itera ree) te ee re Tee Pre Bre fe, ee se, soe ical ¢
ile ieee hee | S| Ow oe ob ee gee alee ces eee | ieee eee (eee
fam
~ |e [ee | at eeeepeieeaetleteeta lee eleteteee late leetyh® Fog 9
Ae:
ae
|e
srs
a g € € € g g € € € g alee pleas
pe
e
€ € € €
Beater ee oe rare tea rat fers lee ah ll
I j € € € £ € € € g € € & € €
€
| i | ra gra ated re ora om a at tea ae
mpm ceeice = |e =|) gs 8 *| Se |e aie) ec yale sie lag eal geen
| te re ae Per elton ieee aac
qusoseiedgo | FZ 19'S10'2Z| T
ee ee ee ee ee ee ee
'OS*("HN)
07H
wauas
uaqaWoNn
NOILOTIG
6 WIAVL
476
Fl+[+4+]+4+4]44++] +++] +4++]+4+4]4+4-4[t4+4l44+4]444]t44l4+4+st]4t4lt4t]444]4+4)444]4+44] 0 uns yoryu0p
Flt +t|t+tltatl+t+[t+t]t+tl+ tt t+st]+t4+lttt|t4t +t ttt [ttt i444 i+44]++4)4+4-4] oe lovelo-g] 1
477
ela rte iter aera air eit Wate esr tres errs err esis Foy vats era Ore sh js Ses Ge
i G € ne alt
eA ete se race rea faa rte rr rere ero Wo Rafe tr err Perit oer a eed
I! a € € € € € 9°elo-zl 11
aft | 4+ H+eteetate tate tt eta e eet eee
I! T G € € € € € €
an re (rere rer eres rer Foerster ars Wrst stra Writs rita res rer Wei se os
P \8°€/0°%) OT
ee ee eee ee a ee le
478 ROKURO UMEMURA
tion is complete at 4.8 saturation. As to the agglutinin present
in the filtrate, we may say that in general the sum of the agglu-
tination values in both the filtrate and the precipitate obtained
at 2.8 to 3.8 saturation equals the original agglutination value
of the serum, but we also find that the agglutination value of
the filtrate at a saturation greater than 4.0 is markedly decreased.
This may be due to the fact that the presence of ammonium
sulphate to an extent greater than 4.0 saturation hinders the
agglutination reaction. The relation of the agglutinin to the
protein bodies which precipitate with the saturated solution of
ammonium sulphate may be expressed diagrammatically as shown
in table 10.
i
Les)
Sm
Of4 2.6 2:8 3.0 3:2 3,4 3.6 (3.8 4.0) 4.2°°4.4 4.6) 40eavoce
=a nnn The precipitation point of agglutinin
The precipitation point of proteins.
The classification of various protein bodies, according to Hof-
meister and Pick, is as follows:
Bibrino slob ulin cs pti sua tween kel eisparee 2.8 saturation and lower
PU GLO UNIT es ee bale etaee vrs ee seieiele oleracea 2.8-3.6 saturation
Pseudoglobulin...:.:.:.........s2000+-.0s00. 0.04.4 Saturation
Thus, contrary to the opinion of E. P. Pick, agglutinin is present
in fibrinoglobulin, while it is absent in albumin.
RELATION OF SERUM PROTEINS OBTAINED FROM DIFFERENT
ANIMALS TO AGGLUTINATION
Does the above relationship of agglutinin to serum proteins
obtain in the protein bodies of other types of immunized animals,
or, as E. P. Pick has stated, is there a difference according to
SEROLOGICAL STUDY OF CHOLERA IMMUNITY 479
the species? This is an interesting question both from the stand-
point of biology and from that of immunology, and the attempt
has been made to determine this point in the following experi-
ments with different animals.
Goat serum immunized against cholera
E. P. Pick examined goat serum immunized against cholera,
which had the agglutination value of 2000, in the same manner
as he did the horse serum, and demonstrated the large amount
of agglutinin in euglobulin as is shown in table 11.
TABLE 11*
DILUTION
ot |[oaltou Ss Sse toe | tar sn lies
i=} So —) So i—] = i] a = o co —]
N = ao -_ Nn vm] — — a = — nN
Ss | ba Isa soles 1 Sb eee sels
(0 Glo) hs ae Seal ae Seika Seo] Sedharal Seal sr] ac |) =
Pseudoglobulin.......... +y)+])+]-
* KK. P. Pick: Hofmeisters Beitrige, Bd. I, p. 380.
TABLE 12*
DILUTION
1/50 1/200 1/500 1/1000
SST +++ | ++4++ | +444 | +444 | +4+4++4+
ipelgbulinas.. 2. .2.....-- —— +
Pseudoglobulin........... SG sa eae Seana | anaes at +
* Gibson and Collins: Journ. of Biolog. Chemistry, Vol. III, p. 243.
But Gibson and Collins reported results quite contrary to
those of Pick, as is shown in the copy of one of their tables
(table 12).
I separated the goat serum immunized against cholera in the
same manner as I dealt with the horse serum and obtained the
results detailed in table 13.
Thus, as there seemed to be no distinct difference as regards
agglutinin content between euglobulin and pseudoglobulin in
ROKURO UMEMURA
480
IaieleanG S € € € € € € € § € & € € § oelosloz| F
= | + oP +t | tt it t4]44+4itt4}4+4+4]t44]+4+4+
EG NG te § € € € € § € € S € § € € ,
+ | +4] +4 fuera rede a) a |) Gen eels
Pprqiny 92|PFE|0ZC| Z@
queosafedg | F'Z |9'¢ |0'Z | T
Bee ee |e | a | re a |e es eee ee ee ea eae ee
Sieve # EB | ek BS | 2 | BciBal Eo ee ee ee ea Sie se
5\S Ss So So S Ss L z 5
DM i)
9
NOILLOTIG
bt HTAVL
+ | ++ | tt [et+]++ti++4)t+tst]44+4i[ttst]4t4] cc ITT WNqopsopnosq
ff AA | AH AEE HH] + H]+ ttt 44i]ttti44tp IT WNqoyZopnosg
+ fae | tt [+t +4ittst]44+4ittaty ] WTNGo[Fopnesg
aesalsateaieed| sctecte al -testeste|ctyreeh latastcstc p= eae ae IIT wyNgopsngy
Ss + +4 +4 +44 +44 scab iio\-e el s].egene\s)6i/ei'e;/e ee iene 6r9ie Il ulynNqo[sny
+ +4 ++ tt +++ ote kes eine oe \sie ie susp si eis 6b ya: 60 0! eas I ulpnqojsngy
— | F | t P44] at fet t]ett]+tt]+4ti+4ti[ttstittst]ttsty] wINIog
NOILOTNG
i NE eee See EE ——————ee ee aes
unias 7006 wosf surajoid fo sanjoa uoynurn) bby
€l ATAV.L
481
elt |EA| ELE HIE lt tlt ttle ttt t]+tti+ttlt+t[t44+l+ttlttt]tet] winzas [oI}UOK
IEE HEH EAH PTH PtHP ttt ttt ttt ttt tsi t+tit4+t]t4+t]t4t]t4t
—|4+t]t+ltttltttit+tit+tlt+t ltt] ttt] t ttt + tt]4+ tit) 4+t+
o'€ |0'S | €I
y'€ |0'S | GI
I € £ ge loz | It
4:5) cera erage eaeasa i Woe vh Perte TOD Wear Poser BETORD eres WPaceTa beer.
T z € € € e ge loz | OF
+f t+ | + [tttit+tytetyt+t]tttyttti ttt tty+ t+ t4i4tt]4+t+
I SG € § € ts tS €
+ fH LH Et tt ttt ttt ttfttt]tt+]t+4]t4]tt+4+]t4+
T I € € € g € g § €
= | +
§ § ; ;
+] FE [E+ tte] t st tt+l4+tt]t+4l+tt]t+sti+t4|t+ Pears
+n
e
€
tS
€
-+-
I G G € € € € € € € € § € € € rFlOZI LZ
+ | 4+ | +t [ttle] +ttietti+tt|+t+l4+t4|4+44i44+
& 3 € & € € €
+
€ § € €
Se ae linn oe Met rece ete aerial ir eel heel
SEROLOGICAL STUDY OF CHOLERA IMMUNITY
VP
vy
O'F |0'F |0'S | 6
8'é
9°€
ve
o'r 0G; 9
Sea j & & € & & & & € & oe &
jar
:
}*
GE i187 108 | ¢ re
§ €
=e =f t | t+] +t [ttt]tt+etti+tt]tt+4+it+4ti+tt
ROKURO UMEMURA
482
O9est/T
OFZOT/T
+ | ++ seek eee ee Heck Peo ter etree ele ore oh ee eee ecg teem eee an
OcIS/T
I
G j € € & g
++ Feo sco aia a. loro Mea eccale maleauicli,
i
G
09SZ/T
0c6I/T
§ € € € € § €
€ § € € € € €
+H J+tt+t+l+t+l+t4+l++4f++4
€ € § € € € 6
++] +4 [t+t]t++4+i+44+]/t+4]444+
§ € § Ge § € €
+] ++) t+ ft+t]4+t4]44-4]444
se fea (Oe teil eee eel ere err eee ee eee
etl
peel
prqany Ajsu0149
NOLLOTIG
St AIAVL
PIQiny,
yuooso[edg
IBID
+}
ee
Bs
ie
ye
*OS*(7 HN)
v |OF
av
VP
9°?
[oe 2)
—
1D 19 19
omnN oS
122
0%
0%
0%
0%
6
wD
aagnon
——————————— St
483
SEROLOGICAL STUDY OF CHOLERA IMMUNITY
nee
Aad fe es 8 YS ee As oo ee
a | See etal ere eae eee a Al ale a le a ee ee ls mies omguog
aa ES Sf Se es es ed es ee ee ee ee ee ee ee ee
— Se eee eee See ee ER al le oo ee la |
a eee eee ere Ee eh el gy lee le) st | E
=. . “a
* ; eee cetewallanerepseelg-7 |-er 8
te i (ee ae foo oa reer es oerm Woerer es Weteoru estere Weetews MusTens| Wunres Wes ora ewes Wevexru WeSTena Enos :
: Bs = * ; ; : é
eae : Pislie toa icine mle Paleo
See + | te] te lett ttle tsetse ttle tees ties eet tt ttt | :
484 ROKURO UMEMURA
the goat serum, the experiments in fractional precipitation of
the proteins were repeated with the results shown in tables 14
and 15.
Thus in the tables 14 and 15, we learn that, in the goat serum
immunized against cholera, there is a relationship between the
precipitation point of the agglutinin and that of euglobulin with
ammonium sulphate similar to that observed for immune horse
serum.
Rabbit serum immunized against cholera
Although we do not find any mention of rabbit serum immun-
ized against cholera in the report of E. P. Pick, we do find table
16 representing the result of an examination by Gibson and
Collins of such serum carried out in the same manner as that in
which they separated the horse serum.
TABLE 16*
DILUTION
1/50 1/100 1/200 1/300 1/1000 1/2000
Rseudoglobulin’...2. 25.2 .-0 t4++4/4+4+44)/4++44]/+444]/4444/ 4+
HBurelobullumye hace. «tis ckeieesee a — = = = a
* Gibson and Collins: Journ. of Biolog. Chemistry, Vol. III, p. 241.
My own results with rabbit serum immunized against cholera
having the following agglutination value is shown in table 17.
Thus the agglutinin in the rabbit serum immunized against
cholera does not exist either in euglobulin or in pseudoglobulin
exclusively. I have repeated with the rabbit serum the experi-
ments conducted with horse and goat serum and obtained the
results in table 18.
In looking over the above table, we note that there is no
distinction in the precipitation points of agglutinin in the rabbit
serum immunized against cholera from those of the horse or
goat sera immunized against cholera.
485
a ee ee
75 oll aig | egal Wie Sea a si ol ft be eo lo (oe IIT UN qo[sopnesg
spo Nore 4 Soe in eae a ope a ee etal eo eceg ooe leiee | pee a a II Wr[Nqojsopnesg
ye eee | la line eng even bret elfuranticlamcneeliecaieleeme joe. ae qa I WyNqo[sopnesg
ae cares oar ocala amc ymerloeaman( eas ee oN IIT WyNqopsngy
+ + +4 ++ +++ +++ i OCC Sint iC ate Cise mcd OC Crs 1 ul[nqosngy
== + ++ +4 +44 +4 CC ed m ulpnqoysniy
“el ie as eae ee bh el ee a Pe oleae kaos wunsog
Ser emeiea) |S | Bo) Sy SB | 8 | eS ls
NOILOTIG
he ae 2 ae Ss
4t ATaVL
ROKURO UMEMURA
486
' Z : . GP |8 € \0'S | OL
SNe tee tesla sleet dlate ate otal ste stecte| ate stect= tect stalls ectallete ae ata|ene Tost siete tel deals ale rt, oe
I I G g g g € 0+ loz | 6
Etameeaties fee |iectctanl ote taee otaatanba ate stratel ote shasta ahemieatr late stents ate teaial teaie stn pty cheats ann
I j & € g g g € is
or is
‘s ; :
+] ++ [4++4]444]t+4]+4+4]+4+4)+4+4]4+4+ Haale ee
tan
I G G € & € € € €
€ € € €
+ | + |4++4+]44+ seater tee eet eee tee tee ee
I SG G € te € € Ss re € € € ¢ 9°F loz | 9
+] + | ++ lt+t+t4it4+t]+4+4]4+4+ pele
Jefe G € € € € € € € € €
pees
gL fib |e rer Paper oh eae fone ea
f°"
+}
|} ore locale |)
jr"
ie
me
Te eles Peele ale Ol ase F< |eee eure 4S | a8. tite. le Ge lhe bees emer eae Brodie eeles
geod rar Pare Fase ase ae ee aT 4
PIqinL, | 8° [2g jos | €
quoosatedQ | 9°% |h'¢ [O'S | Z
way | FZ |9°9 OS | T
Pees Ss |-5 esc e-) ooo) as eae a es eee ae Ele] a
Biseercenecg oe oe he tee ea soe ee ee eee | | eee ©
%0S*0HN)
woud
ugawoNn
NOILOTIG
8t ATAVL
487
eee ee ee ee ee
ee et ttle tlt+tlt tte tt ttt ]t ttt et etait ttt ttt ttl tte ett] ttt] 7 ainsas porquo9
EEE tHE THIET HIF tH t tlt ttl tate tt lttti+tti+t+i+ts+itt+titt+tit+t) 86 [o's [0d | €1
T a § €
SEROLOGICAL STUDY OF CHOLERA IMMUNITY
soy VE (0S | Gl
T SG § g €
Vor 9°€ O'S | IT
488 ROKURO UMEMURA
CONCLUSION
We have thus far described the outline of various experiments
after each of which we added our brief comment. On the basis
of these results, we may draw the following conclusions:
1. Agglutinin begins to come down at the time when the
serum proteins begin to precipitate by the addition of ammonium
sulphate, and is completely precipitated at 4.6 to 4.8 saturation,
and therefore,
2. Agglutinin is present in both euglobulin and pseudoglobulin
and never in either exclusive of the other.
