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. <A 
few examples of quantitative macroscopical and microscopical 
tests are shown in table 5; contrary to expectations the mac- 
roscopical tests frequently gave higher readings than the micro- 
scopical tests, the differences being largely ascribed to the time 
allowed for agglutination, being but fifteen minutes for the latter 
and many hours for the macroscopical tests. As shown later 
fresh horse sera contain larger amounts of agglutinins for human 
cells, but the results shown in table 5 indicate the range of 
agglutination with sera several weeks or months of age and 
preserved with tricresol at fluctuating temperatures. 


RESULTS OF HEMOLYSIN TESTS 


Tests for natural antihuman hemolysins in normal and immune 
horse sera were conducted by mixing, in small test tubes, 0.1 ce. 
serum, 1 cc. of 1 per cent suspension of washed corpuscles and 1 
ec. of a 1: 20 dilution of the hemolysin-free sera of several guinea- 
pigs for complement; incubation was conducted in a water bath 
at 38°C. for one hour and the results read the following day. 

The results observed with 12 unheated sera tested with the 
erythrocytes of 9 persons are shown in table 6; inasmuch as the 
anticomplementary activity of these sera may have inhibited 
hemolysis, additional tests were conducted with 27 sera pre- 
viously heated at 56°C. for one-half hour, with the corpuscles 
of 12 persons; the results are given in table 7. 

As shown in these tables ordinary normal and immune horse 
sera prepared for administration to persons, are practically free 
of hemolysin for buman erythrocytes. 

Fresh horse sera, however, may contain small amounts of 
natural antihuman hemolysins as shown in table 8; this table 
shows the results of macroscopical agglutination tests with 3 
fresh normal horse sera in final dilution of 1:11 with the cor- 
puscles of 24 different persons and hemolysin tests conducted as 


ee 


a 


} 


TABLE 6 


Results of hemolysin tests with unheated horse sera and human erythrocytes 


HUMAN ERYTHROCYTES, 1 cc. OF 1 PER CENT SUSPENSION 


HORSE SERUM 
UNHEATED 0.1 cc. 


Antistreptococcus. .|N.H.*| N.H. | N.H. | N.H. 
Antistreptococcus. .| N.H. | N.H. | N.H. | N.H. 
Antistreptococcus. .| N.H. | N.H. | N.H. | N.H. 
Antistreptococcus. .| N.H. | N.H. | N.H. | N.H. 
Antistreptococcus. .| N.H. | N.H. | N.H. | N.H. 


Antipneumococceus..| N.H. | N.H. | N.H. | N.H. 
Antistreptococcus. .| N.H. | N.H. | N.H. | N.H. 
Antipneumococcus..| N.H. | N.H. | N.H. | N.H. 
Antistreptococcus. .| N.H. | N.H. | N.H. | N.H. 
Antistreptococcus. .| N.H. | N.H. | N.H. | N.H. 


* Complement furnished by mixed guinea-pig sera in dose of 1 cc. of 1:20 
dilution (= 0.05 cc. undiluted serum). N.H. = no hemolysis. 


TABLE 7 


Results of hemolysin tests with heated horse sera and human erythrocytes 


HUMAN ERYTHROCYTES 


INACTIVATED HORSE SERA 


Antistreptococcus....... 
Antistreptococcus...... k 


N 

N 

N 
PCM es ks .,.(c aia... |, N 
Antipneumococcus...... N 
Antistreptococcus....... N 
Antipneumococcus......| N 
Antistreptococcus....... N 
Antistreptococcus....... N 
Antipneumococcus......| N 
Antipneumococcus......| N 
Antipneumococcus......| N 
Antipneumococcus......| N 
Antipneumococcus......| N 
Anti-influenzal..........| N 
Anti-influenzal..........| N 
Anti-influenzal..........| N 
Anti-influenzal..........}| N 
Anti-influenzal..........| N 
Antimeningococcus...... N 
Antimeningococcus...... N 
Antitetanus.............} N 
Antitetanus........,....| N 
Antistreptococcus....... N 
Antistreptococcus....... N 
Antistreptococcus...... SON 
Antistreptococcus....... N 


Ly Ay yA A Ay Ay Ae Ly Ly pee yy eae on ia A 
PEE SED TAL LG LED ABS BN A A A A A A A A 
AABDZAAZAAAZANRANANAANAMNAMAARMAARMAAmMABMAAmMsS 


AZAAAZAAZAYN™NAN™NANANANAAAAANAN™NAMAPMAmMmAmwAmAwme 


AZAAAAZAAZAZAZAANPM™MA™AMAMAANANRMAmAAAmwAwswmy 


AABZAAAAAAAAAAAARAAAAMAPAPAABrPApPDy 


ZAAAZAYN™*A™AAAAN™NA™PRA™AAMAMAMASAAAAmAmBmAwmwAy 


AZAAZAZAAZANA™AARANANA™AAS™ANRMAAMAPMAmMmwawr sys 


AZAAAAAZAA™A™AN™NARNANANRMAAAAAMAmAMAmMAmseewy 


AZAAAAAAZA™AANANANAANAAANAANRMmSASA~AmSwt2sawmsys 


ZAAAZAAAAAA™AYAANA™AAMAAMSASMAmApAmAPMwrswwmys [ 


N, No hemolysis; C, hemolysis. 


oo 
— 


82 JOHN A. KOLMER AND MOTOMATSU MATSUMOTO 


described above, complement being furnished in constant dose of 
1 ec. of a 1:20 dilution of the hemolysin-free sera of several 
guinea-pigs. 

As shown in table 8, a fresh serum may contain hemolysin 
for the corpuscles of a certain individual but not for all persons; 
this indicates that the natural antihuman hemolysins in horse 
serum exist in groups similar to the group of hemagglutinins and 
analagous to the groups of natural hemolysins and hemagglu- 
tinins in human sera for the corpuscles of the lower animals 
described by Kolmer, Trist and Flick (2). 


TABLE 8 


Results of hemolysin and hemagglutination tests with fresh horse sera and human 
erythrocytes 


HUMAN RED BLOOD CELLS 
HORSE SERA 
5| 6] 7} 8} 9 | 10) 11) 12) 13] 14) 15) 16) 17] 18} 19} 20) 21) 22) 23] 24 


1 | 2] 3] 4 


| 
Fresh normal...{{ bal avlae lar aavlay alan lay larly ava oy] lavlaVN 


Fresh normal —|—-J—| 4) —-| 4] -)-} -]-1- | -|-}-J-}-}-|-!-|-|-|-|- 
"|| NININININININININ|IN|N| NI NI NI NI NININ| NI N| N| NININ 


Fresh normal...{! 54] s{aayl wlan NIN MIN NINN SIN| Sin sts 8 


* — Noagglutination; +, agglutination; S, slight hemolysis; N, no hemolysis- 


These experiments have also shown that in a macroscopical 
test hemolysin may be present and hemagglutinins absent or as 
is more frequently observed, hemagglutinins may be present 
and hemolysins absent. 


NATURE OF HEMAGGLUTININS AND HEMOLYSINS IN HORSE 
SERA FOR ERYTHROCYTES 


Natural hemolysins and hemagglutinins are divided into two 
groups according to resistance to the influence of heating sera 
at 56°C. for thirty minutes; those that are destroyed or masked 
by this exposure are designated as thermolabile in contradistince- 


HEMOLYSINS AND HEMAGGLUTININS IN HORSE SERA 83 


tion to those that are more resistant or thermostabile. The 
majority of natural hemolysins and hemagglutinins in human 
sera for human erythrocytes and the erythrocytes of the lower 
animals have been found to be thermolabile (2); heating sera 
apparently destroys some of these antibodies outright, while 
others become masked (3). In general terms thermolabile hemol- 
ysins and hemagglutinins are also quite susceptible to the de- 
teriorating influence of age, desiccation and strong chemical 
germicides and antiseptics. 

These factors have an important bearing upon the present 
subject, inasmuch as horse immune sera prepared for adminis- 
tration to persons are generally weeks or months old, preserved 


TABLE 9 


The influence of desiccation upon hemagglutinins in three horse sera for the 
erythrocytes of four rabbits 


Da 
& 
w 
q 
= 
4 


SERUM II 


@ 
& 
w 
z 
= 


CONDITION OF SERA fol nN ey ~~ _ Nn G} ~~ Sal nN oD ~ 
aifaelZisiai/al/a/2/2/2/21/8 

Sierras PoxlO od poshemes 7.o-| Ors 

Fresh.................--.} #*] $} +/+) +/+} +/+}]+/4+/])+/4+ 
Dried 5 days............,; +] +t}/4}/4/+/+/+/+/+/-/+]- 
DricG 12 GAYS... 6200 s.55% +})/—|- Ss ei eee Lb — 


*+ agglutination; +, doubtful agglutination; —, no agglutination. 


with various chemical germicides notably tricresol and phenol, 
and are frequently kept at ordinary room temperature for varying 
periods of time, all of which may bring about deterioration of 
any natural hemagglutinins and hemolysins for human erythro- 
cytes present in the fresh serum. 

Experiments have shown that the hemagglutinins and hemoly- 
sins in horse sera for human erythrocytes are thermolabile and 
thermostabile; the hemolysins are even more susceptible to heat 
than the hemagglutinins and are almost entirely thermolabile. 

Of interest in this connection is the influence of desiccation 
upon the natural hemagglutinins in horse sera; Karsner and 
Koeckert (4) and Kolmer (5) have shown that drying human 
sera upon cover glasses may result in deterioration of isohemag- 


S4 JOHN A. KOLMER AND MOTOMATSU MATSUMOTO 


glutinins and we have found the same true of the agglutinins 
in horse sera for human and rabbit corpuscles. In table 9 are 
shown the results of tests conducted with rabbit erythrocytes; 
some of the hemagglutinins deteriorated more rapidly than others. 
The tests were conducted with cover glasses bearing one loopful 
of serum dried at room temperature; the corpuscles were used 
in 1 per cent suspensions, a loopful being rubbed up in the dried 
serum on each cover glass and examined microscopically fifteen 
minutes later. 

Additional experiments were conducted with rabbit corpuscles 
and the fresh sera of three normal horses, portions of which were 
kept in a refrigerator at 0° to 2°C. with and without tricresol 
(0.25 per cent), at room temperature (about 20°C.) and in an 
incubator (38°C.). Preliminary microscopical agglutination tests 
showed that these sera contain agglutinins for a mixed antigen of 
erythrocytes from several rabbits in final dilutions as high as 
1:80 to 1: 100, and the influence of temperature and 0.25 per — 
cent tricresol upon natural hemagglutinins in horse serum, was 
studied with tests in which mixed suspensions of cells from two 
or more rabbits were employed. 

The results of these experiments summarized in tables 10, 

11 and 12 have shown that tricresol itself in 0.25 per cent had no 
influence upon the hemagglutinins, neither aiding in their de- 
struction nor materially protecting against deterioration; tem- 
perature, however, had a marked influence inasmuch as sera kept 
in the incubator and at room temperature showed marked 
deterioration of hemagglutinins at the end of one week (titers 
reduced to about 1:20), still more at the end of two weeks 
(titers about 1:8) and almost complete destruction at the end 
of three months (titers about 1:2). Sera kept in a refrigerator 
at 0° to 2°C., however, showed much less deterioration, the 
titers being at the end of three months reduced from about 
1: 80 to 1:32; there was little or no difference between sera kept 
at this temperature with and without 0.25 per cent tricresol. 

These experiments have shown, therefore, that natural hemag- 
glutinins and hemolysins in horse sera are quite susceptible to 
deterioration and when immune horse sera preserved with tri- 


HEMOLYSINS AND HEMAGGLUTININS IN HORSE SERA 85 


TABLE 10} 
The deterioration of agglutinins for rabbit erythrocytes in normal horse sera kept 
at 39°C. 
SERUM I | SERUM II SERUM III 


SERA KEPT FOR SS SSE 
No pre- | Tricresol | No Pre- | Tricresol | N° Pre | Tricresol 


servative servative servative 
DINE WEEK set ee ok. <.-,2:- |) 2.20 1:20 1 fee 
PENWOLWECKM....-.2-ir- v2. 56 15l 1:8 H 
Fourteen weeks............ H 2 H 


*H, Hemolysis due to bacterial contamination. 


TABLE 11 


The deterioration of agglutinins for rabbit erythrocytes in normal horse sera kept at 
room temperature 


SERUM I SERUM II SERUM III 
SERA EEPT FOR N N ea an 
iexvative Tricresol pervades Tricresol dervotive Tricresol 
One week............ Re ees 1:40 | 1:40 | 1:20 | 1:20 H* 
PIWVORWEEKS 15 c.5 0.0 006 05 56 5s None 1:20 H RAG H 
Fourteen weeks............ None 1:4 H 1:8 H 


*H, hemolysis due to bacterial contamination. 


TABLE 12 
The deterioration of agglutinins for rabbit erythrocytes in normal horse sera kept at 
0 to 2°C. 
SERUM I SERUM II SERUM III 


SERA KEPT FOR SS al ae Se 
No pre- | Tricresol | NO Pte | Tricresol | N° P¥e~ | Tricresol 


servative servative servative 
MER WEEK Soc cicsic cise cares 1:80 1:40 1:60 1:40 1:100/ 1:60 
MeWwORWEGKS. 8. 2555.65... 1:40 1:80 1:80 1:80 1:60 1:80 
Fourteen weeks............ 1:16 £232 1:80 1:22 1:60 1:32 


cresol are kept at room temperature or even in an ordinary 
refrigerator (8° to 16°C.), as they are likely to be under com- 
mercial conditions, these substances may be expected to de- 
teriorate and their clinical significance lessened thereby. 


86 JOHN A. KOLMER AND MOTOMATSU MATSUMOTO 


THE INTRAVENOUS INJECTION OF RABBITS WITH HORSE SERA 
CONTAINING HEMAGGLUTININS AND HEMOLYSINS 


As previously stated fresh horse sera generally contain agglu- 
tinins and hemolysins for rabbit erythrocytes; tables 13 and 14 
give the results of macroscopical tests with three fresh normal 
horse sera and the cells of four different rabbits. In the hemoly- 
sin tests complement was furnished by 1 cc. of 1:20 hemolysin 
from guinea-pig sera. 

TABLE 13 


Agglutinins in normal unheated horse sera for rabbit erythrocytes (macroscopical 
tests) 


HIGHEST POSITIVE DILUTIONS 
SERA 


Rabbit 1 Rabbit 2 Rabbit 3 Rabbit 4 
Horse eres «ccc: 1:40 1:20 Less than 1:20 1:20 
ELOTSOORE Eines 1:40 1:40 Less than 1:20 1:40 
FLOESC Hee nse nucleate 1:40 E20) 1:20 1:80 
TABLE 14 


Hemolysins in normal unheated horse sera for rabbit erytrhocytes 


HEMOLYSIS 
SERA ees 
Rabbit 1 Rabbit 2 Rabbit 3 Rabbit 4 


ELOTSO Meera ee nace S.H.* S.H N.H M.H 
EFOTSO eee cn ese en eens S.H M.H S.H C.H 
ELOISE ee ec he eines M.H S.H S.H C.H 


C.H., Complete hemolysis; M.H., marked hemolysis; §.H., slight hemolysis; 
N.H., no hemolysis. 


Rabbits injected intravenously with normal horse sera contain- 
ing agglutinins for their cells in macroscopical tests in dilutions 
as high as 1: 40, also hemolysins, showed no ill effects; 10 ce. of 
serum per kilogram of weight and even as much as 20 cc. per 
kilogram injected intravenously rather rapidly (about 5 cc. per 
minute) produced no discernible immediate or late reactions, as 
respiratory disturbances, hematuria and hemoglobinuria. 

Inasmuch as commercial immune horse sera contain compara- 
tively lesser amounts of agglutinins for human erythrocytes and 


HEMOLYSINS AND HEMAGGLUTININS IN HORSE SERA 87 


practically no hemolysins at all it is very doubtful if their ad- 
ministration to persons even in large amounts intravenously, 
is followed by sufficient agglutination and hemolysis in vivo 
to produce discernible symptoms directly referable to these 
phenomena. 

CONCLUSIONS 


1. Practically all horse sera contain agglutinins for human 
erythrocytes; these hemagglutinins occur in groups. 

2. The amounts of agglutinins for human erythrocytes in horse 
sera are highest in fresh sera; in ordinary horse immune sera 
preserved with tricresol at fluctuating temperatures, the amounts 
of hemagglutinins are relative small, about 50 per cent of such 
sera containing agglutinins in dilution 1:2 to 1:11 and only 
about 20 per cent agglutinating 1: 20 or higher. 

3. Fresh horse sera may contain natural antihuman hemoly- 
sins, but these are found only occasionally in older and preserved 
sera. 

4, Agglutinins in horse sera for human erythrocytes are ther- 
molabile and thermostabile; the hemolysins are almost entirely 
of the thermolabile variety. 

5. Hemagglutinins in normal and immune horse sera de- 
teriorate rapidly and to a large extent in sera kept at body and 
ordinary room temperatures; when kept at a low temperature 
deterioration is slower and less marked. Desiccation tends to 
destroy these natural hemagglutinins. 

6. Horse sera contain larger amounts of agglutinins and 
hemolysins for rabbit than for human erythrocytes; these occur 
in groups. 

7. The intravenous injection of large amounts of horse sera 
containing agglutinins and hemolysins for rabbit erythroctyes 
into rabbits, produced no ill effects and no symptoms referable 
to intravascular agglutination or hemolysis. 

8. It is highly probably that the intravenous injection of pre- 
served horse serum does not introduce sufficient agglutinin and 
hemolysin for human erythrocytes to produce ill effects referable 
to intravascular agglutination and hemolysis; this is particularly 
true if the serum is diluted and slowly injected. 


88 JOHN A. KOLMER AND MOTOMATSU MATSUMOTO 


REFERENCES 


(1) Witirams, H. U., anp Parrerson, H. A.: The agglutination of human red 
blood corpuscles by horse serum. Jour. Amer. Med. Assoc., 1918, 70, 
1754-1755. 

(2) Kotmmr, J. A., Trist, M. E., anp Fuicx, A.M.:A study of the natural ther- 
molabile and thermostabile hemolysins and hemagglutinins in human 
sera in relation to the Wassermann reaction. Amer. Jour. Syphilis, 
1920, 4, 111-135. 

(3) Koumer, J. A.: The nature of thermolabile hemolysins. Jour. Immunology 
(in press). 

(4) Karsner, H. T., anp Korcxsrrt, H. L.: Jour. Amer. Med. Assoc., 1919, 73, 
1207-1210. 

(5) Koutmmr, J. A.: The influence of desiccation on human normal isohemag- 
glutinins. Jour. Amer. Med. Assoc., 1919, 73, 1459-1460. 


AN ATTEMPT TO PRODUCE SPECIFIC IMMUNE AG- 
GLUTININS AND HEMOLYSINS FOR THE FOUR 
GROUPS OF HUMAN ERYTHROCYTES 


JOHN A. KOLMER anp MARY E. TRIST 


From the Dermatological Research Laboratories of Philadelphia and the McManes 
Laboratory of Experimental Pathology of the University of 
Pennsylvania 


Received for publication February 9, 1920 


Antibodies occurring naturally or normally in sera are seldom 
if ever as resistant to deleterious influences as immune anti- 
bodies, those produced by the body cells during active immuni- 
zation; accordingly, the natural agglutinins and hemolysins in 
human sera for the erythrocytes of persons and of the lower 
animals are as a group thermolabile or heat sensitive and like- 
wise highly susceptible to the deterioration of age and desicca- 
tion (1). Some of the natural hemolysins and hemagglutinins 
and particularly those for sheep corpuscles are more resistant 
than others, but the isoagglutinins and isohemolysins, which are 
of particular interest in relation to blood transfusion and the 
grouping of bloods, are quite sensitive. As a general rule the 
natural hemolysins are more susceptible to destruction by heat, 
age and desiccation than the agglutinins and in human serum 
containing an agglutinin and hemolysin for a particular group of 
corpuscles, the hemolysin disappears before the agglutinin. 

Karsner and Koeckert (2) have reported deterioration of iso- 
agglutinins in sera dried on cover glasses after Sanford’s method 
and have made the interesting observation that group specificity 
is lost in dried sera kept for several weeks; we have noted with 
this method considerable reduction in the isoagglutinins in sera 
dried on cover glasses which may amount to a total loss with 
some sera and particularly during the first few days after drying. 
Isoagglutininins persisting in the dried sera over the first one 

89 


90 JOHN A. KOLMER AND MARY BE. TRIST 


to four days were found to keep fairly well over a period of 
several weeks when preserved according to Sanford’s directions; 
accordingly only picked sera, that is, those containing more 
isoagglutinins than the average serum, should be chosen for 
drying on cover glasses and these should be tested a few days 
later to determine whether or not the agglutinins have escaped 
destruction. Probably a better practice is to add a small 
amount of preservative to a serum (0.2 per cent tricresol) and 
keep it at a low temperature, frozen if possible, to preserve the 
isoagglutinins at least, if not the isohemolysins as well. 


PURPOSE OF INVESTIGATION 


Inasmuch as it has been shown that immunization of the 
lower animals with type I pneumococci results in the production 
of antibodies largely or entirely specific for the type and that 
immunization with one of the types of meningococci also results 
in the production of a highly monovalent serum, we have con- 
sidered the possibility of producing agglutinins and hemolysins 
specific for each of the four groups of human erythrocytes (Moss 
classification) by immunizing rabbits with group I, II, III and 
IV corpuscles; we had no reasons for hoping for the production 
of absolutely specific sera and expected the production of smaller 
amounts of group agglutinins and hemolysins but surmised that 
these may be removed by absorption, the specific antibodies 
alone remaining in diminished amounts but yet in greater de- 
gree than found in human sera and more resistant to deleterious 
influences, as are immune antibodies generally. The purpose of 
this investigation was to put this proposition to trial; the technic 
employed and the results observed are briefly stated. 


TECHNIC 


Classification of erythrocytes. Human erythrocytes were se- 
cured from specimens of blood submitted for the Wassermann 
test, the sera being first removed and a small amount of saline 
added to the coagula with gentle breaking up and the release of 
corpuscles. With this method the erythrocytes of twenty or 


IMMUNE AGGLUTININS AND HEMOLYSINS 91 


more persons were grouped at one time and not only furnished 
an abundance of corpuscles satisfactory for immunizing pur- 
poses, but sera as well for further groupings. 

The groupings were made after Sanford’s modification of the 
Moss classification (3) by a microscopic reaction in which group 
IT and group III sera are employed; the tests were conducted on 
cover glasses by mixing two loopfuls of group II serum with 
one loopful of corpuscles suspension and repeating with group 
III serum on a second cover glass; these preparations and a cor- 
puscle control in which saline solution replaced serum, were 
suspended in hanging drop slides and examined after standing 
thirty minutes at room temperature (this period has yielded us 
better results than a ten or fifteen minute period for agglutina- 
tion); the reactions were read after the convenient diagram 
published by Sanford, which may be stated as follows: 


Group II serum + and group III serum + = group I 
corpuscles. 

Group II serum — and group III serum x = group II 
corpuscles. 

Group II serum x and group III serum — = group III 
corpuscles. 

Group II serum — and group III serum — = group IV 
corpuscles. 


Immunization of rabbits. Rabbits were immunized either 
with the corpuscles of one individual or with a mixture of the 
erythrocytes of several persons belonging to the same group; the 
broken up coagula were filtered through cotton to remove fibrin 
and coagula and the corpuscles washed three times with saline 
solution. After the last washing the corpuscles of the four 
groups were kept in formalized saline solution (4) in a refrigerator 
and used as required over a period of several weeks. 

Rabbits were immunized with corpuscles of the four groups 
by injecting intravenously 0.5 cc. of washed packed cells sus- 
pended in 5 cc. of sterile saline solution every five days. 

Agglutination tests. At intervals during the immunization 
each rabbit was bled four days after the previous injection and 


THE JOURNAL OF IMMUNOLOGY, VOL. Y, NO. 2 


92 JOHN A. KOLMER AND MARY E. TRIST 


its serum was tested for agglutinins and hemolysins for groups I, 
II, I11 and IV corpuscles; in conducting these tests a macroscopic 
technic was employed as follows: 

The serum was inactivated at 56°C. for thirty minutes and 
four series of dilutions were made in small test tubes ranging 
from 1:10 upward in amounts of 1 ec.; to each tube of series I 
including a control was added 1 ce. of a one per cent suspension 
of group I corpuscles; to series II was added group II corpuscles 
and so on with groups III and IV corpuscles. The total volume 
in each tube of the four series was 2 cc.; after one hour incuba- 
tion in a water bath at 38°C. the tubes were placed in a re- 
frigerator over night and the readings were made the following 
day with the naked eye. In the tables are given the highest 
dilutions of serum producing definite agglutination; as a general 
rule the readings were quite easy and sharp although microscopic 
tests may have carried the titers somewhat higher. 

Hemolysin tests. These were conducted in exactly the same 
manner except that complement was added to each tube except- 
ing the corpuscle controls, in the form of 0.1 cc. of 1: 14 dilutions 
of the fresh mixed sera of guinea-pigs. In the tables are given the 
smallest amounts of undiluted serum producing complete hemolysis. 


RESULTS 


The results are summarized in tables 1, 2, 3 and 4. 

The results of preliminary experiments conducted with group 
I and group II corpuscles were encouraging and lead us to 
believe that it may be possible to produce specific agglutinins 
and hemolysins for the corpuscles of the four groups (5); the 
results of the first preliminary experiment are shown in table 1 
in which the serum of rabbit 1 immunized with group I corpuscles 
agglutinated group I corpuscles in dilution of 1: 1280 and groups 
II, III and IV in 1: 320, after six injections of corpuscles; the 
results of the second preliminary experiment are shown in table 2 
in which rabbit 1 immunized with the corpuscles of a group II 
person agglutinated group II corpuscles in final dilution of 
1: 640 and groups I, III and IV corpuscles in dilutions of 1: 80, 
after four injections of cells. Why these two animals reacted 


IMMUNE AGGLUTININS AND HEMOLYSINS 93 


as they did is unknown because a larger series of immunizations 
conducted with equal care has not shown similar results, as may 
be seen by glancing over the tables. 

As a general rule the serum of each rabbit contained just a 
little more agglutinin for the erythrocytes of the group used in 
immunization than for the erythrocytes of the other groups; in 
several instances, however, these differences in agglutinin content 
were not apparent. 

It is more difficult properly to interpret the results of the 
hemolysin tests because of variation in resistance of the cor- 
puscles to hemolysis; as a general rule, however, it may be 
stated that the serum of a given rabbit was equally hemolytic 
for the corpuscles of all four groups, although instances are shown 
in the tables when the titer was somewhat higher for the cor- 
puscles of the homologous group, that is, for the corpuscles of 
the group employed in immunization. 

Absorption of the sera removed all agglutinins and hemolysins 
equally; even with sera containing agglutinins 1:1280 for the 
corpuscles of a certain group and 1: 640 for the corpuscles of 
the other group, we have not been able to devise a method of 
absorption that did not remove all or almost all agglutinins and 
hemolysins at the same time. These absorption tests were con- 
ducted by adding to each cubic centimeter of the heated serum 
of a rabbit immunized with group I corpuscles, 0.5 cc. of the 
washed packed cells of each of groups II, III and IV at the same 
time, mixing and placing in a refrigerator over night. On the 
next day the mixtures were placed in a water bath at 38°C. for 
one hour and centrifuged; usually the sera required one or more 
additional absorptions at 38°C. for one hour before microscopic 
tests with undiluted serum showed the absence of all agglu- 
tinins for groups IJ, III and IV corpuscles; unfortunately, no 
agglutinins for group I corpuscles would be found or were present 
in but traces. In absorbing the sera of rabbits immunized with 
group II corpuscles, the corpuscles of groups I, III and IV were 
used and so on with the sera of rabbits immunized with the 
corpuscles of groups III and IV; the results were similar and we 
have not so far been successful in differentiating them by proc- 
esses of absorption, even when one is present in larger amount. 


TRIST 


KOLMER AND MARY E. 


A. 


JOHN 


94 


900°0 0cE + T 900°0 OF9+T $20°0 0r9°T $200 suorzoal Ul § 1095 } Il ‘ON 
Sz10°0 O9T + T 900°0 OdE :T &20°0 OZE-T $200 suolzoaf ul 8 19z;V 
900°0 OF9+T 900°O | O82I:T 900°O | O8ZI-T Sz10°0 suorqoofUl F 1045V OT “ON 
900°0 0cE + T 900°0 0cE - T 900°0 0cE + T 900°0 suorPOOLUT 8 109JV 6 “ON 
$<0°0 Or9-T $20°0 Océ + T $20°0 0F9-T $200 suorqoofUl § 1033V 8 “ON 
gz10°0 Océ + T Sc10°0 OE + T SZ10'O | O8ZI-T GZ10'0 suorzoolul g 1045V LON 
9uou OE: T guou O9T:T d0BI} 0ZE -T a0BI} suolzolul ¢ 1ajJV 9 ‘ON 
900°0 0cE + T SZ10 0 OF9-T 900°O0 | O8I:T $200 suoryolul ¢ 1033 GON 
9081} OSZI:T a0B14 OF9:T o0B1Id OSZI :T a0 B14 suoTzyolUl E 10,JV PON 
$20°0 Or9 + T ¢20°0 O79: T €00°O | O8Z1:T 900°0 suorzooalut ¢ 1045 0 
$0°0 OF9+T ¢0°0 OF9:T GZ0'O | O8Zt:T ¢0°0 suolqoolul g 109JV oN 
€00°0 Or9+T 900°0 OF9-T S100°0 | O8cT-T Sc10°0 suoryotur g 1045V Z ‘ON 
¢T00°0 O9T + T ST00°0 OE + T ST00°0 OE + T ¢100°0 suoryoalul F 1095 
ST00°0 08: T ST00°0 08: T G20 0 Or9 -T ST00°0 suoToofUl F 104FV TON 
SUISA[OUIEH | SUTUTYN[ZZy| sursA[oMapy | SULUTyNIAsy | suIsA[OUTO;T suruTynssy SUISA[OULOTY | SULUIZNIZSYy 
—— ie oe | ee oe ee NOILVZINOWWI SLIGd Va 


saT0soduod AI dnaouo sa19sNduoo Ill daouD sa10soduoo II dnouo Sa T0snduod I doouo 


o_O 


sajosnd.oo Jy dnowb yzum paziunwwur 


s719qQv42 fo Dias ay) pun AT puv [TI ‘TI ‘I sdnosb fo sajzosndioo yzwm pajyonpuod sjsaz urshjoway pun uoyourynpbbouay fo szjnsaw oY J, 


é WITAV.L 
ee a eee 
900°0 O09: T 900°0 000T = T c10'0 OOOT: T 900°0 | O89ZT -T SuorzooLUl ¢ 10}JV @ ‘ON 
Z10'0 028 + T ¢c0'0 Océ + T Gc0'0 0ZE: T Z10'0 OF9=T suoryoofur F 19}5V 
g0°0 O9T + T G00 OZE = T ¢0°0 0ZE + T 00 OF9:T SUOTPIOLUL F 10}JV GON 
$z0'0 0ZE + T Gz0°0 OZE + T $200 OZE + T Z10'0 O8ZT : T suorzoolul g 104FV 1 ‘ON 
gZ0°0 028 + T GZ0°0 0ZE + T $z0'0 OG: T ¢z0'0 OF9-T suolzoaf UT G 104FV 


sulsfjouoeyy | sururyn[say | sursAjouepy | Sururyn[#ay | SursAjowoy | Sururyn[zzy SUISA[OWOH | SUIUTIBNIAIY 


NOILVZINOWWI SLIGA Va 
SaI0soduod AI doOuD Sa TOsoOduod Ill daoup saTosoduod II daouop Sa T0sNOduod I doouo 


sajosnd..09 J dnowb yy pazvunwur sziqqns 


fo vias ay) puo Ay pun qT ‘II ‘I sdnosb fo sajzosndso9 yp payonpuoo sjsaq urshjowoy pup uoynuynibbowoy fo s]]nset oy 


Tl @TaVvo 


IMMUNE AGGLUTININS AND HEMOLYSINS 


900°0 OF9:T €00°0 OF9 = T €00°0 OF9 = T €00°0 O8Z1 + T sesOp J 19}SV ¢ ‘ON 
Gz100°0 OF9=T £00°0 0ZE + T €00°0 028: T €00°0 0dE + T 8880p 9 10}FV Pb ON 
Sz0'0 OF9:T gz0°0 OZE + T g0°0 OF9-T g0°0 OF9 = T 8980p 9 194 V € “ON 
g0°0 OF9=T 9981} O9T:T g0'0 028: T o9B1} 028 = T 8980p F 104 V 6 ON 


£00°0 O83: T €00°0 O8ZT + T $20'0 OF9 = T GZ10'0 O8Z1 + T suoljoaful TT 104FV 
Sz10'0 O8Z1 + T 900°0 O8Z1 + T GZ0'0 "| O8%I:T SZ10°0 O8Z1 + T suolpolul § 10}SV 


Gz10 0 O83 + T 900°0 OF9+T $0°0 O8Z1 + T SZ10 0 O8ZI:T suolyooful g 1azFV LON 
g0°0 O8ZI + T ¢20°0 OSI: T g0°0 OF9 + T ¢20°0 OF9:T suoryolur 2 1043V 
S0°0 OF9+T S0°0 OVO: T S0°0 OF9 + T S0°0 OF9:T suolpoofur ¢ 10}FV 
c0'0 OF9: T BI} 028: T d0B1} OFO: T a0B1} Oro: T suolqoolul g 1904S V 
sursAjomoyy | Surarjpnssy | sursA[Ouyy] | SUTUTZN[DZy | sSUISA[OUWOF] | SULUTIZNIZSy | suIsA[OWOT] | SUlUIyN[AsYy 
a a eee | Ges Sas it ay as. 0 aa SY | Pal seLe ee le alee NOILVZINOWWI SLIdavu 


SaTosoduoo AI ddouo 8a10shduod III dnoup Sa@T0soduoo Il dnouo 8aT10soduod I daoup 


sazosnd.09 AT dnosb yyw paziunwus s.19qns 
fo vias ay) pun AT puv TIT ‘TI ‘J sdnos6 fo sajasndioo yyw payanpuod s}sa7 urshjzoway pun uoyouynpbboway fo sjjnsas ay, 


+ AWIAVL 
czt0'0 | oF9:T | St00';0 | oF ‘Tt | Szt0'O | OF9:T £00'0 | OF9:T | suoroofur F 104;V 9 ‘ON 
ezo'0 | OF9:T | Szt0'O | oze:T ¢z0'0 | Oze:T ez0'0 | Oze:1 |  suoroofut p 103;V ¢ ‘ON 
900°0 | OFO:T 900°0 | O8@I:T 900°0 | O8ZI:T 900°0 | O82: |  suoroofur g 194zy ON 
euou 08:1 euou O9T:T auou 08:1 euou 08:1 suoljoolut ¢ 104JV @ "ON 
GZ0'0 | O8ZI:T eoviy | O8ZI'T 6Z0'0 | O8ZI:T Gz0'0 | O8ZI:1 | suoroofur g 104;y Z ‘ON 
<z0'0 | 0ze:T 6z0'0 | 0F9:1 aoeiy | OFT ov | OFOIT | suorpofur g 1093 TON 


SUISA[OWIOY] | SUIUIZN[SZy | SUISA[OMET | SUIUIZNI SAY | suIsA[OWNY | SUIUTyN[SSy | sursAjOUETy | SUIUIZN[ZIWV 
SS Sa er Se || Le ne Se ee = eee ae a oe ee NOILVZINONWI sLIaavu 
Sa10soduod AI doouo S@I0soduoo III dnaouD sa10soduod II dnaouD sa10soduod I dnouD 


sajosnd.409 JIT dnowB yyw pazvunwur s7z.qqos 
fo vias 94} pun AT pun JT ‘TIT ‘7 sdnow6 fo sajasndsoo yzum pazanpuod s3saz urishjoway pun uorynurynpbboway fo sz)nsa2 ay], 


era lav 


96 JOHN A. KOLMER AND MARY E. TRIST 


CONCLUSIONS 


1. The immunization of rabbits with human corpuscles be- 
longing to groups I, II, III and IV does not result in the pro- 
duction of specific agglutinins and hemolysins for the corpuscles 
of the group employed in immunization. These sera, however, 
frequently show slightly more agglutinin and hemolysin for the 
corpuscles of the group used in immunization than for the 
corpuscles of the remaining groups. 

2. Absorption of these immune sera to remove the group 
agglutinins and hemolysins have generally resulted in the 
removal of all agglutinins and hemolysins. 

3. For the grouping of human erythrocytes it does not appear 
possible to prepare specific immune sera; human sera contain- 
ing isoagglutinins and hemolysins must be used and these are 
best preserved in a fluid state at or near the freezing point. 

We beg to acknowledge the assistance rendered by Mr. Joseph 
Sands in the conduct of a part of this investigation. 


REFERENCES 


(1) Koumer, J. A., Trist, M. E., anp Fuicx, A. M.: A study of the natural 
thermolabile and thermostabile hemolysins and hemagglutinins in 
human serum in relation to the Wassermann reaction. Amer. Jour. of 
Syphilis, 1920, 4, 111-135. 

(2) Karsner, H. T., any Korcxert, H. L.: The influence of desiccation on 
human normal isohemagglutinins. Jour. Amer. Med. Assoc., 1919, 
73, 1207-1210. 

(3) Sanrorp, A. H. A modification of the Moss method of determining isohem- 
agglutination groups. Jour. Amer. Med. Assoc., 1918, 70, 1221. 

(4) Koumrr, J. A., AND Brown, C. P.: The red corpuscle suspension for the 
Wassermann reaction. The preservation of red blood cells. Amer. 
Jour. of Syphilis, 1919, 3, 169-193. 

(5) Kotmer, J. A.: The influence of desiccation on human norma! isohemagglu- 
tinins. Jour. Amer. Med. Assoc., 1919, 73, 1459-1460. 


A COMPARATIVE STUDY OF METHODS FOR THE 
PREPARATION OF TYPHOID 
AGGLUTINOGENS 


JOSEPH E. SANDS 


From the McManes Laboratory of Experimental Pathology of the University of 
Pennsylvania 


Received for publication February 10, 1920 


Various methods have been proposed for the preparation of 
antigens for macroscopic agglutination tests, most attention 
having been given those prepared of bacilli of the typhoid- 
paratyphoid group and B. mallei; in the various methods pro- 
posed, broth cultures and suspensions of the microdérganisms 
washed from solid media in saline solution, have been usually 
employed. In some methods, the antigen is heated for sterili- 
zation and in others it is used unheated; various chemicals have 
been employed for sterilization, when heat has not been applied, 
and also for preservation, including formalin (as in the Dryer 
method), phenol and tricresol, in varying amounts. 


PURPOSE OF INVESTIGATION 


A systematic study of the influence of preparation upon the 
qualities of an agglutinogen, with particular reference to the 
influence of heat and the kind and more particularly the amount 
of chemical germicide and preservative employed, does not ap- 
pear to have been made, although it is well known that the 
technic employed modifies the results to be observed in agglu- 
tination tests, aside from variations due to the microérganism 
itself. At the suggestion of Professor Kolmer, agglutinogens of 
a single strain of B. typhosus have been prepared after various 
methods and tested against the sera of rabbits immunized with 
the same culture, to determine the influence of technic of prepa- 

97 


98 JOSEPH FE. SANDS 


ration upon the qualities of agglutinogens of this bacillus, that is, 
regarding spontaneous agglutination and susceptibility to spe- 
cific agglutination by the immune sera. 


TECHNIC EMPLOYED 


Culture. The culture of B. typhosus employed throughout this 
investigation has been used in microscopical agglutination tests 
for many years and found quite susceptible to specific agglu- 
tinins and usually free of spontaneous agglutination. All agglu- 
tinogens were prepared of this one strain, and all rabbits were 
immunized with the same, being injected intravenously at in- 
tervals of five to seven days with increasing doses of living 
bacilli. 

Agglutinogens. Twenty-one agglutinogens were used in this 
study; the first six were prepared from a ninety-six hour culture 
in plain neutral broth as follows: 


No. 1. Heated at 60°C. for two hours and preserved with 0.5 per 
cent phenol. 

No. 2. Heated at 60°C. for two hours; no preservative. 

No. 3. Heated at 60°C. for two hours; preserved with 0.5 percent 
phenol and 10 per cent glycerin. 

No. 4. Not heated; preserved with 0.5 per cent phenol. 

No. 5. Not heated; preserved with 0.5 per cent tricresol. 

No. 6. Not heated; preserved with 1: 2000 mercurophen.! 


Agglutinogens 7 to 16 were prepared from a suspension of 
bacilli secured by growing the strain on plain neutral agar in a 
large series of Blake bottles for forty-eight hours, removing with 
physiological saline solution, shaking the emulsion with beads 
until a uniform suspension was secured and diluting with suf- 
ficient saline until each cubic centimeter contained about 2,000,- 
000,000 bacilli. 


1Mercurophen is the name applied to sodium oxy-mercury-orthro-nitro phenol- 
ate, a new and superior mercurial germicide prepared by Schamberg, Kolmer 
and Raiziss, Jour. Infect. Dis., 1919. 


PREPARATION OF TYPHOID AGGLUTINOGENS 99 


No. 7. Unheated; preserved with 0.1 per cent formalin. 
No. 8. Unheated; preserved with 0.5 per cent formalin. 
No. 9. Unheated; preserved with 1.0 per cent formalin. 
No. 10. Unheated; preserved with 2.0 per cent formalin. 
No. 11. Unheated; preserved with 5.0 per cent formalin. 
No. 12. Unheated; preserved with 1.0 per cent phenol. 
No. 13. Unheated; preserved with 5.0 per cent phenol. 
No. 14. Unheated; preserved with 1.0 per cent tricresol. 
- No. 15. Unheated; preserved with 1: 1000 mercurophen. 
No. 16. Unheated; preserved with 1: 5000 mercurophen. 


Agglutinogens 17 to 21 were freshly prepared as required by 
removing forty-eight hour growths from slants of plain neutral 
agar with the following fluids; these suspensions were treated 
as agglutinogens 7 to 16, except that preservatives were not 
added: 


No. 17. Unheated; suspension in 0.85 per cent sodium chloride in 
distilled water. 

No. 18. Unheated; suspension in 1 per cent sodium chloride in dis- 
tilled water. 

No. 19. Unheated; suspension in 2 per cent sodium chloride in dis- 
tilled water. 

No. 20. Unheated; suspension in 5 per cent sodium chloride in dis- 
tilled water. 

No. 21. Unheated; suspension in distilled water. 


Antigens 1 to 16 were kept in a refrigerator near the freezing 
point and first tested with immune sera about one week after 
preparation and again about one month later; as previously 
stated antigens 17 to 21 were fresly prepared as required. 

Density of agglutinogens. The strain of B. typhosus employed 
cultivated for ninety-six hours in extract broth neutral to phe- 
nolphthalin, produced a culture of just sufficient density for the 
technic employed which consisted in setting up the tests in tubes 
having an internal diameter of 1 cm. and using each agglu- 
tinogen in amount of 1 cc. with 1 cc. of diluted serum. Sus- 
pensions made of cultures grown on solid media and diluted 
so that each cubic centimeter finally contained about 2,000,000,- 


100 JOSEPH E. SANDS 


000 bacilli, proved just about right in the tests employed. 
Denser cultures very much obscured the results and rendered 
the readings difficult and uncertain; thinner cultures were like- 
wise unsatisfactory and one general result of the study was to 
show the technical importance of proper density of cultures and 
the necessity of having this factor satisfactorily adjusted to the 
size of the tubes and total volume of fluid employed. 

Technic. A uniform. macroscopic technic was employed 
throughout; sera were diluted with sterile physiological saline 
solution (0.85 per cent chemically pure sodium chloride in dis- 
tilled water) and used in appropriate test tubes in amounts of 1 
cce.; each antigen was used in dose of 1 cc., thereby doubling 
each dilution. Controls on each antigen were included in each 
experiment. All tubes were incubated at 55°C. for about 
twenty-four hours and readings were made with the naked eye; 
generally readings were made by two persons in order to reduce 
the error due to the personal equation in the interpretation of 
results. 

The sera of five rabbits were employed with each agglutino- 
gen. In the first tests the sera were used after each rabbit had 
received four intravenous injections of the culture; the second 
series of tests were conducted with sera after five injections and the 
third series after eight injections. In each experiment the sera 
were so diluted as to obtain the limits of agglutination with each 
antigen. 


RESULTS 


The results were recorded according to the maximum dilution 
of each serum in which agglutination was detectable with the 
naked eye, namely, the production of flocculi with or without 
some sediment, and for brevity they are recorded graphically in 
charts 1 to 5. These charts present the results observed with 
the serum of each rabbit in two out of three sets of reactions 
with all agglutinogens, and they show the marked influence the 
method of preparation excited upon the results of agglutination 
tests. 


PREPARATION OF TYPHOID AGGLUTINOGENS 101 


Chart T Curves af at nate ion, oe afl The 


~figglutine gens -_ 
ol 2IY TOTS VPWUYIZIFSEIS IIT NEST 20 


=—=—= First Test. 
*=@e=s Second Test- 


Chart & 


/=/0000 
/= ¥b00 
/~ G00 
l= $600 
[-4§00 
l-H000 
l= 3200 
/- 2900 
1° 2400 
l= 2000 
1- 1600 
te 100 
{+1200 
1=1000 
1= £00 
l= 600 


= 
—= 
= 
- 


JOSEPH E. SANDS 


Curves Of daalaiieison Cana Wi 


- Rag| utinog cass 
10 1] led 13 119 161 


First Test - 
Second Test: 


itt The 


PREPARATION OF TYPHOID AGGLUTINOGENS 103 


Chart TL Curves Pa Ragl ulinalien Observed With the 


abt 


ae UP inogens— 
SIU SF 9/011 1219 14/516 /7 1819202 


/-/0000 
J=- F000 
1-6 400 
15600 
/=¥§00 
= 4000 
/- §,200 
/=2§00 
/- 2400 
/- 2000 
/=/600 
/-/400 
Juo/200 
fm/000 | 
/°§$0 
/=600 
4-00 | 
f= 500 


: em 3 = 
—: = oe 


veo Test. 
cond Test. 


ea 


104 JOSEPH E. SANDS 


Chart WW. Curves of Haq lutination Observed With the 
= RmvUnEe Serym at ab sit to. A 

= Rgqlutinogens— 

LA2BHKSGTP A WIN SAISIGISSOITIE SD 


J- Zfo0 
l= £400 
/= &boo 
l= 1/600 
l=/400 
l= /200 
/ = 1000 
/-F00 
1-400 
/- 400 
/- 300 
4- £200 
Y- 100 
eed Hhay 
(=/00 : 
— = Ficet Taot: 
emwe mm = Second Test. 


