ve
+e
; nya a pisos
ree, if :
vie viene
ie were
ve
toe
+
mas
srs Ad
vv."
Andean ra da ttt
3 Tire
vive ete ny.
vier - aa
vr,
§
or
MeRet es
¥
Sieaty
ele
rt?
peak aes
y
wre
.
\Veree
CV viws
eo pheeerwewengen eds (re Pere rwee
seein tte nit einai
AS
(ia ie Sean
We a ont ' be Vay pi Tr eh te
y ‘5 ’ CAS eG LE
j > { vt Rat Oh reey
fi } Bitty Va eee
. Aad ol 1 ” ; \ *
Ne A ol yy j ft
t : ;
f fit,
% a ’
3 1 i iy
* 7 ? 9
; DFR ek it Wh
Vo hae }
i bs] 5 ‘ a*
i } t
ie
¥
P ‘
‘
i
f i f
,! :
: x >
- ‘ 1
THE JOURNAL
OF
IMMUNOLOGY
VOLUME III
BALTIMORE, MD.
1918
906196
CONTENTS
NuMBER 1, JANUARY, 1918
A Rapid Method for the Production of Precipitin Antigen from Bacteria:
An Attempt to Apply it to the Determination of the Type of Pneumo-
coccus in Sputum. Charles Krumwiede, Jr., and W. Carey Noble... ...
On Von Dungern’s Indigo Test for Syphilis. B. Fujimoto.................
Extracts of Antibodies Obtained from Specific Precipitates of Typhoid-
Antityphoid Serum'Complex. Israel Weinstein........................
The Specificity of Intracutaneous Absorption. G. H. Smith and M. W.
(COOLS Ri ES ge ae RRR ci UGE tine Bir neh aA ULE OD LR CRA Sg OO ep ere RRR
Specific Reactions of the Body Fluids in Pneumococcic Infections. G. R.
ere yasmal © OS Pla chien tase soci ee ciel ae hae nc eis coe aco eee oe ae
NuMBER 2, Marcu, 1918
Studies on the Antitrypsin of Serum. B. Fujimoto................5.....2%
The Constancy of the Protein Quotient During Intensive Digestion and
Prolonged Starvation. Samuel Hanson: /.0.79 ss ee ie ones eee
The Immunologic Properties of Uveal Pigment. Alan C. Woods...........
The Examination of the Blood Preliminary to the Operation of Blood Trans-
1 AUN) PY Wat) Osea a O Tel Raa ie A EO A
Experiments upon the Passive Transfer of Antibodies from the Blood to the
Cerebrospinal Fluid. John A. Kolmer and Shigeki Sekiguchi..........
The Isolation, Purification and Concentration of Immune Bodies. A Study
CisimmunesHemoliysinss.> Mis Kosaka pects sae hia et
A New Method of Estimating the Antitryptic Index of Blood Serum. T.
Brailsford Robertson and Samuel Hanson.....................0ese-00-
The Non-Influence of Injections of Trypsin upon the Protein Quotient in
Blcodsserum.. Samuel iansona era eee eee eae
Effects of Intravenous Injections of a Colloid (Gelatin) upon Rabbit Sera.
LEWES “\ NGG) GN amen we OO aa eae a ea RA US Be Sar Par re Oe er
NuMBER 3, May, 1918
The Influence of Active Normal Serum (Complement) upon Meningococci. I.
The Opsonic Activity.of Fresh Normal Serum Alone and in Combination
with Antimeningitis Serum for Meningococci. John A. Kolmer, Ikuzo
oyamavanG LOLs) VMatsunamt: 2 occ cpp ils eit oom Weld sca onl aeieiae
The Influence of Active Normal Serum (Complement) upon Meningococci.
II. The Bactericidal and Protective Value of Fresh Normal Serum Alone
and in Combination with Antimeningitis Serum for Meningococci.
ois Viatsunami and John, iKolmerien...c sjecia atic an ueeeceeee
lil
101
109
131
139
147
157
1V CONTENTS
The Relation of the Meningococcidal Activity of the Blood to Resistance to
Virulent Meningococci. Toitsu Matsunami and John A. Kolmer....... 201
Experiments on the Production of Antipoliomyelitic Serum in Rabbits.
dean “he H. Dsenmis sos!) nr eee sooo oe oc Rein ee tote ohare ee 213
The Study of Problems of Immunity by the Tissue Culture Method. II. The
Tissue Culture as a Means for Quantitatively Estimating Toxin and
Antitoxin and Determining the Distribution of Antitoxin in Passively
Immunized Animals. Montrose T. Burrows and Yoshio Suzuki........ 219
The Study of Problems of Immunity by the Tissue Culture Method. I. A
Study of the Cells and Blood Plasma of Animals which are Naturally
Resistant and Others which are Susceptible to Diphtheria and Tetanus
Toxins. Yoshio 'Suzuki.l/.c a)... 25 {SUS Ah ee eee 233
NumBErR 4, Juuy, 1918
A Study of the Immunizing Properties of Bacterial Vaccines Prepared after
Various Methods. M. W. Perry and John A. Kolmer....... aay,
The Bactericidal Action of Whole Blood with a New Toctnrane, ae he
Determination. George D. Heist, Solomon Solis-Cohen and Myer
Solis=Cohens. ss .2a-d 5 Sh aes vse ea ee oars oe Eee 261
Complement Fixation with Protein Substances. Reuben L. Kahnand Archi-
bald MeNeil..... eS eee ee Le ee PETS Pes co Soke yoo Nae ase ae< 277
A Note on the Relation Between Proteolysins and Hemolysins. Archibald
MeNeiliand Reuben Il, Kahne. fac y:.<.1-e eee ee eee ao eeeeee 295
The Influence of Arsphenamine and Mercurie Chlorid, upon Complement
and Antibody Production. Ikuso Toyama and John A. Kolmer........ 301
Proceedings of the American Association of Immunologists. Fifth Annual
Meeting, held at the New Medical Laboratories and in the Hygiene Lab-
oratory of the University of Pennsylvania, Philadelphia............... 317
NuMBER 5, SEPTEMBER, 1918
A Contribution to the Study of the Complement Fixation Reaction in Tuber-
culosis: MC A> Wilson. 26. .840555,. ./..< agen ee ee 345
A Contribution to the Study of the Complement Fixation Reaction for
Tuberculosis. Hassow) von. Wedel... ).aeeccee ee eee oe ee 351
The Réle of Immunity in the Conduct of the Present War. John A. Kolmer 371
On the Mode of Action in Vitro and the Preparation of Hemolytic Antibodies.
A. 1K, Balls:and. John i: Korne.. 2280 oe ee eee ee ee 377
A Note on Bleeding Guinea-pigs and on Preserving Sheep’s Erythrocytes.
Sok J: Wenner 5556 oid sod she 2 pnaa's ko oleae & ee I ee 391
Studies in Pneumonia. VIII. A Skin Reaction to Pneumotoxin. Charles
Weiss‘and John A.. Kolmer:. ; ....0 20) saseme ee ee eee 397
A Method of Preparing Bacterial Antigens. James C. Small............... 413
A Study of Saponin Hemolysis. Tanemoto Furuhata...................... 423
CONTENTS Vi
NuMBER 6, NoveMBER, 1918
Active Immunity in Experimental Poliomyelitis. H. L. Abramson and
Herman Gerber........... ty Ao Oe eed cd aa MO 2? ot RE aN COURTS 435
Experimental Pollinosis. Preliminary Report. Henry L. Ulrich........... 453
Prompt Macroscopic Agglutination in the Diagnosis of Glanders. Olga R.
AON ALIE LETS c sols sb POS CORR IO ae Oa aT te tT tir an en aA meen, 463
ae
MAJOR RICHARD WEIL, M.O.R.C. ,
Richard Weil was born in New York City in 1876, next to the
youngest child in a family of seven children. His early educa-
tion was directly under the supervision of his mother, who was
an exceptionably able and clear-headed woman. At the age of
twelve he entered the sehool of Dr. J. Sachs where he prepared
for college, entering Columbia University, class of 1896. Al-
ready at this early age he showed a remarkably brilliant mind,
which reasoned clearly and delved deeply into all subjects that
presented themselves to him. At Columbia his intellectual ver-
satility showed itself in different lines of study—prizes in English
and Latin, honors in Greek, original research in science bearing
testimony to his assiduous study and careful and thorough work.
In recognition of his exceptional ability he was elected to the
Phi Beta Kappa at the close of his junior year and he was in-
vited by three departments, the Latin, the English and the bio-
logical to become a member of their teaching staffs. Before
graduating from Columbia, his mind had been opened to the
great possibilities presented by the study of medicine, and this
led him to decide upon medicine as a career despite all the other
attractive fields that lay open before him. Before entering the
College of Physicians and Surgeons, class of 1900, he spent a
summer at Wood’s Hole at the biological station.
In the Medical School he was an excellent worker and found
time in off hours to do experimental work in the physiological
laboratory. This work led to the degree of Master of Arts. On
graduating from the medical college he became an interne at the
German Hospital (October, 1900, to October, 1902), where he re-
ceived an excellent practical training in medicine and surgery.
After the completion of this period of practical training he went
to Europe for one and a half years study of medicine and its allied
sciences. His clinical work was done chiefly in Vienna under
Nothnagel, Neusser and Naunyn, and all three developed
1
ll EDWIN BEER AND CARY EGGLESTON
boundless enthusiasm in their eager pupil. Under Marchand
and vy. Recklingshausen his scientific longings were further
whetted and he obtained a glimpse of the wide reaches of scien-
tific medicine. While at Strassburg he was detailed to work in
a typhoid epidemic and acquitted himself most creditably.
In 1904 he returned to New York to enter upon a career as
physician-investigator, and in all the years that followed he
lived up to this professional ideal.
Upon taking up practice in 1904, he began his scientific in-
vestigations in the realm of medicine at Cornell Medical Col-
lege and in the following year was appointed Assistant in Experi-
mental Pathology. Four years later he became Instructor in
Cornell in the Department of Experimental Therapeutics,
which position he held until 1911 when he became Assistant
Professor in the same department. During this period he was
active in investigative work, which followed along the lines of
experimental pathology and in 1915 he was given the Assistant
Professorship in the Department of Experimental Pathology.
The following year this department was merged with that of
Experimental Medicine and Dr. Weil was given the chair which
he held at the time of his death.
In 1904 he was also made Adjunct Pathologist to the German
Hospital, which position he held until 1910. From 1908 until
1913 he was an Adjunct on the Visiting Staff to the Mt. Sinai
Hospital, thereafter being a member of the Assistant Attending
Staff. In the same year that he became an Associate Attend-
ing at Mt. Sinai Hospital he was given the positions of Assist-
ant Director of Cancer Research and Attending Physician in
the newly reorganized General Memorial Hospital. Upon his
appointment to the chair of experimental medicine in Cornell
University Medical College he resigned the position of Assistant
Director to the Memorial Hospital but continued there as At-
tending Physician. During these thirteen years Dr. Weil not
only performed his official duties in the positions Just men-
tioned, but found time to conduct a fairly active private prac-
tice and to delve deeply into scientific research. Among the
problems to the solution of which he bent his energies the two
MAJOR RICHARD WEIL ill
to which he devoted the major portion of his time were those
concerned with hemolysis and with anaphylaxis. To the former
of these problems he contributed much of value, but the work
by which he is known in this field is probably that concerning
the variable resistance of human red blood cells to the hemolytic
action of cobra venom.
In the field of anaphylaxis he stood almost alone as an expo-
nent of the cellular theory of its mechanism. Although his
work in support of this theory was vigorously criticized by many
of those who supported the humoral mechanism, and although
he met opposition to his views at almost every turn, he lived
to see some of the leading European investigators won over to
his side. His untimely death found two new studies on one of
the phases of this problem in course of publication. These
studies gave promise of clearing up one of the most hotly dis-
puted points regarding the mechanism of anaphylactic reactions
in the higher mammals.
All of the investigative work that came from his pen was
noted for its remarkable clarity of presentation and logic. He
was never content to approach any of the problems from a
single point of view, but always attempted to see all sides in
their true perspective and to attack the problems that pre-
sented from as many different points as possible. While the
workers in the same field in this country have not yet accepted
the cellular mechanism of anaphylaxis, as have some of those
abroad, the evidence which Dr. Weil adduced in its support is
of such a nature that we feel safe in predicting its ultimate
acceptance.
It is neither necessary nor fitting that we should dwell fur-
ther upon his contributions to medical science in these and
other fields for his published works are known to all who are
interested in the subjects with which they deal.
Dr. Weil was a member of many medical organizations in-
cluding the New York Academy of Medicine, the American
Medical Association, the American Society for the Control of
Cancer, the American Society of Clinical Investigation, the As-
sociation of American Physicians and others of a more restricted
1V EDWIN BEER AND CARY EGGLESTON
and narrower scientific nature. He occupied important offices
in a number of scientific societies, having been vice-president
of the American Association for Cancer Research and president
of the Society for Serology and Hematology, and of the Ameri-
ean Association of Immunologists. He also found time to per-
form the duties of an Associate Editor of the JouRNAL oF Im-
MUNOLOGY and of the American Review of Tuberculosis as well
as having been the Editor of the Journal of Cancer Research
from its foundation to the time of his death.
The range of Dr. Weil’s activities was truly remarkable but
it serves better than anything else to emphasize the striking
features of his character. He was a man of broad interests and
clear vision, and an indefatigable worker in every field that he
entered. But it was not alone in medicine that he stood out
above the average for he was keenly interested in the progress
of the times; he was a lover of art, of history and of literature.
He could talk well on almost any subject of an intellectual
nature and was always ready to sharpen his wits in a friendly
argument. He had a keen sense of humor and a ready wit com-
bined with a manner of easy cordiality. It was not one but
rather the combination of all of these attributes which won him
the lasting friendship of all who came to know him.
Finally, he was imbued with a profound spirit of patriotism
which led him to tender his services to his. country at the
outbreak of hostilities. Starting as a captain in the Medical
Officers Reserve Corps, it was but a few months before he was
made a major and detailed to Camp Wheeler as Chief of the
Medical Service. There his devotion to the great task of com-
bating pneumonia rapidly undermined his health and he him-
self fell a victim to the disease which he was endeavoring to
conquer.
EpwIn BEER,
Cary EGGLESTON.
MAJOR RICHARD WEIL Vv
BIBLIOGRAPHY
1899
Development of the ossicula auditus in the opossum. Ann. N. Y. Acad. Sci.,
1899, 12, 103.
An anomaly in the internal course of the trochlear nerve.. Jour. of Comp.
Neurology, 1899, 9, 1.
1900
On the evidence of the Golgi methods for the theory of neuron retraction. (With
R. Frank.) Arch. of Neurology and Psychopathology, 1900, 3, 265.
1904
Typhoid epidemiology. N. Y. Medical News, 1904, February 20 and March 5.
1907
Concerning a distinct type of hypernephroma of the kidney which simulates
various cystic conditions of that organ. Ann. of Surgery, September,
1907.
Hemolytic properties of organ and tumor extracts. Jour. of Med. Research,
1907, 11, 287.
1908
The hemolytic reactions in cases of human cancer. Jour. of Medical Research,
1908, 19, 281.
The hemolytic reactions of the blood in dogs affected with transplantable
lymphosarcoma. Archives of Internal Medicine, 1908, 1, 21.
1909
On the resistance of human erythrocytes to cobra venom. Jour. of Infectious
Dis., 1909, 6, 688.
On the specific acquired resistance of red blood cells. Proc. of the Soc. for Exp.
Biol. and Med., 1909, 6, 49.
Variation in resistance of human erythrocytes. Proc. of the Soc. for Exp.
Biol. and Med., 1909, 7, 2.
A method of testing the interaction of ferments and antiferments. (With S.
Feldstein.) Proc. of the Soc. for Exp. Biol. and Med., 1909, 7, 61-
Avoidance of hemolysis in transfusion. (With M. Rebling.) Amer. Jour. of
Surgery, 1909, 23, 268.
Serum reactions incancer. Jour. of Amer. Med. Assoc., 1909, 52, 407.
1910
The antitryptic activity of human blood serum. Amer. Jour. of the Med. Sci-
ences, 1910, 139, 714.
Experimental study of the antitryptic activity of human serum. Archives of
Internal Med., 1910, 5, 109.
Properties of ascitic fluids, especially in cases of cancer. Jour. Med. Research,
1910, 23, 85.
1911
On tumor-immunity in rats. Proc. of the Suc. for Exp. Biol. and Med., 1911,
9) 32.
vl EDWIN BEER AND CARY EGGLESTON
1912
Bemerkungen zur Arbeit von P. Kuschakoff. Zeitschr. f. Immunitatsforsch.,
1912, 13, 216.
An experimental study of antianaphylaxis. (With A. F. Coca.) Proc. of the
Soe. for Exp. Biol. and Med., 1912, 9, 102.
1913
The nature of antianaphylaxis. (With A. F. Coca.) Zeitschr. f. Immuni-
taitsforsch., 1913, 17, 141.
On a new factor in passive anaphylaxis. Proc. of the Soc. for Exp. Biol. and
Med., 1913, 10, 110.
The nature of anaphylaxis. Jour. of Med. Research, 1913, 27, 497.
A study of the blood in rats recovered from implanted sarcoma. Jour. of Exp.
Med., 1913, 18, 390.
The effects of colloidal copper with an analysis of the therapeutic criteria in
human cancer. Jour. of the Amer. Med. Assoc., 1913, 61, 1034.
The intravascular implantation of rat tumors. Jour. of Med. Research, 1913,
28, 497.
On antisensitization. Zeitschr. f. Immunititsforsch., 1913, 20, 199.
The relation of anaphylaxis to the problem of human disease. Louisville
Monthly Jour. of Med. and Surg., 1913, 20, 193.
Studies in anaphylaxis:
Ito IV. Jour. of Med. Research, 1913, 28, 243.
V. Desensitization. Jour. of Med. Research, 1913, 29, 233.
The cellular theory of anaphylaxis. Fourth International Congress, London,
August, 1913. Section of Bacteriology and Immunity.
Anaphylaxis in immune animals. Proc. of the Soc. for Exp. Biol. and Med.,
1913, 10.
Studies in anaphylaxis, VI: A study of the cellular theory by the graphic method.
Jour. of Med. Research, 1913, 30, 87.
1914
Studies in anaphylaxis: (Jour. Med. Research, 1914, 30, 299-364):
VII. The relation between antibody content and lethal dose in anaphylaxis.
VIII. The function of circulating antibody and the avidity of cellular antibody.
IX. The relations between partial desensitization and the minimal lethal
dose in anaphylaxis.
X. The presence of intracellular antigen as a factor in immunity.
XI. The share of intracellular antigen in immunity and in desensitization.
Theoretical considerations.
Studies in Anaphylaxis:
XII. Experiments in antisensitization. Zeitschr. f. Immunititsforsch., 1914,
23. 1s.
XIII. The activation of antibody by the cell. Jour. Med. Research, 1914, 32,
107-120.
The mechanism of anaphylactic shock. Proc. N. Y. Path. Soc., 1914, 14, 8.
1916
Sodium citrate in the transfusion of blood. Journ. of A. M. A., 1915, 64, 425.
Chemotherapy andtumors. Journ. of A. M. A., 1915, 64.
MAJOR RICHARD WEIL Vil
The treatment of parotid tumors by radium. Journ. of A. M. A., 1915, 65.
Anaphylaxis to formed or cellular elements. (With B. Denzer.) Proc. of the
Soc. for Exp. Biol. and Med., 1915, 12, 7.
Anaphylatoxin and the mechanism of anaphylaxis. Proc. of the Soc. for Exp.
Biol. and Med., 1915, 18, 2.
Equilibrium in the precipitation reaction. Proc. of the Soc. for Exp. Biol. and
Med., 1915, 18, 2. :
Equilibrium in the dissociation of precipitates. Proc. of the Soc. for Exp. Biol.
and Med., 1915, 18, 2.
Equilibrium in the combination and the dissociation of precipitates. Proc. N.
Y- Path. Soc., 1915, 15, 7.
1916
Note on a skin reaction in pneumonia. Jour. Exper. Med., 1916, 23, 1.
Immunological studies in pneumonia. Jour. Exper. Med., 1916, 23, 1.
Studies in anaphylaxis: (Jour. of Immunology, 1916, 1, 1):
XIV. On the relation between precipitin and sensitizin.
XV. Equilibrium in precipitation reactions, equilibrium in combination.
XVI. Equilibrium in precipitation reactions, dissociation.
XVII. On the coexistence of antigen and antibody in the body.
Chemotherapeutic experiments on rat tumors. Jour. Cancer Research, 1916,
Li.
The analysis of serum sickness. Proc. N. Y. Pathological Soc., N. S., 16, 3
and 4.
Characteristics of the precipitation reaction. Proe. of the Soe. for Exp. Biol.
and Med., 1916, 13, 8.
Studies in Anaphylaxis:
XVIII. The mechanism of delayed shock. Jour. of Immunology, 1916, 2, 95.
XIX. Simultaneous injection of antigen and antiserum. The anaphyla-
toxin theory of anaphylaxis. Jour. of Immunology, 1916, rl
1917
The excretion of congo red by the stomach. (With R. L. Cecil.) Proc. of the
Soc. for Exp. Biol. and Med., 1917, 14, 72.
Further studies in serum sickness. Proce. of the Soc. for Exp. Biol. and Med.,
1917, 14, 60.
Anaphylaxis in the dog. Proc. of the Soc. for Exp. Biol. and Med., 1917, 14, 6.
The immune reaction to tuberculous infection. Jour. of A. M. A., 1917, 68, 972.
The relation between antigen and antibody in the living animal. Jour. of Im-
munology, 1917, 2, 399.
The vasomotor depression in canine anaphylaxis. Jour. of Immunology, 1917,
2, 429.
Studies in Anaphylaxis:
XX. The reciprocal relations of antigen and antibody within the cell. Jour.
of Immunology, 1917, 2, 470.
XXI. Anaphylaxis in dogs; a study of the liver in shock and-in peptone
poisoning. Jour. of Immunology, 1917, 2, 525.
XXII. Anaphylactic reaction of the isolated dog’sliver. (With C. Eggleston.)
Jour. of Immunology, 1917, 2, 571.
A RAPID METHOD FOR THE PRODUCTION OF PRE-
CIPITIN ANTIGEN FROM BACTERIA: AN ATTEMPT
TO APPLY IT TO THE DETERMINATION OF THE
TYPE OF PNEUMOCOCCUS IN SPUTUM
CHARLES KRUMWIEDE, JR. anp W. CAREY NOBLE
From Bureau of Laboratories, Department of Health, New York City
Received for publication October 9, 1917
Various methods have been devised for obtaining a precipitin
antigen from bacteria. Many of these require prolonged incuba-
tion of broth cultures or more or less complicated methods of
extraction where growth on solid media is used. In some at-
tempts to use the precipitin reaction to determine the presence
of antigen in sputa and feces, we developed the following simple
and rapid method of making a bacterial precipitin antigen which
seems applicable to most,and probably to all bacteria with the
exception of the acid fast types.
To a heavy suspension of bacteria in distilled water, add suffi-
cient alkaline hypochlorite solution! to give a final concentration
of 5 per cent; boil over the free flame or heat in a water bath for
several minutes. If the bacteria have not dissolved, add more
of the alkaline hypochlorite solution and reheat. Repeat this
process till the bacteria are completely dissolved. This is shown
by the appearance of translucency. Add several drops of an
alcoholic solution of phenolphthalein. Then add ;-hydrochloric
acid till the color is just discharged To this neutralized solu-
tion, add several volumes of 95 per cent alcohol. A copious
precipitate which increases on standing, should result. The
precipitate is collected by centrifuging and decanting the super-
natant fluid. Add normal saline solution to the sediment to
give approximately a one to twenty solution by volume and
1 For convenience we have used the soluticn marketed under the trade name of
“antiformin.”’
1
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 1
2 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE
extract at the temperature of boiling water or boil over the free
flame for from three to five minutes. Then centrifuge to clear
the solution of the insoluble debris. The supernatant fluid is
the finished antigen. It can be prepared in a half hour or less
if necessary. This antigen can be reheated for sterilization at
100°C., apparently indefinitely.
Certain points must be considered in carrying out the various
steps of the process if a concentrated and economical antigen is
to be obtained Although broth cultures concentrated by sedi-
mentation in the centrifuge may be used with success, it is pref-
erable wherever possible, to use distilled water suspensions of
organisms grown over large areas of agar.
The initial suspension of organisms must be relatively heavy,
in fact, as little fluid as is needed should be used in making the
suspension. Furthermore, it must not be diluted unduly in
adding the alkaline hypochlorite solution or in the process of
neutralization. The alkaline hypochlorite solution should be
added as concentrated as is feasible in measuring the desired
amounts. If the neutralized extract is very concentrated several
volumes of alcohol will give a prompt, copious and almost com-
plete precipitation or so nearly complete that it would not be
economical to try to precipitate the remainder by further addi-
tion of alcohol. If, however, the extract is dilute, more alcohol
is needed proportionately to obtain a precipitate and this separ-
ates more slowly. Even in only moderately dilute extracts the
addition of more than several volumes of aleoho] may only cloud
the solution and flocculation will be very slight and much delayed.
The finished antigen, even after centrifuging, may be slightly
opalescent. ‘This is due to the presence of fine particles of bacte-
rial debris, which are difficult to throw down. As the concen-
trated antigen may be diluted from ten to forty or more times
and as the diluted antigen is very clear this opalescence is not a
factor. If for some special purpose it is necessary to use the
antigen in its concentrated form, the presence of this bacterial
debris must be considered as it is agglutinable and will therefore,
bind antibody and constitute part of the precipitate. In one
instance we collected the insoluble debris after boiling typhoid
PRODUCTION OF PRECIPITIN ANTIGEN 3
bacilli in a 30 per cent solution of ‘“antiformin” for one-half hour;
we washed the debris till no precipitable substance was present
in the washings and then suspended the debris in normal saline.
With dilutions of serum that gave precipitation with a finished
antigen, the bacterial debris was promptly agglutinated. The
control suspension without serum flocculated to some extent
and settled out. Even the severe treatment given did not
destroy the agglutinability of the bacterial “rests.” The presence
of the fine debris can be avoided to some extent by not breaking
up the sediment too vigorously in making the final extract. If
for some special reason the debris is objectionable it can be re-
moved by filtration.
The most striking point in the method outlined above is the
extreme resistance of the precipitable substances to the prolonged
heating in alkaline hypochlorite solutions, which allows the
rapid solution of dense suspensions with subsequent concentra-
tion of antigen. This is really the only original step in the method
as extraction with cold alkaline hypochlorite solution has been
used before. The other steps are all well known. That which
is new in the method is the combination of these steps so as to
allow the rapid preparation of antigen freed from extraneous
substances added in the process of preparation. We have used
the method to prepare antigens from pneumococci, using the
sediment from centrifugalized broth cultures, with agar cultures
of B. typhosus, with types of B. paratyphosus, with B. diphtherice
and with B. mallei and have obtained such satisfactory results
from all that we see no reason why the method is not applicable
to all bacteria soluble in an alkaline hypochlorite solution.
Two questions arise in relation to this method. In the process
how much precipitable substance is destroyed? We attempted to
determine this by preparing antigens with different concentrations
of alkaline hypochlorite solutions and employing different periods
of heating. By preparing these so that the concentration of the
end product was as comparable as possible we could see no
evidence that there was any appreciable destruction of antigen.
The other point is how far specificity is injured if at all, by
the severe treatment. If the factors entering into the immune
4 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE
reaction are considered; that is, standardization of antigen de-
termined by dilution and the addition of varied dilutions of
immune serum, with the subsequent employment of graded
dilutions of the serum against the standard of antigen thus
determined, as well as due consideration of the time factor; viz.,
the rapidity of the reaction, there is no evidence of loss of speci-
ficity. If any of these factors are not considered, especially if
the antigen is too concentrated, cross reactions become numer-
ous and marked, especially with closely allied organisms.
With types of bacteria that can be easily separated because
TABLE 1
Precipitin reactions—pneumococcus sera
SERUM TYPES AND DILUTIONS
PNEU MOCOCCUS
ANTIGEN | TIME Type I Type II Type IJI
| 1-1 1-4 1-9 ie ee 1-9 1-1 er 1-9
| hours
L|++*] + | 2] = pS) =) = pa
Typed... - of ee Pees pineal me = * ae a
} | | |
1} —-|- |} = |++}44+] 4] -— |] —] =|
Type II.... | |
Oo J 2) £ ) - | = 444/444) 44+) - | - | -
| | |
Type lil...4{ 2 [- = | =") cle ee. ee ee
-}-|]-]-]-}- }41} +] -
|
j
Antigen used = 0.2 cc. Serum solution = 0.2 ce.
*++4+ = profuse precipitate. ++,+]|,+, +,+ = decreasing amounts of
precipitate.
of their distinctly different agglutinability even a relatively con-
centrated antigen gives little evidence of cross reaction; that is,
the antigen so prepared loses none of its specificity. The fol-
lowing table illustrates this, only a slight cross being noted after
long incubation. The antigens were made from broth cultures
treated with antiformin and prepared according to the method
already given above.
It is not the place to go into the question of the relative speci-
ficity of precipitins and of agglutinins. Because we encountered
more marked cross reactions with the precipitin reactions with
_ail
be ite
te as
PRODUCTION OF PRECIPITIN ANTIGEN oO
very closely allied bacteria than was evident using the same
serum for agglutination reactions, we feared that we were de-
stroying some specificity in these cases To exclude this possi-
bility we prepared antigens according to the accepted method,
that is, prolonged growth in broth and subsequent removal of
bacteria by filtration and compared the results obtained from the
use of such antigens with the results obtained with antigens pre-
pared according to our method. With due regard to the factors
mentioned above, that is, appropriate dilutions of antigens and
of sera and consideration of the time factor, both types of anti-
gens gave comparable results. The broth antigens being very
much more dilute, did not introduce the factor of concentration
of the antigens somarkedly. The following tables are representa-
tive of the results obtained.
No final method of standardizing the antigens has been evolved.
A relatively satisfactory method, however, is to determine the
volume of sediment obtained by alcoholic precipitation, using
graduated centrifuge tubes, and making the final suspension by
adding up to twenty volumes of saline. With comparable
antigens in this way, and having obtained the appropriate dilu-
tion of one antigen against its homologous serum other antigens
can be similarly diluted in testing cross reactions of this serum.
This was not carried further as the practical application of the
precipitin reaction for differentiation of bacteria is relatively
little used. The main interest of the method, in our mind is
its great value in preparing material for teaching and for demon-
stration purposes and for courses in immunology. It should
also be of value in experimental procedures where a readily
obtainable supply of concentrated antigen is needed.
We have attempted to apply the method for diagnostic pur-
poses for the extraction of antigen from feces and from sputa.
The object in attempting to extract antigen from feces was to
determine whether we could detect typhoid antigen in stools and
in this way determine the presence of typhoid or allied bacilli
without the necessity of a prolonged bacteriological examination.
Thus far such attempts have been without success. Even
extracts of feces in which 95 per cent of the bacterial flora are
ee Ce
6 CHARLES KRUMWIEDH, JR., AND W. CAREY NOBLE
typhoid bacilli fail to yield a precipitable extract. Evidently we
are extracting other substances from the feces which appear ;
TABLE 2
Comparison of broth and of antiformin antigens f
ANTIFORMIN
ANTIFORMIN ANTIGEN ANTIGEN
BROTH ANTIGEN DILUTED 1:10
SERUM (oNDEGEEY) (ANTIGEN A) ees >
DILUTION
ORGANISM (ANTI- = 5 = = Pm = & |
| TYPHOID) | 3 Bye 5 set ae x
OMOEA GOS aaa 96/4 /| 22a ee
Le amir hopantaeneleaeen (dae (arcade lor
B. typhosus......... 126. esi etal ctrl gt ated tate let elton a
thes Ea fle Ne eto ae it) a lfee aey = =
( bebecal Co 222 Ws ihe SE EE TS Ca eels
B. paratyph ‘‘A’’.... 1:6 = = = = + 4] (Sea
1:12 ee ic ee (a be td all ae | Slee
f Pea = == Ss == = — | = | =
B. paratyph “‘B’’.... 126 — _ + — |SLC}+=]—|]-|=
(e125 =) = 4) Sn ee | —|=
1:1 ++ 44+) + | +i [+1] - ae
B. pullorum. .........: 1:6 a 1 ited ee + |+1)—-—|+|=
1:12 + {+] = | + HIl—-|= 72
1a spice mann beedueietr th lear ae | ag
B. sanguinarium. ... 600 [FR EL EEL) CR eae
toa2 “ie pric ler a= Ae el ek fit
a | = SLCFP=" SLC S| =
B. abortus equi...... . 136 = |) i= 1 SLEs) re = a
Oiler 8 Pe a ee es 8a | Fag | -
tae Oe ec esa Mee a a ery
Breas esis 32. | es | a eer eS) eS ee ee
| 1: 12) sulci Sale tle Sun eet | - i -
S1.C = slightly cloudy. C = cloudy. + = slight flocculation. + = dis-
tinct flocculation. + ; = flocculationmore marked. + + = heavy precipitate.