3. It is present also in fibrinoglobulin, but not in albumin.
4. There is a uniform relationship between the agglutinin and
the protein bodies in all animals (horse, goat and rabbit).
The author is deeply grateful to the kind supervision and
criticism of Professor Yokote and Dr. Kawamoto, and to the
helpful assistance of Drs. Y. Kato and Tamiya.
REFERENCES
(1) BurcxHarpt, A. E.: Arch. f. exp. Path. u. Pharm., 1883, 16, 322.
(2) Freunp, E., vu. Joacuim, J.: Zeits. f. physiol. Chem., 1902, 36, 407.
(3) Gipson, R. B., anp Couuins, K. R.: Journ. of Biol. Chem., 1907, 3, 233.
(4) HammarstTen, O.: Lehrbuch d. phys. Chem., Bd. 3.
(5) Kauprr, G.: Arch. f. exp. Path. u. Pharm., 1886, 20, 411.
(6) KoLLE-WASSERMANN: Handb. d. path. Mikrorg., Bd. 2 (1913).
(7) Kiune, W.: Lehrbuch der physiol. Chemie, Leipzig, 1866-68.
(8) Marcus, E.: Zeits. f. physiol. Chem., 1899, 28, 559.
(9) Panum, P.: Virchow’s Archiv, 1851, 3, 251; 1852, 4, 17, 419.
(10) Pick, E. P.: Hofmeister’s Beitrage, 1901, 1, 378.
(11) Pick, E. P.: Handb. von Kraus-Levaditi, 1908, 1, 331.
(12) Pout, J.: Arch. f. exp. Path. u. Pharm., 1886, 20, 426.
(13) Porasgs, O., vu. Saprro, K.: Hofmeister’s Beitrage, 1903, 3, 277.
(14) Wryt, Tu.: Beitrige zur Kenntnis thierischer und pflanzlicher Eiseiss-
k6rper.
(15) Winat, F., vu. Stcarp, A.: Ann. de l’inst. Past., 1897, 11, 353.
(16) WinterRBERG, H.: Zeits. f. Hyg., 1899, 32, 375.
THE VALUE OF THE INTRA-PALPEBRAL MALLEIN
TEST IN THE DIAGNOSIS OF GLANDERS
EDWARD H. MASON anp R. V. B. EMMONS
Received for publication September 4, 1920
While on duty in the Base Laboratory, Base Section No. 1,
American Expeditionary Forces, an opportunity was offered to
study the relation of the intra-palpebral mallein test, as advo-
cated by the French veterinarians upon horses and mules, to
the serological findings.
The work being reported was conducted during the months
of February and March, 1919, with the assistance of the staff
of Veterinary Hospital No. 9, St. Nazaire. It consisted in
a study of the agglutination and complement fixation reactions
of the sera of 94 horses and 8 mules; animals which gave slight
but not definitely diagnostic intra-palpebral mallein reactions.
Similar reactions were performed upon a control series of 51
horses, all of which gave negative intra-palpebral mallein reactions.
The 94 horses and 8 mules were bled on February 7, 1919, and
subsequently ‘‘malleined’” on February 9, 1919, the readings
from that test being the ones recorded in table 1. All of these
animals had been previously ‘‘malleined”’ on January 22, 1919.
The 51 controls had been “malleined”’ on February 9, 1919 and
were bled on February 20, 1919. The intervals elapsing between
the last previous malleination and the bleeding, sixteen days in
the case of the suspicious animals, and eleven days in the con-
trols, is subject to criticism, we think, as regards excluding the
presence of specific agglutinins or complement fixing bodies in
the sera at the time of bleeding.
The complement fixation reactions were performed as directed
by Memorandum No. 34, Office of the Chief Surgeon, Division
of Laboratories and Infectious Diseases, American Expeditionary
Forces, mule sera being inactivated at 62°C. for half an hour.
489
490 EDWARD H. MASON AND R. V. B. EMMONS
TABLE 1
Horse sera
|
MALUEINAS) | GOMeDEMENG ||| oo ee AGGLUTINATION, SERIES 2,
TION, HOURS! “ SIXATION SERIES I, FEB- FEBRUARY 24, 1919
AFTER TEST aa RUARY 18, 1919 ,
SERIAL 29 ea ERED SS Ba
NUMBER! ORSE No.1, | No.2,
Feb- | Feb- 1 1 1 1 1 1 1 1 1
ruary | ruary | 400 | 600 | 900 | 600 | 900 | 1200 | 1800 | 2400 | 3600
14, 1919] 17, 1919
5 1753 | 2) 2)2) + estas |e | ean aie tlc ee
6 iW 0 - U a Pe i iee | ere
Gey Mea Pee eT fee ee pee a TEP A ee |
G0 6 1762710 | 24 ee a a a
BOM 779 \doh Qe) ea eee eet] et
13 1369 We DA La hee +ytl—-}—-}—-|-|]-[-
ER Ds ls WW eT amg tg el ee
15 fe) 04 bap ea a ed ee pS a Bee ee
19 1S9G roe ee | em | eal ele lS
20 GAGE deal ata ol ole chal) ec eat ete ee etal eee
22 393°) Lop P| Rae See aati tonto [ty || ot ele
23 eT We ae ie Osh eh me | Pea Re aime es |
24 1 dW Br ete es a Or lee (ee | cee |)
sec] ies eh fea ey fat en (eee) [mettre eae rN Pee eee
26 SRON dtewe)) dae ECM ae a Pherae ee eee ee
27 70 a VB Se esp isteats |i | ataatn|ens || cil eee
28 9342 /1}/1/1) 2 | 4+4)t) 42) 42)/42/-] -)-]-t-
Tye 4 Weis eal Fad Wel as De) es he eee ay eae Pree eed en er eg fe Pe
303) 2603.) 05) 2) Do) aR | Se te ele el tect le
31 | 2531/1}/1/1/ + | & | +/+]-|-|-|-]|-
32) ) 2089); 1) Pi) + = a fae el eat (see ae ni gael eal Meme)
23. 1-770 Me OR Me i i 2 ce is Gal | eed re ae emt lc
34 B20 Lb Ls hp ete Ae eetaate: lott | aka Sete =
35 POOH TL Lileks see ese ah 2h | ile ea ae [ote ete
36 SJ Wi Oe Be Fa eal atl taltpatehainatghectnal | oc
37 Pollen 1a eae + [44/44+/4+/4+/4/-/]-|]-{-
38 ASSON Le — - PO ae fe 1 ee ee ie | oe
=) 2a) AS WR Pi Rs Fe | Re a er eon
AO 2575 | LE a a Werte ae ciel itl =| oo
BU ADE ONT 1 ye er et re eee Ns
42 BRL Owl On) Dil! bay | Weed Real SES AN reas pe | hoe ee
43 Abate Ol! a et + |+4+/4++} — |J44+)/—|]-|]-
44 sf (i fae an Oe BR Bee leeds P= Ne 9) hae Remain Cee ee
45 570 Ws es (tet a | Rs oe
AGA” 0264 14/1 | 1) fe] ae] aE A Ae See
47 SOM Dhed|-1 i) AE) |e SRNR SEISR et testal dr sty, ctr] at
48 STO a Dc | ate eee Peme
49 Samet a |) oat ae eae 2) = | ae
INTRA-PALPEBRAL MALLEIN TEST FOR GLANDERS 491
we TABLE 1—Continued
MALLEINA- AGGLUTINATION,
TION, HOURS foeaaone SERIES 1, FEB-
AFTER TEST RUARY 18, 1919
AGGLUTINATION, SERIES 2,
FEBRUARY 24, 1919
NUMBER! J orse No. 1, | No. 2,
meena ome | Keb- ) 2. | 2 de epee | | ke
ruary | ruary | 400 | 600 | 900 | 600 | 900 | 1200 | 1800 | 2400 | 3000
14, 1919| 17, 1919
50 =. 53 Pe a ee A ee) ee eS ee
51 352}1/1/1] + ae ger eee) ee | ae ee ee
52 sao }1)/1/1) + + hah ee ee Sa —
53 1329;1);1/;1] + St [Pee et Paha Fe
54 370;0;/1/1] + + Pela p—] Hh af— | =
55 382; 1]/1]}/1] + + |+4+/-/-|-/-]-]-
56 2522|1/1|/1|] + that aa aah eo
57 fea epi) i) + + [4++/4++/-—]4+/—-/]-|-]-]|]-
58 982);1/;1/1) + | + [44+1H4+/-]-]-—|]-]-|]-|]-
59 374 Ma Os a Ghat hh = oe
6O |. 2593; 1/1/1) + ==) ib ech ph bala Se
61 | 1011;/1/1/1} + | + | +]}/4+]-/]-]-J]-|]-|]-]-
62 S10} 2 | L jk - = we |e) ee)
63 1393 | 1} 1-|1] + =Es Wate ctaleats +}/—-}-|-]-|]-
64 LE TO i Sl We cen ct Ver = a I i
65 935 }1/1/1] — Se a ee ae et S| |
66 oem te ae Ee EE i st] Ee
67 1308; 1;);1)1] + + /444+)/-/-/]}-|-|-|-
adios) 2) %)1| + + [44+1—/,—|/—|—-|}=]=
69 eee ts) ee ee | | St Kt
70 eerie he RE eae eae et ate)
71 fet tii} — ee a |
72 SSO tp} 1} 1) + Sig wil amy Th aod | eects Gee) es coal Piteemiel (ie |
73 SE ay AER ae a VR scr al eh a Mee dS
74 844/1/1]/1] —- SG eo oiler == a [it el ce eel eed ie
75 583 }O0/}1/1] + Be epee ee Mee ae) Ge Py ey |e
76 ON ak a Ag a bern Oc bag orl ee eA Ue all Ui
77 602); 1)1)/1] + 2) ARechice ihe te bia
78 atti} i) — BEY a |e ee RE ee eS
79 Merril 1) — aa (eae a eng) i ee a ee ee
80 See yt |) + + ]44+)4+)/-/]/-]-/]-|]-
81 ST Fa oe in ce ld ot oom (rll (tn a co (a toatl cal Wie |
82 565} 1/1/11] + Fe ee estate tel Sac tee baa le bo. ee
83 948 );1);1/1); + + }/+/—/-—-/];-!|-—-|-]-
84 | 2655/1/1/1}/ ++) + }+]/-]}-|/-]-|-|]-
etait Pe) aS |
Be) 2314) t)}1)/1) — a ee eee a |
87 | 215/1/1/1|] — + |4+4+/++)/4/4+]-/]-|-|/]-|-
88} 2404/1/1/1)/4++/4++/+)/-]-/]-J]-|-[-
492 EDWARD H. MASON AND R. V. B. EMMONS
tbe
TABLE 1—Concluded By
MALLEINA- AGGLUTINATION,
~mco| COMPLEMENT =
TION, HOURS FIXATION SERIES 1, FEB
AGGLUTINATION, SERIES 2,
FEBRUARY 24, 1919
AFTER TEST RUARY 18, 1919
SERIAL | NUMBER
NGS ae No. 1, | No. 2,
94 [an | gael Reb | eben cbs yates te 9 _ Ds pce a eg a
ruary | ruary | 400 | 600 | 900 | 600 | 900 | 1200 | 1800 | 2400 | 3600
14, 1919] 17, 1919
89 Boson | etal eel eal eel! Se Se am Se) Se eee) eh =
90 TAU PETES peal if =e + J+-t+h—} —) te} He] eH
av | 7oo}1{/1)1) ++) + [44 eeie4]}4)—-]-—] =] =i
92 PATE ak | idly at aF +}—}]—-/]—-]—-]—-]-
94 ZO G4 ees a + SESE Se le eS eS S|
95 riycabend|| leet |ielien | ea — z= Se realy) || ae
96 PCa le |) al] aa + BE ES) ies PS os
98 2581 |1}/1/1/]4++)/)4+4+ 444+) 4;)-/]-/]-/]-]-
woo | 3szfrfrfa] + | + J++tet) = |4+l-4]+4) 4] -] -
101 Coal ale lenl al Se 35 fe} — |e] —)]—) —
102 CH OMl i lnslaieal eel: Be ee faecal se | SS
103 e-20}1/1)1 + + et | ie ta ia (cf
104 | 977; 1)1 fa) ++) +4 ]4+4) 4] - f+4i+4] - | -
105 S¥(ileh ak |r al + + {44)/+4+/4/—]—]/—-—]—]—-—]—
106 SO2Z aml aetna at + j44+)4+/-]-]-/]-|]-
107 D252 ale Slee = + j++/+4+) —-]); +/-—-] -—] -
108 PSH Soils se: _ + j4+4+/++)/4/;)-—-/-/]-]—-]-]-
TABLE 2
Mule sera
MALLEINA- AGGLUTINATION, A
cron mova] COMPtmeENT | Sore ire’ | ASCLETINARION, SHBIES2,
AERTATH Oe
sR ep ers No. 1, | No. 2,
o4°|93°| go) Feb- |; Rebs.) 1) to te a a
Tuary | ruary | 400 | 600 | 900 | 600 | 900 | 1200 | 1800 | 2400 | 3600
14, 1919} 17, 1919
2 | 1605 [Doubtful | ++ | ++ |++/4++/44144] + | 4+])—|]—]=
3 af be PA + ++/—/—/;/—-|/—-];—-|/]-]-—-
4 DOAN OZ eal, + +4f—}—]| =—}—}] —]—]=
8 HSS le 212 oo + re ig tee Te bast |e fo ce ee
9 1696 242252 _ + oie ih Pages ea eet NL NE nna |
17 TSGor eels eT - 4+ Sc Poe ON een EE te UN ae ae
18 TOGA | tet oo = ei) ee a een ioe Nie ge ee
21 AS 2a ad, 2 _ + fee eee fs eee en ea eo
INTRA-PALPEBRAL MALLEIN TEST FOR GLANDERS 493
The antigen was obtained from the Central Medical Department
Laboratory, it having been made from a single strain of B. mallet
previously isolated at Neufchateau. Two series of reactions
were performed as shown by tables 1 and 2.