PREPARATION OF TYPHOID AGGLUTINOGENS 105 


Charl YW, murine oF Agglutination Observed With the 


m™myne, Serum of Kaspt Nev 
ag lutino gene — 


/~/0000 
l- $000 
/=6@400 
[0 3G00 
/=4§00 
/= #000 
(#3260 
/=2F00 
/-2400 
[22000 
/-/@00 
l=/¢00 
{2/200 
/-/900 
/=€00 

{-600 
/=400 
/=300 
/e 200 
/=-/00 


ins Shan 


s Fits? Test. 
> 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. <A sucrose-positive strain was not 
considered serologically related to the vaccine cultures unless it 
was definitely agglutinated at the 1-500 dilution. The object 
of the agglutination tests was to determine whether the sucrose- 
positive B. coli persisting in the intestinal tract in spite of the 
vaccine treatments were serologically related to the strains used 
in the vaccine or were entirely distinct strains which had grown 
up and replaced the originally prominent sucrose-positive B. colt 
types. 

As has already been intimated, both the salicin-positive and 
the salicin-negative representatives of the sucrose-positive B. coli 
were selected in preparing the autogenous vaccine. Agar slant 
cultures of these strains were emulsified in normal saline and 
killed by exposure at 60°C. for one hour; the count was 
determined by the Wright method. The dog inoculations were 
given subcutaneously generally every two or three days for three 
or four treatments and then after an interval of a week or more 
the series of injections were repeated. The dosage in general 
ranged from one billion to eight billion. Except for a fairly 
marked loss of weight the animals seemed to suffer no ill effect 
from the inoculations. 

In connection with each of the four experimental animals 
evidence was produced that treatment with an autogenous 
B. coli vaccine tended to cause a reduction in numbers and, in 
some instances, an apparently almost complete elimination of 
the particular varieties of B. coli used for immunization. This 
elimination, however, was more apparent than real; for subse- 
quent examinations would reveal the presence of the same colon 
types, although it might be in very small numbers. In fact 
there was no evidence that it was possible, even with very large 
dosage of antigen to cause their disappearance permanently and 
completely. 

In the tabulation the results for each dog is given. As may be 
noted, a considerable individual variation was encountered in 
the readiness and degree to which the animals reacted to the 
vaccine. Dogs 2 and 3 responded quite readily, whereas num- 


EFFECT OF AUTOGENOUS B. COLI VACCINES 137 


bers 1 and 4 were more refractory. With dog 1 a prelim- 
inary examination period of five weeks revealed, on the 
average, an approximately equal number of red (sucrose-posi- 
tive) and white (sucrose-negative) B. coli colonies. Following a 
billion dosage of the autogenous vaccine, repeated four times, 
the relative number of red colonies decreased until a week after 
the last injection they were outnumbered by the white colonies, 
9to1. Soon, however, although an increased dosage was given, 
a change occurred in the ratio which would have been misleading 
unless agglutination tests had been carried out simultaneously. 
Apparently the vaccine had lost its effectiveness in that the 
white B. coli colonies were outnumbered by the red on the 
average 1 to 2. That this inference had no basis in fact was 
revealed in an examination of the agglutination results. For, 
whereas, in the previous period 91 per cent of the red colonies 
surviving the vaccine treatment yielded positive agglutinations 
at 1-500 dilution of the anti-serum, now an average of only 26 
per cent agglutinated positively; indicating that serologically 
unrelated strains of sucrose-positive B. coli had grown up and 
replaced the original varieties with which the vaccine was pre- 
pared. Further examinations indicated that this substitution 
was only temporary as at the end of the period of examinations 
of this dog 86 per cent of the sucrose-positive B. coli aggluti- 
nated positively with the specific anti-serum, although on the 
average the white colonies now outnumbered the red five to one. 
This tendency of serologically unrelated strains of sucrose- 
positive B. coli to grow up was noted also with dogs 2 and 3, 
appearing in each instance at about the same stage of the im- 
munization process, but in no case could these substituting 
strains maintain themselves. 

In experiments of this character it is important to carry out a 
long series of preliminary observations to determine the degree 
to which the types of B. coli selected vary in numbers under 
ordinary conditions. A period of six to eight weeks was con- 
sidered adequate for this purpose. After inoculations were 
started about two to three weeks elapsed before definite results 
were noted in the differential B. coli determinations. In dog 3 


138 JOHN CG. TORREY AND A. H. RAHE 


an apparent rapid elimination of sucrose-positive strains occurred 
within a week after the first treatment with ‘the vaccine. It is 
questionable, however, if this change may be ascribed to the 
specific action of the vaccine. As regards dosage it was found 
than an inoculation with about 2 to 4 billion B. colt secured the 
maximum effect, and that nothing was gained by increasing the 
amount above this point. In one experiment a dosage as small 
as 100 million was employed. After five inoculations covering 
a period of three weeks there seemed to be some slight effect, 
but not as definite as following the larger dosage. 

Observations were not continued long enough to. determine 
definitely how enduring the effect of the inoculations might be. 
In the case of one animal, dog 2, the effect was undiminishéd 
after the lapse of ten weeks following the last of a series of six 
inoculations. With dog 3 the effect seemed to have largely 
worn off in seventeen weeks, but during the interval the diet 
had been somewhat changed. Diet, in fact, is the factor of 
prime importance in the determination of the types of bacteria 
pullulating in the intestinal tract, and without an uniform diet 
the results of these experiments would have been indefinite and 
misleading. An experiment with dog 4 showed how readily 
modification of the diet would upset the relative distribution of 
B. coli types established through vaccine inoculations. A week 
after the animal had completed the immunization process and 
the sucrose-negative B. coli had become predominant, fifty 
grams of lactose was added to the diet of rice and meat. Within 
three days there was noted a very marked increase in the B. cola 
count with about an equal number of red and white B. cola 
colonies on the Endo plates. This apparent inhibition of the 
specific repressive action of B. coli inoculations ao as long 
as lactose was a part of the diet. 

Although the question of the relative effect of aukogensis and 
heterogenous vaccines was not made a feature of this study, 
yet some experiments with dog 2 indicated that to obtain marked 
results it was necessary to use an autogenous vaccine. After a 
preliminary observation period of six weeks, this animal was 
given three weekly inoculations of the B. coli vaccine prepared 


; 
q 


EFFECT OF AUTOGENOUS B. COLI VACCINES 139 


from cultures derived from dog 1 in a dosage of two billion. 
Very little in the way of definite effect, however, was noted until 
an autogenous vaccine was used when a marked and persistent 
decrease in the relative numbers of the sucrose-positive B. coli 
followed. 

With one animal (dog 4) an attempt was made to correlate 
the cultural results with the rise of antibody (agglutinin) in the 
blood following the inoculations. The serum from this animal 
before inoculation agglutinated an autogenous sucrose-positive 
B. coli culture at 1-160 dilution. This amount of agglutinin 
showed first a definite enhancement about two weeks after a 
series of three inoculations had been given, when the titer rose 
to 1-640. Shortly after this increase in anti-body content of 
the blood, a definite decrease occurred in the number of sucrose- 
positive B. coli in the fecal specimens examined. Continued 
inoculations with larger doses of the vaccine finally sent the 
titer up to 1-1280, with a coincident repression of the sucrose- 
positive B. coli as shown in the tabulation. A number of agglu- 
tination tests with extracts of the fecal matter were carried out 
to determine to what extent this antibody found its way into 
the intestinal tract. Positive results were obtained with watery 
defecations and especially with saline extracts of the mucus, 
but not with extracts from emulsions of the formed or semi- 
formed matter. 

These experiments with dogs have established the fact that 
it is possible through the use of autogenous vaccines to effect at 
least a temporary suppression of corresponding strains of B. coli 
naturally vegetating in the intestinal tract. Assuming that 
some varieties of B. coli may be of importance in connection 
with intestinal toxaemia, a certain amount of experimental 
justification is accorded the employment of an autogenous vac- 
cine therapy. In fact it should be much easier to effect by this 
means the suppression of parasitized B. coli than of the normal 
vegetative types which are firmly established in the intestinal 
tract. Such being the case one of the real difficulties in applying 
this form of therapy is to determine just what strains of B. coli 
may be giving rise to toxic products and to incorporate these in 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 2 


RAHE 


H. 


TORREY AND A. 


JOHN C. 


140 


owes “gy Z ‘G-TIT 
ous ““ Z ‘9Z-II 


OULDDBA JT 6I6I ‘ZI 
QUIDDBA SNOUdZOIN} OF] 16 1-2 s0q “@ @ O0G-Il | 24 SG ‘S-Ill ‘22 ‘Sel | % 
6161 ‘61 
porsed uoljeulo9BA-o1g 1g -Z UE ‘St ‘UL ‘6-9-F ‘ct |-S 
IGT ‘OE ‘Lz 
porsed uoryeurs9vA-o1g +6 e-I ‘VG ‘6 ‘0 ‘OT ‘FI ‘6-1 | @ 
s1equinu Ut ‘dF ‘92-11 6I6T 
pasvaloep oAvy suTe.ys eAOqy 98 1-S "ay 00-1 | 2 Ss STi Lee. Seo 
8g 1-4 CIGh ZEaSh eh Il @-iieler 
1[09 “g ‘a & ‘08-1 
@AT}ISOd-9so1ons jo sure.ys ‘Gl GEScl 6I6I 
Joyyo JO Joquinu ur as¥erdUT 92 e-1 a Seal || OV CI 308 26 Fe-E e 
a 1 Ciel 
a Lelie 
16 1-6 ‘a 16] GIGI ‘Zz ‘0% ‘FI-I | I 
‘el Te Geel 
“a 1 O6-1Ix 
"a L 8é-1IX 6I6I ‘8 
porsed uorjeut008 A I-€ “dT 1Z-1TX | ‘FS ‘T SI6T “TS ‘8Z-11X | I 
SI6I ‘0Z ‘OT 
potsed uoryeuIs0BA-a1g I-I “OL ‘OI-IIX *61 ‘ZI-1X | T 
SAHOLTOO ANIOOVA Ss # 
GHL OL 1100 ‘a 
SuUVNGY Gee as Are Lene eouone aati tN Ae fe SNOILVNINVXG JO satva |poa 


1100 ‘@ AAILISOd 
-as0u008 
40 ING) wad 


40 OILVH ADVUAAV 


sbop fo sauysazur ay) uryzim burnj9baa 090 ‘g fo sadfj ay} uo aur99DAa 21409 “g YRIM UOYWD)NI0U2 fo JIA] 
1 @TaVvL 


~~ 


141 


COLI VACCINES 


EFFECT OF AUTOGENOUS B. 


a 


ouLddBA SnoUes 
-oyny “poled UorlyeUIDIBA 


porsod uosul00BA-a1g 


QUIDIGA JO OBBSOP OF1B'T 


1[09 “g 9ATpISOd-aso.190 
-ns jo uonjsulunje juorsddy 


aUuINIBA SNOUuesOINY 


ee i OE aA a SSS ieee eee eae 


SP 


06 

OOT 
(83899 8) ST 
(83804 Z) 08 

€8 

LI 


TZ 


8% 


6€ 


1-9 
dAT}BSOU-9SOIONS [[V 


dATPISOd-asO19NS [TV 


Lest 
dAT}ISOd-aso1ONS [TV 
Tél 
T-L 


I-€T 
I-21 


I-8T 
9AT}BZ9U-9SOIONS [TV 


I-F 
9AI}BZIU-9SOIONS [[V 


1-9 


‘ad F ‘LZ-A 
‘ad 2 ‘OC-A 


"a2 ‘6-A 


"qd 8 ‘SI-IA 
"a8 OT-1A 
‘a 8 ‘OI-IA 


oues ‘g F ‘FZ-I11 
eues ‘"g F ‘61-111 
9UIDIBA TT 


30q ‘2 ‘SI-II1 


6161 
‘18 ‘8S 59S ‘SS ‘IZ ‘6I-A 
6I6I ‘SI-A 


6I6I ‘ZI-A 
6I6I ‘8 ‘S ‘Z-A ‘62-AI 
6I6I “Sz 
"eT? s2T “IL ‘6 L=AI 
6161 ‘E-AI 
6161 
‘I-AI *82 “96 “IZ-III 
6161 ‘81-111 
6I6I ‘FI ‘ZI-III 
GIGI OF ‘2% “8% ‘OZ-IA 


GIGI ‘ST ‘OT ‘ST ‘TI-IA 
6161 ‘6 “9 ‘FIA 

6161 
‘te ‘82 ‘9% ‘8% ‘12 ‘6I-A 
6161 ‘STI-A 

6I6I ‘ZI £8 ‘¢ 
‘SA *S% ‘8% ‘TZ ‘LI-AI 


GIGI ‘FT ‘TI-AI 


6161 
‘6 ‘2 ‘g ‘T-AI ‘82-III 


GIGI ‘9% ‘TZ ‘ST ‘ST-III 


es 
€ 


€ 
£ 


oD OD 


NM OD OD 


RAHE 


A. H. 


AND 


TORREY 


JOHN C. 


142 


1ojje pus ‘g]T JaquIeAON 4eIp 
0} peppe (sulBid 0g) as07,0BT 


-ONYV 


euIDDvA SnOues 
‘ported = U0lyBUlD0BA 


potiod u01yBUl00BA-o1g 


SHuUVANGY 


0€ 


(suory 
-BUIUIGX9 
G JO od BIDAL) 


o& 
v9 


I-8% 


ia 


T-L1 
1-S@ 
9AI}BZ9U-9SO1ONS [TV 


I-1€ 
I-9T 
9AI}BSIU-9SOIONS [[V 


‘a 9 ‘STI-IX 
"a9 ‘8 -IX 
‘O51 -Iix 
“@-9 “L6-X 
‘a § ‘0@-X 
“a9 ‘2I-X 
‘a 8 ‘PI-X 


‘dF ‘OIA 


6161 
“pb ‘S-IIX ‘82 ‘96 ‘FS-IX 
6161 
TiAl Cla Ol LI 
6161 
“g “@-IX ‘18 ‘6% ‘20-X 


‘ 


6161 

‘ez ‘82 ‘TZ ‘OT ‘ST ‘SI-x 
6161 

‘6 ‘9 ‘8-X 562 ‘2 ‘8I-XI 
6I6T ‘ZI ‘8 ‘E-XI 

66 ‘96 ‘TS ‘FI ‘ZI-IIIA 

6161 ‘TT ‘6 ‘Z-IIA 

6161 ‘S-IIA 


6161 ‘ZITA :08 
‘1G ‘GZ ‘8% ‘0% ‘ST ‘9T-IA 
6I6I ‘ST ‘TT ‘6 ‘9 ‘FIA 
6161 ‘E-IA 


oo OD SH 


SAUO L109 ANIOOVA 
GHLOL 
WOUDS-ILNV HLIM 
SONILVNILOTIODV 
1100 *& AAILISOd 
-asoOuon08 
4O INA) aad 


1100 °@ 
GAILISOd-As0u008 
OL GAILVOAN-ASONNO8 
40 OILVY ADVUAAV 


NOILVNIOOVA AO 
apvsod dNV saitva 


SNOILVYNINVXG JO SAaLva 


Soa 


= ae 


EFFECT OF AUTOGENOUS B. COLI VACCINES 143 


the vaccine. Apparently heretofore vaccines of this character 


have been made up in a rather hit or miss fashion. It should 


also be borne in mind that diet is the most important factor in 
the regulation of the intestinal flora. In fact these experiments 
indicate that an attempt to control types of bacteria germinat- 
ing in the intestinal tract through specific vaccine therapy is 
futile unless the diet is also carefully regulated. 


CONCLUSIONS 


1. In the case of a number of dogs it has been found possible 
to effect the temporary suppression of a certain variety of B. 
coli normal to the intestinal tract through inoculations with a 
specific vaccine. 

2. Autogenous vaccines are apparently necessary for marked 
results and the dosage must be large. 

3. Cultural results indicating a decrease in numbers of the 
type of B. coli in question were associated with a coincident rise 
of specific anti-body in the blood. 

4, A uniform diet must be maintained, otherwise the effect of 
the vaccine will be obscured. 


REFERENCES 


(1) SarreR.esn, G. R.: Journ. Amer. Med. Assoc., 1916, 67, 1729. 

(2) JACKSON AND PicHarp: Calif. State Journ. of Med., 1918, 16, No. 7. 
(8) Torrey, J. C.: Journ. Med. Research, 1919, 39, 415. 

(4) Levine, Max: Journ. of Bact., 1916, 1, 619. 


apy 


EXPERIMENTAL STUDY OF THE SENSITIZED 
CHOLERA ANTIGEN 


Y. MIURA 
From the Juntendo Hospital, Tokyo, Japan 


Received for publication February 18, 1920 


Knowing that ordinary heated vaccine has no complete pro- 
phylactic action upon experimental typhoid fever, Besredka (1), 
in 1902, first used the so-called sensitized vaccine, which was 
made by combining the living typhoid bacilli with the useful 
constituents of immune serum. Later he investigated with 
Metchnikoff (2) its prophylactic value against experimental 
typhoid fever, and they found it effective and not dangerous to 
the human body. In 1914 Ichikawa, in Japan, experimented 
with the non-heated sero-vaccine upon typhoid fever patients 
with good results. In 1916, when the outbreak of cholera was 
at its height in Japan, the cholera sensitized vaccine was pre- 
pared by Shiga and his associates (3) and favorable results were 
obtained. Later Yabe published an article on cholera vaccine, 
its weak reaction and rapid development of antibodies, as com- 
pared with the ordinary vaccine. 

We regret, however, that the sensitized vaccine in solution is 
not stable, and thus it cannot be kept for a long time in soluble 
form. In order to keep the sensitized vaccine without losing 
its power, as long as possible, it has been made in powder form 
by drying it in a vacuum desiccator over calcium chloride. We 
found that the sensitized cholera vaccine powder thus prepared 
has nearly the same effect as the bacilli suspension in liquid 
form, even after a period of nineteen months. We shall describe 
the experimental results in the following manner: 


145 


146 Y. MIURA 


I. METHOD OF PREPARING CHOLERA VACCINE 
a. Preparation of sensitized cholera vaccine 


The cholera bacilli from the agar culture were incubated for 
eighteen hours at 37°C. and weighed. Then were added together 
18 cc. of saline solution and 2 ee. of cholera immune serum of a 
rabbit (the bacteriological unit of which was more than 0.0001) 
for each gram of the above. After complete combination took 
place between the bacilli and the immune bodies, it was incu- 
bated at 37°C. for about two hours, and then centrifugalized. 
The supernatant liquid was decanted and the sediment was 
washed twice with saline solution. ‘Then as much saline solu- 
tion was added, containing 0.5 per cent carbolic acid as was 
necessary to make 1 cc. contain 2 mgm. of bacilli. 


b. Preparation of the sensitized cholera vaccine powder 


The bacilli suspension thus prepared was taken upon a watch 
glass which was placed in the middle of the desiccator provided 
with calcium chloride. Then the air in the jar was pumped out 
by means of an air pump connected with it, and the jar was 
allowed to stand for about forty-eight hours at room tempera- 
ture. When the vaccine became a brownish powder, it was 
kept in an ampule or tightly corked bottle. 

The sensitized cholera vaccine powder thus made corresponds 
to 10 mgm. of living bacilli in 1 mgm. Before use, 0.2 mgm. 
vaccine powder is suspended in 1 ce. saline solution containing 
0.5 per cent of carbolic acid, and the suspension is well shaken. 


II. ANIMAL EXPERIMENTS 


The animal experiments with the ordinary sensitized cholera 
vaccine have been published by Shiga and his associates as 
described above. We attempted to get the same results with 
the sensitized cholera vaccine powder nineteen months old, and 
to discover the difference in value, if any, between them. 

Ten healthy rabbits between 2000 and 2500 grams body 
weight were divided into two groups, In the first group 1 ce. 


Pe ee ee 


ee ie <a 
2s 


ie a 
> 


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 <a oa) % 
115 | Wm. Lent + TBA) 4+ | 4+ | 3+ | 2+ | 4+ 
116 | C. Gaynor atone | NIT! SNe —|/ -|] - —-| - 
117 | P. Mills ~ INES ING | SING — | 2+ _ — | 4+ 
118 | McDermott = NE TNE | NE = ~ _ -- -- 
119 | Foster — INET | NGL, |) SINE a cE _ + |-+ 
120 | Angelo — | NT |NT| NT — — — -- — 
121 | De Laney Se NE? INT | INS _ _ - - — 
122 | McNold -- INAS NEE | INE _ -- — - -- 
123 | J. Colbert _ NE | NIE | NQ + -- _ _ + 
124 | Lottie Crews — INES | ING |) INE — _ o — — 
125 | Hansen — Ne Nae |) INE -- aa _ ~ — 


1 Refers to the classification made by the clinician at the time the first speci- 
mens were obtained. 

2 Refers to the subsequent classification made by the clinician three weeks 
later when the second specimens were taken. 

3 Refers to the writer’s classification, of the cases. This was obtained by 
transcribing the first Otisville (National classification) into terms of the writer’s 
classification. 

4 Refers to the first specimens taken from the patients. 

5 Refers to the second specimens taken three weeks later from the same 
patients. 

6 Blank space indicates that only one specimen of serum was obtained from 
the patient. 

7Sputum report obtained within one month of the time the blood specimen 
was taken. 

8 Refers to the finding of many or moderate numbers of tubercle bacilli. 

® No specimen examined for quite a long while. 

10 Refers to the finding of only an occasional bacilli. 

11 Remaining non-tubercular control sera were negative in all tests with all 
antigens. 

122 NT refers to non-tuberculous. 

18 TBA,—Tuberculous arthritis of the ankle. 

Explanation of the above table: The foregoing table gives a résumé of the 
important points noted in the Otisville investigation (series V). In addition 
to the foregoing data, we had von Pirquet tests, and Wassermann reactions on 
quite a few and clinical data on all. 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO, 2 


220 HASSOW O. VON WEDEL 


16. SUMMARY 


The one hour fixation period in the water bath at 37°C. appears 
to be the optimum time and temperature for the complement 
fixation test with our antigens. 

Specially tested complement, if only one guinea-pig is killed 
should be used, but untested pooled complement from six or 
more guinea-pigs apparently gives satisfactory results. 

Natural antisheep hemolysin is markedly thermolabile. 

If the serum is kept sterile the natural antisheep hemolysin 
does not depreciate very rapidly due to ageing. 

Heating tubercle bacillus antigen at 100°C. for three hours 
does not seem to impair its antigenic value in the slightest. The 
addition of 25 per cent alcohol or 4 per cent carbolic acid to the 
finished antigen as preservative tends to make the antigen anti- 
complementary to a greater or less degree. 

The best method for preparing the tubercle bacillus antigen 
seems to be by killing the bacilli with alcohol, and after the 
finished antigen is bottled and corked in small vials, sterilizing 
at 100°C. for one hour on two or three successive days to kill 
all contaminating spores and bacteria. 

An increase in the percentage of positive results can probably 
be obtained by absorption of the natural antisheep amboceptor 
before making the test or by using the fresh non-inactivated 
sera according to the technic advocated by Hecht and Gradwohl. 
However, while both of these modifications give a higher per- 
centage of specific positive reactions than the standard test, 
they also give a moderate percentage of non-specific reactions. 
Therefore, they should only be used as control tests along with 
the standard technic. 

The increase in the strength of the fixation and the increase 
in the percentage of positive findings after preserving the 
patient’s sera for seven days in the ice chest is due, probably, 
to one or more of the following reasons: the formation of thermo- 
stable antilysins in the kept sera, loss of natural antisheep 
amboceptor due to ageing, loss of natural antisheep amboceptor 
due to reheating and other unknown non-specific causes. This 


COMPLEMENT FIXATION TEST FOR TUBERCULOSIS Pv | 


change in sera after preserving in the ice chest for one week is 
apparently non-specific. 

No cross fixation was apparent between syphilitic antibodies 
and the tubercle bacillus antigens which were used in this study. 

The results of the von Pirquet reaction apparently do not 
parallel the results obtained with the complement fixation test. 

The National Association Classification of tuberculosis patients 
appears to be too complicated to be used in successfully inter- 
preting this test. If used in its simplified form, the results of 
the test do not run parallel with the classification. 

By using the writer’s simplified interchangeable classification 
a more accurate interpretation of the results of this test is pos- 
sible. However, even with this simpler classification it was 
found impracticable accurately to tabulate the results by types 
in a routine way as the classification of these same patients 
varied considerably when made by different clinicians. In 
general, we may say that the largest percentage of negative 
and weak positive results are in the cases having few or no 
constitutional symptoms, while the strength and percentage of 
positive reactions increase progressively with the increased 
constitutional symptoms, until the patient reaches the far 
advanced condition, when the reverse is true; i.e., when the 
patient loses his power of resistance, the complement fixation 
reaction becomes negative. 

The age of the patient apparently has no effect on the com- 
plement fixation test for tuberculosis. 

About 70 per cent of all types of tuberculosis patients except 
those clinically healed or inactive gave positive fixation results 
on repeated tests. Normal non-tuberculous cases gave almost 
no positive results on repeated tests. Moderately and far 
advanced cases in good condition showing constitutional symp- 
toms (type III) gave an average of 85.2 per cent positive fixations 
for all six series. 

This complement fixation reaction will not be very valuable, 
as an aid in diagnosis, to the tuberculosis specialist except as 
a confirmatory test. However, a positive fixation reaction will 
be of very great value to the general practitioner not only as 
a confirmatory test but also as an aid in diagnosis and prognosis. 


bo 
bo 
bho 


HASSOW O. VON WEDEL 


17. CONCLUSIONS 


The results of 6128 complement fixation tests made on 1207 
sera from 1000 patients point to the fact that this is not a 100 
per cent test for the diagnosis of tuberculosis. A considerable 
percentage of sera from incipient and far advanced cases appar- 
ently contain insufficient antibodies to fix complement, no 
matter what system or what antigen is used for the test. This 
fact, therefore, precludes the probability of a 100 per cent test. 
based on complement binding antibodies in the patient’s serum. 

A great many antigens and many systems have been tried by 
various workers but even the most favorable reports on large 
series bear out this statement. 

Numerous favorable reports have been published on small 
series but apparently none have been confirmed where the per- 
centage of positive findings, with a series of unselected active 
tuberculous cases of all types, has been more than 80 per cent— 
unless there was at the same time a marked degree of non- 
specific fixation. 

About 70 per cent positive results appears to be the average 
findings, with all types of unselected active tuberculous cases, 
for many thousands of complement fixation tests made by many 
serologists, using tubercle bacillus suspensions or tuberculins 
as antigens. The reactions are weakest when the patient 
exhibits few, if any, symptoms of tuberculosis, while they are 
most definite and strongest in the incipient and moderately 
advanced cases exhibiting marked symptoms. The results are 
therefore more confirmatory than actually diagnostic in the 
largest percentage of cases. However, when used intelligently 
along with the clinical history, the results justify its more 
extended use. 

A positive reaction repeated twice seems to prove fairly con- 
clusively that the patient has an active tuberculous process. 
The weakening of the reaction from a strong positive to a weak 
positive or negative reaction, is apparently a good prognostic 
sign in incipient and moderately advanced cases exhibiting 
clinical improvement, while at the same time this same change 
with a far advanced case in a poor clinical condition is a very 
bad prognostic sign. 


CGMPLEMENT FIXATION TEST FOR TUBERCULOSIS 223 


REFERENCES 


(1) McCrak anp Funk: Jour. Am. Med. Assoc., 1919, 73, 161. 

(2) Wipau AND LEeSourp: Cited by Shennan and Miller. Edinburgh Med., 
Jour., 1913, 10, 81. 

(3) Borper anp Gencovu: Compt. rend. Acad. de Se., 1903, 187, 351. 

(4) WASSERMANN AND Bruck: Deutsch. med. Wehnschr., 1906, 32, 449. 

(5) Caunrretp: Arch. Int. Med., 1911, 8, 440. 

(6) CAULFIELD AND BrarTiE: Jour. Med. Research, 1911, 24, 122. 

(7) Dettmann: Zeit. f. Immunititsf., 1911, 10, 421. 

(8) Larrp: Jour. Med. Research, 1912, 27, 163. 

(9) MéuueRs: Cent. Bact., Ref., 1912, 54, Beiheft, 202. 

(10) Hammer: Deutsch. Tierarzt. Woch., 1912, 20, 593. 

(11) Zwera: Berl. klin. Woch., 1912, 49, 1845. 

(12) Catmetre, Masson anp Méziz: Compt. rend. Soc. Biol., 1912, 73, 122 

(13) Mucu: Munch. med. Wehnschr., 1912, 59, 685. 

(14) Letu.Lue: Thése, Fac. Méd. Paris, 1912. 

(15) Fraser: Ztschr. f. Immunititsf., Orig., 1913, 20, 291. 

(16) Duperon, MEEK anv WEIR: Lancet, 1913, 1, 19. 

(17) Bane anpD ANDERSEN: Cent. Bact., Orig., 1913, 69, 517. 

(18) ARMAND-DELILLE, Rist AND VAUCHER: Compt. rend. Soc. Biol., 1913, 74, 
791. 

(19) WyscHEeLessky: Zeit. f. Tuberk., 1912, 19, 209. 

(20) KINGHORN AND TWITCHELL: Zeit. f. Tuberk., 1913, 20, 11. 

(21) Rorur AND Brersaum: Deutsch. Med. Woch., 1913, 39, 644. 

(22) Harris AND LANnForp: Jour. Inf. Dis., 1913, 13, 301. 

(23) Momossr: Deutsch. Med. Woch., 1913, 39, 1029. 

(24) BesrepKa: Zeit. f. Immunititsf., Orig., 1914, 21, 77. 

(25) WwWEDENSKY (revised by Wutrrius): Zeit. f. Immunititsf. referat., 1914, 
8, 931. 

(26) DEBAINS AND JUPILLE: Compt. rend. Soc. Biol., 1914, 76, 199. 

(27) Kuss, LpEREDDE AND RUBINSTEIN: Ibid., 244. 

(28) Inman: Compt. rend. Soc. Biol., 1914, 76, 251. 

(29) BrerBAUM AND BERDEL: Zeit. f. Immunititsf., 1914, Orig., 21, 249. 

(30) McInrosH AND FiupsEs: Lancet, 1914, 2, 485. 

(31) Rapcuirre: Lancet, 1914, 2, 488. 

(32) Dupgron, MrEeK anp WetR: Jour. Hyg., 1914, 14, 52, 72. 

(83) DupGron, MEEK AND WEIR: Jour. Hyg., 1914, 14, 52, 72. 

(34) BRONFENBRENNER: Arch. Int. Med., 1914, 14, 786. 

(35) BRONFENBRENNER: Proc. Soc. Exp. Biol. and Med., 1914, 11, 180. 

(36) BRONFENBRENNER: Proc. Soc. Exp. Biol. and Med., 1914, 11, 92-93. 

(37) BRONFENBRENNER AND SCHLESINGER: Trans. 12th Meet. Nat. Assoc. 
Study and Prev. Tuberc. 1916, 225. 

(388) Crate: Am. Jour. Med. Sc., 1915, 160, 781. 

(89) Stimson: Bull. Hyg. Lab. U. S. P. H. S., 1915, 101, 7. 

(40) Corprer: Jour. Inf. Diseases, 1916, 19, 315. 

(41) Mituer, H. R., anp Zinsser, H.: Proc. Soc. Exper. Biol. and Med., 1916. 
13, 134. 


224 HASSOW O. VON WEDEL 


(42) Mitier, H. R.: Jour. Am. Med. Assoc., 1916, 67, 1519. 

(43) Woops, BusHNELL, AND Mappux: Jour. of Immunol., 1917, 2, 301. 

(44) McCasxry, G. W.: Am. Jour. Med. Sc., 1917, 154, 648. | 

(45) Brown AND Perrorr: Am. Rev. Tuberc., 1918, 2, 525. | 

(46) Laner, L. B.: Am. Rev. Tuberc., 1918, 2, 541. 

(47) Srrvetman, B.: Amer. Rev. Tuberc., 1918, 2, 546. 

(48) Lewis, P. A.: Am. Rev. Tuberc., 1919, 3, 129. 

(49) PrircHarD AND Roperick: Jour. Am. Med. Assc., 1919, 73, 1879. 

(50) Cooxs, J. V.: Jour. of Inf. Dis., 1919, 25, 493. 

(51) Stott anp Neuman: Jour. Am. Med. Assc., 1919, 72, 1043. 

(52) Mournsunp, W. H.: Jour. Inf. Dis., 1920, 26, 85. 

(53) Moon, V. H.: Jour. Am. Med. Assc., 1919, 71, 1127. 

(54) BRONFENBRENNER: Jour. Lab. Clin. Med., 1917, 3, 50. 

(55) Bonz anp Dunor: Compt. Rend. Soc., Biol., 1919, 82, 559. 

(56) Wiuson, M. A.: Jour. Immunol., 1918, 3, 345. 

(57) Ruepicrer, E. H.: Jour. Inf. Diseases, 1919, 24, 405. 

(58) Nocucut, H., anp BRONFENBRENNER: Jour. Exper. Med., 1911, 18, 217. 

(59) Roncuesz, A. D.: Compt. Rend. Soc. Biol., 1919, 82, 193. 

(60) THomeson: Jour. Am. Med. Assc., 1916, 66, 652. 

(61) Ruamy: Jour. Am. Med. Assc., 1917, 69, 973. 

(62) McNeruu (cited by THomerson): Amer. Jour. of Syph., 1917, 1, 555. 

(63) Kot~mer: Jour. of Immunol. to be published. 

(64) Turetze, F. H., anp Empueton, D.: Ztsch. f. Immunititsf. U. Exper. 
Ther., Orig., 1914, 20, 1-50. 

(65) SHermMan, H.: Jour. Inf. Dis., 1918, 22, 534. 

(66) Mituer, H. R.: Jour. of Lab. and Clin. Med., 1916, 1, 816. 

' (67) CorpEer: Jour. Inf. Diseases, 1916, 19, 315. 

(68) Brown AND Petrorr: Amer. Rev. Tuberc., 1918, 2, 525. 

(69) BRONFENBRENNER, Kaun, M. H., RockMan AND Kaun, M.: Arch. Int. 
Med., 1916, 17, 492. 

(70) Brucr, W. J.: Jour. Lab. and Clin. Med., 1919, 4, 215. 

(71) Lewis anp Newcomer: Jour. Exp. Med., 1919, 29, 351. 

(72) Grapwou., R. B. H.: Jour. Am. Med. Assce., 1917, 68, 514. 

(73) RuBINSTEIN: Compt. Rend. Soc. Biol., 1919, 82, 526. 

(74) RUBINSTEIN AND RADOSSAVLIEVITCH: Compt. Rend. Soc. Biol., 1919, 
83, 361. 

(75) EscopacH AND DunHot: Compt. Rend. Soc. Biol., 1919, 82, 452. 

(76) Bauer, J.: Deutsch. Med. Wehnschr., 1908, 34, 698. 

(77) Bauer, J.: Berl. klin. Wehnschr., 1908, 45, 834. 

(78) Rosst, O.: Ztsch. f. Immunititsf., orig., 1911, 10, 321. 

(79) Stmon, C. E.: Jour. Am. Med. Assoc., 1917, 72, 1535. 

(80) WEcHSELMANN: Ztsch. f. Immunitatsf., orig., 1909, 3, 525. 

(81) Nocucui, H., anD BRONFENBRENNER: Jour. Expr. Med., 1911, 13, 217. 

(82) Jacopanus, H. C.: Ztsch. f. Immunititsf., orig., 1911, 8, 615. 

(83) OumstEAD, M. P.: Medical Record, N. Y., 1914, 85, 341. 

(84) Van Savn, A. I.: Jour. Lab. and Clin. Med., 1917, 3, 59. 

(85) OTTENBERG, R.: Archiv. Int. Med., 1917, 19, 457. 

(86) SeetmMan, J. J.: Jour. Lab. and Clin. Med., 1918, 3, 626. 


COMPLEMENT FIXATION TEST FOR TUBERCULOSIS 225 


(87) Stern, M.: Ztsch. f. Immunititsf., orig., 1914, 22, 130. 
(88) Stimson: Bull. Hyg. Lab. U. S. P. H. S., 1915, 101, 7. 
(89) Neitz, M. H.: Hygienic Laboratory Bulletin No. 110, September 1917, 47. 
(90) Ravorort, F. H.: Jour. Am. Med., Assc., 1919, 72, 633. 
(91) Burns, Stack, CASTLEMAN AND Battery: Jour. Am. Med. Assc., 1917, 
68, 1386. 
(92) BRONFENBRENNER: Jour. Exp. Med., 1912, 15, 598. 
(93) CorperR AND Sweany: Jour. Am. Med. Assoc., 1917, 68, 1598. 
(94) Bonz anp Dunot: Compt. Rend. Soc., Biol., 1919, 82, 559. 
(95) TurBan: Diagnosis of Tuberculosis of the Lung, Edition 1906, 42-49. 
(96) TurBAN: Reference quoted by Rathbun; not found. 
(97) Meissen: Handbuch der Tuberculose, Band 1,743, published by Barth, 
Leipzig, 1914. 
(98) Rarusun, W. L.: N. Y. C., Dept. of Health. Tuberculosis Monograph, 
No. 4, March ,1917. 
(99) von Wepet, H.: Jour. Immunol., 1918, 3, 351. 
(100) Moursunp, W. H.: Jour. Inf. Dis., 1920, 26, 85. 
(101) Grapwout, R. B. H.: Jour. Am. Med. Assc., 1914, 63, 240. 
(102) Snow anp Cooper: Amer. Jour. of Med. Sce., 1916, 152, 185. 
(103) Lewis anp Newcomer: Jour. Exp. Med., 1919, 29, 351. 
(104) Hammer: Munch. Med. Wehnschr., 1912, 59, 1750. 
(105) DupcEon, Mrrex anv Weir: Jour. Hyg., 1914, 14, 52, 72. 
(106) BesrepKa: Compt. rend. Acad. d. sc., 1913, 156, 1633. 
(107) CatmMetTrre anpD Masso: Compt. rend. Soe. Biol., 1913, 75, 160. 
(108) Lucxs, B.: Jour. Immunol., 1916, 1, 457. 
(109) Porter: Jour. Hygiene, 1911, 11, 105. 


pa 


ON THE TRANSFER OF THE SO-CALLED NORMAL- 
ANTIBODIES FROM MOTHER TO OFFSPRING 


I. AGGLUTININS 


G. C. REYMANN 
From the State Serum Institute, Copenhagen; Director, Th. Madsen 


Received for publication March 10, 1920 


By far the greater number of those who have occupied them- 
selves with experiments on the agglutinins (bacterial as well as 
hemagglutinins) present in the blood of normal mothers and their 
newborn young, have arrived at the result that in the blood of 
the offspring these bodies are absent or at any rate less in 
amount than in that of the mother, despite the fact that 
the mother’s serum as well as the mother’s milk shows a strong 
reaction. This conclusion was reached by Gruenbaum (1), 
Halban (2), Halban and Landsteiner (3), Schumacher (4), 
Schenk (5), Zubrezychi and Wolfsgruber (6), for human beings, 
and by Kraus and Loew (7), Kraus and Clairmont (8), G. Muel- 
ler (9), Park (10), Luedke (11), V. Eisler and Sohma (12) for 
various species of animals, whereas Klieneberger (13) draws no 
quantitative conclusions; finally V. Fellenberg and Doell (14) 
arrive at the result, that the serum of a child sometimes shows a 
stronger, sometimes a weaker reaction than that of the mother, 
and they decline to believe in any constant ratio between them. 

According to Pfaundler (15), the fact that the serum of the 
new-born agglutinates faintly or not at all, necessitates, in view 
of the Gruber-Widal reaction, the fixation of the upper limit of 
the normal agglutination for the various ages. 

The mother’s milk was found by Schumacher to contain agglu- 
tinin in varying amounts; Fellenberg and Doell tested the pro- 
portion between the agglutinin content of the mother’s serum 
and milk, but could not demonstrate any constant proportion, 


997 


“al 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 3 


228 G. C. REYMANN 


and Zubrezychi and Wolfsgruber found in human milk rather 
large amounts of hemagglutinin, which, in their opinion, is not 
transmitted to the serum of suckling children. Whereas the 
relative agglutinin content of the blood of mother and offspring 
at the moment of birth is, in the main, clearly elucidated through 
the works quoted above, the time immediately after the birth 
of the young has, to all intents and purposes, been treated only 
_ in the above mentioned, by no means exhaustive, work by 
Zubrezychi and Wolfsgruber. The reason of this is to be sought 
in the fact that the researches are chiefly based upon human 
material or have been undertaken with small experimental ani- 
mals; hence, a regular taking of blood samples at short intervals 
has been in both cases out of the question. 

The object of this work was, through regular series of blood 
and milk tests, to examine the quantitative proportion of the 
agglutinins (normally present in the blood) in the blood and 
milk of the mother and the blood of the offspring, in order to fix 
the relative values of these three. As the experimental animal, 
the goat was selected, it being in every respect suitable for the 
purpose; the samples of blood were taken from the jugular vein. 

The samples were tested for typhoid and coli agglutinin as well 
as for agglutinin against horse and rabbit blood-corpuscles; 
each blood sample was tested on as many as possible of these 
four. As the measurements of bacterial agglutinin is impossible 
in the natural milk, it is necessary to make this fluid translucent 
by precipitation of the casein by means of rennet; all the figures 
below are given for this milk serum. 

The technic followed in these experiments was the one used 
at the institute and originally introduced by Th. Madsen and 
Jorgensen (16) for the determination of bacterial agglutinins, 
and the same method was employed for the measuring of the 
hemagglutinins, only with somewhat different quantitative pro- 
portions; thus 8 cc. of 1 per cent blood suspension per tube were 
used in the tests for hemagglutinin, and only 1 cc. of bacterial 
suspension was used in the bacterial tests. 

In view of the tests undertaken at the same time for normal 
antihemolysins, which will be communicated in a later article, 


oS a 
, : 


TRANSFER OF SO-CALLED NORMAL-ANTIBODIES 229 


the samples were inactivated for half an hour at 56° with the 
exception of kid 7 and 8 (see table 1), after experiments had been 
carried out showing that the weakening was imperceptible at 
this temperature. Only at higher temperatures does the influ- 
ence of heat become perceptible, as was formerly shown by 
Th. MadsenandStreng (17) in the case of the immune agglutinins. 

All of the kids were weighed at the taking of the blood samples, 
but the increase in weight has not been taken into consideration 
in the calculation, because the period within which their blood 
contains agglutinin is so short. 

Nor was the hydrogen ion concentration taken into considera- 
tion during the measuring of the agglutinin, for, as demon- 
strated by Michaelis and Davidsohn (18), fluctuations in pH 
round the neutral point plays no part, and I have made use of a 
broth the pH of which lay at this point, and which was, besides, 
very rich in regulators; furthermore, in the case of the hemag- 
glutinins there have always been rather large quantities of serum 
in each tube and according to Walbum’s (19) researches serum is 
an excellent regulator when present in sufficient concentration. 

In table 1 is presented a summary of the researches which were 
made on 7 pairs of twins. Dates of the bleeding and of the milk- 
ing are given. The results are quoted in agglutinin units (con- 
tent per cubic centimeter) ; the dates of the birth are marked with 
an asterisk. It appears from table 1 that only the serum of one 
kid contained agglutinin at birth, in spite of the fact that in all 
of the cases agglutinin was present in the serum of the mother 
animal. This case may possibly be a parallel to the few instances 
in which previous experimentators, Gruenbaum and Schumacher 
find “‘inherited” agglutinin. The agglutinin is transmitted to 
the kid through the milk, as appears with probability from the 


_table; in the milk as well as in the kid’s serum it decreases quickly, 


and as a rule it first disappears from the milk. The titer in the 
serum of the kid may be above or below that in the mother’s 
serum, but only in a single case, where the milk gave an abnor- 
mal reaction (12 and 13’s mother), was the titer higher than that 
of the milk, which was possibly due to the presence of bodies 
preventing agglutination in the milk, such as those previously 


0] Of} I>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 <i] 2.8 1.0 5.4 
2/22 7.6 <1 2.5 1.0 2.6 
2/23 5.9 <All 1.3 


mals, and it appears that the agglutinin disappears with varying 
rapidity in them; thus samples always ought to be taken from 
the nipple used. 

Experiment 1 shows the significance of the mother’s milk in 
the occurrence of agglutinin in the blood of the kid; in experiment 
2 it seems as if agglutinin had possibly been present in the cow’s 
milk in a disguised form or in quantities too small to be meas- 
ured; in any case the occurrence of agglutinin has been slight, 
considerably less than in the kid that remained with the mother. 


234 G. C. REYMANN 


According to tables 1 and 2 the greatest concentration of agglu- 
tinin in the serum of the kid occurs immediately after birth and 
in order more accurately to determine its place, test samples 
were taken with inter spaces of afewhours from a single kid and 
the milk of its mother. It now turned out, as is apparent from 
chart 2, that within a few hours after birth agglutinin can be 
demonstrated in the blood of the kid and that the height of the 
agglutinin curve hes about eleven hours after birth, where it 


Colon agglutinins ; units per cubic centimeter. 


g — «Kid serum 


The serwm of the mother animal 


contained 36 units at birth. 


CHART 2 


reaches exactly half of the initial titer of the milk, which is in- 
teresting, in so far as a passive immunization through the diges- 
tive tract is otherwise difficult and at any rate uneconomical; 
thus, Wegelius, as mentioned above, could not prove the pres- 
ence of antihemolysin in the blood of the kids, after feeding with 
milk containing antihemolysin. There is thus in all proba- 
bility a difference between the assimilation of artificially trans- 
mitted antibodies and those which in a natural way have been 
produced in the blood of the mother animal. When the agglu- 


TRANSFER OF SO-CALLED NORMAL-ANTIBODIES 235 


tinin has disappeared from the milk,it remains free of agglutinin 
(the longest regular experimental space was seventy-four days), 
and in all the samples taken I found only two with a faint agglu- 
tinating power. 

In the blood of the kid the agglutinin reappears after a few 
months, probably, as indicated by several authors, in conse- 
quence of the colon-immunization from the digestive tract. As 
an example of this I present table 3; only the last of the kids put 
down in the latter is not to be found in table 1,. but it was like- 
wise examined immediately after birth and during the first three 
months of its life. 