Antigen 0.2 cc. and serum dilution 0.2 ec. used.
* Overnight.
+ Symbols in these columns not comparable with those of very dilute antigen
in table 3. The symbols in each table are in comparison with most voluminous
precipitate obtained in series represented by a table.
TABLE 3
Comparison of broth and of antiformin antigens
BROTH ANTIGEN (1:5) ANTIFORMIN ANTIGEN
eae (1:100) (ANTIGEN B.)
ORGANISM EASE SOR!
r (ANTI-
TYPHOID) halt eae Tea bort Walt oo Toa hae
(| 1-10 +l} ] ++] tl ++] ++
LO oilecteslbe actA Metect= ht) be) toe
Bett DROSUSS. nase ener ee 1-50 + +]; |4+ a EA gall tke)
1-75 C. aE = —- + ata
[| 1-100 — + aE = CQ. ae
(| 1-10 — |SLOG|SLC} — Saat
1-20 Sailers 9) agli panes Cay se
Ba paralypie sche 6. <8... 1-50 _ — = ae as “
a eS, |e
|| 1-100 - = ix = a =
1-10 | C C e = =| ten,
omy |e | ot ee ADS oth a
Peemmatypn. | B??..........- 1-50 = as. im BY st es
1-75 = _ = = his =
1-100 = — = 2 aS =~
| Pete tel eta at a ete Metco steaks
20 | +i | +1 | ++] 4! net
1835 FOU UTILS ees Aids SOE 1-50 = + at. + | ata
| 1-75 SLC] = + — =i
1-100 — C + — ae
1-10 anil) |Waeseelarae) Sear = a
1-20 Ses eleltsese cei (se ae
B. sanguinarium*........... 1-50 + }/4+]-] +1 tie | cet
1-75 C C aa = at
1-100 = — te = a
(} 1-10 C C C a = +
1-20 ~ _ = = ke se
B. Abortus equinus.......... 1-50 = = a fea = stir
1-75 _ — = = es a3
1-100 — _ = pe vs) =
(i, 1-10 C C aa = = eet
1-20 - — =— as es ad
ESRACOUPE Rares ere Rate 2 oe 1-50 -- = = = 2 zs
1-75 — - — = = es
1-100 — — = zs a Ze
S1.C = slightly cloudy. C = cloudy. + = slight flocculation. + = dis-
tinct flocculation. + | = more marked flocculation. +-+ = heavy precipitate.
Antigen — 0.2 cc. and serum dilution 0.2 cc. used.
* These two paratyphoid types of fowl origin were selected because of their
very close agglutinative relationship one to the other and to B. typhosus. The
extreme cross precipitation with these types in this series is no indication of loss
of specificity. Compare with similar agglutinative results of Smith and Ten
Broeck. (Journ. Med. Res., 1915, xxxi, 549.)
7 Overnight.
8 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE
in the final antigens and interfere with the reaction. Further
work is in progress to determine whether we can eliminate the
inhibiting factors and to determine their character.
The application of this method in preparing antigens from
sputa was attempted with the hope of evolving a rapid method
of determining the type of pneumococcus present. Modi-
fications have been found necessary and although the method at
present used is far from satisfactory and has only given positive
results in a relatively small percentage of the specimens examined,
it is given in the hope that it will serve as a basis to which further
improvements can be added. We have been handicapped by
lack of satisfactory material and further work will not be possible
till the next pneumonia season.”
The success of the method depends on several factors. One
difficulty is that not infrequently the sputum submitted is saliv-
ary or pharyngeal and not a specimen raised by coughing. Natur-
ally a positive result can only be obtained if the sputum is com-
paratively rich in pneumococci. The technical difficulties are
first, the necessity of removing the albuminous and mucous
material which otherwise passes over to the final extract, making
it thick or gelatinous; secondly, the fact that the alcohol will
not give a sufficiently prompt and complete precipitation where
little antigen is present. The method has been modified in the
attempt to overcome the first difficulty mentioned.
To the sputum add sufficient antiformin to give a concentra-
tion of from 3 to 5 per cent. If the sputum is thin, add the
concentrated reagent; if it is very thick, dilute by adding one-half
its volume of the appropriately diluted antiformin; that is, no
more dilution should be made than is necessary for final digestion.
The mixture is then heated to 100°C. and shaken till it becomes
fluid. Add a few drops of phenolphthalein to the hot antiformin-
sputum and neutralize with acid. Then add a sufficient excess
of acid to cause coagulation of the coagulable substances in the
fluid, centrifugalize and collect the supernatant fluid; neutralize
and add several volumes of alcohol, making a sufficient mixture
to fill a 15 ee. test tube. If a precipitate does not separate, let
2 Three successive Type I sputa just received have been positive.
PRODUCTION OF PRECIPITIN ANTIGEN 8)
the tubes stand for a while; then if a precipitate is formed, cen-
trifugalize and decant the supernatant fluid; add 1 cc. of saline
and heat to extract the sediment and centrifuge to clear. (If
too much saline is added the extract may be too dilute and beyond
the reaction stage.) The supernatant fluid is then added to the
tubes containing the ‘‘type’’ sera and the tubes are incubated
in the water bath. Each tube contains 0.2 ec. of serum and 0.2 ee.
of antigenic fluid The rapidity of the reaction varies with the
antigenic content of this fluid With a fluid that is rich in anti-
gen, there may be seen an immediate clouding followed by a
prompt separation of the precipitate, if serum and antigen are
mixed in the tube, or by a heavy ring of precipitate if the antigen
is carefully “layered” on the serum. With a fluid less rich in
antigen, the reaction is slower; with little antigen, it may be
absent. Should the finished antigen be thick an attempt may be
made to use it after dilution, but this is usually unsuccessful.
As is evident the method as far as evolved is not wholly satis-
factory, but even in its present state, sputa rich in antigen can
be examined and the type diagnosed in from one-half to one hour.
Any increase in the number of positive reactions and in the
rapidity of the reaction will depend on overcoming the technical
difficulties already mentioned. Thus far no false reactions
have been encountered unless the incubation was unduly pro-
longed. These may be found to occur occasionally because of
the presence of much extraneous material in the finished antigen.
One fact has impressed us very strongly in carrying out the
above work; viz., the very marked difference between bacterial
and serum precipitation in regard to the dilution limits of the
antigen. A serum antigen—as is well known—may be diluted
as much as 1 to 50,000 times and still give a precipitate with a
highly potent homologous immune serum. A bacterial antigen
must be much more concentrated. The alcoholic precipitate
as described when dissolved in twenty times its volume of saline
cannot be diluted more than eighty or one hundred times at
most. It may be that this is due to the fact that only a small
fraction of the bacterial protein is precipitable. Another differ-
ence is the extraordinary resistance of the bacterial antigen to
10 CHARLES KRUMWIEDE, JR., AND W. CAREY NOBLE
the action of heat and of chemical agents. Acidification and -
boiling are without effect. In fact we have cleared antigens by
adding abumin and coagulating it by heat and by acidification
and neutralizing the filtrate.
SUMMARY
A simple method of preparing a precipitin antigen from bac-
teria is presented, which allows of the preparation of such anti-
gens with a rapidity and in concentration hitherto impossible.
An attempt to modify the method to extract antigen from pneu-
monic sputum to determine rapidly the type of pneumococcus
present has been partially successful. It is presented as a
possible basis for further improvements.
ON VON DUNGERN’S INDIGO TEST FOR SYPHILIS
B. FUJIMOTO
From the Forensic-Medical Institute of the Imperial University at Tokio
Received for publication, October 10, 1917
Von Dungern’s indigo test for syphilis, which has been de-
scribed in the issue of the Miinchener Medizinische Wochen-
schrift for September, 1915 (no. 36), was recently examined by
Edward P. Flood (1) and it was proved by him that its results
are not constant and therefore not practically applicable to the
diagnosis of syphilis.
A further study of this reaction was undertaken in our labo-
ratory to determine more precisely its value. The reagents,
which were used in our test, were prepared as follows. Though
Flood stated that the preparation of the reagents was not easy
and that he was obliged to introduce a modification of von
Dungern’s procedure in order to get a clear solution of the in-
digo, I was able to prepare them easily in the same manner as
did von Dungern, namely:
One gram of indigo was carefully triturated and 4 cc. of con-
centrated sulphuric acid were added to it. After the mixture
had stood for forty-eight hours (the indigo being then com-
pletely dissolved), distilled water was added up to 100 cc. A
gray sediment, which was caused by unpurity of the material,
was thrown down. The solution thus obtained was finally clear
and indigo blue in color.
The Fehling solution no. 2 was prepared as usual; 173 grams
of Rochelle salt were dissolved in 200 ce. of hot distilled water
and 73 cc. of sodium hydroxide (specific weight 1.5) were added
to this. Distilled water was finally added up to 500 ce.
To 1.5 ce. of the indigo-sulphurie acid there were added 10ce.
of distilled water and to 4 cc. of this solution there was added
1 cc. of Fehling no. 2. The final solution was greenish yellow
11
12 B. FUJIMOTO
and free from any precipitate. This reagent was always used
soon after the final mixture was prepared.
At first, the test was carried out exactly as described by von
Dungern in order to determine whether this test is practically
applicable or not. The sera to be tested were obtained from
the dermatological clinic and from the medical clinic of our
university. The results of these tests with syphilitic and non-
syphilitic sera are shown in table 1. In our tests they were con-
trolled with the Wassermann reaction.
TABLE 1
Dungern’s indigo test
SEMI- NO
7 9 : c -
REAGENT, 0.2 cc. INACTIVATED SERUM, 0.3 cc. EGAGULATION)| 7 iceant on liconaaniaras,
Wassermann ( +) sera..................+--. 1 0 22
Wassermann ( — ) sera.............-...-+-- 1 2
|
Almost all of the sera did not coagulate in our tests.
Now comparing these results of the indigo test with the normal
and the syphilitic sera, we see that the test is not specific, and
that there is no difference between them at all. Such a non-
specific reaction is evidently unavailable for the clinical diagnosis
of syphilis.
In order to determine at what point coagulation of sera occurs,
the reagent was used in various amounts (0.2, 0.1, 0.05 ce.).
The results are as follows.
TABLE 2
Coagulation point of sera, to which the reagent was added in various amounts
SEMI- NO
ACTIVATED SERUM, 0.3 cc REAGENT | COAGULATION | ¢o aGULATION | COAGULATION
0.2 1 0 Peps
Wassermann (+) sera......... 0.1 20 3 0
0.05 23 0 0
0.2 1 2 ol
Wassermann (—) sera......... { 0.1 29 5 0
0.05 34 0 0
VON DUNGERN’S INDIGO TEST FOR SYPHILIS 13
Almost all sera both Wassermann positive and negative coagu-
lated with 0.1 and 0.05 ce. of this reagent. But with 0.2 ee.
most of them do not coagulate; that is to say, their heat coagu-
lation is prevented by 0.2 cc. of the reagent almost in all cases.
When 06.3 ce. of the sera was mixed with 0.2 ce. of distilled
water or of physiological salt solution without any of the re-
agent the heat coagulation occurred soon after fifteen or twenty
seconds.
In order to solve the question why the heat coagulation of
the sera is prevented by the reagent, I have proceeded to the
following tests.
The reagent contains indigo, sodium sulphate, sodium and
potassium tartrate and sodium hydroxide. Which of these sub-
stances plays a part in preventing the heat coagulation of sera;
is it the indigo or the salts or the alkali?
I. INDIGO
The indigo sulphuric acid was neutralized with barium hy-
droxide, and we obtained thus a neutral solution of indigo free
from sulphuric acid. This neutral solution was concentrated
until it possessed the same depth of color as the original indigo
sulphurie acid.
Indigo solution (0.2 ec.) was added to the syphilitic and non-
syphilitic sera in order to see whether the indigo alone has the
power to prevent the heat coagulation of sera. The results are
shown in table 3.
TABLE 3
Tests with pure indigo solution free from acid
L g
. E a z 9 : b SEMI- NO
PURE INDIGO SOLUTION (FREE FROM ACID), 0.2 cc. COAGULATION COAGULATION | COAGULATION
Wassermann (++) sera0.3 cc.:............:. 6 0 0
Wassermann (—) sera 0.3 ce................. 4 0 0
Coagulation occurred always rapidly, in fifteen or twenty sec-
onds, as if only distilled water had been added, which proved
that indigo does not play any part in preventing the heat coagu-
lation of sera.
14 B. FUJIMOTO
II. COAGULATION TESTS WITH SALTS
Comparing the very small quantity of neutral sodium sul-
phate with that of the Rochelle salt in the reagent, we can neg-
lect the former, because it is neutral salt and does not prevent
the heat coagulation; on the contrary, it can even accelerate the
coagulation in a small degree. :
It was proved in test tubes that the Rochelle salt can acceler-
ate coagulation even in the presence of the concentrated sodium
hydroxide. Thus it is evident that neither indigo nor salts pre-
vent the heat coagulation of sera.
Ill. TESTS WITH ALKALI
Finally it remains to test the action of sodium hydroxide.
The alkalinity of the reagent was titrated. Flood states that
1 ce. of it requires 1.21 ce. of acid for its neutralization, but I
have found not only by titrating but also by mathematical cal-
culation from the molecular weights of the ingredients of the
reagent that 1 cc. of it requires 3.3 cc. to 3.4 cc. of 4 acid for its
neutralization.
Acidity of indigo sulphuric acid in the reagent:
at N x 4/100 x 1.5°/ (10 + 1.5) x 4/441) =—0.154 Ni
Alkalinity of Fehling no. 2 in the reagent:
(sodium hydroxide specific weight 1.5 = 17 N)
17 N x 73 / 500 x 1/ (441) = 0.496 N
Therefore, the alkalinity of the reagent, must be
0.496 N (alkalinity) — 0.154 N (acidity) = 0.342 N (alkalinity)
To determine at just what point of alkalinity the sera coagu-
late, several Fehling solutions of different alkalinity were so pre-
pared as its alkalinity in the test tubes may be each 0.12 N,
0.10 N, 0.08 N, 0.06 N, 0.04 N, 0.02 N. These solutions of
different alkalinity were added (a) to the indigo and (b) to the
reagent without indigo (viz., to 4 ec. of concentrated sulphuric
/
VON DUNGERN’S INDIGO TEST FOR SYPHILIS 15
acid there were added 96 ce. of distilled water). The results are
shown in table 4.
TABLE 4
Coagulation tests with solutions of different alkalinities
ALKALINITY N
SERA, 0.3 cc. REAGENT, 0.2 cc.
0.12 0.10 0.08 0.06 0.04 0.02
Syphilitic sera
me kart Mommas, |e] ft
ee ee eee
ee ee | ee | Oe
Non syphilitic sera
Sed 2) (eh eigen aa Esa Mtg Mi
eT ae meee er) (ein oa (mn
ee eee els
(—) = non coagulation. (+) = coagulation. (a) Reagent with indigo, (b)
Reagent without indigo.
As we see in the above table, each serum has its coagulation
point in its 60 per cent concentration (serum 0.3 cc., reagent
0.2 cc.), and we can not find any difference, as to this coagulation
point, between, syphilitic and non-syphilitic sera.
TABLE 5
ALKALINITY IN TEST TUBES
SERUM, 0.3 cc.
0.12 N 0.10 N 0.08 N 0.06 N
{ No. 1 = = umes fe
Wassermann (+) sera INO25.,. + =f Bie +h
Nowoes. = == ao +
{ No. 4 = 2 = si
Wassermann (—) sera Nosot aan. er _ — ze 25
| No. 6 _ at se ae
(+) = coagulation. (—) = non coagulation.
16 B. FUJIMOTO
Now suspecting that not only the alkalinity but also variation
in the concentration of the serum might serve to cause varia-
tion in the heat coagulation, we have tested the sera in its 75
per cent concentration instead of 60 per cent, viz., to 0.3 ce. of
serum was added 0.1 cc. of the reagent. The alkalinity of the
reagent used here was so prepared that the alkalinity in the test
tubes was 0.12 N, 0.10 N, 0.08 N, 0.06 N respectively. The
results are shown in table 5.
Comparing these results with those shown in table 4, we see
that the more diluted the sera, the more easily their heat coagu-
lation can be prevented by alkali.
SUMMARY
1. Von Dungern’s indigo test for syphilis is practically un-
available for the serum diagnosis of syphilis, because we can not
find any difference as to this test between syphilitic and non-
syphilitic sera.
2. Among the substances that the reagent contains neither
indigo nor salts play any part in the inhibition of the coagula-
tion, as von Dungern formerly thought.
3. The coagulation of sera in this reaction is prevented only
by the alkali, as Flood proved.
4. The degree of alkalinity that can prevent the coagulation
of serum, varies according to the individual, not according to
the presence or absence of syphilis.
5. On the other hand, the alkalinity that prevents the coagu-
lation of sera varies according to the concentration of the sera;
that is, the more diluted the serum, the more easily may its heat
coagulation be prevented by alkali.
I wish to express my thanks to Prof. Dr. K. Katayama and
Prof. Dr. S. Mita for suggesting the subject and for their kind
advice.
REFERENCE
(1) FLoop: Journal of Immunology, 1916, 2, 69.
EXTRACTS OF ANTIBODIES OBTAINED FROM
SPECIFIC PRECIPITATES OF TYPHOID-
ANTITYPHOID SERUM COMPLEX
ISRAEL WEINSTEIN
From the Laboratory of Bacteriology and Hygiene, New York University,
New York City
Received for publication October 11, 1917
An investigation was undertaken in this laboratory with a view
to precipitating the antibodies from antityphoid sera by means of
a specific antigen and subsequently dissociating this antigen-
antibody complex, obtaining the antibodies in a solution as free
from foreign protein as possible.
As is well known, a number of antityphoid sera have been pre-
pared and occasionally physicians have reported good results
with them, but their use has never been general. In order to
be successful with a bactericidal serum it seems necessary to
inject large amounts. For example, Cole (1), in the treatment
of pneumonia, by using very large quantities of antipneumococcus
serum (from 190 to 700 ec.) has had success where others have
failed. Owing to the danger of ‘serum sickness’ following the
injection of large amounts of horse serum, physicians have hesi-
tated to treat their patients with it. Thus, the importance of
eliminating as far as possible all extraneous protein and at the
same time the harmful effects that it produces, can readily be
seen. It was with this aim in view that the experiments described
below were begun. |
Within the last fifteen years great interest has been shown by in-
vestigators in regard to the fate of the antibodies in the precipitin
reaction. Gay (2) was the first to demonstrate that complement was
fixed by precipitates. He later (3) amplified his earlier studies and
concluded that ‘“alexin fixation by a mixture of serum and antiserum
is produced by an antigen-antibody complex distinct from precipitin-
ily
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 1
18 ISRAEL WEINSTEIN
ogen-precipitin but usually brought down by the precipitate in its -
formation in such a way as to give the appearance that fixation is
produced by the precipitate itself.” Muir and Martin (4), Toyosumi
(5) and Zinsser (6) likewise showed that precipitates had complement-
fixing properties. In the case of agglutinins, Von Eisler and Tsuru
(7) showed that normal hemagglutinins were removed from the serum
through precipitation while Landsteiner and Prasek (8) proved that
both hemagglutinins and bacterial agglutinins were brought down with
the precipitates. Gay and Chickering (9) found that the protective
bodies in an antipneumococcus serum were carried down with the
precipitate.
Other investigators have shown that it is possible to break up the
antigen-antibody complexes. Matsui (10) succeeded in extracting
bactericidal bodies from cholera vibrios after the former had united
with the latter. Muir (11) removed the hemolytic amboceptor from
red blood corpuscles, while Landsteiner (12) separated agglutinins
from their antigens. Chickering (13) was able to dissociate the pre-
cipitates formed by the union of pneumococcus antigen and antiserum
and get the antibodies into solution. He found that his extracts con-
tained agglutinins and precipitins besides protecting susceptible animals
as efficiently as the original serum.
EXPERIMENTS AND RESULTS
Through the kindness of Professor Park an antityphoid serum
having high agglutinating and bactericidal titers was obtained.
The serum was taken from a horse that had been immunized
with increasing intravenous injections of dead and then living
typhoid bacilli at definite intervals for a number of months.
In order to determine whether antibodies beside the precipitins
were carried down with the precipitate, the original serum, the
supernatant serum and the extract obtained from the precipitate
were, in each experiment, tested for their agglutinin content.
This antibody was chosen in preference to the others because of
the ease with which the tests could be made. However, the
extract that appeared to be most promising was also tested for
bactericidal, complement-fixing and protective bodies. The
macroscopic agglutinin test was used. 1 cc. of a twenty-four
hour typhoid broth culture was added to an equal amount of the
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 19
serum or extract. The tubes were incubated at 37.5°C. for
one hour and then left over-night in the ice-box. Readings were
made the following morning. In the preliminary experiments
extracts were prepared according to the method that Chickering
had found to be best. The precipitates, which had been washed
three times with normal salt solution were emulsified in normal
saline to which had been added a few drops of a 1 per cent sodium
carbonate solution, and were then heated for one hour at 42°C.
with occasional shaking.
The method which Chickering used in preparing his antigen,
namely, shaking in saline pneumococci that had been previously
killed with acetone and dried in vacuo, was found to be entirely
unsatisfactory in the case of the typhoid bacillus. This was
undoubtedly due to the greater resistance of the typhoid bacilli
to autolysis. In attempting to get a suitable antigen, the fol-
lowing procedures were used: (1) saline suspension of bacteria
killed by heating at 54°C. for 1 hour; (2) extracts of bacteria
killed with heat or acetone, dried in vacuo over sulphuric acid
and shaken in saline or distilled water for lengths of time varying
from ten minutes to twenty hours; (3) extracts of bacteria dried,
ground in a mortar and then shaken in saline, (4) extracts of
bacteria shaken in distilled water from seven to sixteen hours
and then allowed to autolyze at temperatures ranging from 37.5°C.
to 54°C. for an additional sixteen hours ; (6) extracts of bacteria
allowed to remain at 37.5°C. in saline or distilled water for ten
days. Whenever distilled water was used the extract was
rendered isotonic before adding it to the serum. None of the
above antigens gave good results. Fair results were obtained
with an extract of bacteria that had been killed with acetone,
dried and ground eight hours a day for six days. This antigen
removed approximately one-half of the agglutinins of the serum.
The best results were obtained with an extract made by digesting
the bacilli in a 2.5 per cent solution of antiformin. The bacterial
growth on an agar slant in a quart Blake bottle was suspended
in 3 cc. distilled water and was added to an equivalent amount
of a 5 per cent antiformin solution. This mixture was allowed
to remain at room temperature for one hour, after which it was
20 ISRAEL WEINSTEIN
centrifugalized at high speed for three-quarters of an hour. The
supernatant liquid was rendered neutral by the addition of nor-
mal hydrochloric acid. Equal amounts of serum and antigen
solution were shaken together and put in the incubator. At the
end of two hours a dense precipitate was observed. After re-
maining in the ice-box overnight the mixture was centrifugalized.
The supernatant liquid was withdrawn, the precipitate was
washed three times and then was extracted according to the
method described above in alkaline saline solution equivalent
to the amount of the serum used. The serum, supernatant and
extract were then tested for their agglutinin content. The
titer of the serum was 1: 20,000; that of the supernatant was
1: 4000, while that of the extract was 1: 1000. It is thus seen
that approximately three-fourths of the agglutinins were removed
from the serum, while one-sixteenth of that amount was recovered
in the extract.
By treating various dilutions of the antigen with normal horse
serum as well as with immune serum, it was seen that the antigen
acted specifically.
At the suggestion of Prof. Charles Krumwiede the antigen
was purified in the following manner. First, the antigen solution
was precipitated in three volumes of absolute alcohol. After
centrifugalization, the precipitate was dried and then taken up
in an amount of normal saline solution equivalent to the original
volume of the antigen. After being shaken at 37°C. for a few
minutes the precipitate went into solution. This purified
antigen was found to be just as potent in carrying down agglutin-
ins from the immune serum as the untreated antigen. It had
the advantage of being rid of the high salt content present in the
untreated antigen due to the large amount of alkali in the anti-
formin and its subsequent neutralization with hydrochloric acid.
DETERMINATION OF THE AMOUNT OF ANTIGEN NECESSARY TO BRING
ABOUT THE MAXIMUM PRECIPITATION OF ANTIBODIES
In order to determine the optimum amount of antigen for
precipitating the antibodies from the serum, various amounts
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 21
of antigen solution and serum were mixed together. The amount
of serum remained constant, while the amount of antigen varied
from eight times the volume of the serum to one-eighth of its
volume. Various amounts of diluted antigen were also used in
order to determine whether the process took place better in
weak than in concentrated solution. The tubes containing
the mixtures were incubated for two hours and then left in the
ice-box overnight. Abundant. precipitates appeared in all of
the tubes with the exception of those that contained 0.5 ce.
and 0.25 ce. of antigen. After centrifugation, the supernatant
liquids were pipetted off and titrated for their agglutinin content.
In the last column of table 1 the actual titers of the supernatants
are given, allowances having been made for the dilution of the
serum by the antigen.
TABLE 1
Comparative agglutinating power of whole serum and sera treated with various
quantities of antigen. Titer of whole serum equals 1: 20,000
ACTUAL DILUTIONS
Sie | cenilion| eran OF Pr SE Sa SNE |
SERUM USED GEN USED s 8 s s S Ss 8 S 3 3 S S S SOLUTION
cc. cc.
1 8.0 +/+/+/+}+|—|/—|—|—|-—|-/—|-| 1: 8000
2 6.0 +I+]+/+]+]— —|—|—|—|—| 1: 6400
3 4.0 +/+]+)/+|+]+]—|—|—|—|-—|—|—] 1: 6000
4 2.0 +/+/+/+|+]+]+/+!-!—|/—I-I-| 1: 8000
5 1.0 2 (1:2) |-+/-+}+/+/+/+/+/+]—!-!-/-l-| 1: 8000
6 0.5 2 (1:4) +/+ /+|+}-+/+/+]+]+/+]—|-|-] 1: 16,000
7 0.25 2 (1:8) | +/+|+)+/+/+]+]+]+]+/+/-|—| 1: 20,000
8 2.0 4 (1:2) |+/+/+/+/+/+]-/-/-]-|-]-]-] 1: 6000
9 2.0 16 (1:8) = |+)+/+/+/-|—|—/|-|-|-|-|-]-| 1: 9000
10 2.0 20 (1:10) |+)+|+/—|—|—|-|—|-|—|-|-]— 1: 8800
The results of this experiment show that little is to be gained
by the use of large quantities of antigen. An amount of antigen
solution, prepared as described above, equal to that of the serum
seems to give as good results as can be obtained. If too little
antigen is used there is, as might be expected, little precipitation.
The diluted antigens seem to work a little better than the un-
diluted ones; e.g., 2 cc. of a 1: 2 dilution reduces the agglutinating
Ze ISRAEL WEINSTEIN
titer of the serum to the same degree that 2 cc. of undiluted
antigen does.
DETERMINATION OF THE BEST METHOD OF EXTRACTING THE
ANTIBODIES FROM THE PRECIPITATE
Effects of alkali
The effect of weakly alkaline and strongly alkaline solutions
in dissociating the precipitates was tested. Twelve tubes, each
containing 2 ec. of antigen and 2 cc. of serum were incubated
for two hours and then put in the ice-box overnight. The pre-
cipitates were washed. Those in the first seven tubes were
suspended in 2 ce. saline with small amounts of a 1 per cent
solution of sodium carbonate. The precipitates in the remain-
ing tubes were suspended in 2 cc. of a solution of sodium carbon-
ate or sodium hydroxide as indicated in table 2. All of the
TABLE 2
Comparative agglutinating power of extracts treated with various amounts of alkali.
Titer of whole serum equals 1: 20,000. Titer of supernatant serum equals 1: 8000
DILUTIONS
TUBE ALKALINE SOLUTION AMOUNT pe TIME Se S
= |8|8|2 /2/8|8
ce. 46 hour |
1 1.0 per cent Na,CO; 0 42 1 aelsaeeae i=
2 1.0 per cent NasCO; 0.01 42 | ee ee ee
3 1.0 per cent Na,CO; 0.02 42 1 Stas estea| lca at | al el
4 1.0 per cent Na,CO; 0.03 42 1 cote eitea| atte al el
5 1.0 per cent NazCO; 0.04 42 1 +{+/+)4+)—|—)—
6 1.0 per cent Naz:CO; 0.05 42 | a a et feed =| |
7 1.0 per cent Nas,CO; 0.1 42 1 See ase) |
8 1.0 per cent Na,CO; 2.0 42 1 +/+}+/+]—|-—|/—
9 0.5 per cent NaOH 250) 42 1 +/+)+)+ 5 a
10 1.0 per cent NaOH [2228 42 1 +/+/+/+)/4+/—|/—
ib 2.5 percent NaOH | 2.0 42 1 j+/+/-|-—|-!-!-
12 5.0 per cent NaOH 2.0 42 | LN | ase fed Ge rd a >
tubes were put in a 42°C. water-bath and shaken gently for one
hour. The precipitate in tube 12 (5 per cent NaOH) went into
solution at once; the precipitate in tube 11 (2.5 per cent
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 23
NaOH) and half of that in tube 10 (1 per cent NaOH) were
dissolved at the end of an hour.
The small amounts of alkali in the extracts evidently exert no
beneficial effect, for the control tube (table 2) shows as high a
titer as the others. Larger amounts of alkali are helpful. The
results also clearly indicate that a strongly alkaline solution de-
stroys the antibodies.
Effect of time and temperature
Ten tubes, each containing 2 ce. of antigen and 2 ce. of serum,
were incubated for two hours and then put in the ice-box over-
TABLE 3
Comparative agglutinating power of extracts of precipitates treated with weakly
alkaline solutions for various lengths of time at different temperatures. Titer
of whole serum equals 1: 20,000. Titer of supernatant serum equals 1: 6000
DILUTIONS
a = = = = <— TEMPE RA-
TUBE ALKALINE SOLUTION AMOUNT eee TIME a : 3 3 3 s
ce. OF hours
1 1 per cent Na.CO; 0.05 42 1 oe ee 2S ee ee
2 1 per cent NasCO; 0.05 42 2 eee |e eh
3 1 per cent Na.CO; 0.05 | 49 3 2) tS a a P|
4 | 1 per cent Na,CO; 0.05 42 4 0 |+)+/+/+})-}—
5 | 1 per cent Na,CO; 0.05 ADO Esp ay feta eee eres | a
6 | 1 per cent Na.CO; 0.05 ADs Vt C6) sic Rl al el
7 | 1 per cent Na.CO; | 0.05 54 1 aE | pe
: a per cent Na.CO; 0.05 54 2 Se a ee
9 1 per cent Na.CoO., 0.05 OM 24 I+/+)/+ ai alr:
10 | 1 per cent Na,CO. | §60.05 20 240 |+/+)—/-|-|-
night. The precipitates were washed three times with normal
saline solution and then each was suspended in 2 ce. of saline
with 0.5 ce. of a 1 per cent sodium carbonate solution. The
tubes were left at various temperatures for varying lengths of
time.
It can readily be seen from table 3 that practically nothing
is gained by incubating the mixture for more than one hour at
42°C. At the end of that time a certain amount of dissociation
of the precipitate has taken place. A longer time or higher tem-
perature evidently adds nothing.
24 ISRAEL WEINSTEIN
Effect of washing the precipitate
The next point considered was whether any of the agglutinins
were lost in washing the precipitate. Accordingly, the precipi-
tates which had formed in four tubes, through the mixture of
2 cc. of serum and 2 cc. of antigen in each, were treated as fol-
lows: No. 1 was suspended in 2 ce. of a 1 per cent sodium carbon-
ate solution: Nos. 2, 3 and 4 were washed once, twice and three
times respectively (2 cc. of normal saline solution being used
TABLE 4
Comparative agglutinating power of extracts of washed and unwashed precipitates.