The agglutination reactions were carried out in dilutions
varying from 1:400 to 1:3600 inclusive. The antigen used was
obtained from Neufchateau. ‘The strain possessed a character-
istic morphology, giving as well a typical growth upon potato
media. The tests were carried out as follows: a twenty-four
to forty-eight-hour growth on glycerin agar slants was emulsified
in physiological saline. ‘Two series of reactions were performed,
in the first the dilutions ranging from 1:400 to 1:900 inclusive,
while in the second series the dilutions ran from 1:600 to 1:3600
inclusive. In the first series the suspension was employed
unheated, while for the second series it was heated at 60°C.
for one-half hour. A uniform light suspension was used in both
series. After being set up the tubes were kept for one-half hour
in a water bath at 56°C., then for twenty-four hours at 37°C.,
likewise in a water bath, and finally in an ice box over night,
when final readings were made. ‘The horse sera used were heated
at 56°C. for one-half hour, while the mule sera were exposed to
62°C. for the same time. The controls were carried through
with the same technic as was used for the first series. Being
all horse sera, they were heated at 56°C. for one-half hour.
From tables 1 and 2 it will be seen that serological reactions
were carried out on 94 horses and 8 mules. The mallein tests
recorded were done upon February 9, 1919, being read by Majors
McKillip, Ratigan, and Gould of the Veterinary Corps. They
are recorded by figures 0, 1, or 2, it being the old system of read-
ing these reactions, 1 being doubtful, and 2 suspicious. The
complement fixation reactions were performed in duplicate, the
same technic being used for both series. The agglutination
reactions were likewise carried out in duplicate, the second series
differing in that the emulsion was inactivated for one-half hour
in a water bath at 60°C., as previously explained. Readings:
a double plus means complete fixation or complete agglutination,
while one plus expresses almost complete fixation or agglutination;
494 EDWARD H. MASON AND R. V. B. EMMONS
a plus-minus means slight fixation or agglutination, while a dash
represents complete hemolysis or no agglutination.
Complement fixation reactions. Horses: From a study of table
1, it will be seen that 71 of the 94 sera from horses gave a positive
complement fixation. ‘Twenty-three of the sera were negative.
TABULATED: HORSES 94 NUMBER PER CENT
Complement fixation (---++) ‘twice... .... cid. ccc cece wees 13 13.8
Complement fixation (++) once, (+) or (+) once........ 6 6.3
Complement fixation (++) at least once.................. 19 20.1
Complement fixation, (-)itwice 2/0. eee ae sas as, see 33 S02
Complement fixation (—) or (+). AOR 23 24.4
Complement fixation (++) or (ay at leases once. shee 71 75.4
It was impossible to control these complement fixation reactions
with absolutely known positive and negative sera, but the fact
that in both series there were reactions where there was complete
fixation and no fixation would indicate that they controlled
themselves. Throughout both series, two sera, serial nos. 20
and 91, were the only ones found to be anticomplementary and
these were both in the second series done on February 17, 1919.
Agglutination reactions. Horses. In consideration of the fre-
quent presence of agglutinins even in a dilution of 1:400 to 1:500
in normal horse sera, and the fact that the sera were heated to
56°C. for one-half hour, we have adopted as a dividing line
between a positive and a negative result a dilution of 1:600.
Reactions to the degree of double plus or plus with a dilution
of 1:600 have been considered as positive.
TABULATED: HORSES 94
1ECOYSb RT A yet ee EG Ras MRO nee oli La ol a a IPAS
Borderline. )).3:..
Negative
Borderline reactions are those which gave a single plus in one
series at a dilution of 1:600 while failing to give a plus or more
in the same dilution in the other series. Considering the border-
ee ere ee ee ee
,
.
a ol al nae
INTRA-PALPEBRAL MALLEIN TEST FOR GLANDERS 495
line cases in conjunction with the definitely positive ones a
percentage of 57.4 is obtained.
Complement fixation. Mules. The sera from eight mules
giving mallein reactions of the same type were carried through
the same tests. Certain difficulties were experienced with the
mule sera, they being partly due to their being inactivated at
62°C. for one-half hour while the horse sera were inactivated at
56°C. for the same time. One of these sera coagulated at this
temperature; animal no. 1391, necessitating its being discarded
from the series. Others showed certain physiochemical alter-
ations resulting in the sera becoming of a viscid consistency,
although still remaining liquid. The discrepancies between the
two complement fixation reactions we cannot explain. One
serum, serial no. 2, was anticomplimentary on the second test.
Another interesting fact is that due to an oversight, the serum
of mule, serial no. 2 was inactivated at 37°C. instead of at 62°C.
Both fixation reactions in this case resulted in a (++) fixation,
although the second test was anticomplementary. Further this
was the only mule serum that showed any agglutination, in this
case being positive up to a dilution of 1:1200.
Agglutination reactions. Mules. In our opinion the agglu-
tination reactions as carried out upon these eight mule sera
are subject to severe criticisms. It has been stated previously
that they were heated at 62°C. for one-half hour. We think that
the heating of the sera to this temperature for that time would
destroy almost all, if not completely, the agglutinins present.
This opinion is rather corroborated by the agglutination reactions
in the case of mule, serial no. 2, where the serum was heated to
only 37°C. for one-half hour.
Complement fixation and agglutination. Taking these two
reactions in conjunction we find that 38 times out of the 94
horses both reactions were positive in the same animal. This
gives a percentage of 40.4. In 15 instances both reactions were
negative in the same animal giving a percentage of 15.9.
Controls. As a control upon the above work, complement
fixation and agglutination reactions were carried out upon 51
horses which gave a negative intra-palpebral mallein test. There
496 EDWARD H. MASON AND R. V. B. EMMONS
were no mule controls. 'The complement fixation reactions were
earried through with the same technic as used for the horse
series. The agglutination reactions were accomplished by the same
methods as were used in the first horse series. Out of the 51
tests, 3 gave a one plus reaction, while 2 others gave a plus-minus
result. In no case was the serum anticomplementary. These
results are satisfactory in that they indicate that there would
be an error in only 5.8 per cent of the cases.
Agglutination. The controls are very satisfactory since in not
a single case do they show a one plus reaction with a 1:600
dilution. Considering the agglutinins that are often present in
normal horse sera, this series would tend to indicate a very low
average agglutination titer among the normal horses. Also it
would indicate that our choice of a 1:600 titer, reading one plus,
was well inside the safe limit as a diagnostic line, which would
indicate that the borderline cases in the suspicious horse series
were probably positive cases.
SUMMARY
There are certain factors that should be taken into consider-
ation in interpreting these serological results. The fact that the
sera of all the suspicious horses and controls were heated at 56°C.
for one-half hour would tend to cut down the agglutinating titer
of those sera. However, as the same technic was carried out in
both test and control animals the results have a true relative
significance. This fact may account for the absence of agglu-
tinins in such a large percentage of the normal control horses.
The results definitely indicate that in horses giving a suspicious
intra-palpebral mallein reaction of this type the complement
fixation test is of more value than the agglutination reaction. In
the former 75.4 per cent were positive while in the latter only
57.4 per cent were positive, even when all the borderline cases
were included.
These observations agree fairly well with published statistics.
In Povitzky’s recent article (this Journal, 1918, 3, 463) she records
work done in the laboratories of the New York City Department
INTRA-PALPEBRAL MALLEIN TEST FOR GLANDERS 497
of Health on 123 horses which were proven at autopsy to have
glanders. The complement fixation reaction was positive in
75.6 per cent of these cases, while the agglutination reaction was
positive in 64.2 per cent. In her work she used horse sera which
had not been heated to 56°C. for one-half hour, but again this
would be partly neutralized by her using a titer of 1:1000 as
necessary for a positive result. The consensus of opinion is
that the agglutination reaction gives higher positive results in
acute cases, while in chronic or subacute cases the complement
fixation is the more reliable.
CONCLUSIONS
In consideration of the problem in view, that is, whether doubt-
ful or suspicious intra-palpebral mallein reactions in horses are
confirmed by the serological findings, our facts would indicate
that the complement fixation reaction is of greatest value, it
being positive in 75 per cent of the 94 horses examined. The
agglutination reaction ranks second, confirming a suspicious
mallein test with a definitely positive reaction in 44 per cent
of the cases in the same series. ‘Therefore we would conclude
that the complement fixation reaction is of the greatest benefit
in confirming a doubtful intra-palpebral mallein test, but that
this reaction should be considered in conjunction with an agglu-
tination test, one to act as a check upon the other.
j j H ay i :
Lan hs, TE Ue .
‘ ® vl arr ary 7
i" 5? >| 5 : '
ie
: ,° | ;' i's »
Vine Le aALY
!
ys > wea, .
OS oe ok C0) ROR
- ‘ :
j ;
. } \o Le i i aa
: ‘ fy cg tie
? ae 4 7
~ dis fig ch iets b>,
4 . oe
ra , ety
ptt,
i -
; ; \
‘
Z i ye pees rh p ere © | Fk “Seay
r . i Tr ie we to MD ue ah ie
‘ F oq r L¥3 or eT Cae f Po aed tues or oe ras ei!
a oa Ne y tat \
, . j APSA ta
‘ ae - } — 4 ee ye
.
COMPARISON OF SMEAR, CULTURE AND COMPLE-
MENT FIXATION IN CHRONIC GONORRHOEA
IN WOMEN
A PRELIMINARY REPORT
JAMES D. SMITH! ann M. A. WILSON?
From the Kingston Avenue Hospital, Brooklyn, New York, and the Bureau of
Laboratories of the Department of Health, City of New York
Received for publication August 31, 1920
CLINICAL
The clinical material, serums and film preparations on which
this report is based, were obtained from patients under treat-
ment on the Venereal Disease Service of the Department of
Health Hospitals of New York City. Ninety per cent of these
patients are public prostitutes, the remainder being of the clan-
destine type. Their ages range from sixteen to sixty, the average
being in the early twenties.
For convenience in outlining treatment and presenting the
necessary data for the bacteriologist, the cases on admission are
divided into three general classes, acute, subacute and chronic.
Acute
Vulva. Active inflammation and swelling of the parts. In-
volvement of Bartholin glands with or without abscess formation,
but exhibiting some evidence of inflammation in the duct.
Urethra. Evidence of active inflammation as witnessed by
mucopurulent or purulent discharge and swelling of the mucous
membrane at the mouth of the urethra and ability to express
a drop of purulent material on massaging Skenes glands.
1 Clinical study.
2 Laboratory studies.
. 499
THE JOURNAL OF IMMUNOLOGY, VOL. v, NO. 6
500 JAMES D. SMITH AND M. A. WILSON
Vagina. Inflammation present in varying degrees; mucous
membrane swollen, reddened and sensitive.
Cervix utert. Presents a swollen and reddened appearance
with or without erosions and a muco-purulent discharge.
Subacute
Those cases presenting any or all of the symptoms of the acute
stage but in less intensified form and showing usually erosions
of the cervix. The discharge is free and mucopurulent in
character.
Chronic
Those cases presenting no evidence of active inflammation
and exhibiting some or all of the following symptoms in varying
degrees. |
Discharge. Slight or profuse; mucopurulent in character.
Bartholin glands. Evidence of involvement as witnessed by
fibrous change in the gland itself, with or without patent duct,
and showing purulent discharge.
Urethra. No inflammatory condition apparent. On massage
pus obtained from Skenes glands.
Cervix. May or may not be enlarged. Canal usually unduly
patent. Erosions fairly constant. Mucopurulent discharge.
Evidence of prior adnexa involvement such as thickened tubes,
loss of mobility of uterus; evidence, or history, of previous pelvic
or abdominal surgery.
In table 1 are given a few of our cases which are representative
of all. If plotted these would show a remarkable similarity to
the Wassermann curve in a luetic responding to treatment. In
the cases shown, the complement deviation has at no time been
at variance with the clinical picture.
In the acute and early subacute cases, we have had positive
cultures and smears when the complement fixation reaction was
negative. In each instance there was no doubt of the infection
being a recent one in very young girls from sixteen to twenty
years of age and of the clandestine type of prostitute.
eee
CHRONIC GONORRHOEA IN WOMEN 501
In the chronic cases, our experience has shown the complement
fixation test to give as high a percentage of positives as does the
Wassermann reaction in old luetics.
TABLE 1
Showing the pranval reduction of the complement fixation. Bleedings were made
at seven day intervals
PATIENT GONOCOCCUS COMPLEMENT FIXATION REACTIONS
1 smatemiain lageate ta |, Soe \amdeoh | Sie pecan | |) ef)
2 SRE an a Tr Fela ee ote eal d= a =
3 she sie si aise
4 se = Sa ee ee ose
5 Tar eee ae a Fah) | gabe lie inte be area hie
6 aime ete a tt | | to
7 = 9° | Gees la aa Ali mS ie ao ie =a = Ea
8 SPAS) | A a la i a i wal = = far ee
9 =PS°A eo | ar ee al i a (ee Pe =
10 = == +f] +/ -
LABORATORY STUDIES
Our laboratory studies have been concentrated upon experi-
ments to improve the methods for diagnosis by smear, culture
and complement fixation. A trial of many methods for each
of those forms of diagnosis led us to adopt the following technic
as our standard for routine examinations.
Technic and diagnosis
Smear. ‘The smears were made by the clinician at the time
the cultures were made. A drop of the discharge material was
taken with a sterile platinum loop and smeared upon the surface
of a sterile slide. When very little material was obtainable, it was
found to be an advantage to place a drop of sterile water on the
slide and gently manipulate the loop containing scant secretion
in that. By this means the leucocytes were not destroyed and
we succeeded in making a good smear in many instances where
we had failed to do so without the water. After drying and fixing,
the smear was stained by Gram’s method. During the first
502 JAMES D. SMITH AND M. A. WILSON
months of our studies when stable staining powders were avail-
able, we used the Nicolle modification of the Gram method.
Later we used the Leitz aniline gentian violet powder withStirling’s
method. Before staining the smears, we tested all of the staining
solutions with twenty-four-hour cultures of Staphylococcus pyo-
genes aureus and Bacillus coli. Without these control tests, we
consider the Gram method to have no value as a differential
stain.
Smear diagnosis. We have followed the Williams (1) rules for
smear diagnosis:
1. Positive smears. Those showing leucocytes filled with
morphologically typical gonococci.
2. Suspicious smears. Those showing any suspicious intracel-
lular diplococci and 50 per cent, or more, of polymorphonuclear
leucocytes.
3. Observation smears. Those showing 50 per cent, or more,
of polymorphonuclear leucocytes, but no suspicious intracellular
diplococci; or, those having the clinical symptoms of discharge
and inflammation and showing less than 50 per cent of poly-
morphonuclear leucocytes.
4. Negative smears. Those showing less than 50 per cent of
polymorphonuclear leucocytes, no suspicious intracellular diplo-
cocci and no clinical evidence of the disease.