TABLE 3 
Samples taken June 10 in the year of birth 


AcGLUUNIN | “UNITS PER 
KID NUMBER | DATE OF BIRTH KIND OF AGGLUTININ CUBIC CENTI- dt ties ae 
METER OF KIDS MOTHER'S 

BEE SERUM 
a 2/21 Typhoid agglutinin 5.0 3.9 
10 2/24 Typhoid:agglutinin ee 4.0 
13 2/27 Horse blood agglutinin Fal 2.0 
27 3/5 Horse blood agglutinin 0.7 2.8 
27 3/5 Colon agglutinin 371 11.0 
27 3/5 Typhoid agglutinin 1S 2.0 
36 3/12 Horse blood agglutinin 0.8 2.5 
36 3/12 Rabbit blood agglutinin 0.7 2.0 
36 3/12 Typhoid agglutinin 4.0 11.0 


It appears from the above tables that there is a parallelism 
between the increase and decrease of the tested agglutinins in 
the kids’ blood, and the circumstance that after having been 
absent for a considerable period they all reappear might suggest 
a mutual relationship; consequently an experiment was made to 
see whether it was possible through the injection of a bacterial 
agglutinin antigen to develop the production of a hemagglutinin 
and vice versa. The experiment was undertaken in the following 
manner: of two nearly six months old kids the one was injected 
with colon culture, the other with washed horseblood corpuscles, 
both subcutaneously with two injections of each antigen of 5 


236 G. C. REYMANN 
and 10 cc. respectively, on October 14 and 16. The graphic 


representation (chart 3) shows that both kids reacted by producing 
anti-horse blood as well as anti-colon agglutinin. 


a 35 Kid al injected with horse's blood stested fer hemasghitinins. 


e—eoltida. tested pr Colon aggltinins: 
ee eid: in jected with B.coli , tested for hemaggltinins- 
o---o iat, te stad pr colon agglutinin s. 


agglutinin units 


CHART 3 


SUMMARY 


1. By examining the transmission of the agglutinins normally 
occurring in the blood from mother to offspring in goats, I have 
only in one case of fourteen been able to prove their recurrence; 
in all the others the kids were born without agglutinin and prob- 
ably derived it from the mother animal through the milk, in 
which it is to be found accumulated at parturition. From the 
milk as well as from the kid’s serum it disappears in the course 
of a few days; then follows a period of a few months in which the 
blood of the kid is free of agglutinin, and then it appears again, 
probably in consequence of an immunization from the flora of 
the digestive tract. The research was concerned with coli and 
typhoid agglutinins, as well as agglutinin against rabbit and 
horse blood corpuscles. 


a 


TRANSFER OF SO-CALLED NORMAL-ANTIBODIES 237 


2. In the blood of the kid there was in some cases more, in 
others less of the ‘‘normal antibodies” than in the blood of the 
mother animal, and only in one case the blood of the kid con- 
tained more than the colostrum, so that it is not possible to deduce 
any quantitative rule from these experiments. 

3. In the colostrum the titer was higher than in the serum of 
the mother animal. 

4. By nursing experiments it was shown in one case that the 
agglutinin is probably transmitted to the kid through the 
mother’s milk, in another case the result was questionable. 

5. The agglutinin maximum in the blood of the kid may occur 
as early as about 11 hours after birth. 

6. By injections of horse blood corpuscles into kids it is pos- 
sible to increase the agglutinating power as well against these 
corpuscles as against coli bacilli; vice versa by injection of coli 
bacilli the agglutinating power is increased also against horse 
blood corpuscles. 


REFERENCES 


(1) GruenBaAum, A. S.: Muench. med. Wochenschr., 1897, 44, 330. 

(2) Haxpan, J.: Wien. kl. Wochenschr., 1900, 13, 545. 

(3) HaLBan, J., AND LANDSTEINER, K.: Muench. med. Wochenschr., 1902, 49, 
473. 

(4) ScoumacHER, H.: Zeitschr. f. Hygiene u. Infektionskrankh. 1901. 

(5) Scuenx, F.: Monatschr. f. Geburtsh. u. Gynaekol., 1904, 19, 159, 344, 568. 

(6) ZuBRzycx1, J.. AND WoLFSGRUBER, R. Deutsche med. Wochenschr., 1913, 
39, 210. : 

(7) Kraus, R., anp Lorw, L.: Wien. klin. Wochenschr., 1899, 12, 95. 

(8) Kraus, R., anp Lipscnurrz, B.: Zeitschr. f. Hyg. u. Infektionskr., 1904, 
46, 49. 

(9) MurtuerR, G.: Diss. Darmstadt, 1901. 

(10) Park, W. H.: Proc. Soc. f. Exper. Biol. & Med., 1903, 1, 19. 

(11) Lurepxer, H.: Zentralbl. f. Bakt., Orig.. 1904, 37, 288, 418. 

(12) v. Erster, M., anp Souma, M.: Wien kl. Wochenschr., 1908, 21, 684. 

(13) KureneperGcer, C.: Deutsche Arch. f. kl. Med., 1907, 90, 267. 

(14) v. FELLENBERG, R., AND Dortu, A.: Zeitschr. f. Geburtsch. u. Gynaekol., 
1913, 75, 285. 

(15) Praunputer, M.: Muench. med. Wochenschr., 1899, 46, 473. 

(16) Mapsen, TH., AND JORGENSEN, A.: Festskrift. Statens Seruminstitut., 
1902, no. VI. 

(17) Mapsren, TuH., AND StRENG: Communicat. d. l'Institut therap. d. Etat 
Danois, 9. 


238 


G. C. REYMANN 


(18) Micnasuis, L., AND DAVIDSON, H.: Biochem. Zeitschr. 
(19) Wazsum, L. E.: Biochem. Zeitschrift., 1914, 63, 221. 


(20) WEGELIUS, 


W.: Arch. f. Gynaekol., 1911, 94, 265. 


(21) Moreenrots, J., AND Bravn, H.: Die Vererbungsfrag 
taetslehre, Hbd. d. pathogenen Mikroorganismen V. 
Koiie AND WASSERMANN: 2 aufl., 1913, 2, 1155. 


STUDIES IN ANAPHYLAXIS 


THE RELATION OF CERTAIN DRUGS TO THE ANAPHYLACTIC 
REACTION, AND THE BEARING THEREOF ON THE 
MECHANISM OF ANAPHYLACTIC SHOCK 


MAURICE I. SMITH 
From the Pharmacological Laboratory of the University of Nebraska, Medical 
College, Omaha 


Received for publication April 1, 1920 


Comparatively few investigations have been recorded in the 
literature relative to the behavior of pharmacologic agents in the 
course of anaphylactic shock. Most of the studies of this kind 
have been made with a view of finding a substance which would 
relieve anaphylactic symptoms, and save the sensitized animal 
from the usually fatal results following the reinjection of the 
antigen in adequate doses. Thus, Besredka (1) found that 
guinea-pigs sensitized to ox serum recovered from a lethal dose 
of the antigen if the animals were under the influence of ether 
when the reinjection was made. Auer (2) has likewise shown 
the protecting influence of atropine in the course of anaphylactic 
shock. According to this author 3 mgm. of atropine injected 
into the subcutaneous tissues of sensitized guinea-pigs before 
the reinjection of the antigen exerted a favorable influence upon 
the outcome, as 3 out of 5 experimental animals thus treated 
were saved from immediate death. This effect of atropine, while 
interesting from the therapeutic standpoint, is also taken to 
throw light on the mechanism of the anaphylactic response. It 
is taken to mean that anaphylactic shock, in the guinea-pig at 
least, is due to a fatal broncho-constriction, which is peripheral 
in nature, since sectioning or degeneration of the vagi has no 
effect, while atropine partially prevents it. The results of Auer, 
however, did not find confirmation in the hands of Mita (3), 
who states that if the minimal lethal dose of the reinjected antigen 

, 239 


240 MAURICE I. SMITH 


be quantitatively determined per unit weight of animal, and if 
one makes sure that the full lethal dose be given in each instance, 
atropine exerts no prophylactic influence on the course of the 
anaphylactic reaction. 

The work of Banzhaf and Famulener (4) may also be cited 
in this connection. These authors investigated the effects of 
chloral on the course of anaphylactic shock, and found that 
100 to 150 mgm. of this drug injected intramuscularly into sensi- 
tized guinea-pigs saved 90 to 100 per cent of the experimental 
animals from anaphylactic shock. k 

Except for the experiments of Auer (loc. cit.) no work appears 
to have been done with pharmacologic substances with a view 
of a possible elucidation of the mechanism of anaphylactic shock. 

The present studies deal in the main with the observation 
made some years ago of the peculiar influence of quinine on the 
course of the anaphylactic reaction. During the course of some 
experimental studies made on the antipyretic action of quinine 
in anaphylactic fever, it was found that this substance had no 
effects on sensitized rabbits if injected at various periods of the 
incubation interval. But if the drug was administered about a 
half hour before the reinjection of the foreign protein the animal 
had been sensitized to, a dose which ordinarily produced merely 
a rise in body temperature, and no other appreciable symptoms, 
now caused grave symptoms of anaphylactic shock, often termi- 
nating fatally. This observation appeared to have sufficient 
clinical importance, aside from the possibility that it might help 
to elucidate the mechanism of the anaphylactic response, and 
the matter was deemed worthy of a detailed study. 

Rabbits and guinea-pigs were used throughout this research. 
The animals were in all instances actively sensitized. In some 
experiments horse serum! was used, while in most of the experi- 
ments ox serum was employed. Sensitization was effected in 
the guinea-pig with great regularity, by the intraperitoneal route, 
while the rabbits were usually sensitized by the subcutaneous 
route. Usually 2 cc. of ox serum injected subcutaneously in the 


1 Normal horse serum of Parke, Davis and Company, containing 0.4 per cent 
cresol. 


STUDIES IN ANAPHYLAXIS 241 


rabbit is sufficient to render the animal anaphylactic within a 
period of fourteen to sixteen days. Occasionally rabbits are en- 
countered that, for some unknown reason, are but incompletely 
sensitized by this method. In some experiments therefore three 
sensitizing doses, a subcutaneous, intraperitoneal and intra- 
venous, were given at two to three day intervals, in the hope of 
effecting more complete sensitization. The anaphylactic dose 
was always injected intravenously; into the external jugular 
vein in the case of the guinea-pig, and into the marginal ear vein 
in the instance of the rabbit. The minimal anaphylactic dose 
of the specific antigen was determined in each series of experi- 
ments without and with the previous administration of quinine 
hydrochloride subcutaneously, in doses varying between 100 and 
300 mgm. per kilo. 

Tables 1 and 2 show the extent to which quinine influences the 
course of the anaphylactic response. In table 1 are given some 
of the results obtained with rabbits, while those obtained with 
guinea-pigs are presented in table 2. Series A of table 1 shows 
that while the minimal lethal dose of the reinjected antigen in 
sensitized non-treated animals is about 2 cc., that in animals pre- 
viously treated with quinine is about 0.2 cc., giving a ratio of 
about 10:1. In series B and C of the same table, in which the 
rabbits were refractory to the specific protein, so that in the un- 
treated animals at least 3 cc. of the serum were required to pro- 
duce symptoms, animals previously treated with quinine suc- 
cumbed to doses of 1 or 1.5 cc., thus presenting a ratio of 3:1 or 
over. . 

Reference to table 2 will likewise show a marked difference in 
the reaction to the reinjected antigen between normal sensitized 
animals, and those previously treated with quinine. Thus the 
minimal lethal dose of the reinjected antigen in guinea-pigs sensi- 
tized to horse serum is 0.2 cc. per 100 grams of body weight, 
while in those treated with 100 mgm. per kilo of quinine hydro- 
chloride subcutaneously the dose is 0.075 cc. per 100 grams 
series A.B). Similarly non-treated animals sensitized to ox serum 
succumb to 0.03 to 0.04 cc. per 100 grams thereof, while animals 
treated with quinine die in anaphylactic shock from doses of 


SMITH 


MAURICE I. 


242 


suoydurds aloavg | ¢'() 00z 0c | 06 
sInoy [BIeAVS UI paid | O'T 00G 06 | 08 
suroydurAs a10A09 Or 00z 0z os 
imoy [T Ul perd | 04 OSG 02 | OF 
SognUlUl 7 UI peld | OG 006 ~ | O08 
sinoy [viaAes Ul pslq | 0% 0SZ 1G |0e 
SINOY [BIOAVS UI poiqd GAT O08 Yi OT 
suioyduixg | Q'T 00g 0G |-0'T 
soynuIul Z UL pslqd | 0'Z 00g OG 20-1 
aynuIul [ Ul psig | FO 008 LE ROG 
sanoy [B1dA0S UI pold ZO 008 Mil 0G 
samoy [BViaAes UI paIq | 70 00€ 9T | 06 
sinoy [B419A0S UL psIg | ZO 00S 9T 0% 
suioydumkg | T'¢0 00z 91 | 0G 
sumogduxg | [0 00z IIE] Whee 
pone A Be dey) es 

asop potsed | oasop 

4[Nsayy Sent RES aaa BUIZt} 


-1suag | -Wedxny 


8G 
LG 
9% 


Iw 
G 


requinu 
quour 


suroydurAg 
suoydurks a.10A0g 
sutoydurds a.taAag 


suroydurscg 
sm0ydurAg 
sutoydursés ON 
suroydurAs ON 
aynurar [ Ul paiq 
suroydutds ON 


sanoy [B1eA0s UL poiq 
sanoy [B19AVS UL poiqd 
sumoydursg 

soynurlul Z UL pod 
suroydwuAs ON 
suroydurds ON 
suroydurds ON 


aAsoy 


ATSOOFUNVLINOHONS [OH ANININD 


T Tava 


ail 61 0'€ 
0'§ ST 0O'€ 
0G H! 0'€ 
O'& 1 OT 
cs IZ OT 
Us 16 OT 
Gi 0G OT 
0€ 0z | 
0% 0z OT 
0G LT 0G 
0G oT 0G 
0% 9T 0G 
0G 91 0% 
OT 91 0% 
g°0 9T 0G 
10 oT 0% 
ie sfiipp ‘99 
asop porsed 9sop 
peqoot uory BuUIz1} 
gla, | -Bqnouy | -1suag 
ANINIOD ON 


aunumnb fig payipow sp spiqqn. pazryisuas ur wabyun pajoaluras fo fipvorwo,y, 


Noto Oo wt 
NANNANN OH OO 


IG 


re Om Ho © 
oD OD sH SH SH SH 
——— ee Se ——— 
aa) 


Jequinu 


quouw 
-119d xf 


SdL0g 


245 


STUDIES IN ANAPHYLAXIS 


_—_——— ere 


soynurur g UL perqT | 800°0 02 Z'0 IZI 
soynuru g ur psi, | 0100 02 Z°0 ZI suojdurkg | ¢0°0 02 3'0 LIT ‘Ss 'O 
soynurur g Ul pol, | 0z0'°0 0 Z'0 611 soynuru f ur por | £00 02 Z°0 OT ai 
soynurm F Ul pod | g0'°0 ra gO 98 
suroydurkg | 70'0 ar ¢'0 8 
soynur F UL psd | ¢0'O rat G0 8 
suroydurkg | 200°0 GT ¢'0 1OT soynurm ¢ ul psd | $0°0 91 10 08 
suroydurAg | 900°0 cI ¢'0 O01 suioydurss ON | 10°0 91 10 6L ‘SO 
suoydurAg | ¢00°0 GT ¢'0 66 soynur F Ul par | ¢o'd 91 10 8 @) 
soynurul F UI psi | g00'0 GI ¢'°0 86 soynulM ¢ Ul psd | OL'O 91 1'0 LL 
suroydurdg | ¢00°0 91 10 8 soynurur g Ul pad | O10 OL 10 92 
smojyduAg | ¢09'0 9 mo Z8 soynurul ¢ UL por 0z°0 9 10 cL 
soynurm F Ul psiq | O10'O 91 10 18 soynur Zul pol | 0Z'0 91 10 PL 
suoyduAs pir | OL'd IZ gO 6F 
sumoydurAs ppl | Glo 1Z cy) QP ‘SH 
oynural T Ul ped | OOT'O 1% ¢°0 og soynurul ¢ Ul parq | Ez" IZ 01 PE a 
sopnural QT Ur pord | ¢20°0 IZ Or 68 soynuIl GE ul psd | 0z'0 IZ Or ee 
suroydurds pry | O10 PI ¢'0 0g 
suroydurds pl | 100 PI ¢'0 62 
suroydurkg | 0g'0 a OT 8Z ed 
suioydurAs ON | O10°0 pI ¢'0 1g oynurur T ur paid | O80 tT Or Za V 
soynurut 6 Ul por | OOT'O iat o'r Ke suroydurés ON | OL'O PI OT £Z 
soynurur Z UL Prd | 00Z'0 v1 OT o% suroydursés ON | 020 v1 O'T 2% 
oot tod "0 any oo Pe ane ie 
oe porsot | Pub. [eon san] “aon” - nek te aeaaal Saeco ston ceils 
-UleyY -Bqnouy sjeues -110d xny -UloYy -eqnouy -Isuag “10d xo 
ATSQOANVLINOGOS NWVUNOTIY UAd SNVUDITUN OOL ‘oH GANINIOO ANININO ON 


ee ee ee ee ee) ee ee ee ee ee a ee ae 
auvuinb fiq pafrpow sv shid-vouinb ur wabyun pajoaluras fo hpoveo,y, 


@ ATAVL 


THE JOURNAL OF IMMUNOLOGY, VOL. Vv, No. 3 


244 MAURICE I. SMITH 


0.008 ce. per 100 grams, or less (series C and D). Thus quinine 
increases the susceptibility of sensitized rabbits and guinea-pigs 
from 3 to 10 times to the effects of the reinjected antigen. 

The incubation period has not been found to be materially 
altered by quinine. That is to say, if the incubation period in a 
given instance was too short for sensitization to be effected, and 
no symptoms were produced by the reinjected antigen, the previous 
administration of quinine did not alter the reaction. In some 
experiments on the rabbit to determine this point, it was found 
that not until eight or nine days have elapsed since the primary 
injection of the foreign protein did quinine influence the reaction 
to the reinjected antigen. This also appears to be the minimal 
period for effecting sensitization. 

The question that naturally presented itself was, what is the 
cause of the peculiar influence of quinine on the course of the 
anaphylactic reaction? Why do animals sensitized to a foreign 
protein react with such a degree of hypersensitiveness to the re- 
injected antigen if previously treated with quinine, even in doses 
too small to show any definite symptoms, while in non-sensitized 
animals treated with large doses of quinine the injection of a 
foreign serum produces no effects whatever? 

Experiments were undertaken in the hope of finding an answer 
to the above questions on the basis of the prevailing theories of 
the mechanism of anaphylactic shock. In the main there are 
two theories for the explanation of this phenomenon. Briefly 
stated, they are the humoral or chemical theory, and the cellular 
or physical theory. The former assumes that the anaphylactic 
reaction is an intoxication brought about by toxic substances 
elaborated in the blood stream of the sensitized animal as a result 
of interaction between antigen and antibody, it being held that 
the latter possesses the nature of an enzyme. ‘The cellular theory 
assumes that the reaction is due to attached cellular antibody 
binding the antigenic substance. 

As stated, in terms of the humoral theory anaphylactic shock 
is an intoxication, the poison being produced by the digesting 
action of the specific proteolytic enzyme on the reinjected anti- 
gen (Vaughan 5). More recently the view has been expressed 


i i ai Be Ow 


STUDIES IN ANAPHYLAXIS 245 


that it is the proteins of the serum of the sensitized animal that 
are split rather than those of the reinjected antigen (Jobling, 
Petersen, and Eggstein 6). The last named authors assume that 
the proteolysis is caused by a deficiency of the antiferments in 
the serum of the sensitized animal. 

Since quinine is known to have a powerful influence on all 
ferment action, it seemed probable that an investigation into 
the influence of this substance on the rate of proteolysis resulting 
from the interaction of sensitized serum and specific antigen 
might not only explain the altered susceptibility of sensitized, 
quinine-treated animals to the reinjected antigen, but it might 
also add further evidence in favor of the humoral conception of 
anaphylaxis. With this aim in view rabbits and guinea-pigs 
were sensitized in the usual manner to ox serum. After a suffi- 
ciently long incubation period the animals were bled, the sera 
collected, and definite amounts thereof were treated with definite 
amounts of the antigenic serum, and were incubated for some 
hours at 38°C., without or with the addition of known amounts 
of quinine. The degree of proteolysis in the various specimens 
was determined by a chemical analysis of the non-coagulable 
nitrogen in the respective sera, and in the mixtures after incu- 
bation. 


The method employed for the determination of the non-coagulable 
nitrogen was essentially that of Folin and Wu (7), with some slight 
modifications which became necessary in the course of this work. 
Briefly the method employed was as follows. One cubic centimeter 
of the respective sera or mixtures thereof was pipetted into a large 
pyrex glass test tube, and the coagulable proteins were precipitated 
by the addition of 1 ce. of 10 per cent acetic acid in 20 per cent sodium 
chloride solution, and gently boiling this for about five minutes. This 
deviation from Folin’s method was necessary since it was desirable 
to look not only for the non-protein nitrogen, but also for that of the 
higher split products. The boiled mixture was filtered, the precipitate 
washed with boiling distilled water, and the filtrate made up to a defi- 
nite volume. An accurately measured portion thereof was used for 
digestion in pyrex glass test tubes. Here another deviation from the 
Folin method had to be made. The digestion could not be successfully 


246 MAURICE I. SMITH 


carried out by means of the phosphoric-sulphurie acid mixture 
recommended by Folin; presumably the acetic acid interfered. The 
digestion mixture employed, therefore, consisted of sulphuric acid, 
copper sulphate, and sodium sulphate, as recommended by Peters (8). 
The digested mixture was allowed to cool, about 10 cc. of water added, 
then 2 cc. of 5 per cent Rochelle salt to prevent turbidity (8), 10 per 
cent sodium hydrate added to slight alkalinity, 5 ec. of the undiluted 
Folin-Nessler reagent, and water to make a definite volume. The 
amount of nitrogen was determined by comparing the color of this 
solution in a Duboseq colorimeter with that of a definite amount of 
nitrogen in the form of ammonium sulphate, treated similarly in every 
respect, except for the digestion, which was omitted. 


The results obtained in these experiments are given in table 3. 
Most of the experiments were made in duplicates. The amounts 
of sensitized serum and of the antigenic serum used are indicated 
in their respective columns. Wherever quinine was used the 
amount indicated as the hydrochloride was dissolved in 1 ce. of 
water. The same volume of water was also added to those mix- 
tures in which no quinine was employed, except in experiment 26, 
in which no water was added. In the ‘‘found” column the fig- 
ures represent the amounts of nitrogen per cubic centimeter of 
the mixture, actually obtained by analysis. ‘The figures given in 
the ‘‘caleulated”’ column represent the amount of nitrogen per 
cubic centimeter of the mixture, theoretically calculated, including 
the theoretical amount of nitrogen in the quinine, wherever this 
was added.” Yo illustrate: In experiment 12, 2 ce. of sensitized 
rabbit’s serum containing 1.12 mgm. of nitrogen per cubic centi- 
meter were mixed with 1 ec. of ox serum containing 0.52 mgm. of 
nitrogen per cubic centimeter. One cubic centimeter of water 
containing 5 mgm. of quinine hydrochloride was added to this 
mixture, and the whole was incubated at 38°C. for four hours. 
One cubic centimeter of this mixture after incubation was found 
on analysis to contain 0.77 mgm. of non-coagulable nitrogen. 
The calculated amount of nitrogen for 1 ce. of this mixture is 
21.12+0.52+0.3 (the theoretical equivalent of nitrogen in 5 


2 A blank experiment was made to determine the nitrogen content of the qui- 
nine used, and it was found to be practically identical with the theoretical figure. 


. | STUDIES IN ANAPHYLAXIS 247 


) mgm. of quinine HCI.3H.O), and the whole divided by 4, giving 
| 0.77 mgm. 

In series A of this table the sensitized serum was obtained by 
| exsanguination under ether of a rabbit that had been sensitized 
to ox serum in the usual manner. The incubation period was 
sixteen days. This serum contained 1.12 mgm. of non-coagulable 
nitrogen per cubic centimeter. The antigen used in this and in the 


TABLE 3 


Non-coagulable nitrogen of sensitized serum incubated with antigen and quinine 


NON-COAGULABLE 


EXPERI- QUININE NITROGEN PER CUBIC 
SERIES Prin ait yee ie ed eceriis passer akY INCUBATED CENTIMETER 
Found | Calculated 
ce. cc. mgm hours mgm mgm, 
(oe 2 1 0 4 0.70 0.69 
12 2 1 5 4 0.77 0.77 
13 1 2 0 4 0.61 0.54 
A 14 1 2 5 4 ONE 0.62 
15 5 15 0 4 0.60 0.64 
16 5 15 10 4 0.65 0.66 
17 5 15 0 0 Oot 0.64 
B 20 3 2 0 3 0.80 0.87 
21 3 2 5 3 1.00 0.92 
C 22 5 5 0 4 0.96 1.16 
23 5 5 5 4 1.00 1.18 
D 24 2 2 0 4 0.80 0.77 
25 2 2 5 4 0.87 0.83 
E 26 2 1 0 4 0.58 0.56 
| 27 iL 2 i 4 0.40 0.51 


next series contained 0.52 mgm. per cubic centimeter. The sensi- 
tized serum used in series B was obtained from a rabbit similarly 
treated, but bled without anesthesia. This serum contained 
1.40 mgm. of nitrogen per cubic centimeter. In series C the sen- 
sitized serum was obtained from another rabbit, similarly treated, 
and bled under ether. The non-coagulable nitrogen content of 
this serum was 1.75, while that of the antigen used in this and in 


248 MAURICE I. SMITH 


the subsequent series was 0.80 mgm. per cubic centimeter. The 
sensitized sera of series D and E were obtained from pooled blood 
of three guinea-pigs each, that had been sensitized to ox serum, and 
bled without anesthesia after an incubation period of thirteen and 
fifteen days respectively. The sensitized serum of series D con- 
tained 1.12 mgm. of non-coagulable nitrogen per cubic centimeter, 
and that of series E contained 0.44 mgm. per cubic centimeter. 

It will be seen that the amounts of non-coagulable nitrogen 
found in the various mixtures approach very closely the calculated 
amounts for those mixtures. In some of the experiments (13, 14, 
and 21) the figures are somewhat higher, while in several experi- 
ments they are decidedly lower than the calculated amounts. In 
some of the mixtures in which the ‘‘found”’ figures are lower 
than in the corresponding “calculated”’ figures, precipitates had 
formed on incubation, the mixtures were centrifugalized, and the 
clear serum was used for analysis. 

De Kruif and German (9) in studying the relation between the 
production of anaphyiatoxin under the influence of agar, and 
serum autolysis as measured by the aliphatic amino acid nitrogen 
in the serum, also observed a diminution of the latter upon incu- 
bation with agar. They ascribe this diminution to adsorption 
of some of the amino acids by the agar. Their attempts to re- 
cover this supposedly adsorbed amino acid was not successful. 
A more plausible explanation, for these experiments at least, 
would seem to be that some of the non-coagulable nitrogen might 
have been thrown down in the precipitate, perhaps in some 
chemical combination. 

The present results do not give evidence of proteolysis taking 
place as a result of interaction between sensitized serum and 
antigen, whether such mixtures be treated with quinine or not. 
The altered susceptibility of sensitized animals to reinjected 
antigen when under the influence of quinine, is therefore not 
explainable on the basis of the action of quinine on ferments. 
Nor do these experiments lend support to the theory that pro- 
teolysis is an essential factor in the causation of anaphylactic 
shock. 


STUDIES IN ANAPHYLAXIS 249 


The evidence heretofore adduced in favor of proteolysis result- 
ing from the interaction of antigen and antibody has been rather 
contradictory. Zunz and Gyorgy (10) determined the amino acid 
nitrogen of the serum of guinea-pigs, rabbits and dogs that had 
been sensitized to ox serum, protalbumose, and heteroalbumose. 
They incubated the sensitized sera with the respective specific 
proteins at different stages of the incubation period, and deter- 
mined the aliphatic amino acid nitrogen in the incubated mixtures. 
In some cases they obtained an increase, in others a decrease. 
The results, indeed, were so variable in the various sets of their 
experiments that the authors are not willing to draw any con- 
clusions, and state that the changes in the proteolytic power of 
sensitized blood do not suffice to explain the phenomenon of 
anaphylaxis. Auer and Van Slyke (11) determined the amino 
acid nitrogen content of the lungs in sensitized guinea-pigs be- 
fore and during shock, and found no difference. De Kruif and 
German (loc. cit.) failed to get an increase of the amino acid 
nitrogen in the serum of rabbits sensitized to egg white, and in- 
cubated with the latter at 38°C. Jobling, Petersen and Eggstein 
(loc. cit.) determined the non-coagulable nitrogen in sensitized 
dog’s serum before and during shock, and found an increase in 
the latter. They believe the proteins of the animal’s own serum 
to undergo proteolysis owing to disturbed ferment antiferment 
equilibrium produced by the antigen. There are undoubtedly 
other factors than anaphylactic shock that may alter the amount 
of non-coagulable nitrogen in the blood, and it is not impossible 
that the changes noted by these authors may have been the effect 
rather than the cause of shock. Manwaring, Kusama, and 
Crowe (12) in an investigation of the fate of the foreign protein 
in acute anaphylactic shock conclude that there is no evidence 
of a measurable destruction of the foreign protein by the blood 
serum of the sensitized animal. 

Failing to find an explanation for the peculiar influence of 
quinine on the course of anaphylactic shock on the basis of the 
humoral theory, experiments were next undertaken in a different 
direction. Coca (13) has recently shown that upon perfusion of 
the pulmonary vessels of sensitized rabbits with Locke’s solu- 


250 MAURICE I. SMITH 


tion containing some of the antigen, there is a marked obstruc- 
tion to the flow of the perfusion fluid, for the pressure required 
to initiate a flow of the perfusion fluid under these conditions was 
found much higher than when such a fluid was perfused through 
the vessels of a normal animal. Coca regards anaphylactic shock 
in the rabbit as a cellular reaction to the antigen, especially of the 
pulmonary arterioles, resulting in pulmonary obstruction, and a 
consequent dilatation of the heart. I have been able to confirm 


TABLE 4 


Effect of antigen and of quinine on the caliber of the pulmonary vessels in normal 
and sensitized rabbits 


HYDRO- | OUT- 
EXPERI- STATIC |FLOWIN 
OUTFLOW 
MENT PRESSURE;| 3 MIN- ox QUININE ; 
NID ; IN 3) N 
NUM- ZANE WEIGHT) RinGER- | UTES | SERUM HCl aie AE SESS 
BER LOCKE | AVER- 7 : 


SOLUTION AGE 


i) 


Sensitized IVE ¢ 25 40 
1: 400 0 14 | Discontinued 


(ae) 
ou 
[e,0) 
iw) 


3 Sensitized 1.8 


1:40 0 30 | Discontinued 


4 Sensitized 2.5 45 1 
1:40 0 58 Discontinued 


the above observation, by means of a somewhat different 
technique. 


Briefly my experiments were as follows. Sensitized rabbits were 
bled from the carotid artery, a cannula was inserted into the left pul- 
monary artery, and one into the left auricle. Perfusion of Ringer- 
Locke solution at 38°C. under a constant hydrostatic pressure was 
immediately instituted, while the lungs were rhythmically inflated 
about 25 times per minute. When the return of the perfusion fluid 
remained constant, a definite amount of the antigen (ox serum) was 


SS ——* ~ 


eS ee 


a es ee eee 
’ t 


STUDIES IN ANAPHYLAXIS 251 


added and the rate of outflow noted. Since serum is known to have 
an effect on smooth muscle, control experiments were made on normal 
animals and it was found that as high a concentration as 1:40 of ox 
serum in Ringer-Locke solution had no influence on the rate of flow of 
the perfusion fluid through the pulmonary vessels. Higher concen- 
trations were not tried. If such a solution is perfused through the 
pulmonary vessels of rabbits sensitized to ox serum, a considerable 
diminution of the rate of outflow occurred, as will appear from the 
subjoined table 4, 


The effects of quinine on smooth muscle in certain situations is 
well known, and need not be discussed here. It was thought 
that quinine might exert an action on the smooth musculature of 
the pulmonary vessels in the same direction as the specific antigen 
in the sensitized animals and thus explain the augmented sus- 
ceptibility of sensitized animals to the specific foreign protein 
when under the influence of this drug. Quinine was found, how- 
ever, to have no appreciable effect on the caliber of the pulmonary 
vessels when perfused through them in such concentration as 
might attain in the blood stream. In higher concentrations it 
appears to relax the pulmonary vessels, rather than constrict 
them. The influence that quinine has on the course of anaphy- 
lactic shock cannot therefore be referred to the altered pulmonary 
circulation caused by the foreign protein in the sensitized animal. 


THE RELATION OF HISTAMINE (8-IMINAZOLYLETHYLAMINE) TO 
ANAPHYLAXIS 


In a recent important contribution on the occurrence of hista- 
mine in various animal tissues, and its probable réle in a variety 
of physiological and pathological processes, Abel and Kubota 
(14) suggest that this substance may also play an important 
part in the mechanism of anaphylaxis. They suggest that this 
may be the hypothetical “poison” producing the symptoms of 
anaphylactic shock. They indeed, express the view that his- 
tamine may be identical with Vaughan’s toxie cleavage product 
obtained from various proteins. Vaughan has long called atten- 
tion to the close similarity between the physiological action of 


252 MAURICE I. SMITH 


histamine as described by Dale and his collaborators, and the 
physiologic behavior of his protein poison (5). Thus Vaughan 
pointed out that the fatal dose of his purified “poison” is prac- 
tically the same as that of histamine for the guinea-pig. Both 
substances cause a lowering of the blood pressure in the dog. . 
The striking broncho-constriction, and lung immobilization in 
the guinea-pig in anaphylactic shock and the similar picture in 
histamine poisoning is a familiar thing. However, such super- 
ficial similarities in the symptoms cannot be taken as final proof 
of the identity of the two conditions producing them. In order 
to determine more definitely to what extent histamine is related 
to anaphylactic shock, experiments were undertaken with this 
substance to see whether it is capable of producing phenomena 
generally regarded characteristic of anaphylactic shock. It is 
assumed in these experiments that if anaphylactic shock is asso- 
ciated with proteolysis with histamine as one of its products, then 
the characteristic phenomena occurring during shock should also 
be produced by histamine. 

The experiments described below relate to the effects of his- 
tamine on body temperature; its influence on the coagulability 
of the blood; its relation to quinine; and its relation to desensi- 
tization or antianaphylaxis. 


a. Effects of histamine on body temperature 


The effects on body temperature produced by the injection of 
a foreign protein into an animal sensitized thereto have been 
described and studied by Pfeiffer (15), Friedberger and his col- 
laborators (16), Hashimoto (17), Smith (18), and others. It has 
been shown that rabbits sensitized to ox serum and reinjected 
therewith after the incubation period, will have a rise in body 
temperature if the dose is small, and a fall in body temperature if 
the dose is large but sublethal (18). To determine whether his- 
tamine produces similar changes in body temperature, rabbits 
were injected intravenously with this substance in varying doses, 
and their temperature taken for some time thereafter. The 
minimal lethal dose of histamine on intravenous injection in 


STUDIES IN ANAPHYLAXIS 253 


the rabbit was found to be 0.4 mgm. per kilo. When histamine 
was injected in amounts from 0.01 to 0.3 mgm. per kilo, in single 
or repeated doses, no appreciable and constant effect was pro- 
duced on the temperature of the rabbit. With the smaller doses 


.no symptoms whatever could be elicited, while with larger doses 


(0.1 to 0.3 mgm. per kilo) there was considerable respiratory dis- 
tress, contraction of the pupils, muscular weakness and convul- 
sions, followed by recovery in a few minutes. No definite effect 
on body temperature could be elicited even with these massive 
doses. 


b. The influence of histamine on the coagulability of the blood 


The diminished coagulability of the bood of animals in ana- 
phylactic shock is well known. This phenomenon of anaphy- 
lactic shock is most pronounced in the dog, to a less extent in the 
rabbit and the guinea-pig. The coagulation time of the blood 
in shocked rabbits is however sufficiently prolonged from the 
normal to make this animal of service in comparative studies of 
this phase of anaphylaxis. As a result of several experiments in 
which the same procedure was employed to determine the com- 
parative coagulability of the blood in rabbits the following con- 
clusions have been arrived at. The blood of normal rabbits 
coagulates in from five to seven minutes. The blood of rabbits 
in anaphylactic shock coagulates in from fifteen to twenty min- 
utes. The coagulation time of the blood of rabbits shocked with 
histamine does not materially differ from the normal; indeed, 
it appears to be somewhat shorter than the normal. 


c. The relation of quinine to histamine 


It was shown in the earlier part of this paper that quinine pro- 
foundly influences the susceptibility of sensitized animals to the 
specific protein. If anaphylactic shock is to be interpreted in 
terms of the liberated histamine as a result of proteolysis or any 
other mechanism, it seems reasonable to suppose that quinine 
should equally alter the susceptibility of animals to this sub- 
stance. In other words, there should be a considerable degree of 


254 MAURICE I. SMITH 


synnergy between quinine and histamine. Accordingly, normal 
and sensitized rabbits and guinea-pigs were treated with quinine 
hydrochloride subcutaneously in doses of from 100 to 300 mgm. 
per kilo, and in about a half hour were injected intravenously 
with histamine in varying doses. Quinine was not found to have 
any effect on the course of histamine intoxication. In fact, it 
was frequently noted that a toxic but not lethal dose of histamine 
injected into rabbits previously treated with quinine produced 
milder symptoms than a corresponding dose of histamine given 
alone. 

Of course, it may be argued that quinine causes the hypothet- 
ical poison (histamine) to be produced in greater amounts or at 
a greater rate, without necessarily having to act synnergistically 
with it. This is not probable, however, in the light of the results 
obtained from the experiments on the non-coagulable nitrogen 
in various mixtures of sensitized and antigenic sera. 


d. The relation of histamine to antianaphylaxis 


It is a well known fact that if a minute dose of a foreign protein 
be injected into an animal sensitized thereto, a state of antiana- 
phylaxis is induced, the animal becomes desensitized, and with- 
stands a lethal or several times the lethal dose of the foreign pro- 
tein. This phenomenon is so constant that it is often regarded 
as a criterion in the determination whether or not a given ana- 
phylactoid reaction belongs to the class of true anaphylaxis. 
Thus Besredka states that peptone injected into guinea-pigs 
sensitized to horse serum does not protect them from this foreign 
protein; and he uses this argument in controverting the con- 
tention made by some authors that peptone shock is ideytical 
with anaphylactic shock (1). 

It appeared justifiable to assume that if anaphylactic shock is 
caused by histamine liberated by proteolysis, and since the pre- 
liminary injection of a small dose of the specific foreign protein 
protects the sensitized animal from larger doses of this protein, 
then the primary injection of a small amount of histamine should 
equally protect the sensitized animal from a lethal dose of the 


STUDIES IN ANAPHYLAXIS 255 


foreign protein. The converse, that is to say, the protecting 
action of a desensitizing dose of a specific foreign protein in a 
sensitized animal against a lethal dose of histamine is not essen- 
tial, for it may be assumed that desensitization in some manner 
inhibits further proteolysis, but cannot protect the animal from 
a lethal dose of the preformed poison. 

Several experiments were made on rabbits and guinea-pigs to 
see whether preliminary injections of histamine into sensitized 
animals could protect them from the effects of the specific foreign 
protein, and whether desensitization would influence the toxicity 
of histamine. The results were entirely negative. Neither did 
the preliminary injection of histamine affect the course of ana- 
phylactic shock, nor did sensitization or desensitization in any 
manner influence the toxicity of histamine. 

It may be added that several experiments were also made on 
guinea-pigs to determine whether a given fraction of the lethal 
dose of histamine could replace a corresponding fraction of the 
lethal dose of the specific foreign protein. This was found to be 
the case. In one series of guinea-pigs sensitized to ox serum, the 
minimal lethal dose of the antigen was found to be 0.03 cc. per 
100 grams. The minimal lethal dose of histamine was determined 
at 0.02 mgm. per 100 grams. Fifty per cent of the minimal 
lethal dose of each administered at the same time proved fatal. 
Indeed, 50 per cent of the one combined with 35 per cent of the 
other proved fatal in some experiments. It would seem, therefore, 
that the point of attack of histamine and of the toxic factor in 
anaphylaxis is the same in the guinea-pig. This point was not 
investigated in the rabbit. It is not unlikely that a similar 
synnergy between histamine and the specific foreign protein 
might exist in this animal too, for Dale and Laidlaw (19) have 
shown that the cause of death from histamine in the rabbit is 
due to pulmonary obstruction leading to acute dilatation of the 
right heart, and a similar mechanism for anaphylactic shock in 
the rabbit is suggested by the experiments of Coca (loc. cit.), 
which I have been able to confirm. 


256 MAURICE I. SMITH 


SUMMARY AND CONCLUSIONS 


Guinea-pigs and rabbits sensitized to ox serum or horse serum, 
and treated subcutaneously with moderate doses of quinine pre- 
ceding the reinjection of the specific antigen, have their suscep- 
tibility increased from 3 to 10 times to the specific protein, as 
compared with control sensitized animals. 

No appreciable degree of proteolysis could be demonstrated to 
occur in vitro by treating sensitized serum with the specific anti- 
gen, whether incubated alone or with quinine. The augmented 
susceptibility of sensitized animals to the specific protein when 
under the influence of quinine cannot, therefore, be referred to 
the well known action of this drug on ferments. 

Quinine added to Ringer-Locke solution perfused through the 
pulmonary vessels of sensitized or normal rabbits does not cause 
any noticeable constriction of these vessels. Specific foreign 
protein added to Ringer-Locke solution perfused through the 
pulmonary vessels of sensitized rabbits produces pulmonary 
obstruction to a marked degree. The altered susceptibility of 
sensitized animals to the foreign protein produced by quinine 
cannot be referred to any synnergy between quinine and the 
anaphylactic process on the pulmonary circulation. 

It has been suggested that histamine might be the causative — 
factor of anaphylactic shock. On closer investigation it does not 
appear that histamine is identical with the anaphylactic process, 
for the following reasons: 

a. Histamine does not produce in animals the temperature 
reactions observed in anaphylaxis. 

b. Histamine does not alter the coagulability of the blood, as 
is noted in anaphylactic shock. 

c. Quinine alters markedly the course of the anaphylactic 
reaction, by augmenting the susceptibility of sensitized animals 
to the foreign protein; it has no harmful effect on the course of 
intoxication with histamine. 

d. There is no relation between histamine and antianaphylaxis. 
Neither does desensitization influence the toxicity of histamine, 
nor does the preliminary treatment with histamine alter the lethal 
dose of the specific foreign protein in sensitized animals. 


STUDIES IN ANAPHYLAXIS 2ay 


A synnergetic relation is shown to exist between histamine 
and the specific foreign protein in sensitized guinea-pigs. This 
is probably best explained on the assumption that some points of 
attack of histamine and of the anaphylactic process are identical. 


REFERENCES 


(1) BesrepKa; Anaphylaxis and Antianaphylaxis. 1919, Mosby & Co. 
(2) AvER: Jour. Exp. Med., 1910, 12, 638. 

(3) Mira: Zt. f. Immunitaetsf., 1911, 11, 501. 

(4) BANZHAF AND FAMULENER: Jour. Inf. Dis., 1910, 7, 577. 

(5) VAUGHAN: Protein Split Products. 1913, Leaand Febiger, Jour. Am. Med. 

Assn., 1914, 62, 583. 

(6) JoBLING, PeTERSEN AND Eaastein: Jour. Exp. Med., 1915, 22, 401. 
(7) Fotin anp Wu: Jour. Biol. Chem., 1919, 38, 81. 

(8) Peters: Ibid., 1919, 39, 285. ; 

(9) De Kruir anp GerRMAN: Jour. Inf. Dis., 1917, 20, 833. 
(10) Zunz anp GyOrey: Zt. f. Immunitaetsf. Orig., 1914-15, 23, 402; Ibid., 1913, 

17, 241; 265; 279. 

(11) AvER anD Van Stryke: Jour. Exp. Med., 1913, 18, 210. 
(12) Manwarine, Kusama, AND Crown: Jour. Immunol., 1917, 2, 511. 
(13) Coca: Ibid., 1919, 4, 219. 
(14) ABeL anp Kusora: Jour. Pharm. and Exp. Ther., 1919, 13, 243. 

(15) Prerrrer: Wien. Klin. Woch., 1909, 22, 14. 

(16) FRIEDBERGER anp Mira: Zt. f. Immunitaetsf., 1911, 10, 216. 
(17) Hasuimoro: Arch. f. Exp. Path. u. Pharmak., 1914, 78, 370. 

(18) SmitH: Jour. Lab. Clin. Med., 1915-16, 1, 902. 
(19) Date anp Larptaw: Jour. Physiol., 1910, 41, 318; Ibid, 1911, 43, 182. 


| 


_THE ANTIGENIC PROPERTIES OF HEMOCYANIN! 


CARL L. A. SCHMIDT 
From the Department of Biochemistry and Pharmacology of the University of 
California 


Received for publication April 12, 1920 


Comparatively little is known concerning the chemistry of 
hemocyanin, the copper-protein compound found in the circu- 
lating fluid of many Crustacea, Mollusca and other invertebrates. 
Functionally it appears to play the réle of oxygen carrier, being 
in this respect analogous to hemoglobin. Hemocyanin unlike 
hemoglobin is not confined to particular cells but appears free in 
the circulating fluid of which in many of these animals it forms 
the chief protein component. 

Concerning its chemical nature no unanimity exists since dif- 
ferent workers have obtained results with hemocyanin obtained 
from different sources which are at variance to an extent greater 
than usually found in proteins prepared from allied species (1) 
Henze (2) prepared crystalline hemocyanin from the circulating 
fluid of the octopus (Octopus vulgaris) by precipitation with am- 
monium sulphate according to the procedure of Hopkins and 
Pinkus. On dialysis a 3 per cent solution of hemocyanin was 
obtained. On addition of acid, part, but not all of the copper 
could be split off, but splitting in the sense that hemoglobin is 
split to yield a protein and a chromogenic substance did not take 
place. Henze believes, from a study of its reactions, that hemo- 
eyanin is a copper albuminate. 

Alsberg and Clark (3) carefully went over the ground covered 
by Henze but used the blood of Limulus polyphemus as a source 
of hemocyanin. Not only did their product differ from Henze’s 
with respect to the concentration of ammonium sulphate re- 

1 Aided by a grant from the George Williams Hooper Foundation for Medical 
Research. 

259 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 3 


260 CARL L. A. SCHMIDT 


quired for precipitation, it behaving in this respect like a globu- 
lin, but they also found a marked difference in elementary com- 
position and a decidedly lower copper content. Like Henze, 
they were unable to obtain from Limulus hemocyanin a substance 
analogous to hematin. It appears that the hemocyanins from 
these two animals are distinctly different and this conclusion is 
supported by a comparison of the stability of the oxygen com- 
pounds, hemocyanin from Limulus being but little dissociated 
(4) under diminished pressure while that from the octopus gives 
up its oxygen more readily. Dhéré (5) has also noted differences 
in the copper content and oxygen capacity of the blood of various 
invertebrates. 

Immunological experiments with pure proteins have shown 
that antigenic property is intimately connected with their chemi- 
eal structure. Thus amino acids, polypeptides, protamines, his- 
tones, certain proteoses, and gelatin fail to give rise to demon- 
stratable immune bodies when repeatedly injected into animals. 
Gliadin from wheat and hordein from barley resemble each other 
in their chemical makeup and are closely related immunologic- 
ally (6). The immunological method can at times be used for 
the purpose of gaining further insight into the chemical makeup 
of a particular protein. Gross chemical analyses yield but little 
information since they do not take into consideration the stereo- 
chemical configuration in the protein molecule, nor is it at times 
possible by chemical methods to demonstrate differences in closely 
related proteins which by immunological methods can readily be 
shown. However it is of interest in this connection to mention 
that a high histidin content is characteristic of both hemoglobin 
and hemocyanin (7). 