Titer of whole serum equals 1: 20,000. Titer of supernatant serum equals 1: 4000
DILUTIONS
TUBE PREPARATION OF SOLUTIION
2/8/8/8/=|2/8/8/8
1 | Precipitate treated with1 per cent sodium carbo-
nate solution +/+)+/+/+/+]+]-|-—
2 | Precipitate, washed once, treated with 1 per cent
sodium carbonate solution fff jj | — | — | —
3 | First washing +)+/+/+)4+/—|—-j-|—
4 | Precipitate, washed twice, treated with 1 per cent
sodium carbonate solution aac pl bus mailed
5 | Second washing +)/+/—|—|—|-—|-|-|—
6 | Precipitate, washed three times, treated with
1 per cent sodium carbonate solution +/+)+/+)+)—|}—|—|—
7 | Third washing +)/—)}—|-—}|—|-—|-—|-|-—
for each washing) and each was then suspended in 2 cc. of 1
per cent sodium carbonate. The extracts and the washings
were tested for their agglutinin content.
It will be seen from table 4 that there is a reduction in the
agglutinin content of an extract of a washed precipitate. We
believe this to be due to the fact that a small amount of serum,
which remains on the unwashed precipitate, is carried over into
the extract. The washings remove all of this serum. In other
experiments it was found that when the precipitate was tightly
packed in the centrifuge tube and the supernatant serum care-
fully removed with a capillary pipette, the titer of the first wash-
ing was not above 1:800. It may further be seen from table
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 25
_4 that the titers of the second and third washings are in no way
comparable to that of the first. We may conclude that the
amount of agglutinins lost through the dissociation of the precipi-
tate by washings, when the process is carried on within a few
minutes and at room temperature, is negligible.
Tests for bactericidal bodies
An extract was prepared in the usual way by treating the
precipitate with an amount of slightly alkaline saline solution
equal to the volume of the original serum. The whole serum,
the supernatant and the extract were tested in vitro for their
bactericidal power.
Method. The three solutions were inactivated by heating at 56°C. for
one hour and were then diluted with normal saline. One cubic centi-
meter of each dilution was mixed with 1 ce. of a 1: 10 dilution of com-
plement and 0.1 cc. of a 1: 2000 dilution of a twenty-four-hour typhoid
broth culture. The complement was obtained from the pooled sera
of three normal guinea-pigs. The tubes were incubated at 37°C. for
two hours, at the end of which time the contents of each tube were
plated in agar. The plates were incubated at 37°C. for twenty-four
hours and then the colonies were counted. Controls were set up as
follows: (1) culture with saline solution, in order to determine the
number of bacilli used in each test; (2) culture with complement, in
order to see whether the complement of itself had any bactericidal
action; and (3) culture with each of the inactivated solutions, in order
to see whether all of their native complement had been destroyed.
By consulting table 5 it is seen that the whole serum shows
a definite bactericidal action in a dilution of 1:40,000, the
supernatant in a dilution of 1: 30,000 and the extract in a dilu-
tion of 1: 20,000. We may roughly say, therefore, that about
half of the bactericidal bodies present in the whole serum are
recoverable in the extract.
. Tests for complement-fixing bodies
The whole serum, the supernatant and the extract were next
tested for their complement-fixing power. The antigen pre-
NUMBER OF COL-
ONIES
3
26 ISRAEL WEINSTEIN
TABLE 5
Comparative bactericidal power of whole serum, supernatant and extract
| gs 7
| BR Sa
Bo ous
a B SOLUTION beageieed D Z
oR ase
ae ae
& je}
ce ce.
@ontrols-eosae ee
0.1
(cout fe
Complement controls. . < 0.1 1
Serum control......:.... | “Ot Whole serum 1: 10,000
= Ord Whole serum 1: 10,000 1
0.1 Whole serum 1: 20,000 1
0.1 Whole serum 1: 30,000 1
0.1 Whole serum 1: 40,000 1
| O20 Whole serum 1: 60,000 1
|} On Whole serum 1: 80,000 1
0.1 Whole serum 1: 100,000 1
Supernatant serum con-
CLO Sere aye eee | 0.1 Supernatant serum | 1: 5,000
| 0.1 Supernatant serum | 1:5,000 | 1
0.1 | Supernatant serum | 1: 10,000 1
0.1 | Supernatantserum | 1: 20,000 1
0.1 Supernatant serum | 1: 30,000 1
| 0.1 | Supernatant serum | 1: 40,000 1
| 0.1) Supernatant serum | 1: 60,000 1
| 0.1) Supernatant serum | 1: 80,000 1
| 0.1 | Supernatant serum | 1: 100,000 Reig
Extract control........ 0.1 | Extract 1 100
0.1 Extract 1: 100 1
| 0.1 | Extract 17; E000) |. wad
0.1} Extract 1: 5,000 il
0.1 Extract 1: 10,000 1
0.1 | Extract 1: 20,000 1
| On! Extract 1: 30,000 1
| 0.1 | Extract 1: 40,000 1
0.1 | Extract 1: 60,000 1
0.1] Extract 1: 80,000 1
0.1 Extract 1: 100,000 1
LO oe ee ge ee eee ae
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 27
pared with antiformin proved to be anticomplementary and
therefore could not be used in making the tests. A suspension
of ground bacilli was used in its place.
Method. In preparing the antigen,! typhoid bacilli that had been
erown on salt-free veal agar were suspended in normal saline solution
and left in flowing steam in the Arnold sterilizer for one and one-half
hours. The mixture was then centrifugalized and the supernatant
liquid discarded. The bacteria were washed once with 5 volumes of
absolute alcohol, twice with 5 volumes of ether and then allowed to
dry at room temperature for forty-eight hours. The dried bacilli were
ground in a mortar for one hour. 5 cc. of this bacterial powder were
taken up in 55 cc. of normal saline solution. This suspension was used
as antigen. 0.1 cc. of 1:50 dilution of the antigen, which constituted
2 units, was used in each test. For complement, the sera from ten
normal guinea-pigs were pooled and 0.1 cc. of a 1:10 dilution of this
serum was used in each test. This represented 1# units of complement.
Dilutions of the serum, supernatant and extract were made and de-
creasing doses of each dilution from 0.1 cc. to 0.01 cc. were put into
tubes. To these, 0.1 cc. of antigen and the same amount of comple-
ment were added. The volumes of the mixtures were equalized by the
addition of normal saline solution. In each rack there was one tube
which contained 0.2 cc. of antigen and 0.1 cc. of complement, with
no serum. This constituted the antigen control. There were also two
other tubes which contained 0.04 and 0.02 cc. serum respectively and
0.1 cc. of complement with no antigen. These served as the serum
controls. After incubating the tubes for thirty minutes in a 37°C.
water-bath, the hemolytic system was added. This consisted of 0.1 ce.
of a 5 per cent suspension of washed sheep corpuscles plus two units of
hemolysin. The unit of hemolysin was found to be 0.04 ce. of a 1: 6000
dilution. “The tubes were then put in the water-bath for an additional
ten minutes, at the end of which time the results were read.
In titrating the various solutions, the whole serum and supernatant
were used in dilutions of 1: 10 and 1: 100; the extract, undiluted and in
dilutions of 1:10 and 1:100. In the case of the supernatant serum,
the serum controls were found to be anticomplementary to a certain
extent, due undoubtedly to the fact that it contained some of the
1 This antigen was prepared in accordance with directions given by Miss M.A.
Wilson of the Research Laboratories of the New York City Health Department.
28 ISRAEL WEINSTEIN
“antiformin” antigen in it. This must be taken into consideration
in interpreting the results.
It would seem from table 6 that the extract had about one-
tenth the complement-fixing power of the serum.
Animal experimentation
Historical note. No true typhoid infection can be produced in labo-
ratory animals, with the possible exception of the anthropoid apes.
The inoculation of small amounts of stock culture has little, if any,
effect. The use of Jarge quantities, especially of a culture whose viru-
TABLE 6
Comparative complement-fixing power of whole serum, supernatant serum and extract
DILU-
SOLUTION TION CONTROLS RESULTS OF TESTS
1:10 | Hemolysis 0.001 cc. Complete fixa-
Whole serum. . tion
1: 100| Hemolysis 0.0008 cc. Weak fixation
1:10 | 0.04 Strong fixation 0.005 ec. Strong fixation
0.02 Weak fixation
Supernatant
Se en ear 1: 100} 0.04) 0.002 cc. Weak fixation
Trace
0.02 f :
Un- | Hemolysis 0.01 ce. Complete fixation
dilut-
Extract: = 25:2: ed
1:10 | Hemolysis 0.002 cc. Weak fixation
1: 100| Hemolysis 0.001 ec. Weak fixation
lence has been increased by successive passages through animals, is
followed by death. That death is due to intoxication was proved by
Sirotinin (14), who showed that the animals succumbed to doses of
dead as well as living bacilli. Investigators have reported definite
lesions in animals that have died following an injection of typhoid
culture. There is usually a congestion of the abdominal organs, especi-
ally of the spleen, liver, kidney and intestinal lymph nodes. But
Beumer and Peiper (15) have shown that these lesions are not specific.
They were able to produce the same lesions through the injection of
non-pathogenic soil and water bacteria. Metchnikoff and Besredka
- Sil Wa. =
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 29
(16) have shown that a true typhoid infection can be produced in the
_ higher monkeys. In the case of fifteen chimpanzees and one gibbon,
that were treated with mixtures of pure typhoid culture and the fecal
material of typhoid patients, they had only one negative result. Chan-
temesse and Widal (17) increased the virulence of their culture by in-
jecting it into guinea-pigs simultaneously with large doses of killed
streptococci. They passed the culture from animal to animal, using
decreasing doses of streptococci, until the virus acquired fixed
characteristics.
Protective power of the extract
The object of the first experiment was to determine whether
the virulence of the stock typhoid culture could be increased by
following the method of Chantemesse and Widal (17), ice.,
by the simultaneous injection of typhoid and killed streptococ-
cus cultures. Accordingly three guinea-pigs, were inoculated as
follows:
No. 1. 4 cc. of typhoid culture subcutaneously and 8 cc. of
streptococcus vaccine intraperitoneally.
No. 2. 4 ec. of typhoid culture subcutaneously.
No. 3. 8 cc. of streptococcus vaccine intraperitoneally.
No. 1 was dead within less than eighteen hours. No: 2 died
after forty-eight hours. No. 3 survived. Cultures taken from
the peritoneum and heart of no. 1 were positive. 1 cc. of the
peritoneal exudate was mixed with 5 cc. of broth and incubated
for five hours. Guinea-pig 4 was given 3 cc. of this mixture
subcutaneously and at the same time 7 ce. of streptococcus
vaccine intraperitoneally. The animal died within fourteen
hours. Cultures from the heart and peritoneum were found to
be positive. The culture obtained from the heart was inocu-
lated into broth. Guinea-pig 5, which received 2 cc. of this
broth culture intraperitoneally, became very sick within six
hours after the injection. A general paralysis set in which was
soon followed by death. The bacilli were obtained in pure
cultures from the peritoneal exudate and heart blood. It was
this virulent culture which was used in the later experiments.
Eight guinea-pigs were each given intraperitoneally 2 cc. of
culture. Three of them received in addition 1, 2 and 3 ce.
30 ISRAEL WEINSTEIN
respectively of serum, while three others received equivalent
amounts of extract. The two control animals died within
twenty-four hours after the injection. 1 cc. of the extract, as
well as 1 ec. of the serum, protected against a fatal dose of
culture.
In the case of mice, the minimal lethal dose was found to be
0.2 cc. injected intraperitoneally. 0.2 cc. of the extract, as well
as 0.2 cc. of whole serum, was found to protect mice against the
minimal lethal dose of culture.
Kjeldahl determinations?
Determinations of the nitrogen content of the whole serum
and of the extract were made. The whole serum was found to
contain 1.064 grams of nitrogen per 100 cc. The content of the
extract varied from 0.018 to 0.028 grams per 100 cc. In other
words, the nitrogen content of the serum was reduced to from
3: to = Of its original amount.
DISCUSSION OF RESULTS
In attempting to obtain an antigen that would precipitate
antibodies from an antityphoid serum, it was soon found that -
methods that gave good results with micrococci were of no use
whatever with typhoid bacilli. Micrococeci autolyze rapidly in
distilled water and even in normal saline. Typhoid bacilli,
on the other hand, disintegrate very slowly. The best method
of preparing a typhoid antigen appeared to be the digestion of
the bacteria with 23} per cent antiformin and the subsequent
neutralization of the solution with hydrochloric acid. Such an
antigen not only produced an abundant precipitate when mixed
with antityphoid serum, but also carried down with the precipi-
tate antibodies other than the precipitins. In order to free the
antigen of its high salt content and of any free chlorine that
might have been present, it was precipitated with three volumes
* These determinations were kindly carried out by R. L. Kahn of the Monti-
fiore Home and Hospital Laboratory, New York City.
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 31
of absolute alcohol, the precipitate being subsequently redis-
solved in normal saline solution.
An equal amount of antigen solution and immune serum gave
the best results. Larger amounts of antigen did not bring down
greater quantities of antibodies from the serum. This is in
keeping with the results of Dean (18) who found that the maxi-
mum precipitate was obtained by combining equivalent pro-
portions of antigen and serum.
The mixture of antigen and serum produced the precipitation
of large quantities of agglutinins, bacteriolysins, complement-
fixing and protective bodies. Landsteiner and Prasek (8) found
that bacterial agglutinins were carried down with the precipitate.
Chickering (13) found the same thing in the case of the precipi-
tates formed from mixtures of pneumococcus antigen and serum.
That complement-fixing bodies are brought down with the pre-
cipitate has been shown by the work of Gay (3) and Zinsser
(6). The removal of protective bodies from an immune serum
was demonstrated by Gay and Chickering (9) and later by
Chickering alone.
It was possible to dissociate the antigen-antibody complex
to a certain degree and to obtain a portion of the antibodies in
solution (about 5 per cent of the agglutinins, 50 per cent of the
bacteriolysins and 10 per cent of the complement-fixing bodies
that were present in the original serum.) Chickering had a
similar experience. He found that while it was possible to
remove most of the antibodies with a single extraction, a residue
alwaysremained. This residue could not be removed by repeated
extractions.
Slightly alkaline solutions were found to be best for extraction.
Strongly alkaline solutions would dissolve the precipitate and
would also destroy the antibodies. A temperature of 42°C. was
found to be necessary. Room temperature was entirely un-
satisfactory. Muir (11) observed the same thing while attempt-
ing to dissociate hemolysin from red blood corpuscles.
Because a comparatively small proportion of the agglutinins
are recoverable, it does not necessarily mean that the same
thing is true of the other antibodies. Chickering, for example,
32 ISRAEL WEINSTEIN
found that his extracts contained practically all of the pro-
tective bodies that were present in the original serum but only a
small part of the agglutinins.
The work of a number of investigators shows that animal
experimentation in the case of typhoid fever is a rather uncertain
procedure, inasmuch as no true infection can be produced in
‘laboratory animals, with the possible exception of the anthropoid
apes. The death of animals inoculated with typhoid is most
probably due to intoxication. If large amounts of the culture
are given, especially if the virulence of the culture has been
previously raised by passage through animals, there is an invasion
of the blood stream and probably an increase in the number of
bacteria. The method of obtaining a virulent culture, described
by Chantemesse and Widal (17), namely, the simultaneous
inoculation of typhoid bacilli and streptococcus vaccine, was
found to be effective.
The enormous reduction in the protein content of the serum
that is possible by extracting the antibodies, compensates for
the antibodies that are lost. Such a solution as the extract,
because of its low protein content, could be injected into typhoid
fever patients without running the risk of having the serious
reactions on the part of the patient that often follow the injec-
tion of the whole serum.
CONCLUSIONS
1. An extract of typhoid bacilli, formed by digesting the
bacteria in 2.5 per cent antiformin and neutralizing the solution
with normal hydrochloric acid, produces a voluminous precipi-
tate when mixed with antityphoid serum.
2. An equal amount of such an antigen solution and serum
produces the maximum precipitate.
3. Such precipitates contain not only precipitins, but also
agglutinins, complement-fixing, bactericidal and protective bodies.
4. About 5 per cent of the agglutinins, 50 per cent of the bac-
tericidal bodies and 10 per cent of the complement-fixing bodies
present in the original serum can be extracted from the precipi-
tate in a slightly alkaline solution at 42°C.
a a
PRECIPITATES OF TYPHOID-ANTITYPHOID SERUM 33
5. The extract contains from 3 to 39 the amount. of nitrogen
that is present in the whole serum. The nitrogen content of the
extracts having been reduced on the average to ;, the amount
present in the whole serum, the agglutinins were concentrated
approximately twice, the bactericidal bodies twenty times?
and the complement-fixing bodies four times.
6. By passing stock cultures of typhoid bacilli through guinea-
pigs, whose resistance is lowered by the simultaneous injection of
streptococcus vaccine, it is possible so to increase the virulence
of the culture that small doses will kill mice and guinea-pigs in
a short time.
7. 1 ce. of extract will protect guinea-pigs against 2 cc. of
typhoid bouillon culture, a fatal dose. 0.2 cc. of extract will
protect mice against 0.2 ec. of culture, which is likewise a fatal
dose.
REFERENCES
(1) Coxtz, R.: Jour. Am. Med. Assn., 1913, 61, 663.
(2) Gay, F. P.: Centralbl. f. Bakteriol., 1905, 39, 603; Ann. de l’Inst. Pasteur,
1905, 19, 593.
(3) Gay, F. P.: Univ. Calif. Pub. Path., 1911, 2, 23.
(4) Muir, R. anp Martin, W. B. M.: Jour. Hyg., 1906, 6, 265.
(5) Torosum1, H.: Centralbl. f. Bakteriol., 1908-1909, 48, 325.
(6) Zinsser, H.: Jour. Exper. Med., 1912, 15, 529.
(7) Von Etstmr, M. anp Tsuru, J.: Ztschr. f. Immunitiitsforsch., 1910, 6, 608.
(8) LANDSTEINER, K. anp PraseEk, E.: Ztschr. f. Immunitiitsforsch., 1911,
10, 68.
(9) Gay, F. P. anp CuHIcKERING, H. T.: Jour. Exper. Med., 1915, 21, 389.
(10) Marsut, I.: Ztschr. f. Immunititsforsch., 1914-1915, 23, 233.
(11) Murr, R.: “‘Studies on Immunity,’’ London, 1909, p. 12.
(12) Lanpsrriner, K.: Miinchen. med. Wehnschr., 1902, 49, 1905.
(13) Curckerine, H. T.: Jour. Exper. Med.; 1915, 22, 248.
(14) Srrotinin, W.: Ztschr. f. Hyg., 1886, 1, 465.
(15) Brumer anp Perper: Centralbl. f. klin. Med., 1886, 7, 633.
(16) Mretcuntixorr, E. anp BrsrepDKA, A.: Ann. de l’Inst. Pasteur, 1911, 25, 193.
(17) CHANTEMESSE AND Wrpat: Ann. de l’Inst. Pasteur, 1892, 6, 755.
(18) Dean, H. R.: Proc. Roy. Soc. Med., Path. Sect., 1912, 5, part 3, p. 62.
* We do not believe that a very good method for testing bactericidal bodies
has yet been devised. The plate method, used in our experiments, is not suited
for accurate work.
THE JOURNAL OF IMMUNOLOGY, VOL. II, NO. 1
THE SPECIFICITY OF INTRACUTANEOUS
ABSORPTION
G. H. SMITH anv M. W. COOK
From the Research Laboratories of the H. K. Mulford Company,
Glenolden, Pennsylvania
Received for publication October 17, 1917
In previously reported experiments (1) we have shown that
the absorption of antigen from the cutaneous tissues of specifi-
cally immunized animals proceeds at a rate markedly in excess of
that occurring in normal animals. Horse serum, injected in-
tracutaneously into guinea-pigs immunized against horse serum,
disappeared from the site of injection much earlier than did horse
serum similarly injected into normal pigs. The question was
not determined, however, as to whether the heightened reactivity
of the tissues, attendant upon the immune state, pertained to
the specific antigen only or whether it indicated an activation
of a general mechanism of elimination extending to non-specific
antigens. The experiments here reported deal with this point.
In the present work guinea-pigs immunized to one antigen
were tested by the intracutaneous injection of the specific and
also of a heterologous antigen to demonstrate to what extent
absorption depends upon a specific factor.
The technic employed was much the same as that already
described. In several particulars, however, it was amplified so
that a better check on the results might be obtained. Guinea-
pigs immunized to normal horse serum were used with normal
guinea-pigs as controls. As antigens for intracutaneous in-
jection we employed both horse and goat serum, the former
derived from a horse immunized to B. dysenteriae (Y-Hiss), and
the latter serum secured from a goat immunized to the gonococcus.
With such antigens check titrations could be made. The
antidysentery horse serum could be detected by the precipitin
35
36 G. H. SMITH AND M. W. COOK f
test with rabbit antihorse serum and by the demonstration of the
dysentery agglutinins. Similarly, by the use of a rabbit anti-
goat serum and the agglutination test with the gonococcus the
presence of goat serum could be demonstrated. Not only were
these tests made upon the tissue extracts prepared from the
places of injection but titrations were also made by both tests
upon the sera of the injected animals to determine the amounts
of antigen absorbed into the circulation. Thus a measure was
secured of both the rate of disappearance of antigen from the
skin and the rate of appearance of antigen in the circulation.
The table, which follows, gives the results obtained by the
use of these titration procedures in several preliminary tests
upon immunized and normal guinea-pigs, all of which received
for the intracutaneous injections horse serum, the antigen speci-
fic for the immunized guinea-pigs. It is evident from the data
given that precipitin and agglutination tests are satisfactory
means for measuring the transference of antigen. In addition,
the data show the marked difference between normal animals
and immunized animals.
TABLE 1
The absorption of specific antigen
IMMUNIZED PIGS NORMAL PIGS
HOURS
Tissue | Serum Tissue | Serum
a. Precipitin tests
6 1: 25,600 1: 1,600 1: 25,600 ee LOO
12 1: 6,400 1: 6,400 1: 12,800 1: 1,600
24 1: 2,400 1: 9,600 1: 6,400 1: 3,200
48 1 400 1: 4,800 1: 2,400 1: 600
b. Agglutination tests
6 1: 12,800 1: 1,600 1: 12,800 Ls 150
12 1: 9,600 1: 12,800 1: 12,800 1: 2,400
24 1: 2,400 1: 12,800 1: 6,400 1: 8,000
48 1: 15200 1: 12,800 1: 1,200 1: 1,600
Examination of the figures given above shows that the agglu-
tination titrations run parallel with the precipitin tests. Not
only is this true of the tests conducted upon the tissue extracts,
a
&
SPECIFICITY OF INTRACUTANEOUS ABSORPTION Si
but also of those in which the serum of the injected animals
was used.
In addition, precipitin and agglutination tests were made
upon the serum of both normal and immunized guinea-pigs,
which had received no intracutaneous injections of antigen, in
order to demonstrate normal agglutinins, if present, and to
detect any horse serum remaining in the circulation from the
immunizing treatment. Normal agglutinins were not present
in the serum in amounts capable of having an appreciable bear-
ing upon the titrations to be made upon the tissues and sera of the
injected animals, nor was there a significant residue of horse
serum remaining from the immunizing injections, since a positive
reaction with a potent rabbit antihorse serum was never obtained
in dilutions greater than 1: 200. Tissue extracts from both the
normal and the immunized pigs, in the dilutions tested—1 : 50—
failed to show either normal agglutinins for B. dysenteriae or
precipitable horse serum.
TABLE 2
The transference of precipitable horse serum. Precipitin tests
HOURS | DISAPPEARANCE OF HORSE APPEARANCE OF HORSE CHANGE IN PRECIPITIN
SERUM FROM THE TISSUES SERUM IN THE SERUM VALUE, GUINEA-PIG SERUM
6 1: 12,800 1: 1,600 1: 3,200
12 1: 6,400 1: 6,400 1: 1,600
24 1: 1,600 1: 12,800 1: 600
48 1: 1,600 1: 6,400 1: 400
Tests upon the sera of the immunized pigs showed that a
certain amount of precipitin had been elaborated in the course
of the immunizations. In this connection it is interesting to
note that the precipitin of the serum decreased as the absorption
of antigen proceeded. This point is illustrated in the following
table which gives the results secured in one series of titrations
made upon pigs immunized to horse serum.
That the greater part of the intracutaneously injected serum
passes into the circulation and is not immediately taken up by
the tissues, cutaneous tissue being the index, is shown by the
following. At the time of removal of the portion of skin into
38 G. H. SMITH AND M. W. COOK
which the injection was made a section of uninjected skin from
the other side of the animal was removed and extracted. Titra-
tions conducted upon these extracts in parallel with extracts
from the injected tissue gave values as indicated in table 3.
TABLE 3
The distribution of absorbed antigen in immunized pigs
a HORSE SERUM IN THE HORSE SERUM IN THE HORSE SERUM IN THE
ELORIEE INJECTED TISSUE UNINJECTED TISSUE CIRCULATION
a. Precipitin tests
6 1:12,300 | 1: 200 1: 1,600
12 1: 6,400 | 1: 400 1: 6,400
24 1: 2,400 1: 400 1: 12,800
48 1: 1,000 1: 400 1: 6,400
b. Agglutination tests
6 1: 12,800 1: 400 | 1: 1,600
12 1: 12,800 1: 200 1: 6,400
24 1: 3,200 1: 1,600 1: 12,800
48 it 800 1: 1,600 1: 12,800
The facts made evident in tables 1, 2, and 3 indicate clearly
that the process of immunization has developed a mechanism
whereby the specific antigen is removed from the cutaneous
tissues to reappear in the serum at a rate far in excess of that
occurring in normal animals. This conclusion is based solely
upon the ability of normal and specifically immunized guinea-
pigs to absorb horse serum.
Before continuing the work ‘by introducing a second, heterol-
ogous, antigen it was necessary to determine whether or not
there exists in normal animals a selective action for a particular
antigen. Accordingly an experiment was conducted in which
normal pigs were injected intracutaneously with the antigens
previously mentioned, that is, with antidysentery horse serum
on one side of the body and with antigonococcus goat serum on
the other. In making the intracutaneous injections every pre-
caution was taken to insure the retention of equal amounts of
each antigen by the tissues and an effort was made to introduce
the needle at the same depth into the skin. The general ap-
SPECIFICITY OF INTRACUTANEOUS ABSORPTION 39
pearance of the elevation, which remained after injection and
the removal of the needle, indicated that the procedure was
earried out with considerable uniformity.
In the table which follows data are given showing the reactions
secured with normal pigs.
TABLE 4
The absorption of horse serum and goat serum in normal animals
a. Precipitin tests
HOURS HORSE SERUM IN THE SKIN GOAT SERUM IN THE SKIN
6 1: 12,800 1: 12,800
12 1: 6,400 . 1: 6,400
24 1: 2,400 Lid;200
48 1: 2,400 1: 2,400
HORSE SERUM IN THE CIRCULATION GOAT SERUM IN THE CIRCULATION
6 is 150 ta 200
12 1: 1,600 1: 3,200
24 1: 3,200 1: 3,200
48 1: 600 I<. 4 000
b. Agglutination tests
DYSENTERY AGGLUTININS IN THE SKIN | GONOCOCCUS AGGLUTININS IN [HE SKIN
6 1: 12,800 1: 12,800
12 1: 12,800 1: 12,800
24 1: 6,400 1: 4,800
48 200 1: 1,600
DYSENTERY AGGLUTININS IN THE GONOCOCCUS AGGLUTININS IN THE
CIRCULATION CIRCULATION
6 1 100 1 150
12 1: 1,600 1: 1,600
24 1: 8,000 1: 9,600
48 1: 1,600 1: 2,400
This experiment, as well as those that follow, was repeated
several times, in each case duplicate animals being used upon
each time interval. Minor variations occurred which could
readily be ascribed to errors in technic, particularly in the intra-
cutaneous injection, but the general type of reaction remained
40 , G. H. SMITH AND M. W. COOK
constant with each of the factors used. A study of the results
shows that the elimination of horse serum from the skin of nor-
mal animals is not more readily brought about than the elimina-
tion of goat serum. In fact, if curves based upon an average of
the several animals are plotted for each of the several factors
titrated, it is found that they coincide almost exactly through-
out their entire course.
We therefore felt secure in assuming that the normal mechan-
ism of eliminating foreign protein from the skin operates equally
well with either the horse or goat serum. This gave us sufficient
warranty for attempting to detect a specificity of elimination
due to an alteration in this mechanism brought about through
immunizing treatment.
To this end a series of guinea-pigs was immunized by repeated
injections of horse serum. ‘Titrations, similar to those indicated
above, were then made upon those animals, a parallel series of
normal pigs being tested as controls. The results of such a test
may be tabulated as follows. .
But one conclusion can be drawn from this work, namely,
that the immune state so alters the process of antigen absorp-
tion from the skin that the specific antigen is removed more readily
than in the normal state. Moreover, it is of especial interest
that in an immunized animal the non-specific antigen is not
eliminated as rapidly as is the same antigen from normal tissues,
for in the animals immunized to horse serum the rate of dis-
appearance of goat serum was delayed over that found in normal
pigs. It would seem, therefore, that in immunized animals, a
specific mechanism has been developed at the expense, in some
measure, of the normal process. To test this point in partic-
ular many repetitions of this experiment were made and in
every case such a relationship appeared.
Throughout the work with this series of animals it was re-
peatedly noted that in the course of the preparation of the
tissue extracts the portions of tissue removed from the immu-
nized animals differed greatly in appearance. Those sections
into which horse serum, the specific antigen, had been injected
were much reddened and somewhat thickened, while those from
SPECIFICITY OF INTRACUTANEOUS ABSORPTION 41
the same pig into which goat serum had been injected remained
free from thickening and inflammation. A similar condition
has also been noted in connection with our work upon the absorp-
tion of antigen from the cutaneous tissues of sensitized pigs.
TABLE 5
The absorption of horse serum and goat serum in immunized and normal animals
a. Precipitin tests
IMMUNIZED PIGS NORMAL PIGS
Horse serum in skin | Goat serum in skin | Horse serum in skin | Goat serum in skin
6 1: 12,800 1: 12,800 1: 12,800 1: 12,800
12 1: 6,400 1: 12,800 1: 6,400 1: 6,400
24 600 1: 6,400 Be 837400) 1: 3,200
48 Lg 400 1: 6,400 1: 2,400 1: 2,400
Horse serum in Goat serum in Horse serum in Goat serum in
circulation circulation circulation circulation
6 1: 1,600 Li: 50 Le 200 1 150
12 1: 6,400 is 100 1: 1,600 1: 2,400
24 1: 9,600 1 100 ee 33200 1: 2,400
48 1: 4,800 1 200 Ais 800 1: 1,600
b. Agglutination tests
| Dysentery agelutinins Gonococcus Dysentery agglutinins Gonococcus
in skin agglutinins in skin in skin agglutinins in skin
6 1: 12,800 1: 12,800 1: 12,800 1: 12,800
12 1: 6,400 1: 12,800 1: 12,800 1: 9,600
24 1: 1,600 1: 6,400 1: 6,400 1: 4,800
48 1 800. 1: 6,400 1: 1,600 1: 1,600
Dysentery agglutinins} Gonococcus agglu- |Dysentery agglutinins! Gonococcus agglu-
in circulation tinins in circulation in circulation tinins in circulation
6° 1: 1,600 None 1 100 Rs 200
12 1: 2,400 Le 200 1: 1,600 1: 2,400
24 1: 9,600 1: 1,000 1: 6,400 1: 4,800
48 1: 12,800 1: 3,200 1: 2,400 1: (3,200
The specific antigen here, also, brought about inflammation
such as never followed the injection of a non-specific antigen.
Moreover, this reaction could not be due to an inherent primary
toxicity, for the injection into normal pigs resulted in no reaction
characteristic of either the horse or the goat serum.
42 G. H. SMITH AND M. W. COOK
It is an established fact that the process of immunization
produces an altered reactivity of the body cells. Experimenta-
tion upon the reactions of the cutaneous tissues indicates that
this change is specific and produces no effect upon the reacting
properties of the cells for other antigens, except perhaps to
lessen their activity. From the point of view of absorption of
antigen, then, the immune state with the changes dependant
upon it, is the result of a heightened reactivity for the specific
antigen only, and does not stimulate the mechanism of elimina-
tion of heterologous antigens.
REFERENCE
(1) Smrra AND Coox: J. Immunol., 1917, 2, 269.
SPECIFIC REACTIONS OF THE BODY FLUIDS IN
PNEUMOCOCCIC INFECTIONS!