Culture. Plates of glycerine-veal-horse-serum-agar, streaked
with rabbit, human, or horse blood, were inoculated by means
of a platinum loop. The discharge material was streaked across
the surface of the plate. The plates were incubated at 36°C.
for one week and examined daily. The average optimum growth
of the gonococcus in the isolation culture is forty-eight hours. In
two of the cases, the gonococcus was isolated from a three day
plate that had shown no growth within forty-eight hours. Occa-
sionally the colonies develop within twenty-four hours. All
colonies appearing suspicious were fished under the microscope
with a magnification of three hundred diameters. ‘The pure
cultures of gonococci were grown upon glycerine-veal-horse-
serum-agar for several generations until they could be grown
upon a serum-free medium. Some of the gonococci isolated in
SS eS eee *
CHRONIC GONORRHOEA IN WOMEN 503
this study made a fair growth upon disodium phosphate agar
in the fourth generation, but most of the cultures had to be
kept on a serum medium a longer time. ‘This is necessary
before transplanting to the North medium, but once induced to
grow upon it they have continued to grow well. None of the
gonococcus cultures have grown upon ordinary agar.
Culture diagnosis. We have diagnosed as positive those cul-
tures from which we have isolated in pure culture a biscuit-
shaped Gram-negative diplococcus, dividing at right angles, that
does not grow in early generations without the aid of blood-serum.
This rule for diagnosis excludes the Micrococcus catarrhalis and
other Gram-negative cocci that grow immediately upon ordinary
agar. At this point, it may be well to emphasize the importance
of obtaining a growth from the fishing transplant before making
a diagnosis of the culture. We consider a diagnosis unreliable
which is made upon the macroscopic or low-power microscopic
appearance of a colony, unless it is corroborated by a fishing
showing typical gonococci in smear. This is essential for two
reasons: first, that several other organisms make similar colonies
upon this medium; and, second, that the gonococcus does not
always make the typical colony.
We have diagnosed as suspicious all cultures in which we have
found colonies showing suspicious Gram-negative diplococci in
a smear of the colony and from which we have obtained no
growth after fishing; we have considered suspicious also, all cul-
tures from which we have obtained mixed subcultures containing
typical or suspicious Gram-negative diplococei which have not
been isolated in pure culture. In some instances the Gram-
negative cocci have not survived the repeated transplants neces-
sary to free them from the associated growth of the other organ-
isms and, therefore, cannot be definitely classed.
We have diagnosed as negative all cultures from which we have
failed to obtain typical or suspicious colonies that have given in
fishings suspicious Gram-negative cocci. A large number of sus-
picious colonies having marked central granulation, have given
in fishings pure cultures of a minute Gram-negative bacillus and
also pure cultures of a lancet-shaped Gram-positive coccus.
504 JAMES D. SMITH AND M. A. WILSON
Complement fixation. Antigen. The complement fixation
tests reported here have all been made with an antigen prepared
from the strains of gonococci isolated by Torrey several years
ago. ‘The gonococci were grown upon disodium phosphate agar
for forty-eight hours at 37°C. They were then treated with
alcohol and ether to remove the lipoids; dried, weighed, powdered
and suspended in saline in the proportion of 1 gram of the dried
powder to 200 ce. of physiological saline solution. This suspen-
sion was heated in the water-bath at 80°C. for one hour and was
then ready for standardization. The standard dose of our antigen
contains two fixation units and one-fourth, or less, of the anti-
complementary unit. This antigen is stable when kept at
ordinary ice-box temperature. Freezing does not injure it. No
preservative is used.
Serum. The patient’s serum was inactivated before the test
and was used undiluted.
Complement. Guinea-pig complement was used. The serums
from all guinea-pigs were tested for natural antisheep ambo-
ceptor, for hemolytic activity, for anticomplementary reaction
with the antigen and the control serum and for fixability with
the antigen and the control serum, before being pooled for tests.
These preliminary tests are essential for the reason that some
guinea-pig serums are not fixable by gonococcus antigen and
serum; if non-fixable serums are included in the pooled com-
plement, false negative reactions of the patient’s serum may be
obtained.
Hemolytic system. The antisheep system was used. The dose
of complement contained two hemolytic units. The hemolytic
unit was obtained by titrating a 10 per cent dilution of the pooled
complement with sensitized cells. The sensitized cell dose con-
tained two hemolytic units of amboceptor and 0.1 cc. of a 5 per
cent suspension of sheep-cells. The complement unit was read
at the end of thirty minutes in the water-bath at 37°C.
The test. One-tenth of the original Wassermann volumes of
all reagents was used. The tests were made in duplicate with
controls for hemolytic system, for anticomplementary reaction
of the patient’s serum, for natural antisheep amboceptor in the
SS ee ee
CHRONIC GONORRHOEA IN WOMEN 505
patient’s serum and for fixation unit and anticomplementary
reaction of the antigen.
Complement fixation diagnosis. The Citron (2) method for
diagnosis was followed.
TABLE 2
Showing the incidence of the complement fixation reaction in comparison with smear
and culture findings
|
CUL- COMPLEMENT FIXATION DIAGNOSIS
NUM- SMEAR
CLINICAL DIAGNOSIS BER OF| piac- | TUBE |} AAA _—
CASES | NOSIS Ghrs Strong| Posi- | Weak |Doubt-| Nega-
positive} tive |positive| ful tive
21 | Neg. | Neg 2 5 9 a 1
: : 22 | Ob.* | Neg 5 a 7 2 1
Mild chronic gonorrhoea. Ae | Pad Pas 0 9 0 1 0
4: Ob. ||-2os 1 a 1 2 0 0
TUGUDI Re Ooo 4 See aoe eee 50 8 15 18 ‘ih he
Controls: Clinically nega-
tive
@inldrense ns 5 ance. ss 8 on oa 0 0 0 0 8
PANGIULU SHEE ets! «= Peary Ree ig
aX? al 7 +
we i 7
i we hs i ayes al
‘ maid! 3 Line ' : et
a é
Rin 5 i
Pete ™~ yank ;
a - : ‘ _ : A i ane '
¥ *, Ok (iter eee NEE ln ton! b A.
" At i : oe i \ j
; rf fue}, Oot ee Ab |
mn . i % ue { PF ike? 4 4 7 _
Y - A a tad Wily. ee 8 ” Jet OO ri 2 kod tly s ie a % iN
i" i es A ens adi hy MVS: J
a / } ; j su ici wiik. -
} Aa Ae , Wh 1
a ‘ i J Ap ete
mara. De peony ; tea SA
x; . *, :
e at - . ak:
: id toys \ fics ' J we
Cia 4) : ~ } é
- re. Ve: . : ray ‘
' ’ =A " ti {
x : : Fouls ‘ f a 7
- ’ - Aa pit A. Jat
pi, 5 ,
; YSERA ie i
‘a { a ; ar Nes
oo f ¥ s i +
* Wi a ; Sh, iy 4 ol ' : i, Sip Gow ee Pi hry .
yk ge Beaslih a er a AW ST ea Nes iges gmat ave
ay - 7 }
4 Por, ; ,
3 5S Perea aes. > ee be Vai shee pve | oa Wi il baie iu,
> - . :
F 4
‘ae AAO):
‘yma fir i Sa) Th};
ae ikgits, See ivi
: $0: UDRP Y
iy PTE a cre i bare? win —_ ‘ida sand | ae
mA 4 4
dior ety eT et Laetares, re Lit
Gert, enti Yale ¥ Ua
Ys
Po
At
ee
aalaaacerxe’ bapa alvin ia tn ty ie oe
tries vATLIT? it feat “LAR
Hue dalg
aS
Hi
0 a 1;
THE SACHS-GEORGI PRECIPITATION TEST FOR
SYPHILIS
THOMAS G. HULL anp EVA E. FAUGHT
From the Illinois Department of Public Health
Received for publication September 23, 1920
For several years investigators in various parts of the world
have given more or less time and effort in attempts to find a
test for syphilis less elaborate and complicated than the Wasser-
mann test, and, if possible, more delicate and specific. These
attempts have been partly successful, but with the exception of
the Lange colloidal gold test on spinal fluid, none of the sug-
-gested procedures have been generally adopted.
One reason for the rather slow advance is the fact that we are
uncertain as to the exact nature of the mechanism of the Wasser-
mann reaction—whether it is caused by a specific antibody or
by some other substance. The problem is still further compli-
cated by the extremely difficult task of growing Treponema pal-
lidum either on artificial media or in the lower animals. By
close clinical studies of cases, however, and carefully correlated
laboratory investigations much progress is being made.
In 1912, Klausner (1) reported his findings in the reaction
between blood serum and distilled water. When these two are
mixed together in the proper proportions a flocculent precipi-
tate separates out, the amount of the precipitate being much
greater in syphilitic than in normal sera. _
The same year Lange (2) attempted to demonstrate a dif-
ference between normal and syphilitic sera by their relative
precipitating properties for colloidal gold solutions. He was
unsuccessful in this but discovered that spinal fluid in various
dilutions gave most characteristic reactions.
About the same time the Porges-Hermann-Perutz reaction (3)
was brought out in which equal parts of sodium glycocholate
521
522 THOMAS G. HULL AND EVA E. FAUGHT
(2 per cent solution) and an alcoholic cholesterol suspension (0.4
per cent) were used with inactivated patient’s serum, a precipi-
tate forming with syphilitic serum, while none forms with normal
serum.
In 1914 Hirschfeld and Klinger (4) reported that tissue ex-
tracts digested with syphilitic serum lose their ability to coagu-
late blood.
In 1917 Vernes (5), in working on the precipitation of colloids,
found that ferric hydroxide would precipitate the colloids of
blood serum in certain dilutions, the dilutions being different in
syphilitic sera from those in normal sera.
Of late considerable attention has been given to the Sachs-
Georgi (6) test. This is essentially a precipitation caused by
the mixture of syphilitic serum and cholesterinized antigen. In
normal serum no precipitate occurs. The antigen (7) consists
of an alcoholic extract of normal beef heart, to each 100 ec. of
which is added 200 ce. of alcohol and 13.5 ce. of a 1 per cent
solution of cholesterin. For use this is diluted one to five with
physiological salt solution. In the test 0.5 cc. is mixed with
1 cc. of serum which has been diluted to 10 cc. with physiological
salt solution. The results are read after twenty-four and forty-
eight hours.
Because of the large volume in the Sachs-Georgi test, the rather
scant precipitate often formed, and the length of time elapsing
before the test is read, the present authors have devised a modi-
fication which has proven in their hands, much easier to manipu-
late and at the same time, no less sensitive in giving positive
results with syphilitic sera. This modification differs from the
original technic in that the antigen may be either cholesterinized
or not; a much smaller volume is used—1.3 cc. instead of 10.5
cc.; and the tests may be read immediately, if centrifuged.
The antigen. The dilution of the antigen is of the greatest
importance, careless addition of the salt solution often giving
erroneous results. The required amount of undiluted antigen
is placed in an Erlenmeyer flask and physiological sodium chloride
added drop by drop with vigorous shaking until at least 5 ce. of
the diluent has been run in. Larger quantities at a time can
SACHS-GEORGI PRECIPITATION TEST 523
then be added but the vigorous shaking must be continued.
The resulting suspension should be very turbid and milky; an
opalescent suspension is an indication that the salt solution was
added too rapidly without sufficient shaking. Such a suspen-
sion should be discarded as it will not give sufficiently sensitive
results.
The antigen should be roughly titrated so that the optimum
amount of precipitation will be obtained. To accomplish this,
the antigen is diluted 1:10, 1:20, 1:40, 1:60, and 1:80, 1.0 ce.
of each dilution being used with 0.3 cc. of a known positive
serum. The dilution giving the optimum result should be used
in the test. ,
Several different antigens have been used with little difference
in results; three different lots of cholesterinized and two lots of
plain alcoholic antigen. As a rule, the antigens giving the best
results in the Wassermann test have served best here, and the
dilution used for the Wassermann test has been proper in the
precipitation test.
The serum. The prime requisite for the serum is that it must
be clear. So far as age of the serum goes, no difference has
been noted. Inactivation at 56°C. for thirty minutes appar-
ently makes no difference in the results, therefore, in most of
the work reported here, this procedure was omitted.
In the test, 0.3 ec. of serum is used, it having been found by
trial to give optimum results. As low as 0.05 ec. gave a trace
of precipitation and as high as 0.5 ec. has been used, but the
amount indicated has been found best.
Time of reaction. The reaction between serum and antigen
probably takes place instantaneously, though no change is appar-
ent until after some hours of standing. If the tubes are whirled
in the centrifuge a few moments, however, they may be read
immediately. For tests in which the centrifuge is not employed,
the mixtures are allowed to remain overnight when all the posi-
tive tests will show a precipitation. Occasionally a tube will
require as long as forty-eight hours for the precipitate to come
down, but after forty-eight hours no change has ever been noted.
524 THOMAS G. HULL AND EVA E. FAUGHT
Effect of temperature. Optimum results are obtained with low
temperatures for incubation. If the tests are heated at 56°C.
for thirty minutes immediately after mixing, no precipitation
takes place on long standing either at room temperature or in
the ice box. Thirty minutes at 37°C. does not retard later pre-
cipitation in the ice box, but no reaction has been obtained when
the tests were allowed to remain at body temperature overnight.
So far no appreciable difference has been noted in overnight
incubation at room temperature and at ice box temperature,
though the latter has usually been used.
THE PRECIPITATION TEST
In setting up the test, to each tube is added 0.3 ce. of clear
serum and 1 cc. of antigen properly diluted. The tubes are
shaken and either centrifuged and read immediately or allowed
to stand overnight in the ice box.
TABLE 1
Comparison of Wassermann and precipitation tests on 296 sera
OGD tests POsitlVe. 4. 6.3. wee hoes eo eee oe Meee oe SOE OE REI ee Renee eee 65
Bothitestsinegative:. 000s 1G base Ae ST Pe an hae Ser 195
Positive precipitation, negative Wassermann......................eeeeeeee 22
Negative precipitation, positive or doubtful Wassermann................ eek
Agreement with the Wassermann test has been obtained in a
large percentage of cases. (Wassermann tests were run with
two antigens; cholesterinized incubated at 37°C. for thirty min-
utes, and alcoholic incubated in ice box over night. The sheep
cell system was used with previous extraction from the patient’s
serum of natural anti-sheep amboceptor.)
Out of 296 samples tested, results were obtained as shown in
table 1.
It will thus be seen that 260 or 88 per cent of the samples agreed
in the two tests; in 7 per cent of the cases, the precipitation test
showed a more delicate reaction, being positive where the Was-
sermann was negative; in 5 per cent a negative precipitation
reaction was obtained with a positive or a doubtful Wassermann.
~
SACHS-GEORGI PRECIPITATION TEST 525
It should be said here that of the 14 such instances, only 6 gave a
definitely positive Wassermann (three or four plus reaction with
both antigens), while 8 gave doubtful reactions (one or two plus
_ reaction with cholesterinized antigen and negative with alcoholic
antigen). About 2 per cent, therefore, of the tests gave a nega-
tive precipitation test with a definitely positive Wassermann.
Of this 2 per cent (six tests), three of the patients had undergone
vigorous treatment while histories of the other three were unob-
tainable. It is apparent, therefore,. that vigorous treatment
affects the result of the precipitation test under certain conditions.