It has been shown that neither globin (8), hematin or their 
combination, as in hemoglobin (9), is antigenic. The inability 
of globin to give rise to immune bodies when injected repeatedly 
into animals cannot be attributed to lack of aromatic amino 
acids (10) and in this respect differs from gelatin, which, accord- 
ing to an hypothesis (11), is non-antigenic because of this de- 
ficiency. 


ANTIGENIC PROPERTIES OF HEMOCYANIN 261 


It appeared to us that the immunological method might be 
used to establish the relationship of the hemocyanins derived 
from different sources and also to throw further light on its 
chemical makeup. Nuttall (12) and von Dungern (13) were 
able to immunize rabbits against the circulating fluid of cephalo- 
pods and other invertebrates and von Dungern and Cohnheim 
(14) found that copper was contained in the specific precipitate 
obtained with octopus serum, indicating that part of the precipi- 
tate consisted of hemocyanin.2. While this appears to indicate 
that hemocyanin is antigenic, it does not exclude the possibility 
that hemocyanin from other species may not possess this pro- 
perty especially in view of the differences in chemical behavior. 
Unfortunately for the purpose in mind neither octopus nor limu- 
lus blood was available and so no biological comparison could be 
made. As a source of hemocyanin use was made of the abalone 
(Haliotis) previously also used as a source for taurin (15), and 
the fluid was obtained in the manner recently described by 
Myers (16). 

The opaque fluid was saturated with oxygen, filtered and an 
equal volume of saturated ammonium sulphate was slowly added, 
the fluid being constantly shaken to prevent local zones of high 
ammonium sulphate concentration. The blue precipitate was 
filtered off on hardened filter paper, then dissolved in a large 
volume of distilled water, and saturated ammonium sulphate 
again added to make a concentration of 4 cc. to each 10 ce. of 
solution. The precipitate was again filtered off, dissolved as 
before and reprecipitated by addition of ammonium sulphate to 
a concentration of 4.3 cc. per 10 cc. of fluid. The precipitate 
was filtered, redissolved, centrifuged to remove a trace of insolu- 
ble matter, and dialyzed first against running and then distilled 
water, toluol being used as a preservative. The salt-free solution 
contained a small amount of a white precipitate, probably hemo- 
eyanin from which the copper had been split. This was removed 
by centrifuging. The hemocyanin solution was concentrated by 
blowing warm dry air over the surface, made isotonic by addition 


2 These experiments came to my notice after the present work had been com- 
pleted. © 


262 CARL L. A. SCHMIDT 


of sodium chloride and for the purpose of injection preserved by 
the addition of phenol to a concentration of 0.25 per cent. The 
portion used for the final tests was preserved at a low tempera- 
ture without addition of preservative. 

It will be noted that with respect to its behavior towards am- 
monium sulphate this preparation of hemocyanin was the same 
as that obtained by Alsberg from Limulus polyphemus. It might 
also be of interest in passing to mention that in comparison with 
the hemocyanin content of the fluid, the amounts of albumin 
and globulin were very small and it is very possible part of these 
consisted of hemocyanin. 

Four normal rabbits were immunized by giving each 8 injec- 
tions in doses of 80 mgm., a period of five days elapsing between 
the fourth and fifth injections. Ten days later the animals were 
bled, the sera inactivated and used for the subsequent tests. 
Fixation experiments using the well-known hemolytic system in 
the manner described in previous work with pure proteins (17), 
gave positive results, the limits of fixation, in terms of serum 
dilution, being as follows: Rabbit no. 25, 0.1 ec. of 1:250; rabbit 
no. 27, 0.1 cc. of 1:10; rabbit no. 29, 0.2 cc. of 1:250; rabbit no. 
23, 0.4 cc. of 1:50. The usual controls were run to eliminate 
the possibility of inhibition of hemolysis by factors other than 
antibody. Positive precipitin tests were likewise given by the 
above sera. Addition of hemocyanin solution to a suspension of 
sheep red cells results neither in agglutination nor hemolysis of 
the red cells. 

Three normal guinea-pigs were sensitized by giving each 8 
mgm. subcutaneously and five weeks later were reinjected with 
results as follows: No. 1, 110 mgm. intravenously, death resulted 
in about ten minutes, symptoms typical of anaphylaxis; no. 2, 
100 mgm. intraperitoneally, animal became very sick, tempera- 
ture dropped 5°C., death resulted six hours after injection; no. 
3, 100 mgm. intraperitoneally, slight drop in temperature, no 
visible symptoms noted. 

Hemocyanin, like hemoglobin, is non-toxic. ‘Two guinea pigs 
were given respectively 75 and 90 mgm. of hemocyanin solution 
intracardially and a third animal received 105 mgm. intraperi- 


ANTIGENIC PROPERTIES OF HEMOCYANIN 263 


toneally. Visible symptoms were not shown and the tempera- 
ture variation was within the normal variability. 

These experiments furnish direct evidence that hemocyanin is 
antigenic and confirm the previous work of von Dungern and Cohn- 
heim. Its chemical makeup must be very different from hemo- 
globin since the latter is non-antigenic. 


SUMMARY 


1. Hemocyanin prepared from the circulating fluid of the aba- 
lone (Haliotis) was precipitated by ammonium sulphate within 
the limits found by Alsberg and Clark from Limulus polyphemus. 
It appears to be a globulin. 

2. The sera of rabbits immunized with this preparation of 
hemocyanin gave positive fixation and precipitin tests. It was 
toxic for guinea-pigs previously sensitized to this substance but 
not toxic for normal animals. 

3. These experiments support the chemical viewpoint that the 
chemical makeup of hemocyanin is very different from hemo- 
globin since the latter is non-antigenic. 


REFERENCES 


(1) Kosset, A., anp Kutscuer, F.: Z. physiol. Chem., 1900-1901, 6, 31, pp. 165- 
214, have shown that considerable differences exist in the amino acid 
content of salmin, clupein, cyclopterin, and sturin. Nevertheless these 
substances give the general reactions characteristic of the protamines. 
The differences in composition of the proteins of horse and ox serum 
noted by Harttey, P., Biochem. J., 1914, 8, pp. 541-552, are not strik- 
ing, and GorTNER, R. A., AND WuERTzZ, A. J., J. Amer. Chem. Soc., 
1917, 39, pp. 2239-2242, find that within the experimental error the com- 
position of fibrin obtained from various animals is the same. 

(2) Henze, M.: Z. f. physiol. Chem., 1901, 33, pp. 370-384. 

Henze, M.: Z. f. physiol. Chem., 1904, 43, pp. 290-298. 
(3) Ausprre, C. L., anp Cuark, E. D.: J. Biol. Chem., 1910, 8, pp. 1-8. 
AusBERG, C. L.: J. Biol. Chem., 1914, 19, 77-82. 
(4) Atspere, C. L., anp Cuark, W. M.: J. Biol. Chem., 1914, 19, 503-510. 
(5) Dur, C.: J. Phyiol. et Path. gen., 1915, 16, 985-997. 
Duére, C.: J. Physiol. et Path. gen. 1919, 18, 222-243. 
(6) Wetts, H. G:, anp Osporne, T. B.: J. Infect. Dis., 1911, 8, 66-124. 
(7) Van Styxe, D. D.: J. Biol. Chem., 1911-12, 10, 15-55. 


264 CARL L. A. SCHMIDT 


(8) Gay, F. P., anp Ropertson, T. B.: J. Exp. Med., 1913, 17, 535-541. 
Scumipt, C. L. A.: Univ. Cal. Pub. Path., 1916, 2, 157-204. 
Brownine, C. H., anp Witson, G. H.: J. Path. & Bact., 1909, 14, 174-183. 
(9) Forp, W. W., ann Hatssy, J. T.: J. Med. Res., 1904, 11, 403-425. 
LEVENE, P. A.: J. Med. Res., 1904, 12, 191-194. 
Scumipt, C. L. A., anp Bennett, C. B.: J. Infect. Dis., 1919, 25, 207-212. 
BrabDLey, H. C., anp Sansum, W. D.: J. Biol. Chem., 1914, 18, 497-506. 
(10) ScsrrreNHELM, A., AND WeIcHarp?, W.: Z. Immfrsch. & exp. Therap. Orig., 
1912, 14, 609-636. 
(11) Weuts, H. G.: J. Amer. Med. Assoc., 1908, 50, 527-528. 
(12) Nurrauy, G. H. F.: Blood Immunity and Blood Relationship, Cambridge, 
1904, p. 358. 
(13) von DuncErN, E.: Centralbl. Bakt. Ref., 1904-1905, 35, 253. 
(14) von DuncerN, E., anp ConnuumM, O.: Die Antikérper, Jena, 1903. 
(15) Scumipt, C. L. A., anp Watson, T.: J. Biol. Chem., 1918, 33, 499-500. 
(146) Myers, R. G.: J. Biol. Chem., 1920, 41, 119-135. 
(17) Scumipt, C. L. A.: Univ. of Cal. Pub. Path., 1916, 2, 157-204. 


ON THE NATURE OF BACTERIAL TOXAEMIA! 


HANS ZINSSER 
Department of Batteriology, Medical Department, Columbia University 


Received for publication April 14, 1920 


I 


On an occasion like this, when opportunity arises for students 
of immunity to meet for the discussion of their mutual problems, 
it seems fitting that, at least once during the meetings, the de- 
liberations should take the form of a reéxamination of some of 
the fundamental premises on which our science is based. It is 
the business of most of us here to cast about continually for new 
leads, to leave the main highways of established fact and to fol- 
low more faintly marked paths to their endings. True progress 
‘can be made, and time and effort justified, only if we are sure, at 
each stage of our investigations that, so far, we have travelled 
straight. I assumethatthere are many of us here who have wasted 
effort on numerous occasions in trying to build at the top when the 
base needed support; and we all learn to recognize that endless 
repetition of experimental facts already well established, by 
slightly different methods and new manipulations, will bear dry 
fruit for science. 

The astonishing progress which was made in the field of im- 
munity during the last twenty or thirty years, was, in the main, 
based upon the careful and detailed ‘‘following through” of 
fundamental observations made by Pasteur, Behring, Richet, 
Metchnikoff, Bordet, Ehrlich, Theobald Smith and a few others. 
The application of these principles to the wealth of material 
which was made available by the rapid succession of etiological 
discoveries has kept students of this subject joyfully busy for 


1 Presidential address at the Annual Meeting of the American Association of 
Immunologists, April 1, 1920. 


265 


266 HANS ZINSSER 


two scientific generations. More recently, too, the possibilities 
of our field have been particularly enriched by the introduction of 
chemical and physical methods into the study of antigen-anti- 
body reactions and by a development of the remarkable group of 
phenomena which we classify together as anaphylaxis. It is 
unnecessary to dwell upon the important yield of diagnostic, 
therapeutic, and prophylactic discoveries which has been con- 
tributed to medicine and biology as a result of these investiga- 
tions. Indeed, these things constitute our entire scientific capi- 
tal, and it is far from our intentions to minimize their great and 
fundamental importance. But it will be profitable, I think, at 
this stage of our development to sweep our eyes across the achieve- 
ments of the past and to recognize that after all, this great edifice 
of facts is built upon the foundation of a few great principles, 
and that we are as ignorant as we were twenty years ago, and 
perhaps a little more confused, regarding some of the most im- 
portant phenomena that accompany infection. 

In the diseases which are caused by bacteria that secrete true 
antigenic toxins, diphtheria, tetanus, and a few others, we have 
arrived at a fairly clear understanding of the mechanism by which 
the animal body is injured and of that by which it is protected. 

In practically all of the remaining infections, however, there is 
uncertainty concerning many of the fundamental factors of attack 
and of defense. And in order to progress, we must first systema- 
tize out ignorance so that new work may not “take off” from 
premises of vague opinion or uncritically accepted assertion. 

As regards the mechanism of injury, the problem about which 
there is most confusion is that concerning the nature of toxaemia. 
In practically all infectious diseases there are manifestations of 
local and general disturbances which cannot be attributed to 
mechanical injury by the growing bacteria. These take the form 
of fluctuations of temperature; of alterations, both quantita- 
tive and qualitative, of the leucocytes; of symptoms referable 
to the nervous system, such as depression, muscular and somatic 
pains, headache, nausea, weakness; and finally of local changes 
such as the inflammatory reactions in neighboring tissues, lym- 
phaties and joints. In the last named instances it is of course 


NATURE OF BACTERIAL TOXAEMIA 267 


debatable whether we are dealing with toxic disturbances rather 
than with local extensions of bacterial growth. But the toxic 
nature of the lymphangitis and oedema surrounding pyogenic 
foci or cellulitis is rendered likely by the speed with which these 
conditions subside upon drainage after incision, the proximal dis- 
appearance of the red lines marking lymph channels after tourni- 
quet application, and such observations as that recently made 
by Auchincloss, in the case of a severe haemolytic streptococ- 
cus infection, in which he incised an intensely oedoematous area 
on the dorsum of the hand above the lesion, the copious fluid 
from which proved sterile in blood plate cultures. 

It is hardly necessary to discuss these points at greater length, 
_ for it is obvious to all who have studied infections that the ele- 
ment of toxaemia, that is, the occurrence of disturbances remote 
from the actual seat of bacterial growth, is apparent in practically 
all diseases of bacterial origin, and there is a striking similarity 
in the nature of these manifestations, whatever the variety of 
infection. In such cases as typhoid fever, cholera, influenza, 
etc., the toxaemia is severe, while in surgical infections with pyo- 
genic cocci it is relatively feeble; or, again, it is almost entirely 
absent, as in anthrax where the infected animal remains in 
astonishingly unchanged general condition until the later pre- 
lethal stages, when the body is flooded with bacteria. Never- 
theless, except in a few instances like the last named, the toxic 
element is hardly ever entirely absent. Now when all is said 
and done, we know very little about the nature of the poisons 
which give rise to these disturbances, and about the mechanism 
by which they injure the body cells. 

As regards defense reactions against infections in which true 
exotoxins play no role, we are a little better off in that we are in 
possession of accurate analyses of the serum antibodies and of 
their reactions with the bacterial antigens. But this after all is 
only a small part of the story. The reasons for the limited suc- 
cess of passive immunization with antibacterial sera; the nature 
of the participation of fixed tissue cells, both in temporary and 
permanently acquired resistance; these are only a few of the 
problems that must be solved before we can possess a clear un- 
derstanding of the pathology of infection. 


268 HANS ZINSSER 


The obvious direction which empirical investigation has taken 
toward the solution of these problems has been the attempt to 
protect passively by the administration of sera with higher and 
higher antibody concentrations, in the hope that eventually 
this might change failure to success by quantitative improve- 
ment alone. To some extent it has seemed that this might 
actually solve the problems, inasmuch as the discovery of anti- 
genic subgroups of organisms like the pneumococcus, the men- 
ingococcus, and the streptococcus indicated that early failures 
might have been due to lack of specificity on the part of the 
sera. But though this recognition has led to partial success 
in such conditions as type I pneumonia, yet in principle the diffi- 
culties have remained unchanged. ‘To take the last named dis- 
ease aS an example, while a potent specific serum seems to pos- 
sess distinct value in the treatment of type I pneumonia, yet the 
painstaking investigations of Cole and his associates have failed 
to obtain success in type II pneumococcus infections, although 
the organism is in most other respects entirely analogous to 
type I. 

To be sure we are in possession of many important empirical 
observations, but to a large extent these are uncorrelated, and 
in some of the most important phases of immunity we are de- 
pendent upon trelisses of theory so flimsy that few of us have 
had the courage to climb them. At the bottom of most of our 
difficulties after all is the unclearness concerning the exact 
mechanism by which most of the pathogenic bacteria cause 
injury to the animal body. If we could thoroughly understand 
this, we might reason more clearly and plan our experimental 
projects more purposefully and in a less groping manner. 

To some extent we have hitherto shirked this issue by assum- 
ing and teaching the existence of endotoxins in the original 
sense of Pfeiffer, although much has appeared in the literature 
which would indicate that the toxic constituents of the cell body 
do not tell the whole story. To formulate a clear judgment of 
this basic problem and to plan for further progress with a clear 
mind, it will be necessary to analyse critically the work that 
has been done in attempted explanations of bacterial toxaemia. 


‘- 


] 
: 
"y 


- a 


NATURE OF BACTERIAL TOXAEMIA 269 


II 


It is not necessary to recapitulate in detail the views of Pfeiffer 
It will be useful for the purposes of our subsequent discussion, however, 
to refresh our memories on certain phases of his experimental work which 
are often neglected when the existence of endotoxins is discussed. 

In addition to showing that dead cholera spirilla, extracts of the spi- 
rilla and filtrates of old cultures, were more toxic than fresh filtrates, 
Pfeiffer obtained results which indicated that, by gradual reduction of 
the intraperitoneal dose we could kill guinea pigs with practically no 
living spirilla in the peritoneum at death. He also found that he 
could bring about rapid death in some of his animals by following the 
administration of considerable amounts of organisms with an injection 
of potent antiserum, a procedure aimed at the rapid dissolution of bac- 
terial cells. These facts Pfeiffer interpreted as indicating that in 
cholera the injury to the body was brought about by the absorption of 
poisons liberated from the bacterial cells upon dissolution, and that 
this disintegration took place under the influence of the bacteriolytic 
serum constituents. This, in brief, extended to other bacteria, is the 
essence of the endotoxin theory. 

From the beginning there was justifiable doubt regarding the specifi- 
city of these toxic substances. Differences in the symptoms induced 
by the poisons derived from different bacteria may be justly regarded 
as largely dependent upon quantitative variations. With the excep- 
tion of the late paralytic effects observed with certain strains of dysen- 
tery and the frequent intestinal symptoms occurring in rabbits on 
inoculation with dysentery, typhoid paratyphoid, or colon extracts, 
there is no evidence of specificity, and even in these examples the 
specificity of such lesions requires more proof. At any rate, there is 
little evidence of tissue selection by the poisons themselves which can 
be regarded as in any way analogous to that exhibited by the true exo- 
toxins of tetanus, diphtheria, and botulismus, and such selection as is 
evident in the human disease can be referred to selective distribution 
of the growing organisms. 

Characteristic from the immunological point of view is the fact that 
the endotoxins are not generally conceived as possessing antitoxin 
inciting properties. The dissolved cell protoplasm gives rise to 
agglutinating, precipitating, etc., antibodies, and if any poison neutral- 
izing properties are at all acquired by the serum of immunized animals, 
this never exceeds neutralizing power for a very limited number of 
lethal units; at any rate the law of multiples cannot be applied. 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 3 


270 HANS ZINSSER 


The first serious experimental objections to Pfeiffer’s conception 
were those which regarded the poisonous properties of the bacterial 
cell not as the effects of preformed liberated poisons, but rather as the 
results of the proteolytic cleavage of the bacterial cell plasma. This 
point of view derived its intellectual stimulus from the work of Victor 
C. Vaughan upon protein split products. His observations on protein 
fever and the subsequent work of Ulrich Friedemann, who succeeded 
in rendering serum toxic in the course of specific hemolysis, gave a 
practical direction to this line of thought. Although Vaughan ad- 
vanced the first experimental proof of toxic split products and their 
bearing upon the field of immunity, neither he nor Friedemann car- 
ried their theories beyond the realm of experimentation into that of 
fancy, a conservatism of which Friedberger, Embleton and Thiele and 
their associates cannot be accused. The conclusions of these workers 
have had much influence upon modern conceptions of toxaemia and 
must, therefore, be briefly discussed. 

Friedberger produced his so-called ‘“‘anaphylatoxins” from practically 
all varieties of bacteria, first by digesting them with sensitizer and 
alexin, then by treating them with normal guinea-pig serum alone. It 
is a very significant fact that he obtained them not only in the test 
tube, but on certain occasions within the guinea-pig peritoneum, a 
point significant in connection with Pfeiffer’s earlier work. Fried- 
berger and his collaborators regarded these toxic substances as split 
products of the bacterial protein which were formed whenever bac- 
teria and active serum constituents were brought together under con- 
ditions of proper quantitative adjustment. 

The opinion of Friedberger may be summarized as follows: Endotox- 
ins in the original sense probably do not exist. When bacteria enter 
the body they are subjected to proteolytic action in which specific 
sensitizers and alexin play a part, the sensitizers present in normal 
serum of most animals and man being sufficient to initiate the process. 
In the course of cleavage a toxic protein split-product is formed, anal- 
ogous to that obtained by Vaughan with chemical methods. This 
toxic substance is produced only during the early stages of the cleay- 
age, but becomes non-toxic as proteolysis continues. All bacteria fur- 
nish the matrix for sucha poison, and there is nothing specific about the 
process except the manner of its production by homologous antibodies. 
Whenever antibody and its antigen meet, the poison is produced; but 
in the animal body under conditions of infection this never occurs with 
sufficient speed nor in sufficient quantity to lead to acute death. 


NATURE OF BACTERIAL TOXAEMIA 271 


However, the symptoms caused by the repeated or continuous produc- 
tion of small amounts of this poison, constitute the basic manifestations 
of every infection; namely, local inflammation, fever, and injury to the 
central nervous system; and this is the explanation of systemic reactions 
occurring in the infected body, reactions which differ from one an- 
other not because of any specific differences in the nature of the poisons, 
but because of the varying abilities of the bacteria to invade, their ac- 
cumulation and distribution, the degree of specific antibody formation, 
and the consequent speed of cleavage. The toxicity of these poisons 
is so great that minute amounts may give rise to symptoms. But 
when a relatively high concentration of antibodies occurs, as in re- 
covery or immunization, the poisons are further broken down so rap- 
idly that intoxication is prevented. In actual disease the accumula- 
tion of bacterial protein and the formation of antibodies are gradual 
processes during which the formation of “anaphylatoxins’” and their 
further cleavage into non-toxic split products takes place constantly, 
and the balance between these two processes determines the intensity 
of the toxaemia. 

Thiele and Embleton, though approaching the subject from a slightly 
different point of view, have gone even farther than Friedberger and 
his associates. They, too, believe that the so-called endotoxic action 
of bacteria is referable to products resulting from the cleavage of bac- 
terial protoplasm by antibodies. They develop a theory of patho- 
genicity from this basic idea, the substance of which is that non- 
pathogenic bacteria are those against which antibody activities are 
either ‘‘so low” that toxic substances are liberated only in extremely 
small amounts; or “‘so high” that the further degradation of the pro- 
toplasm takes place with a rapidity that precluded the accumulation 
of the intermediate toxic substances. Between the two are the true 
pathogens. 

Thiele and Embleton attempted to demonstrate the actual forma- 
tion of these toxic substances in the bodies of infected animals by ad- 
ministering massive doses of bacteria to rabbits and guinea-pigs, bleed- 
ing them twenty-four to thirty-six hours later, and intravenously 
injecting the defibrinated or citrated blood, or even peritoneal exudate, 
into normal pigs. From the fact that these animals often died in 
acute convulsions they drew the conclusion that they had actually 
transferred poisons of bacterial origin. 

Friedberger and Thiele and Embleton’s work resulted in the con- 
struction of a very attractive and simple scheme for the explanation 


272 HANS ZINSSER 


of toxaemia, virulence and the course of bacterial infections in general. 
Though they based their work upon the fundamental ideas advanced by 
Vaughan, they far outstrip him in comprehensiveness of the conclusions 
drawn. Vaughan, in summarizing the subject in 1918, mades the 
following statement: 

“During the active progress of an infectious disease, the body cells 
supply the ferment, the infecting organism constitutes the substrate, 
the process is essentially destructive, the protein poison is set free, the 
symptoms of disease appear and life is placed in jeopardy.” 

It will be noted that Vaughan does not take for granted the forma- 
tion of the poisons under the influence of the sensitizer-alexin mechan- 
ism in the circulation; and does not assume that the entire process 
necessarily takes place in the circulation. 

The first important experimental objection brought forward against 
Friedberger’s views was the observation of Keysser and Wassermann 
who showed that ‘‘anaphylatoxin” could be produced in guinea-pig 
serum by the use of kaolin and barium sulphate, without the presence 
of bacterial protein. Bordet showed that the same thing could be ac- 
complished by incubating fresh guinea-pig serum for two hours with a 
small amount of weak agar solution. The same thing was done by 
Nathan and by Tschernoroutsky with amounts of agar as small as 0.01 
mgm. per cubic centimeter of guinea-pig serum. Jobling and Peterson 
rendered serum toxic by treating it with chloroform. And finally 
Novy and De Kruiff in a very thorough and painstaking study con- 
firmed these results in various ways, showing in addition that the im- 
portant factor in the agar experiments is the physical state of the agar, 
and that ferment action is probably not at all involved. They found 
that inert sol-gel solutions injected intraveously into guinea-pigs, 
rats and rabbits gave rise to shock and other symptoms indistinguish- 
able from anaphylatoxin effects; that the transfusion of blood from ani- 
mals so treated occasionally showed toxic effects in the recipient guinea- 
pigs. Meanwhile, Kohler, Moldovan, Doerr, and others found that it 
is not even necessary to introduce foreign substances like kaolin or 
agar into the experiment. If blood is taken from animals in some 
manner by which clotting is delayed or prevented, as for instance by 
receiving it into paraffined vessels or by defibrinating or citrating, it 
may become toxic automatically, and cause death upon reinjection 
into animals of the same species or even into the same animal from 
which it was taken. In experiments done by Slatineau and Ciuca, 
guinea-pig serum reinjected into guinea-pigs within forty-five min- 
utes, in large amounts, was shown to be occasionally toxic. De Kruiff 


NATURE OF BACTERIAL TOXAEMIA 273 


especially confirmed and extended these observations, first by showing 
that rabbit blood can be rendered toxic and even fatal to guinea-pigs 
and white rats if transfused in the preclot period or after defibrination. 
Guinea-pig blood transferred unchanged within three minutes, or de- 
fibrinated or, on occasion, even the serum of such blood obtained by 
rapid clotting and centrifugation was often fatal to animals of the 
same species. We may summarize his more significant conclusions as 
follows: The spontaneous toxicity of normal blood develops in a simi- 
lar manner as does the toxicity of blood treated with agar, peptone, 
etc. Poison production and fibrin formation go hand in hand and 
occur in the preclot stage. 

The work of the writers mentioned, as well as much similar investi- 
gation of others renders untenable the theories of infection, virulence 
and bacterial toxaemia advanced by Friedberger and by Embleton and 
Thiele. The ‘‘anaphylatoxins” of Friedberger are quite surely not 
derived from the matrix of the bacterial cell plasma. Attempts to 
show bacterial poisons in the circulations of the infected animals by 
transfusion into other animals are not valid since the sera or the blood 
of normal animals may prove toxic if transferred in the same way. 
Moreover, it has never been shown and it is extremely unlikely that 
amboceptor or sensitizer and complement induce any sort of proteo- 
lytic process. 

It should be distinctly understood, however, that these considera- 
tions do not eliminate the possibility of reactions between the bacterial 
antigen and the body cells and their ferments, as suggested by Vaughan 
in his book. Nor do they exclude the possibility that an accumula- 
tion of bacteria in the body, or their introduction into the blood from 
lesions might not cause injury in exactly the same way as does the 
injection of agar, peptone, bacterial protein, etc., in the anaphylatoxin 
experiment in animals. The mechanism, however, would be a similar 
one; that is, not a poisoning with a bacterial protein or its derivatives, 
but an injury due to a chemical or perhaps purely physical change in- 
cited in the blood, in which the poison is derived from the blood con- 
stituents themselves, and the bacteria play a réle which De Kruiff 
has spoken of as analogous to a catalyst, but are otherwise inert.? 


2 Incidentally, the work of Moldovan, Mita and Ito, Novy and De Kruiff, and 
others throws much interesting light upon the toxic effects which have been ob- 
served after the intravenous injection into human beings of such substances as 
bacterial vaccines, the much heralded phylacogens, and occasional similar ob- 
servations with gum acacia solutions. They also promise to elucidate some of 
the hitherto obscure toxic effects of blood transfusion. 


274 HANS ZINSSER 


Til 


Let us now turn our attention to another line of investigation also 
dealing with the problem of bacterial toxaemia, but carried on by 
methods more directly aimed at the discovery of possible exotoxic 
antigenic substances from bacterial cultures. When we survey the 
literature which has accumulated upon this phase of the problem, we 
are struck by the fact that there is hardly a mircoérganism in the 
realm of pathogenic bacteria for which the claim of the existence of an 
antigenic toxin has not been made, at one time or another, and curi- 
ously enough, though none of these numerous claims have been gener- 
ally accepted, and certainly none of them have been utilized to any 
extent in practical therapeutics, many of them seem based upon 
apparently valid experimental data, and in some cases such claims have 
been confirmed at intervals of a great many years by consecutive 
workers. The relatively slight impression which much of this work 
has made upon the development of immunity, is perhaps due to the 
fact that the antitoxic potency of the sera produced by immunization 
with such poisons has never been very high, and the results of clinical 
application have never been sufficiently impressive to carry conviction. 

We cannot possibly undertake here to review the entire literature 
on this subject which has accumulated within the last twenty years, 
nor would it help us much in our analysis to do so. Much of it con- 
sists of repetition and confirmation of early results and many of the 
observations are so incomplete that it is difficult to evaluate them. 
However, it seems unquestionable that a large number of observers 
have actually obtained toxic substances from filtrates of extracts of 
many of the bacteria of which we are speaking, under circumstances 
which make it doubtful whether they were dealing with true cell pro- 
teins, that is endotoxins in the sense of Pfeiffer, or with bacterial prod- 
ucts elaborated by the living bacteria and excreted into the culture 
fluids. Some of these observers have also presented reasonable evi- 
dence that the sera of animals immunized with their poisons have 
possessed true neutralizing properties. 

Before we proceed with the analysis of such work, however, it seems 
to us worth while to consider briefly some of the underlying principles 
which should be kept in mind whenever a solution of the toxaemia 
problems by direct search for a poison is undertaken. 

Whenever workers have turned their attention to investigations of 
this sort they have been to a certain extent under the spell of the 


NATURE OF BACTERIAL TOXAEMIA D7 


analogy to the classic poisons of diphtheria, tetanus, etc. This is worth 
considering since it is important in clarifying our conceptions regarding 
these researches. The essential feature of poisons of this nature is 
not their potency, but rather whether or not they are bacterial secre- 
tions which are yielded as results of the metabolism of the living 
microorganisms, and whether they are truly antigenic. In the case of 
diphtheria, and in that of tetanus as well, we are dealing with bacteria 
that are essentially more saprophytic than parasitic organisms by 
which the living tissues are actually invaded, either not at all, or to a 
slight extent only. It is clear, therefore, that unless these bacteria 
which can multiply in the body locally only, produced very powerful 
poisons, they would not be pathogenic at all, and would not be brought 
to our attention in connection with human disease. 

It is indeed quite conceivable that there may be a great many bac- 
teria in nature that are as little invasive as these, and may produce 
true toxins of a weaker potency; these would never functionate as 
pathogens, simply because the weakness of their poisons and their lack 
of invasive power, taken together, render them entirely incapable of 
establishing a foothold in or upon the living body; or should this hap- 
pen, of producing a sufficiently powerful poison, to incite symptoms. 
It is conversely not only conceivable, but rather likely that bacteria 
which possess that property which we describe as virulence or invasive- 
ness, do not need to produce poisons of the degree of potency possessed 
by tetanus and diphtheria in order to cause symptoms of toxaemia in 
the invaded body. Small amounts of poisons of moderate potency 
would serve cumulatively to intoxicate the animal body if the bacteria 
can actively grow within it and it is, therefore, not essential to find 
poisons in broth or other cultures of such bacteria of anything ap- 
proaching the potency of the classic ones mentioned in order to explain 
the toxaemias which accompany infection with these organisms. 

It must be remembered, too, in analysing this literature and in 
carrying on further research, that it is by no means out of question 
that soluble poisons may be produced by bacteria in the course of their - 
growth which are neither extract products nor antigenic. They may 
not even be specific. If we insist upon the conformity of many poisons 
with these criteria before we include them in our considerations of the 
causations of bacterial toxaemia, we may be missing the truth. For 
just as the peptic, tryptic, and other intestinal enzymes which per- 
form similar tasks in many different animals, are functionally alike, 
so the products of many bacteria may be similar to each other, what- 


276 HANS ZINSSER 


ever the species. And, indeed, such similarity of product seems par- 
ticularly likely in a group of living things which are so essentially alike 
in basic metabolism and morphology as the bacteria. How much this 
last possibility should be considered will become clearer as we go on. 
At the present moment we wish simply to insist that in seeking for 
the cause of bacterial toxaemia we must emancipate ourselves from the 
suppositions that all the poisons we observe must fit into the mold of 
earlier investigations, if they are to be at all included in our principles 
of reasoning. 

For the purposes of our presentation today, we have tabulated these 
researches since it would take a needless amount of time and space to 
analyze each one separately. 

So many important bacteriological principles are involved in the 
researches just tabulated that no superficial general analysis can do them 
justice. A considerable degree of clearness, however, can be obtained 
by comparing the results arrived at in regard to the more basic facts. 

Two main principles of poison production have been employed. 
One of these has been aimed at the extraction of the substances of the 
cell bodies either by grinding and extracting growths on solid media 
or by filtration of old liquid cultures. These poisons, quite obviously, 
are analogous to the endotoxins of Pfeiffer and, although the litera- 
ture is not entirely consistent, there seems to be a more or less general 
agreement that this class of poisons has exhibited little specificity, that 
they have been toxic for both rabbits and guinea-pigs, and have been 
found relatively stable in their resistance to heat and deterioration 
on standing. 

Other methods have aimed to produce poisons with as little extraction 
of the bacterial cell body as possible, either by filtration of younger 
liquid cultures or by the rapid washing up of young agar growths with 
salt solution, and immediate filtration of these substances. 

The manifold investigations in which the nature of the experiments 
makes it quite impossible to determine to what degree extraction prod- 
ucts may have been admixed with the products of metabolism of the 
living microdrganisms, is the feature which has led to so many diver- 
gent results and renders general analysis so difficult. 

However, in those cases in which the manner of poison production 
was such that it is reasonable to believe that cell body extracts played 
a minor part only in the final toxicity of the filtrates, there has been a 
certain amount of regularity of behavior which serves to characterize 
such poisons as different from the endotoxic products obtained with the 


NATURE OF BACTERIAL TOXAEMIA 277 


dead bacteria or cell extracts of the same species. Thus, while Bes- 
redka’s typhoid cell extracts are stated as possessing considerable sta- 
bility, Kraus’s nine day filtrates deteriorated rapidly, Aronson’s ten day 
filtrates were destroyed at 80°C. and deteriorated in a short time, and 
Yamanouchi’s seven day filtrates were partly destroyed at 60°C. in 
thirty minutes. 

Similarly, in dysentery, cell body poisons have generally been found 
to resist heating up to 90°C. and above. But Kraus and Doerr’s 
poisons from agar washings were destroyed at 80°C. in a few minutes; 
Todd’s 10 day filtrates were unstable on standing, and were destroyed 
by 70° to 80°C. in an hour; and Olitsky and Kligler’s ‘‘exotoxins’’ were 
destroyed by 75°C. in one hour, while 90°C. for the same time was 
required for the destruction of their cellular poisons. 

With streptococci, endotoxic cell extracts have been satisfactorily 
demonstrated, and the toxic substances occasionally observed in broth 
filtrates have usually been unstable. 

The B. influenzae filtrates of Mrs. Parker in which the time of culti- 
vation varied anywhere from six to eighteen hours, have been found to 
lose potency at 70°C., as have similar filtrates produced with other 
organisms in our own laboratory. 

Again, these supposedly exotoxic substances have in many cases 
failed to show toxicity for guinea-pigs. Kraus and Doerr’s dysentery 
agar washings, with which they produced neutralizing sera were entirely 
innocuous for guinea-pigs and in Kraus and Stenitzer’s work on ty- 
phoid, potency for guinea-pigs was very weak. This is likewise true of 
Arima’s typhoid exotoxins, and has been the experience generally with 
streptococcus poisons. ‘The same thing is true of Mrs. Parker’s poisons 
of B. influenzae, and all of the toxic substances produced by filtrates of 
early cultures of various organisms which we will described below. 

It seems more than probable, therefore, that in broth filtrates of rela- 
tively young cultures as well as in the filtrates of agar washings, we are 
dealing with poisons distinct from the endotoxic substances, and such 
supposedly exotoxic products have shown remarkably points of simi- 
larity in regard to instability, and relatively low toxicity for guinea- 
pigs. Furthermore, it is significant in this connection that different 
organisms have occasionally shown analogous differences in the symp- 
tom complexes produced by the stable and the unstable poisons. With 
the Shiga dysentery bacilli, Kraus, Bessau, Dopter and more recently 
Olitsky and Kligler have shown that filtrates of the younger cultures 
of this organism produced paralytic symptoms in rabbits, while the 


OUIN[OA 
Sj YQuey-ou0 04 
oGf 78 poye1odeas 


sues puv peie4[y ‘sanoy ZE "TOA 
00g jo said : Sf UOTyN[OS 4] es ‘o061 = ‘*ayas 
-voulns @ ul pezAjoyne ‘ain} =| -UaYyOO MA “pour 
(2) (é) PTET 9 70 -[no proydéy anoy OZ | “yneq ‘tperu0yH 
dn punois 
SUOISO] [VUTY ‘aaoqe jo 4ueul 
-SOJUL OI4SIIOjIBIBYD (é) Soules OT, | -Ipag “UIXOIOPUT 
[BA10q UI 
spIqqer qNOyYM Suryesny 
“eoyIIvIp pus pues ‘syvoS 10} Ayjvrodsa | -114j3U00 pue ouUITes "EQ "JOA 
SIqqBl Ur sIsATBI (2) | ynq ‘sje08 ‘syiqqea ‘s3id | ur ypoeq jo dn | ‘Zier “de 
-ed oaAvS UIx0JOxY | pres SuryyON | -vourns ‘sort soy ‘++4-+4 | Surysey ‘werojorg |} "OQ ‘euIy 
posnjir7 
-u90 uey} ‘sinoy 
yinryeq pur Z 19Y}030} wines 
SAON BI & Bl10j08q ‘09 § puv UOTINIOS 
ey} Aq undies 8s104 q[vs JO ‘00 Z pur 
oy} JO UO0IyBOyIXO4 sesop [Bey wievis CTO “qBS 61 “JOA 
Ie[luis B urxoye[Ayd CI ysurede $4004 SWIBIS OYE YUM punois “9,09 | ‘SOBI ‘48d 
-eue YIM Burpevep | -o1id ‘00 G{’Q wines simoy GT 105 jo Ss1d-vouins (3) ew 48 pety “Wyploeq “4sUuy | ap 


useq 8ABY OY p[noyH)/ counuIUII ssIOFT ‘SOX 'O,1G SJSISOY) S[[OT “99 |] WOTeqnouUyT ({)| Aip ‘aingynd wsy | ‘UUy ‘vypoisog 
sa1d-vounr) Ssyqqey 
saiva 
aNV SUdMUOA 


SHUVAGH NOILONGOUd NIXOLIINV ALITIAVIS NOILONGOUd AO GOHLAW 


woud ALIOIXOL 


proydhiy 


: Ip “JOA 
‘9061 “A “SO 


‘UdAPB OBI 
HOM 4ueo aod pus ‘(g "JOA 
ssid $yBos TO YM uonoR.4 "TORT asstlins cd 
poez1[B1y -woulns 10J 10} Osje | -xo puv uoTyrpuod ‘Sy ‘pue[Moyy 
-nou sosop Qe ‘sox aiqey | A[Qysis Asad | puw ‘4++4+-+4+- | uozoay ur Surputsy | pue usApeovyy 
UNIS §,Bypoiseg Aq U0I}BZI[ 8.1} NOU 
pezi[vazneuyoudiom | OU SBM 91049 poqool 
suosiod-vivd pur -ur Ajoqyearedas {soqn *).00T 4soul Avp pag uo dt 
‘guou poonpoid snyt | -urlul Og 1OJ 1OyRqnd =| 4B AByyoRO4 4B sunoy ZI -x07 soulodaq {sap 
-[loeq JURY “g -Ul Ul UInIes qyIM | -[B fseqnuru ‘sinoy 9 09 | -UBd puR[LoquIRYyD Igor ‘d 
pus y proyddAqeind VABIT YSOq {vypor Og Ul “9,09 Z Ul yIqqRa ysno.ryy 194 [Y ‘19 "10A “6061 
Aq peonpoid uosiod -sogq jo urxojopuo 48 Ayjaed OI, T ste feuojyded yueo sed “TOIq ‘009 "YY 
JaTqee}) ynq $ ‘repluig | -1yue Aq pozt[BiqynaN poAor4soq pesn JON | 00 [T 0% GQ] gG UO soyRIy[y ABp J} “CO ‘TyonoUBUIE 
oO 
I 
‘asnyip you sanoy = 
sosop £°9,08 Q) 01% Ur oyRIy[ YE SYIMOIS ZIGea 
a1ojaq ye  poasory4s SUIvIS (QOOT sfep QOL OF 8 ‘Fh TOA ‘LOGIT 
sinoy $Z poeyoolur | -ap fAlpidea jo yiqqer =| fyjorq uo YyMOIS “OOM xT 
eq ysnu foyerepoyy | soyv101leqyoqy pasn JON | s[Py] 90 G04 Z| epotjod §=jo yulog | ‘[aeg = ‘uosuOry 
iaded 
. ysnory} = pexoyy ty 
‘smmnoy 7% OF ZT 
puszs 0} PeMOT[V 
‘epos yBeom 10 (MOTOq) "YOO A 
yes ‘10yeBM YYLM lM “UST AA 
yo poysem o1ny ‘SNB.y OSTB 90g 
ueAI3 sso'T +++-+ -[no 1e3v Moy FZ (Z) @ “JOA 
Sesop [BYy4e[, =| 918 eyep roded ysnoigqy ‘6061 ‘“UIMT “J 
€ 10 Z pezijBignou yey} poureygo a10M Bios | oYTOeds OU peiyy  ‘eangyno “nosy ‘197410 


-o}g pu sneiyy 


cS ra 


$}OBI}X9 GING[ND IVBV 9Y4 YIM Burzriunurwy] | ynq ‘oyqeysug, BOM [838j ‘00 Z] yjorq Avp 6 (T) 


sesop 0% SB YSIY sv ozI[vaqnou 
p[noo wuinies ay} fuses oy} JO “09 GQO'O 
Aq pezijeiynou svM esop [eyyo] B {uInJes 


BUUDIA OY} YIM poyst[qeyso soydiypnur JO Mey] ye paXoaqysoq | ssid 


“030 UOTPVQNOUI 104 JB 
1944050} = paqoofur 
JOYJOYM ‘UOTZBZI[V1Y 
-neu jo oanbruyoe4 
JO S[IByop 0} SB SIV] 
-noijied OU SeAIS yng 


und 
-OS S,Bypoiseg YIM 
puv UMO SIy YQIA 
sosop [ByB} FI jo 
avisep 0} suostod 


sty peziyeineu ofy 


sesop 
1eq}°] Z ~paziypery 
-nou s}uNoUIe ]]BUIS 
UL YOIYM UINJes pa 


-p[aih asioy poqoofuy 


SUUVANaY 


NOILLONGOUd NIXOLIINV 


you =sa4ery -dutds ory AT -[Oqaivo 19}Jv aod IT ‘sjoa ‘yonq 
“5 wyoig |-vied ‘++ |-vd ysnor1yy ysoq |-pueyE Ip 
"OIXO} = OAV sfoyuour | {syoom ¢ 04 Z 10qFU -BAO'T pue 
soynuiul MO} | SBTPO [RL.I04 ‘aTqiydaosns pe1e}y aingpNo sneiy ‘1100 
® UI °9,08 |-98q AluUO ysoul— SqIq ‘SISOTqOIeB =f Aes pues snery 
vouinry | -qeat ‘+-+4+-+-+ | -so00U AyUITeypPy) pue L190 
fidaquastiq: 
09 T 
YPM sunoy G 
YyooM | Ul ul YQvep jo Pre “d 
yseyo aor UL syeods {104s ‘0% “[OA ‘LO6T 
pe} R10L1o}op aun} uoT}eVq “yoo “UlPy 
fopiqey = AOA -nour‘+-+-+-+ | soyey[y Avp OF 09 6 | “Usryy  ‘snery 
U0T}B} 
-tdieid uw 
10U UOTZRxXY 19}[y purl 
yueure;du109 -laquivyy Yysnoryy 
ul = asuRya paiey[g ismoy gp 
ON  ‘sanoy a1njei19dtie} U001 
Fo 20s UTM {19}VM —- POT] TYSIP ggg ‘d ‘pF [oa 
sulpurys simoy ZL | ouNpeyT’ APYSYS | LOGT “YOOAA 
uo poye UL YYBep ‘00 ‘09 OZ ur dn pa ‘y eq ‘]jos 
-101ayaq; f+ (2) poye‘+++-+ | -ysva soing[no 1esy | -leg pues JssAoTY 
said-vournr sziqquy 


ALITIAVIS 


panuyuo j—proydhi 7, 


c ABP [CTY 
“OI SITTOA,, 
‘ssid-vourns 
1O0J dIxO4 


art 
-GBa UT SUIO4 


SUIZI 


G pues 


wou ALIOIXOL 


NOILONdOuUd AO GOHLAN 


Satva 
aNV SYUaMUOM 


280 


97qQD 
-40apf 889) a4am spoof 
-fa 0y) pazoelur fipazp 
-pawmwmr pup parr 


way A “10yBqnour UI 
Inoy [ 10} puvys 04 
PEMO][B O1OM UTX0Z4 
-UB pUB UXO} Woy 
sqy[nser yseq fuos 
-10d jo sasop [By49] 
OS JSUIvSB pojooq01d 
‘00 T ‘fuinses os10F] 


SYQUOW fF 104jB 
[[9A\ Loyqoue ! poy [Ly 
esi0oy oud {s}vos 
10 = =69si0y =BuIziu 
“nur ur Aqpnowgid 


@AIJVINI 1OZUO] OU 919M WINIOS 
ey} JO seryUenb paesverour seynurm ¢ Aq 
peARfop SBM UIxXOjIQUB BY} Jt ynq ‘poqoofur 
Ajeyeisdes 10M UIX0O} PUB UIxXOyWUB 9Yy 
usyM oov[d Yoo} uoTyZezZI[VIyNeU 9[qQBo0T4 
-OU OS[B JNq JOyYBVOM YONU wv ‘UOTPeZzTUNUT 
-ul pesuojoid sajje queyod AyuoO uwinuos 
‘sosioy puB s}vos ur paonpoid urxoyyuy 


yors 
aul099q ~PBY 4IGQGBI 9Y} 1Aq;8 syoayo Aue 
pemoys Bios 4sedu0Iys oy} ATUQ ‘“19Y 403 
-0} pozoofur Ajsnoeueynoqns jr ozijeaynou 
0} poyrey AOY4 103B] USAW O1ZIA Ul poxrur 
ways [UO poziTeynou Burposjq Aro 


s 9S oud 48U4 98I14G890 


pa BS BAUD VU+ + 417 oUt . 