G. R. LACY anp C. C. HARTMAN
From the William H. Singer Memorial Research Laboratory, Pittsburgh,
Pennsylvania
Received for publication November 17, 1917
The problem of the reaction of the body fluids in pneumo-
coccus infections has claimed our interest for some time. To-
gether with the grouping of the pneumococci which was begun in
our laboratory in the fall of 1916, the following studies were
made to determine if possible, the extent and the way in which
the body fluids reacted to the pneumococci and to antipneumo-
coecic sera. The materials examined were obtained from pa-
tients admitted to the Allegheny General Hospital. Sputum
from persons presenting clinical manifestations of pneumonia
were sent to the laboratory at the earliest opportunity and the
erouping made according to the method of Dochez and Gillespie
(1). When the type of pneumococcus to which the infection
belonged had been determined, tests were begun for the purpose
of demonstrating any defensive antibody formations that might
be present in the body fluids. For this purpose, examinations
were made of the blood serum, spinal fluids, urine, pleural fluid
and, in those instances where autopsies were held, pericardial
fluids. These fluids were tested for agglutinins, precipitins,
precipitable substances, etc.
The sera, urines and other materials were collected at varying
intervals with reference to the time of onset of the infection and
to the condition of the temperature curve. From some of the
patients we obtained these materials as early as the second day
of the illness and from then repeatedly until complete recovery.
In some instances where death occurred early, only a single test
1 Read before the Pittsburgh Academy of Medicine, April 24, 1917.
45
G. R. LACY AND C. C. HARTMAN
44
a
0
0
0
0 |++++l0
0
0
AI dQouop
0
0
0
++++[t+++4+]+++4]+4+4++]4++4+4+
SIoquINU e41Nz[NP)
o|o |++++]++++)++++\++++4]++++4] 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
t+++tl+t4+tl[++t+]+++4t] T pur al
0)
0
0
0
0
+++4+[44++4]4++4+4]4+4+4+4+
[pus Tie 1G Paech
1a!
68
€€
ONALLVd
JO
uqddWoNn
BODY FLUIDS IN PNEUMOCOCCIC INFECTIONS 45
was made. In all instances where we could get sufficient ma-
terial, the agglutination tests were made against at least four
different cultures of types I, II and IV and one culture of type
III. We invariably used the organism isolated from the pa-
tient whose serum was being tested. These multiple cultures
were used with the hope of finding one or more cultures that the
serum would agglutinate.
The method of procedure which we used is very well illus-
trated in table 1, wherein the reactions of a typical example of
each group except group II is shown.
In this table the number of each culture corresponds to the
number of the patient from whom it was isolated; 1.e., ‘‘culture
33” was isolated from “‘ patient 33,’ ‘‘19” from “‘patient 19,” ete.
The patients were grouped according to the type of organism
isolated from them. Patients 33 and 19 were typical group I
and group II patients respectively and their sera agglutinated all
cultures belonging to their respective groups and no others. All
of the other patients were atypical as shown by the table. Pa-
tients 11, and 12 and 25 were especially interesting in that each
had two separate and distinct attacks of pneumonia before leav-
ing the hospital and that each recovered fully from both attacks.
These two patients we will take up somewhat in detail.
Patient 11 had as his first infection a member of group IV
and as his second a member of group I. He was exposed to a
group I infection by having occupied a bed in the ward on two
occasions of some days duration, adjacent to a patient from whom
a pneumococcus, type I, had been isolated. His serum, as will
be seen in the table, agglutinated all group I organisms or those
corresponding to his last attack, and did not agglutinate any of
the group IV organisms. Unfortunately his first culture had
been lost before this phase of the work was taken up, so we do
not know whether his serum would have agglutinated it or not.
Patient 12 and 25 had as his first infection a member of
group II and as his second a member of group I. We have no
definite history of such intimate exposure in this instance, as in
the case of patient 11, but he was kept in a large ward with
patients suffering from infections with the other types of pneu-
46 G. R. LACY AND C. C. HARTMAN
mococci. His serum showed agglutinins only for organisms be-
longing to group II, or those corresponding to his first infection,
and none for those belonging to group I—not even for his own
last culture, no. 25 of table 1. We believe that if one were
fortunate enough to follow a sufficiently large number of reinfec-
tions or recurrences, much more light might be thrown upon the
immunity or lack of immunity established by infections with
pheumococci.
Table 2 shows the results of the agglutination tests carried out
with the sera from fifty patients.
TABLE 2
GROUP NUMBER POSITIVE pete |
Di cioiete aban soe md Sargon Seek ee © eee eee 21 15 71.4
Dc Ackte ster ooo 2.6 sc as ee OE ee ene 13 10 77.9
| Ie te de PE re me ree yer ns At ¢ 0 0 0
DIV 5 POE IS. al dled oe, See eee ee 10 5 50.0
Unclassified (clinically pneumonia)..... 6 0 0
Motels as. ccs pee ee ee eee 50 30 60.0
This differs somewhat from the report by Chickering (2) who
found specific agglutinins in 100 per cent of no. I’s, 53.8 per
cent of no. II’s and 55.5 per cent of no. IV’s.
Of twenty-one patients with no. I infections six did not show
specific agglutinins in their sera. Two of these died and four
recovered; one of the ones who recovered had a blood stream
infection and another an empyema necessitating a stay of some
months in the hospital. Only two patients with fatal infections
showed agglutinins in their sera and both had received anti-
pheumococcic serum for group I before the test was done. Of
the three patients with no. II infections who did not show spe-
cific agglutinins in their sera—one ran a very mild course; an-
other, a small child, developed an empyema from which it re-
covered after many weeks; and the other one died. Of the five
patients with group IV infections, showing no specific agglu-
tinins in their sera, one had a very mild infection, two developed
empyema and two died.
BODY FLUIDS IN PNEUMOCOCCIC INFECTIONS 47
In the thirteen patients having infections with organisms of
group II, whose sera were tested, there were three sera which
showed that the infection was caused by organisms belonging to
sub-group II, one to sub-group IIx, and the other two to sub-
group Ila or IIb.
In regard to the appearance and duration of the agglutinins
it may be said that in all instances in which the serum was ob-
tained during the fastigium the agglutinin tests were negative,
except in the two patients mentioned above who had received
antipneumococcie serum therapeutically. The agglutinins ap-
peared only during or after defervescence. As would be ex-
pected, the time during which they were demonstrable varied
greatly. There was an apparent tendency for the agglutinins for
organisms of group I to persist for a longer time than for the other
groups (II and IV), and for those for group II to persist longer
than those forgroupIV. We were unable to test any for group IIT.
As has been noted before (Chickering (2), Lister (3) ) the
agglutinins in instances of human infections correspond with
the experimental findings resulting from immunization of ani-
mals against the various groups of pneumococci in being spe-
cific for the organism or group to which the infecting organisms
belongs. That is, a patient having an infection with a group I
organism shows agglutinins only for the organisms in that
group (or sub-group); a patient having an infection with an or-
ganism belonging to group IV possesses agglutinins only for his
homologous organism and no other.
We have tested a number of urines from patients whose sera
gave strongly positive reactions and were unable to demonstrate
at any time the presence of agglutinins. Agglutinins were not
demonstrable in the spinal fluid in those patients having a
meningitis, except in two who had received antipneumococci
serum both intravenously and intrathecally. In several of the
patients with meningitis we were unable to carry out aggluti-
nation tests with their sera, so that we cannot say whether
agglutinins were present or not in their blood. All of these
patients died, so that/as a group they were not favorable for the
development of agglutinins.
48 G. R. LACY AND C. C. HARTMAN
In carrying out agglutinin tests on spinal fluids which con-
tained a great number of pneumococci we noted that, upon the
addition of the antipneumococcic serum, a precipitate was formed
almost immediately. We did not appreciate the significance of
this reaction until Blake (4) reported his observations upon the
precipitin reaction on the urine. Since this precipitin reac-
tion offered a rapid and accurate method of diagnosis, we took
advantage of it and found it to be a reliable test, if a sufficient
number of dilutions of both no. I and no. II sera were used. In
order to explain our findings in the tests on spinal fluids and
urines we made the following cross tests. Cultures were made
in broth and after sufficient growth had occurred we centrifu-
galized them and pipetted off the supernatant, clear fluid.
Serum for groups I and II were added in varying dilutions to
the spinal fluid, urine and broth separately. Precipitates oc-
curred in all three with the corresponding serum. Mixtures of
spinal fluid and urine, spinal fluid and broth and urine and
broth showed no precipitates. Therefore, the precipitating sub-
stance, ‘‘precipitin,’’ was shown to be present in the serum and
precipitable substance ‘‘precipitinogen’”’ in the urine, spinal
fluid and broth. We would advise, therefore, as a point of
practical value—that, in cases of meningitis of indefinite etiology,
a precipitin test with antipneumococcic serum be made as a
rapid method for determining or excluding a pneumococcus
infection with either group I or group II..
DISCUSSION
The agglutinin tests on the sera of fifty patients showed that
humans react to their acquired infections with the pneumo-
coccus in a specific manner similar to that experimentally pro-
duced in animals; i.e., the patient produces agglutinins only
for the homologous organism or group to which the homologous
organisms belongs. We did not obtain, however, as high a
percentage of positive results as Chickering in the instances of
group I infections. The very mild infections and the fatal
infections were less apt to show specific agglutinins.
BODY FLUIDS IN PNEUMOCOCCIC INFECTIONS 49
The agglutinins did not appear until during or after defer-_
vescence. They remained demonstrable in the serum for vari-
able periods of time. There seemed to be a tendency for them
to persist longest in group I, next in group II and shortest in
group IV.
We were unable to find any evidence of the presence of specific
agglutinins in the urine, even in those patients having strong
agglutination reactions in the serum. In the precipitin tests
with antipneumococcic serum and urine we believe that the
specific “‘precipitin” is in the serum and the precipitable sub-
stance, “‘precipitinogen,”’ in the urine. In patients with pneu-
mococcic meningitis, the precipitable substance may be found
both in the urine and cerebro-spinal fluid.
CONCLUSIONS
1. Specific agglutinins usually appear in the serum of patients
during or just after defervescence.
2. The agglutinins in the serum seem to persist longest in
group I; somewhat less in group II and for the shortest time in
group [V. .
3. The urine contains specifie precipitable substances but
does not contain agglutinins or precipitins.
4. The spinal fluid in cases of pneumococcic meningitis also
contains specific precipitable substances similar to those in the
urine, but neither agglutinins nor precipitins.
5. The precipitin reaction may be applied to the spinal fluid
in pheumococcus meningitis as a practical means of rapid classi-
fication of the infection.
REFERENCES
(1) Docurz anp Avery: Journal American Medical Association, 1913, 61, 727.
(2) CuickreRING: Journal Experimental Medicine, 1914, 20, 599.
(3) Lister: South African Institute for Medical Research (Publications—De-
cember 22, 1913, 1).
(4) Buake: Reported at the meeting of the American Association of Immunolo-
gists, April 6, 1917.
Vi
=) ae
‘| , : ; ne
; |
F a
co
’
‘;
tae
4 _
\ ‘ : ; ok ; bite
~ , =n
yh, ui
= \
‘
‘
.
<
1
STUDIES ON THE ANTITRYPSIN OF SERUM
B. FUJIMOTO
From the Forensic-Medical Institute of the Imperial University at Tokyo
Received for publication November 23, 1917
INTRODUCTION
As to the nature of the antitrypsin of serum, a great deal has
been written, and even at the present day various views are held
in regard to it. In our laboratory several tests were performed
in order to see how it can be inactivated and which part of the
serum exerts the antitryptic action.
PREPARATION AND METHOD OF OUR TESTS
In our experiments von Bergmann’s casein method was modi-
fied a little to make the reaction finer. Namely:
1. Solution of casein: 0.5 gram of casein was dissolved in 50
ec. of ; NaOH solution and the solution was warmed a while.
Then the solution was neutralized with 4 muriatic acid and
finally diluted to 500 ce. with normal salt solution. This solution
of casein in our case is, therefore, less concentrated than that
usually employed; i.e., it is of 0.1 per cent instead of 0.2 per
cent.
2. Solution of trypsin: 0.05 gram of trypsin was dissolved in
5 ec. of physiological salt solution, to which 0.5 cc. of 7 NazCOs
solution was added. To make it 100 cc. (0.05 per cent) nor-
mal salt solution was then added to the mixture up to 100 cc.
thus making a 0.05 per cent solution. This solution is also much
more diluted than that ordinarily used (0.1 per cent).
3. Solution of acetic acid: Acetic acid, 5 ec.; alcohol, 45 cc.;
aqua destillata, 50 ce.
4. The doses in each test tube were always: Solution of casein,
2 cc.; solution of trypsin in graded doses; blood serum (1 per
cent), 1 cc.
ol
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2
52 B. FUJIMOTO
The reason why all of these solutions were used in less con-
centrated form than ordinarily, is that the reaction may be
more easily and clearly distinguished. These solutions being
mixed, and after being allowed to stand half an hour in the
thermostat (37°), the series were examined with the solution of
acetic acid, which was poured slowly on the surface of reagents
in the test tube. When the casein is not yet digested completely
and a trace of it exists still in the reagents, a beautiful circle of
gray precipitate is easily to be seen at the plane of contact be-
tween the acetic acid and the reagents. In our following tables
antitrypsin is not expressed in units; but if it is desired to ex-
press antitrypsin content in units, it may be calculated in the
same manner as in the original method, because the relation be-
tween the concentrations of casein and that of the trypsin
solutions is not changed, though both solutions are less con-
centrated.
I. INACTIVATING TESTS OF THE ANTITRYPSIN OF BLOOD SERUM
1. Inactivation by heating
Inactivating experiments for the antitrypsin of blood serum
have been performed by several investigators. Vandevelde (1)
saw that the antitrypsin may be destroyed in greater or less
degree by heating it at 55°, 60° and 65°C.; Achalme (2) proved
that they became completely inactive by heating at 65° to 70°C.
Jochmann and Kantorowicz (3) found that the antitrypsin
of human serum may be completely inactivated by heating at
66°C. for half an hour. Kurt Meyer (4) has shown that the
antitrypsin of serum is destroyed in but a slight degree (only
one-fifth of it) by heating at 56°C. for half an hour. Kam-
merer (5) compared the resistance of the antitrypsin of human
serum with that of ox serum, and he saw that the former is
more resistant than the latter, and that the latter is inactivated
almost completely by heating at 65°C. for half an hour, but that
it can scarcely be inactivated by heating at 56°C. for half an hour.
Doblin (6) did not confirm these results, but he has, on the con-
trary, proved that the antitrypsin is thermostable and of a
STUDIES ON THE ANTITRYPSIN OF SERUM 03
colloid nature. In short, the temperature that is necessary to
inactivate the serum antitrypsin has been described differently,
and there is one investigator that denies its thermolability.
In order to see whether the antitrypsin of serum is thermo-
stable or not, and if it be thermolabile, at what point of tem-
perature it may be completely inactivated, we have begun with
the following tests.
Before proceeding to the main tests, we have had to determine
at what point the sera coagulate. If they begin to coagulate by
heating, the factors of the reaction are changed by it and it be-
comes hard to judge the results correctly, so that the coagula-
tion of the sera must be avoided as much as possible. The
tests with regard to this, were performed as follows.
First, the undiluted serum was heated at various points;
secondly, it was diluted with distilled water or with normal salt
solution and then heated at various points. The results are
shown in the following table.
TABLE 1
Coagulation tests of rabbit serum in its different concentrations at 72°C. and 80°C.
av 72°C. at 80°C.
CONCENTRATION OF
SERUM
Not diluted
per cent
100 +r “te
Diluted with
Aqua destillata Salt solution Aqua destillata Salt solution
50 + = hsp oe
20 — — + _
10 — — —
Fs = el ns _
1 as Bi es =
+ = Coagulation; — = no coagulation.
Thus it was decided that the coagulation of the sera can be
avoided by diluting it. In our tests we have, therefore, diluted
all sera to 1 per cent, and then heated them at the high tem-
perature that is necessary for their inactivation.
54 B. FUJIMOTO
a. Serum of rabbit. Sera of rabbits were heated at various
temperatures, i.e., at 65°, 70°, 72°, 75°, 80°C. for from one min-
ute to three hours. The results are shown in the following
tables.
TABLE 2
Rabbit serum heated at 65°C.
Serum heated
SEMIN ULES se. ae eee
HIM INUMGESs nek lee. Ree
IOsmimutest-e. os. Lee
—
a
se
SO MMINUtES.. 62 ere a
>
--
-—
It
|
|
|
|
GOlmmutes:. 2... oe eee
120 minutes......... Nite AER
USOeMINUbeS ee... eee
|
|
|
|
|
+++++++ +4
+t+++4+4++ 41
+t+tt+4++ 41
++t++h++ +1
+ = digestion incomplete; — = digestion complete.
TABLE 3
Rabbit serum heated at 70°C.
0.4") (0:5° | 0-6 4] 0:7 || 0:8-) 0:9.) 1.1) 1) at 25) as
INorserlmky.. Wiehe: see eae oh See |S eS SS S| S | =
Serum nob heated... nwck eee oe +/+}+]}4+]/4+]/+}4+/4+/]41]-
Serum heated
SAMINULES ¥. nsihsriel aS 6 +it/+)4+]/+]}4+/—-/]-!/]-|]-
SAMINNUES se .-42 ssa eer Se ee fe a fe a en ee fe
LO MINUTES os 22s ew ns + fp] ef] ey | Hf SH SS
30 MINUTES: 3 soe oe oe yep) = bh | HP a pe ES SS
Ol aT BUtEs Ese: aim ee +/4+/+)+=]—-;-]-/;]-]-]-
AZO-MINUGES 0.- 165 eetesc4s Bes ees eo} ee ee a a
TABLE 4
Rabbit serum heated at 72°C.
0.5 | 0.6 | 0.7 | 0.8] 0.9} 1:0) 2-1 | 1.2 | 1.3 | 1-4
INORSeTUNY S320 oe. nes t+ye]—f]/—f]—-}—f}]—-f—-]—-] =
perum not heated: :...0.........8 +/4+/)4+)4+]/4+]+}]4+]}4+/)4+/]-
Serum heated
SMILES Sere ae 6 Hee eR eek +}/ty+y4ty)+]+=;)-}]-]-] -
DMINUeSsey eee eee eee t+y+ti4+s}4)y4+]4+}]—-];-)]-]-
LO PMIMUbeS eee ee a +/+} +y4+s)/+}]—-};-—-}]-]-]-
ANMMINIbEs Lee ss le Be +ytey+y) =] —-]—-]-}—-}]-] -
SU THMUGES: | lacey ae ee ee +)t}—}—y—]—)] —] —-] =] =
GOamimUtests apes oes +yt})+=]—-/—-}]/-]-/]-]-]-
120 TITLES Aci tee |
STUDIES ON THE ANTITRYPSIN OF SERUM 55
TABLE 5
Rabbit serum heated at 75°C.
DOSES OF TRYPSIN
INGTSCRUTMeermeerei eo... s be || a P| a
Serum not heated........ Es te esa ital lectee sto 3 stan |
Serum heated for
3 NWI Oooo ee eee eee nee ey ec | es | ee
EMPIMITIAUIGES irae fale Gis eis ce oe ees So ole: lc |) => |) || RS a
10; DOM So eee ee eee SN eel the |e a
3{0) TTUTNOOSS Oe nee ee eee as ep a all em 1 a ee
At 80°C. it took only seven and one-half minutes to inacti-
vate the serum antitrypsin completely.
It may be concluded from the above tables that the sera of
rabbits can be inactivated completely by heating at 75°C. for
ten minutes, or at 80°C. for seven and one-half minutes, and
that at temperatures lower than 72°C. the serum of rabbit cannot
be inactivated completely even if it be heated for two hours.
b. Serum of horse. Several experiments similar to the foregoing
were performed with sera of horses. The results are given in
table 6.
TABLE 6
Inactivation tests with the sera of horses
DOSES OF TRYPSIN
0.7 | 0.8 019) |) 9-00) a-0 e (Pies |e
INTO) GICTATIT SS 5 rence eee Ge SIERO on LAPS aoe ep ae = |e | | |
Mend NOL NEAEMS. 2.05056. oe ss se ees eee: +}/4+})4+/4+/4+/4+/4+/]-
Serum heated at
50°C. for OpMINULES awe tases et tpt] tl] ty ste] st le] =
“er BOMMAINMPES 2:8 - owes cae ke Spee +i+it+i/+/4+/4+/.+]-
6 LOnmInuUtesss sn seens. eh eee Se ee ae a |
fi aes FATHULESEL) salted 55 AE ee) se sees SE = ye ee
z POMGINIGES, oiiellars he os ow ot See Ee fat | Fa ee lees a
cern tap MUNUMbES ee SHE |p ee ee
s LO SATINUIGES| et pees ct o. Loa = |) = SSeS | Sj i=
aoa i HUUMUIGES 5-2-5) ayn cess +}/—-/;/-—-]/-—-/|}-j;-]|-|-
56 B. FUJIMOTO
As we see in above table, the sera of horses can be completely
inactivated by heating at a lower temperature than those of
rabbits.
c. Serum of sheep. Further tests were undertaken in the same
manner with the serum of sheep and the following results were
obtained (table 7).
TABLE 7
Inactivation tests with the serum of sheep
DOSES OF TRYPSIN
0.7 | 08] 09 | 10 {1.2 | 14|16 | 18
INIOTSCLUMI ofan: tothe toca eee ee ey Sf S|] = | = |
Serum not heated... ....:......:eeeneease +)/4+/}/4/]/4+)/)4)4/)4+)/)-
Serum heated for ten minutes at
G5cO ee ee ee ee p24 sare AM Er arabs MA ie Cee
TO°W) Gb vic as ee oe ee ee Pee ea eg Ree poten eer =
UD Cp GS ahs. BSS a ee +|/—|/=- —|—-|-|-
SOS Re ie anne Sets dae eee ae ee +)/—-/-—-/;/-|-|]-|]-|-
Though Dédblin maintained that the antitrypsin of sera is
thermostable, it is certain that it may be inactivated by heat-
ing; i.e., it is thermolabile. And the temperatures which must
be applied for the inactivation differ in each case. Namely:
Rabbit serum at 75°C. for ten minutes
Horse serum at 65°C. for ten minutes
Sheep serum at 75°C. for ten minutes
Human serum at 65° or 66°C. for one-half hour (Jochmann and Kammerer) —
Ox serum, higher than human serum (Kimmerer)
Conclusion. It may be concluded that the temperatures,
which must be applied for the inactivation of antitrypsin of
sera, differ not according to individuality but according to spe-
cies and are between 65° and 75°C. As sera never coagulate at
75°C. in their lower concentration, it would be better at first
to dilute them and then heat them at 75°C. for ten to thirty
minutes for their complete inactivation.
Incidentally, a brief experiment for the inactivation of the
human urine was undertaken. The results obtained are shown
in table 8.
STUDIES ON THE ANTITRYPSIN OF SERUM A
TABLE 8
Inactivation tests with human urine
DOSES OF TRYPSIN
OL6r jaOem INOS meO-O0) 2.0) || 1.4 | 1.9°) 1,3) 14
ING) Wile. 5 <6 5a02 CRs =f || eee ee (en |r pe |
Urine not heated, 30 per cent, lce.....} + | +] +/+]/+/4+/+/+]-
Urine heated, 30 per cent, 1 cc. at
MethOUT. |... 202. on 24) ser) SEU ea ee bs se a ste
UE mo ee +)+)+)+)/+/+/]+] +] -
x PO OURY as ects ote SE Ngee if |) es, ee | ey = Ee
ae tor i oaant, fcd3% Avcdeeserschaces Safe ey eS PS ee
The antitryptic action of urine cannot be affected so easily as
that of sera; urine must be heated at 100°C. for one-half hour
for its complete inactivation.
2. Inactivation by shaking
We can find no record of the inactivation of serum antitrypsin
by shaking. It is well known that complement may be inacti-
vated by shaking at 37°C. for one-half hour. If the antitrypsin
of serum also could be inactivated by shaking as well as com-
plement, many interesting experiments might be undertaken as
to its nature. One cubic centimeter of serum was diluted to
10 cc. with normal salt solution and then shaken at various
temperatures.
TABLE 9
Inactivation tests by shaking (rabbit serum)
DOSES OF TRYPSIN
EOS Oe Nu OO. I Ree | ai || a
ING@ EGIING 5 5e8 See eee eee o _ — - — — — — =
Serum neither heated nor
Sha kehiaerne iets oe toe ber. - — — - bh + + + + —
Serum shaken for two to
3 hours: at
Be AC ee Ae ae ee ap it SF ae soul Seill Se ad = —
AGL CMe ete ees ar = Fill lie allo S| bale Se = =
Controlserum (notshaken) at
Sib rie Hn oe yen ok 2 + ae (arse se |) are ll sr = ==
AOE Creeps te marcas holcayees ei Sos + ar ie + ar |) ae i) ar = =
58 B. FUJIMOTO
Horse serum was similarly treated, the results being the same.
Conclusions. Antitrypsin of serum cannot be inactivated at
all by shaking.
II. EXPERIMENTS WITH SERUM GLOBULIN AND SERUM ALBUMIN
In regard to this problem, there are several different views.
Landsteiner (7), Miiller (8) and Opie and Parker (9) reported
that antitrypsin is mostly fixed to the albumin fraction of serum.
Glissner (10) stated that the antitrypsin of serum is all fixed to
the euglobulin, not to albumin. But Dodblin (6) and Kimmerer
(5) maintained that both the globulin and the albumin fractions
exert an antitryptic action, but the former less than the latter.
Beside this, Kiimmerer (5’) has recently made interesting ex-
periments, in which he has proved that the globulin fraction in-
creases after heating serum at 56°C. for one-half hour, while the
albumin fraction decreases, and that the antitryptic action of
the albumin fraction may be more easily inactivated than that
of the globulin fraction.
In our laboratory different tests were performed bearing upon
this question. To fractionate serum in globulin and albumin,
we have employed several methods.
1. Serum treated with ammonium sulphate
For the fractionation of 50 ce. of horse serum, there was added
an equal volume of saturated solution of ammonium sulphate and
the mixture was filtered. The precipitate was washed five times
with a half saturated solution of the same salt.
As ammonium sulphate precipitates casein, both fractions
must be dialyzed for its elimination. However, as it takes a
week or more to make them completely free from the salt by
dialysis, their antitryptic action might, also, be weakened by
the treatment. For this reason we were satisfied with a day’s
dialysis, after which time the concentration of the ammonium —
sulphate in the globulin and albumin fractions was so little that
the casein was not precipitated by it. Fortunately it was al-
ready proved by Kimmerer that ammonium sulphate does not
STUDIES ON THE ANTITRYPSIN OF SERUM 59
prevent the action of trypsin in any way; and Doblin saw that
the antitrypsin was not lost at all in the first week of dialysis.
Afterward the same result as Déblin’s was obtained by us as to
this property of trypsin (see the following chapter). It is, there-
fore, clear that only a day’s dialysis exerts no influence upon
the antitryptic action of serum.
If the doses of serum globulin (100 per cent) in the test tubes
are more than 0.2 ec., the globulin may be precipitated by
TABLE 10
Tests with serum globulin and albumin
DOSES OF TRYPSIN
No serum.. i MGetsterseetee | P= —shSo Sea =e
Peer cen cont Mice. veeeeeeees/ Le] eEE] +H] +] +l stir
Serum globulin*
ae S| a ee a
Not eae ce per cent, I cc..... to} te} py] +
(eaeper cent, 1 cc:. +y—]—f—F}—f—)] —]—-] —-—
Heated (75°, one-half hour) 20°
per cent, 1 cc.. ee es ae
Serum albumin
Not heated, 2 per cent, 0.5cc....| +] +] +/s=]|—/]—|—-|-|/-
Heated (75°, one-half hour) 2
per cent; 0:5 COs jae. s6 see soe = +/—}—}/—/]—-/]—-]—-]|]-]-
* The serum globulin was dissolved in normal salt solution in a volume equal
to that of the original serum.
acetic acid. Therefore, we must be careful in our experiments
not to misjudge the reaction that is caused by the precipitation
of serum globulin and not of casein. The precipitation of
globulin may be easily distinguished by its coloring.
As the antitryptic action of serum is not weakened by dialysis
at all, we can say that the antitryptic action of serum is consid-
erably weakened by precipitating with ammonium sulphate, and
that the antitrypsin is to be found not only in the albumin frac-
tion but also, in a small degree, in the globulin fraction.
60 ; B. FUJIMOTO
2. Serum treaied with carbonic acid
By the preceding experiment it was shown that the globulin
fraction of serum also shows antitryptic action. To confirm this
fact through some other experiments, we have treated the serum
with carbonic acid and 5; muriatic acid.
To 5 ce. of horse serum were added 95 cc. of aqua destillata,
which was cooled in ice, and then carbonic acid was led from
Kipp’s apparatus into it until the precipitate no longer in-
creased (about one-half hour). The precipitate consists solely
of globulin. It was separated from the albumin portion by
TABLE 11
Tests with serum globulin and albumin obtained by means of COz
DOSES OF TRYPSIN
0.7 | 0.8 | 0.9 | 1.0 | 1.2 | 14 | 16 | 2.6 | 2.8
ING SCRUM 4206 fice coos se ee teen aa ie —|—-—-|/—-|;-|-
Serum!) per cent, 1. Ce ..25 eee eee +/+)+/+}]+/4+)/-
Serum treated with CO:
Albumin part, 1 per cent, J ce:.:-) | ae) eo | aie
1.0 per cent...... SS sel SS eS
220) DED CEDbae-i1F: SP | == a el
Globulinpart, } 5.0 per cent...... Se Wa eee
1 ce. } 10.0 per cent...... +/+/+/+=]-|]-|-
Ex per cent...... +i/+/+{/+/+]/4+|{- :
50.0 per cent...... Se ee oa ee feo ol listen a alk
means of a centrifuge, and it was washed more than five times
with distilled water to render it free from albumin.
Both the globulin and the albumin thus obtained were exam-
ined as usual, the results being shown in table 11.
Such an experiment was repeated with the serum of the horse
and rabbit, and it was always followed by the same results. Of
course, both the globulin and albumin fractions could be inacti-
vated by heating.
Thus it has been demonstrated that the antitryptic action of
serum may be weakened only a little by precipitation with car-
bonic acid, and that the globulin fraction undoubtedly has
antitryptic action.
pa OR Se
~ oy ae Ag Re ye -. s
ice wee =.
STUDIES ON THE ANTITRYPSIN OF SERUM 61
3. Serum treated with 35) HCl.
Sachs employed this method to separate the complement of
guinea-pig serum into two components. To 0.5 cc. of horse
serum, there was added 4.1 ce. of 55) muriatic acid; after stand-
ing an hour at room temperature it was separated by means of
the centrifuge into two portions. The globulin fraction (..e.,
the precipitate) was washed with distilled water several times
carefully. The results of the tests were mostly the same as we
have seen above (table 12).
TABLE 12
Tests with serum globulin and albumin obtained by Sach’s method
DOSES OF TRYPSIN
0:5 | 0.6) | O67. O-85 (10h feuko! | stst| sina e7
PIGESE GUNN te yet. cick De « Wvintoie ere doce we a es ee (ee
Sexumplapermcent. 1 Ck. so:.- 2-sece 0 a ee ee ee ee
Serum treated with HCl
Albumin) partel per cent, 1 -ces.4.2\osh stee | Mates ste | ete ete
( 5 per cent.., +] =/—]|—|]/—/]—-—|]-
: 10 per cent..} + | +/+] =/—-|]-|-
Globulin part, 1 ce. Solpen cont.) bo| erie Mee eee ea me
(50 per cent..); + | + | +]+]+ pe eee
Globulin part obtained according to Sach’s method shows an
antitryptic action, but in less degree as in the case before.
Conclusion. Both globulin and albumin fractions have anti-
tryptic action, but the latter in less degree than the former.
III. THE NATURE OF SERUM ANTITRYPSIN
It has been shown by the above experiments that the anti-
trypsin of serum is to be found in both the globulin fraction and
the albumin fraction. To determine whether these proteins
themselves have such antitryptic action or whether they are
accompanied by some specific substances which exert an anti-
tryptic influence, further studies were undertaken.