Of the 22 instances in which a positive precipitation test was
obtained with a negative Wassermann, all of these cases, so far
as could be learned, gave a history of syphilis. The information
was taken from the cards submitted to the laboratory with the
blood samples and was not complete in all instances.
RELATION BETWEEN WASSERMANN AND PRECIPITATION TESTS
Whether this precipitation test is identical in the factors in-
volved with the first part of the Wassermann test, has not been
determined. Friedberger (8) claims that the amount of the
actual precipitate is no criterion of the degree of the complement
fixation. In fact the precipitating power of a serum may be
destroyed by moderate heat without the destruction of comple-
ment-fixing antibodies. Dean (9) states that the proportion
of antigen and antibody favorable for rapid and complete pre-
cipitation does not favor the most complete complement fixation.
The two phenomena do not run parallel courses, but they prob-
ably represent different phases of the same phenomenon. Wells
(10) sums up in this fashion:
A favorite interpretation of the Wassermann reaction, which seems
to harmonize with the known facts, is that there is a precipitation
of serum globulin by the lipoidal colloids of the antigen, and adsorp-
tion of the complement by this precipitate.
An attempt was made by the present authors to separate the
precipitate formed in the precipitation tests and to determine
whether it was entirely responsible for the Wassermann reaction.
526 THOMAS G. HULL AND EVA E..FAUGHT
To this end, the precipitation tests were centrifuged at high speed
for an hour, in order that as much as possible, if not all, of the
precipitate might be thrown down. The supernatant fluid was
then pipetted off and the sediment washed three times with
salt solution. Wassermann tests were run on both, using of the
supernatant fluid 0.2 cc. and all that was obtainable of the
sediment, with two units of complement and thirty minutes
incubation in the water bath.
The results are shown in table 2.
TABLE 2.
Result of Wassermann tests run on the supernatant fluid and the sediment of the
precipitation tests
SPECIMEN NUMBER WASSERMANN TEST ON WASSERMANN TEST ON WASSERMANN TEST ON
SERUM SUPERNATANT FLUID SEDIMENT
1 4+ 4+ 4+
63 4+ 4+ 4+
2 4+ 2+ 4+
3 — _ No precipitate
43 — = No precipitate
64 - - No precipitate
The precipitate in each case bound the complement, as was
to be expected. In only one instance, however, was there a
reduction in the strength of the Wassermann test run on the
supernatant fluid. From these tests it is impossible to say
whether the precipitate is entirely responsible for the binding
of the complement or not. Further work needs to be done on
the subject.
SUMMARY
1. A modification of the Sachs-Georgi precipitation test for
syphilis is described, using clear blood serum and an alcoholic
extract of beef heart, either cholesterinized or not.
2. The precipitation test agreed with the Wassermann test in
88 per cent of the cases; in 7 per cent it was more delicate, giving
positive results where the Wassermann was negative; in 3 per
cent it was negative where the Wassermann was doubtful; in 2
SACHS-GEORGI PRECIPITATION TEST 57
per cent it was negative where the Wassermann was positive.
‘Treatment of the patient apparently affects the results of the
precipitation causing it at times to come negative while the Was-
sermann still is positive.
3. Attempts to determine whether the precipitate formed in
this test was entirely responsible for the Wassermann reaction
were unsuccessful.
J REFERENCES
(1) Kuausner: Biochem. Zeitschr., 1912, 47, 36.
(2) Lanes: Berl. klin. Wochenschr., 1912, 69, 897; Zeitsch. f. Chemiotherap.,
1913, 1, 44.
(3) GamMMELTAFT: Deutsche med. Wochenschr., 1912, 38, 1934.
(4) HirscHreLp aND KuincrerR: Deutsche med. Wochenschr., 1914, 40, 1607.
(5) Vernes, A.: Presse Medicale, 1917, 25, 704.
(6) Sacus AND Groret: Munch. med. Wochenschr., 1920, 67, 66.
(7) Gauui-VaLERIo: Corresp. Blatt f. Schweizer Aertze, 1919, 49, 1977; Abstr.
Jour. A. M. A., 1920, 74, 563.
(8) FrinpBERGER: Deutsche med. Wochenschr., 1906, 15.
(9) Dean: Zeitsch. f. Immun., 1912, 13, 84.
(10) Wretts, H. G.: Chemical Pathology, 3d edition, p. 237.
nll vb vii Wega? wit Re ie eal Ge rr yea iynteun as a og
ee OVA. RIE ws) Pinan eee ee Pcie
| ‘ | Soe i he GR Ded Verh A vei ehteris | Ger aD anne wohteald
nl pt ial ms a fo ene per emai at
| ‘a "vay OFS RY mere OnE eh GE OME Lise: y ty ue SB) ar sci ; ‘
+, Re oie a) coer inet: . Ae eee
rei dei ee: Beat.
oe NAAN RED: Piri itt itis .
wt 5 P : ‘ i . aK i ewes ie ‘aie iia eae jstatele
ili” | A daetian Cie i at aac 4) heer
i Ne Oy | A 0
nq a ; nik ORO? A obey ui ios Fie AS ry, dit | 4 *$ ye; a A \ r ;
, ; ale iret ATR. 5 Rit veered ate da Ty Me sate 2 RAKE: Wh ON A eas se: ¥ i Mes
Dy ie at re Pal a KGa ie A phbiaabane ye itn Cee a
rae cane a mint PyAiant tis tf So a yh Wee
| stat Al TOE: OF RIE ds onder debi pad Ah sare cae Roehh
ig | bins i) eee
oe hat Oe eas Bh hohe Bai, fy ;
Ana | a bee Ai aabaey | b eibahdagaiees
ae : oe SRA ay tg teed HE ew ied hy Gah geen elt
¥ ~ 4§
fT tae : .
Lf . % ale » :
Bi ee $
be |
s i
Lin i F AY 7 i) ek bios
. Be
iO 4 1 x) i ~ <4 yh a
. ¥ - ’ p
* 7 x = ve » V 49
ne om et s { ‘ 4 ary ae
oN y : ‘
y i. { aa i iL i i Nea)
i i fae
ae r, ifn Aan
Ate sip | ae ia ea weg
' u ¥ Z |
Ma | if 7 in
j - 7) % 5
we j rr. 7 =
is | -¢ ous
Tee , ‘1 )
ay 7 rt i oT
er i j ind 4
a a
a er >
> un
Bite ete Ey me as ere trey ieee eat) 2 a ‘adie a jt
ae ay ‘ a ‘ at i) CS ee oth ray is “a he Cf pe : i
a wea + } ‘
en ¢ ees wet Bs Wait ¥ (a ee ma | may
> ah
a whee res ve } rah LY. DA aay ae
a ie nc oe ‘pug Veet ah Se) Sa eae
a 4 PNCTREN I at ere UME
ee il ¥ me .
‘ 4 i
{
“ ;
ar) ey aaa
A STUDY OF THE MECHANISM OF HUMAN
ISOHEMAGGLUTINATION!
HERBERT L. KOECKERT
From the Department of Pathology, School of Medicine, Western Reserve University,
Cleveland, Ohio
Received for publication August 18, 1920
The mechanism of isohemagglutination has been interpreted
in different ways by the various investigators of this phenomenon.
The assumption of differences of affinity as explanatory of the
purely chemical nature of such antigen-antibody reactions, as
Ehrlich hypothecated, has given way to the widely accepted
principle that these bodies are of colloid nature and their inter-
actions analogous to those of colloids in general. On the other
hand, Gay (1) has offered the explanation that these reactions are
governed by physicochemical laws; but this conception has not
been supported by sufficiently convincing proof to warrant its
acceptance.
Bordet (2) showed that bacterial agglutinins can be absorbed
and specific agglutinins isolated by thismethod. The absorption
of hemagglutinins was demonstrated by Malkoff (3), and later by
Hektoen (4); since then, this has been observed repeatedly by
others. Shibayama (5) and Ottenberg (6) showed that agglutinable
cells will absorb more agglutinin than is necessary to agglutinate
them, and the latter demonstrated this quantitatively by show-
ing that one volume of agglutinable cells will absorb all the agglu-
tinin from 16 volumes of agglutinative serum, whereas complete
agglutination occurred with 13 volumes of serum. Healso pointed
out that when the ratio of serum to cells is 4 volumes to 6 vol-
umes all the agglutinin is absorbed without producing agglutina-
1 Read before the twentieth annual meeting of the American Association of
Pathologists and Bacteriologists, New York, April 2, 1920.
529
530 HERBERT L. KOECKERT
tion, the explanation being that each of the cells absorbed some
agglutinin but not enough to bring about clumping. —
By means of similar absorption experiments various investiga-
tors have attempted the classification of the bloods of humans and
animals into groups according to their isoagglutination reactions.
The efforts with human bloods have been successful, but no clas-
sifications of the bloods of other mammals have been definitely
demonstrated. On the basis of these classifications assumptions
have been made regarding the number of normal isohemaggluti-
nins in human bloods and their specificity for particular agglu-
tininogens or receptors of cells of particular groups. ‘The spec-
ificity of normal human ischemagglutinins has been generally
accepted, but has been denied by Landsteiner and Sturli (7) and
others, and Karsner and Koeckert (8) recently pointed out that the
specificity of these bodies in normal serum may be lost during
the desiccation of such serums.
Isoagglutination of human erythrocytes was discovered inde-
pendently by Landsteiner (9) and Shattock (10) in 1900. Land-
steiner’s classification provided for three groups, and he, followed
by Descatello and Sturli (11), postulated two agglutinins and two
agglutinophilic substances, or receptors, according to the following
distribution:
Group I. Serum contains agglutinins A and B; cells possess no
receptors. ;
Group II. Serum contains agglutinin A; cells possess receptor b.
Group III. Serum contains agglutinin B; cells possess receptor a.
Hektoen, in 1907 observed from the results of his absorption
experiments, that
Cells I do not absorb agglutinin and therefore possess no receptors.
Cells II do not absorb the agglutinin of serum II for cells III, but
do absorb the agglutinin of serum I for cells II and III.
Cells III do not absorb the agglutinin of serum III for cells I, but
do absorb the agglutinin of serum I for cells II and III.
Cells II and III therefore possess distinct receptors and the failure
of agglutination by some serums is due to the absence of suitable
receptors.
HUMAN ISOHEMAGGLUTINATION 531
He assumed the presence of three main agglutinins—the agglu-
tinin in serum I for cells II and III, the agglutinin in serum II
for cells III, and the agglutinin in serum III for cells I]. By the
independent discovery by Jansky and Moss (12), a fourth group,
whose serum contains no agglutinins but whose cells are agglu-
tinated by all the other serums, was added. This phenomenon
was also observed by Hektoen but not recognized by him as the
basis of a separate group. Moss made the following classifica-
tion, which has been widely adopted: -
Group I. Serum is non-agglutinative; cells are agglutinated by
serums II, III and IV.
Group II. Serum agglutinates cells I and ITI; cells are agglutinated
by serums III and IV.
Group ITI. Serum agglutinates cells I and II; cells are agglutinated
by serums II and IV.
Group IV. Serum agglutinates cells I, II, and III; cells are non-
agglutinable.
In his explanation of these reactions, Moss assumed the pres-
ence of three different isoagglutinins and three isoagglutinophilic
substances, as follows:
Group I. Serum contains no agglutinins; cells possess receptors
a, b, and c.
Group II. Serum contains agglutinin A; cells possess receptors b
and c.
Group III. Serum contains agglutinin B; cells possess receptors a
and ec.
Group IV. Serum contains agglutinins A, B and C; cells possess no
receptors.
The contradiction between the hypotheses of Landsteiner and
of Moss, as well as the questions raised by our experiments
quoted above, determined the further investigation of the prob-
lem by the method of differential absorption. Throughout the
remainder of this article the following classification, that of Jan-
sky, will be adhered to:
THE JOURNAL OF IMMUNOLOGY, VOL. Vv, NO. 6
532 HERBERT L. KOECKERT
Group I. Serum agglutinates cells II, HI and IV; cells are not
agglutinable.
Group II. Serum agglutinates cells III and IV; cells are agglu-
tinated by serums I and III.
Group III. Serum agglutinates cells II and IV; cells are agglu-
tinated by serums I and II.
Group IV. Serum agglutinates no cells; cells are agglutinated by
serums I, IJ and III.
Or graphically,
SERUMS
CORPUSCLES
I Il III Vi
I os 123 a =
II 4 = a Z
UI + =: = 2
IV at + = =
Serums and thrice washed cells of the four groups were ob-
tained, and the proof of their grouping established by testing
with standard hemagglutinative serums of known groups. The
agglutinative titer of the serums for suspensions of agglutinable
cells of equal concentrations and the agglutinability of the
washed cells by standard agglutinative serums were determined in
order to permit accurate observations of the quantitative differ-
ences in the agglutinative power of the serums and the agglutin-
ability of the cells used in the experiments. When cells IV were
added to serum I, titer 1 to 32, in the ratio indicated by Otten-
berg as proper for the absorption of all the agglutinin and com-
plete agglutination of the cells, the mixture incubated at 37°C.
for one hour, and centrifuged, it was seen that the serum had lost
its agglutinative power for any cells and the cells would absorb
no more agglutinin. When these agglutinated cells were washed
with a quantity of saline equal to that of the serum used, the
mixture heated at 50°C. for thirty minutes and centrifuged, the
supernatant saline now possessed the power to agglutinate cells
II, Ilf and IV and behaved exactly as did the weakest potent
dilution of the original serum I—that is, its agglutinative power
HUMAN ISOHEMAGGLUTINATION 533
was very slight; and the sedimented cells IV, after repeated wash-
ing with saline, reacted qualitatively with normal serums of the
four groups exactly as normal cells IV, but quantitatively did
not reabsorb all the agglutinin from an amount of serum I equal
to that used in the original agglutination of these cells. Further,
such cells after repeated washing did not absorb as much agglu-
tinin from agglutinative serums of the other groups as was ab-
sorbed by an equal amount of normal cells TV. It was seen from
this experiment that some, but not all of the absorbed agglutin-
ins could be recovered from the agglutinated cells, and that the
repeatedly washed cells reacted qualitatively exactly as did the
normal cells. In other words, the original serum I was changed
to a serum IV, the non-agglutinative saline reacted as a weak
serum I, and the cells, repeatedly washed after agglutination,
again became group IV cells, qualitatively.