A[snouoa 

LOpeqnout -81yUl "00 T 
ul Surpueys qyIM pal P] ;,potsed 
ynoqe JIQGBs jueye[,, JO Yysue| 
pres Suryqyou ‘+++4+-+ oAIs §=YOU sa0qy 
‘anoy [ 10j VAT} eS0U as10y ‘peley[y  yymous 
‘Do08 9F .0L Aoyuoul a Aep OL =! 4yyo1q 

ye pofoiysoq | puv asnoyyy yiqqey | ouoyded aul[eyxTV 


(doZu0] 10 Sk¥p OZ 0 8 WIZ dAdYMATIB UMOIB) Sy[NSe1 OAT 
-Bd0U VAVD SNI[IOVq AOUXE[,J 9Y} JO serngynd poredeid Ayxveprumig 


‘poyyour siyy Aq pepnjoxe sea sishjoyny 


sdid-vourns 10j ssoywuey Ayoynjosqe suos 
-1od osayy, ‘[eorddy swmoydurks ory Apeieg 
‘Awssovou sesop osiv, ynq ‘uoroofur [vou 
-oylIodeiqul pue snosuRynoqns uo YyBweq 
‘SINOY OE OF FZ UI YYBIP fsunoYy Moy B UIYIIM 
suostod osoy}y yyIM avodde ueqyo pynom 
SIsA[Bleg ‘Ayyornb o10u payoe soyeyTy 
ey} 3Nq fSUISIUBSIO po], JO suoryoofur se 
smojyduks owes oy} A[jovxe aABS suorzool 
“UL 9FVIZG ISUIBIZS JUOIOYIP Usaajeq saouo 
-layip ‘seingjnd sesunok 10 skep 6 uByy 
[njsomod a10ul YyonuT sainyjnd pjo syoeM 


peonporyur AT[BUTySeqUT 
usyM snonoouul fueAId AjsnoouRyNogns jt 
yvos AOA {sayNUIUT YG OF OF UIYZIM suI04 
-duiAs uve prnom sityy suostod ano yy 
‘sinoy Q Ul YYvop Jo ywods s10yyne SuorTy09 


Game (2) VIC 5 >! 


‘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 
<q GHL dO NOIUoOdOUd 


ILOfAQ SD IWDS WOYDZYDLINIV [DYLVG 
“(97+ 967) Untas pozyouynau fijjoywod 947 pun yourbr1o ay} fo 4a; woYynaif-yuauajdwuos pun L974 UOYDpdL0Nd ay) fo Wostundwog 


6 ATAVL 


306 


i a i i es ei 


<< i 


STUDIES IN ANAPHYLAXIS 307 


One quantitative discrepancy is apparent between the results 
of the animal experiment and those of the test tube experiment. 
The ratio between the normal minimal anaphylactic dose of the 
antigen (0.002 cc.) and that after partial desensitization with 
0.00025 ce. of the egg albumin solution (0.035 cc.) is not the 
same as the ratio between the normal minimal precipitating 
dose of the antigen (0.000002 cc.) and that after partial neutral- 
ization in vitro with 0.00025 cc. (0.00001 cc.). 

This discrepancy could be largely explained by a retest of the 
immune serum on the seventh day after it had been obtained 
from the rabbits. At this retest it was found that the minimal 
anaphylactic dose of the antigen for guinea-pigs passively sen- 
sitized with 0.3 cc. of the immune serum had increased to at 
least 0.005 cc. A retest of immune serum 170-397 after it had 
been kept for one month in the ice-box showed a similar increase 
in the minimal anaphylactic dose of the antigen; that is from 
0.01 ce. to 0.02 ce. 

As the desensitization experiments had been carried out on 
the fifth or sixth days after the serum had been drawn, it is likely 
that the minimal anaphylactic dose of antigen had already 
increased. 

The experiments with horse serum pseudoglobulin were 
carried out with the pooled sera of two rabbits (822 and 388) 
which had received two intraperitoneal injections of pseudoglob- 
ulin as follows: first day 1 cc.; fifth day 5 cc. The rabbits were 
bled seven days after the second injection. 

The minimal sensitizing dose of the pooled sera was found 
to be 0.8 cc. The minimal anaphylactic dose of the antigen 
for guinea-pigs that had been sensitized with four doses of the 
pooled immune sera 322-388 was found, as is seen in table 10, 
to be 0.02 ce. 

In two series of animals we studied the effect of partial neu- 
tralization of the precipitin on the minimal anaphylactic dose of 
antigen. In the first series of animals the partial neutralization 
was carried out in the test tube and the supernatant fluid ob- 


60.004 cc. produced only slight symptoms; 0.005 ce. killed the only animal 
tested. — 


308 ARTHUR F. COCA AND MITSUJI KOSAKAI 


tained after the partial precipitation was tested with the ana- 
phylaxis reaction and also with the precipitin reaction and the 
complement fixation method in vitro. 


TABLE 10 
Determination of the minimal anaphylactic dose of pseudoglobulin after passive 
sensitization. Sensitization: with 3.2 cc. of serwm 322-+-388. _ 
eee PSEUDOGLOBULIN OF HORSE = 
ERTL S SERUM INTRAVENOUSLY INJECTED poNERe 
cc. 

1 0.01 Very severe symptoms 

2 0.015 Very severe symptoms 

3 0.02 


The results of these parallel tests are presented in tables 11 
and 13. | 


TABLE 11 


Determination of the minimal anaphylactic dose of antigen after partial neutrali- 
zation of the precipitin (in vitro). 18 cc. serum (322+888)+0.4 cc. pseudoglobu- 
lin (1-86). After two hours in the incubator and twenty hours at room tem- 
perature, the mixture was centrifuged and the supernatant fluid was used for 
sensitization. 3.2 cc. of the fluid was injected intraperitoneally into each animal. - 
The tests were made on the following day 


GUINEA-PIG PSEUDOGLOBULIN OF HORSE 


SERUM, INTRAVENOUSLY Basson 
cc 
1 0.05 Very slight symptoms 
2 0.1 Severe symptoms 
3 0.1 Death (delayed shock) 
4 0.15 ae 
5 0.15 | <2 


TABLE 12 


Determination of the minimal anaphylactic dose after partial desensitization. Sen- 
sitization: with 3.2 cc. of serum 322+888. Partial desensitization: on the third 
day with 0.284 cc. of pseudoglobulin (1-112), intraperitoneally. 


GUINEA-PIG PSEUDOGLOBULIN OF HORSE 


SERUM, INTRAVENOUSLY ES 


Moderate symptoms 
Moderate symptoms 
Moderate symptoms 
Very severe symptoms 


oe OO NO re 
ooooo 
Hm BH Ww WH bo 


a i SO ee ee SS ee 


Sp fp Pt at | et ttt tit tit lg tlt lt uoIyR} poywory 
: : -1d1001 7 windes | -ugQ (Z) 
Se te a eee oe Pte EP se | bb | ese fe | Sp ees) Se ee fe ee worye} (2Z100°0) (ag) 


“1d199Iq | (9-1) °99 F'0|'99 QT (1) 


(ayn 
100000 0/ 200000" 0) 200000" 0} T0000°0 | z0000°0 | ¢0000°0 | 1000°0 ¢000°0 | ¢000°0 100°0 600°0 £00°0 100 |20°0/20°0| 1'0| 70} awsQuy -qo|d0pnasd) uIndag 
ce a a a eS le oe a ee 


ussyU0y 
(440% HOVT NI ‘00 [0 WONASMINY) NOMMVEIdIOaUd AHL JO Ma 


309 


daXIW NASIINV GNV 
WOUdS ANOWAI 
GHL AO NOILHOdOUd 


: Pa}sa} SL punyf junjousadns ayy puv pabnfrsuaa 
82 aunjanU OYJ, “d.inyodadwaj Wood UL sunoy fiyuam) ya] Uay? puD SANOY 7 paynqnour puv Yan paxnu aio (wnsas asuoy fo uynqopbopnasd) uabywo 
PUD UNAS FO4A UL UOYDZYO.NIU JOY *(SSE+Eeg) wnaas pazypinau iynysod ay? puv yourbrso ayy fo 197) wounjidraasd ay) fo uosrspdwo,) 


&l ATAVL 


310 ARTHUR F. COCA AND MITSUJI KOSAKAI 


It is seen that the minimal anaphylactic dose of the antigen 
was increased by the partial neutralization about 74 fold, while 
the minimal precipitating amount of the antigen was increased 
apparently about 10 fold. This slight discrepancy is, no doubt, 
due to the inherent inaccuracy in the precipitation method. 

In another series of animals which had been passively sensi- 
tized with the untreated serum mixture 322-388, the effect of 
partial desensitization on the minimal anaphylactic dose of the 
antigen was determined. 

The results of this determination are presented in table 12. 

The amount of antigen used for the partial desensitization 
was relatively twice that used for the preceding partial neutral- 
ization experiment. 

It is seen that in this experiment the minimal anaphylactic 
dose of the antigen was somewhat more than twice that after 
the partial neutralization experiment. This apparent dis- 
crepancy can be explained in the same way as those previously 
noted. In this instance the immune serum was four days old 
at the time of the test-tube experiment and seven days old at 
the time of the desensitization experiment, at which time, 
therefore, the minimal anaphylactic dose of the antigen had 
increased for the untreated immune serum. 

All of the experiments that have been described above point 
to the conclusion that precipitin remains unaltered quantita- 
tively and qualitatively in the guinea-pig for several days and 
that it reacts in the animal body with its antigen in exactly the 
same manner as it does in the test tube. Furthermore, the 
experiments offer additional evidence pointing to the identity 
of precipitating and sensitizing antibody. 


Il. ON ANTISENSITIZATION 


A condition of resistance to passive sensitization with a heterol- 
ogous immune serum was observed by Richard Weil to result 
from the previous injection of the normal heterologous serum 
and this condition was designated by Weil as “‘antisensitization.”’ 
The following facts were adduced by Weil regarding the phe- 
nomenon: 


‘7 Ly OS a are 7 eee 


STUDIES IN ANAPHYLAXIS 311 


1. The previous injection of rabbit’s serum obstructs the 
passive sensitization with rabbit’s immune serum but not with 
a homologous (guinea-pig’s) immune serum. 

2. The interference is established after an incubation period, 
which is longer (eight days) after small injections (0.1 to 0.5 
cc. repeated) than after large injections (1 to 8 ec.); that is, four 
days. 

3. The duration of the normal period of heterologous passive 
sensitization in guinea-pigs is six days. If the guinea-pigs have 
received 0.1 cc. of normal rabbit’s serum two to eight days 
previously, the period of passive sensitization with heterologous 
serum is shortened to barely five days. This period may be 
shortened, also, by the injection of 0.6 ce. of normal rabbit’s 
serum made on the day following the passively sensitizing 
injection. 

4. The refractory condition of ‘‘antisensitization” persists for 
at least sixty-eight days. 

5. Active sensitization is unaffected by the injection of large 
amounts of normal rabbit’s serum. 

Weil believed the condition of antisensitization to be due to 
the interference of antibodies and he supported this belief with 
the demonstration of an obstructive action on the part of the 
serum of guinea-pigs that had been immunized with normal 
rabbit’s serum. This view of Weil was apparently contradicted 
by his own observation that the resistance of antisensitization 
is not specific; it could be induced by the previous injection of 
the serum of the sheep and the dog or that of man. 

Weil met this difficulty by demonstrating that guinea-pigs 
sensitized to rabbit’s serum are actually hypersensitive to large 
injections of the three apparently unrelated sera; that is, to 
sheep’s serum, dog’s serum and human serum. 

As it has been pointed out elsewhere (2), this observation of 
Weil is in disagreement with those of Ehrlich and Sachs, who 
reported that their so-called antiamboceptor was specific. 

It may be remarked that if Weil’s explanation of the apparent 
non-specificity of his anti-antibodies is to be accepted it must be 
upon the assumption that the sensitizing antibodies of the 


312 ARTHUR F. COCA AND MITSUJI KOSAKAI 


immune rabbit’s serum are identical, in their antigenic spec- 
ificity, with the common partial antigens of the three non- 
related sera, an astonishing coincidence. | 

If, as Weil believed, the phenomenon of antisensitization is 
due to the action of anti-antibodies, it should be possible to 
demonstrate a corresponding interference with passive sensi- 
tization when the immune serum, previous to its injection, is 
mixed with an anti-immune serum. Moreover, it should be 
possible to show that this interference is actually due to the 
specific precipitation (‘‘neutralization’’) of the sensitizing 
antibodies. 

With the purpose of applying this test to Weil’s theory of the 
mechanism of antisensitization, we have carried out the suc- 
ceeding experiments. On April 2, 7 and 12, three normal guinea- 
pigs received intraperitoneal injections of 0.5 cc., 1 ee. and 1 
ee. respectively of the serum of a rabbit that had been highly 
immunized with egg albumin. On April 18 these three animals 
were bled to death and the sera obtained from the defibrinated 
blood were pooled. The mixed’ serum was tested, as to its 
precipitating power, with rabbit’s serum. The result of this 
test is shown in table 14. 


TABLE 14 
Determination of the precipitation titer of the pooled immune guinea-pig serum 
7 CUBIC CENTIMETERS OF RABBIT SERUM 
IMMUNE GUINEA-PIG 8 
SERUM IN EACH TUBE 
| 0.1 0.05 | 0.03 0.01 0.005 | 0.0025 | 0.001 1.0005 
ce. | 
0.1 | Cloudy |-Cloudy |-----\--4-+-|)-+-+ | + = _ 


With the mixed immune guinea-pigs’ serum the ‘curative 
experiment” of Weil (3) was carried out as follows: on April 18 
three normal guinea-pigs received, by intraperitoneal injection, 
0.3 ec. each of the serum of rabbit 440. (This rabbit had been 
injected a number of times with egg white; the minimal sensitiz- 
ing dose of its serum was 0.1 ce.) On April 19 each of the three 
passively sensitized guinea-pigs was given 2 cc. of the pooled 
immune guinea-pigs’ serum by intravenous injection and two 


STUDIES IN ANAPHYLAXIS 313 


days later the three animals were tested as to their hypersensi- 
tiveness to egg-albumin by intravenous injection. ‘lwo ani- 
mals responded with very slight symptoms to injections of 1.0 
ec. and 0.4 cc. respectively of the undiluted egg albumin solu- 
tion; the third animal died with the typical symptoms and 
anatomical changes of anaphylactic shock, after an injection of 
0.6 cc. of the egg-albumin solution. The inflation of the lungs in 
this animal was not maximal. 

This result confirms in a general way that obtained by Weil 
in the single animal employed in his experiment. It reveals, 
however, an irregularity in the ‘‘curative” action of the immune 
guinea-pigs’ serum, which had escaped Weil and which is diffi- 
cult to harmonize with Weil’s view of the mechanism of anti- 
sensitization. 

It is difficult, in other words, to explain why, notwithstanding 
the identical treatment of the three guinea-pigs, the “anti- 
antibodies” of the immune guinea-pigs’ serum should have 
“‘neutralized” the antibodies of the sensitizing rabbit’s serum in 
two animals but not in the third. 

Having demonstrated the ‘“‘antisensitizing’ action of the 
mixed anti-rabbit guinea-pigs’ serum with the technic of Weil’s 
curative experiment, we were in position to examine the mecha- 
nism of that action with the test tube experiments. If,as Weil 
believed, the phenomenon of antisensitization is due to a specific 
interference by anti-antibodies, then it must be possible to pre- 
cipitate the sensitizing antibodies from immune rabbit’s serum 
by mixing the latter with the anti-rabbit guinea-pigs’ serum 
in vitro; the antibodies of the rabbit’s serum should be found to 
enter into the resulting specific precipitate as indicated by their 
absence in the supernatant fluid obtained by centrifugation. 

On April 21, 12 cc. of the pooled anti-rabbit guinea-pigs’ 
serum were mixed with 1.8 cc. of the anti-egg albumin serum of 
rabbit 440. This mixture was kept for one hour at 37°C. and 
then overnight in the ice-box. The precipitate which had 

6 The minimal lethal dose of this solution for guinea-pigs sensitized with three 


doses of anti-egg albuminimmune rabbit’s serum had been found to be not more 
than 0.003 ce. 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 3 


314 ARTHUR F. COCA AND MITSUJI KOSAKAI 


formed was removed by rapid prolonged centrifugation and the 
clear supernatant fluid was examined: (1) As to its content of 
rabbit’s serum proteins; (2) as to its power passively to sensi- 
tize against egg albumin; (3) as to its content of anti-egg albu- 
min precipitins. 

The examination for the presence of rabbit’s serum proteins 
was made on April 22 by mixing 0.4, 0.2, 0.1, 0.05 and 0.025 ce. 
of the supernatant fluid with 0.1 ce. of the pooled anti-rabbit 
guinea-pigs’ serum. In none of these mixtures had any precip- 
itation taken place after they had stood overnight. All of the 
rabbit’s serum protein that was precipitable with the anti-rabbit 
guinea-pigs’ serum had been removed. 

In examining the supernatant fluid as to its power passively 
to sensitize against egg albumin, account had to be taken of the 
dilution of the rabbit’s immune serum (approximately 1-8) that 
had occurred as a result of mixing it with the guinea-pigs’ immune 
serum. 

On April 22 a series of normal guinea-pigs were given intra- 
peritoneal injections of 0.8 cc. or 1.6 ec. or 2.5 ec. of the super- 
natant fluid; that is, amounts of the mixture representing 1, 2 
and 3 minimal sensitizing doses of the immune rabbit’s serum. 
On the following day the animals were tested with the intra- 
venous injection of not less than 33 lethal doses of the egg 
albumin solution. The results of this test are presented in 
table 15. 

TABLE 15 


“Antisensitization’”’ in vitro 


| 
AMOUNT OF THE ar , 
+ SUPERNATANT FLUID ONS os S 
GUINEA-PIG EGG-ALBUMIN USED FOR RESULT 


USED FOR saa 
SENSITIZATION EBA EO 


Ss 
is) 


cc. 


1 0.8 0.4 No symptoms 

2 1.6 0.4 Slight symptoms 
3 2.3 0.4 oe 

+ 2.3 0.4 oe 

5 2.3 0.2 Slight symptoms 
6 2 0.1 Slight symptoms 


STUDIES IN ANAPHYLAXIS 315 


These results are susceptible of the interpretation that the 
effect of the anti-rabbit guinea-pigs’ serum had been to increase 
threefold the sensitizing dose of the immune rabbit’s serum and 
also to increase by more than 66 fold the killing dose of the egg 
albumin for the passively sensitized guinea-pigs. However, in 
view of the outcome of our ‘‘curative’’ experiment, it seems at 
least possible that in the test tube experiment we have encoun- 
tered the same irregularity of action that we observed in the 
curative experiment; that is, it is possible that guinea-pigs num- 
ber 5 and 6 would not have been killed with even 0.4 ce. of the 
egg albumin solution. 

In any case, if Weil’s hypothesis were correct it must have 
been possible to demonstrate, in the supernatant fluid, a change 
in the precipitating action against egg albumin corresponding 
with the change in the sensitizing action. In accordance with 
the results of our experiments in the partial neutralization of 
precipitin in vitro it could be expected that the change in the 
present instance would consist in an increase of at least 66 fold 
in the minimal precipitating amount of the egg albumin. How- 
ever, the examination of the supernatant fluid in this respect 
showed no such change. Indeed, as the protocol of the compara- 
tive titration presented in table 16 shows, precipitation titer of 
the fluid corresponded exactly with that of the untreated rabbit’s 
immune serum. 

TABLE 16 
Showing the comparative precipitation titer of the supernatant fluid and the original 


anti egg white rabbit’s serum 


CUBIC CENTIMETERS OF EGG-ALBUMIN 


| 0.002 0.0001 0.00002 0.00001 


Supernatant fluid in each tube, 0.4 cc...... sgcam! Ssae9enGe) 47 
Control (serum 140) in each tube, 0.05 cc. . | +++] ++ | cr) + 


The experiment just concluded demonstrates that the phenom- 
enon of the anti-sensitization is not due to the interference of 
‘“anti-antibodies;’ the power of the rabbit’s precipitin of re- 
acting with egg albumin was not perceptibly altered by contact 


316 ARTHUR F. COCA AND MITSUJI KOSAKAI 


with the “‘anti-sensitizing’’ immune guinea-pigs’ serum. Fur- 
thermore, the irregularity of the action of the anti-sensitizing 
guinea-pigs’ serum suggests a non-specificity of that action which 
is in harmony with the observation of Weil that the active form 
of the anti-sensitization can be induced by the previous injection 
of heterologous sera. 


Ill. EXPERIMENTS WITH SPECIFIC PRECIPITATES 


In one of his studies in anaphylaxis (4) Weil attempted to 
demonstrate directly the identity of precipitin and sensitizing 
antibody (‘‘sensitizin’’). Weil’s procedure was to inject washed 
specific precipitates into guinea-pigs and to test the injected 
animals, after a proper interval of time, as to the development 
of passive sensitization. 

In such experiments Weil was usually successful, though in 
some instances the precipitate failed to sensitize the normal 
animals. 

Our experiences with the partial neutralization of precipitin 
in vitro and in vivo made it seem unlikely that the combination 
of protein and precipitin would be more dissociable in the ani- 
mal’s body after the test-tube reaction than after the reaction 
that occurs in vitro. 

We have, therefore, employed Weil’s technic in an effort to 


duplicate his results. According to the procedure in his first ~ 


experiment a constant amount of the precipitating serum—1.5 ce. 
—was mixed with different quantities of the antigen, crystalline 
egg albumin, and after an incubation of one hour at 37°C. and 
forty-eight hours in the ice-box, the mixtures were centrifugated 
and the precipitate was twice washed with 2 cc. of sterile saline 
solution. The whole of the washed precipitate from each mix- 
ture was injected intraperitoneally into a normal guinea-pig. 
Three days later each of the injected guinea-pigs was given a 
test injection of the egg albumin intravenously. None of the 
animals exhibited any symptoms of anaphylactic shock. Fur- 
ther details of this experiment are presented in table 17. 


STUDIES IN ANAPHYLAXIS 317 


TABLE 17 


Attempted passive sensitization with washed precipitates 


SERUM WASHED p 
322 + 388 F: PRECIPITATE pe ny ala: 
(MINIMAL SEN-| EGG-ALBUMIN | PRECIPITATE | INJECTED INTO aeienernicie RESULT 
SITIZING DOSE GUINEA-PIG aS 
0.1 cc.) ON THIRD DAY 
ce ce. ce 
1.5 0.001 + 1 1.0 No symptoms 
15 0.005 Seerae 2 1.0 No symptoms 
1.5 0.02 ateafat 3 0.6 No symptoms 


This experiment was repeated with the sera of two rabbits 
that had been immunized with egg white. The results of these 
experiments are presented in table 18. 


TABLE 18 


Attempted passive sensitization with washed specific precipitates 


: Serum 440 ! 
Experiment IT. eam re Immune serum v. egg white 
WASHED | 
PRECIPITATE 
aeEuM 0.) INTRAPERI- | ALBUMIN 
(SENSITIZING ALBUMIN PRECIPITATE | TONEALLY INTRAVE- RESULT 
Dose) 0.1 cc. G ae petite 
ON THIRD 
DAY 
cc. ce ec 
1.5 0.001 + 1 0.5 No symptoms 
tea 1.5 0.005 |+++++ 2 0.5 No symptoms 
15 0.02 | ++++ 3 0.5 Moderate 
symptoms 
SERUM 440 
Getoa'h 1.5 0.02 ; ++++ 4 0.5 No symptoms 
1.5 0.04 | ++++4+ 5 0.5 No symptoms 
SERUM 419 
MINIMAL 
(SENSITIZING 
DOSE) 0.1 cc. 
1.5 0.001 + 6 0.5 No symptoms 
Series ¢ 1.5 0.005 |+++++ vi 0.5 No symptoms 
1.5 0.02 } ++++ 8 0.5 No symptoms 


It is seen that a moderate sensitization was attained in only 
one animal and that when that test was repeated with the pre- 


318 ARTHUR F. COCA AND MITSUJI KOSAKAI 


cipitate from an identical mixture (series B) not the least sen-. 
sitization resulted. Thus, it appears that although our one 
animal may be looked upon as confirming Weil’s observation of 
the passively sensitizing property of specific precipitates, the 
exhibition of that property is the exception rather than the rule. 

The negative results appear the more striking when it is re- 
called that the precipitates were obtained from ten sensitizing 
amounts of the immune serum. 

In the study of antibodies of various kinds, numerous efforts 
have been made to separate these from the serum proteins of the 
immune serum containing those bodies. Such studies have had 
either a practical end in view or they have sought information 
as to the chemical nature of the antibodies. 

For technical reasons these efforts have been made with anti- 
bodies against formed antigens such as bacteria and blood cor- 
puscles and the procedure has consisted in the absorption of the 
antibodies out of the immune serum with the antigenic elements 
followed by an extraction of the antigen-antibody combination 
with various substances at various temperatures. In the extrac- 
tion fluid nearly all of the experimenters have demonstrated only 
the antibodies. Weil, alone, reported that he found antigen but 
no precipitating antibody in his “‘extracts.’’ Weil mixed horse 
serum with the anti-horse serum of an immune rabbit and, fol- 
lowing Chickering, he extracted the resulting precipitate with 1 
per cent sodium carbonate. The extract, when mixed with 0.01 
ee. of horse serum, remained clear, but when mixed with 0.5 ce. 
of the anti-horse serum, it formed a precipitate. This unique 
result was complicated by the successful passive sensitization of 
normal guinea-pigs, which Weil produced with the same extracts 
in which he could find no precipitin. 

Veil explained this paradoxical outcome with the assumption 
of a haptophore or sensitizing group and an ergophore or pre- 
cipitating group in the precipitin molecule. 

We have investigated this phenomenon with an anti-egg white 
rabbit’s serum. We mixed the serum with crystalline egg- 
albumin and dissolved the resulting precipitate, obtained by cen- 
trifugation and decantation of the supernatant fluid, in 2 cc. of 


STUDIES IN ANAPHYLAXIS 319 


1 per cent sodium carbonate. In confirmation of Weil’s obser- 
vation it was found that the addition of 0.5 cc. of rabbit’s immune 
serum did produce a precipitate when mixed with a certain 
amount of the carbonate solution of the precipitate. However, 
this effect was shown not to be specific because an equal degree 
of precipitation could be produced in the carbonate solution of 
the precipitate with 0.5 ec. of a normal serum or by simply neu- 
tralizing the extract. The phenomenon thus represented merely 
a reprecipitation of a dissolved precipitate and not thespecific 
precipitation of free antigen, as Weil thought. 


SUMMARY 


1. Experiments are presented which demonstrate that pre- 
cipitin remains unaltered quantitatively and qualitatively in the 
guinea-pig for several days and that it reacts in the animal body 
with its antigen in exactly the same manner as it does in the 
test tube. The experiments furnish further evidence that pre- 
cipitin and “‘sensitizin’’ are identical. 

2. The phenomenon of “‘anti-sensitization’”’ (Weil) is not due 
to the action of anti-antibodies. It is a non-specific effect the 
nature of which is obscure. 

3. Passive sensitization with washed specific precipitates has 
generally failed in our hands. 

4. The sodium carbonate extract or solution of specific pre- 
cipitates do not contain free antigen but represents the whole 
precipitate in solution. 


REFERENCES 


(1) Wein, Ricnarp: Jour. Immunol., 1917, 2, 469. 

(2) Tick: Practice of Medicine, vol. 1, p. 143. 

(3) Wert, Ricuarp: Zt. f. Immunitatsf., 1914, 23, 1 (p. 13). 
(4) Wert, Ricuarp: Jour. Immunol., 1916, 1, 1. 


A STUDY OF THE PRECIPITIN TEST IN CASES OF 
PNEUMOCOCCUS EMPYEMA 


CLEVELAND FLOYD 


From the Department of Bacteriology, Harvard Medical School, Boston, 
Massachusetts 


The large measure of success attendant upon the use of im- 
mune sera in the treatment of certain types of pneumococcus 
pneumonia has given rise to the hope that other infections of 
similar origin might be benefited. 

The invasion of the pleural cavity by the pneumococcus, 
secondarily to pneumonia, is common—in fact, it is the rule 
rather than the exception. Whether this infection gives rise to 
dry pleurisy, or empyema, depends upon the virulence of the 
micro-organism, the amount of chemiotactic action exerted and 
the destructive action of elaborated toxins. The very nature of 
the lymphatic circulation of the pleura with the opportunities 
for the inflow of immune substances, unquestionably tends to 
reduce materially the number of secondary cases of empyema. 
Where, however, the lymph channels are blocked with fibrin 
and cells as the result of acute pleural infection, there may be a 
great difference between the amount of circulating antibodies in 
the body at large and their abundance in the pleural exudation. 

Often after the pathological process is well established, neu- 
tralization or absorption of all agglutinins, precipitins and bac- 
teriolysins takes place, and the enclosed infection may die out 
from the effect of the products of cellular degeneration. More 
- often, however, pleural drainage by surgical means is necessary 
to permit of a rapid interchange of toxin and antitoxin, and 
bacteria, agglutinin and precipitin. 

A study of pleural exudation in empyema is instructive in 
that it gives a measure of the power of the forces of the body to 
combat a localized infection. 

321 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 4 


322 CLEVELAND FLOYD 


The crisis of pneumonia is commonly accepted to be the 
time when the body rides over the infection through its maxi- 
mum production of immune substances. The crisis of the 
infection of the pleura in empyema probably takes place in a 
few instances before surgical intervention, with a resulting sterile 
empyema—in others with the influx of fresh substances after 
thoracotomy or of the pointing of empyema necessitans. 

Our study of a series of twenty-two cases of empyema has 
been in the nature of observations on the presence of demon- 
strable precipitins in the exudation and their fluctuation; and 
the effect upon them of the introduction of appropriate immune 
serum. At the same time, the nature of the infection, the 
entrance of secondary invaders, and the amount of pleural 
phagocytosis was noted. 

These cases were under observation for a period of a few days to 
several weeks. The action of the serum on the course of the dis- 
ease was only incidental to other observations, because it was not 
possible to systematically carry it out to its logical conclusion. 

The method of procedure was as follows: Following the opera- 
tion, pus from the chest was obtained, the type of organism 
determined, and the presence of precipitin sought for. Subse- 
quently, every other day, when it was possible, the chest cavity 
was irrigated with sterile saline solution, and 4 ounces of equal 
parts of appropriate immune serum and saline solution were 
put into the chest. 

This procedure was repeated until the child recovered, or some 
intercurrent disease interrupted the period of observation. No 
untoward results were noted from the introduction of the serum. 

The type of pneumococcus obtained was chiefly that grouped 
as I and IV. Only two cases of type II were obtained, and 
type III organism was not seen. As serum for type IV was 
not available, only types I and IJ were treated. 

In all of this series of cases the precipitin reaction with the 
exudate was present following the operation, and, as a rule, con- 
tinued with increasing positiveness in favorable cases. In cer- 
tain instances, however, following the injection of considerable 
quantities of serum, the test occasionally was negative indicat- 


PRECIPITIN TEST IN PNEUMOCOCCUS EMPYEMA 323 


ing that even following drainage and artificial support all pre- 
cipitin was absorbed. The amount of phagocytosis and the 
duration of the presence of pneumococci ran a parallel course 
with the precipitin reaction. The presence of agglutinins in the 
exudate proved to be uncertain, and we failed frequently to 
get satisfactory results. In many instances the pathogenic 
process was definitely prolonged by the invasion of other organ- 
isms than those originally present—such as members of the 
staphylococcus and diphtheria groups. 


Case 1. Boy. Age, thirteen months. Unilateral empyema with 
history of cold for two weeks before admission. 

February 18. Thoracotomy. 

February 19. Thick pus received. Smears showed pneumococci 
and very little phagocytosis. Precipitin test showed the organism to 
be type II. 

February 28. Agglutination test not successful. Chest irrigated 
and pneumococcus serum, type II, injected into chest cavity. Fairly 
thick pus obtained. Precipitin test positive. Smears and cultures 
showed pneumococci, staphylococci and B. hoffmaniz with not much 
phagocytosis. 

March 8. Chest irrigated and serum injected. Pus obtained. 
Precipitin test positive. Smears showed few pneumococci, staphylo- 
¢occi and B. hoffmanii with increased phagocytosis. 

March 13. Child discharged relieved. Very little discharge. 


Case 2. Girl. Age, six years. Unilateral empyema with history 
of pneumonia three and one-half weeks previous. 

February 19. Thoracotomy. 

February 20. Thick pus obtained. Precipitin test showed organ- 
ism to be pneumococci type I. Smears showed pneumococci with 
very little phagocytosis. Agglutination test not successful. 

February 25. Chest irrigated and pneumococcus serum type I 
injected. Pus obtained. Precipitin test positive. 

February 26. Chest irrigated and serum injected. Pus obtained. 
Smears show increased phagocytosis and diminution in the number of 
pneumococci. 

February 27. Button removed from chest. 

March 1. Chest irrigated and serum injected. 


324 CLEVELAND FLOYD 


March 4. Child discharged to Boston City Hospital with measles. 

March 16. Child readmitted from Boston City Hospital. 

March 18. Child’s condition good; wound still draining. 

April 6. Child’s chest healed. Drain reinserted as temperature 
had been septic in type for past few days. Thick pus evacuated from 
chest. Smears show no pneumococci. 

April 12. Drain removed. 

May 6. Discharged well. 


Case 3. Boy. Age, three years. Admitted March 5, 1918. Uni- 
lateral empyema with history of pneumonia two weeks previous. 

March 5. Thoracotomy. 

March 6. Pus obtained. Precipitin test showed organism to be 
pneumococci type I. Smears showed pneumococci in abundance. 

March 8. Chest irrigated and serum type I injected. Pus ob- 
tained. Precipitin test positive. Smears and cultures show pneumo- 
cocci, staphylococcus aureus and albus. 

March 11. Chest irrigated and serum injected. Pus obtained. 
Precipitin test not so clearly positive as at first. Smears showed some 
phagocytosis and only a few pneumococci. 

March 15. Button removed. 

March 17. Discharged relieved. Slight drainage from sinus. 


Case 4. Boy. Age, six years. Admitted March 20, 1919. Uni- 
lateral empyema with history of pneumonia twenty-four days previous. 

March 20. Thoracotomy. 

March 21. Pus received. Precipitin test showed organism to be 
pneumococcus, type I. Smears showed pneumococci. Phagocytosis 
slight. 

March 22. Chest irrigated and serum injected. 

March 28. Chest irrigated and serum injected. Pus obtained. 
Smears showed no pneumococci and B. hoffmanii. 

April 2. Practically no discharge from the chest. 

April 7. Cultures taken showed very few pneumococci. 

April 9. Child discharged to Boston City Hospital with measles. 

April 15. Child readmitted. 

April 16. Chest healed. 

May 6. Discharged. 


Case 5. Girl. Age, five years. Admitted March 22, 1919. Uni- 
lateral empyema with history of pneumonia three weeks previous. 


PRECIPITIN TEST IN PNEUMOCOCCUS EMPYEMA 325 


March 22. Thoracotomy. 

March 23. Heavy pus obtained. Precipitin test showed organism 
to be pneumococcus, type I. Smears showed pneumococci in chains 
with little or no phagocytosis. 

March 25. Chest irrigated and type I serum injected. Muco- 
purulent pus obtained. Smears showed a number of pneumococci 
and slight increase of phagocytosis. 

March 28. Chest irrigated and serum injected. Pus obtained. 
Precipitin test unsatisfactory. Smears and cultures showed a few 
pneumococci and staphylococcus albus. Phagocytosis more marked. 

April 1. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test faintly positive. Smears show few pneumococci, 
numerous bacilli and staphylococci. 

April 3. Chest irrigated and serum injected. Button removed. 
Bloody fluid, no pus, obtained. Precipitin test negative. Cultures 
showed heavy growth of a diphtheroid, and a few pneumococci. 

April 7. Chest irrigated and serum injected. Specimen obtained 
from chest. 

April 8. Cultures showed B. diphtheroid and a few colonies of 
staphylococcus albus. 

April 11. Child discharged well. 


Case 6. Boy. Age, four years. Admitted March 23, 1919. Uni- 
lateral empyema with history of pneumonia three weeks previous. 

March 23. Thoracotomy. 

March 25. Heavy pus received. Precipitin test showed organism 
to be pneumococcus, type II. Smears showed a moderate number of 
pheumococci—some in chains—and no phagocytosis. 

March 28. Chest irrigated and serum type II injected. Pus ob- 
tained. Smears showed very few pneumococci, well phagocyted, and 
cultures showed only colonies of B. xerosis. 

April 1. Chest irrigated and serum injected. Thin pus obtained. 
Precipitin test negative. Smears showed very few ea 
numerous bacilli, probably B. xerosis. 

April 3. Chest irrigated and serum injected. Button removed. 
Bloody fluid obtained. Precipitin test negative. Cultures showed 
very few pneumococci. 

April 7. Chest irrigated and serum injected. Mucopurulent fluid 
obtained. Culture showed an occasional colony of pneumococcus and 
B. xerosis in large numbers. 


326 CLEVELAND FLOYD 


April 13. Child discharged to Boston City Hospital with measles. 

April 20. Child readmitted. 

April 26. Marked recrudesence of pus since April 13 Swab 
showed no pneumococci, and a fair number of typical diphtheria bacilli. 

April 28. 1500 units of antitoxin given. 

May 14. K. L. bacilli still present. 

May 20. Patient improving, but discharge still considerable. 

June 3. Discharge lessened. 

June 11. Child practically healed. Child discharged. 


Case 7. Boy. Age, two years. Admitted April 2, 1919. Uni- 
lateral empyema with history of pneumonia one week previous. 

April 2. Thoracotomy. 

April 3. Thick, heavy, greenish pus received. Precipitin test 
showed pneumococci to be type I. Smears and cultures showed 
numerous pneumococci with little or no phagocytosis. 

April 7. Chest irrigated and serum type I injected. Pus obtained. 
Precipitin test markedly positive. Smears and cultures showed 
pneumococci with great increase of phagocytosis. 

April 12. Chest irrigated and serum injected. Very little secre- 
tion from chest obtained. Precipitin test negative. Smears show 
only an occasional pneumococcus. Cultures showed rare colony of 
pneumococci, and several colonies of staphylococcus albus. 

April 16. Chest irrigated and serum injected. Slightly opaque, 
watery solution obtained. Precipitin test positive. Smears showed 
almost no organisms and only an occasional leucocyte. 

April 18. Temperature septic in type. 

May 1. Temperature still septic. Wound draining slightly. 

May 7. Cultures from discharge show staphylococcus aureus. 

May 15. Temperature still septic. Wound closed. 

May 19. Wound reopened and much pus evacuated. 

May 29. Septic temperature still present. 

May 31. X-ray showed no pus in the chest. 

June 19. Child improved. Chest entirely closed. 

June 23. Discharged relieved. 


Case 8. Girl. Age, five years. Admitted April 11. Unilateral 
empyema with history of pneumonia twelve days previous. 

April 11. Thoracotomy. 

April 12. Heavy pus received. Precipitin test slightly positive 
for type I. Smears and cultures showed numerous pneumococci with 


PRECIPITIN TEST IN PNEUMOCOCCUS EMPYEMA 327 


a fair amount of phagocytosis. Agglutination test with cultures 
showed marked positive reaction to serum 1. 

April 16. Chest irrigated and serum type I injected. Watery pus 
obtained. Precipitin test positive. Smears showed a moderate num- 
ber of pneumococci with a fair amount of phagocytosis. Cultures 
showed staphylococcus albus. 

April 18. Chest irrigated and serum injected. Pus obtained. 
Precipitin test negative. 

April 22. Chest irrigated and serum injected. Fairly heavy pus 
obtained. Smears showed pneumococci with phagocytosis increased 
and an increase of staphylococci. 

April 26. Chest irrigated and serum injected. Thick greenish pus 
obtained. Smears showed relatively few pneumococci—many mor- 
phologically typical diphtheria bacilli. Cultures showed staphylo- 
coccus albus and streptococci. 

May }. Chest irrigated and serum injected. Heavy pus obtained. 
Precipitin test positive. Smears showed few pneumococci, a consider- 
able number of phagocyted diphtheria bacilli, and staphylococci. 

May 6. Chest irrigated and serum injected. Thinner pus ob- 
tained. Precipitin test faintly positive. Smears showed moderate 
number of pneumococci and fairly well phagocyted diphtheria bacilli. 

May 8. Chest irrigated and serum injected. Watery pus obtained. 
Precipitin test negative. Smears showed few pneumococci. 

May 16. Diphtheria bacilli still present. Still considerable dis- 
charge from chest. 

June 8. Wound practically healed. Child discharged. 


Case 9. Boy. Age, nine years. Admitted April 20. Unilateral 
empyema with history of pneumonia three weeks previous. 

April 20. Thoracotomy. 

April 22. Thick pus received. Precipitin test showed positive 
reaction to type I serum. Smears showed moderate number of pneu- 
mococci, slightly phagocyted. 

April 26. Chest irrigated and serum type I injected. 

April 27. Button removed. 

April 29. Chest irrigated and serum injected. Watery pus re- 
ceived. Precipitin test positive. Smears showed few pneumococci. 

May 1. Chest irrigated and serum injected. Thick, yellow pus 
obtained. Precipitin test positive. Smears showed some pneumo- 
cocci, well phagocyted. Cultures showed pneumococci and staphylo- 
coccus albus. 


328 CLEVELAND FLOYD 


May 3. Drainage still considerable. Chest irrigated and serum 
injected. Smears obtained showed few leucocytes, no pneumococci, 
and few other organisms. 

May 6. Chest irrigated and serum injected. Thinner pus ob- 
tained. Precipitin test faintly positive. Smears showed few pneu- 
mococci with moderate phagocytosis. 

May 8. Chest irrigated and serum injected. 

May 10. Chest irrigated and serum injected. Pus obtained. 
Precipitin test faintly positive. Smears show a moderate number of 
pneumococci well phagocyted. 

May 13. Chest irrigated and serum injected. Pus obtained. Pre- 
cipitin test negative. Smears showed pneumococci in chains and no 
phagocytosis. 

May 15. Chest irrigated and serum injected. Pus obtained. 
Precipitin test negative. Smears showed broken down leucocytes; 
few pneumococci; little phagocytosis. Cultures show moderately 
numerous pneumococci and some streptococci. 

May 20. Patient’s temperature was septic in type—drainage still 
profuse. Serum discontinued. 

June 11. Temperature normal. 

June 16. Child discharged. 


Case 10. Boy. Age, nine years. Admitted April 23. Unilateral 
empyema with history of pneumonia three weeks previous. 

April 23. Thoracotomy. 

April 24. Pus with heavy film received. Precipitin test showed 
slight positive reaction with pneumococcus type I. Smears showed 
many pneumococci; no phagocytosis. 

April 26. Chest irrigated and type I serum injected. 

April 29. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test negative. Smears showed moderate number of pneu- 
mococci; slight phagocytosis, many bacilli and staphylococci. 

May 3. Chest irrigated and serum injected. Fairly heavy pus 
obtained. Smears showed few pneumococci, fairly well phagocyted, 
numerous diphtheroid bacilli and staphylococci. 

May 6. Chest irrigated and serum injected. Very little pus ob- 
tained. Precipitin test negative. Smears showed few organisms not 
phagocyted. 

May 8. Chest irrigated and serum injected. 


PRECIPITIN TEST IN PNEUMOCOCCUS EMPYEMA 329 


May 10. Chest irrigated and serum injected. Watery pus ob- 
tained. Precipitin test faintly positive. Smears showed a moderate 
number of phagocyted pneumococci. 

May 13. Chest irrigated and serum injected. Button removed. 
Very thin pus obtained. Precipitin test positive. Smears showed no 
organisms. 

May 15. Chest irrigated and serum injected. Almost no dis- 
charge present. Smears showed few broken down leucocytes and no 
organisms. 

May 16. Child discharged. 


Case 11. Boy. Age, four years. Admitted April 26. Bilateral 
empyema with history of bronchitis four weeks previous. 

April 25. Thoracotomy. 

April 27. Thick pus obtained. Precipitin test showed positive 
reaction to type I pneumococcus. Smears and cultures showed pneu- 
mococci. 

May 3. Chest irrigated and serum type I injected. Heavy pus 
obtained. Precipitin test positive. Smears showed few pneumococci 
moderately phagocyted and many diphtheroid bacilli. 

May 5. Serum treatment discontinued. 

May 6. Second operation. 

May 10. Child died. 


Case 12. Boy. Age, five years. Admitted April 26. Unilateral 
empyema with history of pneumonia four weeks previous. 

April 26. Thoracotomy. 

April 27. Thick pus received. Precipitin test positive for serum 
typeI. Smears and cultures showed pneumococci, but no phagocytosis. 

May 1. Chest irrigated and serum type I injected. Watery pus 
obtained. Precipitin test not satisfactory. Smears and cultures 
showed a fair number of pneumococci, staphylococcus aureus and 
albus, and moderate phagocytosis. 

May 3. Serum treatment discontinued. 

May 9. Child died. 


Case 18. Girl. Age, five years. Admitted May 6. Unilateral 
empyema with history of pneumonia five days previous. 
May 6. Thoracotomy. 


330 CLEVELAND FLOYD 


May 6. Fairly heavy pus received. Precipitin test positive for 
pneumococcus type I. Smears showed pneumococci with little phago- 
cytosis. 

May 18. Chest irrigated and serum I injected. Thin pus obtained. 
Precipitin test negative. Smears showed scattered pneumococci not 
phagocyted. 

May 15. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test negative. Smears showed few pneumococci, mostly 
phagocyted. Cultures showed numerous colonies of staphylococcus 
albus, diphtheria bacilli and a few of the pneumococcus. 

May 17. Chest irrigated and serum injected. Watery pus ob- 
tained. Precipitin test positive. 

May 20. Chest irrigated and serum injected. Fairly thick pus 
obtained. Precipitin test positive. Smears showed fairly numerous 
pneumococci slightly phagocyted. 

May 22. Chest irrigated and serum injected. Thin pus obtained. 
Precipitin test positive. 

May 24. Chest irrigated and serum injected. Thin pus obtained. 
Precipitin test negative. 

May 27. Chest irrigated and serum injected. Thin pus obtained. 
Precipitin test positive. Smears showed a few chains of pneumococci. 

May 29. Chest irrigated and serum injected. Very thin pus 
obtained. Precipitin test positive. Smears showed no pneumococci. 

May 31. Chest irrigated and serum injected—discharge decreased. 

June 2. Discharged. 


Case 14. Boy. Age, two years. Admitted May 8. Unilateral 
empyema with history of pneumonia ten days previous. 

May 8. Thoracotomy. 

May 8. Thick pus received. Precipitin test positive for pneumo- 
coccus type I. Smears showed moderate number of pneumococci well 
phagocyted. 