62 B. FUJIMOTO
1. Is the antitrypsin dialysable?
In the first place, the globulin and albumin fractions of serum
were dialyzed. As we know that globulin and albumin are not
dialysable, if the antitryptic action of sera be lost by dialysis,
there must be some specific acting substances; if it be not lost,
antitryptic acting substances in sera must be undialysable and it
TABLE 13
Dialyzation tests
DOSES OF TRYPSIN
INO SELUM 2k folasc- occ eee eas eee eer eeeROES +)}—/]—]—-/]-—-];-|]-
Serum not dialyzed 1 per cent, 1 cc............ Jt}+)4+)]+},4+]4+)]-
Serum (1 per cent) dialyzed for
didays ll .€C2.\:..05 sk teeeiee eee eee +)/+y+ty ti +] =i] -—-
PORWR; il CC's. 2. /cac: ee eee ere ee +i4+)/+)4+/4+)={/-
| goidays, 4 ce... 2522-52 -e te see: eee eee +/+) +]/4+/4+)+=—-
ge dideys, 1166 -<.302 Sk ce ere oe ee ee eee +)/+]+]}]+)})+)+{-
Serum (1 per cent) not dialyzed but otherwise
under the same conditions for
tday:. 1 ces, 24at- Pelee ete Pee SS eamlae i) == |) =
QiGays, ErGCs 202 ran oS)
. choke te eae Cee Se Se ea ea
Albumin (1 per cent) dialyzed for
DOAYS* 1sCOee sgancie aceon eee Se eel! sel he
4 days, 1 eels 4-2 Pee. ee eee +it+y+]/+]/4+)]+4+/-
Albumin (1 per cent) not dialyzed, 1 cc. for
2:£0 A Gays. ot he oie ss speech eee ee: “eal eal ea S| ee
might be said that antitryptic acting substance is nothing but
protein itself. The following tests were always carefully con-
trolled by undialyzed samples which were exposed otherwise to
the same conditions.
It is seen that the antitryptic action of serum as well as that
of globulin or albumin is not weakened at all after one to four
STUDIES ON THE ANTITRYPSIN OF SERUM 63
days’ dialysis. We can say, therefore, that the antitryptic sub-
stance of serum at least is not dialysable.
Human urine was similarly tested but in this case the
antitryptic property disappeared completely after two days’
dialysis.
; TABLE 14
Dialyzation test with human urine
DOSES OF TRYPSIN
0.3 | 0.4 | 0.5 | 0.6 | 0.7 | 0.8 | 0.9 | 1.0 | 1:2
MAMVMTITEITN GC Pee PENIS clei) See yavors avacdeee Bes ae +/+/—/—;-—-/—-];-]/-—-]—
Wrme 50) per cent, 0:6 ce.............. +i i+ty4+y)+)}4+]/4+)4+);)4+)-—-
Urine 50 per cent, 0.6 cc. dialyzed for
2) CES HEbs Gace oe ee nena a eee +/+ ;—-}]/—/]—]—-]—-]-] =-
Urine 50 per cent, 0.6 cc. (control)
not dialyzed (only standing for
PEGA S acess oats, sslveenes s SEaROR Yh eee Se se dl ae ar eae | sr ae) aS
2. Experiment with pepton
A brief experiment was carried out with pepton, which is
dialysable. But pepton itself was found to have no antitryptic
action, and it neither prevented nor accelerated that action.
3. Ether extract of serum
Serum was extracted with ether to see whether the substances
that are soluble in ether, have an antitryptic action or not. O.
Schwarz (11) stated that serum became inactive after the lipoids
were extracted from it and that a mixture of the lipoids with
protein exerts an antitryptic action. Kammerer (5’) refuted
Schwarz’s theory by his experiment, in which he saw only a
slight decrease of antitryptic action (only one-third) after ex-
tracting the serum five to six times with ether.
Five cubic centimeters of horse serum were treated with 50
ec. of ether in our experiment. The extract was redissolved in
ether to make it pure and was made into emulsion with physio-
logical salt solution. The results were as follows (table 15):
64 B. FUJIMOTO
TABLE 15*
Tests with ether extract of serum
DOSES OF TRYPSIN
0:5 | 0.6 | 0:7 | 0:8 | 09° |) 1.0) 2
INOW SEGUINY 29S = Shi 2 2 are se rele Soe eee er eae a ee | = |) ==
Serum 1. per cent, 1 ce.......5+...2:9sgeeeee-eeal ae | 1 SE} Sete eee
Ether extract
20 per cent? 1 \CCs..5. : a... ee Ee eer ee =i aS | Ht = || =
HOspericent, 1 6C 122s 04.02 oe eee +) =};/—-—-}—-]—-}y-]-
* See the note to table 10.
It was established through the above experiments that the
antitryptic substance is not dialysable and that the ether extract
-of serum does not contain such a substance. We have, therefore,
proceeded to the following tests.
4. Crystalline albumin
It seems now probable that the antitryptic substance of serum
may be nothing but the protein of the serum itself. This ques-
tion is not yet definitely decided, and we can not find a record
of any previous experiment which was undertaken to settle it.
Taking advantage of the property of serum albumin to crystal-
lize, we have undertaken an experiment with crystalline serum
albumin and, also, with crystalline ovoalbumin. As albumin
that is not denatured and that is free from any admixture can
be obtained by crystallization, it seemed that an experiment
with crystalline albumin might bring a decision with regard to
this problem. To obtain the crystalline albumin, I have em-
ployed the following method.
Fifty cubic centimeters of horse serum and the same quantity
of saturated ammonium sulphate solution were mixed and then
filtered. To 50 cc. of the filtrate was added more saturated
ammonium sulphate solution until further precipitation began,
and then 1 to 1.25 ec. of saturated ammonium sulphate solution,
to which acetic acid to 10 per cent had been added. After
twenty-four or forty-eight hours (in the ice box), we obtained
typical crystals of serum albumin; the crystals were separated
STUDIES ON THE ANTITRYPSIN OF SERUM 65
from the fluid by centrifugation and washed five or more times
with a half saturated solution of ammonium sulphate, which
contains the salt and acetic acid in the same relation as above.
The crystals were finally dissolved in distilled water or physio-
logical salt solution.
TABLE 16*
Test with solution of albumin crystal
DOSES OF TRYPSIN
0.5 | 0.6 | 0.7 | 0.8 | 0.9 | 1.0 | 1.2 | 1.4] 1.6
ING SERUM eee ae be See On Oe ee +/—}—}]—};—}]—-);—-]-—-
Senumeolepenicent, lice: «4. ....%...... +) ] +] eye ey ae ey K—
Crystalline serum albumin, 1 cc.
[OSE CAN se.5:6 cole hse eee +/+;)+/) +=); —-/—-/;—)];-]-—
iQ), JOGTEGEI tess on 0 6.0 SDI eee +/)+/)+)/)+;+/4+)/)-/]-/-
Crystalline ovoalbumin, 1 ce.
5) jOXeNP OSTA a Sbrotoit cs IOR Ee ae en Ree +)4+})+)]/+=;);—-—-]-—-]-]-]-
NORMGTRCENGer en ete ee oa cae se +)4+)4+)+);)4+/+/-/]-)|]-
* See the note to table 10.
As the crystalline serum albumin is pure and free from any
admixture, it must be accepted that the serum albumin itself has —
an antitryptic action.
Conclusion. There may be many kinds of substances in
serum which exert an antitryptic influence, as Oppenheimer (12)
has said; but there is no doubt that albumin itself exerts such
action. It is still open to question, whether it is the serum
protein alone that is the antitryptic agent or whether some
other substances beside the protein are also antitryptic. Fur-
ther quantitative experiments are necessary to solve this question.
RESULTS
1. The temperatures that are necessary to inactivate sera,
are different according to species, and are between 65° and 75°C.
2. The antitrypsin of serum can not be inactivated by shaking.
3. Both the globulin and the albumin fractions are antitryptic,
but the former is less so than the latter, as Déblin and Kim-
merer stated.
66 B. FUJIMOTO
4, The antitrypsin of serum is not dialysable.
5. Ether extract of serum has no antitryptic action.
6. Crystalline serum albumin is antitryptic.
7. It is not yet decided whether serum contains antitryptic
substances beside the serum protein or not.
It is a great pleasure to me to thank Prof. Dr. K. Katayama
and Prof. Dr. 8. Mita for suggesting this problem and for their
kind advice.
REFERENCES
(1) Vanpevetpe, A. J. J.: Uber die Wirkung der Erwirmung auf Protease.
Biochem. Zeitschr., 1909, 18, 142.
(2) AcHALME: Propr. pathol. de la trypsine. Annale de 1’Institute Pasteur,
1901, 15.
(3) JocHMANN AND KantTorowicz: Antitrypsine und Antipepsine im mensch-
lichen Blutserum. Zeitschr. f. klin. Med., 1908, 66.
(4) Kurt Meyer: Uber Trypsin und Antitrypsin. Biochem. Zeitschr., 1909,
Nr. 23.
(4’) Kurr Meyer: Uber die Natur. d. Serumantitrypsins. Berl. kl. Woch-
enschr., 1909, Nr. 42.
(5) KAmMeprerR, H.: Studien iiber die Antitrypsine des Serums. Deutsch.
Arch. f. kl. Med., 1911, 103.
(5’) KAmMmerRER, H., anp Lupwic ANBRy: Untersuchungen iiber die Bezieh-
ungen der Serumeiweisskérper zur Antitrypsinwirkung. Biochem.
Ztschr., 1913, 48, 247.
(6) Désiin: Untersuchungen iiber die Natur des Antitrypsins. Ztschr. f
Immunitiatsforsch., 1909, 4, 229.
(7) LANDSTEINER: Zur Kenntniss der antiferment. lytischen und agglutinier.
Wirkung des Blutserums und der Lymphe. Zentralblat f. Bakt.,
1900, 27.
(8) Mituer: Uber das Verhalten des proteolyt. . Leukozyten-fermentes.
Deutsch. Arch. f. kl. Med., 1908, 91.
(9) Opin AND Parker: Leucoprotease and antileucoprotease of mammals and
of birds. Journ. of Exp. Med., 1907.
(10) GuLAssNzER: Antitrypt. Wirkung des Blutes. Hofmeister’s Zeitschr., 1904.
(11) O. Schwarz: Die Natur des Antitrypsins im Serum und der Mechanismus
seiner Wirkung. Wiener kl. Wochenschr., 1909, Nr. 33.
(12) OppENHEIMER. C.: Die Fermente und ihre Wirkungen. 4. Aufi., 1913, I,
478.
THE CONSTANCY OF THE PROTEIN QUOTIENT
DURING INTENSIVE DIGESTION AND
PROLONGED STARVATION
SAMUEL HANSON
From the Department of Biochemistry and Pharmacology, Rudolph Spreckels
Physiological Laboratory, University of California
Received for publication, January 8, 1918
It is usually assumed that the fluctuations of the protein
quotient under normal and pathological conditions are due
directly or indirectly to variations in the general nature and
especially in the rapidity of the metabolic processes.
Thus, Cervello (1) ascribes to a retardation of metabolism
the rise in globulins he obtained under the administration of
antipyrin. Hurwitz and Meyer (2) suppose that the increase
in the serum globulins in certain actue infections is due to meta-
bolic disturbances. Schmidt and Schmidt (8) are inclined to
believe that the serum proteins have no relation to immunity
and would rather attribute the increase in globulins to changes
in the metabolism resulting from the toxemia and tissue waste
of acute infections. |
In view of the fact that alterations in the protein quotient
have been so frequently attributed by various authors to dis-
turbances in the metabolism, it was considered necessary to
investigate the more precise relations of the globulin albumin
ratio to difference in the rapidity of metabolism. From this
standpoint it was shown in a previous communication (4) that
the protein quotient of blood serum is not changed by the admin-
istration of certain drugs which are known to retard or accelerate
the metabolic processes. The present work is a continuation of
the former experiments, and its object is to determine the influ-
ence on the protein quotient of a disturbance of metabolism
when such a disturbance is produced by periods of digestion
alternating with prolonged periods of starvation.
67
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2
68 SAMUEL HANSON
Although many determinations of the serum proteins during
starvation and digestion have been reported by various inves-
tigators; yet, the inadequate methods for estimating the serum
proteins, which were then available, have so precluded accurate
and thorough work, that a wide disparity in the results obtained
by different workers is rather the rule than the exception. The
chief disadvantage in the older methods is that large quantities
of blood are required for a determination. ‘This made frequent
estimations in the smaller animals impossible, and as a result
the normal variation was not known, and the intermediate
effects remained undertermined.
To illustrate the marked lack of uniformity in the findings of
different investigators, the following reports have been taken
at random from the literature.
Salvioli (5) found that the quantity of globulins in the blood
serum of two dogs starved for twenty to twenty-four hours is
less than in two other dogs during a period of digestion. This
change he attributed to individual variations, since he found no
difference in the globulins of the same dog during fasting and
digestion.
Burckhard (6), however, obtained in the serum of fasting dogs
a marked increase in the globulins and a considerable decrease
‘in the albumins: He explained these findings on the hypothesis
that during starvation the globulins pass from the organs into
the blood in order to cover the deficiency in the albumins.
Lewinski (7) reported only a slight increase in the serum
globulins of dogs after prolonged periods of starvation.
Briggs (8) obtained a considerable decrease in the globulins
of certain species of birds starved for a period of at least twenty-
four hours. This work faces the criticism, however, in that the
normal values were ascertained for one group of animals and
the values in starvation for another; hence the comparison of
the latter data with the former is not fully justified.
METHOD AND MATERIALS
The experimental conditions were very much the same as
those of the previous work (4). Robertson’s micro-refracto-
ee
THE CONSTANCY OF THE PROTEIN QUOTIENT
TABLE 1
Rabbit 1. Weight: June 8, 3332 grams; June 21, 2023 grams
69
k GLOBU-| 320-
DATE TREATMENT aes sae rea aay aonar: seed
: TEIN TEIN ae TIENT
per cent|per cent|per cent|per cent\per cent
LUTE (Cae ere ae Normal LEG een ete nos510 2080) 0625
DUES a site oss 205s, » Normal LS) Peon oe ee OMnle4.8 |) 210) 20L26
WVisivaelOQW is oe sass sss Normal ISOS PSe Shellie eee OF F230 R29
1G A Normal 163) 5 4735 ete 4-4) 2304) 0382
LWiGl7 12 a. ae Normal 1.55) SS Shiba sO). | 24.0) 0.32
IVER Semis. 2elsi> «ss « Normal 155 | eon ellen one | 2O0LOn | On26
Vietnygail Aaa ie eiete sec) Normal ay | 22550). |) Uoth || eal WP PaO) O77
ECG Normal LON 4 oe ele Omn Ronen | OeOn FOn24
June 8 Normal 15505220 Ossie eGe Orel 13.0) 0216
Jean 1) Sea ear Starved, 1 day 1.4 oma OR Gate 11620) 0220
| 0 Starved, 2days | 1.4| 6.0/0.7 | 6.7 | 10.5 | 0.10
IME AO shee ee oS. Fed, 2 hours 1:6) 4:35) 1A bee | 24:0) 0:33
CITY ee ee Starved, 1 day 1 On| oe OR lez Geen 1920) O24
ume wl Ane ys eyes. Starved, didays | 1.4 | 4.97/35 la652) . || 21:0) 10.27
JUL oe Ce Fed, 43 hours 1.4) 4:9) Tet 76.0) || 18.0) | 0:22
4G (Or ee Starved, 2days | 1725) Soi MAO 625 slel7 On Ons
MERCY OO se tao S oot Starved, 5 days |> 1:4 | S25yiL2o6.6 17.0) 0820
ume e2 0 sey Pe yarns ae Fed, 33 hours 12251, 15)5OR | Te hOe teen SO) One
AEE B AT ee a tat Normal 140) 4s 5s O85 5.2504 On Ona
LUG 2-355 Ab enone bore Normal IAP “4 eSa Olden |S). One el 42 ON ORIG
TABLE 2
Rabbit 2. Weight: June 8, 2261 grams; June 21, 1666 grams
ate GLOBU-| p20-
DATE TREATMENT ee gael me eas Sante penal
TEIN TEIN ae TIENT
per cent|per cent|per cent|per cent|per cent
REC RY, 8 eB Icis.o2. 516 b> Normal 13s ih 2365 | SIRO 6: 20 e262 OulOeS>
dunes Ue Ae ae eee Starved, 1 day 1 Aa pol a5 9 (6.65 |) 23200 |t0e29
JUTE Ce eae aa a Starved, 2days | 1:33) 520) })230) | 7.0) || 285010840
Meter OMe. ota... fesse Fed, 2 hours 4) |e el 2EOR | eSe On | son On mOnar
dors) 11275 ee ao ee Starved, 1 day 1.4) )) soe Te S1e7 |) 68) 2520") 10233
PUNE EASE 5. Ba nis Starved, 3days | 1.2] 5.6|1.8 | 7.4 | 24.0] 0.32
itive 7: i Fed, 4 hours 1.3) S245 154 7628 | -20755 0526
(CT UP ee Starved, 2:days | We25) soe25 (eT) 6.9) 25-05" 0633
“JT Xs) 24 Stamved.,o days | 158 |) 4x7 eran | 6. 192370 W0r380
AUTTS VDE Ie aes ee Fed, 3 hours 1.3 MOI) IEA 6.1 23.0 | 0.22
AGS 21 Ree Normal 10s 42 4a RA 58.0) 24-0. 0832
divi DE} eRe ee Normal heal Aioed |), I Ne ecstsy || eal ay || (027/
70 SAMUEL HANSON
“metric method (9) was employed for the estimation of the
serum proteins. Determinations were made in all cases through
a fore-period of several days, to serve as a safe standard for
comparison with the values to be obtained during the subse-
quent period of starvation and digestion. For the analysis four
cubic centimeters of fresh blood drawn from a manera ear vein
was employed.
_Rabbits were selected as the experimental animals.
© During the fore-period the rabbits were fed once daily with a
moderate quantity of alfalfa hay and crushed barley. Suffi-
TABLE 3
BE!
a NON TOTAL Sane wee
: DATE TREATMENT ices coe ean ae peice avo.
is . TEIN NU
Ve..( per cent|per cent|per cent|per cent\per cent
May O 7.0.24 6..000.4 Normal TRA 42 ee Oe 2) o) LOROMAOR24
May O60 840 200: Normal 12" |. 4.2 100.0 15.2. ||. 19502 Oaes
May A0i82 94. ist Normal 1.2))) 4.4 | 1.0.) 5.4--| 18:5-)0%23
May 2 5...55204 Normal 152!) 74.4.)'91.1_)-5.5--| 1925 Oat
Moly 42 Seis. ses. Normal 1.5 |, 4.1} °1.0|-5.1- | 19.5 |. 0024
Mays 2285 3... Urt.. Normal 1.3.) 45:10 1.11) .5.6- 4) 20:0) 0e4
May-l4- eo ees: Normal 1.24)—4.2 | -1.051-5-2—|-19-0-| 0824
May 1G tec. aad - Normal 13.) 4.6 |) 0.9. |: 5x55") 62010520
IMs MES eared aceon Normal 1:4.) 3.9) 1.4.) 5.34) 26.0 |)0235
1A) I i aL RSet cae Normal 13 | 44 Oost eal SOR oet
VMsiy DUM RO eaten 52 Normal 1.3.| 4:4) 1:5 | 5.95 | 25.0 | 0.34
May 24 x: sinasers, a cetwis Normal 151A eb Slee 2a onder ele ON ROE
Mia t29)3 Scns shins oe iie Normal 1.3:| 4:55) 0 Ls2elcoads 220-0 0s 26m
June Ssa.c8. cade: Normal 123.|- 4.6.) 1.0, |-5s7—}-20-0-|50224
AWS aah SaaS or Starved, 1 day 1227 VASA Ta oe8 | ZOO ROE23
JumewlO ste... cee Starved, 2days | 1.2] 5.0} 1.0] 6.0 | 17.0} 0.20
JUBCUONM Feiss Fed, 2 hours 08'| 40° |e 1) 5-1 - |. 21.0. 0227
Juma Geek. sess Starved, 1 day Ma Ale e122 1} O29 - +) 20. 020326
Jide ML Mays ste: Starved, 3idays |) 1.3 * 5.2 91-11) 6.3. | £725-)/0228
JuMeULa Me... See: Fed, 33 hours WAN ALTO | 528 > 2020) 0223
BS Co ly GA ee Starved, 2days| 1.2] 5.0| 1.4] 6.4 | 22.0 | 0.28
Jthe 20...05.........| Starved, 5days |) 1.37/50 |i 1596-1 | 18:0) 0x22
Jee v20.4 sCo. coe sBat Fed, 2} hours 163} |~ 4564/5108) -5 26. | 1820-) O222
JUMeyAT esse geet ote Normal 1:2') 420 17039 \-5-0--| 18:0 | 0822
JURE 23 .\...060.2520..| ‘Normal 125 4 ae ee Pe 9522 |) 210M ORT
Rabbit 3. Weight: June 8, 3332 grams; June 21, 2380 grams
THE CONSTANCY OF THE PROTEIN QUOTIENT
TABLE 4
Rabbit 4. Weight: June 8, 4165 grams; June 21, 3570 grams
71
GLOBU-|
on | PRO-
DATE TREATMENT aoe ene Bea | Beet we per
TEIN a TEIN eae: ecbars
per cent|per cent)per cent|per cent|\per cent
Aprlldee sss. 2s. Normal Ld) eae 2; |.620.), 20.2) Or25
Apron 6.0.6... Normal 14a) oral 20. 166,)||. 18.9! (OF22
AprtGreke |. .22.). 5. Normal (EAD OP Zaeel eon iG | «22a OFZ
Li he ee Normal Lebnioxaa at 2.1.66.) 18i9h 0522
BTN QOS 6c. sss’. s. Normal Wd Roe Ped 24 41!65.5.4),..,.22-S0 OL2e
aune 1Ss-.2)......-+...) Normal LSE POBZa esl aluGh sil. 22h OLZ9
PMFC TT On tentet 3. 5b 5 ):. Starved, 1 day IEPA 25541 (POR ROLOW|aezone hOLoO
me NO ease. e's 3. 2: Starved, 2-days| 1.2] 5.5| 1.4] 6.9.) 20 | 0.25
2 OS Oe _ Fed, 2 hours LAC). COMO Ne iGe 1 +! < L6-O Or20
Uwe ee ane ee ee Starved, 1 day 1eSHR4E Sale Gr enO24, || 25,0. O38
RUIN WA ras bev eae. starved, 3:days'| 1:1 | ‘4:80 1.57]. .6.3.|. 25-1)! 0:32
ume, AP ee Fed, 3 hours 123°) T4530 le GaeGe4 6.25.0 Oxss
JUG I Ses as eaeCae Starved, 2days'| 1.1) 5.0} 1.5 -|- 6-5 | 23-0.30
JUINS MAAR Reo ce ee Starved, 6 days LO) ASS ee ON Gaal Zen Ok40
UIE BZ: arama ee Fed, 4 days IL PAI Gye Sal G6. 9alaeeo | 0235
JUIN, 7B sememcana eae Normal 12 | 42 eee oe Onl 2940840
; TABLE 5
Rabbit 5. Weight: June 8, 3094 grams; June 21, not weighed
agg ERO PRO-
DATE TREATMENT ane er bes eae DOL ae!
TEIN TEIN ee TIENT
per cent\per cent|per cent|per cent|per cent
gegotal U2 oe t.. zee Normal V4 eo O eleSe ol) ond 33 | 0.50
Pept eAe ies os. . Normal EY al ess ed er? eT | 36 | 0.56
POT Ufc jaan. ss Normal 1.4 | 4.3 2) bk 9} |. 6:2 31 | 0.44
/Nr over oe See Normal 1.4 | 42 eS) 6.0 30 | 0.42
AJ TiySs (Se ee Normal Dan leeean ORO 32 3) O47
‘huis eo he Starved, 1 day 1 aFOSSue eer |) (6 30 | 0.43
dane 10. ce. 2. 4:545..|: Starved, 2.days'|. 1:2) 4.8) |.2-2) | 7:0 31 | 0.46
NS Os eek «agra. - Fed, 2 hours 19) AS Sbelo2e28 wile eOdale Sleny OL46
ANT e De en ee Starved, 1 day Ls2\ 4.9 22k | val 35 | 0.47
Jno ee ae ee ae Starved, 3 days Sample lost
AJiiinven 11 Ee ee ee Fed, 22 hours 1342 Gono) i fol 385 | 0.54
NO fh ores As Sarena « Starved, 2idays |' 1:24 4.7..|.2.55 | 7.25 | 37 -| 0.54
June dA9. ee.) .c5.5. 6% Starved, 4 days Died
fe SAMUEL HANSON
TABLE 6
Rabbit 6. Weight: June 8, 2618 grams; June 21, 2023 grams
daa RO-
DATE TREATMENT aes eee onal SOEOr, one ea
TEIN TEIN eae TIENT
per cent|per cent|per cent\per cent|per cent
JUTE! ISURAe seth ss : Normal 12 eae oboe 9 > 1) 7-4 Sie OnOsos
JUN eee Nek. . 3. Starved, 1 day ese oe2enl eel 7.3 | -29s0nnOEeG
ue wl OMRRe saan oe. Starved, 2 days | 1.2| 4.9 | 2:2 | 7:1 | 31.0-) 045
mel ON ets exc: Fed, 2 hours 159) 421 42223. 19624. “| 36408 BORDG
Vibna yes Ae Pees Se ee Starved, 1 day 1.3) | 425. 10224 1.6.9 | 3520870253
ol jieun sl 2: Uy ae oa Starved, 3days | 1.3) 4.8 | 2.55 | 7.35 | 35.0 | 0.538
merase eee oy. ©... Fed, 23 hours ds osOm 222. | 7.2 | 30.59) One
JUVE ets elackes <0: Starved, 2days | 1.2 | 5.0 | 2.05 | 7.05 | 29.0 | 0.41
Tue: ZOE es noes a Starved, 5days| 1.2) 5.0''| 2:2 | 7.2 | 30.5) 0:44
Jumen20 See ane ie ..3- Fed, 2 hours 122A SEO 222) aie? N30. bn | ORs!
ditine Wle. aesnnon ane Normal 15529) Axoonle2e2e | O-oDNlooeom Oso
UM CR23 eee thsstic Normal 1.3.) 4.2 19) WOR {| 3IR OF ORAS
TABLE 7
Rabbit 7. Weight: June 8, 3094 grams; June 21, 2618 grams
; GLOBU-| p25.
DATE TREATMENT aoe fered eee rae Toni, per
TEIN TEIN cent TIENT
per cent per cent|\per cent|per cent\per cent
Mia eile cis tare pice Normal 123.) 4.284) ©29% || Gate |) 2830580839
Misty: 134.8: bck ee ss ‘Normal Be AZo 20) Gat {eZee ON eae
Mutya Occ pikecprrcac Normal 1.4) 4.8} 2.0 | 6.8 | 29.0) 0.41
Ming HONS bes acct. Normal 15) “26a Toe eo ie 29R Om sOks
11 in ga a ar a Normal 15 | 4:9) Love e526) 1 S0k0sOeS4:
IMlaiy LD: mehr tes ote ae Normal 125). 4269) 5 56st) 29255 t0rSs
11) Eid ee rege Normal 1-4. |) 4207) 208915680" 1330980250
Maye la mene Normal 14°) P4577) WGI) 2523" 1) S008 80243
IMiavalG see cee Normal VA | Ag8e) Wea G6 25" 4 2705 tO eso
UN Guerre ben. cere Normal 132!) *5O 54 Dede IA ebe 42820 sORs
ne Oe Ss es ees Starved, 1 day M3 Osae2. LZ 25) | 280 Oess
“TT a ee ei a Starved, 2days| 1.3] 5.2|2.4 | 7.6 | 32.0 | 0.46
PUNE MOL es eect Fed, 2 hours 14] 455 122741629) | (3520) (0835
jini PAP ae se Seneione Starved, 1 day 1.3] 5.6] 2.75 | 8.35 | 33.5 | 0.49
dl UboVeh 3 BAe noeee ears Starved, 3 days | 1.1] 5.2) 2:5-1 7:7 | 32.0 | 0-48
jumevla ee cance t eer Fed, 13 hours QM NOeSe| oO! Weve: 932 20m| Oo!
AJ Shee Geers ae eee Starved, 2 days| 1.0| 5.0] 2.5 | 7.5 | 33.0] 0.50
Jume 2h. o.20.5. Starved,6days| 1.0] 4.8/2.8 | 7.6 | 37.0 | 0.51
MAMIE 20 Pecos ears otoes: Fed, ? hour 120) || 48 °8288 187.6" *| 37-0 |) Oza
THE CONSTANCY OF THE PROTEIN QUOTIENT 73
cient fresh water was always in the cages during the fore-period
and during the period. At the beginning of the period, all the
straw bedding and other possible food was removed from the
cages. Where the interval of starvation exceeded two days,
determinations of the serum proteins were made to show the
intermediate effect of the fasting. At the termination of the
starvation interval, blood was drawn for determinations. The
animals were then given an excess of alfalfa hay, crushed barley,
cabbage, and carrots. This. interval of feeding was allowed to
vary from three-quarters of an hour to four hours and a half,
in order that several gradations of the intensity of digestion may
be included. All the food was removed again, at the end of
this feeding interval, and blood was immediately once more
drawn for determinations. A longer interval of fasting then
followed, and so on.
DISCUSSION
From a perusal of the tables presented, it will be seen that
the protein quotient remains normal during digestion periods
alternating with prolonged starvation periods.
The negative data for digestion are fully in accord with the
findings of Rowe (10) who showed that high protein diets have
no immediate effect on the globulin-albumin ratio.
It will be recalled that the retardation of the nitrogenous
metabolism as shown by the diminution in the elimination of
nitrogen during starvation (11), is due to the intrinsic change,
which consists of a gradual decrease in the exogenous metab-
olism, accompanied but not equalled by an acceleration in the
endogenous metabolism. Viewed in the light of these profound
metabolic disorders, the constancy of the protein quotient during
prolonged starvation is not an easy matter to explain.
An analogous phenomenon is the constant and normal per-
centage of glucose in the blood even during an extended period
of fasting (12). It is assumed that this constancy of composition
is maintained chiefly by the action of enzymes elaborated in
the liver cells, which convert dextrose into glycogen, or glycogen
74
SAMUEL HANSON
into dextrose, depending respectively on whether the glucose
concentration in the blood is in the upper or lower limits of
the normal variation.
Is it not then also probable that a ‘similar mechanism serves
to adjust the constancy of the protein quotient?
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
REFERENCES
CERVELLO, C.: Arch f. exper. Path. u. Pharm., 1910, 42, 357.
CrRVELLO, C.: Arch. f. exper. Path. u. Pharm., 1911, 64, 403.
Hurwitz, S. H. anp K. F. Meyer: Jour. Exper. Med., 1916, 24, 515.
Scumipt, E. S. anp Cart L. A. Scumipt: Jour. of Immunology, 1917, 2, 343.
Hanson, S. ano I. McQuarrie: Jour. of Pharm. and Exper. Therap., 1917,
10, 261.
Satviou1, G.: Arch. f. Physiol., 1881, S. 269.
Burcxuarp, A. E.: Arch. f. Exp. Pathol. u. Therap., 1883, 16, 322.
LewinskI, J.: Pfluger’s Arch. f. d. gesammte Physiol., 1903, 100, 611.
Briees, R. S.: Jour. Biol. Chem., 1915, 20, 7.
Rosertson, T. BrRArLsrorpD: Jour. Biol. Chem., 1915, 22, 233.
Rowe, A. H.: Arch. of internal Med., 1917, 19, 499.
Voit, E.: Hermann’s Handbuch, 1881, 6, 1, p. 89.
TIGERSTEDT, Rosert: Lehrbuch der Physiologie des Menschen, 3 Aufl.,
Bd. 1, p. 111, Leipsic, 1905.
(12) Howetu, W. H.: Text-Book of Physiology. 4th edition, pp. 888-889.
THE IMMUNOLOGIC PROPERTIES OF UVEAL
PIGMENT
ALAN C. WOODS
From the John Herr Musser Department of Research Medicine, University of
Pennsylvania
Received for publication January 12, 1918
In the course of a detailed study of the anaphylactic theory
of sympathetic ophthalmia, a study has been made of the anti-
genic properties of the uveal tract, especially the uveal pigment,
of the eye. From the standpoint of the ophthalmologist the
results are of little import beyond their possible correlation with
clinical disease. From the standpoint of the immunologist,
however, the results appear to be of more scientific interest, in
so far as they indicate rather striking immunologic properties
for a native body protein. It is, therefore, the purpose of this
paper to report the results of this phase of our studies.