By using cells II instead of cells IV with serum I, it was seen
that the supernatant serum no longer agglutinated normal cells
I, but did agglutinate cells III and IV; the agglutinated cells
would absorb no more agglutinin from normal serums I or III;
the agglutinated cells after heating at 50°C. gave up some of the
absorbed agglutinin to the saline in which they were washed;
the supernatant saline after centrifuging possessed the power to
agglutinate cells II and IV, although only slightly; and the sedi-
mented cells after repeated washing reacted qualtitatively as
normal cells II, but failed to do so quantitatively. In this in-
stance the original serum I may be said to have become a serum
U; the supernatant saline reacted as a weak serum III; and the
sedimented cells, after repeated washing, again reacted qualita-
tively as normal cells II. When this serum previously used to
agglutinate cells II was added to cells III or IV, agglutination
and complete absorption of the remaining agglutinins occurred;
the agglutinated cells after heating gave up some of the absorbed
agglutinin as evidenced by the subsequent agglutination of nor-
mal cells III or IV, by the supernatant saline and the reabsorp-
tion of agglutinin from normal agglutinative serums; and the
sedimented cells after repeated washing recovered their original
qualitative reaction to normal agglutinative serums. ‘The results
534 HERBERT L. KOECKERT
in this case were the same as those observed by treating normal
serum II with normal cells III or IV; but by using a serum I
previously treated with cells II, it is seen that, in addition to the
above observations, serum I can be deprived successively of the
agglutinins for cells U1 and III, or vice versa, or its entire agglu-
tinative power may be lost by treatment with cells IV alone;
and further, those agglutinins which agglutinate cells IV also
agglutinate both cells If and III. In other words, serum I may
be altered to react as serum II or III and then as serum IV, or
as serum IV directly, according to the succession of the agglu-
tinable cells used. Therefore, it is obvious that such fractional
isolation of agglutinins from what may be called a polyvalent
serum can be accomplished.
When cells II were treated with serum IIT in the same ratio as
that in the above experiments, it was found that the serum had
lost its agglutinins for any cells and the agglutinated cells would
absorb no more agglutinin when treated with any agglutinative
serum; so that, the combination of antigen and antibody was
complete—no additional receptors remained uncombined, nor did
the serum contain any additional agglutinms. However, when
cells IV were treated with serum III, incubated, centrifuged, and
a quantity of saline equal to that of the serum was added to the
cells and heated, it was found that the cells, although agglutin-
ated, absorbed other agglutinins from serums I or II; the serum
IIT became non-agglutinative for any cells; the supernatant saline
caused slight agglutination of normal cells II and IV; and the
cells after repeated washing reacted qualitatively as normal
IV cells. Before washing, however, the cells [V used in this —
instance reacted as normal cells IIT after treatment with serum
IfI. When serum II was added to these agglutinated cells, the
cells absorbed more agglutinin from that serum, and failed to
absorb any additional agglutinin from serum I. These original
cells IV were thus altered successively to react as cells III and
then as cells I. Similar results were obtained by using serum
II first, and then serum III; but when the cells were treated with
serum I they did not absorb any further agglutinin from serums
Il or III. Therefore, it is also obvious that fractional saturation
HUMAN ISOHEMAGGLUTINATION 535
of receptors of what may be called polyvalent cells can be accom-
plished in a manner similar to the fractional absorption of agglu-
tinins from polyvalent serums.
In order to prove the correctness or incorrectness of Moss’
assumption, experiments were done, based upon this principle
of fractional absorption of agglutinins and fractional saturation
of receptors. Serum I was treated with cells II, incubated, cen-
trifuged, and the serum recovered. According to Moss’ hypo-
thesis, this serum should be free from agglutinins B and C, and
it was found that it no longer agglutinated cells IJ but did agglu-
tinate cells III and IV. An equal volume of the same original
serum I was treated with cells III, which, by the same reasoning,
should then be free from agglutinins A and C. This serum was
recovered after incubation and shown to have lost the agglutinin
for cells II]. The two recovered serums, one containing agglu-
tinin A, and the other, agglutinin B, were then mixed and treated
with cells IV. If these cells IV possessed receptors a, b, and e,
the last would remain uncombined until further treated with a
serum said to contain agglutinin C, namely, serum I. However,
it was found after such treatment that no additional agglutinin
had been absorbed from normal serum I which, after recovery,
reacted qualitatively and quantitatively with equal amounts of the
same suspensions of cells II, III and IV, used in the original
determination of the titer of the serum, exactly as before. ‘There-
fore, cells IV do not possess a receptor ec which remained uncom-
bined, and serum I does not contain an agglutinin C which could
have combined with such an agglutininogen.
The same results were obtained when cells IV were treated suc-
cessively with serums II and III, and then with serum I. After
treatment with serum II, the cells absorbed agglutinin from se-
rums I and III only; but after subsequent treatment with serum
III, the cells no longer absorbed agglutinin from serum I by vir-
tue of an uncombined agglutininogen c of the cells and an agglu-
tinin C in the serum. Similar experiments showed that cells II
and III also do not possess more than one receptor.
That two agglutinins operate to produce the complete combi-
nation of receptors of cells of group IV can be shown quantita-
tively by the following experiment. ‘The titer of serum I was
536 HERBERT L. KOECKERT
determined for equal amounts of equally concentrated suspen-
sions of cells II, II] and IV and found to be 1:32, 1:32, and 1:64,
respectively. When cells II were added to serum I in quanti-
ties sufficient to remove all the agglutinins for cells of that group,
thus rendering that serum innocuous for the further agglutination
of normal cells II, it was. found that the titer of the recovered
serum for cells [V was reduced to that of the original serum I for
cells III, but the titer for cells LI] was unaltered. The same ob-
servation was made when cells III were used instead of cells II.
This observation was also made in the preceding experiment
(see page 535) in which one volume of serum I was treated with
cells II and another volume with cells III, the serums recovered
and mixed. ‘The titer of each portion for cells [V was lower than
that of the original serum I, but after combination of bothpor-
tions the titer of the mixture for cells IV was equal to that of
the original serum I.
CONCLUSIONS
1. By what may be designated the fractional absorption of
agglutinins from so-called polyvalent serums and the fractional
saturation or combination of receptors of polyvalent cells, it can
be shown that there are two distinct normal human isohemagglu-
tinins and two agglutininogens, which operate to produce the
group distribution of human bloods.
2. These agglutinins can be isolated by the method of frac-
tional absorption.
3. Isohemagglutinins may be recovered from agglutinated cor-
puscles, but not completely.
4. By empirically designating the human isohemagglutinins A
and B, and the agglutininogens or cell receptors a and b, their
distribution may be conveniently charted as follows:
GROUP
I II III IV
SeTUM ee See eee AB A B oO
Cell sewerage eres fe) b a ab
5. The agglutinins in fresh normal serums are specific for par-
ticular agglutininogens or receptors.
HUMAN ISOHEMAGGLUTINATION ES
REFERENCES ’
(1) Gay: Jour. Med. Res., 1908, 17, 321.
(2) Borpret: Ann. de |’ Inst. Pasteur, 1901, 318.
(3) Matxorr: Deutsche Med. Wochenschr., 1901, 915.
(4) Hextorn: Jour. Inf. Dis., 1907, 4, 297.
(5) SarpayaMA: Centralbl. f. Bakt., 1902, 760.
(6) OrrenBERG: Jour. Exp. Med., 1911, 13, 925.
(7) LANsSTEINER & StuRLI: Wien. klin. Wochenschr., 1902.
(8) Karsner & Korckert: Jour. Am. Med. Assoc., 1919, 73, 1207.
(9) LANDSTEINER: Zeitschr. f. Med., 1902, 3.
(10) SHarrock: Jour. Path. and Bact., Feb., 1900.
(11) DescaTELLo & Sturt: Miinch. Med. Wochenschr., 1902, 49, 1090. +
(12) Moss: Bull. Johns Hopkins Hosp., 1910, 21, 63.
ed ie ae rane
. ‘ \ Srl i NaNO Pe t wig
Wo OT ee ae oT Seal ahhh
Mei iy te i Vey CLANS Re yh Tay ya v4 - hi ve i hae *. mI iy
vat uA Bar ese ; et h » ANA AY ree bite, NG ha ress)
Wy aver ¥,
ere ci ey Pa bute af He Sit
High r yh a } Ay ae Veep :
Ue Bs this F Prat #U
‘
‘ y Witten d ; / ee
+ , ee
Bs) aes aad ‘nice a
ae Se ey
Bee Pane in jeter Von ay ean
is imader i RRR USN OM MT aes tea .
‘ ae : A sey " 7 ‘al i vee 1) my yal ve
8, Pe Pe RAT ty jae at Te Se, ne
robes nea) ftps ‘ i 2 Cie Wadeni
: Sia |
Y Saar es «
& ,
: ’ : Pe oy? Nz
(Sti itis" * eee ht te, age ay : bites an ony . Bri
beg ais nplaledne i, 7 aie “LORE cur ate i ee Lif ag
. a tae ti! pont al ee cheated Saee
TF, PTT. y ih ann, tart cae en trae diaseton + hd ey ate
Laty Healy, LLP PG TIO TER ae Ne! et “nets ER
wrest! “sy itt gh af ings fOr, a
eins oh a x 4
‘ af 0) i oa a
his eae HCO iat:
AT Be. «| «.' ae ede aT tat RE aes Ms Villa erie,
wy vsanabona chambers cially bes
bie Wo Ob: ‘shat nae eat
vida oH ae
STH yt % Reerere
Baby stat Bits ii
rf 7 r n we 7 -r ie 7 f
‘tah Wie she ae nee waar er eH on ROR gS AL a
’ 1 cs
‘i? oi Sl? ae
1
7 Y Aj ous ane ae Lat ~ Wi ied me gee '
Ra 1 DV Shek, Seah TN Sa Saar ai oat
2 NOY ) o o o
) ° oo ° mS oO
res OH Sx = NX N » + ~
So o So o —) [—) q q i=]
ABSORBING DOSE...... 4 2i|2 0 sil 8 il gi} 3 il Sm Sm = ao
n n n n n
22/22 |22 ai |sela8| of oe oe
Sa Sa Sa Sa eS sie eSee oe 62
—Q foal joa faa) fQ ea] a) jee} faa)
Qn n n n n n n 2 a
=| 5 =) | cs} 3 =| 3 3
Q n nm n n n n hu 2
6 rs} rs) 6 3 rs} } ro} rs)
™ reels alte s ie | icles 7h oD rey oD a oO
EST ANTIGENS........ o8 £ > E
Pigs | Ble Biles| B les] Bi s3| Brest) 2 $3 ts 83 acy $3
Ae SaaS | OS | eS lo] So See Ss ce | OS ee
Bul Bil Galea el penile obec vear mes es Reelin toes. |, eel eae
Mimimimiai(a) a laialsigig| ma | ml ml} am} a] A
Serum dilutions:
1:3 +\t |—|—|—|—|4a a f4)—|-]-|++) + |++]4+1 [44/41
1:6 +|—|—|—|-|—|41 |—|-]—|-|-|++) + |++]+1 |++] +
1:12 Vp feo Pee es fs) ee fe Se Reese pes er) =
1:24 a Fn FP an FF fa fe
1:48 —|/—j—J—|—|/—] — J—-j-|-—|-|J-| 1] - 1/-);-j/-
ABSORPTION OF ANTI-BACTERIAL PRECIPITINS 557
TABLE 10
Agglutination tests—Anti-paratyphosus B serum. Serum absorbed by bacterial
mass,* and unabsorbed serum control
ABSORBING TYPE...........0.-000+- B. paratyphosus B* | 8B. ee ot sper heerbnce
TEST ANTIGENS.........2..-000+ a { Poa Alege pperat qe 33 Biperat =| 8888
Serum dilutions:
1:50 a _ +1 - +--- a ai
1:100 = = 5 al os = SRE age
1:200 = = Leia z = aeqe q55r
1:400 ' - — ara = SS qeae
1:800 — _ S55 = 5335 Sa
1:1600 = = Seac = S555 a =a
1:3200 _ = Sets = a a Sa =
1:6400 - - aril - Sat aq
1:12,800 “= ~ +1 — +1 -
1:25,600 _ — + -- > +
1:51,200 - oe _ = 1 _
Control _ _ _ _ _
* To one volume of packed bacteria was added three volumes of serum diluted
1:15, the mixture incubated at 37°C., shaking frequently, ice chest overnight and
centrifuged. The clear supernatant fluid used for further dilutions.
TABLE Il
Precipitin tests—Antiparatyphosus B serum. Effect of dilution of precipitin anti-
gens on the degree of cross or group reactions
tina detearieeic «ce cioee “Antiformin”’ Broth
B.
STITCIMNA SN Olinciicicc sfacicseclcns oss B. paratyphosus B B. suis 333 phigsus 333
TION sets cnc sta 0 1;10 1:50 0 1:10 1:50 0 0
Serum dilutions:
1:3 Tecbenitte oes 1 +1/ + i a+ 1
1:6 Simatal = = 1 Sm + - 1 1
L2 +1 1 a 1 = 1 =
1:24 + il _— os _ = 1 ma
1:48 1 _- _ _ — _ — —
In answer to this we would say that the degree of difference
elicited in these tests (before absorption), even where most
marked, is not appreciably greater than that we have previously
558 CHARLES KRUMWIEDE AND GEORGIA M. COOPER
encountered with two bacteria which were identical, except as
regards their sensitiveness to antibody action. Furthermore, as
these differences might be accentuated by inequalities in the
antigens, it is evident that the direct reaction could not be
accepted as conclusive unless it showed a more marked differ-
entiation.
The second question would be in relation to the dosages
recorded. Would the moderate differences in the absorption
dose result in a loss of specificity with agglutinin absorption
similar to the apparent loss of specificity with the precipitin
absorptions recorded in table 9?
On the basis of a very extensive experience with the agglutinin
absorption method, we feel justified in answering in the negative.
For instance, if a dose of 0.1 cc. of a heterologous bacterium
absorbed the group agglutinins for itself and because of very
close relationship with the serum strain reduced the titer some-
what for the latter, double or ‘triple this absorption dose would
only result in some further reduction, not necessarily very marked.
In no instance, thus far, even with excessively large doses, have
we observed a complete removal of the specific agglutinin from
a highly potent serum by the use of a heterologous strain.’
In other words, the agglutinin-absorption technic leaves us a
practical and relatively wide working range as regards dosage.
The precipitin absorption method on the contrary, as the experi-
ments indicate, has a narrow range and may fail to reveal distinct
differences, especially where strains are closely related, even with
scrupulous regard as to the size of the absorption dose.
We have already referred to our previous failure to obtain
more specific precipitation results by dilution of the antigens of
B. typhosus, B. sanguinarium and B. pullorum. As these results
might not hold true with other closely related bacteria, it seemed
desirable, as a further control, to determine this point for other
3In this connection it is worthy of note that some workers have utilized a
method for agglutinin absorption which does not include testing below 50 per
cent or 25 per cent of the original titer of the serum. This inadequate method
has even been employed to separate “‘antigenic varieties’’ of a definite type of
bacterium, such as B. typhosus, for instance. Such reports have been given a
prominence which critical scrutiny does not warrant.
-—
ABSORPTION OF ANTI-BACTERIAL PRECIPITINS 559
types. The supernatant fluid of centrifuged broth cultures was
also used as an added control antigen on the specificity of the
heated antigens.