May 13. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test positive. Smears showed few pneumococci—no phago- 
cytosis. 

May 14. Button removed. 

May 15. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test negative. Smears showed numerous pneumococci and a 
fair amount of phagocytosis. Cultures show pneumococci. 


PRECIPITIN TEST IN PNEUMOCOCCUS EMPYEMA 331 


May 17. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test positive. 

May 20. Chest irrigated and serum injected. 

May 22. Chest irrigated and serum injected. 

May 24. Chest irrigated and serum injected. Very little pus 
obtained. Precipitin test positive. 

May 28. Child discharged. 


Case 15. Girl. Age, eight years. Unilateral empyema with his- 
tory of sickness since measles one year previous. 

May 14. Operation (patient has had empyema before). Tho- 
racotomy. 

May 15. Thick, bloody pus received. Precipitin test positive for 
type I pneumococci. Smears showed pneumococci in chains moder- 
ately phagocyted. Cultures showed some colonies of staphylococcus 
albus, and only a few pneumococci. 

May 17. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test faintly positive. 

May 20. Chest irrigated and serum injected. Fair amount of pus 
obtained. Precipitin test negative. Smears showed many pneumo- 
cocci well phagocyted. 

May 22. Chest irrigated and serum injected. Pus obtained. 
Precipitin test negative. 

May 24. Chest irrigated and serum injected. Pus obtained. 
Precipitin test negative. 

May 27. Chest irrigated and serum injected. Pus obtained. Pre- 
cipitin test negative. Smears showed few pneumococci and a few 
diphtheroid bacilli. 

May 29. Chest irrigated and serum injected. Fairly thick pus 
obtained. Pecipitin test faintly positive. Smears showed moderate 
number of pneumococci, staphylococci and many diphtheroid bacilli. 

May 31. Chest irrigated and serum injected. Fairly heavy pus 
obtained. Precipitin test faintly positive. Smears showed numerous 
pneumococci with slight phagocytosis. 

June 1. Button removed. 

June 5. Chest irrigated and serum injected. 

June 7. Chest irrigated and serum injected. Fairly thick pus 
obtained. Precipitin test faintly positive. Smears showed few 
organisms, mostly pneumococci moderately phagocyted. 

June 10. Chest irrigated and serum injected. 


332 CLEVELAND FLOYD 


June 11. Child running high temperature and her condition poor. 
Serum discontinued. 

June 23. Child much worse, having had no improvement since 
June 11. 

June 28. Child died. 


Case 16. Boy. Age, four years. Admitted May 14. Unilateral 
empyema with history of sickness and cough for two weeks previous. 

May 14. Operation. Thoracotomy. 

May 15. Thick greenish pus obtained. Precipitin test faintly 
positive, pneumococcus type I. Smears showed a very few pneumo- 
cocci. Cultures showed pneumococci in pure culture. 

May 17. Chest irrigated and serum injected. Thick pus obtained. 
Precipitin test slightly positive. 

May 20. Chest irrigated and serum injected. Fairly thick pus 
obtained. Precipitin test positive. Smears showed very few pneu- 
mococci well phagocyted and many diphtheroid bacilli also phagocyted. 

May 22. Chest irrigated and serum injected. Fair amount of pus 
obtained. Precipitin test negative. Agglutination test positive. 

May 24. Chest irrigated and serum injected. Pus obtained. 
Precipitin test negative. 

May 27. Chest irrigated and serum injected. Pus obtained. 
Precipitin test positive. Smears showed almost no organisms. 

May 29. Chest irrigated and serum injected. Pus obtained. 
Precipitin test positive. Smears showed a few diphtheroid bacilli and 
practically no pneumococci. 

May 31. Chest irrigated and serum injected. 

June 3. Button removed. 

June 5. Chest irrigated and serum injected. Very thin pus ob- 
tained. Precipitin test faintly positive. Smears showed some staphy- 
lococci and diphtheroid bacilli. 

June 7. Chest irrigated and serum injected. Very thin pus ob- 
tained. Precipitin test faintly positive. Smears showed many 
staphylococci and diphtheroid bacilli, and a few pneumococci. 

June 11. Child died. 


Case 17. Girl. Age, thirteen months. Admitted June 3. Uni- 
lateral empyema with history of pneumonia four weeks previous. 
June 3. Operation. 


: 
' 


PRECIPITIN TEST IN PNEUMOCOCCUS EMPYEMA 333 


June 4. Thick pus obtained. Precipitin test showed positive 
reaction to pneumococcus type I. Smears and cultures showed pneu- 
mococci, staphylococcus aureus and unidentified bacilli. 

June 7. Chest irrigated and serum injected. Pus obtained. Pre- 
cipitin test positive. Smears showed moderate number of pneumococci. 

June 10. Chest irrigated and serum injected. Very thin pus 
obtained. Precipitin test faintly positive. : 

June 11. Button removed. 

June 15. Chest irrigated and serum injected. Practically no pus 
obtained. Precipitin test negative. 

June 16. Chest irrigated and serum injected. Very little pus 
obtained. Precipitin test positive. Smears showed very few pneu- 
mococci. 

June 17. Chest irrigated and serum injected. 

June 19. Chest irrigated and serum injected. Very little pus 
obtained. Agglutination test negative. Smears showed few pneu- 
mococci, fairly well phagocyted, and no secondary infection. 

June 21. Chest irrigated and serum injected. Very little pus 
obtained. Precipitin test negative. Agglutination test negative. 
Cultures showed staphylococcus aureus. 

June 24. Chest irrigated and serum injected. Very little pus 
obtained. Chest practically closed. Precipitin test positive. Smears 
showed a few pneumococci. 

June 28. Chest closed. 

July 9. Discharged. 


Case 18. Boy. Age, twenty months. Admitted June 4. Uni- 
lateral empyema with history of pneumonia three weeks previous. 

June 4. Thoracotomy. 

June 5. Thick pus received. Precipitin test positive with pneu- 
mococcus type I. Smears showed a few pneumococci only slightly 
phagocyted. 

June 7. Chest irrigated and serum injected. Pus obtained. Pre- 
cipitin test positive. Agglutination test positive. Smears showed a 
few organisms, mostly bacilli with a few pneumococci. 

June 10. Chest irrigated and serum injected. Very little pus 
obtained. Precipitin test faintly positive. 

June 15. Chest irrigated and serum injected. Little pus obtained. 

June 17. Chest irrigated and serum injected. 


334 CLEVELAND FLOYD 


June 19. Chest irrigated and serum injected. Very little pus 
obtained. Agglutination test faintly positive. Smears showed moder- 
ately numerous pneumococci with marked phagocytosis and no sec- 
ondary infection. 

June 21. Chest irrigated and serum injected. Very little pus 
obtained. Smears showed no organisms. 

July 9. Discharged well. 


Case 19. Boy. Age, one year. Admitted June 13. Unilateral 
empyema with history of pneumonia eighteen days previous. 

June 13. Thoracotomy. 

June 14. Thick pus received. Precipitin test positive for type I 
pneumococcus. Smears and cultures show only pneumococci. 

June 17. Chest irrigated and serum injected. 

June 19. Chest irrigated and serum injected. Fairly thick pus 
obtained. Agglutination test positive. Smears showed many pneu- 
mococci and marked phagocytosis. 

June 21. Chest irrigated and serum injected. Very little pus 
obtained. Smears showed very few organisms. Button removed. 

June 24. Chest irrigated and serum injected. Very little pus 
obtained. Precipitin test positive. Smears showed few pneumococci 
with considerable phagocytosis. 

June 26. Chest irrigated and serum injected. Very little pus 
obtained. Precipitin test positive. Smears showed very few pneu- 
mococci. 

June 28. Chest irrigated and serum injected. Very little pus 
obtained. Precipitin test positive. Serum treatment discontinued. 

July 22. Discharged. 


Case 20. Boy. Age, two years. Admitted June 18. Unilateral 
empyema with history of pneumonia four weeks previous. 

June 18. Operation. Thoracotomy. 

June 19. Thick pus received. Precipitin test positive for type I 
pneumococcus. Smears showed pneumococci in fair numbers. Little 
phagocytosis. 

June 21. Chest irrigated and serum injected. Thick pus obtained. 
Smears showed very few pneumococci, mostly phagocyted. Serum 
treatment discontinued. 

July 15. Child died. 


PRECIPITIN TEST IN PNEUMOCOCCUS EMPYEMA 3395 


CONCLUSIONS 


1. The presence of an easily demonstrable amount of pre- 
cipitin in the exudation of empyema is generally a favorable 
prognostic sign. 

2. The amount of demonstrable agglutinin in the exudation 
of empyema is often very small. 

3. Phagocytosis runs a parallel course with increasing pleural 
resistance. 

4. The pleura will clear itself of a pneumococcus infection in 
a much shorter time than the average, if the secondary invaders 
can be excluded. 

5. Pleural irrigation with an appropriate immune serum in 
pneumococcus empyema is suggested as a means of treatment in 
order to increase locally those immune substances that tend to 
limit the duration of the infection. 


I wish to take this opportunity to thank Dr. William E. Ladd 
on whose service at the Children’s Hospital these cases were 
observed. 


ae iid ae” Wh ay re) ) ees eae 


Aa ha 


. 
ates - 
i j 
i 
‘ 
_ 
‘ 
~ 
« 
= « 
* 
: 
. 
y 
oT 
on ‘ 
Ri * 
’ 
' 
? ~ 


A NEPHELOMETRIC METHOD OF ESTIMATING THE 
NUMBER OF ORGANISMS IN A VACCINE 


GEORGE C. DUNHAM 
From the Typhoid Vaccine Department, Army Medical School, Washington, D. C. 


Received for publication May 6, 1920 


In the manufacture of a bacterial vaccine, the determination 
of the number of organisms per cubic centimeter is a very im- 
portant part of the work. For this determination two methods 
are in general use, the microscopical count and the measurement 
of the opacity of the bacterial suspension as compared with the 
opacity of a bacterial or chemical standard. The microscopical 
count is usually made either by the Wright method or with a 
Halber counting chamber. The measurement of the opacity, 
or the turbidity, of a vaccine is made, in the older methods, by 
comparing the turbidity of a tube of bacterial suspension, of 
unknown concentration, with that of a series of tubes containing 
‘standard suspensions. The standards consist of bacterial sus- 
pensions, or of suspensions of a chemical such as barium sulphate, 
in which the turbidity of each suspension represents a certain 
bacterial concentration. More recently a method has been 
devised for estimating the opacity of a suspension by measuring 
the depth at which a wire loop disappears from view in the 
suspension as compared with the depth at which it disappears 
in a suspension of known concentration (1). 

Counting the bacteria in a vaccine with the microscope is 
laborious, time consuming, and in routine work it is liable to 
considerable error. Estimation of the concentration of a vaccine 
by the comparative measurement of the turbidity of the opacity 
of the suspension, by the methods in use at the present time, is 
subject to error due to the differences in the acuteness of vision 
of different individuals, and the variations in the color of the 

337 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 4 


338 GEORGE C. DUNHAM 


suspensions. With the nephelometer the errors caused by the 
differences in the acuteness of vision of the different observers, 
and variations in the color of the suspensions, are, to a large 
extent, eliminated. This reduction in the percentage of error 
is due to the accuracy with which the depth of the suspension 
can be measured and the accuracy with which the two fields 
of reflected ight can be matched. 

In the performance of this work the Kober (2) nephelometer 
was used. This instrument was primarily devised for the esti- 
mation of the amount of suspended matter in analytical work, 
such as fats in blood and milk, and the results obtained are based 
upon the measurement of depth of the suspension that is necessary 
to reflect the same amount of light as that reflected by a standard 
suspension at a fixed depth. With this apparatus artificial light 
passes into the suspension in the cups on either side and is re- 
flected through the plungers and eyepiece to the eye of the 
observer. ‘The cups can be raised or lowered independently of 
each other and the extent of this movement is accurately meas- 
ured on the corresponding vernier millimeter scale on the back 
of the instrument. By means of the scale the depth of the 
suspension required to reflect a certain amount of light can be 
measured directly to 0.1 mm. and by interpolation to 0.01 mm. 
The arrangement of the cups, plungers and eyepiece gives two 
flat, side by side, fields of light which can be accurately matched. 

Standards consisting of suspensions of bacteria have been 
found to give the most satisfactory results. These are made 
by suspending the killed bacteria in normal salt solution to which 
a small amount of antiseptic has been added. An accurate 
microscopical count, with the Halber counting chamber, is made 
of the bacteria in this suspension, which is then diluted until 
it contains the desired number of bacteria per cubic centimeter. 
Several recounts should be made. 

Standards which consist of material other than bacteria, such 
as barium sulphate, may be used. However, barium sulphate 
will settle out of an aqueous suspension too rapidly for accurate 
work. When it is suspended in glycerin, or a similar substance 
of sufficient density to prevent this settling, the air bubbles 


ESTIMATING NUMBER OF ORGANISMS IN VACCINE 339 


remain in the fluid for some time and interfere with the readings. 
On the other hand, bacterial suspensions will remain accurate in 
the instrument for from twenty to thirty minutes and then only 
require to be shaken. 

This work has been done in connection with the preparation 
of monovalent typhoid and paratyphoid suspensions in the 
manufacture of triple typhoid vaccines. Monovalent standards 
have been used in which the standard for the paratyphoid A 
vaccine contained two billion bacilli per cubic centimeter, that 
for the paratyphoid B vaccine contained two billion bacilli per 
cubic centimeter, and the standard for the typhoid vaccine 
contained one billion bacilli per cubic centimeter. 

In order to determine the number of bacteria in a vaccine of 
unknown concentration, 10 to 12 cc. of the standard for that 
vaccine are placed in the cup on the left side of the instrument. 
This cup is then raised until the vernier scale on the left side 
is set at 20 mm. indicating that the column of the standard 
suspension between the bottom of the plunger and the bottom 
of the cup is 20 mm. in depth. A proper dilution of the vaccine 
to be counted is then placed in the cup on the right side. This 
cup is raised until the light that is reflected through the eyepiece 
from the bacteria in the two suspensions is exactly equal on both 
sides of the field. The depth of the suspension of the unknown 
vaccine, that is required to reflect the same amount of light as 
20 mm. of the standard suspension, can then be read on the 
right vernier scale. 

The concentration of the unknown suspension is not exactly 
inversely proportional to readings of the scale. To supply the 
necessary correction Kober has suggested the following equations 


(3): 


x — SH+SE+V(S + Sk)? — 48khY 4 _ S  U-—X)Sk 
rab ia xX x 
In which, Y = the reading of the unknown on the millimeter 


seale. 
S = the reading of the standard on the millimeter 
scale. 


340 GEORGE C. DUNHAM 


X = the ratio of the concentration of the two sus- 
pensions. 


r 


K 
k= 5 where K = a constant obtained by the sub- 


stitution of the standardization values of S, Y, 
and X. 

To use the formula for each calculation would, however, 
involve a considerable expenditure of time. It is much more 
practicable to use a nephelometeric curve. To make this curve, 
the left cup containing the standard is set at 20 mm. Various 
dilutions, such as 0.9, 0.8, 0.7, and 0.6, of the standard are read 
against the undiluted standard. The readings thus obtained are 
plotted on cross section paper and a curve drawn through the © 
established points (fig. 1). Once this curve has been plotted, 
it may be used for all future work with that standard. Before 
each series of readings, the standard should be read against 
itself and the vernier scale on the right adjusted so that the 
light reflected from one suspension matches that reflected from 
the other when both scales are set at 20 mm. When this is 
done, the original curve is applicable. 

It is necessary to dilute the vaccine to be counted so that the 
readings will come within the limits of the curve. 

As all parts of the curve represent a known number of bacteria 
per cubic centimeter, it is only necessary to apply the reading 
of the millimeter scale to the curve to obtain the number of 
bacteria per cubic centimeter in the dilution that is being read. 
This result multiplied by the dilution will equal the number 
of bacteria per cubic centimeter in the undiluted vaccine. For 
example, assuming that a paratyphoid A vaccine, diluted to 
one part in four, gives a reading of 24.7 mm. on the scale, reference 
to the curve for this vaccine (fig. 1) shows that this reading 
represents 0.75 of the concentration of the standard (2000 million 
bacilli per cubic centimeter) or 1500 million bacilli per cubic 
centimeter. This figure multiplied by four gives 6000 million 
bacilli per cubic centimeter, which is the concentration of the 
original vaccine. 


ESTIMATING NUMBER OF ORGANISMS IN VACCINE 


The bacterial standards should be renewed each month. 


341 


A 


fresh vaccine is diluted until it exactly matches the old standard 


in the nephelometer. 


and avoids any danger of autolysis. Standards 
old have been found to be accurate. 


Eee aa HH FEC 


ecaaia =r Lt festa) pa at 


BEEa 7 aan oo 


ESE ATE ae eeeecnSE 


odes taney sasen teaza 
COAL 


, y, a 
Popa em ert At Zaemy 


EE EEH Seis ooatcaterisecd AV tre an 
er Sri fosteceteciior daa anfosteeninedt 


This procedure requires only a short time 


three months 


Fig. 1. NEPHELOMETRIC CURVE FOR THE PARATYPHOID—A SUSPENSION 


Concentration of the standard, 2000 million bacilli per cubic centimeter. 


The ordinates represent the millimeter readings on the scale. 


The abscissae 


represent the ratio of the concentration of the {unknown suspension to that of 
the standard. 


1.0 = 2000 million bacilli per cubie centimeter 
0.9 = 1800 million bacilli per cubic centimeter 
0.8 = 1600 million bacilli per cubie centimeter 
0.7 = 1400 million bacilli per cubic centimeter 
0.6 = 1200 million bacilli per cubic centimeter 


342 GEORGE C. DUNHAM 


The results that are outlined below are typical of those ob- 


tained by this method. 


NEPHELOMETRIC 
VACCINE ESTIMATION 
PER CUBIC CENTIMETER 


5440 million 
6080 million 
5220 million 
5070 million 


PAE MM PAGE AS. ws Seshn console tiehereraicdeleun seek 


7320 million 
9500 million 
9200 million 
10000 million 


PATARYDHONE 1D), 5 oso. = Acia o-108 ciohaieing see 


4236 million 
3440 million 
3700 million 
3780 million 


ASD UGI faa lo akstetoh cinkactgebaseicidame coe 


Color, unless of sufficient intensity to cut off the rays of light, 
To demon- 


does not interfere with the accuracy of the results. 
strate this point, the following examples are given: 


- Read against the usual standard 


Per cubic centimeter 
7800 million 
7600 million 
7080 million 
7080 million 


Before coloring -6)) sb jesse aes Coe ee oT ee wae 
After coloring with methylene blue........................ 
Before (coloring yt. hog coe ayactaievin yo ove Nev ae eee atte oscicia ss 
ATCET COLOTING WIth. DICTICIACIG 4.2.4.8 latest aie ee 6a 


MICROSCOPICAL COUNT 
PER CUBIC CENTIMETER 


5600 million 
6000 million 
5280 million 
5120 million 


7500 million 
10000 million 
9000 million 
9600 million 


4200 million 
3420 million 
3760 million 
3520 million 


Vaccine that had been grown on dark agar media, and vaccine 
that had been grown on blood agar media when read against 
a standard that had been grown on light agar media gave accurate 
results. For example: 

NEPHELOMETRIC 


ESTIMATION 
PER CUBIC CENTIMETER 


MICROSCOPICAL COUNT 
PER CUBIC CENTIMETER 


3520 million 
2640 million 


3480 million 
2670 million 


Pyploid Gdark appar). < sass Baia #) oes 
ay phorde(bloodsagar))\.cn cs eee eee 


It has been found that certain cultures of paratyphoid B take 
the stain very poorly and also that some of the cultures exhibit 


ESTIMATING NUMBER OF ORGANISMS IN VACCINE 343 


a tendency towards fragmentation. ‘These factors render the 
bacilli very difficult to count and introduce a considerable error 
into the microscopical count. With the nephelometer the error 
due to the poor staining qualities is eliminated. The fragmented 
bacilli reflect the light in proportion to their mass as well as 
the whole bacilli. 

In general, the advantages of the nephelometric method over 
the microscopical are the saving of time and labor. The average 
reading can be made and calculated in less than ten minutes. 
The error should not be greater than 10 per cent plus or minus. 
This is as accurate as the microscopical count, if not more so. 


SUMMARY 


A method is described for estimating the number of organisms 
per cubic centimeter in a vaccine by means of the nephelometer. 

The estimation can be made in less time and with less labor 
than a microscopical count, and the results are fully as accurate 
as those obtained with the microscope. 


REFERENCES 


(1) Gates: Jour. Exp. Med., 1920, 29, 105. 

(2) Koper: Jour. Biol. Chem., 1917, 29, 155. 

(3) Koper: Jour. Am. Chem. Soc., 1915, 37, 2379. Jour. Biol. Chem., 1913, 13, 
485. 


ea 


i 


? 
i 


EFFECT OF ULTRAVIOLET RAYS ON ANTIGENIC 
PROPERTIES 


I. STUDIES ON MENINGOCOCCUS 


FREDERICK EBERSON 


From the Washington University School of Medicine, Saint Louis, Missouri 


Received for publication May 10, 1920 


As early as 1885, Duclaux (1) showed that light exerted a 
destructive action on spores. His work was followed by the 
researches of Arloing (2) who studied particularly the B. anth- 
racis. Koch (3) in 1890, applied these results to the tubercle 
bacillus and succeeded in demonstrating the same principles. 
Confirmatory studies were published by Mignesco (4) in 1896. 

The bactericidal power of sunlight is not dependent upon heat 
but as Ward (5) has reported, is a function of the ultraviolet 
rays. Indeed, the earliest studies on the effect of sunlight upon 
bacteria suspended in water (6) tend to disprove the idea that 
prolonged heating is responsible for the destruction. 

The role which photochemical action of light plays in the 
phenomenon is not definitely established. Peroxide of hydrogen 
may be formed in fluid media under aerobic conditions, yet the 
action of ultraviolet rays is too marked to be explained in this 
manner. In addition to this, there is direct evidence that dried 
bacteria are similarly influenced (7). On the other hand, much 
has been written on the photodynamic influence of fluorescent sub- 
stances on microérganisms and the analogy between the action 
of light and photodynamics presented. The studies of Jodl- 
bauer and Tappeiner (8), Tappeiner (9) and Finssen (10) contain 
pertinent data. Contradictory evidence is given, however, by 
Cernovodeanu and Henri (11) who contend that pigmented as 
well as unpigmented organisms do not differ in their resistance 
to ultraviolet rays. Later work by Burge (12) compared the 
rate of death of fluorescent and non-fluorescent (12) bacteria 


345 


346 FREDERICK EBERSON 


when subjected to ultraviolet light and advanced the theory 
that coagulation of bacterial protoplasm was responsible for the 
bactericidal effect. 

More recently, the literature on ultraviolet rays has been de- 
voted to the practical application of water sterilization. Buch- 
ner (13) in 1893, presented data which paved the way for later 
studies among which may be mentioned those of Desfosses (14) 
and Nagier (15). 

Apparently no attempt has been made to study the action of 
the ultraviolet ray from the standpoint of bacterial attenuation 
or physicochemical modification. In this paper the problem has 
been approached from these angles and an effort made to rule 
out the factor of heat in the experiments, so that no extraneous 
influence other than that of the ultraviolet ray might be involved. 


EXPERIMENTAL 


The most suitable organisms for a study of this kind are those 
which belong to a group characterized by a serologic relationship 
of some of its species. The effect of any external influence such 
as ultraviolet light ought to be manifested readily owing to the 
unusual susceptibility to change which any transitional group of 
organisms possess. These experiments must be considered as 
secondary in importance to the problem for which they were 
intended; namely, that of finding a method for producing potent 
sera and vaccines by the use of living cultures, without the usual 
untoward manifestations which attend present methods. For 
such a study any group of organisms may suffice. We have 
employed the pneumococcus and have found that the principles 
laid down as a result of the experiments with meningococci are 
applicable to the pneumococcus and that it is possible to use 
living organisms for active immunization with great success. 

It is essential in studying the action of the ultraviolet ray to 
suspend the organisms in a menstruum which will insure a water- 
clear suspension free from protein or pigment, and having a 
definite density of the organism. Since the penetrating power 
of the rays varies inversely as the turbidity of the bacteria! sus- 


STUDIES ON MENINGOCOCCUS 347 


pension and the transparency of the suspending fluid, these 
factors must be subject to reasonably accurate control. 


Experiment 1. The effect of sublethal doses of the ray upon menin- 
gococcus was studied and the first step was to determine the minimal 
lethal dose of ultraviolet light. 

Technic: Meningococcus cultures representing the regular, irregular 
and para types were subcultured from an eighteen to twenty hour old 
sheep serum agar slant to glucose agar. After twenty hours incubation 
these were again transferred to glucose agar and the growth treated on 
the next day as follows. Standard suspensions were prepared in physi- 
ological salt solution to contain 2000 million cocci per cubic centimeter. 
These clear suspensions in amounts of 15 to 20 ec. were placed in sterile 
quartz Erlenmeyer flasks of 50 cc. capacity and exposed to ultra-violet 
rays for varying periods after which the flask was well shaken and 0.5 
cc. of the suspension was withdrawn and pipetted on to sheep serum 
glucose agar. These subcultures were incubated for twenty-four 
hours and negative tubes were re-incubated for an additional day in 
order to determine possible attenuating action which might result from 
the different exposures. An “Alpine Sun Lamp,” manufactured by the 
Hanovia Chemical Company was used with a current of 110 volts, 3 
amperes, at a distance of 36 inches. The temperature of the flask 
when exposed directly to the mercury are did not exceed 38°C. at any 
time. 

Minimal lethal dose of ultraviolet ray on types of meningococcus 


TIME OF EXPOSURE (MINUTES) 


Gulture..| 4} 1113] 2/23 | 3] 33] 4) 43| 5] 54] 6 | 64] 7 |zal slio 
Regular..|+-+}+-+/4+--+}--+}+-+1-4+-+|-++/++/+4+/4++/++/4++] + | + |-|-|- 


Control..| Still fully viable after one hour 


Irregu- 
lar... ..|-+-E}--E]E IE I+ IE 444+ 4) 44+)+4+)44+)4+4+/44/4+)/4/)+ 
Control..| Still fully viable after one hour | 


— elyrilniel MEAL Pa ee +|+]-|-|-|- 


Control..| Still fully viable after one hour | 


Growth after twenty-four hours incubation indicated by ++. Negative 
growth after twenty-four hours but positive after forty-eight indicated by +. 
No growth indicated by —. Repeated three times with approximately same 
results. 


348 FREDERICK EBERSON 


In order of their susceptibility to the ultraviolet ray a para- 
meningococcus comes first, followed by the regular and next by 
the irregular type. This, in general, appears to be in keeping 
with the resistance of each of these strains to environmental 
conditions as observed in the laboratory. 


Effect of ultraviolet ray on agglutinogenic properties of meningococcus 


Any change effected in the bacterial protoplasm by the physico- 
chemical action of the ray should be coincident with a modifica- 
tion of a salient immunologic property. A monovalent serum 
prepared in the rabbit by means of a single type of meningococcus 
is highly specific, especially so when a rapid method of immuni- 
zation is used. The studies of Gordon (16) and other workers 
have established this point and our own experience has not been 
otherwise. An irregular type serum may agglutinate to a slight 
degree a regular or a parameningococcus, but the reaction is 
never marked. On the other hand, a regular meningococcus 
produces no agglutinin for the para type—at least not above 
1—25 in our observations—and slight agglutinins now and then 
for an irregular strain. The para type is highly specific, possess- 
ing slight group agglutinins for the irregular strains at times. 

The object of the succeeding experiments was two-fold: first, 
to show, if possible, that the meningococcus antigen is complex 
and contains elements common to numerous types; secondly, to 
demonstrate the physicochemical influence of ultraviolet rays 
on a bacterial cell. This action should manifest itself either by 
a rearrangement of the internal structure of the organism with 
an alteration of its normal serologic reactions, or by bringing 
into prominence the group-radicals already present within the 
cell. The organism might then be shown to possess a wider 
range of agglutinating properties. 

A regular no. 4, irregular no. 30 and a parameningococcus no. 
60 were exposed to the action of ultraviolet rays for varying 
periods which corresponded to subminimal lethal doses deter- 
mined previously. Duplicate sera were prepared in rabbits with 
each of these treated strains, a modified rapid method of immuni- 
zation being used in which two daily injections were given fol- 
lowed by a rest period of six days after which the final injection 


STUDIES ON MENINGOCOCCUS 349 


was given. The amounts of culture used were respectively 
1/100, 1/50 and one-quarter of an eighteen to twenty hour 
growth prepared as described in the technic for ultraviolet 
treatment. The animals were bled from the ear six days after 
the last injection and agglutination series set up against the 
untreated, living homologous and heterologous cultures. All 
agglutination tests were made at 56° C., and the results were 
read after eighteen to twenty hours incubation in the incubator. 


Experiment 2. Regular meningococcus No. 4 was exposed to the 
ultraviolet ray for 1, 2, 3, 4, 5 and 6 minutes, respectively, and sera 
for each of these modified cultures were prepared as described above. 
The animals showed no ill effects from the injections. 


AGGLUTINATING RANGE 


ULTRAVIOLET 
Bee EXPOSURE 


| } 
1-100 | 1-200 | 1-400 | 1-500 | 1-800 | 1-1000 | 1-1200 |Control 
| 


4 + + + + — = = = 

176 1 30 - + — os = = zB 4 
SUS a sw a + | + - - — 

Zoe \\ lige a= = = a - = = 

177 2 S00 see | ea | 4 =: = = = 
SU “ta ta = — = — = 

eet te ih oe = — ~ - - 

178 3 7) | pate tal ee a | Te re 
BO Mies cll) Ses nck) = dh ES eee ah =, fp 

OY Mopar! Baia |liecn | ars || ar =k — - 

179 4 UN es alo mel cr | + a - - 
60 + t+: ck _ = = = = 

4 a5 35 ck ok a _ - - 

180 5 30 }t++i4+/ +/+] +]—-{]-4]- 
6c }++}/ +] +)/+]/+]-—-]/-]- 

oA A oa a5 == = = = = 

181 6 oy goa et a TER Mg TE oe ee eee 
60 + a = - = —_ — ae = 


In all protocols the following symbols are used to denote degree of agglutina- 
tion. ++, complete, with clear supernatant fluid; +, sediment with slightly 
cloudy supernatant fluid; +, flocculi in fine suspension; —, negative. 


2 


350 FREDERICK EBERSON 


The conclusions to be drawn from the experiment are that 
exposures of one to five minutes to the action of ultraviolet light 
under certain conditions renders the antigen more inclusive; also, 
that a toxic organism, such as the regular meningococcus which 
was used in the experiment, may become less toxic by the treat- 
ment and may be tolerated in much larger doses on the second 
day without giving rise to severe reactions or death. The serum 
produced with such altered cultures agglutinates heterologous 
strains equally as well as or better than its specific antigen, an 
indication that a regular type meningococcus contains antigenic 
elements common to the irregular and para types. 


Experiment 3. Irregular meningococcus No. 30 was exposed to the 
ultraviolet ray for 2,4, and 10 minutes. Reference to the first. table 
will show that distinct inhibition occurred with an exposure lasting 
ten minutes. 


| AGGLUTINATING RANGE 
ULTRAVIOLET 


SEE EXPOSURE 


1-100 | 1-200 1-400 | 1-500 | 1-800 | 1-1000 |Control 


minutes 


20 SP eee | Se 

182 2 Ac apa rates We Se ||: Shea 
I apa |) oo +] +] +] - 

30-2 E eee | 

183 4 4 at = = 7 a a = 
UGO Woes er crise 4) =o 

Otel cisco) |icrap pcre as sae a 

186 10 4 ate =e = = = = = 
UY 5600s Nes steals ots ||| ot | ea is 


All three sera showed marked agglutinating properties for the 
heterologous strains. The parameningococcus no. 60 was ag- 
glutinated in all cases to a greater degree than the specific antigen 
no. 30. Exposure to the rays brings out pronounced agglutino- 
gens for the regular and para types. 


STUDIES ON MENINGOCOCCUS 351 


| Experiment 4. Parameningococcus no. 60 was exposed to ultraviolet 
rays for periods varying from one-half to six minutes. 


ULTRA- AGGLUTINATING RANGE 


| 
SURE 1-100 | 1-200 | 1-100 1-500 1-800 | 1-1000 | 1-1200 | 1-1500 |Control 


188 | 13 


190 | 43 


~ 
Sis 
He 
| 
| 
| 
| 
| 
| 


S 
ae 
ae 
w8 
as 
at 
a 
at 
a8 
ter 
a 
[str 
(ate 
| 


191 | 53 


3;++)/ +), +]+i;}—-{/-;/-]-|]- 


The parameningococcus is not altered agglutinogenically by 
the ultraviolet ray with exposures ranging from one-half to six 
minutes. A serum prepared with such treated cultures is highly 
specific and indicates that the parameningococcus is a distinct 
type the serum of which rarely, if ever, possesses group agglu- 
tinins for the regular strain. However, it does include some 
agglutinin for the irregular type which is evidently more closely 
related to the parameningococcus. By the same token, a para- 
meningococcus no. 60 possesses no agglutinogen radicals common 
to a regular meningococcus, whereas the regular type includes 
within its protein molecule the elements of the parameningo- 
coccus. Another point brought out by this experiment is the 
possibility of developing in rabbits sera of high titer by so few 
as three injections. This is especially marked in the last series 


352 FREDERICK EBERSON 


where parameningococcus was studied. Exposure of this or- 
ganism to the rays for four and one-half to six minutes appears 
to have made the antigens especially agglutinogenic for the 
homologous strain. 

The effect of exposing a regular type of meningococcus to 
ultraviolet light was next studied in the monkey instead of the 
rabbit. The results were even more striking. <A typical experi- 
ment is given. 


Experiment 5. June 17, 1918. Macacus rhesus. Regular menin- 
gococcus no. 4 was grown for two generations on glucose agar, then 
exposed to ultraviolet radiation for five minutes. The technic of 
immunization and serum study were the same as those described pre- 
viously. One-fourth culture of an eighteen to twenty hour growth 
injected intravenously. June 18, one-half of a similar culture injected 
intravenously. June 25, an entire culture of the same strain treated 
as before was injected intravenously. On July 1, the animal was bled 
from median vein. -The agglutination test was set up on the same day. 


AGGLUTINATING RANGE 
CULTURE 


1-100 1-200 1-400 1-500 1-800 1-1000 Control 
4 oe uy a ts ze ss me 
30 + + o os - - - 
60 ++ ] +4 |] ++ | + ok - - 


In summarizing these experiments the conclusion may be drawn 
that exposure to the action of ultraviolet rays results possibly 
in a definite protoplasmic modification of microérganisms. This 
change may be shown in part by immune properties which be- 
come apparent in sera obtained by the injection of such altered 
bacteria. The action of ultraviolet radiation appears to be a 
distinct principle unrelated to the influence of heat upon the 
bacterial protoplasm, since a physical change from this source 
has been ruled out by careful control of temperature. 


The nature of the bacterial modification induced by ultraviolet rays 


The next series of experiments was designed to illustrate the 
nature of the change induced by treating meningococci with the 


STUDIES ON MENINGOCOCCUS aa 


ultraviolet ray. With a definite exposure it was possible to 
repeat regularly the antigenic modification observed in the first 
experiments. ‘Therefore such intervals were selected as served 
to bring about the most inclusive range of agglutination. The 
agglutinins were studied by means of absorption tests. 


Technic: Suspensions of meningococci were prepared and treated 
exactly, as in Experiment 1. Rabbits were injected intravenously on 
two successive days with 1/100 and 1/50 of the treated culture, respec- 
tively, then given a rest period of six days, following which one-quarter 
of a treated culture was given. Seven days later the animals were bled 
from the ear, the serum was heated at 53°C. for fifteen minutes and 
agglutination tests were made on the following day. The method of 
Tulloch (17) for absorption of meningococcus agglutinins was not used 
since in that procedure the absorbing suspensions are too thin and are 
insufficient even when doubly exposed to an agglutinating serum. To 
insure satisfactory absorption, sera were twice absorbed according to 
the following procedure. The serum, diluted 1-50, was added in 
amounts of 5 cc. directly to glucose agar slants of eighteen-hour old 
cultures of the meningococcus, the suspensions being made up to a 
turbidity standard of about 8000 million per cubie centimeter. After 
thorough emulsification, the mixtures were placed in centrifuge tubes, 
incubated for two hours at 37°C., then stored in ice chest at 4 to 8°C. for 
sixteen hours. After centrifugation at high speed for twenty minutes, 
the clear supernatant serum was again treated as before. The second 
centrifugation was prolonged to twenty-five minutes at high speed in 
order to remove all possible clouding. Absorbed sera were then tested 
for agglutinins against the homologous types and against those which 
were found to be agglutinable as a result of the action of the ultraviolet 
rays on the given antigen. All series were set up with eighteen hour 
old cultures grown on glucose agar and suspended in physiological salt 
solution to contain 2000 million meningococci per cubie centimeter. 
Each serum was controlled with a non-absorbed specimen which was 
subjected in parallel to every detail of the experimental conditions im- 
posed by the method. A few typical results have been selected from 
duplicate sets of experiments. 

Experiment 6. Regular meningococcus no. 4 was exposed to ultra- 
violet rays at a distance of 36 inches for four minutes. 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 4 


AGGLUTINATING RANGE 


CULTURE 
1-100 1-200 1-400 1-500 1-800 1-1000 1-1200 1-1500 | Control 
4 marae Wisma linus Meanie: |e. cur ae a +E ri 
30 Grae il Dicer et el ee iba me re ts es Ss a 
60 + ok + - - - — - - 
After absorption with parameningococcus no. 60 
4 + aie ote ais = 25 ae a aa 
ORs, a5 si i 2s _ — = = = 
60 pr + — ps a — — — a 
After absorption with irregular no. 30 
4 ae aii a ta = aS == a oat 
30 _ - — _ _ — - - = 
60 = 7 a = a = = a = 
Control serum (not absorbed) 
4 ais + aie ts an Bie + - a 
30 oe ah cSt = _ a = = = 
60 ate Fi a Wi; = = ry 7 


Experiment 7. Irregular meningococcus no. 30 was exposed to ultra- 
violet rays for two minutes at a distance of 36 inches. 


AGGLUTINATING RANGE 
CULTURE 


1-100 1-200 1-400 1-500 1-800 1-1000 1-1200 Control 
4 ++ | + + + +: - - - 
30 ++ | ++ | +4 | +4 |] +4 ] 4+ + - 
60 ++ | ++ | +4 | +4 ] 4+ +- - - 


4 aa hs) eg ul ey, wes va eu 
30 ate + a Fi 25 ~ — = 
60 at als ar ata zo +: — = 

After absorption with para no. 60 i 

4 te a ai ee + + - - ' 

30 + + a: os ae Zs _ = ‘ 

60 - - = = 7 a ~ = , 

Control serum (not absorbed) i 

i 

4 a ae aie + a _ pat ee i 
30 oF te = 5F se + — = 
60 + an os xs + ob _ = 


STUDIES ON MENINGOCOCCUS 355 


Agglutinins newly developed in consequence of treating menin- 
gococci with ultraviolet light could be removed by a method of 
absorption, after which the sera so tested were still capable of 
agglutinating the other types with no appreciable diminution in 
antibody content. The results attest to some profound change 
in the bacterial cell. 


The presence of group-agglutinogens in single bacterial types 


If the theory be correct that micro-organisms contain radicals 
common to related groups of bacteria, then it should be possible 
to show that certain serologic manifestations developed artificially 
in a type organism are common to the untreated specimen. Ac- 
cordingly, representative type meningococci were utilized in the 
following experiments as the homologous absorbing agents of 
different sera. These sera were prepared by injections with the 
same strains exposed to ultraviolet light and at the time the 
absorption tests were made they were one month old. The 
technic of absorption was the same as that previously outlined. 
Protocols representative of duplicated experiments are given. 
The series included the results obtained with different degrees 
of exposure to the rays, but these are omitted for the sake of 
brevity. 


Experiment 8. Sera obtained from rabbits by injection with regular 
meningococcus no. 4 which was treated with ultraviolet rays, were 
absorbed with untreated meningococcus no. 4. 


AGGLUTINATING RANGE AFTER ABSORPTION 
CULTURE 


1-100 1-200 1-400 | 1-500 1-800 1-1000 1-1200 Control 
4 = i ie os & ae ae = 
30 _ — + 2 a — -- = 
60 + - ae ch = — _ _ 
Control serum (not absorbed) 
4 Seon a eae ---- ARSE = + + — 
30 aie aa ar ok _ — _ -- 
60 - + =e) jes - - - - 


2 


356 FREDERICK EBERSON 


Experiment 9. Sera similarly produced by systematic treatment 
with ray-exposed irregular meningococcus no. 30 were absorbed with 
the corresponding untreated strain. 


AGGLUTINATING RANGE AFTER ABSORPTION 


CULTURE 


1-100 1-200 1-400 1-500 1-800 1-1000 1-1200 Control 
30 - — — — — — — ~ 
4 + 25 = = = = = _ 
60 Bec sal eats ste = 25 Ze - _ 


Control serum (not absorbed) 


30 ++ | ++ | + - + — = = 
4 ++ | ++ | + + + ~ - - 
60 ++ | ++ | + + a + = - 


When sera containing newly formed agglutinins are absorbed 
with homologous types of meningococcus, the group-agglutinins 
are materially reduced for those organisms which appear to have 
antigenic radicals locked up within the major agglutinogenic 
complex. ‘The relationship of some of the types may thus be 
shown. A regular type meningococcus contains antigenic ele- 
ments common to the other groups. An irregular strain such as 
no. 30 appears to have in its make-up radicals largely of a regular 
type such as the no. 4 and to a slight degree those of a para 
strain represented by the no. 60. These results are instructive 
if compared with the behavior of sera which have been absorbed 
with organisms known to contain normal group-agglutinogens. 
As is well known, the secondary or minor agglutinins cannot be 
detected after the serum has been absorbed with the organism 
which has been used for immunization. 

To carry this conception a step farther it is important to show 
if possible, that the immune properties developed in a micro- 
organism truly represent changes in the bacterial complex. If 
the theory be correct, we should be able to establish the relation 
between newly induced immune reactions and those present in 
the inciting agent. To test this point, the following experiments 
were devised. 


STUDIES ON MENINGOCOCCUS 357 


Experiment 10. Sera were prepared in rabbits by injections of regu- 
lar meningococcus no. 4 which was subjected to various exposures of 
ultraviolet light. These sera were then absorbed with the same strain 
of meningococcus modified by identical treatment. The results are 
summarized from a number of tests. 


AGGLUTINATING RANGE AFTER ABSORPTION WITH REGULAR NO. 4 


CULTURE 
1-100 1-200 1-400 1-500 1-800 1-1000 Control 
4 =a = a s. LS 3 
30 + _ _ — — — _— 
60 = = = of = = 
Control serum (not absorbed) 
4 7 are sir So =f = = 
30 — ae + + + — =— 
60 == == = = = = = 


Experiment 11. Sera were prepared as in the preceding experiment, 
with the use of an irregular meningococcus no. 30 treated with ultra- 
violet light. Absorptions were made with the same strain similarly 
treated. 


AGGLUTINATING RANGE AFTER ABSORPTION WITH IRREGULAR NO. 30 


CULTURE 
1-100 1-200 1-400 1-500 1-800 1-1000 | Control 
30 pe = S. = = 4. a 
4 = = =: = x a a3 
60 +: Spi giit, = - = = as 
Control serum (not absorbed) 

30 ++ | ++] + + | A 
4 aed a: =e = = — 
60 ++] +4 | ++ LIne tae 


The results are quite decisive. Preceding experiments have 
shown that newly formed agglutinins are specific for homologous 
or heterologous antigens possessing corresponding agglutinogenic 
radicals. The same tests when applied to an antigen so modified 
as to show its group affinities, establish the fact that the serum 


258 FREDERICK EBERSON 


reactions may be traced to profound changes in the bacterial 
protoplasm, as evidenced by the complete disappearance of newly 
developed agglutinins for the heterologous strains. 


DISCUSSION 


The experiments recorded in this paper establish several points 
which bear directly upon the problem of immunity and group- 
relationships of bacteria. In the first place it has been shown 
that ultraviolet light is capable of modifying bacterial antigens 
in such a manner as to improve their immunizing properties. 
Secondly, their serologic manifestations are rendered more in- 
clusive by this treatment. This method of study followed in 
analyzing the reactions of supposedly unique types of menin- 
gococcus has indicated that certain representatives of the species 
are composite antigens which contain cell radicals common to 
other members of the group. The relationships which may be 
brought out by a suitable absorption technic appear to be defi- 
nitely correlated with profound changes exerted upon the bac- 
terial cell by a physicochemical agent. ‘These results are contrary 
to those of Gordon (18) whose ‘‘superimposition” tests argue 
for specifically distinct types. 

The mechanism by which ultraviolet rays act upon bacterial 
protoplasm is not clear. The general impression gained from 
work done on this subject is that the rays coagulate the cell 
proteins. Those who have studied the problem, however, have 
limited their observations to the bactericidal action of the rays 
and have concluded that death of the organism results from 
coagulation of the protein. This is undoubtedly true for the 
method of experimentation followed by these workers since they 
failed to take into account the action of heat and therefore were 
not dealing with the effect of ultraviolet rays. By removing 
the factor of heat and working with ray exposures which are 
sublethal, one is able to arrive at a different conclusion. In 
this connection it is of interest to remark the close analogy which 
exists between the action of heat and some physicochemical 
agent in bringing about marked alteration of antigenic properties. 


STUDIES ON MENINGOCOCCUS 359 


That agglutinogenic properties may be influenced by heat has 
been shown by Joos (19). A serum produced with heated 
bacteria agglutinates both heated and unheated organisms, 
whereas a serum obtained by injection of unheated bacteria 
agglutinates heated organisms less highly. Similar results have 
been obtained by Kraus and Joachim (20) and Scheller (21). 
Of especial importance is Scheller’s observation that heated 
bacilli (typhoid) absorb agglutinins out of the sera more actively 
than do the unheated and that the highest agglutination titers 
can be obtained by injecting heated bacteria. The experiments 
with ultraviolet rays seem to embody some such change which 
is analogous to the alteration of bacterial cell bodies and, further- 
more, they offer a method which leaves the bacteria intact and 
thus capable of exerting their maximum immunizing activities. 

For want of a better term the idea of “molecular configur- 
ation” is suggested for the internal structure of bacteria and 
it is supposed that the characteristics of organisms are distributed 
uniformly throughout the cell. A physicochemical influence, 
such as ultraviolet light, acting upon the living cell, changes its 
internal structure by modifying all of the molecules which define 
the organism. The result is either a new combination of the 
elements already present, or a qualitative change exerted uni- 
formly throughout the cell substance. The new characters which 
are so imparted must not be taken to imply mutation or a trans- 
formation of one type of organism into another. 