The anaphylactic theory of sympathetic ophthalmia, se
and advocated by Elschnig, assumes that sympathetic ophthal-
mia is an anaphylactic uveitis brought about in the follow-
ing manner. The injury to the uvea in the exciting eye, by
trauma, intraocular tumor, etc., leads to a destruction or disin-
tegration of uveal tissue. This uveal tissue is absorbed and
acts as an antigen, producing a hypersensitiveness of the organ-
ism and especially of the other eye. A reaction now takes place
between the sensitized cells of the uvea of the second eye and
the antigen circulating in the blood or lymph. This anaphy-
Jaetic reaction or intoxication is manifested clinically as a sym-
pathetic ophthalmia.
This: theory, of course, assumes that the cells of the uveal
tract, or some constituent of these cells, possess antigenic prop-
érties in the homologous animal, and further assumes organ
75
76 ALAN C. WOODS
specificity and lack of species specificity for the protein
involved.
Elschnig fully realized these assumptions and in his presenta-
tion of the theory in its present form, presented experimental
work to substantiate these assumed points.
HISTORICAL
Elschnig’s work (1), appearing in a series of papers, may be
summarized as follows. Using as his methods the intraocular
injection of sheep erythrocytes and cholera vibrio extracts, and
as his indices the hemolytic titer and agglutination reactions
of the blood serum, he established the fact that absorbtion from
the eye could lead to immune body production. He then sought
to determine three points, to give scientific support to the ana-
phylactic theory. These points were: (I) Does uveal tissue
possess antigenic properties? If so, in the homologous animal?
(Il) What constituent of the uveal tract is responsible for such
properties? (III) What are the antigenic properties as regards
organ and species specificity? Elschnig immunized rabbits
against various heterologous uveae, and against homologous
uvea, and finally against so-called ‘‘chemically pure’ pigment
from the uveal tract of various animals. Thus he obtained
hetero-immune and iso-immune uvea serum, and pigment-
immune sera. He then used the complement fixation reaction
for the detection of immune bodies in these sera and the study
of their properties. He found that heterologous uvea produced
immune bodies, as would be expected. Homologous uvea, also,
acted as antigen and produced immune bodies. The sera of
animals immunized to uvea fixed complement with any uvea
antigen, irrespective of species—the immune bodies were organ
specific. As regards species specificity, the sera of animals
immunized to an emulsion of whole uvea were species specific
in their reactions. The sera of animals immunized to pigment,
however, were not species specific. The reason for this apparent
anomaly is evident. Emulsion of whole uvea contains two
elements, (1) pigment which is not species specific and (2) such
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT a
tissue as blood, smooth muscle and connective tissue which is
species specific. Immunization with heterologous uveal emul-
sion gives an immunity not only to non-species specific pigment,
but also to the other species specific elements in the uveal tract.
The species specificity of uvea immune sera is due to these
last elements.
Elschnig therefore concluded that uveal tissue could act as
antigen in the same animal, and that the pigment was the
constituent responsible for this property. The pigment was
organ specific and not species specific.
The remaining work on this subject may be quickly sum-
marized, Weichardt and Kummel (2), using the doubtfully
valuable epiphanin reaction, substantiated Elschnig’s findings.
Likewise, with the epiphanin reaction, and complement fixation,
Kummel (3), demonstrated uveal antibodies in a percentage of
the sera of patients with sympathetic ophthalmia. Wissman
(4), supported the anaphylactic theory by experimental work
and showed uvea immune bodies in the serum of two sympa-
thetic ophthalmia patients by the precipitin reaction, but failed
to substantiate Kummel’s work with complement fixation, on
the sera of these patients. Rados (5), substantiated Elschnig’s
work only partially. Fuch and Meller (6), were unable to
demonstrate uveal antibodies in the sera of patients with sym-
pathetic ophthalmia. Von Szily (7), after lending much valu-
able support to the theory criticizes the antigenic properties of
pigment.
OUTLINE OF WORK
In our study of the various phases of the anaphylactic theory
of sympathetic ophthalmia, we have studied primarily the
immunologic properties of uveal tissue, and especially of uveal
pigment. This has been done in two ways, first by a repetition
of Elschnig’s work with the complement fixation reaction of the
sera of immunized animals, and secondly this has been con-
firmed by direct perfusion experiments upon the eyes of sensi-
tized animals.
78 ALAN C. WOODS
I. THE ANTIGENIC PROPERTIES OF UVEAL TISSUE AS SHOWN BY
COMPLEMENT FIXATION
This, as before mentioned, is substantially a repetition, in
toto, of Elschnig’s work.
Dogs were immunized, by intraperitoneal injections repeated
every six days, against cow’s uveal emulsion, cow’s uveal pig-
ment and dog’s uveal emulsion and uveal pigment.! In order
to avoid non-specific fixation, the sera of the dogs selected for
this work were all examined, in the complement fixation reac-
tion, with the various uveal emulsion and uveal pigment anti-
gens, before immunization was begun. Eight dogs giving
completely negative reactions with all antigens in this prelimi-
nary work, were selected for immunization. To immunize,
intraperitoneal injections of uveal emulsion and uveal pigment
were given to respective dogs at six day intervals. The initial
injection was 5 ce. which was increased 1 ec. at each injection
until 10 ec. was reached. A total of eight immunizing injections
was given each animal.
At suitable intervals after the last injection, the sera of
these dogs was tested again against all the antigens. The results
are shown in the tables 1 to 4. In order to quantitate the
strength of the reaction, three different quantities of sera were
used with every antigen. A +-+-+ reaction indicates complete,
er practically complete, fixation of complement with all quanti-
ties of sera, a ++ reaction similar fixation of complement in the
tubes containing two largest amounts of sera, and a single +
reaction indicates that such fixation was observed only in the
tube containing the maximum amount of serum.
The first two complement fixation reactions were done with
the same antigens used in the preliminary reaction. The last
reaction was done with freshly prepared antigens. All antigens
were used in one-third the anticomplementary dose.
All animals showed immune sera, differing, as might be ex-
pected, in the various dogs. Dogs 16-96 and 16-98 showed
the lowest immunity.
i For the details of technique used to prepare the uveal emulsion and uveal
pigment consult references (8) and (12).
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 79
Table 1 illustrates the complement binding phenomena shown
by heterologous immune sera. As would be expected these sera
gave complement binding with their specific antigens. More-
over, the uvea-immune sera gave fixation of complement with
pigment antigens. Pigment immune sera fixed complement
with pigment antigens, showing here that the pigment itself is
capable of acting as antigen.
TABLE 1
Heterologous sensitization (cow’s uvea and pigment)
FEBRU-
ea) _|ARY 21,
SERA ANTIGEN DECEMBER 14 ea nee aa FRESH
3 ANTI-
- GENS
(| Cow’s uvea | Negative el ele
16-89. Uvea immune... /
++) $/44+4+
_| Cow’s pigment | Negative
+++\|4++)/+++
+) +4+]+++
Cow’s uvea Negative
16-90. U 1 aa eel : é :
6 20. Pires mamune Cow’s pigment | Negative
+++] +4]4+4++
++| ++/+++
Cow’s uvea Negative
Cow’s pigment | Negative
Period of immunization
16-91. Pigment immune
BMG cit i {| Cow’s uvea Negative sates (ines eel Peace
a a \| Cow’s pigment | Negative +++) +4]/4+++
Table 2 illustrates the complement binding reactions of the
sera of dogs immunized to dog’s uveal emulsion and dog’s pig-
ment—iso-immune sera. The iso-immune sera gave fixation of
complement with their specific antigens, showing the presence
of iso-immune bodies. In other words, uveal tissue possesses
the power to act as a foreign protein—as antigen—in animals
of the same species. This is the first, and cardinal point, to be
proven in order to establish the anaphylactic theory. Further-
more, the purified pigment of the uvea possesses this power of
immune body production in the homologous animal, and uvea
immune sera gives fixation of complement with pigment anti-
gens. This substantiates Elschnig’s contention that it is the
pigment which is the responsible factor for the peculiar antigenic
properties of uveal tissue.
80
ALAN C. WOODS
TABLE 2
Homologous immunization (dog’s uvea and pigment)
SERA
ANTIGEN DECEMBER 14
16-95. Uvea immune.
16-96. Uveaimmune.
16-98. Pigment im-
FINN G5 hee evi syns ase
16-99. Pigment im-
Dog’s uvea Negative
Dog’s pigment | Negative
Dog’s uvea Negative
Dog’s pigment | Negative
Dog’s uvea Negative
Dog’s pigment | Negative
Dog’s uvea Negative
Dog’s pigment | Negative
SERA
16-89. Cow’s uvea
LINIMNUNES. -o eee
16-90. Cow’s uvea
IMMUNE S - eee
16-91. Cow’s pigment
(MMUNE.22- a. eee it
16-92. Cow’s pigment
ATTN Cone eee eae
16-95. Dog’s uvea
IMMUNE. oss ge see
16-96. Dog’s uvea
ATMAMUING ea ee
16-98. Dog’s pigment {
ATMUITUUN G2 5,28 eee
16-99. Dog’s pigment
IMTS 2.5 = ho creree
if
\| Dog’s pigment
TABLE 3
Organ specificity
ANTIGEN DECEMBER 14
Negative
Negative
Dog’s uvea
Negative
Negative
Dog’s uvea
Dog’s pigment
Negative
Negative
Dog’s uvea
Dog’s pigment
Negative
Negative
Dog’s uvea
Dog’s pigment
Cow’s uvea
Cow’s pigment
Negative
Negative
Negative
Negative
Cow’s uvea
Cow’s pigment
Negative
Negative
Cow’s uvea
Cow’s pigment
Cow’s uvea
Cow’s pigment
Negative
Negative
Period of immunization
Period of immunization
ae
ae iek FEBRUARY 14 Pres
GENS
Soa Bist el (i aes ie =
i ios (LS er
++] eae
+] Negative| ++
aE 3 +
aE ate ++
+] ° a een
tale. aia =
JAN- FEBRU-
teed FEBRUARY 14 pete
ARY 6 a
+) eae
a ca ++
Sisk + 4
+} Negative +
++) Negative|. ++
+| Negative} ++
seme
|, sh aie
“Ee SE ee
SE | GEN eaatogee
shF+] +a) | Ice
saan Neca |r
aber west aie
seapaet Gren |E=55 =
feasts | ais ois Seas
SESRS5i| Sea eee
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT Sl
Table 3 illustrates the organ specificity of the immune sera.
Iso-immune sera gave positive complement binding reactions with
heterologous antigens, and the heterologous immune sera gave
complement binding with the iso-antigens. In other words the
immune bodies were organ-specific, showing an affinity for anti-
gens of the specific tissue, without regard to the species from
which obtained. This substantiates Elschnig’s second point of
the organ specificity of uveal tissue. The pigment-immune
sera show this property also and it is evidently to the pigment
constituent that this organ specificity of uveal tissue is due
TABLE 4
Species specificity
l
JANUARY 31 eae
SERA ANTIGEN AND FEBRUARY 14 aaSCie
FEBRUARY 6 ANTIGEN
5
16-89. Cow’s uvea im- {| Cow’s liver = Saas + ++
(20.6 > Sea || Cow’s kidney " apse See Sear
16-90. Cow’s uvea im- Cow’s liver +++ +++ +++
MAUITENS «SPAM ALN Satins fie! de’ « Cow’s kidney| (= | +++ +++ +++
©)
16-91.Cow’s pigment im- Cow’s liver S Negative | Negative ++
PAE eee tl. . Cow’s kidney 2 + | Negative +
16-92. Cow’s pigment im- if Cow’s liver + + | Negative
PAINE hy EIN Ses. 2s || Cow’s kidney + “ob +4
Table 4 illustrates the question of species specificity. With
the sera of animals immunized to cow’s uveal emulsion and
cow’s uveal pigment, it was determined whether or not fixation
of complement occurred with antigens of cow’s protein, other
than uveal tissue. It will be seen by a study of this table that
the uvea-immune sera give complete fixation of complement with
antigens of cow’s liver and kidney extract. The reason for this
is that already stated—that these dogs were immunized to the
whole uvea, which contained not only the uveal pigment, but
also some blood, smooth muscle and connective tissue of the
cow. The fixation of complement with the liver and kidney
82 ALAN C. WOODS
extracts is evidently dependent upon the immunization to this
last factor.
On the other hand, the sera of dogs immunized to cow’s pig-
ment alone gave either weak or negative reactions with cow’s
liver and kidney. The pigment was purified as much as possible,
but it probably still contained traces of the other elements of
the uveal tract—the blood, smooth muscle and connective
tissue. The weak reactions occasionally observed are probably
due to this impurity.
However, the difference in the degree of fixation of comple-
ment between the uvea immune and the pigment immune sera
with the same antigens, is striking. Although in this phase of
the work, our results are not so clear cut as those of Elschnig, it
seems probable that the pigment acting as antigen, lacks at
least in a degree, species specificity.
These observations of the complement binding reactions of
uvea and pigment immune sera corroborate Elschnig’s findings.
Heterologous and homologous uveal tissues have the power of
acting as antigens. In the case of homologous uveal tissue, the
pigment is the factor responsible for its antigenic properties in
animals of the same species. ;
Whole uveal emulsion is both organ specific and species specific,
the species specificity probably being due to the blood, smooth
muscle, and connective tissue in the emulsion. The pigment,
however, is organ specific and probably not species specific,
and in this respect is analogous to lens protein and differs from
other common body protein.
These properties of uveal pigment—ability to act as foreign
protein to animals of the same species, organ specificity, and
lack of species specificity—are as before emphasized, funda-
mental properties uveal tissue must needs possess to make the
anaphylactic theory a possibility.
_— niet
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 83
Il THE ANTIGENIC PROPERTIES OF UVEAL TISSUE AS SHOWN BY
PERFUSION OF THE EYES OF SENSITIZED ANIMALS
a. The reaction of the sensitized eye to perfusion with specific
antigen
With the confirmation of Elschnig’s fundamental work we
next sought, both for the purpose of again confirming this work,
and of more closely correlating these findings to the clinical
disease, to determine whether these peculiar antigenic properties
indicated by complement fixation, could be manifested in a
direct anaphylactic reaction upon the eye, through general
sensitization and vascular intoxication. To this end, we resorted
to the perfusion of the eye. Figure 1, shows diagramatically
the operation employed and the position of the inflow and out-
flow cannulas. The apparatus employed has already been
reported (8). Fresh defibrinated dogs blood to which sufficient
Ringer’s solution was added to bring the red corpuscle content
approximately to normal, was the perfusion fluid used.
With the technique used we were dealing with what was to
all purposes a living eye, maintained on an artificial circulation
with defibrinated blood, oxygenated by an artificial lung. AI-
though the dog’s heart always ceased to beat after the final
ligature was placed above the heart, nevertheless the winking
reflex persisted in the eyes often for an hour or more.
In this experiment the eyes were perfused for three hours,
constant observations being made throughout that period.
It was found that when a sensitized dog was perfused with
the defibrinated blood of a normal dog, there was no ocular
reaction. Wheh, however, the specific antigen (the antigen to
which the dog was sensitized) was added to the perfusion fluid,
a prompt contraction of the pupil occurred, and as the perfusion
continued, small hemorrhages appeared throughout the fundus.
The contraction of the pupil was marked, usually from a dilated
pupil 10 to 12 mm. in diameter to a pupil from 2 to 4 mm. in
diameter. This observation is in direct accord with those
of Dale (9) and Schultz (10) who observed the contraction of
. sensitized smooth muscle in the presence of specific antigen, and
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2
84 ALAN C. WOODS
is another observation in support of the cellular theory of ana-
phylaxis. The hemorrhages observed cannot be so easily
SpiensesQe E
o pattern’ meee es
é i
i g oe
<
> :
PA TBE EPEC PELL i 3
ta sae - - -
ml ue . Scat
; yee i,
Inflow
Cannula 4 FE
7; Brachio |
Cephalic
oe Ny,
~ :
~* Aorla
Cannula
Fia. 1
explained. Petechial hemorrhages have been observed over
the peritoneum of animals recovering from anaphylactic shock,
and it may be that these are analogous. From their method of
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 85
formation it appeared that they were caused by an alteration
of the endothelium of the capillaries, allowing a diapedesis of
red cells.
In this experiment, the antigens used were horse serum and
cows uveal emulsion. Before proceeding with the second phase
of the perfusion work a number of perfusions were done to con-
trol this observation. These controls are illustrated in table
5. One of these controls requires a little explanation. Normal
dogs perfused with horse serum showed a slight contraction of
the pupil, never over 3mm. Schultz (11) previously has shown
that fresh horse serum possesses the power to produce a slight
contraction of smooth muscle, and this is evidently what oc-
TABLE 5
Contrgls
RESULTS
SENSITIZATION PERFUSION FLUID =
Contraction of the HeMIOR
: rhages in
pupil fundi
None DEB None None
None D. B. + Uveal emulsion None None
None D. B. + Horse serum Never over 3 mm.| None
Uveal emulsion Dees None None
D. B. is defibrinated blood of normal dogs.
eurred with us. This contraction of the pupil in normal dogs,
perfused with horse serum, was never over 3 mm. while the con--
traction in the sensitized dogs was from 8 to 10 mm. Normal
dogs perfused with uveal tissue showed no contraction of the
pupil. In these and in all subsequent perfusion experiments
here reported, each individual observation was controlled by at
least two perfusions. In those experiments where for any reason
(clots, thrombosis) the nature of the reaction was not clear
after two perfusions had been performed, subsequent perfusions
were done until the presence or absence of the anaphylactic
phenomena was established beyond question.
From this it was evident, therefore, that the eyes may be
sensitized as a part of general sensitization, and that anaphy-
86 ALAN C. WOODS
lactic phenomena may be elicited in the eyes by means of antigen
carried by the blood stream. ‘These anaphylactic phenomena
consist in a marked contraction of the pupil, and in small extra-
vasations of blood throughout the fundus.
With the artificial condition, under which we were working,
it is manifestly impossible to expect inflammatory phenomena
of any kind, but the establishment of the fact that ocular ana-
phylaxis, however manifested, may be demonstrated by a gen-
eral sensitization and vascular intoxication, gives us the second
important point in the establishment of a scientific basis for the
anaphylactic theory. Moreover, this observation affords us a
means of studying the anaphylactic properties of uveal tissue,
in vivo, by direct observation of the eye.
b. The antigenic properties of vveal tissue in the production of the
perfusion anaphylactic reaction
Having shown that an anaphylactic reaction can be obtained
through the perfusion of the eye with specific antigen, this
reaction has been used to determine the antigenic properties of
uveal tissue. We first sought to determine whether whole
uveal tissue possessed the antigenic properties necessary to make
an anaphylactic uveitis a possibility. When this point was
established, we sought to determine the constituent of uveal
tissue responsible for its peculiar antigenic properties. Pre-
liminary experiments with uveal pigment showed that there
was much reason to believe, with Elschnig, that the pigment
was the responsible factor. A pigment solution was finally
prepared which was suitable for use in perfusion (12). Experi-
ments similar to those in which uveal emulsion was used as
antigen were then performed to determine whether the pigment
is the responsible factor. The reactions given by the pigment
were in every way similar to those given by whole uvea, except
as regards species specificity. Moreover, dogs sensitized to
homologous uvea, reacted to perfusion with pigment, establish-
ing more conclusively the fact that the pigment is responsible
for the peculiar antigenic properties shown by uveal tissue.
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 87
_ The results of this work are shown in the following tables
(6 to 9).
Table 6 represents the results obtained by sensitization and
perfusion with heterologous (cow’s) uveal emulsion and uveal
pigment. The anaphylactic reaction was observed in all of
these perfusions, establishing the power of both uveal emulsion
and pigment to act as antigens.
TABLE 6
Heterologous sensitization (cow’s wea and pigment)
RESULT
SENSITIZATION PERFUSION FLUID Contrac- Hemor-
tion of rhages in
the pupil fundi
Cow’s uveal emulsion D. B. + Cow’s uveal emulsion! Marked | Marked
Cow’s uveal pigment D. B. + Caw’s uveal pigment | Marked | Marked
D. B. is defibrinated blood of normal dogs.
Table 7 illustrates the fundamental fact necessary in the
anaphylactic theory—namely the power of homologous uveal
tissue to produce an ocular anaphylactic reaction in a sensitized
eye when carried there through vascular channels. It shows
also that it is the pigment which is responsible for this remark-
able antigenic property. Dogs sensitized to homologous uveal
emulsion give an anaphylactic reaction when perfused with
normal defibrinated blocd to which pigment is added, and dogs
sensitized to pigment give the reaction when perfused with the
pigment-containing blood.
TABLE 7
Homologous sensitization (dog’s uvea and pigment)
RESULT
SENSITIZATION PERFUSION FLUID Contrac- Hemor-
tion of rhages in
the pupil fundi
Dog’s uveal emulsion D. B. + Dog’s uveal emulsion| Marked | Marked
Dog’s uveal emulsion D. B. + Dog’s pigment Marked | Marked
Dog’s uveal pigment D. B. + Dog’s pigment Marked | Marked
D. B. is defibrinated blood of normal dogs.
88 ALAN C. WOODS
Table 8 illustrates the organ specific property of uveal emul-
sion and uveal pigment. As shown before by the complement
fixation reaction, uveal tissue and pigment are organ specific.
The sensitization resulting from the introduction of uveal tissue
into an animal, is specific for uveal tissue, without regard for
the species from which the tissue is taken. Dogs sensitized to
cow’s uveal emulsion react to perfusion with dog’s uveal emul-
sion, and vice versa. Similarly dogs sensitized to cow’s pig-
ment react to perfusion with dog’s pigment, and vice versa.
There evidently results from sensitization with uveal tissue, a
strong chemical affinity for similar tissue,—organ specificity.
TABLE 8
Organ snecificity
RESULT
SENSITIZATION PERFUSION FLUID Contrac- Hemor-
tion of rhages in
the pupil fundi
Cow’s uveal emulsion D. B. + Dog’s uveal emulsion| Marked | Marked
Cow’s uveal pigment D. B. + Dog’s pigment Marked | Marked
Dog’s uveal emulsion D. B. + Cow’s emulsion Marked | Marked
Dog’s uveal pigment D. B. + Cow’s pigment Marked | Marked
D. B. is defibrinated blood of normal dogs.
Table 9 shows the species specificity reaction of both uveal
emulsion and pigment. The same properties are shown here
as were indicated by complement fixation, the uveal emulsion
is species specific, while the pigment is not species specific.
Dogs sensitized to cow’s uveal emulsion give an anaphylactic
reaction when perfused with other cow protein, for here the
sensitization is not alone with the pigment, but also with the
other species specific protein contained in the whole uveal emul-
sion. On the other hand, dogs sensitized to the pigment alone
show no reaction when perfused with other cow protein.
The last perfusion illustrated in this table is largely a control
perfusion. Dogs sensitized to uveal emulsion give no reaction
when perfused with other dog protein. Perfusion with the other
elements of the uveal tract—blood, smooth muscle and connec-
tive tissue, evokes no anaphylactic reaction.
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 89
To sum up the work thus far, both the complement fixation
reactions of the immune sera, and the perfusion reactions, have
shown that uveal tissue possesses the power to act as antigen in
animals of the same species, and that the pigment is the con-
stituent of the uvea responsible for this property. In its anti-
genic action, uveal pigment is organ specific and not species
specific. The sensitization resulting from the absorption of
uveal pigment is specific. Other body protein can produce no
anaphylactic reaction in animals so sensitized. Pigment alone
may produce intoxication.
TABLE 9
Species specificity
RESULTS
SENSITIZATION PERFUSION FLUID eee eae
tion of | rhages in
the pupil fundi
Cow’s uveal emulsion | D. B. + Cow’s serum, liver and kidney} Marked] Marked
extracts
Cow’s uveal pigment | D. B. + Cow’s serum, liver and kid- | None | None
ney extracts
Dog’s uveal emulsion | D. B. + Dog’s liver, and kidney ex- | None | None
tracts
|
c. Ocular sensitization from antigen absorbed from the other eye
To complete our study of the antigenic properties of uveal
tissue as manifested in the ocular perfusion reaction, and as an
incidental point in our study of the anaphylactic theory, we
sought to determine whether uveal tissue, absorbed in one eye,
could create a hypersensitiveness to uveal tissue in the second
eye. To demonstrate this we sensitized dogs by the injection
of uveal tissue, both heterologous and homologous, into the
vitreous of one eye. After a suitable period had elapsed for sen-
sitization to occur, the injected eyes were enucleated, in order to
remove any factor which could give a possible intoxication.
One week after this, the remaining eye was perfused with the
specific antigen. In every case, the perfused eye gave an ana-
phylactic reaction, indicating that ocular sensitization had
90 ALAN C. WOODS
taken place as a result of the absorption of antigen from the
fellow eye. Table 10 illustrates this experiment.
TABLE 10
Ocular sensitization e
RESULTS
SENSITIZATION PERFUSION FLUID Contrac- Hemor-
tion of rhages in
| pupil fundi
Vitreous injection. | Enucleation left | D. B. + Cow’s | Marked | Marked
Left eye. Cow’s | eye uveal emulsion
uveal emulsion
Vitreous injection. | Enucleation left | D. B. + Dog’s | Marked | Marked
Left eye. Dog’s eye uveal emulsion
uveal emulsion
DISCUSSION
As is well known, Uhlenhuth (13) and Haendel (14) have
established the fact that animals can be sensitized to their own
lens protein, and Rosenau and Anderson (15) have found that
guinea-pigs can be sensitized to guinea-pig’s placenta. From
the evidence brought out by the researches of Elschnig and this
work, it seems that uveal pigment possesses these same anti-
genic properties. However, we realize that we have not demon-
strated the actual point of auto-sensitization, the sensitization
of a dog to the pigment from his own eye. The technical diffi-
culties in the way of such a proof are enormous. One attempt
to demonstrate such auto-sensitization by general anaphylactic
reactions has been unsuccessful. The eye of a dog was removed,
the uvea excised and macerated, and injected to sensitize. Two
weeks later the second eye was removed, the uvea excised and
macerated, and this again injected intravenously under ether
anesthesia, and this dog was observed for a fall in blood pres-
sure, change in the coaguability of the blood, and drop in body
temperature. No such signs were observed and we have little
hope of demonstrating auto-sensitization by this method. A
more delicate method must be devised. For the present we
must content ourselves with the demonstration that dogs uvea
can produce ocular anaphylactic phenomena in the dog.
IMMUNOLOGIC PROPERTIES OF UVEAL PIGMENT 91
CONCLUSIONS
The pigment of the uveal tract of the eye possesses the prop-
erties of acting as antigen in homologous animals, and in its
immunologic reactions is organ specific and not species specific.
These findings can be demonstrated by the complement fixation
reaction with the sera of properly immunized animals, and by
perfusion experiments on the eyes of sensitized animals. In
the case of the perfusion experiments, the anaphylactic reaction
is manifested by a marked contraction of the pupil, and the
occurrence of small hemorrhages in the fundus. This reaction
was used to study the antigenic properties of uveal pigment,
and the results shown by complement fixation confirmed.
REFERENCES
(1) Etscunte, A.: Studien zur Sympathischen Ophthalmia. JI. Wirkung von
Antigenen vom Augeninnem aus. Arch. f. Ophth., 1910, 75, 459.
II. Die antigene Wirkung des augenpigments. Arch. f. Ophth.,
1910, 76, 509. III. Teil Arch. f. Ophth., 78, 549. IV. (Elschnig,
A. and Salus, R.) Wirkung des Augenpigments, Arch. f. Ophth., 1911,
79, 428.
(2) Wetcuarpt, W., AND Kummet, R.: Studien itiber Organ-Spezifitit des
Uvaeiweisses. Miinch. Med. Wochenschr., 1911, 58, 1714.
(3) Kummet, R.: Versuche einer Serumreaktion der sympathischen Ophthal-
mie. Arch. f. Ophth., 1912, 81, 486.
(4) Wissan, R.: Ueber Versuche mit Augen-extracten. Arch. f. Ophth.
1911, 80, 399.
(5) Rapos, A. : Ueber das Auftreten von Komplementbindenden Antikorpern
nach Vorbehandlung mit arteigenen Gewebezellen, nebst Bemer-
kungen tiber die anaphylaktische Entstehung, der Sympathischen
Ophthalmie. Zeitsch. f. Immunititsf. 1913, 20 (orig.), 305.
(6) Fucus, A., AND Metuer, J.: Studien zur Frage einer Anaphylaktischen
Ophthalmie. Arch. f. Ophth. 1914, 88, 280.
(7) von Satuy, A.: Anaphylaxie versuche mit sog. chemisch reinem Augen-
pigment (von Rind, Schwein, und Kamnchen), nebst Pathologisch
anatomischen Untersuchungen. I. Teil. Klein. Monatsbl. f. Augen-
heilk., 1916, 56, 79.
(8) Woops, A. C. : Ocular anaphylaxis. I. The reaction to perfusion with
specific antigen. Archives of Ophth., 1916, 45, 557.
(9) Daun, H. H.: The anaphylactic reaction of plain muscle in the guinea-pig
Jour. of Pharm. and Exp. Therap., 1913, 4, 167.
(10) Scuurrz, W. H. Physiological studies in anaphylaxis. I. The reaction of
smooth muscle of the guinea-pig sensitized with horse serum, Jour.
of Pharm. and Exp. Therap., 1910, 1, 549.
92 ALAN C. WOODS
(11) Scuunrz, W. H. : Physiological studies in anaphylaxis. IV. Reaction of
the cat towards horse serum. Jour. of Pharm. and Exp. Therap.,
1912, 3, 219.
(12) Woops, A. C. : Ocular anaphylaxis. III. The réle of uveal pigment.
Arch. of Ophth. 1917, 46, 283.
(13) Ustennurn P. : Zur Lehre van des Unterscheidung verschiedener Eiweis-
sarten mit Hilfe- spezifischer Sera. Festschrift zum _ sechsigsten
Geburtstage Robert Koch, Jena 1903.
(14) Untennuth anp Harnpex. Untersuchungen iiber die praktische Verwert-
barkeit der Anaphylaxie zur Erkennnng und Unterscheidung ver-
schiedener Eweissarten. Zeitschr. f. Immunititsf., 1910, 4, orig., 761.
(15) Rospenav, M. J. anp ANDERSON, J. L. : Further studies upon anaphylaxis.
U.S. Public Health and Marine Service Hygienic Laboratory Bulletin,
no. 45, 1908.
THE EXAMINATION OF THE BLOOD PRELIMINARY
TO THE OPERATION OF BLOOD TRANSFUSION
ARTHUR F. COCA
From the Department of Bacteriology in Cornell University Medical College and
the New York Orthopedic Hospital
Received for publication January 16, 1918
Before the operation of blood transfusion is carried out it is
necessary to test the compatibility of the prospective donor’s
blood with that of the patient; that is, it must be determined
whether the blood corpuscles of the donor will remain intact in
the circulation of the patient. It is altogether probable that
no two human individuals possess blood having exactly the
same chemical composition; indeed, the possible different com-
binations of demonstrated different substances in the corpuscles
alone have been estimated to be about 4000 (1).
Fortunately, however, the differences in the blood plasma and
most of those in the corpuscles, although they are, perhaps, the
cause of certain unpleasant symptoms, such as a chill followed
by a rise of temperature, do not constitute an incompatibility
such as could contraindicate the use of the blood of an individual
for transfusion. Incompatibility in the latter sense is always
due to the presence in the blood of the one individual of isoag-
glutinins or isohemolysins or both of these agencies against the
corpuscles of the other individual. As the isohemolysins are
never present without associated isoagglutinins and since the
latter are sometimes found alone, it is necessary, for the purpose
under consideration, to examine the blood of the respective
individuals only for the presence of isoagglutinins.
Through the studies of Landsteiner and of von Dungern it is
known that there are two different isoagglutinins in human
blood.! These have been designated with the capital letters A
1 A discussion of the large body of evidence supporting this assumption would
lead too far from the practical purpose of this paper.
93
94 ARTHUR F. COCA
and B. The respective agglutinable elements in the corpuscles
are designated with the small letters a and b. An isoagglutinin,
eg., “A” and its corresponding agglutinable substance ‘“a”’
cannot be present in the same person, as this would result in the
agglutination of the individual’s corpuscles by his own plasma.
All of the possible combinations of the substances A, B, a and
b are found in human beings and individuals have been separated
into four groups in accordance with the different combinations
of those substances. This grouping is shown in table I.
TABLE 1
Plasma’... v8 4. oe A and B (Plasma... :0.3-2 nee A
SrepP iy esas pal ep satiaae omer gan eeue dt \'Corpuseles?:.9.255= ee b
Pigsma 35515. eee B 7 { Plasma... 13245. Speer 0
Sn ee ba Slee Peer Eee a See da \ Corpuscles.........:. a and b
The grouping is schematically shown in figure 1, which
consists of four test tubes containing blood of the four groups
of individuals.