Table 11 gives the results of this experiment. ‘There is little
suggestion of greater specificity in the dilute antigens, with the
exception, possibly, of the clarified broth antigen which in itself
is a dilute antigen. Even if we exclude the probable influence
of differences in concentration in the case of the broth antigens,
it is evident that these antigens would be of little value in carry-
ing out the absorption technic because the initial differences
before absorption are very slight, and founded on the readings
of the minimal observable reaction. A very small absorbing
dose would suffice to reduce this reaction beyond the limit of
certainty of observability. In other words, dilution reduces the
degree of reaction obtained with the antigens and leaves us no
working range for absorption.
The results of precipitin tests made fier absorption of anti-
meningococcus serums type I and type II (Gordon) are given
in tables 12, 13 and 14.
The results with type I serum are sharply specific with the
absorbing dose employed. The results with type Il serum are
TABLE 12
Precipitin tests—Antimeningococcus serum, type I.* Serum 1 cc. absorbed by
bacterial mass, and unabsorbed serum control
Salt solution
ABSORBING TYPE............-.- I II Ill contol
ABSORBING DOSE..............- 15 billiont 15 billion 15 billion _
TEST ANTIGENS..............-.- I II | 1 I TE | VOR jal [ot Oi a 18 hy yes 6 8 |
SS es es ee — — SSS |
|
Serum dilutions:
is Th Oe eg ele 504 Ne ade
1:6 seal st seme el ese eet Nero) | Gooey Vn ec = =
1:12 —}/—}]—-)}t+]}]—-/]—-—-]4+/]-/]- 41/2 i
1:24 —/—=—-}]—)}4/—-]- tN | Pe |
1:48 a re ee | =, (ea | a ee
* Similarly specific results were obtained with another set of antigens in spite
of the fact that the type I antigen was decidedly weaker than those of the other
types.
{ Estimated by opacity standards.
560 CHARLES KRUMWIEDE AND GEORGIA M. COOPER
TABLE 13
Precipitin tests—Antimeningococcus serum, type II. Serum 1 cc. absorbed by
bacterial mass and unabsorbed serum control
Salt solution
ABSORBING TYPE........-...--: I II III cdntzsl
ABSORBIMG DOSE...........-.--- 15 billion 15 billion 15 billion —
TEST ANTIGENS*........-------- I JOE | BU posad | EEE Li ea |e P| enn
Serum dilutions:
11233 aso iles= lise || a= 1);4]—-—{]+7441)/+
1:6 ee ee en fe eeemnted | — |) eae eaten at tae
2 — 1};—]|-—- 1}; -—]- 1};-—-/—-—;}+]/-
1:24 Se pees | ee eee eS)
1:48 Ea) Wesnpaok| esc eB mn mi be FE a
* Antigens, Lot no. 1, of which type I is distinctly weaker than the other two.
TABLE 14
Precipitin tests—Antimeningococcus serum, type II. Serum 1 cc. absorbed by
bacterial mass and unabsorbed serum control
ABSORBING TYPE.........0.--.- I Il III Balt eolntion
ABSORBING DOSE............... 20 billion 20 billion 20 billion —
TEST ANTIGENS*...............- I IBEW OG e | ae TES ))) TS ae JOC" P08 Eiken DOGS Henne
Serum dilutions:
1:3 1 1} —- 1 1} — |} +1) +1) -— |4+4+)/4+4+/) +
1:6 — 1|;/-|- 1}/—{]+/+] —-— J4+4+/41 1
112 —/}/—;—-]—-—-f]J-]- 1 1|}—{|+/]+
1:24 —{/—f—-]—-frosyry-}y-f] 1 1] —-
1:48 SS SS a | SS aS ae Sh eS |] aS =
Antigens, Lot 2, of which type II seems somewhat the weaker.
irregular and to some extent contradictory. In one case, table
13, we knew that the type I antigen was less strong than the
other but we used it to see what results would develop. Here
there is an apparent specificity, but obviously this specificity
may be wholly illusory and due to inequalities in the antigen.
It will be noted that the absorption by the homologous strain is
incomplete. This test could not be repeated to complete exhaus-
tion as the supply of these antigens was depleted.
A repetition of the experiment with another set of antigens,
table 14, led to results so nearly, if not quite non-specific, that
further tests were not made.
ABSORPTION OF ANTI-BACTERIAL PRECIPITINS 561
The finished antigen of type II in this series seemed somewhat
less concentrated than the others. This may be a factor in the
apparently non-specific results. This again brings forward the
question as to the influence of the antigen on the results obtained.
The results already recorded depend on the use of similar masses
of culture and their manipulation in a similar way, the end anti-
gens heing assumed therefore to be closely comparable. In
other words, we have at present no practical method by which
we can determine the concentration of precipitable substance
in an antigen and its consequent power to react. One cannot
exclude the possibility that if we had such a criterion, more nearly
specific results might be obtained. We wish to make clear, there-
fore, that the non-specific results presented are limited to our
present available methods. Considering our results as a whole,
we are inclined to believe that even with a satisfactory method
for antigen standardization, non-specific results would still occur
_ where cross-precipitation was marked.
The results obtained parallel the non-specific results occasion-
ally noted with a low titer agglutinating serum, which phenom-
enon may be explained by the assumption that the increased
- agglutination titer is due to an accumulation of group and normal
agglutinins, with only a negligible associated increase in specific
- agglutinins. It would seem, therefore, that the non-specific
results obtained with the precipitin absorptions were likewise
due to a low titer of the serum. The sera employed were not of
a low titer in the usually accepted sense. They were from
animals receiving intensive immunization and compared favor-
ably with the most active precipitating serums we have been
able to produce in the past. We should, therefore, prefer to
-consider the range of action as narrow, not low. The narrowness
of the working range may have influenced the results, but it
should be noted that the methods employed included a possibility
of reduction from 48 to 3 or one-sixteenth of the reacting titer.
A similar ratio of possible agglutinin reduction with an agglu-
tinating serum would yield specific results if the same care were
employed as regards the absorbing dosage.
In this connection, and also to illustrate the statements already
made, we have given the agglutinin absorption results with B.
562 CHARLES KRUMWIEDE AND GEORGIA M. COOPER
paratyphosus B serum in table 10. This table shows the
clear-cut results obtainable even when the cross-agglutinaticn
approximates 100 per cent. These results are not selected and
show the clean cut results obtainable even when the absorbing
dose is estimated and not determined by titration.
CONCLUSIONS
A precipitin absorption method has only a limited application
in the differentiation of closely related types of bacteria which,
because of such relationship, show marked cross-precipitation.
The tendency in such a case is toward non-specific results; that
is, the bacterium, heterologous to the serum, in removing the
precipitins active against itself may also remove the precipitins
active against the homologous type. Where the cross-precipi-
tation is less marked, more specific results tend to appear after
precipitin absorption.
The results are the same whether a precipitin antigen or the
bacteria themselves are used for absorption.
The results may be influenced by our lack of a suitable method
for the standardization of precipitin antigens of bacterial origin.
The conclusions as given, therefore, are stated in terms of this
limitation.
The non-specific results obtained may be referable, in some
degree, to the narrow working range of precipitating sera.
The inadequacy of the precipitin reaction as a primary or
single method for the differentiation of related bacteria, which
are serologically still unclassified, is again emphasized. While
similar failures in differentiation may occur with the agglutina-
tion reaction, such failures may be easily corrected by resorting
to the absorption method.
REFERENCES
(1) Gay, F. P., anp Cuickerine, H. P.: Journ. Exp. Med. 1915, 21, 389.
(2) CuickERING, H. P.: Journ. Exp. Med. 1915, 22, 248.
(4) WetnstTEIN, I.: Journ. Immunol., 1918, 3, 17.
(3) Gay, F. P., anp Stones, R. L.: Journ. Immunol., 1916, 1, 83.
(5) Smrru, ‘T., AND TEN Broeck, (.: Journ. Med. Res., 1915, 31, 503 and 547.
(6) KRUMWIEDE, C., AnD Nosiz, W. C.: Journ. Immunol., 1918, 3, 1.
INDEX TO VOLUME V
Abortion in cattle, infectious, Simplification of the factors involved in the
ReMICINEITU AATEOU TORU LOL... 5. as acn'ee Bebe aisle SoM Eb aileic de Ceviewes 399
Absorption of anti-bacterial precipitins, A study of the specificity of the.. 547
—— of antigen, The relation of the rate of, to the production of immunity.. 39
Agglutination, A dropping bottle as an aid in macroscopic slide............ 155
RE 5 EE OR 2 Co a 465
Agglutinins. I. On the transfer of the so-called normal-antibodies from
ORE SU GPS Oe ee ee Ace ae de RY {
—— and hemolysins for the four groups of human erythrocytes, An attempt
nonmaducesspeeite wamune. . 6.8 ee ol ee es .. 89
Agglutinogens, typhoid, A comparative study of methods for the preparation
Do dontodtoen bot! gle ahh IMS SS RE ae idee Rah ante tant os Bhan 1s A Rema 97
Apetcesiny piackley, Immnninization with.................0..0.ecceeceeescee O89
Allergy, Anaphylaxis and: Hypersentitiveness. . a8 .. 363
Anaphylactie reactions, The relation of certain veins to the, and fie peacne
thereof on the mechanism of anaphylactic shock. Bed os. 230
— shock, The relation of certain drugs to the an: palwiacue ee aa
fPeipearing tuereor on the mechanism of... 2... s. ci. cnaee ees csecesaee 230
Anaphylaxis and allergy: Hypersensitiveness ..... 363
——, Studies in. I. On the quantitative reaction of asker ee eed
SE PME ITLO ONG ID. VIVO'p2 icc. .ceruic «feu ad 56 2 ha aye twie steleie wih alojeinye 297
—, Studies in. The relation of certain drugs to the anaphylactic reaction,
and the bearing thereof on the mechanism of anaphylactic shock...... 230
Anti-bacterial precipitins, A study of the specificity of the absorption of.. 547
Antibodies, A note on the non-specific production of..............-.....-. 517
~—., On the placental transmission of so-called normal. Bo NI ... 391
—, On the transfer of the so-called normal-, from ay Fes to Ey, . 227
—, On the placental transmission of so-called normal. sik FAS . 455
body, from the globin of an immunised animal, The eae onectne
of globin, with a note on the independence of the properties of serum
and tissue proteins, as exemplified by the absence of . Sc ton spe
Antigen, The relation of the rate of absorption of, to the Goadieban. of
SMA. 7, 21) Nee wand SER NRE FSR a aie VR Ske Wa a's 39
Antigenic properties, Effect of ultraviolet rays on.................0- eee 345
—— properties of globin, The, with a note on the independence of the prop-
erties of serum and tissue proteins, as exemplified by the absence of
antibody from the globin of an immunised animal..................... 417
ST RODETIICHIOMMEMOCYANIN, LNG. 5... ae cage cee we nen case cece secccnees 259
Antihuman hemolysin, Experiments upon the production of, with special
reference to immunization with erythrocytes sensitized with heated
FCA ee Na MIS E> ira sna R GENS atclom cahe Rioed Spee aeS: vias’ crsleSe dalle «jk -aiv'a's 507
— hemolysins, Natural, and hemagglutinins in horse sera in relation to
SECU TOE OM. 7) oa a i Ro 75
564 INDEX
AM GRL BIRR s elect tele = fo ake ootey shot otal ste) Neots ee 455
Antitryptic-acting bodies. «0.2.2.0. 2cee ere eect eee ce eo seeinnese ss aameeeee 391
Arnstein, Natalie, Fleisher, Moyer S., and Hall, T. G. Serological relation-
ships, of liver and kidney... eee c ce cence ee pecs ween eee eee 437
Asthma, The relationof sputum bacteria to. ...........-.-. +. esse eee e eee 373
Autogenous B. coli vaccines on the intestinal colon bacilli of dogs, An experi-
menval study. Of the elect 2. .7prteacee ee vee «o> steers oa vee 133
Bacterial toxaemia, On the nature of... ..... 00.0... eee eee eee meee cssees 265
Bactericidal immune sera, On the so-called Neisser-Wechsberg inhibiting
GOHSMOMI|MON 1105 22 fo, 2 op cst. at aloe age eels oR @ .- 2 es nieiei ee nergeh 1
B. coli vaccines, autogenous, An experimental study of the effect of, on the
intestinal colon bacilli of dogs'..: ai. <\tenem meric.» » > +, sets oeuserers sete pains 133
Blackleg aggressin, Immunization with............-..+-2- eee eee eee 539
Blood, Studies on the meningococcidal activity of...........-..-.--++.+++- 51
Browning, C. H., and Wilson, G. Haswell. The antigenic properties of
globin, with a note on the independence of the properties of serum and
tissue proteins, as exemplified by the absence of antibody from the
globin of an immunised animal...............-.- sss eeeece cece cer eceee 417
B. typhosus antigen, A study of different methods for the preparation of.... 111
Cattle, infectious abortion in, Simplification and partial revision of the
factors involved in the complement fixation test for.................. 399
Cholera antigen, sensitized, Experimental study of the.................... 145
= immunity, A serological stindiy Of 2). fc... a/ nee eee cies! = <)c\e 2 piel rree ee 465
Chronic gonorrhoea in women, Comparison of smear, culture and comple-
TOW ERATION UM cic iets th coud bok bee Scale beep ecehats ele eke eaeenpleteieie« Sisk sete tary 499
Coca, Arthur F. Hypersensitiveness: Anaphylaxis and allergy............ 345
—— and Kosakai, Mitsuji. Studies in anaphylaxis. I. On the quanti-
tative reaction of partially neutralized precipitin in vitro and in vivo.. 297
Colon bacilli, intestinal, of dogs, An experimental study of the effect of autog-
enous ‘Bb. colt. vaccmes: on the. 2... 2. .2eeaseas tne. «one een er 133
Complement fixation test for infectious abortion in cattle, Simplification
and partial revision of the factors involved in the...............-.-+-- 399
= fixation test, ihe, for tuberculosis © .Fo aro cms see oie eee 159
Complements of different animals, Some observations on the constitution
Pela 1 CV ot ae RE gaa te SONY Aa ay EAI SS 3 5 Sea IONS 379
Constitution of the complements of different animals, Some observations
OTT H LY by AAR es Ue PO er or cE RN LS Seo oo ios AEROS Oe & 379
Cook, Marjorie Ww. The relation of the rate of absorption of antigen to the
production of immunity . eR EAS aor. - SI SS 2 = = 39
Cooper, Georgia M., and [oem meals. Charles. A study of the specificity
of the absorption of anti-bacterial precipitins..............--.-.----- 547
Dogs, An experimental study of the effect of autogenous B. coli vaccines on
theAnbestinal colon bacillivol Ns. a5 . eee = 2 < ios o e 133
Dropping bottle, A, as an aid in macroscopic slide agglutination... ........ 155
Dunham, George C. A nephelometrie method of estimating ‘ihe number
Of OLEAMISMS IMA VACCINES . J )rs x nicr Oe eies> «+ 12 oo reve lente 337
INDEX 565
Eberson, Frederick. Effect of ultraviolet rays on antigenic properties. I.