Another point which might be discussed is the possible effect 
which sublethal doses of ultraviolet light may have upon the 
least resistant organisms. Assuming that these were killed off 
by the treatment, it would still be impossible to accept this in 
explanation for the altered immunity response to bacteria so 
exposed, since a reduction in numbers can have no influence on 
specificity to the extent observed in these experiments. 

A consideration of the physiologic action of ultraviolet light 
is linked with the question of enzymes (22). Since. these are 
concerned in the processes which we are able to study in vitro 
there are reasons for believing that changes effected by physico- 
chemical agents may be correlated with alterations of such 


360 FREDERICK EBERSON 


enzymes or with an activation of precursory enzymes in the cell. 
That the ultraviolet ray may have a marked effect on tryptic 
and diastatic ferments has been noted by some investigators, 
among whom may be cited Hertel (23). The method by which 
the rays are able to change the action of enzymes is not explained 
easily. We believe that we have given experimental evidence 
that some such change may be responsible for unusual serologic 
reactions. 

The practical application of these findings is suggestive. In 
view of the alterations which ultraviolet rays can produce in 
micro-organisms without impairing their antigenic value, new 
fields are opened for the investigation of bacteria which have 
thus far proved refractory as antigens. The advantage of 
employing living, yet innocuous organisms is obvious. In a 
further communication we shall discuss, from this viewpoint, 
some experimental work with Pnewmococcus. In this connection 
the effect of ultraviolet light on B. tuberculosis is of interest. 
It has been found, for example, that after exposure to these rays 
the organism loses its acid-fast properties (24). That the direct 
action on bacteria is of a comprehensive nature follows also from 
the loss of Gram-staining characters (25). 

There are further corollaries to be deduced from the experi- 
ments. It should be possible to develop a degree of immunity 
in animals by the injection of organisms which for each dose 
have been subjected to a correspondingly shorter exposure to 
ultraviolet rays. This simulates, in a measure, the cycle through 
which an infective agent passes in the animal body in which a 
gradual and a correspondingly diminished attenuating influence 
operates prior to the overwhelming of the body by virulent 
organisms. 

From the nature of group antigens it appears that a few 
strains may be used for vaccination against a large heterogeneous 
group of bacteria. We have here what seems to be a manifesta- 
tion of a single protein (antigenic) structure which gives to 
related organisms the properties whereby they may be classified 
within a certain group. Bearing this in mind, the explanation 
for much of the success with non-specific vaccine and protein 
therapy becomes relatively simplified. 


aie 


STUDIES ON MENINGOCOCCUS 361 


SUMMARY AND CONCLUSIONS 


Types of meningococci have been subjected to ultraviolet 
treatment and changes in antigenic properties observed. By a 
technic of agglutinin absorption it has been shown that such 
modifications are attributable to the physicochemical action of 
the rays. 

Carefully regulated exposures which exclude the influence of 
heat on the bacterial proteins exert a definite action on the cells. 
Such treatment appears to favor agglutinin response and to 
diminish the toxic effects of certain organisms. 

The group relationships of bacterial types are brought out 
quite clearly when under the influence of a physicochemical 
agent which “unlocks” or alters antigenic properties within the 
cell. 

The action of ultraviolet rays on micro-organisms as observed 
in these experiments suggests a method for building up an im- 
munity in animals by injections of bacteria exposed to the rays 
for constantly diminishing periods of time. The importance of 
working below the lethal dose of ultraviolet rays is obvious. 

Certain types of bacteria while supposedly containing only a 
major antigenic structure actually contain minor antigens also. 

There is evidence that a few strains, or probably a single strain 
of bacteria may sufhice for immunizing against a heterogeneous 
group of organisms. 

The experiments suggest that a single protein (antigenic) 
structure represents the element common to groups of biologically 
related organisms. 


REFERENCES 


(1) Ductavx: Comptes rendus de |’Academie de Sciences, 1885, 100, 119. 

(2) ArRLoInc: Comptes rendus de |’ Academie de Sciences, 1885, 100, 378. 

(3) Koca: Tenth International Med. Congress, Berlin, 1890. 

(4) Mienesco: Archiv fiir Hygiene, 1896, 25, 361. 

(5) Warp: Proceedings Royal Society, 1893, 52, 393. 

(6) Bucuner: Archiv fiir Hygiene, 1893, 17, 179. 

(7) CERNOVODEANU AND Henri: Comptes rendus des Séances de |’ Academie des 
Sciences, 1910, 150, 52. 

(8) JoDLBAUER AND TAPPEINER: Miinch. Med. Wochenschr., 1904, p. 737; ibid., 
p. 1096, p. 1139. 


362 FREDERICK EBERSON © 


(9) Taprpeiner: Miinch. Med. Wochenschr., 1900, p. 5; 1901, p. 1810; 1904, p. 714. 

(10) Frnssen: Chemiche Strahlen des Lichtes fiir Medicin und Biologie, Leip- 
zig, 1899. 

(11) CERNOVODEANU AND HENRI: Loc. cit. 

(12) Burge: Am. Jour. Physiol., 1915, 38, 399; 1917, 43, 429. 

(13) Bucuner: Loe. cit. 

(14) Desrossrs: La Presse Medicale, 1913, p. 453. 

(15) Naaier: La Presse Medicale, 1913, p. 612. 

(16) Gorpon: Special Report no. 3, Eng. Med., Research Committee, London, 
1917, p. 27. 

(17) Tuttocn: Jour. Roy. Army Med. Corps, 1917, 29, 66. 

(18) Gorpvon: Loe. cit. 

(19) Joos: Centralblatt fiir Bakteriologie, 1903, 33, 762. 

(20) Kraus anp Joacuim: Ibid., 1904, 36, 662. 

(21) ScHEenueR: Ibid., 1904, 36, 427, 694; 1905, 38, 100. 

(22) RicHarps AND Woopwarp: Am. Jour. of Roentgenology, 1917, 4, 564. 

(23) Herre: Zeitschr. fiir Allgemeine Physiologie, Verworn, 1905, 5, 95; 1904, 
ee 

(24) CERNOVODEANU AND HENRI: Loc. cit. 

(25) CERNOVODEANU AND HENRI: Loc. cit. 


HYPERSENSITIVENESS: ANAPHYLAXIS AND 
ALLERGY 


ARTHUR F. COCA 
From the Department of Bacteriology in Cornell University Medical College, 
New York City 


Received for publication June 3, 1920 


The subject of hypersensitiveness dates from the earliest 
observations upon human idiosyncrasy when it was noted that 
certain persons reacted, with peculiar symptoms, to contact 
with substances, such as foods, drugs, animal emanations and 
pollens, which to the great majority of individuals were entirely 
innocuous. Considerable added interest was infused into the 
study of this natural hypersensitiveness of human beings by the 
observations of Richet, Arthus, Theobald Smith and many 
others upon the experimental hypersensitiveness of lower ani- 
mals. One of the outstanding results of this stimulated interest 
was the generally accepted inclusion of all of the known phenom- 
ena of peculiar or unusual physiological reactivity under one 
heading—that of anaphylaxis. 

In the designation of such phenomena in one group the quality 
of peculiar reaction has generally been the only one taken into 
account, notwithstanding obvious and often radical differences 
among some of them as to their nature and causation. At 
present no etiological subdivision of these various phenomena 
is recognized. 

It is the purpose of this article briefly to present restrictive 
definitions of the terms in general use in the literature of the 
subject and to classify the known phenomena of peculiar physio- 
logical reaction in accordance with these definitions.! 


1A fuller exposition of this question is contained in the article on ‘‘ Hyper- 
sensitiveness’”’ in the forthcoming Practice of Medicine edited by Dr. Frederick 
Tice. 
363 


364 ARTHUR F. COCA 


It may be that the term hypersensitiveness will always be 
needed for use in a general sense applicable to conditions of non- 
specific, exaggerated or unusual reactivity on the part of a living 
organism. Such conditions are illustrated in the exaggerated 
sensitiveness to light of the eye in conjunctivitis. 

There are, however, two groups of the phenomena of peculiar, 
physiological reactivity, which, on account of definite constant 
features, can be separated from the others etiologically and 
these, though themselves of different etiology,’ can be associated 
under the heading ‘‘Hypersensitiveness.’”’ These two groups 
have been designated with the familiar terms ‘‘Anaphylaxis”’ 
and ‘“‘Allergy.”’ 

From the determining features of these two groups has been 
drawn the following definition of true hypersensitiveness; namely, 
a condition of specific or particular reactivity, with character- 
istic symptoms, to the administration of or contact with any 
substance in a quantity which to most of the individuals of the 
same species is innocuous. 

This definition should be amplified by the following restrictive 
explanation: 

1. The ‘“‘characteristic symptoms” are generally different in 
the different animal species for the same group of substances. 

2. They are uniform in any one species for various substances. 

3. Where the exciting agent possesses a normal physiological 
action; for example, the drugs, the symptoms of this action are, 
with few exceptions, different from those of hypersensitiveness 
to that agent.’ 

By the terms of the foregoing definition the tuberculin reaction 
of; Koch and the so-called toxin hypersensitiveness are excluded 
from the category of true hypersensitiveness; because: 

1. The symptoms of the tuberculin reaction are the same in 
all animal species. 


2 Etiology refers, here, to the origin of the underlying conditions, not to the 
exciting agents, which are often identical in both groups. 

3 All of these facts concerning the symptoms of true hypersensitiveness point 
to the conclusion that the phenomena under consideration are not dependent 
on any special property of the exciting agent, but solely upon a peculiar ad- 
justment of the hypersensitive individual toward that agent. 


HYPERSENSITIVENESS 365 


2. The symptoms of toxin hypersensitiveness are not different 
from those of the normal physiological effect of the agent. 

The phenomena of true hypersensitiveness are classified by 
the writer as those of anaphylaxis and those of allergy and, as 
it will presently appear, these terms are not used as synonyms 
but as mutually exclusive terms. 

Anaphylaxis is a state of true hypersensitiveness that is due 
to the presence, in certain tissues, of specific antibodies. The 
symptoms of anaphylaxis are caused by the meeting of these 
antibodies with the respective antigen in those tissues. This 
definition may be elucidated by the following facts: 

1. The state of anaphylaxis does not occur spontaneously but 
as a result of an immunological procedure (active or passive). 
It is experimental. The exciting agents are always antigens 
capable of inducing the formation of precipitating antibodies. 

2. Anaphylactic hypersentiveness is not inheritable; that is, 
it is not transmissible to the offspring from the male parent. 
The hypersensitiveness that is sometimes transmitted from the 
mother to the offspring is always identical in both with respect 
to the exciting agent. In such a case the transmission of the 
hypersensitiveness does not occur by inheritance but by a trans- 
feral of specific antibodies from the blood of the mother to that 
of the offspring through the placenta. 

3. An animal that is anaphylactic can always be rendered 
specifically and completely insensitive by a suitable (gradual) 
neutralization of the antibodies which are responsible for the 
hypersensitiveness. This procedure is known as “desensitiza- 
tion.”’ 

If we examine the facts concerning the causation and symptoms 
of human hypersensitiveness or allergy, we find full conformity 
with the definition of true hypersensitiveness. 

1. The condition is highly specific in the sense that it is exibited 
in about half of the hypersensitive individuals to only one 
substance (Cooke and Vander Veer), sometimes indeed, to only 
one chemical group or element. 

2. The symptoms are generally different from those of hyper- 
Sensitiveness in other animal species (the coryza, the gastro- 
intestinal symptom complex, eruptions). 


366 ARTHUR F. COCA 


3. The symptoms are the same for a great variety of different 
substances. 

In these three characteristics the hypersensitiveness of allergy 
resembles that of anaphylaxis. In all other respects, however, 
the two conditions are diametrically dissimilar. This dis- 
similarity is evident in the following comparison: 

1. The exciting agent in anaphylaxis is always an antigenic 
substance; that in allergy is often non-antigenic and cannot, 
therefore, induce a condition of anaphylaxis. 

2. The hypersensitiveness of anaphylaxis is always induced by 
previous introduction of an undigested antigenic substance into 
the body of an experimental animal. Such a procedure has not 
been shown to induce allergic hypersensitiveness in human beings 
and the weight of evidence is overwhelmingly against such an 
assumption. Allergy is often exhibited immediately upon the 
first contact with the exciting agent. 

3. The hypersensitiveness of anaphylaxis is not a heritable 
condition. That of allergy has been proven to be inherited, 
and when the mother is affected the hypersensitiveness of the 
offspring seems to be more often exhibited to a different sub- 
stance than the one that affects the mother (Cooke and Vander 
Veer). 

4. The phenomenon of desensitization, which never fails in 
anaphylaxis, is entirely wanting in allergy. It is true that a 
certain degree of lessened sensitiveness to natural contact with 
the exciting agent is often attained in allergy, which may be 
termed, clinical insensitiveness. In all such cases, however, the 
suitable administration of the exciting agent, by intracutaneous 
or subcutaneous injection, will demonstrate the persistence of 
the hypersensitiveness (Cooke). 

The differences between these two conditions, which have just 
been mentioned, may be summarized as follows: 

Anaphylaxis is an experimental, or induced, non-heritable 
hypersensitiveness due to the presence of specific antibodies in 
certain tissues. 

Allergy is a natural inherited condition of hypersensitiveness, 
which affects only human beings and is not dependent in any 
way on immunological antibodies. | 


HYPERSENSITIVENESS 367 


It is clear from this comparison that the true hypersensitiveness 
of human beings, which we have designated as “allergy,’’ must 
rest on an etiological basis that is quite different from that of 
anaphylaxis. 

The question naturally presents itself, cannot the hypersensi- 
tiveness of anaphylaxis be induced in human beings as well as 
in lower animals? 

It may be pointed out, here, that not all lower animals have 
been rendered anaphylactic.‘ The negative experiments of 
Yamanouchi and also by Uhlenhuth and Haendel with apes have 
been confirmed with a much larger number of animals by Auer 
in unpublished observations. ° 

For the certain recognition of anaphylaxis in human beings 
it would be necessary that peculiar and characteristic symptoms 
develop only after a reinjection of an antigenic substance. If 
the same symptoms are seen to be produced often by a primary 
injection of the antigenic substance or, indeed, more frequently 
after a primary injection than after a reinjection, then such evi- 
dence could hardly be used as indicating the existence of a 
condition of anaphylaxis. The mere absence of effect at the first 
injection by no means proves that the effect of the reinjection is 
due to the immunological mechanism. The difficulty of recog- 
nizing the hypothetical anaphylactic symptoms following the 
reinjection of a true antigen must increase if, as is actually the 
case, the symptoms observed under such circumstances are never 
different from those of allergy. 

It becomes necessary at this point to identify the allergic 
symptoms; that is, to determine what symptoms in human beings 
may be considered to be allergic. 

On the one hand allergic conditions must conform to the 
general criteria of hypersensitiveness and, on the other hand, 
they must present some feature which will exclude them from 
the category of anaphylaxis. 

Under the definition of hypersensitiveness laid down above, 
allergic symptoms may be identified by their specificity, their 


4 Only the usual criteria of outward symptoms and death are considered here. 
5 Personal communication. 


368 ARTHUR F. COCA 


unusual occurrence, their uniformity with different exciting agents 
and their difference from the normal physiological effect of the 
agent. 

It is a striking feature of allergy that these requirements are 
met by several symptoms’ or symptom-complexes. Thus, coryza 
bronchial asthma, gastro-intestinal disturbance, multiform cu-. 
taneous eruptions and sudden death following dyspnoea and 
respiratory failure, are all seen, in some individuals, to result 
from natural contact with or the administration of substances 
that are, to most individuals, entirely innocuous. 

All of these symptoms are found to occur under circumstances 
that exclude the possibility of an anaphylactic etiology. Such 
circumstances are: 

1. The non-antigenic nature of the exciting agent as in drug 
allergy (‘‘idiosyncrasy’’). 

2. The development of the symptoms at the first contact with 
the agent, as in many cases of “‘serum sickness” without incu- 
bation period. 

3. The demonstration of the factor of heredity in the causation 
of the symptoms. 

Having identified the various symptoms of allergy we may 
turn again to the question whether anaphylactic hypersensitive- 
ness does actually occur in man. We can consider this question 
most conveniently by examining the effects in human beings of 
the injection of therapeutic sera; because the material injected 
is a known and commonly employed anaphylactogen and because 
the conditions under which such sera are administered often 
approximate those of the anaphylaxis experiment. 

The injection of therapeutic serum, which is practically always: 
derived from the horse, is often followed by a reaction known as 
“serum disease.” The symptoms of this condition are those of 
allergy and they include the multiform eruptions, fever, edema, 


SIt is possible that there are other clinical manifestations of allergy that 
have not yet been recognized as such. 

The character of the symptoms in any particular individual is determined in 
part by the mode of contact with the exciting agent. For example, the exciting. 
agent of animal emanations if inhaled may cause coryza; if injected under the: 
skin it may cause sudden death. : 


~ 


HYPERSENSITIVENESS 369 


joint pains and sudden death. All of these symptoms and some 
other minor manifestations have been seen to follow a primary 
injection of the serum; that is, under a circumstance which, alone, 
would seem sufficient to rule out an anaphylactic (antibody- 
antigen) mechanism. Indeed the fact should be emphasized 
that the vast majority of the instances of serum disease occur 
after a primary injection. 

Additional evidence of the non-anaphylactic nature of the 
“serum disease” that results from a primary injection of serum 
is found in the constant absence of the phenomenon of desensiti- 
zation in that condition. 

It is important, here, to note the effect of a reinjection of serum 
upon the course of serum disease in individuals that presented 
that condition at the first injection. Under these circumstances 
the symptoms following the reinjection do not differ in character 
from those that were produced by the first administration of the 
serum. However, there is no doubt that the period of incubation 
is generally shorter at the reinjection and that the severity of the 
symptoms is often increased. It must be emphasized that no 
evidence exists to indicate that this difference between the re- 
actions occuring at a first injection and at a reinjection depends 
- on any immunological influence. 

There are left to be considered the relatively few instances 
in which, as in the anaphylaxis experiment in the lower animals, 
symptoms resulted in human beings from the reinjection but 
not from the first injection. These circumstances are analogous 
to those of active anaphylaxis and if allergy did not exist they 
could be accepted as representing anaphylaxis in man. How- 
ever, in none of these instances have symptoms been noted which 
were characteristically different from those of allergy. They 
have been either the usual eruptive manifestations or collapse 
followed, in some cases, by death. 

Hence, the only reason for suspecting these cases to be instances 
of anaphylaxis and not of allergy is the absence of symptoms at 
the first administration of the serum, and this circumstance 
cannot be satisfactorily explained until the mechanism of allergy 
is known. Certainly, it is not sufficient basis for a separation 
of these instances from the category of allergy. 


THE JOURNAL OF IMMUNOLOGY, VOL. V, NO. 4 


370 ARTHUR F. COCA 


In this connection it is well to bear in mind the instances in 
which the symptoms of drug allergy were absent at the first 
administration of the drug but appeared after a repeated admin- 
istration. Only very few such instances have been reported, 
probably because many of those who have observed them have 
overlooked their theoretical interest. 

I am permitted by Dr. John A. Fordyce to refer, in advance 
of their publication, to some instances of this kind that have 
occurred in his experience. In as many as ten or twelve cases 
no symptoms were caused by the first four or five intravenous 
injections of salvarsan. The fifth or sixth injection was followed 
by allergic symptoms (eruption) which recurred at all subsequent 
administrations, although sometimes long intervals of time (one 
to three years) elapsed between successive injections. In a 
personal communication, Dr. Fordyce remarks: 


The amount of the drug which produces the relapsing eruption 
apparently has no special significance. The amount of drug injected 
after these intervals [one to three years] has usually been a minimum 
quantity, about 0.25 to 0.3 of a gram of salvarsan. 


An exactly similar case has been observed in the dermatological 
clinic in Cornell University Medical College, which will be 
reported by Dr. Oscar L. Levin. 

Thus, it is evident that the mere absence of Sy onea upon 
a first injection of serum is not sufficient to indicate that the 
symptoms occurring upon subsequent injection are of anaphy- 
lactic origin; there was previous evidence, which is now con- 
firmed by the more numerous observations of Fordyce and that 
of Levin, that this occurrence is a characteristic phenomenon 
of drug allergy—a condition obviously unrelated to anaphylaxis. 

Even in the few cases, therefore, in which the possibility of 
the operation of an anaphylactic mechanism could be considered; 
that is, those in which symptoms developed only upon a rein- 
jection of serum, there appears to be no good ground for looking 
upon these as anything else than less usual forms of allergy. 

As is well known, the proof that the manifestations of anaphy- 
laxis are due to an antibody-antigen reaction was provided in 
Otto’s demonstration of passive sensitization. 


HYPERSENSITIVENESS 371 


The principle embodied in the experiment of passive sensiti- 
zation has been applied in serum and food allergies and even 
in non-antigenic drug allergy and the ultimate purpose of such 
applications of the principle has been to prove the anaphylactic 
nature of those allergies. However, the purpose of such experi- 
ments has been nullified by a fallacy which seems to have been 
overlooked. The technic of passive sensitization can be used 
to detect the presence of precipitin in an individual’s blood, 
but if the presence of precipitin has been demonstrated by this 
means it does not follow that the individual under examination 
is anaphylactic. For example: guinea-pigs have been passively 
sensitized with serum from an immunized monkey, yet in not a 
single instance in numerous experiments has the monkey itself 
been rendered anaphylactic. 

In applying the technic of passive sensitization in the study 
of allergy, the investigators have transferred the serum of the 
human individual almost exclusively to the guinea-pig, in which, 
then sensitization to the exciting agent of the allergy was looked 
for. However, in one instance reported by Ramirez,’ a quantity 
of blood (600 ec.) was transfused from an individual that was 
hypersensitive to horse dander to an anaemic patient who had 
previously exhibited no symptoms of allergy. Two weeks after 
the transfusion the recipient went for a carriage drive and was 
seized at once with an attack of asthma. The usual cutaneous 
test revealed in the patient a hypersensitiveness to horse dander. 
Unfortunately, it is not known whether the cutaneous hyper- 
sensitiveness existed previous to the transfusion or not. 

This observation is unique in the records. If it should be found 
that under similar circumstances hypersensitiveness could be 
regularly or often transferred from one human individual to 
another, it would be necessary to revise the conception of allergy 
that is presented in the foregoing pages. For the present, it 
seems proper to look upon the occurrence reported by Ramirez 
as an accidental coincidence. Some support is given to this 
view by the fact, which appears in the paper of Ramirez, that 


7 Journ. A. M. A., 1920, 73, 984. 


37 #2 ARTHUR F. COCA 


the same donor had supplied a larger quantity of blood (800 cc.) 
to another recipient, who, however, did not develop hypersensi- 
tiveness to horse. 

The age of onset of the clinical manifestations of allergy is 
different in different individuals. Cooke and Vander Veer have 
shown that the age of onset depends entirely upon a hereditary 
influence. It seems more probable that the recipient observed 
by Ramirez had just reached the age of natural onset of the 
horse allergy when the transfusion was carried out, than that 
the transfusion itself was the cause of the allergy. 

In view of the facts that form the basis of the foregoing dis- 
cussion, it seems necessary to conclude: first, that if anaphylaxis 
does occur in man, it does so only very rarely and secondly, that 
there is no positive evidence that anaphylaxis occurs at all in 
human beings. 


THE RELATION OF SPUTUM BACTERIA TO ASTHMA! 


FRANCIS M. RACKEMANN 
From Harvard Medical School and the Medical Clinic of the Massachusetts General 
Hospital, Boston, Mass. 


Received for publication May 5, 1920 


It is now recognized that asthma is a symptom complex which 
depends on one or the other of two great groups of causes: either 
the cause is a foreign protein which exerts its influence from out- 
side the body—“‘extrinsic’’—or it produces its effect from some 
focus, usually of bacterial growth and action within the body 
— intrinsic.” 

The treatment of this last group consists either in eradicating 
the focus by surgical means if possible or in the use of bacterial 
vaccines made preferably from the same strains which are causing 
the focus of infection or finally by surgery and vaccination 
together. 

Inasmuch as the search for a “focus” is usually futile, it is 
assumed that a chronic infection of the bronchi is responsible 
for the asthma and vaccines are prepared from the sputum 
bacteria. 

Treatment with such vaccines has been used with success by 
several workers. Montgomery and Sicard (1) isolated strepto- 
cocci, either hemolytic or non-hemolytic from each sputum and 
claim that out of 16 cases treated with autogenous vaccines, 
12 were cured, 3 improved and 1 unimproved. The work of 
Walker (2) is of greater interest: Out of 178 patients treated 
with vaccines, 28 (15.7 per cent) were found to be sensitive to 
bacteria and were treated accordingly with 75 per cent good 
results. Seventy-five other patients were treated only with the 
predominant organism with only 40 per cent good results. 

1 Read at the meeting of the American Society for the Advancement of Clini- 


cal Investigation, May 3, 1920. 
373 


374 FRANCIS M. RACKEMANN 


In an attempt to place the vaccine treatment of asthma on 
a more definite basis the following method of study was carried 
out in 40 cases. A smear from a suitable nugget of sputum 
obtained in a sterile vessel was made on a blood agar plate. 
Colonies from this first plate were then streaked in the sections 
of a second blood agar plate to insure purity and from these 
sections, tubes of dextrose broth were inoculated. The twenty- 
four hour growth in broth was then washed three times with 
saline containing 0.5 per cent carbolic acid, was finally suspended 
in the same and was killed by heating to 56°C. for one hour. 
The vaccines thus made were diluted so as to each have the 
same degree of cloudiness, as a suspension containing in 1 ce. 
500,000,000 bacteria. 

One hundred and twenty-nine organisms were isolated in this 
way. ‘There were: 


per cent 
Wi Non-hemolytic’streptococel) 1. fata Mies eros os oo on re 60 
l/AHemolyticustreptococeiss- siete Eee Eee eee 13 
Waspapbylococeus albus ss2cyc ere erect eickeul ic leioh Gees 13 
if, GOLAIN-NEPATIVE (COCK! i: .c.+5.0'- cep nee meee ayels «0-018 5.6 8D eee ene 6 
5 Staphylococcus aureus | 
3 Pneumococci (HINOPRERSH aE Breer satchls 5. .cneeaaene 8 


2 Gram-negative bacilli 
1 Diphtheroid 


Attention is called to the fact that of these organisms most 
numerous in the sputum of asthmatics, streptococci, chiefly 
non-hemolytic streptococci were found to make up nearly 75 
per cent. Also that in the entire group, pneumococci were found 
only three times. That this last finding is dependent entirely 
on the technic used and probably does not represent the true 
condition, is shown by the work of Stillman (3) who in studying 
the organisms present in normal saliva, was able to recover 
pneumococci by mouse passage in 47 instances in none of which 
were these organisms recovered by direct sputum culture on 
blood agar plates—indeed pneumococci were recovered by the 
two methods at the same time in only 11 out of 116 specimens 
of sputum studied in this way. 


RELATION OF SPUTUM BACTERIA TO ASTHMA 375 


With pure vaccines made from these organisms, intradermal 
tests were performed in the usual manner injecting about 0.02 
ec. of the carbolized bacterial suspension. ‘The tests were con- 
trolled by simultaneous injections of heterologous vaccines 
placed alternately with the autogenous. They were also con- 
trolled by injection into other patients. 

The early reading of the tests was made in one-half hour and 
the late reading in twenty-four hours. Positive early tests were 
found to consist of an urticarial wheal surrounded by erythema 
much like the usual tests to pollens, but positive late tests were 
found to resemble an inflammation with its area of redness, 
swelling and slight tenderness. In the interpretation, the com- 
parison of each test with its fellows was considered to be of far 
greater importance than any actual measurement of its size. 

Skin tests with autogenous vaccines were done in this way in 
39 patients; 25 of this group of 39 patients gave a positive test 
to one or more vaccines. There were 20 positive tests to an 
autogenous vaccine and 15 to a heterologous vaccine. In 17 
additional patients, tests were done with only heterologous 
vaccines and 9 of these patients reacted positively. Thus in 
the whole group of 56 patients, 34 or 60.7 per cent gave a positive 
skin test. This figure includes those who gave only a slight 
test and is therefore higher than Walker’s figure of 15.7 per cent 
positive. Attention is called to the fact that these tests were 
made by the intradermal method. 

Of the 358 individual intradermal injections in the whole 
series of 56 patients, there were 70 (19.5 per cent) which were 
classed as positive. This 19.5 per cent includes 7.5 per cent 
which were early positives, 7.2 per cent which were late positives 
and in addition 4.8 per cent which were positive both early and 
late. Thus the occurrence of positive tests was by no means 
limited to autogenous vaccines, nor except in one case, were 
more tests obtained with one organism or group of organisms 
than another. 

It is of interest to examine these tests with reference to the 
individual organism used. For this purpose only those organ- 
isms which were tested on 5 or more different persons, a total 


376 FRANCIS M. RACKEMANN 


of only 32, are here included. However with these 32 organisms, 
a total of 283 tests were made—an average of 9 each. 

Fourteen of these 32 organisms (almost half) gave no test 
at all and of the 4 organisms which gave more than 2 positives 
1 was a hemolytic streptococcus with 3 positives, 2 were staphylo- 
coccus albus with 4 each and 1 was a gram negative bacillus 
which seemed to be irritative because 10 positives were obtained 
in 18 tests. 

Treatment of these patients was carried out with small doses 
of pure vaccines given at seven-day intervals, each succeeding 
dose being regulated according to the amount of local reaction 
from the previous dose. If this local reaction was large, the 
next dose was made less. Any successful results appeared 
usually after 5 or 6 treatments, although in many of the un- 
successful cases as many as 12 or 15 treatments were given. The 
word “‘successful”’ is here used to denote a definite improvement 
in the subjective and objective signs and symptoms. It includes, 
however, only 2 instances of a virtual cure. 

It was found that treatment was successful in fairly close 
accordance with the presence of a positive skin test. Ten 
patients who gave a positive test to their own organism were 
treated with this organism, all of them with success. Six patients 
were treated with a heterologous organism, 4 of them with 
success. However, in the face of a negative test 10 other patients 
were treated, 8 of them with their own organism and 2 of them 
with a heterologous organism, but with no success whatever. 
The distinctions in prognosis between an early and late test 
were found to be unimportant. Good results in treatment were 
obtained after both; in fact of the 2 patients who were ‘‘cured”’ 
(in quotation marks) 1 gave an early while the other gave a 
late test. 

The organisms used in the treatment of the 14 successful 
cases included: non-hemolytic streptococcus 6, hemolytic strepto- 
coccus 3, staphylococcus albus 3, staphylococcus aureus 2, gram 
negative bacillus 1, Gram negative coccus 1; while those used 
in the 10 unsuccessful cases included: non-hemolytic strepto- 
coccus 7, staphylococcus albus 2, Gram negative coccus 1. In 


>. ~— — 


-_ 


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</L9 1° 129 |0°9 |’ |2°% |8°9 |2'F |2°2zj0'9 |¢°% joc | ® GT |2°9 |St {St joc joc |1'6 |Se joc joc ge  |(+T6r) 
¥/9 |F/T |8/90 ¥/9 |¥/1 |€/82\£/92 &/1Z|8/L1|8/PT\S/L S/F |8/'S |8/Ts P/ITIF/9 |8/1 |8/82\8/S0|S/TCIS/L118/FT\8/2L |8/F |8/%e | ST 
: aint 
TUN wauas sara WAUGS SUATHLOW oe 
> -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! «=<! =\- 50 0 0 0 0 50 


Totals in percentages: 82 per cent gave some degree of positive fixation; 14 
per cent gave doubtful fixation; 4 per cent gave negative reaction. 
* Observation. 


We have charted fifty cases that were diagnosed as clinically 
mild chronic gonorrhoea and fifty-eight control cases that were 
considered to be clinically negative. The chronic gonorrhoeal 
cases were examined by smear, culture and complement fixation. 
The control cases had the complement fixation test only. The 
incidence of complement fixation with the gonococcus antigen 
compares closely to that reported by McNeil (8) (4). We have 
not been able to group these reactions in relation to the duration 
of the disease because we have not been able to obtain such data 
from the patients. 


CONCLUSIONS 


The gonococcus complement fixation test is of undoubted 
value in chronic gonorrhoeal infections. 


506 JAMES D. SMITH AND M. A. WILSON 


In acute and early subacute infections it is on a par with the 
Wassermann test in the initial lesion stage prior to development 
of the secondary. 

A non-gonorrhoeic does not give a positive complement fix- 
ation test. (Table 2.) 

A gonorrhoeic may give a negative test in certain stages of the 
disease. 

REFERENCES 


(1) WituraMs: Collected Studies. Res. Lab. N. Y. H. D., 1911, 6, 29. 
(2) Crrron: Immunity. Translated by Garbat, ed. 1914, 184. 

(3) Scawartz AnD McNetu: Amer. Jour. Med. Sci., 1911, 141, 693. 
(4) McNett: N. Y. Med. Jour., 1914, 100, 350. 


EXPERIMENTS UPON THE PRODUCTION OF ANTI- 
HUMAN HEMOLYSIN WITH SPECIAL REFERENCE 
TO IMMUNIZATION WITH ERYTHROCYTES SENSI- 
TIZED WITH HEATED SERUM 


MOTOMATSU MATSUMOTO 
From the McManes Laboratory of Experimental Pathology of the University of 
Pennsylvania 


Received for publication September 23, 1920 


Undoubtedly the usual difficulties encountered in the prepara- 
tion of highly potent antihuman hemolytic sera by the immuni- 
zation of rabbits with human erythrocytes, have checked the 
wider adoption of an antihuman hemolytic system for comple- 
ment-fixation tests. It would appear to be the common experi- 
ence of most serologists that rabbits withstand immunization 
with human erythrocytes rather poorly, many animals losing 
rapidly in weight and succumbing before their sera are suffi- 
ciently hemolytic for practical purposes, either suddenly with 
symptoms of anaphylaxis or more slowly with general wasting. 

Three causes have been generally assigned for these results, 
namely: (1) that animals succumb to anaphylactic shock; (2) 
that fatal embolism may occur due to agglutination in vivo of 
the injected cells and (3) that human erythrocytes are highly 
toxic for rabbits. Various methods have been proposed from 
time to time for the preparation of antihuman hemolysin and 
in a recent study of these by Kolmer and Rule (1), the daily 
intravenous injection of 0.1 cc. washed human erythrocytes was 
found to give best results, the mortality being lower and hemo- 
lysin production good after twenty-one or more injections. 

Kolmer and Rule have assigned toxicity of human erythro- 
cytes for rabbits, and probably agglutination in vivo, as being 
the more prominent causes of deaths during immunization rather 
than anaphylaxis, which is generally held responsible. An 


507 


508 MOTOMATSU MATSUMOTO 


experiment by Vedder (2) tends to show that agglutination in 
vivo may play an important réle; Vedder heated the diluted 
serum of an immunized rabbit at 70° to 80°C. for one hour to 
destroy agglutinins and produce agglutinoids and found that 
washed cells exposed to this heated serum were rendered insus- 
ceptible to agglutination, and probably better borne when 
injected intravenously into a highly sensitized rabbit than 
untreated cells. 

At the suggestion of Professor Kolmer I have repeated Ved- 
der’s experiment and have also made comparative studies on 
the production of antihuman hemolysins and hemagglutinins in 
rabbits by the intravenous injection of unwashed cells carry- 
ing serum, thoroughly washed cells free of serum and washed 
cells rendered unsusceptible to agglutination either totally or 
partially by exposure to heated immune serum, to determine 
more definitely the roles of anaphylaxis and agglutination in 
vivo as the probable causes of fatalities among rabbits during 
immunization with human cells. 


TECHNIC 


Human corpuscles were secured by defibrination and centri- 
fuging; to one series of rabbits 0.5 ce. of these unwashed and 
~ packed cells suspended in 5 ce. of saline solution were injected 
intravenously every five days. 

A second series of animals received similar injections after the 
corpuscles had been washed with large volumes of saline solu- 
tion at least five times, the last fluid yielding negative reactions 
for proteins. 

A third series received similar injections of corpuscles washed 
three times, then exposed for one to one and one-half hours to 
10 cc. of antihuman sera which had been diluted 1:10 and 
heated at 70° to 80°C. for one hour, followed by two additional 
washings to remove serum. 

Four days after each injection each animal was weighed and a 
small amount of blood removed from an ear vein; the sera were 
heated at 56°C. for thirty minutes and tested for agglutinins and 
hemolysins as follows: 


Ee ————— =< a 


~ 


PRODUCTION OF ANTIHUMAN HEMOLYSIN 509 


Agglutination tests: 1 ec. of varying dilutions of serum were 
added to 1 cc. of 1 per cent suspensions of mixed washed cor- 
puscles from several persons, incubated in a water bath at 38°C. 
for one hour and read macroscopically after standing in a refrig- 
erator over night. - 

Hemolysin tests: 1 ce. of varying dilutions of serum were mixed 
with 0.9 ce. of 1 per cent suspension of mixed washed corpuscles 
of several persons and 0.1 cc. of 40 per cent dilution of mixed 
guinea-pig complement sera; these.were incubated in a water 
bath for one hour and the results read after standing in a refrig- 
erator over night. 


RESULTS 
When 0.5 cc.:of washed packed human corpuscles are mixed 


with 10 ec. of 1:10 dilution of immune antihuman serum heated 
at 70° to 80°C. for one and one-half hours and left standing at 


TABLE 1 
The agglutinability of human erythrocytes after exposure to various antihuman 
sera diluted 1:10 and heated at 70° to 80°C. for one and one-half hours 


RESULTS WITH IMMUNE SERA FROM 


CORPUSCLES 
Rabbit 1 | Rabbit2 | Rabbit3 | Rabbit 5 | Rabbit 6 
CUTTS. a 1:800 1:800 1:600 1:1600 | 1:2400 
Sensitized (exposed)......... 1:300 | 1:100* | 1:100* | 1:100* | 1:100* 


* Less than this. 


room temperature for one hour, the corpuscles are usually 
rendered less susceptible to agglutination by unheated immune 
serum. This is shown in table 1, giving a summary of the results 
of experiments with five immune sera; presumably the decrease 
in susceptibility of the heated corpuscles to agglutinins is due to 
the union of agglutinoids to the corpuscles, as explained by 


‘Vedder. 


These exposed corpuscles are, however, more susceptible to 
hemolysis, as shown in table 2, summarizing experiments with 
nine immune sera. This is due to the fact that immune hemo- 
lysins are probably more resistant to heat than the agglutinins 


510 MOTOMATSU MATSUMOTO 


and that the corpuscles were actually sensitized with unchanged 
hemolytic amboceptors rather than with amboceptoids. 

Table 3 gives a summary of the results of agglutination tests 
with the sera of rabbits tested four days after each of nine injec- 
tions of unwashed, washed and sensitized (to heated serum) 
corpuscles; table 4 gives a summary of the hemolytic titers of 
the sera. 

Charts 1 and 2 show in a graphic manner the average aggluti- 
nating and hemolytic titers of the animals immunized with the 
unwashed, washed and sensitized corpuscles. 

Table 5 shows the hemolytic titers of the sera of those rabbits 
surviving nine injections of the unwashed, washed and sensi- 
tized corpuscles; each serum was diluted 1:40 after heating at 


TABLE 2 


The susceptibility to hemolysis of human erythrocytes after exposure to various anti- 
human sera diluted 1:10 and heated at 70° to 80°C. for one and one-half hours 


HEMOLYTIC ACTIVITIES OF SERA FROM 


CORPUSCLES ss 
Rabbit | Rabbit | Rabbit | Rabbit | Rabbit | Rabbit} Rabbit 
2 3 5 6 7 9 10 


Tampa meks cue ceca aaate mate 1:100] 1:200) 1:100) 1:200} 1:200/1:100*} 1:100 
Sensitized (exposed)...........{ 1:800} 1:200} 1:800] 1:800} 1:800|1:800 | 1:800 


* Less than this. 


56°C. for thirty minutes and tested in amounts of 0.1 to 0.5 ce. 
with 1 cc. of 1 per cent suspension of corpuscles and 0.1 cc. of 
40 per cent of guinea pig complement. 

Table 6 shows the weights of the animals during the period of 
immunization as an index of the influence of immunization upon 
the general condition of rabbits receiving injections of unwashed, 
washed and sensitized corpuscles. 

These results may be summarized as follows: 

1. Rabbits receiving intravenous injections of 0.5 ec. of packed 
unwashed, washed and sensitized corpuscles (exposed to immune 
serum heated at 70° to 80°C.) every five days showed no differ- 
ences in body weight and the death rate was the same. 


—————._ sleet Or t—“—<—‘“‘“—afhM| 


PRODUCTION OF ANTIHUMAN HEMOLYSIN 511 


TABLE 3 


The production of immune antihuman hemagglutinins 


AGGLUTINATION TITERS FOUR DAYS AFTER 


ee 
oe First |Second| Third | Fourth| Fifth | Sixth Seventh! Fighth| Ninth 
: injec- | injec- | injec- | injec- | injec- | injec- | injec- | injec- | injec- 

tion tion tion tion tion tion tion tion tion 


1:8 | 1:80 {1:800 |1:2000] D.7+ 
Unwashed corpus- 4} 1:32 | 1:256/1:800 |1:4000!1:2000|1:8000/1 :2000/1:2000/1:3000 
cles 1:16 | 1:128)1:800 |1:4000|1:4000/1:4000)1:2000/1:4000/1:5200 


1:32 | 1:200)1:1600/1:8000/1:4000)1:4000)1:2000| D.+ 
Washed corpuscles 4} 1:32 | 1:256/1:6400/1:8000)1:8000/1:8000)1:8000/1:4000|1:3000 
1:32 | 1:256/1:800 |1:4000)1:2000/1:8000/1:4000/1:4000/1 :4000 
1:8 1:128|1:3200|1:4000)1:4000)1:4000)1:4000/1:4000/1:4600 
Sensitized corpus- || 1:4 | 1:64 |1:800 | D.+ 
cles* 1:64 | 1:128)1:400 |1:4000 
1:64 | 1:256)1:3200)1:4000 


1:4000/1:4000}1 : 2000) 1 :4000/1:4000 
1:4000/1:4000/1 :4000)1:4000/1:3400 


* Corpuscles exposed to immune serum diluted 1:10 and heated at 70° to 80°C. 
for one and one-half hours and then thoroughly washed. 


7 Died. 
TABLE 4 
The production of antihuman hemolysin 
HIGHEST DILUTION IN 1 CC. PRODUCING COMPLETE HEMOLYSIS 
IMMUNIZATION 


After | After | After | After | After | After | After | After | After 
first |second| third | fourth} fifth sixth {seventh} eighth! ninth 


{ OF 15-0 0 | 1:10 | Died 
Unwashed corpus- 0 0 O | 1:20 | 1:50 |, 1:50 | 1:50 | 1:50 | 1:100 
cles | 0 0 QO | 1:40} — | 1:50 | — | 1:50} 1:100 
O | 1:20} 1:80 | — ; 1:50 } 1:50 | Died 
Washed corpuscles 0 O | 1:20 | 1:40} — | 1:50 | 1:50 | 1:50 | 1:200 
0 E28 |) £240) T80" | 2350982507 1-0250: | 12100) 1200 
0 OO LS1Os ee On e50F E50: | -1:250' | 12200 
Sensitized corpus- 0 0 0 | Died 
cles ) 0 Oo 10) [ESO 1250 1 ts50) | 1:50.'|- 2:50 
0 | O | 1:40 | 1:80] — | 1:50 | 1:50 | 1:50 | 1:100 


* Hemolysis absent or incomplete in 1 cc. of 1:10 dilution. 


Average Four Days after Injections of Corpuscles. 


z| 
aN 
cr 
a 
a 
A 
2 | 
@ 
ct 
a 
~“] 
oe 
a 


sitera _| aot_| ona | sea | stn | sen] cen | rtm | oom | oon | 
1:7000 Fa is | aS eh 
118000 pe tet 
afeooe Wisse | cP a ZN a 
1:4000 Pe YS 2 SS ee 
1:3000 ae a a i Se ees 
112000 |} pp} +} Nef 
coe Ege Eire 20 a eS a ee 
So SRA mr See 
1:250 ae PAO, dee pes Fe See ae: 
rere VA TOR 708 FS EE | a es 
1150 | ae 
1125 cA eS eee ee 
1:10 ea ae 2 ee hes) 


CuHaArtT 1. PrRopucTION OF IMMUNE ANTIHUMAN HEMAGGLUTININS 


By intravenous injection of rabbits with unwashed corpuscles. 
~------- By intravenous injection of rabbits with washed corpuscles. 
meee n eee eee By intravenous injection of rabbits with corpuscles, exposed 

to immune serum 1:10 heated at 70° to 80°C. for one and a half hours. 


Averago Four Days after Injections of Corpuscles. : 

11190 ea Ce a a ae ee 
1:160 a 
1160 ee A ss es Be 
1:40 (cmd bes cos ID. is ae aS 
a aa eect | ee 
1:20 A "4 Loi 


Tare : 7? ‘ j 
cate — z Pens! 
1210 a ace li ed ieee) 


CHartT 2. PRopucTION OF IMMUNE ANTIHUMAN HEMOLYSINS 


By intravenous injection of rabbits with unwashed corpuscles. 
-------- By intravenous injection of rabbits with washed corpuscles. 
mentee ne eee By intravenous injection of rabbits with corpuscles exposed 

to immune serum 1:10 heated at 70° to 80°C. for one and a half hours. 


512 


| 


PRODUCTION OF ANTIHUMAN HEMOLYSIN 513 
TABLE 5 
Final hemolytic activity of heated immune sera after nine injections of human 
erythrocytes 


DOSES OF HEATED SERUM DILUTED 1:40 
IMMUNIZATION 


Minweasited COrpuscles........2..-..-+2-+s+00% \ M M C C 
C 

\WECUEG UGelg chs Cl xe tei. Orn , sf 3 C 
M M M C C 

DenatEMed (COUPUSCIES 0... 5.5... eee ee wei eee M M M M M 


* M = marked hemolysis; C = complete hemolysis. 


TABLE 6 
The influence upon body weight of rabbits receiving intravenous injections of human 
erythrocytes 


BODY WEIGHTS IN GRAMS FOUR DAYS AFTER INJECTIONS 
IMMUNIZATION = 


Before | First |Second} Third | Fourth] Fifth | Sixth |Seventh| Eighth 


Unwashed corpus- | 1700 | 1800 | 1900 | 1800 | 1960 | 2000 | 2100 | 2200 | 2100 
- cles 1650 | 1500 | 1700 | 1600 | 1600 | 1700 | 1400 | 1500 | 1350 


Washed tenis! 1700 | 1400 | 1500 | 1500 | 1400 | 1400 | 1300 | 1500 | 1500 
| 


{| 2400 | 2100 | 2000 | 1700 | 1800 | 1700 | 1700 | 1850 | 1800 
Sensitized corpus- | 2300 | 2400 | 2200 | 2100 | Died | — — — — 
cles 2250 | 2400 | 2500 | 2100 | 2300 | 2300 | 2500 | 2560 | 2400 


2. Under these conditions, apparently, unwashed corpuscles 
carrying human serum were not any more likely to kill rabbits 
than washed cells, which tends to minimize the importance 
attached to anaphylaxis as a cause of death among immunized 
animals when the injections are closely spaced. 