Since the plasma of individuals of group I contains both A
and B isoagglutinins it is able to agglutinate the corpuscles of
the individuals of all of the other groups all of which contain one
or both of the agglutinable substances a and b. Similarly it is
clear that a group II plasma will agglutinate group III or group
IV corpuscles but not those of group I. Group III plasma will
agglutinate group II or group IV corpuscles but not those of
group I. Finally, group IV plasma is not able to agglutinate
any corpuscles since it lacks both isoagglutinins.
The proportional (percentage) representation of the four
groups among human individuals has been estimated in three
different studies as follows:
VON DUNGERN
AN W. L. Moss (3)
HIRSCHFELD (2) OLMSTEAD (4)
Group els a42 :.3.) 22 clse ee oo eee 36.0 43 46
Croup wll e en. a een peee accep eee 47.0 40 39
Groups ne se Sc sateen vee 11.0 i 13
Group Vs in. eee eee ote eee 5.7 10 2
These results show that by far the greater number (about
a
EXAMINATION OF BLOOD BEFORE TRANSFUSION 95
83 per cent) of human individuals are nearly equally divided
between groups I and II.
It is evident that donors that are of the same group as the
patient should be preferred for the transfusion, but if such a
donor is not available then one of another group must be selected
whose corpuscles are not agglutinated by the patient’s plasma.
In such a case the donor’s plasma will usually contain agglutinins
for the patient’s corpuscles but this fact does not constitute a
contraindication to the use of the donor’s blood for the trans-
Fic. 1. Bhoop Grourina or HumAN INDIVIDUALS (SCHEMATIC)
The corpuscles are represented by the circles, the plasma by the space sur-
rounding the circles.
fusion because the donor’s plasma is so diluted in the patient’s
circulation that its agglutinating or hemolytic power is reduced
below the point of injury to the patient’s corpuscles.
The testing of patient’s and donor’s blood preliminary to
transfusion may be done directly or indirectly.
DIRECT TEST
For carrying out the direct test it was formerly customary to
obtain several cubic centimeters of blood by venepuncture from
both patient and donor a small part of which was mixed with
96 ARTHUR F. COCA
sodium citrate solution the remainder being allowed to clot.
The citrated blood was washed with normal saline solution and
in a dilute (1-10) suspension it was mixed with the clear serum
of the other individual. After an incubation of one or two
hours at 37°C. the presence of agglutination and hemolysis was
determined macroscopically.
Weil (5) took the first step in the simplification of this time
consuming procedure by mixing the citrated blood of the two
individuals in three proportions: namely, one part of A with
nine parts of B; one of A with one of B; and nine of A with one
of B. He found that the phenomenon of agglutination was not
interfered with by the ‘presence of the even more numerous
corpuscles not taking part in that reaction, and that the reaction
could be observed macroscopically after the usual incubation
period. Hemolysis, also, could be detected by the presence of
hemoglobin in the supernatant fluid of the mixtures after
centrifugation.
A further modification in the interest of economy of bloed was
contributed by Rous and Turner (6). These authors using the
ordinary white blood cell mixing pipet drew up first, 10 per cent
sodium citrate to the mark 1 and then blood from the finger to
the mark 11 and blew out the citrated blood into a narrow test
tube. The two bloods were then mixed with the use of the
Wright capillary pipet in the proportions suggested by Weil,
and the mixtures sealed in the capillary pipet by fusing the tip.
After fifteen minutes the tips of the sealed pipets were broken
and a drop of the mixtures examined microscopically in normal
saline solution for agglutination.
The following method, which is based on a suggestion of Dr.
James Ewing, permits the mutual tests to be made with prac-
tically a single drop of blood from each individual and with the
use of little more than the usual apparatus employed for a white
blood cell count.
The patient and the donor or donors being in the same or
adjoining rooms, a finger of each individual is prepared as for
an ordinary blood count, for which purpose soap and water
meet every requirement. Three ordinary glass slides are placed
oe OS
See RR es cag
ey =
EXAMINATION OF BLOOD BEFORE TRANSFUSION 97
in order as shown in figure 2. For convenience the right and
left ends of all of the slides or of only slide 3 may be marked
A and 5, the first letter referring to the patient, the second one
referring to a donor. The stem of a white blood cell mixing
pipet is filled with normal saline solution up to the ninth mark
and the fluid thus measured (nine divisions) is deposited on
slide 1 at the right (position B). The pipet is then washed out
with a 10 per cent solution of sodium citrate, enough of this
solution being left in the stem to fill the terminal division. The
1
9-Nacl
~1-Blood-B
rs
9-Blood-A }- _J9-Bl 00d-B
L 1-Citrate 1-Citrate
3
3-cit. Blood-A }- {| ae _j3-cit. Blood-B
$-dil. Blood-B : = $-cit. Blood-A
Fic. 2. Ssaowrna ARRANGEMENT OF SLIDES IN THE BLoop TEST.
patient’s finger is punctured as for a blood-count and the blood
is drawn up into the stem of the pipet to the mark 1. This
mixture of blood and citrate solution is deposited on slide 2 at
the left and a similar mixture of the donor’s blood with citrate
solution is deposited at the right of the same slide, the pipet
being thoroughly washed, between the two operations, first with
saline solution and then with the citrate solution. One division
of the donor’s citrated blood (B) is well mixed with the nine
divisions of saline solution at the right of slide 1, the resulting
mixture thus representing a 1 to 10 dilution of the donor’s blood
B. If equal parts of this diluted blood and of the undiluted
citrated blood of the patient A are mixed, the resulting mixture
98 ARTHUR F. COCA
will represent 9 parts of blood A and 1 part of blood B in 10
parts of saline diluent. Such a mixture is made by drawing up
with the white blood cell mixing pipet, first, 3 divisions of ‘ci-
trated blood-A at the left of slide 2 and then 3 divisions of diluted
blood-B, the two portions being then deposited together at the
left of slide 3, stirred once with the tip of the pipet and covered
with a ¢ inch cover-glass. At the right of slide 3 is placed a
mixture of 3 divisions each of the two undiluted citrated bloods
at right and left of slide 2, this mixture, likewise, being covered
with a cover-glass. In making this second mixture, which, in a
sense, is intended as the reverse of the first mixture, it is not
necessary to dilute the patient’s citrated blood —-A— ; first, be-
cause, in most instances, the corpuscular content of the patient’s
blood is reduced sufficiently to meet the purpose of the dilution
and secondly, because the quantitative relation in the mixture,
as far as the ratio of donor’s plasma to patient’s corpuscles is
concerned, is nearer that at the actual transfusion than if the
patient’s blood were used in a 1 to 10 dilution. If, however,
mutual tests are being made in two normal individuals in order,
for example, to determine their group relationship, then blood-A
must be diluted 1 to 10 before being mixed with blood-B in
the second mixture.
Agglutination usually begins at once and can easily be detected
with the microscope as it has been described by previous authors.
If agglutination does not appear the observation should be con-
tinued for fifteen minutes, at the end of which time only the
borders of the covered blood films will have begun to dry.
To recapitulate: three ordinary glass slides are placed in order
as shown in figure 1 and designated 1, 2 and 3, the third slide
being marked left and right for identification with the letters
A and B. With a white blood cell mixing pipet a mixture of
9 divisions of the patient’s blood and 1 division of 10 per cent
sodium citrate is placed at the left on slide 2 and a similar mix-
ture of the donor’s blood and the citrate solution is placed at the
right of the same slide. A mixture of 1 division of the citrated
donor’s blood and 9 divisions of normal saline solution is placed
at the right end of slide 1. Three divisions of the citrated
EXAMINATION OF BLOOD BEFORE TRANSFUSION 99
blood-A and 3 divisions of the diluted blood-b are drawn up
into the pipet deposited at the left of slide 3 and covered with a
cover-glass. Three divisions each of the two undiluted citrated
bloods are similarly deposited at the right of slide 3 and covered
with a cover-glass. These two latter mixtures are immediately
examined microscopically for agglutination.
If more than one donor is to be tested the series of slides must
be duplicated or triplicated et cetera according to the number
of prospective donors. In such case two divisions of the citrate
solution are left in the end of the pipet before obtaining the
patient’s blood the latter being then taken twice up to the mark
1. This amount suffices for the examination of three donors.
INDIRECT TEST
If the blood-grouping of patient and donors has been deter-
mined it is unnecessary to carry out the direct test because the
availability of the donor’s blood for the transfusion can be
learned by referring to the constitution of the blood of the dif-
ferent groups as shown in table 1 or in figure 1. A patient of
group I, for example, requires a donor of group I, the blood of
all other groups being incompatible. The indirect test is based
on the foregoing principle and for its performance, as Moss
pointed out, it requires the storing of sera of the two groups II
and III. If the corpuscles of an individual are not clumped by
either of these sera, that individual belongs to group I; if they
are clumped by both sera they must be group IV corpuscles;
if they are clumped only by the group II serum or only by the
group III serum then they belong respectively to a group III
individual or to a group II individual.
The technical procedure of the direct test is applicable also to
the indirect test. If the patient and a single donor are to be
examined the mixtures on slides 1 and 2 (figure 1) are prepared
exactly as for the direct test. Two slides—3a and 3b—must
take the place of slide 3 in figure 1 since each blood is to be mixed
with both stock sera. Three divisions each of the patient’s
undiluted citrated blood are mixed respectively with three
divisions of the two sera and placed under cover-glasses at the
THE JOURNAL CF IMMUNOLOGY, VOL. III, NO. 2
100 ARTHUR F. COCA
two ends of slide 3a and similar mixtures of the diluted donor’s
blood with the two sera are placed on slide 3b. Clumping, if
present, can be seen at once with the microscope and in a few
minutes it becomes apparent to the naked eye.
REFERENCES
(1) von DuNGERN AND HirscHrELD: Muench. Med. Wochenschr., 1910, p. 741.
(2) von DuNGERN AND HirscHFe.pD: Zeitschr. f. Immunitaetsf., 1910, 8, 526.
(3) Moss, W. L.: Bull. Johns Hopkins Hospital, 1910, 21, 62.
(4) MELENEY, STEARNS, FoORTUINE AND Ferry: Am. J. Med. Sci., 1917, 154, 733.
(5) Weitz: Journal of A. M. A., January 30, 1915, p. 425.
(6) Rous anp TuRNER: Jour. Amer. Med. Assoc., 1915, 64, 1980.
EXPERIMENTS UPON THE PASSIVE TRANSFER OF
ANTIBODIES FROM THE BLOOD TO THE
CEREBROSPINAL FLUID
JOHN A. KOLMER anv SHIGEKI SEKIGUCHI
From the McManes Laboratory of Experimental Pathology of the University of
Pennsylvania
Received for publication January 24, 1918
Since the cerebrospinal fluid under normal conditions is gener-
ally free of certain normal or natural antibodies or other con-
stituents which may be present in the blood, as, for example,
hemolysins for the erythrocytes of various animals, agglutinins
for various micro-organisms, diphtheria antitoxin and comple-
ment, the mechanism concerned in the production of the cere-
brospinal fluid under normal conditions is regarded as an effectual
barrier against the entrance of antigens, antibodies and various
chemical substances into the cerebrospinal fluid. For this rea-
son the presence of antibodies in the cerebrospinal fluid during
disease is generally interpreted as indicating infection of the
tissues of the central nervous organs with the production of
antibodies by tissues in direct communication with the cerebro-
spinal fluid or, by an injurious effect upon the choroid plexus and
subarachnoid villi, facilitating the passage of antibodies from
the blood to the cerebrospinal fluid. As shown by Wile and
Stokes (1, 2 and 3) and Hauptmann (4) the cerebrospinal fluid
may show the presence of syphilis reagin (the antibody concerned
in the Wassermann reaction) and present abnormal chemical
and cytological changes in the late primary or early secondary
periods of that infection, with or without demonstrable clinical
evidences of infection of the nervous system. For this reason
Wile and Stokes very properly emphasize the importance of care-
ful clinical examinations of the nervous system during the early
stages of syphilis, coupled with a thorough examination of the
101
102 JOHN A. KOLMER AND SHIGEKI SEKIGUCHI
cerebrospinal fluid and regard the presence of syphilis reagin in
the fluid as an indication of spirochaetic invasion of the tissues
of the central nervous organs.
As agglutinin for B. typhosus may be found in the cerebro-
spinal fluid during typhoid fever without clinical manifestations
of involvement of the central nervous organs, it would appear
possible for 7. pallida or other micro-parasites lodged in tissues
other than the nervous tissues to stimulate the production of
antibodies to such extent as to reach a high degree of concentra-
tion in the blood with a passive or forcible transfer of these anti-
bodies into the cerebrospinal fluid in a manner analogous to the
transfer or elimination of the syphilis reagin in the secretions of
the mammary glands and kidneys. The object of our experi-
ments was to determine whether antibodies introduced into the
venous blood of normal experimental animals passed into the
cerebrospinal fluid and more particularly if this occurred after the
introduction of the syphilis antibody or reagin, in order to study
by experimental means whether the presence of antibodies in the
cerebrospinal fluid is to be accepted as indicating the presence
and activity of the respective antigen or antigens in the tissues
of the central nervous organs, or, whether antibody in high
concentration in the blood may be passed into the cerebrospinal
fluid by the uninjured mechanism governing the production of
cerebrospinal fluid.
EXPERIMENTAL
Dogs were employed in all experiments because of the ease
with which amounts of blood-free cerebrospinal fluid up to 2 ce.
may be secured by spinal puncture. In experiments concerning
the passive transfer of the syphilis reagin from the blood to the
cerebrospinal fluid, normal dogs were bled from the carotid artery
under ether anesthesia and an equal amount of human serum.
from active cases of syphilis was injected intravenously. Was-
sermann reactions were conducted with the blood and cerebro-
spinal fluid of each animal prior to injection and at varying
intervals afterward; in conducting these tests, the sera were used
in amounts of 0.1 ec. and the cerebrospinal fluid in amounts of
PASSIVE TRANSFER OF ANTIBODIES 103
0.5 cc. with each of three antigens; namely, an alcoholic extract
of beef heart re-enforced with cholesterin; an alcoholic extract
of syphilitic liver and an extract of acetone insoluble lipoids.
The following protocols of several experiments express the results
observed.
a. The passive transfer of human syphilis reagin in the blood of a normal
dog to the cerebrospinal fluid. Weight 6250 grams; preliminary Wasser-
mann reactions with serum and cerebrospinal fluid negative with all
antigens. Under ether anesthesia 150 ec. of blood was removed from
the carotid artery and 210 ce. of human syphilitic serum yielding
++-+-+ Wassermann reactions with all antigens, were injected into the
jugular vein.
Wassermann tests with cerebrospinal fluid removed twenty-two
hours later were weakly positive.
Wassermann tests with blood removed two and four and one-half
hours later were moderately positive, twenty-two hours after, the
tests were weakly positive and seventy hours later, completely negative.
The tests with urine secured four and one-half and twenty-two hours
after the injection of syphilitic serum were negative in doses of 0.5 ce.
urine with each antigen.
b. A second dog weighing 5420 grams was bled 150 ce. under ether
anesthesia and 250 ec. of human syphilitic serum yielding ++-+-+
Wassermann reactions with all antigens, were injected into a femoral
vein. Preliminary Wassermann tests with blood and cerebrospinal
fluid yielded negative reactions with all antigens.
Cerebrospinal fluid removed four hours later yielded doubtfully
positive reactions with each of the three antigens; fluid removed twenty-
four hours after transfusion yielded negative reactions.
Wassermann tests with blood serum secured two and four hours after
injection yielded moderately positive reactions, while tests with serum
secured twenty-four and seventy-two hours after injection were negative
with all antigens.
c. The passive transfer of human syphilis reagin in the blood of a dog
to the cerebrospinal fluid after the preliminary injection of sterile horse
serum into the spinal canal. Since the experimertts of Flexner and
Amoss (5) with the virus of acute anterior poliomyelitis have indicated
that the injection of sterile fluids into the spinal canal of monkeys
facilitates the passage of the virus from the blood to the central nervous
organs, due presumably to injury to the mechanism governing the pro-
104 JOHN A. KOLMER AND SHIGEKI SEKIGUCHI
duction of cerebrospinal fluid, we have conducted experiments similar ~
to those already summarized except that the animals received an intra-
spinal injection of 1 cc. of sterile horse serum twenty-four hours before
transfusion with human syphilis serum, followed by tests of the cere-
brospinal fluid and blood for the syphilis reagin.
Dog, weighing 9150 grams, was given an intraspinal injection of 1 ce.
sterile horse serum. Wassermann tests with blood and cerebrospinal
fluid were negative with all antigens. Twenty-four hours later 150 ce.
of blood were removed from the carotid artery under ether anesthesia
and 150 ec. of human syphilitic serum yielding ++-+-+ Wassermann
reactions with all antigens, were injected into a femoral vein.
Cerebrospinal fluid removed four hours later yielded moderately
positive Wassermann reactions with all antigens; cerebrospinal fluid
removed twenty-two hours after the injection of serum yielded com-
pletely negative reactions with all antigens.
Blood serum secured four hours after transfusion yielded moderately
positive reactions and twenty-two hours after transfusion weakly posi-
tive reactions. Tests made with the serum forty-eight hours after
transfusion were negative with all antigens.
While in this experiment the amount of human syphilitic serum
injected was less per kilogram of body weight than employed in the
former experiments, the amount of reagin in the cerebrospinal fluid
appeared to be somewhat greater as judged by the degree of complement
fixation with each of the three antigens; a similar result was observed
in the following experiment.
d. A dog weighing 7050 grams was given an intraspinal injection of
1 cc. of sterile horse serum. Preliminary Wassermann tests with cere-
brospinal fluid and blood serum were negative with all antigens.
Twenty-four hours later 150 cc. of blood were removed from the carotid
artery under ether anesthesia and 250 cc. of human syphilitic serum
yielding +-+-+-++ Wassermann reactions with each of the three antigens,
were injected into a femoral vein.
Cerebrospinal fluid removed three hours later yielded moderately
positive reactions with each of the three antigens; fluid secured twenty-
two hours after the transfusion yielded doubtfully positive reactions,
after forty-eight hours the fluid yielded completely negative reactions.
Blood serum secured three hours after transfusion yielded strongly
positive reactions; serum secured twenty-two hours after yielded weakly
positive reactions while with serum secured seventy hours after the
reactions were completely negative.
PASSIVE TRANSFER OF ANTIBODIES 105
e. The passive transfer of typhoid agglutinin (dog) from the blood of a
normal dog to the cerebrospinal fluid. As in the experiments already
mentioned a heterologous immune serum (human) was employed, it
was not surprising that elimination of the syphilis reagin from the blood
stream of the dogs was rapid and usually completed within seventy-two
hours to such extent that the reagin could not be demonstrated in the
blood serum by the Wassermann tests. We have immunized a dog
with typhoid bacilli until the amount of agglutinin in the blood reached
a high degree of concentration. During the process of immunization
the blood serum and cerebrospinal fluid were examined at intervals for
the presence of agglutinin. At the completion of the period of im-
munization the animal was bled to death under ether anesthesia and
the serum was injected into a second normal dog followed by examina-
tions of the cerebrospinal fluid for the presence of agglutinins.
A dog weighing 9450 grams yielded negative agglutination reactions
with an emulsion of typhoid bacilli in final dilutions of 1: 2 with blood
serum and cerebrospinal fluid. After a series of fourteen intravenous
injections with typhoid vaccine at intervals of four and five days, the
serum agglutinated in final dilution of 1: 5120 and the cerebrospinal
fluid in final dilution of 1:8 (microscopic technic). The animal was
now bled and 150 cc. of serum secured.
From a second dog weighing 5000 grams 100 ec. of blood were taken
from a carotid artery under ether anesthesia and 150 ce. of the typhoid-
immune serum were injected into a femoral vein. Preliminary aggluti-
nation tests with the serum and cerebrospinal fluid of this animal yielded
negative agglutination tests in final dilutions of 1: 2 with an emulsion of
typhoid bacilli.
Cerebrospinal fluid removed three hours after transfusion yielded
complete agglutination of B. typhosus in dilutions up to 1: 5; fluid
removed twenty-four hours later agglutinated 1:4 while forty-eight
hours after transfusion agglutination was not in evidence with the
lowest dilution; namely, 1: 2.
The blood serum removed three and again twenty-four hours after
transfusion yielded complete agglutination in final dilutions up to
1: 160; forty-eight hours after transfusion the titer was 1: 40 and ninety-
six hours later only partial agglutination occurred in final dilutions of
1:4 and 1:8.
106 JOHN A. KOLMER AND SHIGEKI SEKIGUCHI
DISCUSSION
These experiments have, in our opinion, demonstrated two
facts with special reference to syphilis reagin, namely, that large
amounts of this antibody in the blood may result in the passage
of small amounts into the cerebrospinal fluid with a normal
condition of the central nervous organs and of the mechanism
governing the production of the cerebrospinal fluid; secondly,
that subarachnoid injection of sterile serum appears to facilitate
the passage of antibody by injury to the mechanism governing
the production of the fluid or by the production of localized
congestion with increased transudation of blood constituents
into the cerebrospinal fluid. Accordingly it would appear
possible that in syphilis or other acute infections accompanied
by a high concentration of antibody in the blood, small amounts
of antibody may be found in the cerebrospinal fluid and that this
finding does not of itself necessarily indicate infection of the
central nervous organs. These observations, however, should
have no further significance and do not by any means lessen the
value of the studies of Wile and Stokes in syphilis, because they
have found the reagin in the cerebrospinal fluid during the pre-
roseolar period when the concentration of reagin in the blood
cannot be regarded as having reached a point of high concentra-
tion and, furthermore, found in some cases an increase of protein
and cells in the cerebrospinal fluid, which cannot be explained
at present on any other basis than actual infection of the nervous
tissues with 7’. pallida.
SUMMARY
1. The removal of blood from normal dogs followed by the intra-
venous injection of human syphilitic serum in amounts varying
from 30 to 50 ce. per kilogram of body weight was followed by the
presence of small amounts of syphilis reagin (the antibody con-
cerned in the Wassermann reaction) in the cerebrospinal fluid.
2. The reagin was found in the cerebrospinal fluid as early
as three hours after transfusion with syphilitic serum; tests at
shorter intervals were not made. The amount of reagin found in
0.5 ee. of cerebrospinal fluid was small in all experiments as based
upon the degree of complement fixation with all antigens.
PASSIVE TRANSFER OF ANTIBODIES 107
3. After irritation of the spinal meninges by the preliminary
injection of sterile horse serum the amount of reagin gaining
access to the cerebrospinal fluid after transfusion of syphilitic
serum, appeared to be somewhat greater.
4. All traces of syphilis reagin in the cerebrospinal fluid of
dogs following transfusion of human syphilitic serum apparently
disappeared after 22 to 48 hours as determined by completely
negative Wassermann reactions.
5. The intravenous injection of dog-typhoid immune serum
into a normal dog in amount of about 30 ce. per kilogram of body
weight, was followed by the appearance of traces of agglutinin
in the cerebrospinal fluid within three hours after transfusion;
fluid removed forty-eight hours later was free of agglutinin.
6. These experiments demonstrate the possibility of the
passage of antibody from the blood into the cerebrospinal fluid
without primary involvement of the central nervous organs or
injury to the mechanism concerned in the production of cere-
brospinal fluid, when the amount of antibody in the blood has
reached a point of high concentration. While it is possible that
in human syphilis the presence of traces of reagin in the cere-
brospinal fluid may be due to the passive transfer of this sub-
stance from the blood, as shown by Wile and Stokes the presence
of the reagin with or without other changes in the fluid, as an
increase of protein and cells usually indicates the presence and
activity of 7’. pallida in the tissues of the central nervous organs.
REFERENCES
(1) Wie, Upo J., anp Strokes, Joun H.: A study of the spinal fluid with reference
to involvement of the nervous system in secondary syphilis. Jour.
Ciuian. Dis., 1914, September, 607.
(2) WiuE, Upo J., AnD Stoxkss, Joun H.: Involvement of the nervous system dur-
ing the primary stage of syphilis. The Jour. A. M. A., 1915, 64, 979.
(3) Wir, Upo J., anp Stokes, Joun H.: Further studies on the spinal fluid
with reference to the involvement of the nervous system in early
syphilis. The Jour. A. M. A., 1915, 64, 1465.
(4) Hauprmann, A.: Die Diagnose der “‘irith-luetischen Meningitis’’ aus dem
Liquor befund. Deut. Ztsch. f. Nervenheilk., 1914, 51, 314.
(5) FLExNER, S., AND Amoss, H. L.: The relation of the meninges and choroid
plexus to poliomyelitic infection. Jour. Exp. Med., 1917, 25, 525.
THE ISOLATION, PURIFICATION AND CONCENTRA-
TION OF IMMUNE BODIES: A STUDY OF
IMMUNE HEMOLYSIN
7 M. KOSAKAT
From the Forensic-Medical Department of the Imperial University of Tokyo, Japan
Received for publication February 8, 1918
The numerous attempts to isolate immune bodies from their original
sera and to concentrate them have, up to this time, not given satis-
factory results. To obtain pure immune substance free from serum
protein is, on the one hand, a most important condition for the solu-
tion of problems as to the biophysical and biochemical properties of
immune bodies and it has, on the other hand, an even more important
relation to the therapeutic use of antibodies, as, for example, in the
possibility that by such means the uncomfortable results of serum dis-
ease, which are said to be engendered by serum protein, may be avoided.
Many workers have inquired into this problem with various methods,
which will be classified according to their purpose in two groups. One
group of methods endeavors to eliminate all of the serum protein ex-
cept the immune substance from the immune serum, while the other
seeks to extract only the immune bodies from the respective sera.
1. The elimination of serum protein
With the purpose of facilitating the practical use of immune bodies
the attempt to eliminate the serum protein has been chiefly performed
with respect to diphtheria or tetanus antitoxin.
a. Fractionating method. By saturation with magnesium sulphate
or half saturation with ammonium sulphate the antitoxin in serum
is precipitated with some kinds of globulin. Thus, by treating certain
immune sera with ammonium sulphate, Pick (1) found that from diph-
theria antiserum produced in the horse the antitoxin is precipitated
with the pseudoglobulin whereas from the antiserum of the goat it
is precipitated with the euglobulin. By this means the total amount
of serum protein ca” be remarkably diminished and the concentration
109
110 M. KOSAKAI
of antitoxin presumably can be expected. Pick states that by the isola-
tion of those fractions it is possible to concentrate the protective power
ten to fifteen times. This method of Pick was proved thereafter by
many workers not to be practicable.
Brieger and Krause (2) state that they succeeded in eliminating 75
per cent of total nitrogen from diphtheria serum, while preserving the
original antitoxic power. The antitoxic serum is diluted with an equal
volume of distilled water and it is then treated with ammonium sul-
phate. The resulting precipitate is placed in a 10 per cent watery solu-
tion of glycerin and treated with sodium chloride. The precipitate
produced by the sodium chloride contains no antitoxin, while the fluid
portion has protective power.
Gibson (3) prepared a refined and concentrated diphtheria antitoxin
in a similar way. By means of half saturation of ammonium sulphate,
and treatment of the precipitate with a saturated solution of sodium
chloride containing acetic acid, he could concentrate the number of
units of antitoxin per cubic centimeter from 200 to 500. But in carry-
ing out this process there was a loss of about 30 per cent of antitoxin
units. Gibson applied this method of concentration not only to
diphtheria antitoxin, but also with his co-worker Collins (4) to the con-
centration of agglutinin. ;
Frouin’s (5) method is more widely known. He added 2 to 5 per
cent watery solution of glycerin to diphtheria immune serum and then
saturated it with sodium chloride. The filtrate was placed on a water
bath at 75° to 80°C. for ten to fifteen minutes. The coagulated serum
was placed in 2 to 3 volumes of a half saturated solution of sodium
chloride and macerated in the ice-box for twelve to thirteen hours.
The maceration was repeated at least three times, each time with a
fresh saline solution. After this manipulation the total decanted fluid
was dialyzed and concentrated in vacuo at about 35°C. to the original
volume.
All of the above mentioned or other so-called fractionating methods,
however, do not give constant and satisfactory results, though they
have been more frequently employed in the practical preparation of
antitoxin than the other methods of concentration that are immediately
to be described.
b. Evaporation and freezing also have been employed for the con-
centration of immune bodiesas the works of Bujwid (6), Ernst, Coolidge
and Cook (7) and Hata (8) show, but the use of these methods has not
been continued, because by these methods not only the immune sub-
stance, but also all the colloidal substances in serum are concentrated.
woe
A STUDY OF IMMUNE HEMOLYSIN Lit
c. Reasoning from the assumption that antitoxin and other immune
bodies are of non-proteid nature, many workers have attempted to
eliminate the serum protein by trypsin digestion. Thus Préscher (9)
treated diphtheria serum with extract of pancreas at 32°C. and expected
to peptonize all of the protein substance associated with the immune
bodies. Then by half saturation with ammonium sulphate he precipi-
tated the antitoxin out of the solution of pepton. The last trace of pep-
ton was eliminated by dialysis. Similar attempts, which were made by
Belfanti and Carbonne (10), Pick and Brieger have not given satis-
factory results, because the antibodies, also, appear to be attacked by
trypsin, though their destruction is very slow.
All of these methods designed to eliminate serum protein from immune
serum are too incomplete to be used in general, because there are diffi-
cu!ties in the technique of handling the blood proteins and our knowledge
of serum proteins is still in confusion. Whether immune substances
are of proteid nature or not is quite unknown. Though an increase in
the globulin content of the blood as the result of immunization was
proved by Atkinson (11), Moll (12) and recently by Hurwitz and Meyer
(13), which may be indicative of the serum-globulin nature of immune
body, Joachim (14) has observed that the increase is manifested in the
non-protective fraction. Glassner (15), also, states that immunization
can be accomplished without any essential globulin change.
These methods have, therefore, only a practical interest their use
having resulted, to some degree, in a concentration of the immune bodies,
but they have no biological interest, because the purification of the im-
mune bodies with these methods is not possible. It seems likely, then,
that it will not be*possible to solve this great problem by any process
of elimination of the serum protein.
2. Extraction of immune bodies from their original sera
a. By employing their antigens. The question whether antigen and
antibody once combined are again separable or not has been treated
by many workers. When antigen has been mixed with antibody, the
two unite at once if the conditions are favorable, but this union differs
from a neutralization in a chemical sense inasmuch as it is not so difficult
to separate the former combination again. Thus Calmette (16) made
a neutral mixture of snake venom and its antitoxin again toxic, by
heating it at 68°C. v. Wassermann (17) also obtained the same result
with respect to a neutral mixture of pyocyaneus toxin and its anti-
112 M. KOSAKAI
toxin. Morgenroth (18) extracted the toxic substance from the toxin-
antitoxin mixture of both cobra venom and diphtheria toxin by means
of weak hydrochloric acid.
Hahn and Trommsdorf (19) treated agglutinated bacteria with 1/100
normal sulphuric acid and succeeded in separating the active agglutinin
again.
Landsteiner (20), and Landsteiner and Jagic (21) proved that agglu-
tinated red blood corpuscles washed with physiological salt solution
gave off some of the agglutinin again, when they were brought into
physiological salt solution and that the quantity of liberated agglutinin
is proportional to the temperature applied to the medium. They also
ascertained that the same relation could be established with respect to
precipitin.
Bail and Tsuda (22), and Spaet (23) proved that immune bodies which
combined with cholera vibrios can easily be separated from the latter
in physiological salt solution at 40° to 42°C. and also that serum of
various animals can separate that combination. Tsuda (24) sueceeded
also in separating the so-called normal immune substances in normal
sera from their union with the respective antigen, but he was not able
so well to separate the immune bodies in immune sera.
The preceding group of experiments demonstrate that a union of
antigen and antibody is separable without the loss of their respective
functions; however, they are far from effecting the purification and
concentration of immune bodies.
b. By employing «norganic surfaces. From the standpoint that the
antigen-antibody combination may be explained as an adsorption in a
merely physical sense, many workers have studied the phenomena of
adsorption of immune bodies by inorganic surfaces. Thus Biltz, Much
and Siebert (25) ; Andrejew (26); Landsteiner and Reich (27) have found
that, in common with other colloidal substances, immune bodies also
could be adsorbed by surfaces such as caolin and animal charcoal.
These surfaces act selectively and there is some distinction with respect
to the adsorbing power between electropositive and electronegative
surfaces.
Then, in logical sequence, the effort has been made to separate the
adsorbed immune substances from the inorganic surfaces. But these
experiments have always ended in negative or unsatisfactory results.