mnrerMIEHIpCOCCUS 2. io... 2.0) bo es ase clea ow hiele nts see slewssclee 345
Emmons, R. V. B., and Mason, Edward H. ‘The value of the intra-palpebral
mille unsthe dinguosis of glanders. 12...) cceee eee s cece ce ceceeee 489
Empyema, pneumococcus, A study of the precipitin test in cases of....... 321
Erythrocytes, human, An attempt to produce specific immune agglutinins
Pepeermelvaris fOr ime) four PTOUDH' OF 2.) 22 i... do cen vive oe een ec eee seen. 89
— sensitized with heated serum, Experiments upon the production of
antihuman hemolysin with special reference to immunization with... .. 507
Faught, Eva E., and Hull, Thomas G. The Sachs-Georgi precipitation
test for Sahilis . ei nt aie ee Pe 5 521
Fleisher, Moyer S., Hall, T. G., and Arnstein, Natalie. Serological relation-
ships of liver and Pees . Su jy. beat Sate hoe Wega ieke ae GS «F< ss 437
Floyd, Cleveland. A study of the precipitin test in cases of pneumococcus
oR Pa RE SEA NGU REP PREM aoc. 5 v's we 3a ayaUGr Nu, ate fa Dad rated >» aidlnie 3 we «ai eld vie = 2 9' 321
Gibbs, Charles S., and Rettger, Leo F. Simplification and partial revision
of the factors involved in the complement fixation test for infectious
So oceha 2 fic io. 6 eee ee Sele eee as merece 399
Glanders, The value of the intrapalpebral mallein in the diagnosis of...... 489
Globin, The antigenic properties of, with a note on the independence of the
properties of serum and tissue proteins, as exemplified by the absence
of antibody from the globin of an immunised animal. ry ari’ ace Ae
Gonorrhoea in women, chronic, Comparison of smear, ealuure a panne
2) elec, [nth MO ae eee ean tern earner eerie re 499
Hall, T. G., Fleisher, Moyer S., and Arnstein, Natalie. Serological relation-
PEMUEIICTEAMOUIONOY «2.5 2 50 nie oe vc re ess et eens talvegescs woheee 437
Haslam, Thos. P. Immunization with blackleg aggressin...............-. 539
Hemagglutinins in horse sera in relation to serum therapy, Natural anti-
PPM IRPMRTE eT SISTA an 2 2. ws diese eon ene selesinosle eae ccas ses sersus te
Hemocyanm, Lhe antigenic properties of... .........-e.+seee eee e eee n eee 259
Hemolysins for the four groups of human erythrocytes, An attempt to pro-
@ice specific immune agglutinins and...................5-+.2-2- eee eee 89
—, Natural antihuman, and hemagglutinins in horse sera in relation to
MINTER ct. be crsiviacisiee sinless areas eve gamma ie areas enews + 8 75
Horse sera in relation to serum therapy, Natural antihuman hemolysins and
hemagglutinins in. Adri wmecaye ar ones ns ot bet cee oo 2 Bini hh > cee 75
Hull, Thomas G., Anh Faught, Eva E. The Sachs-Georgi precipitation
test for ee «rab limpet Var ha OME a SS hE oe ean ta el lr 521
Human erythrocytes, An attempt to produce specific immune agglutinins
suuenemongemmtor the four groups Of ...... 2... -. 5 cece eee eee 89
— isohemagglutination, A study of the mechanism of.................-- 529
Immune agglutinins and hemolysins, An attempt to produce specific, for the
faur groups of human, erythrocytes. « . <2 nth lide eee lene cece eee eee: 89
—— sera, bactericidal, On the so-called Neisser-Wechsberg inhibiting phe-
(2 Lae TO PS Or ee a cee ee er ee 1
566 . INDEX
Immunity, cholera, A serological study of... ............- cee secescseecse 465
, production of, The relation of the rate of absorption of antigen to the.. 39
Immunization with blackleg ‘aggressin .: 2 jacedied .ia\o5 o. bale de oe eee 539
with erythrocytes sensitized with heated serum, Experiments upon the
production of antihuman hemolysin with special reference to.......... 507
Infectious abortion in cattle, Simplification and partial revision of the
factors involved in the complement fixation test for.................. 399
Inhibiting phenomenon in bactericidal immune sera, On the so-called Neisser-
Wiecheberg é. iclesiaen Lusk Sarde bee SE ISG eo «cree cee 1
Intra-palpebral mallein in the diagnosis of glanders, The value of the...... 489
Isohemagglutination, human, A study of the mechanism of................ 529
Kidney, Serological relationships of liver and....................00..e005- 437
Koeckert, Herbert L. A study of the mechanism of human isohemagglu-
UTD EAE LORD SRN Arey a desl Ow Shekiba whe ta Bake Pade Uae rane thls Siz 66 Gi aes oe ee 529
Kolmer, John A., and Matsumoto, Motomatsu. Natural antihuman hemol-
ysins and hemagglutinins in horse sera in relation to serum therapy.... 75
——,and Trist, Mary E. An attempt to produce specific immune agglutinins
ae Pomorie for the four groups of human erythrocytes....... 89
Kosakai, Mitsuji, and Coca, Arthur F. Studies in anaphylaxis. L On! fhe
quantitative reaction of partially neutralized precipitin in vitro and in
VV Ol cece erect ie one eiciagevels re ite 6 fe: eueteve se, cds tenegen amnpeMeMe ea eye tt asi oes tae te eee 297
Kermode: ‘Charles. A dropping bottle as an aid in macroscopic slide
Cyurea LUAU 0 (0) | ep De ri RRR reo SM REI 6 © SISA aR NIN ois 2 155
— and Cooper, Geureia M. A study of the specificity of the absorption
Of anti-bacterial precipitins)icc.ectce codeine orem eae te tiene eee 547
Liver and kidney, Serological relationships of. ....................steeecees 437
Mackie, T. J. Some observations on the constitution of the complements
of different animals 5.2) Ye fae. Pid ones 1 ORE eee 379
Macroscopic slide agglutination, A dropping bottle as an aid in............ 155
Mallein, intra-palpebral, The value of, in the diagnosis of glanders........ 489
Mason, Edward H., and Emmons, R. V. B. The value of intra-palpebral
mallem)an. the diagnosis of glandergiew) 4. 4.7m. osieck es eeeeeainie sae © 489
Matsumoto, Motomatsu. A note on the non-specific production of anti-
| 51616 b=) OE PRR On EM Sa Rn Cpe eA Ae. 2 SOM fe 517
—. Astudy of different methods for the preparation of B. typhosus antigen 111
——. Experiments upon the production of antihuman hemolysin with spe-
cial reference to immunization with erythrocytes sensitized with human
SOLUTE oh ocd canis dak rele tamer she pepe cicero ca Le Ree ols > 1a. cach aera 507
—— and Kolmer, John A. Natural antihuman hemolysins and in ge
tinins in horse sera in relation to serum therapy.... . aie, Seek
Matsunami, Toitsu. Studies on the meningococcidal none oe Bioal® SI
Mechanism of anaphylactic shock, The relation of certain drugs to the ana-
phylactic reaction, and the bearing thereof on the................... 280
Meningococcidal activity of blood, Studies on the...................+-05. 51
Meningococeus; Studies On &.iis:+ o 2'y avs @U Rp AOR EON Cs Cos toe aye ee Ee 345
INDEX 567
Mice and rabbits, The protective value of pneumococcus vaccination in.... 429
Miura, Y. Experimental study of the sensitized cholera antigen.......... 145
Natural antihuman hemolysins and hemagglutinins in horse sera in relation
PEMALMIEREC ESTING fel. vic) s/s Siisirvh i ada made SAME Anes eke s «cde ces ode 75
Neisser-W echsberg inhibiting phenomenon, On the so-called, in bactericidal
7 UTI T EAE, Oly GG Saar cee Pe Re eee i a 1
Nephelometric method of estimating the number of organisms in a vaccine, A 337
Neutralized precipitin, partially, On the quantitative reaction of, in vitro
and in vivo. Sig ee OMe Re cea Rice atc art Cee re 297
Normal antibodies, On the placental transmission of so-called......... 391, 455
——-antibodies, On the transfer of the so-called from mother to offspring. . 227
Placental transmission of so-called normal antibodies, On the........ 391, 455
Pneumococcus empyema, A study of the precipitin test in cases of.......... 321
— vaccination in mice and rabbits, The protective value of. Oe een Ao
Precipitation test for syphilis, The Sachs-Georgi.........................2 521
Precipitin, partially neutralized, On the quantitative reaction of, in vitro
and in vivo. 2 0 ¢ OD GEERT CEE OEE oon Sane) Some eee Sew ee ena 8 297
— test, A study ae fhe, in cases of pneumococcus empyema.............. 321
Precipitins, pafeameterial, A study of the specificity of the absorption of.. 547
Production of immunity, The relation of the rate of absorption of antigen
Gl MEE- 02s 22+ abo ee ee 39
Quantitative reaction of partially neutralized precipitin in vitro and in
ReMANO 2S. 2 bis ban eets ate aw he a whrrewatreesits Sebo vceaele tot 297
Rabbits, The protective value of pneumococcus vaccination in mice and.. 429
Rackemann, Francis M. The relation of sputum bacteria to asthma........ 373
Rahe, Alfred H., and Torrey, John C. An experimental study of the effect
of autogenous B. coli vaccines on the intestinal colon bacilli of dogs... 133
Rate of absorption of antigen, The relation of, to the production of immunity 39
Rettger, Leo F., and Gibbs, Charles S. Simplification and partial revision of
the factors involved in the complement fixation test for infectious abor-
IRM aT CE tS OES Sig ck de ewe OE pee baa weg cased wee eee BIDS o 399
Reyman, G. C. On the placental transmission of so-called normal anti-
Pree erAnitroryptic-acting DOGIES. 2... 6. ee ce cee eects seeecees 391
—. On the placental transmission of so-called normal antibodies. III.
oo Edn Oe Geek A eee een ene 455
—. On the transfer of the so-called normal-antibodies from mother to
Pp CIUTINING. . .... . cc ales con ace meaclncas dees ccessdecceces 227
Sachs-Georgi precipitation test for syphilis, The....................0.4.. 521
Sands, Joseph E. A comparative study of methods for the preparation of
Sar REE STIPE TRY TUERETIEESTAS 9. 5, :c) «- 5 6's du 5: 4's & ergs tale vn erie MWe wrnpn'ee wi o's 97
Schmidt, Carl L. A. The antigenic properties of hemocyanin............. 259
Sensitized cholera antigen, Experimental study of the.................... 145
Sera, bactericidal immune, On the so-called Neisser-Wechsberg inhibiting
PIRTESYIPALNECSTT COTISITI NM PENN a oe 5 cvc.0 s areture wiarercsieta nscale Rime ae patatg eat eral sus ins =p Anint 1
568 INDEX
Serological relationships of liver and kidney. .........2.0..0. 005 ceeneeueee 437
study. of cholera immunity. A; bo cics.c hia aaakee Latte tiaie ne cee 465
Serum therapy, Natural antihuman hemolysins and hemagglutinins in horse
sera invrelation: tox’. 22-885 CUS yeeceen Woe eek. Ste, Ee ey ee 75
Smith, James D., and Wilson, M. A. Comparison of smear, culture and-
complement fixation in women. A preliminary report................ 499
Smith, Maurice I. Studies in anaphylaxis. The relation of certain drugs
to the anaphylactic reaction, and the bearing thereof on the mechanism
oflanaphy lactic shockiy: 22) teie sian ae ates. Ae: dina chats Sie aoe 230
Sputum, bacteria, The relation of, to asthma... %......-.....-25--0 es ven ae 373
Studies in anaphylaxis. I. On the quantitative reaction of partially neu-
tralized:precipitin’ in‘vitro: and! in-VivOrr. O65... 6t es eek ee ete te 297
— in anaphylaxis. The relation of certain drugs to the anaphylactic
reaction, and the bearing thereof on the mechanism of anaphylactic shock 230
Syphilis, The Sachs-Georgi precipitation test for...................----60- 521
Thjgtta, Th. On the so-called Pra 8 inhibiting phenomenon
in bactericidal immune sera. OL Sa: hes Fg EAE hee (es 1
Torrey, John C., and Rahe, Niered! H. UM experimental study of the effect
of pitoneneus B. coli vaccines on the intestinal colon bacilli of dogs.... 133
Vexaemia} ‘bacterial,.@n the mature ofs).1. he ee. alls olen 2 2 ele eee 265
Transfer of the so-called normal- antibodies from mother to offspring, On the 227
Transmission of so-called normal antibodies, On the placental . wits erp eos
Trist, Jary E., and Kolmer, John A. An attempt to produce Epecisen immune
Recline and pomoieee for the four groups of human erythrocytes.. 89
Tuberculosis, The complement fixation test for...................eeeeeees 159
Typhoid agglutinogens, A comparative study of methods for the preparation
Of S.:..
Ultraviolet rays, Effect of, on antigenic properties...................++++ 345
Umemura, Rokuro. A serological study of choleraimmunity. I. Agglutinin 465
Vaccination, pneumococcus, The protective value of in mice and rabbits.. 429
Vaccine, A nephelometric method of estimating the number of organisms in a 337
von Wedel, Hassow O. The complement fixation test for tuberculosis.... 159
Wadsworth, Augustus B. The protective value of pneumococcus vaccina-
Hlonwin, MiICe sand TADDIGS ws .<.8t his: ccs pais tins ee edo ee EE rie eo 429
Wilson, G. Haswell, and Browning, C. H. The antigenic properties of
globin, with a note on the independence of the properties of serum and
tissue proteins, as exemplified by the absence of antibody from the
red Kolounay Cory Pu dbontaniduoNls(eyel CobbOnMly Gy gong ecacedcdacobucidcodpemedobvduoce: 417
Wilson, M. A., and Smith, James D. Comparison of smear, culture and
complement fixation in chronic gonorrhoea in women. A preliminary
TRS) 0X01 POS CTS EUR E SEMI an Gna eno outed CDA Gm OemGeas Goud Oe OSs eSo 499
Women, chronic gonorrhoea in, Comparison of smear, culture and comple-
ment fixation in
Zinsser, Hans. On the nature of bacterial toxaemia...................... 265
CR The Journal of immunology
PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET
UNIVERSITY OF TORONTO LIBRARY
Te eat Orewa ee Pe .,
oe CEN Oe eee
ane
a et: TN a MN
er Artal
fol pags ted
A ye Eat y tte La" Ae tr
pad oy -verentatt