3. Apparently corpuscles exposed to heated immune sera and 
thereby rendered less susceptible to agglutination but more 


514 MOTOMATSU MATSUMOTO 


susceptible to hemolysis, are just as toxic for rabbits as plain 
cells; this tends to minimize the importance attached to agglu- 
tination in vivo as a cause of death of rabbits during immuniza- 
tion with human corpuscles but does not exclude the possibility 
of fatal agglutination in vivo inasmuch as the corpuscles were 
not totally insusceptible to agglutination. 

4. The agglutinin and hemolysin production by unwashed, 
washed and sensitized corpuscles were practically the same, 
allowing for marked individual variations among animals. 

When rabbits were immunized with human corpuscles injected 
intravenously and intraperitoneally at intervals of seven days or less, 
the gradual loss in weight and high mortality are apparently due 
primarily to the toxicity of human corpuscles for these animals; 
this appears particularly true of animals succumbing one or more 
days after an injection. Sudden fatalities with convulsions are 
probably most frequently caused by agglutination and hemolysis 
in vivo rather than by true anaphylactic shock. 

Additional experiments have shown that the stromata of 
human corpuscles when injected intravenously into fresh rabbits 
(not previously injected) may produce immediate death and 
presumably by agglutination in vivo and embolism. Vedder had 
described a method of immunization with ‘‘shadow corpuscles,” 
hemolyzed after preliminary treatment with illuminating gas, 
which he believes superior to the usual method of injecting whole 
cells. Kolmer and Rule found this method fairly satisfactory 
for the production of hemolysins but in a series of comparative 
experiments the mortality rate was practically the same as that 
accompanying immunization with whole cells. 

Solutions of human erythrocytes prepared by dissolving each 
cubic centimeter of washed packed corpuscles in 5 ce. of sterile 
distilled water for two hours at room temperature followed by 
thorough centrifuging and rendering the fluid isotonic with 
sodium chlorid, have proven somewhat less toxic for rabbits 
upon intravenous injection than suspensions of whole cells. It 
is also noteworthy that the injection of these solutions results in 
the production of much less agglutinin than whole cells (table 7) 
and likewise somewhat less hemolysin (table 8); these results 


PRODUCTION OF ANTIHUMAN HEMOLYSIN 515 


are similar to those reported by Kolmer and Rule (1) and they 
support the findings of Bordet (3) and Levene (4) indicating 
that hemoglobin stimulates the production of hemolysins rather 
poorly. 

According to these results it would appear that the stromata 
of human corpuscles are more dangerous for the rabbit either by 
agglutination in vivo or by direct toxicity than solutions largely 
free of stromata; accordingly, immunization with watery solu- 
tions of human erythrocytes largely free of cellular fragments 
may prove the method of choice, although more prolonged 
immunization may be required for the production of sera of 
sufficient hemolytic activity for complement fixation tests. 


TABLE 7 
The production of immune antihuman hemagglutinins 


TITERS AFTER INJECTIONS 
ANTIGENS 


Wiiiete’corpuscles. ................| 1:2 1:40 1:640 | 1:700/ 1: 
Hemolyzed corpuscles.............. 4:2" 1:2* 1:64 1:45 ie 


* Less than this. 


TABLE 8 , 
The production of immune antihuman hemolysins 


TITERS AFTER INJECTIONS 
ANTIGENS 


Second Third Fourth Fifth Sixth 


Meaa@lereerpuceles.-..........-...:| 1:2* 1:30 1:30 1:45 1:24 
Hemolyzed corpuscles.............. 2 5 1:15 1:13 1:18 


* Less than this. 


SUMMARY 


-1. Immunization of rabbits with unwashed and washed 
human erythrocytes by intravenous injection every five days, 
produced similar results in influence upon body weight, mor- 
tality, hemagglutinin and hemolysin production. 

2. Exposure of human erythrocytes to antihuman serum 
diluted 1:10 and heated at 70° to 80°C., reduces susceptibility 


516 MOTOMATSU MATSUMOTO 


to agglutination and increases susceptibility to hemolysis; 
immunization of rabbits with these sensitized erythrocytes pro- 
duced results similar to those observed with plain washed 
corpuscles. | 

3. Stromata of human erythrocytes injected intravenously 
into rabbits may prove fatal; solutions of erythrocytes in dis- 
tilled water largely free of stromata (shadow corpuscles) were 
better borne. 


CONCLUSIONS 


1. The high mortality among rabbits immunized with human 
erythrocytes is probably due primarily to direct toxicity of the 
cells and agglutination and hemolysis in vivo, rather than ana- 
phylaxis. . 

2. The stromata of human corpuscles (shadow corpuscles) 
may prove fatal for rabbits when injected intravenously, prob- 
ably by agglutination in vivo. 

3. Solutions of human erythrocytes in distilled water largely 
free of shadow corpuscles are much less toxic and better borne 
by rabbits than suspensions of whole cells; these solutions pro- 
duce much less hemagglutinin and slightly less hemolysin than 
whole cells, but may prove preferable to the latter for the immu- 
nization of rabbits for the preparation of hemolysin. 


REFERENCES 


(1) Koumer, J. A., anpD Rue, A.: A study of methods for the preparation and 
preservation of hemolysins. Amer. Jour. Syphilis, 1920, July. 

(2) VeppEr, E. B.: The Production of Anti-human Hemolysin. Jour. Immu- 
nology, 1919, 3, 141-146. 

(3) Borpet, J.: Sur l’agglutination et la dissolution des globules rouges par le 
sérum d’animaux injectes de sang défibrine. Ann. Inst. Pasteur, 1898, 
12, 688. 

(4) Levenz, P. A.: On the production of hemolytic serum by injecting animals 
with different constituents of erythrocytes. Jour. Med. Research, 
1904, 12, 191-194. 


A’ NOTE ON THE NON-SPECIFIC PRODUCTION OF 
ANTIBODIES 


MOTOMATSU MATSUMOTO 
From the McManes Laboratory of Experimental Pathology of the University of 


Pennsylvania 


Received for publication September 23, 1920 


During the course of experiments upon the production of im- 
mune agglutinins and hemolysins by the immunization of rabbits 
with sheep and human erythrocytes,! the sera where examined 
for the presence of agglutinins and hemolysins for sheep and 
guinea-pig corpuscles and for agglutinins for B. typhosus as well, 
as bearing upon the svecificity of antibody production. Several 
investigators have reported experiments tending to show the 
non-specific production of such antibodies as agglutinins and 
hemolysins for sheep corpuscles during immunization of rabbits 
with human corpuscles; unfortunately these experiments were 
not always carefully controlled by a preliminary study of the 
sera of rabbits for natural antibodies and the selection of animals 
whose sera were free or almost so of these substances in order to 
avoid falling into the error of interpretating the presence of these 
natural antibodies as due to stimulation of body cells by a specific 
antigen. 

In these experiments macroscopical agglutination tests were 
conducted with 1 ec. of varying dilutions of sera heated at 56°C. 
for thirty minutes, and 1 ce. of 2 per cent suspensions of corpus- 
cles of persons, sheep and guinea-pigs; in addition forty-eight 
hour broth cultures of B. typhosus were employed. ‘These mix- 
tures were incubated at 38°C. in a water bath for one hour and 
the readings made after the tubes had stood in a refrigerator over 
night. 

1 Matsumoto, M. Experiments upon the production of antihuman hemolysin 


with special reference to immunization with erythrocytes sensitized with heated 
serum. Jour. Immunol. (this number). 


517 


518 MOTOMATSU MATSUMOTO 


The hemolysin tests were conducted with 1 cc. of varying dilu- 
tions of unheated sera, 0.1 ec. of 40 per cent guinea-pig comple- 
ment and 0.9 ec. of 2 per cent suspensions of corpuscles. These 
mixtures were incubated for one hour and the results read the 
following day. 

The tables give the averaged highest dilutions of sera producing 
well marked or complete agglutination or complete hemolysis. 

Rabbits were immunized with intravenous injection of washed 
human and sheep corpuscles every five to seven days; the sera 
were tested four days after each injection for agglutinins for 
human, sheep and guinea pig corpuscles and B. typhosus and for 
hemolysins for sheep, human and guinea pig corpuscles. 

The results are given as averages in tables 1, 2, 3, and 4. 


TABLE 1 
The influence of injections of human corpuscles in rabbits upon agglutinins for 
human, sheep and guinea-pig corpuscles and B. typhosus 


AGGLUTINATION TITERS FOR 
INJECTIONS 


H Sh Guinea-pi 
SoRpuacles Soronienlen spEunolee B. typhosus 
Pre limunanyee cree ne oe —* _ — = 
SOCOM. A ptshinny ct sere OF 1:4 _ —_ = 
PSE oh ee Nec 28 perms ooh 1:40 ~ — — 
Hourthee st pees coe ce 1:640 — = _ 
BEG Ree ser oe eo ete car 1:800 _ — — 
Siete wenn seis aes 1:480 ~ _ 7 
* — = No agglutination or incomplete agglutination in 1:2. 
TABLE 2 


The influence of injections of human corpuscles in rabbits upon hemolysis for 
human, sheep and guinea-pig corpuscles 
LE pORIERR Se) 9 ul hls 0 eieeay Pap Geek aes eee 
HEMOLYTIC TITERS FOR 


INJECTIONS ; : 

Human corpuscles Sheep corpuscles plein 
Preliminanyayacieiyie se esc —* 12 12 
SCCONGMM I eics acon octane _ 2 1 
AE) sirb io Wea so ae ae 1:30 1:4 12, 
IMO indy Meco we COE tee a 1:40 1:4 12, 
Batt hoet peeeer ye cc St asp eet 1:50 1:4 1:2 
SURG i  eereeree rcs avers 1:24 1:4 a2 


* — = No hemolysis or incomplete hemolysis in 1:2. 


NON-SPECIFIC PRODUCTION OF ANTIBODIES 519 


TABLE 3 
The influence of injections of sheep corpuscles in rabbits upon agglutinins for sheep, 
_ human and guinea-pig corpuscles and B. typhosus 


AGGLUTINATION TITERS FOR 


INJECTIONS ? : 
Sheep Human Guinea-pig 


corpuscles corpuscles corpuscles B. typhosus 
Preliminary ra SR eee —* _ - - 
SECGIG lod, ee 1:3 _ - £2 
UDI: a 1:26 _ — fe 
DQURUR: 1626s eee 1:40 — _ = 
DH 0s 1:80 — — _ 
Sissies = - (ee ee 1:480 _ - - 
*— = No agglutination or incomplete agglutination in 1:2. 
TABLE 4 


The influence of injections of sheep corpuscles in rabbits wpon hemolysins for sheep, 
human and guinea-pig corpuscles 


HEMOLYTIC TITERS FOR 


INJECTIONS 


Sheep corpuscles | Human corpuscles lal 
lel Min Aye oe aoe 12 —* 32 
Sariig|- 2-346 eee 1:6 _ te 
UNG. 5 63 S8 Ss 66e eee 1:40 — 1:4 
INCE 6 4 2 sae 1:480 — 1:12 
* — = No hemolysis or incomplete hemolysis in 1:2. 
SUMMARY 


1. The immunization of rabbits with washed human and sheep 
corpuscles results in the production only of specific agglutinins 
and hemolysins for human and sheep corpuscles respectively. 

2. Occasionally rabbits showed an apparent slight increase of 
non-specific hemolysin or agglutinin, but these were so slight 
and irregular as to be explained by the normal variation of 
natural antibodies or experimental error. 


a4 a at Rye 5? AG ee GST Sleek a 4 Al ‘ 
hee te 2, 7 
a ba fe i AT 19 eh Pe ile hs ahs 7 x 5 
at) ; ‘ 
[A - ‘ ‘ Dey i 7m 
2 a ha Lhe ay aa ; 
Aa ors Jee Aik } dot dhe le Lis 
= + i ¢ J ‘ ‘ aioe 
a Yer 4 > 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 ah i 


‘ cle J 7 i ee er 4 
> 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 <A nite is hi pheeeS2 Sat i Wit un 


sv Wace Dae 7 
on, j 
‘ r Pi) red 
« ¥. 
ely dip rs Ndpentg Ne t * 


Net AAD 9 alan: aly ; i 


7 é « ‘| 
‘ae, ai , 
‘ F : ‘ Pye 
nae a a ie Paal bd ee - ed Rie wives ae : n/t wey wikia 


nt 


hy “hy eat ii ae, i iwen = mie ese Bo 
sa ae Merit) igi) Re 


IMMUNIZATION WITH BLACKLEG AGGRESSIN 


THOS. P. HASLAM 


From the Purity Biological Laboratories, Sioux City, Iowa 


Received for publication September 6, 1920 


In the last three years the methods of immunizing calves 
against blackleg have been changed. The older methods of 
immunizing calves with small doses of virus, based on the work 
of Arloing, Thomas, and Wright have been largely replaced by 
sterile, soluble products derived from the blackleg bacillus. In 
1917 Haslam and Franklin published a preliminary report (1) 
on the preparation and properties of natural blackleg aggressin. 
The publication did not have wide circulation, and several articles 
on the subject appearing subsequently (2) have failed to place 
the credit for the first practical use of a sterile blackleg immuniz- 
ing agent in the United States. The work of Haslam and Frank- 
lin was the first demonstration of the practical value of sterile 
blackleg aggressin in actually controlling losses from blackleg 
on infected premises, although Roux (3) in 1888 and later Schobel 
(4) had shown by laboratory experiments that the sterile antigens 
possessed marked immunizing properties. 


Roux collected the oedematous fluid from guinea-pigs dead after 
inoculation with blackleg, and filtered it through a Pasteur-Chamber- 
land filter. He succeeded in immunizing other guinea-pigs against an 
otherwise lethal dose of blackleg virus. Roux suggested that the 
product might have value in the control of blackleg. 

Schobel extended and confirmed Roux’s conclusions, using calves in 
addition to guinea-pigs in his experiments. Schobel’s technic of 
filtration and testing seems to have been at fault inasmuch as he stated, 
“The safest and quickest method (for testing sterility) is animal experi- 
ment, for it frequently occurs that the anaerobie culture process may 
prove the blackleg aggressin sterile, and at the same time guinea-pigs 
inoculated with 0.5 ec. of the aggressin may die of blackleg, in spite of 

539 


540 THOS. P. HASLAM 


the test.”” He showed doubt concerning the sterility of his product in 
a lengthy argument in which he proves that the immunizing value of 
blackleg aggressin cannot be due to the presence of a few organisms, 
but to specific soluble products. 


MEDIA FOR TESTING STERILITY OF BLACKLEG AGGRESSIN 


When our series of experiments on a sterile product for immun- 
izing calves against blackleg was undertaken, we first concerned 
ourselves with the best method of ascertaining the sterility of 
blackleg aggressin. The use of ordinary culture media, such as 
glucose bouillon in fermentation tubes or glucose bouillon under 
oil, was speedily abandoned, because it required heavy inoculation 
of very viable material to produce growth of the blackleg bacillus. 
Following the work of von Hibler (5) and of Foth (6), we sub- 
stituted the brain recommended by von Hibler for the meat 
pieces used in Foth’s medium, and obtained a medium which 
seemed superior to either ‘‘Hirnbrei’’ of von Hibler, or ‘‘Meat 
piece liver bouillon” of Foth. 


Grind liver, add one liter of water to each 500 grams of liver. Bring 
to a boil slowly and cook about 30 minutes or until the liquid is clear. 
Pour off the clear broth, add 1 per cent peptone and 0.5 per cent salt, 
adjust to a pH of 8, and it is ready to combine with the brain. Grind 
the brain, using the coarse grinder, and cook- at 3 pounds pressure 
about one and one-half hours. Fill flasks three-fourths full with 
mixture of brain tissue and liver broth, and autoclave at 6 pounds for 
three hours. 


This media will grow B. chauveaui and other anaerobes without 
protection from the air, and without preliminary boiling to drive 
off oxygen. Its reliability as a means of detecting minute traces 
of B. chauveaui in blackleg aggressin has been repeatedly checked, 
but one characteristic experiment will be cited. Dilutions of 
B. chauveaut culture were made in blackleg aggressin so that 
20 cc. of the aggressin contained from one one-hundredth of a 
cubic centimeter to one one-billionth of a cubic centimeter of 
blackleg culture. For the guinea-pig tests the same culture and 
the same aggressin were used but the dilutions were prepared 


IMMUNIZATION WITH BUACKLEG AGGRESSIN 541 


so that 10 ec. of aggressin contained from one-tenth to one five- 
thousandth of a cubic centimeter of the culture. The data are 
presented in table 1. 

TABLE 1 


Detection of B. chauveaui in aggressin by the use of brain medium and by the use 
of guinea-pigs 


BRAIN FLASKS GUINEA-PIGS 
Per one in 
‘ whic 
rs ioe a u mer me aes 7 aes Number Time oe roam 
aaceiddad asks observed pecs de- site guinea-pigs | observed guinea-pig 
blackleg 
cc hours ce. days 
0.1 0.1 3 4 335 
0.05 0.05 3 4 663 
0.015 4 48 100 0.01 3 4 none 
0.005 4 48 100 0.005 3 4 none 
0.001 4 48 100 0.001 3 4 335 
0.0005 4 48 100 0.0005 3 4 663 
0.0001 4 48 100 
0.00005 4 48 100 
0.00001 4 48 100 
0.000001 4 48 100 
0.0000001 4 48 100 
0.00000001 4 48 | 100 
0.000000001 4 48 75 


From the data in table 1 we conclude that the brain medium 
js many times more delicate and more dependable than guinea- 
pigs as a test for the presence of B chawveawi in aggressin. A 
test on three guinea-pigs is not sufficient to demonstrate the 
absence of B. chauveaui in small quantities, but the test on brain 
medium is capable of detecting exceedingly minute amounts with 
a great degree of regularity. 


METHOD OF PREPARING BLACKLEG AGGRESSIN 


Blackleg aggressin is prepared from the oedematous fluid of 
calves dying in one to three days after inoculation with a pure 
culture of B. chauveaui. The methods of identifying the culture 
have been reported in a previous paper (7). The pure culture 


542 THOS. P. HASLAM 


is prepared for calf inoculation simply by growing in brain 
medium, and straining through gauze. The oedematous fluid 
from calves dying of blackleg is collected partly by catching the 
fluid which comes out when the skin is removed and partly by 
pressing the affected tissue The oedematous fluid is cooled, 
centrifuged, and filtered through Berkfeld or Mandler filters. 
After filtration blackleg aggressin is tested for sterility and 
for absence of toxicity. The sterility of blackleg aggressin is 
proven by adding a sample of at least 10 cc. to each of three 
brain flasks containing 250 cc. of medium, and incubating for 
at least sixty hours Absence of toxicity is established by inocu- 
lating guinea-pigs with 10 ec. of the aggressin, or calves with 
50 to 100 cc. of the aggressin In these amounts blackleg aggres- 
sin rarely produces any local reaction, and very seldom any 
systemic reaction In our laboratories we have inoculated 
approximately 5000 guinea-pigs with 10 cc. each of sterile black- 
leg aggressin, and 150 calves with 100 to 50 cc. each of blackleg 
aggressin. In no instance have we demonstrated toxicity. 


IMMUNIZING EXPERIMENTS ON GUINEA-PIGS 


With this non-toxic, sterile product a number of immunization 
experiments have been carried out. On guinea-pigs the immun- 
ity, within limits, is proportional to the dose. This is illustrated 
by the experiment recorded in table 2, in which guinea-pigs were 
given varying amounts of blackleg aggressin, and later given 
one lethal dose of pure culture blackleg virus Guinea-pigs are 
immunized with difficulty with the smaller amounts of aggressin, 
although even 5 cc. of backleg aggressin gives marked protection. 
In 1917 we found that out of 1068 guinea-pigs inoculated with 
10 ec. blackleg aggressin and later given 5 lethal doses of blackleg 
culture, 83.5 per cent lived. The amount of culture given was 
0.5 cc. This proportion of immunity was later confirmed on 
about 4000 other guinea-pigs. 


IMMUNIZATION WITH BLACKLEG AGGRESSIN 543 


TABLE 2 
Immunizing experiments on guinea-pigs 


NUMBER NUMBER 


AGGRESSIN Pearman VIRUS oe Hal IMMUNE 
ce. ce. per cent 
0.5 10 0.1 8 20.00 
2.0 12 Oct 8 33.33 
3.0 ial 0.1 5 55.60 
5.0 33 0.1 11 66.66 

10.0 32 0.1 6 81.25 

Controls 18 0.1 17 5.55 


IMMUNIZATION EXPERIMENTS ON CALVES 


In calves as in guinea-pigs the result of an immunity test will 
depend on the amount of aggressin given, and the size of dose 
and the potency of the virus used in testing. Most of the tests 
on immunization of calves with blackleg aggressin have been 
made to determine some practical point in the preparation of the 
product and we have attempted to use the smallest amount of 
aggressin which would immunize against the largest amount of 
virus. After a number of experiments we found that 5 cc. of 
blackleg aggressin would usually protect against 5 ec. of pure 
culture virus. The virus used was of such strength that 0.1 ce. 
would kill two out of three guinea-pigs in forty-eight hours. In 
our preliminary report (1) a typical experiment was cited. We 
gave 8 cc. of aggressin to calves against 1 gram of dried meat 
virus which would kill guinea-pigs in 2 mgm. doses Under these 
conditions we got 100 per cent protection This method was 
satisfactory in general for demonstrating the protective action 
of the blackleg aggressin, but we did not feel it would show 
finer differences. That is a more sensitive test was required to 
determine such points as these; whether a certain preservative 
affected the potency of the product, or whether the oedematous 
fluid of a calf dying in seventy-two hours differed in value from 
the oedematous fluid of a calf dying in thirty hours. To obtain 
information on such points the dose of virus was increased so 
that occasionally a calf would succumb, and even the protected 


544 THOS. P. HASLAM 


ones would show a moderate reaction. In an experiment to deter- 
mine the effect of any factor on the potency of blackleg aggressin, 
a series of at least five calves is inoculated with aggressin prepared 
in the manner under consideration. Another series of the same 
size is inoculated at the same time with aggressin prepared in 
the usual manner. The two lots of aggressin should be of the 
same origin, or as nearly of the same origin as the conditions of 
the experiment will permit. The two series of calves should be 
inoculated with the same virus at the same time, along with 
at least five control calves. Although somewhat cumbersome 
and very expensive, the accuracy of the results seems to justify 
this method in ascertaining the fundamental principles involved 
in the production of blackleg aggressin. 


TABLE 3 


Immunization experiments on calves 


AGGRESSIN AGGRESSIN CONTROLS 
METHOD A METHOD B VIRUS ONLY 
DATE GIVEN DATE 
AGGRESSIN GIVEN VIRUS eT 
Died | "sey | pied | Tee | pied | pee 
May 29, 1919 | June 18, 1919 1 4 0 5 5 10 
June 6, 1919 | June30, 1919 2 4 1 4 4 5 
June 21, 1919 | July 8, 1919 0 6 0 5 4 5 
July 1, 1919 | July 16, 1919 1 5 0 5 5 5 
July 5, 1919 | July 22, 1919 2 4 0 5 3 5 
ECOG AR GO SBME Bae Oe 6 23 1 24 21 30 


In table 3 is given a summary of five sets of experiments. 
Each set consisted of three groups of calves; one group which 
received no aggressin, and two groups vaccinated with 5 ce. 
of blackleg aggressin originally the same but subsequently treated 
in different manners. Two weeks after inoculation these two 
groups as well as the first were inocuated with 5 ce. of pure 
culture blackleg virus. It will be noted that 77 calves were used 
in the experiment. Of those receiving virus only, 21 out of 30 
died in seventy-two hours. Of the 47 calves which received 
aggressin and virus only 7 died. Two of these, in the set inocu- 
lated June 6, 1919, one in each group, lived for over a week after 


IMMUNIZATION WITH BLACKLEG AGGRESSIN 545 


receiving virus, and they must have had a considerable degree 
of immunity. The experiments given in the table satisfactorily 
demonstrate the protective action of blackleg aggressin, in 
addition to yielding certain information concerning the effect 
of different methods of treatment. 

In the last three years natural blackleg aggressin has been 
used on five million calves, as nearly as can be estimated, and 
the field results have been highly satisfactory. The field results 
confirm the laboratory findings that in natural blackleg aggressin 
we have a sterile, non-toxic product, which confers active 
immunity against blackleg. 


CONCLUSIONS 


_ 1. A medium is reported which is much more sensitive than 
guinea-pigs for detecting the presence of B. chauveaui in blackleg 
agegressin. 

2. The immunity conferred on guinea-pigs by blackleg aggres- 
sin is proportional to the size of dose given. Ten cubic centi- 
meters of blackleg aggressin will immunize about 80 per cent of 
guinea-pigs against five lethal doses of pure culture virus. 

3. In calves as in guinea-pigs the result of an immunity test 
will depend upon the amount of aggressin given, and the size 
of dose and the potency of the virus used in testing. For detect- 
ing degrees of difference in aggressin, experiments were carried 
out in which the virus given was strong enough to cause a marked 
reaction in the protected calves. Under these conditions 74 
per cent of calves receiving 5 cc. of blackleg aggressin were 
immune to 5 ce. of pure culture virus. 

4, The findings set forth in the preliminary report concerning 
the immunizing value of blackleg aggressin have been confirmed 
and extended. 

5. As closely as can be estimated, natural blackleg aggressin 
has been used on five million calves, and the field results have 
been highly satisfactory. 


546 THOS. P. HASLAM 


REFERENCES 


(1) Hastam AND FRANKLIN: Circular No. 59, Agri. Exp. Sta., Kansas State 
Agri. College, January, 1917. 

(2) Warp, H. C.: Jour. Am. Vet. Med. Ass., 1919, 55, 394. 

(3) Roux, L. E.: Ann. de l’Instit. Pasteur, 1888, 2, 49. 

(4) Scnospe.: Centralbl. f. Bakt. 1912, 62, 296. 

(5) Von Hrster: Handbuch der Pathogenen Microorganismen by Kolle and 
Wassermann, 1912, 4, 800. 

(6) Fors: Ztschr. f. Infectionskrankh. der Hausthiere, 1909. 239. 

(7) Hasutam anp Lumps: Jour. Inf. Dis., 1919, 24, 362-365. 


A STUDY OF THE SPECIFICITY OF THE ABSORPTION 
OF ANTI-BACTERIAL PRECIPITINS 


CHARLES KRUMWIEDE anp GEORGIA M. COOPER 
From the Bureau of Laboratories, Department of Health, New York City 


Received for publication November 6, 1920 


In planning some experimental work on the immunological 
identity or non-identity of certain bacteria, it was found that a 
suitable agglutination antigen, that is, a stable suspension of 
uniform turbidity, could not be prepared from the cultures under 
consideration. This difficulty as is well known occurs with some 
bacterial types. ‘The question therefore arose whether under the 
circumstances we could employ the precipitin reaction. Should 
sharp differences be noted with this reaction, non-identity would 
be established. Should, however, the antigens of two or more 
cultures be found to give equal or nearly equal precipitation with 
an immune serum it would be necessary to resort to the antibody 
absorption method to determine the ultimate identity or non- 
identity of such strains. How far would such absorptions reveal 
specific differences? 

This question was raised because of the well known tendency 
of the precipitin-precipitinogen complex to bind immune bodies 
of all types as well as complement; a phenomenon utilized by 
Gay (1, 3), Chickering (1, 2) and by Weinstein (4) to separate 
the antibodies of immune sera. The question of the specificity 
of such an adsorption has apparently received little attention. 
Chickering found that an antigen of pneumococcus, type I or 
type II, removed only the immune substances for itself when a 
bivalent pneumococcus serum was employed. As these types 
are relatively very distinct and as they ‘‘cross-precipitate’’ to 
only a slight degree, these observations are of no significance as 
regards the question we have raised. Although we have not 
reviewed exhaustively the existing literature on bacterial pre- 


547 


548 CHARLES KRUMWIEDE AND GEORGIA M. COOPER 


cipitins, we do not believe that the fundamental question of the 
specificity of precipitin absorption has been studied with regard 
to bacteria which are closely related and therefore give similar 
or nearly identical reactions with precipitin sera. 

We were in possession, fortunately, of cultures and serums 
which were especially suitable for a study of the above question. 
Smith and Ten Broeck (5) called attention to the extreme degree 
of cross agglutination between B. typhosus and the avian types 
B. sanguinarium and B. pullorum when an immune serum for 
any one of these varieties was employed. 

Extending these observations to the precipitin reaction (6) 
we found that an anti-typhoid serum would precipitate the 
antigens of the above three types equally in the same dilutions 
and for the same time periods. The similarity of precipitation 
was so close that the strains could not be differentiated one from 
the other by this reaction (see following tables). We have also 
found that with an anti-pullorum serum similar cross reactions 
occur. Evidently, therefore, the serums and antigens of the 
three bacterial types mentioned above offer suitable reagents 
in testing the specificity of precipitin absorption. As a control, 
an antigen of B. paratyphosus A was used because antigens 
from this type were precipitated by the anti-typhoid serum in 
relatively low dilutions only. This type was therefore easily 
differentiated from B. typhosus by this reaction. 

It was planned to carry out the absorptions in two ways; 
first, by utilizing the precipitin antigen; second, by utilizing the 
bacterial suspensions as is done in agglutinin absorptions. 

The precipitin antigens' were prepared according to the 
method we (6) have previously reported. Briefly, this consists - 
of growing the bacteria on large areas of agar; collecting and 
suspending them in distilled water; dissolving them by the addi- 
tion of alkaline hypochlorite solution? and boiling; neutralization, 
precipitation by alcohol and the extraction of the sediment with 


1 The term antigen is used in the sense of areagent. The ability of antibody 
stimulation is not implied. 

2 For convenience we employed the commercially prepared solution, marketed 
as ‘‘Antiformin.” 


~ 


ABSORPTION OF ANTI-BACTERIAL PRECIPITINS 549 


0.8 per cent salt solution at 100°C. The final extract is clarified 
by centrifuging. This method, as has been noted, yields a very 
concentrated antigen. The different antigens were made as 
comparable as possible by utilizing similar amounts of culture 
and extracting the end-products in the same volumes of saline. 
The anti-typhoid serum employed was from a horse (no. 659) 
which had received intravenous injections first of killed and then 
of live B. typhosus over a considerable period of time. The 
anti-pullorum serum was similarly prepared by utilizing rabbits. 

The method employed in carrying out the absorptions with 
a precipitin antigen is illustrated by the following: 


To 0.5 ec. (or other quantity as noted) of antiserum was added the 
absorbing antigen in the amounts stated in the tables. The mixtures 
were incubated for three hours at 45°C. with frequent shaking of the 
tubes, which were then placed in the refrigerator over night. The 
mixture was then centrifuged and the clear supernatant fluid used for 
the tests. The tests were carried out by mixing 0.1 cc. of the test 
antigen and 0.1 cc. of the absorbed serum in small tubes and incubating 
them at 45°C. for three hours, after which the results were read. The 
dilutions in the tables are the final dilutions of the original serum and 
include the dilution by the test antigen. Where the serum was ab- 
sorbed, it includes the dilution by the absorbing antigen. The test 
antigen was the same as the absorbing antigen; both were used undi- 
luted. The antigen was not diluted as this would reduce the precipi- 
tability of the antigen and therefore would reduce the range of action 
of the serum. Our previous work had shown that with the three closely 
allied types mentioned, dilution of the antigen gave no greater specificity 


Tables 1 to 6 give the results obtained with the use of an 
anti-typhoid serum (horse no. 659) and an anti-pullorum serum 
(rabbit no. 339). Two different batches of antigens were em- 
ployed to exclude the possibility of error due to irregularities in 
the strength of these reagents. As a further safeguard, similar 
experiments (not tabulated) were carried out with another anti- 
typhoid serum (horse no. 652). The results check those obtained 
with serum no. 659. 

Tables 1 to 4 represent the results of two groups of absorption 
experiments. It will be noted that for each group of experiments 


550 CHARLES KRUMWIEDE AND GEORGIA M. COOPER 


TABLE 1* 
Precipitin tests—Anti-typhoid serum. Serum 0.6 cc. absorbed by 0.26 cc. of the 
precipitin antigens, and unabsorbed serum control 


ABSORBING B. sangui- : 
Rance B. typhosus nagiiiGn B. pullorum! B. paratyphosus A Salt solution control 
<q < < < < 
mn 2 S 2 n n 
i] * ~ I ~ 
Teer a) $/ 8/5) 3/8] s/a/3is/sial 2] 2|8|2e] 3/8] 814 
ANTIGENS...|| 6]°3] &| P| 6/3] &| B] O]'s] &| B S 5 im P ro) | & 2 
S|! S| Sls] &| 2] 3]S] ws} 3s] 4 a | & 3 a fon 3 
& s1S/ 8) FS els] a) S) ssa; & | 3 3 = I 3 a 
BIS! 6) S1213| a) a/ 2) Slater eins | a | Sl 2 Sal ee 
c[aicdisiaiaimlaim isis iol ato | | | Ml a] a] 
Serum dilu- 
tions: 
1:3 | 1} 4J-|-| 1) 1/-|-|-|-|-|- +4)44|+4+) — |++]+4+/4+4]41 
1:6 © |-|-|-|-|-|-|-|-|-|-|-|- 4+4]44)44+) — [44/44/44] + 
ete) eles PS Se eS Sorte | a |= a ee see 
1:24 = |-|-/-|-|-|-)-|-|-|-|-|- +\/+]}+}—-|+]+]+] 1 
1 ee Sele ia es ie en ee ape! 2 
1:956 9 jJj-|-—jJ-—J-—/-—)-/-/-!-/-I/-!- —-j—-lJ-l]- if 1 1) - 


* This table is representative of the results obtained with larger absorbing 
doses of antigen. A duplicate experiment using also 0.25 ec. of antigen showed 
no reaction in a 1:3 dilution with the first three types. 

Symbols: ++, +1, +, +, 1, degrees of reaction from strongest to weakest. 
Degree of reaction obtained with the 1:3 serum dilution with its homologous 
antigen taken as ++ for comparative readings. 


TABLE 2 
Precipitin tests—Anti-typhoid serum. Serum 0.5 cc. absorbed by 0.1 cc. of the 
precipitin antigens, and unabsorbed serum control 


ABSORBING ANTIGENS ......... B.typhosus |B.sanguinarium| B. pullorum Salépelapon 
§ | 8 | 
= & i = 
n 5 rs] a m n & mn & 
TEST ANTIGENS..........-.-.. EU |e lace, |W | ha 2 2| 8 A 3 a : 
3 |. Biles). Su) a leeumee le 8" |. Su) iaeee 
(Ge Pe Re ei) So A MBER 05) 
fe PR ree Mey Ih EPs || es || SR PSS eet |] os 
gio |e | a) oe eae ta lm | eee 
Serum dilutions: 
13 = i ee se | | ae 
1:6 T}+] 1) 1] 1] - 1) +) — |44+]44+/4+4+ 
i181 1} 1i/—-|}-|]-|J- 1} 2) =94+4+ he 
1:24 a eee | (ee en | ee ey ee ee Se 
1:48 Seah Vesoetl sca aes ay Ve eel rtd tee ieee (te == 


eS ee 


ABSORPTION OF ANTI-BACTERIAL PRECIPITINS 551 


TABLE 3 


Precipitin tests—Antityphoid serum. Serum 0.5 ce. absorbed by bacterial mass, 
and unabsorbed serum control 


eS B. typhosus B. sengul- B. pullorum| B. paratyphosus A Salt solution control 

Ansonoms {\" Macomb? | Sfass0.08 | Mass=0.15 |  f.agar lant - 
< < < < | < 
B) 13i |e] |2) |Sl (2 E = E = 
4) 8/8) 8/4| /2|el2)s/2/2| 2) 2/3) 2] 4] #| 4] 2 
BISlalalel sl alsl2/sialal =| | a] als] e] al a 
Slalalalalalaigisicianig| ol alalalalalala 

Serum dilu- 
tions: 

Sr | | itl] — (AFI EA +141 
or | | | | | | = iors ba (ll nd ecm LL GL LL (oe 
| — || — | —|—|—|— +l ae) [it] = 
1:24 J—|—J—|—J—|—|-J}-|-J]-|-!|- +) +)4+) —- /41441/4+ 1 
es | | — | — ||] = |— ie le ne ee 
1:96 3 |j—|—|-—|-—|-—|-|-—/-/-|-|-|- Se read aay 1 1 1| — 


* Approximate mass of bacilli after sedimenting suspension of growth from 
agar slants. Larger absorbing masses gave the same results. 


TABLE 4 


Precipitin tests—Anti-typhoid serum. Serum 0.5 cc. absorbed by bacterial mass, 
and unabsorbed serum control 


Salt solution 
control 


= 
s|ils 
ai\a|s 
B|2| 2 
sje | a 
++\4+/4++ 
++\4++/4++ 
++/+4+)++ 
+1 |+1| + 
=a) == == 


ABSORBING BACTERIUM........ B.typhosus (|B.sanguinarium} B. pullorum 
6 agar slants 3 agar slants 6 agar slants 
ABSORBING BERL be cos GP eRe { Mass=0.08 ce. Mass=0.04 ec. Mass=0.08 ce. 
ior 1 (ie: a ee ee = (a cl ee SS es 
= = os 
~ a _ ~ 
Stel) 2) 8-8) 2ial 8 
TEST ANTIGENS............... cS) SEN, SOON Wea. tenet el ces bags 
2, a = Fe = = r= a = 
Be TS aC Sa vee eet | oR 
AlmalmalAiaA lama lalalea 
Serum dilutions: 
1:3 +|/+]}+])+]+ +)+]+ 
1:6 1 1 ee ee ee i 
P12 —}/—J—o-l]—-Jo-t|- 1} = 1 
1:24 rl ae etd teen ef See ee 
1:48 Ne ey ect ee eo, ee 
1:96 Sey) = |) Se ee pb 


1} 1 1 


552 CHARLES KRUMWIEDE AND GEORGIA M. COOPER 


TABLE 5 
Precipitin tests—Anti-pullorum serum. Serum 0.6 cc. absorbed by 0.1 cc. of the 
precipitin antigens, and unabsorbed serum control 


Salt solution 
control 


ABSORBING ANTIGENS.......... B. typhosus 
| 3 
42 = 
2 a I a a 
il bes telecine aay | : 
TESWANTTGENS Ss: «oc s/ocln2s/0 lye: £ 2 6 & 2 i) 
Be |e eae 
ig S a | ss = A, 
A aia |) mea a 
Serum dilutions: 
1:3 —-}—/;/-—-/;-—-|]-—-] 1 
1:6 2452S) el eee eee fa 
1:12 ee ee ee ee ie 
1:24 — fo |e 
TABLE 6 


g 

Ae 
n 5 g 
a ees 
Eafe |S 
e| 2 | 3 
+ eS Q 
fea) fea} fea) 
1} —- 1 


Precipitin tests—A nti-pullorum serum. Serum 0.6 cc. absorbed by bacterial mass, 
and unabsorbed serum control 


ABSORBING BACTERIUM........ B. typhosus 


B.sanguinarium| B. pullorum 


Salt solution 
control 


{ 3 agar slants 


ABSORBING DOSE............. Mass=0:1co! 


3 agar slants 


3 agar slants 


Mass=0.1 cc. 

& 

s 

i] 
n 3 S 
fa & | 
° 3 ~ 
a & £ 
[27 re] Zz 
@ ||-¢ a 
fee) faa] [ea] 


one table gives the smallest absorbing dose which 


Mass=0.1 ce. 
Bie 
= 
nm - 
a | 
= | 8 | 2 
rel Ibe [ib S 
Ser Ps 
» a 2 
—Q —Q fea) 
_ 1 


| 
Bm 
3 | alee 
° 3 & 
SS op 2 
B| 2/2 
aima|a 
+4iti |4+ 
+1] + |41 
1 _ — 


resulted in 


complete or practically complete absorption and the succeeding 
table gives the next smaller absorbing dose which failed to 
absorb completely. The experiments are presented in this way 
to show that we have carefully considered the size of the absorb- 
ing dose in relation to the specificity or non-specificity of the 
results. It may seem that we should have used further inter- 


ABSORPTION OF ANTI-BACTERIAL PRECIPITINS Bao 


mediate doses. But even with the doses recorded, repeated 
experiments showed some variation in the degree of absorption; 
the limits of the delicacy of the method were, therefore, evidently 
reached. Furthermore, with the dose which did not wholly 
absorb, it is evident that the reduction is approximately the 
same with each of the antigens. In one instance (table 2), 
there is a greater reduction for one antigen (pullorum), which, 
at first glance, seems to indicate a specific difference. As this 
reduction is almost equally marked in the case of absorption by 
antigens of the other two types, this difference must be referable 
to the test antigen. ‘This brings up the question of our inability 
to standardize adequately precipitin antigens, which may be a 
factor in the non-specificity shown with absorption by antigens 
which precipitate equally. This question is considered more in 
detail below. 

It is evident from the tables 1 to 6 that the absorption of 
precipitins from the immune serums by the use of precipitin 
antigens, or even by whole bacteria, was specific only within 
narrow limits. With a heterologous type which gave little pre- 
cipitation, the absorption by the antigen of this type removed 
only the precipitins against itself (see B. paratyphosus A, 
tables 1 and 3). In the case of the heterologous types giving 
profuse precipitates, absorption by one of these types resulted 
in the complete removal of all the precipitins, even those active 
against the type homologous to the serum. 

Assume in the case of the above cultures, that satisfactory 
agglutinating antigens were impossible of preparation, and reli- 
ance had to be placed upon the precipitin reaction. If we allow 
this assumption, it is obvious that the precipitin absorption would 
have given unsatisfactory results where its application would be 
required because of failure of differentiation due to cross-precipi- 
tation and would have given satisfactory results where it was 
not required because differentiation was evident on direct 
precipitation. 

The results thus far recorded as regards non-specificity of 
absorption by the precipitin antigen itself were more or less 
expected. The result with absorption of precipitins by the 


554 CHARLES KRUMWIEDE AND GEORGIA M. COOPER 


bacterial bodies (tables 3 and 4) was so unexpected that it 
raised the question as to how much these sera cross-agglutinated 
and consequently what influence an equal absorbing mass of 
bacteria would have on the agglutinin content. The degree of 


TABLE 7 
Agglutination tests—Anti-typhoid serum. Unabsorbed serum 


HORSE 659 HORSE 532 
SERUM DILUTIONS B. B. . B B. j 
typhosus pee pullorum typhosus Epa er pullorum 
1:150 “= Fina sient “teat apar =o- 
1:1500 aisoin =f oie Sra orate Sinaia 
1:15 ,000 ateatg 1 1 ++ 1 1 
1:150 ,000 APaE = = =r = = 
1:300 ,000 ail — — =f _ _ 
1:600 ,000 ae = = — _ — 
1:1,200 ,000 + - — - ~ - 
Control — - _ _ _ = 
Symbol +-- = complete reaction, other symbols indicate decreasing degrees 


of reaction. ‘ 
Method.—Broth antigens (Dreyer) employed. Tests incubated for three hours 
at 45°C. Readings made after sedimentation in ice-chest over night. 


TABLE 8 
Agglutination tests—Anti-pullorum serum. Unabsorbed serum 


SERUM DILUTIONS B. TYPHOSUS B. SANGUINARIUM B. PULLORUM 
1:200 Sate AP AF qRSr 
1:400 +1 oe ++ 
1:800 + +1 +1 
1:1600 + SE + 
1:3200 it 1 -: 
1:6400 _ _ + 
Control — - - 


cross-reaction as regards agglutination is given in tables 7 and 8, 
and is of interest in showing how the cross-reaction by these 
two antibodies may or may not run parallel. In the case of the 
anti-typhoid serum the cross-agglutination is very much less 
marked than that recorded by Smith and Ten Broeck. 


ABSORPTION OF ANTI-BACTERIAL PRECIPITINS 655 


When the effect of a similar mass of bacteria (as above) on 
the agglutinin content of the serums was investigated, it was 
found that absorption of the serums by the homologous bacteria, 
using the same masses as in connection with the precipitin tests 
described above, reduced the specific titer only to a slight degree. 
Absorption by a heterologous strain, in the doses noted, also 
had relatively little influence on the height of the specific or 
group agglutinations. However, if the group titer was low, the 
heterologous strain might serve to remove the agglutinins for the 
absorbing type, but would leave the specific titer nearly unchanged. 

The non-specific results obtained with precipitin absorption 
in the above experiments induced us to try other related varieties 
of bacteria and their antisera in the attempt to verify the aston- 
ishing results recorded above. 

Antigens were prepared from three strains; a typical B. cholerae 
suis no. 131, an atypical suis variety no. 333 and a B. paratyphosus 
B. These were tested with an anti-serum for B. cholerae suis. 
The results are not tabulated. Only a moderate direct cross 
precipitin reaction was obtained. Precipitin absorption of this 
serum was also carried out by using bacterial masses of each 
variety for absorbing. The absorption was incomplete in some 
instances. Where it was complete, specificity was shown and 
where absorption was incomplete, the indication of specificity 
was so marked that further experiments with this serum were not 
made. These results indicate again that with slight or moderate 
cross-precipitation, specific results can be obtained by absorption. 

The atypical strain of porcine origin mentioned above was 
then tested with B. paratyphosus B serum and, as expected 
on the basis of our previous agglutinative study of these strains, 
rather marked cross-precipitation was obtained. Absorptions 
of this serum was carried out with the atypical B. cholerae suis 
type and with B. paratyphosus B, using bacterial masses for 
absorption. 

The precipitin experiments with this serum and these two 
bacterial types are recorded in table 9. All the tests made 
are given for comparison. ‘These experiments are of especial 
interest because they record the greatest fluctuation in the degree 


556 CHARLES KRUMWIEDE AND GEORGIA M. COOPER 


of cross-reaction that we have encountered. This variation in 
the direct reaction seems to parallel the degree of specificity 
obtained after absorption. Thus, in the ‘‘three slant”’ series the 
cross-reaction is least marked in the salt solution control and 
likewise the specificity after absorption is noticeable. In the 
‘“‘six slant”? series where the degree of cross-reaction in the salt 
solution control is more marked, the specificity after absorption 
is so slight as to be neglible. Inthe ‘‘nine slant” series, although 
the cross-reaction is again less, the absorption is complete and 
non-specific. 

The degree of cross-agglutination as well as the results of 
agglutinin absorption with the above two strains are given 
in table 10. 

Two questions have probably occurred to the reader in a 
consideration of the above protocols. First, why resort to 
absorption when a difference in direct precipitation is encountered 
comparable to that shown by the two strains given in table 9. 


TABLE 9 


Precipitin tests—Anti-paratyphosus B serum. Serum 1 cc. absorbed with bacterial 
mass, and unabsorbed serum control 


ABSORBING TYPE........ B. paratyphosus B B. Be ical Salt solution control 
° ° ro) 
oO o [>) 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 = <rasafnae. covers e's eter reveal eneve sie fale oti le > 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