Thus, the well known work of Jacqué and Zunz (28) has shown that after
the antitoxin or the toxin has been adsorbed by such surfaces they are
by no means easily separable zn vitro, although zn vivo a separation, in
A STUDY OF IMMUNE HEMOLYSIN Pie
greater or less degree, is observed. According to Zunz, however, if
antitoxic serum is subjected to adsorption by animal charcoal, after a
great deal of the protein substance in the serum has been diminished by
the above mentioned method of Frouin, only a small quantity of anti-
toxin can be separated again into physiological salt solution.
Nicholas Ssbolew (29) states that when the immune serum of typhus
or cholera is treated with iron hydroxide, the precipitate that forms
contains all of the immune bodies, but attempts to separate the immune
bodies from that precipitate by means of various physical or chemical
processes succeeded only zn vivo, never in vitro.
The use of these inorganic surfaces for the concentration of immune
body can not give satisfactory results, because such surfaces adsorb
not only the immune substances, but also the serum protein. Those
agents have not a selective power to adsorb only the immune substance
such as the specific antigens have. Therefore, even if a separation has
succeeded, the purification must be quite questionable.
Studies on hemolysin
To solve the problem of the isolation, purification and concentration
of immune bodies, it is most convenient to use hemolysin, because its
biological reaction is more easily recognized zn vitro than that of other
immune bodies such as precipitin, agglutinin or bacteriolysin. And
since the antigen for hemolysin is red blood cells, we can use them as
ideal surfaces for our purpose, because they can take up their ambocep-
tors without adsorbing any blood protein. Moreover since this antigen-
amboceptor union can exist without hemolysis taking place so long as
complement is absent, the manipulation must be accordingly easy. But
on the other hand, as red blood cells are very susceptible to slight
physical or chemical changes in their environment, there are, in this
respect some technical difficulties connected with their use.
1. Studies on normal hemolysin. Normal blood serum of many
animals causes hemolysis to a greater or less degree when mixed with the
red corpuscles of another species of animal. We call this hemolytic
- substance “normal hemolysin.’ Bail and Rotky (30) showed that the
blood cells of the horse or guinea pig, that were sensitized with normal
active human serum could render hemolytic the physiological salt
solution, in which they were afterward digested; that is to say, the nor-
mal hemolysin of human serum which was combined with the red blood
cells can be separated again into physiological salt solution. Bail (31)
114 M. KOSAKAI
also reported a similar experiment with the blood corpuscles of sheep
that had been sensitized with imactive pig serum. Similar attempts,
however, with immune hemolysin have never been successful, though
the biological property of immune hemolysin appears to be quite analo-
gous with that of normal hemolysin.
2. Studies on immune hemolysin. Von Liebermann and Fennyvessy
(32) have succeeded in the extracting immune hemolysin against pig
blood cells from its union with the latter by means of 1/100 normal
hydrochloric acid. They state that the pure immune hemolysin that is
isolated by this method shows no protein reaction with most sensitive
protein tests and moreover it does not go through animal membrane.
But this isolation of immune hemolysin, as was pointed out by Pietro
Rondoni (33), cannot be said in the strict sense of the word to be a true
separation of immune body and antigen union, for hydrochloric acid
destroys not only red blood cells but also the immune body, especially
when the acid is of high concentration.
Pietro Rondoni extracted active immune hemolysin from sensitized
red blood cells by means of diluted NaOH solution, which has been
added to physiological salt solution. He says that in that case there is
no alteration of blood cells and immune hemolysin and that the extrae-
tion of combined amboceptor is accomplished in a short time either at
0°C. or at 37°C. But this method has one and the same failing as that
of von Liebermann, for alkali also effects hemolysis, though in a different
degree from that of acid, and Rondoni was not able to purify the iso-
lated hemolysin.
With the encouragement of Professor Mita I have taken up the
study of this problem and I have employed, as the most con-
venient material, the immune hemolytic serum of the rabbit
against sheep’s blood. The first question to be settled was
whether the union of amboceptor and antigen is completely sep-
arable or not. The method of Liebermann and Fennyvessy has
its value only with respect to the isolation and concentration of
immune hemolysin. That of Rondoni has also the disadvantage
that alkali cannot be used in the higher concentrations on account
of its hemolytic action. Furthermore, the experiment of Bail
and his co-workers proves that the same manipulation employed
for normal hemolysin cannot be applied to immune hemolysin.
It occurred to me that in order to separate the immune hemoly-
A STUDY OF IMMUNE HEMOLYSIN 1S
sin (hemolytic amboceptor) and red corpuscle combination, it
must be necessary to use some other agent beside sodium chloride
in the medium, because it seemed very unlikely that antigen and
amboceptor combined in ClNa solution could be separable again
in the same solution, even if a little alkali or acid be added or
if the temperature be modified.
From this point of view I examined various salts and non-
electrolytes and I found, after long investigation, that some
kinds of sugar have a surprising property of separating the united
blood cells and hemolysin.
Technique
By means of the usual repeated administration of the red blood cells
of the sheep, I obtained from rabbits an immune hemolysin for my
experiments the hemolytic power of which was 1: 10,000; which
means that 1 ce. of a 1 to 10,000 dilution of the inactivated immune
serum with 0.1 ce. of fresh guinea-pig’s serum could lake all of the
blood cells in 1 ce. of a 5 per cent emulsion of washed sheep’s blood cells
in two hours at 37°C.
I employed this serum for my study, diluting it to 100 volumes with
physiological salt solution. When 4 cc. of washed sheep’s blood is
mixed with 5 ce. of this diluted inactive immune serum and left at room
temperature for fifteen to twenty minutes, the red blood cells will unite
with all the amboceptors in it. This fact is in accordance with the
results of the work of Morgenroth (34), who found that red corpuscles
ean combine with far more amboceptor than the minimal quantity
by which the former will be completely laked.
The antigen and amboceptor union thus obtained was repeatedly
washed with physiologic salt solution, till the last trace of blood protein
was removed. If, under this condition, it is possible to separate the
combination, the amboceptor thus isolated must be pure. The union
was digested in solutions of various chemical substances and the ex-
tracted fluid after centrifugalization was tested as to its hemolytic
power. The results that were negative will be omitted from this report.
EXPERIMENTAL RESULTS
With the above mentioned technique, two test tubes of pure
antigen-amboceptor union were made and then digested at 37°C.
A. In 5 ee. of physiological salt solution.
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2
116 M. KOSAKAI
B. In 5 ce. of 10 per cent watery solution of saccharose.
After two hours, during which the tubes were shaken several
times, both were centrifugalized and red tinged extracts were thus
obtained. The hemolytic test was performed in the usual way,
1 ce. of a 5 per cent emulsion of washed sheep’s blood cells and
0.1 ec. of fresh guinea-pig’s serum being used in each tube.
The results of this experiment are shown in table 1.
TABLE 1
EXTRACT A (WITH SALT EXTRACT B (WITH SALT SOLUTION Junta coNMROG GHae
SOLUTION UP TO 1 cc.) uP TO 1.0 cc.) SALT SOLUTION
HOURS vP To 1.0 cc.)
| 0.5cc. | 0.3cc. | O.1ce. | O0.5ec. 0.3 ce. 0.1 ce. 0.03 ec. | 0.01 ec.
05 | — | — | — | Compl.| Weak | Weak- | Weak | Weak
20 tee = = — | Compl. | Compl. | Very Almost | Very
strong| compl.| strong
15 |) =) «= — | Compl. | Compl. | Almost | Compl. | Almost
| compl. compl.
2.0 | — _ — | Compl. | Compl. | Compl. | Compl. | Compl.
It is seen that the hemolytic amboceptor that had been united
with the red blood cells was easily separated into the solution of
saccharose, while the physiological solution was quite indifferent.
Beside the saccharose I used also glucose and lactose, and
found that these sugars also have the property of separating the
amboceptor-corpuscle union. I should like to note here that
these sugars must be pure and especially that the glucose must
be that of Merck. The following experiments were performed
with saccharose.
It was important to determine whether the separation of anti-
gen-amboceptor union by sugar could be made complete and it
was necessary, therefore, to examine the various physical factors °
that favor the separating power of the sugar.
Since the volume of the sugar solution used in our first experi-
ment (table 1) for extracting the hemolysin from the sensitized
corpuscles is the same (5.0 cc.) as that of the original hemolytic
serum used for the sensitization of the cells, we can estimate the
proportion of the absorbed hemolysin recovered by the digestion
A STUDY OF IMMUNE HEMOLYSIN I 7;
in the sugar solution by directly comparing the hemolytic titer
of the original serum and of the extraction fluid B. This com-
parison (0.1 ec.: 0.01 ec.) shows that 10 per cent of the absorbed
amboceptor was separated from the sensitized cells during the
digestion in the sugar solution. This calculation does not take
into account the inhibitory action of sugar on the hemolysis
(see under “‘Discussion’’) but as the amount of sugar present in
the test mixtures was relatively small, as compared with the
amount of salt present, its inhibitory influence may be disregarded.
FACTORS
THAT INFLUENCE THE SEPARATING POWER OF SUGAR
The separation of amboceptor from its antigen combination
with sugar depends upon various physical factors.
1. Influence of temperature
Pietro Rondoni states that the separation of the antigen-
amboceptor union by alkali is the same either at0°C. or at 37°C.
But as Landsteiner showed that the quantity of the separated
TABLE 2
Showing the effect of temperature on the extracting power of sugar solution.
EXTRACT A EXTRACT B
HOURS
0.5 cc. 0.3 cc. 0.1 ce. 0.5 cc. 0.3 ce. 0.1 cc.
0.5 Compl. | Weak Weak — _ --
1.0 Compl. | Almost Strong _ | =
compl.
125 Compl. | Compl. Almost Weak Trace =
compl.
2.0 Compl. | Compl. Compl. Almost Strong Weak
compl.
agglutinin from agglutinated bacteria is proportional to the tem-
perature of the medium, it is most probable that the separation
of hemolytic amboceptor from the sensitized blood cells is also
proportional to the temperature applied.
With the usual technique two test tubes of pure antigen
amboceptor union were made and digested:
118 M. KOSAKAI
A. In 5ee. of 10 per cent saccharose, at 37°C.
B. In 5 ce. of 10 per cent saccharose, at 5°C.
After two and one-half. hours, during which the tubes were
shaken several times, both tubes were centrifugated and the
hemolytic power of the extracts was tested in the usual way.
The results of this experiment are shown in table 2.
The experiment makes it evident that the higher temperature
Quantity
ssolated hemolysen
+
wl
ro/
Sneee
ec
AlN
Temperature nee (Is?) eg as" oe g8>" 3 6
Cc
Cuart I. SHOWING THE INFLUENCE OF TEMPERATURE ON THE SEPARATION OF
HEMOLYSIN FROM SENSITIZED CORPUSCLES IN SUGAR SOLUTION
Concentration of saccharose, 10 per cent; quantity of solution of siccharose, 20
cc.; time, fifteen minutes.
favors the separation of the hemolysin. It is known that the
blood cells are laked in physiological solution when certain high
temperatures are applied; we call this phenomenon heat-hemolysis.
According to Gros’s study, red blood cells are destroyed at 56°C.
after eighteen minutes in 0.9 per cent ClNa solution. But the
heat hemolysis, as he states, depends upon the concentration of
the emulsion of blood cells. A concentrated emulsion of blood
A STUDY OF IMMUNE HEMOLYSIN 119
cells is dissolved more slowly than a diluted one. In my second
experiment temperatures up to 65°C. were examined, a 25 per
cent emulsion (in this case, for only fifteen minutes) and the
macroscopic appearance of the extract after centrifugation was
about the same as that obtained at lower temperature. The
experiment makes it seem probable that the heat hemolysis is
inhibited by sugar. The results of this experiment are shown
in Chart 1.
Quantity
of
(solated hemalysin
Jime> os £ Ss 2 2s z +s <
hour
Cuart Il. SHowING THE INFLUENCE OF TIME ON THE DISSOCIATION OF CORPUSCLE
AND AMBOCEPTOR IN’ SUGAR SOLUTION
Curvea: Concentration of saccharose, 10 per cent; quantity of solution of sac-
charose, 20 cc.; temperature, 56°C.
Curveb: Concentration of saccharose, 10 per cent; quantity of solution of sac-
charose, 20 cc.; temperature, 45°C.
If temperatures above 70°C. are applied, the red corpuscles
will be altered in few minutes. If the same experiment is per-
formed with agglutinin or bacteriolysin it may be possible to
apply the higher temperature. After all, it is quite apparent
that higher temperatures greatly favor the dissociation of the
120 M. KOSAKAI
corpuscle-hemolysin combination and it also appears that the
separation can be made complete, if the conditions are favorable,
though in my experiment the separation has been carried out only
up to 84 per cent.
2. Influence of time
It might be thought that at the lower temperatures a longer
period of digestion would result in the same degree of separation
Quantity
c.
csoloted hemoly sr
of
Sucrose
Ye
Cuarr III. Showing THE INFLUENCE OF THE CONCENTRATION OF THE SUGAR
ON THE DISSOCIATION OF THE CORPUSCLE-AMBOCEPTOR COMBINATION
Curve a: Temperature, 56°C.; time, thirty minutes; quantity of solution of
saccharose, 20 cc.
Curve b: Temperature, 40°C.; time, thirty minutes; quantity of solution of
saccharose, 20 ce.
of the hemolysin as would a shorter period at a higher temperature.
But within the limit of four hours the length of time, according
to my experience, does not exercise any influence upon the separa-
tion of the immune body. In this fact the dissociation of the
hemolysin differs from other biological reactions, such as hemol-
A STUDY OF IMMUNE HEMOLYSIN 121
ysis by heat. Whether, as seems improbable, any different effect
would be obtained by a still longer digestion I am unable to say.
A more prolonged digestion of the cells causes considerable
hemolysis which, moreover, interferes with the further manipula-
tion of the extract.
5. The influence of the concentration of saccharose
It is, of course impossible to use a hypotonic solution, on ac-
count of its hemolytic action.
The isotonic solution of saccharose is theoretically a 7.8 per
cent watery solution. I examined the effect of different con-
centrations from the isotonic solution up to a 25 per cent solu-
tion, but I found that concentrations above 20 per cent are also
unavailable, because in such concentrated solutions there appears
also partial destruction of the red corpuscles at higher tempera-
tures.
The result of this experiment, graphically presented in chart
III shows that the concentration of the sugar does not exert any
considerable influence upon its separating power, but it seems as
though the isotonic or slightly hypertonic solution is the optimum
for the action of the sugar.
4. Influence of the quantity of sugar solution
It would be most convenient for subsequent study if all the
combined hemolysin could be separated in a small quantity of
sugar, but when the quantity of sugar solution is little, the
separated hemolysin also is relatively small in amount; it seems
as though the hemolysin becomes saturated, to a certain extent,
in the solution. In such case a new solution should act effectively
to separate more of the combined hemolysin, and the following
experiment shows that this is true.
With the usual technique one test tube of pure antigen-ambo-
ceptor union is prepared. This is digested for two hours at 37°C.
in 5 ee. of a 10 per cent watery solution of saccharose; after which
it is centrifugated and a red tinged extract is obtained. (Extract
A.) To the sediment of sensitized red blood cells a second por-
122 M. KOSAKAI
tion of 5 ec. of 10 per cent saccharose is added and after 1 hour
incubation also at 37°C., it iscentrifugated. The second diges-
tion caused considerable hemolysis. (Extract B.)
The hemolytic power of the two extracts was tested in the
usual way. The results of this test are shown in table 3.
TABLE 3
Showing the effect of a second treatment of the corpuscle-amboceptor combination
with sugar solution
EXTRACT A EXTRACT B
HOURS —— eee
0.5 cc. 0.3 ce. 0.1 ce. 0.5 ce. 03ce. | O1ee.
0.5 Compl. | Weak Weak Compl. | Almost Weak
compl.
1.0 Compl. | Almost Strong Compl. | Compl. Almost
compl. compl.
15 Compl. | Compl. Compl. Compl. | Compl. Compl.
2.0 Compl. | Compl. Compl. Compl. | Compl. Compl.
It is seen to be possible to extract a considerable quantity of
hemolysin by a second digestion and it may also be possible by
a third or more treatments to dissociate all of the combined
hemolysin. But as even the second extraction caused more or
less destruction of red blood cells, it was necessary to employ
such a quantity of the sugar solution that by first treatment
most of the united immune hemolysin will be separated. Experi-
ments were carried out to determine what that quantity is and
the result of this study is shown in chart IV.
The experiment shows that it is necessary to use at least 20 ce.
of sugar solution in order to extract almost all of the amboceptor
that was combined with 4 ce. of red blood cells.
To summarize the foregoing experimental results: when 20-30
ec. of a 10 per cent watery solution of cane sugar are mixed with
4 ee. of sheep’s red blood cells that have taken up the amboceptor
in 0.05 cc. of immune hemolytic rabbit’s serum and this mixture
is incubated at 60°C. (1: 10,000), after fifteen to twenty minutes
almost all of the immune substance, at least five-sixths of it, will
be transferred from the corpuscles into the sugar solution.
Thus the separation of antigen-amboceptor union is practically
A STUDY OF IMMUNE HEMOLYSIN 3
complete and in that separating action not only the quality of
medium, but also its quantity and the temperature play impor-
tant roles.
Quantity
¢,
cseialed hemolysin
‘
Quantity of 4 § a. 07S a Son wee se EO Gnd
Solution of Susecrese
ee,
CuHart LIV. SHowINa THE INFLUENCE OF VOLUME ON THE DISSOCIATING POWER
OF THE SuGAR SOLUTION ON THE AMBOCEPTOR-CORPUSCLE COMBINATION
Concentration of saccharose, 10 per cent; temperature, 56°C.; time, thirty
minutes.
PURIFICATION AND CONCENTRATION OF THE ISOLATED IMMUNE
HEMOLYSIN
As shown above, I was able to isolate immune hemolytic ambo-
ceptor from the union with its antigen. It was, however, very
important to purify and concentrate the isolated hemolysin in
order to investigate its biological properties.
a. Elimination of corpusclar substances including hemoglobin
It is true that the sugar extract of the sensitized corpuscles
does not contain, beside the immune hemolysin, any serum pro-
tein, but it does contain a small quantity of destroyed red blood
124 M. KOSAKAI
cells. Therefore the extract has always a slight red nuance of
hemoglobin. To purify the hemolysin this trace of corpuscular
substance must be excluded. For the solution of this second
difficulty I examined various means.
1. Dialysis. Though the hemolysin does not pass through
parchment, hemoglobin, also, is not dialyzable; therefore we can
re use dialysis in order to exclude hemoglobin.
2. Adsorption. By inorganic surfaces, such as caolin, hemo-
Babin is completely adsorbed but at the same time the hone
is also adsorbed.
3. All other methods except the following have ended in a
negative result.
4. Successful method. It is well known that the hemolysin in
immune serum can not be extracted with fat solvents such as
ether, petroleum ether, aceton and chloroform and it is by no
means susceptible to them, while complement is easily attacked
by them. By shaking the sugar extract with ether in the manner
to be described I have succeeded in removing from the extract all
traces of the corpuscular substance thus obtaining the hemolysin
in a pure condition.
If, to the sugar extract, 5 to 10 volumes of pure ether are added,
and the mixture is shaken in a separatory funnel for one to two
hours, there appear three layers in it. The upper layer is of
ether; the middle layer consists of a coagulated mass of corpuscular
substance, which is colored red and looks gelatinous; the lowest
layer is the sugar solution, which is now quite colorless and con-
tains the same quantity of the extracted hemolysin as before the
treatment with ether.
When, instead of ether only, ether and hydrochloric acid are
used together and likewise shaken in a separatory funnel, the
exclusion of the hemoglobin is far easier, but in that case some
of the hemolysin is destroyed, especially when the acid is con-
centrated.
If the exclusion of the hemoglobin by shaking with ether is not
accomplished by one treatment then the lowest layer is separated
into another funnel and treated again with fresh ether. After
repeating this manipulation twice or thrice °the last trace of
hemoglobin will be excluded.
A STUDY OF IMMUNE HEMOLYSIN 25
b. Elimination of sugar and the trace of salt
The solution thus purified with ether contains still a certain
quantity of sugar and a trace of ClNa. These substances are
quite easily eliminated by dialysis; i.e., the purified solution is
poured into parchment and kept in running water for about two
days.
c. Concentration of purified hemolysin
Finally the watery solution of the hemolysin is brought into
an exsiccator and concentrated in vacuum to the required
volume.
d. The nature of pure immune hemolysin
A further study is necessary to determine this. I can only
report here that the isolated hemolysin preparation does not
react so sensitively as true protein with protein tests such as
that with sulphosalicylic acid or with ClNa + acetic acid.
DISCUSSION
1. The influence of the medium upon the action of hemolytic serum
The action of compound hemolysin depends upon the reaction
of its environment or the quantity of ions init. It is well known
that the action of hemolysin is prevented by a higher concen-
tration of salts and von Liebermann (35), Michaelis and Skwir-
sky (386), and Hecker (37) proved that both acid and alkali
prevent the hemolytic action of hemolysin. Lisler (38) found
that in a non-electrolyte medium the complex immune hemolysin
cannot produce its characteristic effect. But as regards the
explanation of these phenomena, all of the authors are agreed
that alkali and acid and the absence of electrolytes act in such
a way as to prevent the interaction of amboceptor and comple-
ment upon the corpuscles.
On the basis of his observation, however, that the combined
antigen and amboceptor are separable in alkaline reaction, Ron-
doni explains the above mentioned inhibitory phenomena as
follows: those agents which prevent the hemolytic action of
126 M. KOSAKAI
hemolysin act upon the combination of antigen and amboceptor
instead of the combination of the latter and complement. This
explanation appears to be correct, in the light of my own observa-
tion, that the sugar solution or more correctly the non-electrolyte
medium can separate the union of hemolytic amboceptor and
its antigen.
But Rondoni’s statement that the preventive power of alkali
is the same at 0°C. as at 37°C., cannot hold for other agents,
because my experiments show that at the higher temperatures the
antigen-amboceptor union is better separated than at lower
temperatures. Indeed it seems likely, that the other inhibitory
agents also will act more effectively at higher temperatures and
this belief is supported by the observation of Landsteiner, that
the quantity of agglutinin separated from agglutinated bacteria
is proportional to the temperature applied.
The general conclusion appears justified that the combination
of antigen and antibody of any kind is more or less influenced
not only by the quality of the medium, but also by the degree of
temperature applied.
2. Reversibility of antigen and amboceptor union
Though it had been proved that the union of antigen and anti-
body is reversible, as I have already stated, it had not yet been
determined whether that reversibility is complete or partial.
The experiments of von Liebermann and Fennyvessy have no
bearing on the question of reversibility and those of Pietro
Rondoni tell us nothing as to the degree of the reversibility.
(See, also, the work of Morgenroth (39). )
My own experiments demonstrate that the reversibility of
antigen and amboceptor is almost or quite complete, so far as
immune hemolysin is concerned.
Those antigen antibody unions, of which the reversibility was
proved only in vivo, would probably be found separable also
un vitro, if the quality of medium and its temperature were suit-
ably modified.
a7 ots
A STUDY OF IMMUNE HEMOLYSIN 127
3. Indispensability of salt in the production of immune reactions
It is quite evident that salt is indispensable to the carrying
out of the biological reaction between immune hemolysin and
red corpuscles. If this fact is taken together with the observa-
tion of Bordet (40), that the agglutination of bacteria with spe-
cific agglutinin does not occur in a medium free from salt, it may
be said in general that an immune body can not exert its charac-
teristic effect without salt. As to the reason for this, I should
like to suggest that salt mediates the combination of antigen
and antibody.
4. Difference between normal and immune antibodies
This question is an old one and it has been often discussed and
the difference between the two sorts of antibodies has been shown
by several workers; for example, by Shibayama (41) by means
of dialysis of hemolytic serum, by Eisenberg and Volk (42)
from the standpoint of absorption of agglutinin and by Land-
steiner and Reich (43) from the standpoint of separability of
hemagglutinin and blood cells.
To this evidence may be added the experiments of Bail that
I have already referred to dealing with the separability of the
hemolysin and blood cell combination. In view of all of these
demonstrated differences between the normal and the immune
antibodies, it must appear doubtful that the normal hemolysins
are merely increased by immunization with red blood cells.
5. Concentration of the immune hemolysin
Any method of isolation and concentration of immune bodies
must be, in its principle, simple and certain and also there must
not be any loss of immune body during the manipulation. The
method of Rondoni is not applicable to a concentration because,
with that method, a complete reversibility of the antigen and
amboceptor union was not shown to be attainable. Though the
method of von Liebermann and Fennyvessy was heretofore the
most convenient, these authors succeeded only with the hemoly-
sin against pig’s blood and there was necessarily some loss of
hemolysin, because they employed a very strong hydrochloric
128 © M. KOSAKAI
acid, which is able to destroy hemolysin even in its weaker
concentration.
As for my own method, since the reversibility is complete,
there is no loss of immune body and though the method has
hitherto been applied only to the immune hemolysin of rabbit
against sheep’s blood, there can be no doubt that it would pro-
duce the same results with other hemolysins.
6. The biological properties of pure rmmune hemolysin
It has been shown that hemolysin does not pass through animal
membranes and that it is not susceptible to ether; but as to its
chemical nature or other physical properties further experiments
are required. As, however, von Liebermann and his co-worker
say, it is certain that hemolysin does not react so sensitively
to the protein reagents such as sulphosalicylie acid or NaCl +
acetic acid as is the case with the known protein. Nevertheless
I hesitate to assert now that immune hemolysin or immune sub-
stance generally are of non-proteid nature.
7. The pure isolation and concentration of other immune bodies
As to the isolation of agglutinin or bacteriolysin a similar
method will probably be available, though such attempts have
succeeded hitherto chiefly in vivo. If the antigen is not formed;
e.g., toxin, the purification of the corresponding antibody can-
not be accomplished by the method under consideration.
If antitoxin is of non-proteid nature, then, with the fractionat-
ing method when the serum globulin is precipitated, it is possible
that the immune bodies are adsorbed by the surfaces of smallest
particles of globulin by a mere physical process so that a separa-
tion of the antitoxin might be expected, though this was at-
tempted by Pick with the use of alcohol and other chemical
agents with negative result.
As the works of Zunz and others show, it is difficult to separate
immune bodies only, after they have been adsorbed by inorganic
surfaces, because if the active substance is separated again,
some serum protein is separated with it. The study of the separ-
ability of immune substance from inorganic surfaces has, there-
A STUDY OF IMMUNE HEMOLYSIN 129
fore, a different biological interest from the one that engaged our
attention in this paper.
SUMMARY
1. The reversibility of the antigen and amboceptor union is
proved to be practically complete, so far as the immune hemolysin
of the rabbit against skeep’s blood is concerned.
2. The isolation of hemolytic amboceptor from its antigen
union is accomplished by a simple method: When the hemolytic
power of the original immune serum is 1: 10,000, it is diluted to
100 times its volume with physiological salt solution; 5 ce. of this
diluted serum is poured into 4 cc. blood cells, which are washed
free from serum protein with physiological salt solution. A/ter
15 to 20 minutes at room temperature all of the hemolytic ambo-
ceptor has been adsorbed by blood cells; and the antigen-ambo-
ceptor union is thus obtained. After the sensitized corpuscular
sediment is washed with physiological salt solution several times,
till the last trace of serum protein has been removed, this pure
antigen-amboceptor combination is mixed with an isotonic or
shghtly hypertonic watery solution of saccharose, glucose or
lactose and left at 55°C. for fifteen to thirty minutes, during which
time the vessel is shaken several times. The sugar extract,
which contains nearly all of the hemolysin used to sensitize the
cells, is obtained by centrifugation.
3. In order to purify this sugar extract, which contains sub-
stances from destroyed blood cells, it is placed in a separatory
funnel and shaken for one to two hours with 5 to 10 volumes
of ether, this treatment, if necessary, being repeated twice or
thrice, till at last the solution becomes quite colorless. This
colorless solution is dialysed in parchment against running
water in order to eliminate the sugar and traces of ClNa.
4. The solution thus obtained is concentrated in vacuo to the
required volume.
I wish to express my indebtness to Professor 8. Mita for the
facilities and encouragement, which he has extended to me in
carrying out these experiments.
130 M. KOSAKAI
REFERENCES
(1) Pick: Beitrige z. chem, Physiol. u. Pathol., 1901, 1, 351.
(2) BrreGER AND Krause: Berliner klin. Woch., 1907, Heft 10, p. 30.
(3) Gipson: Journal of Biol. Chem., 1906, 1, 161.
(4) Gipson AND Co.tuins: Journal of Biol. Chem., 1907, 3, 233.
(5) Frovurin: Compt. rend. d.1. Soc. Biol., 1908, 65, 444.
(6) Buswip: Centralblatt f. Bakt., 1892, Bd. 12, p. 287.
(7) Ernst, CooLipGe AND Cook: Journal Boston Med. Soc., 1898, 2, 166.
(8) Hara: Centralblatt f. Bakt., 1909, Abt. I, 48, 203.
(9) Préscuer: Miinch. med. Woch., 1902, p. 1176.
(10) BeLFANTI AND CaRBONNE: Centralblatt f. Bakt., 1898, 23, 906.
(11) ATKInson: Journal of Exper. Med., 1901, 5, 67.
(12) Mout: Zeits. f. exp. Path. u. Therap., 1906, 3, 325.
(13) Hurwitz AND Meyer: Journ. of Exper. Med., 1916, 24, 515.
(14) Joacuim: Archiv f. d. ges. Physiol., 1903, 93, 558.
(15) GLAssNER: Zeits. f. exp. Path., 1905, 2, No. 1.
(16) Catmerte: Annales d. |’Instit. Past., 1895, 9.
(17) von WASSERMANN: Zeits. f. Hyg., 1896, 22, 263.
(18) Morcenrotu: Minch. med. Woch., 1903, Nr. 2, p. 61.
(19) Hann AND TrommsporF: Miinch. med. Woch., 1900, Nr. 18, p. 413.
(20) LANDSTEINER: Wien. klin. Woch., 1902, Nr. 40.
(21) LANDSTEINER AND JaGcic: Wien. klin. Woch., 1904, 17.
(22) Barn AND Tsuba: Zeits. f. Imm., 1909, 1, 546.
(23) SpaEet: Zeits. f. Imm., 1910, 7, 712.
(24) Tsupa: Zeits f. Imm., 1909, 2, 225.
(25) Brtrz, Mucu anp Srepert: Beitr. z. exp. Therap. v. Behling, 1905, H. 10,
p. 30.
(26) ANDREJEW: Arb. aus d. kaiserl. Gesundh., 1910, 33.
(27) LANDSTEINER AND ReEicH: Centralb. f. Bakt., 1905, 39, 83.
(28) JAguE AND Zunz: Arch. intern. d. Physiol., 1909, 8, 227.
(29) SsBpoLew: Zeits. f. Imm., 1912, 13, 507.
(80) Barn AND Rorky: Zeits. f. Imm., 1913, 17, 566.
(81) Baru: Zeits. f. Imm., 1914, 21, 202.
(32) von LigEBERMANN AND FENNyvEssyY: Centralb. f. Bakt., 1908, 47, 274.
(383) Ronpont: Zeits. f. Imm., 1910, 7, 515.
(384) Morcenrotu: Berl. klin. Woch., 1905, No. 50.
(35) LIEBERMANN: Biochem. Zeitschr., 1907, 4, 25.
(36) MicHarELis AND SKwirsky: Zeits. f. Imm., 1909, 4, Heft 5, p. 357.
(37) Hecker: Arb. a. d. Inst. f. exp. Ther. i. Frankfurt a. M. 1907, Heft 3.
(38) Ester: Zeits. f. Imm., 1909, 2, 159.
(389) Morcenrotu: Miinch. med. Woch., 1903, Nr. 2.
(40) Borpret: Ann. d. |’Inst. Past., 1899, 13, 225.
(41) Suipayama: Centralb. f. Bact., 1902, Nr. 30.
(42) EISENBERG AND VOLK: Zeits. f. Hyg., 1902, 40, 161.
(43) LANDSTEINER AND ReicuH: Centralb. f. Bakt., 1905, 39, 712.
A NEW METHOD OF ESTIMATING THE ANTITRYPTIC
INDEX OF BLOOD SERUM
T. BRAILSFORD ROBERTSON anp SAMUEL HANSON
From the Department of Biochemistry and Pharmacology, Rudolph Spreckels
Phystological Laboratory, University of California
Received for publication February 18, 1918
I, INTRODUCTION
The methods that are at present employed for the estimation
of the antitryptic indices of blood sera yield only an arbitrary
measure of antitryptic power, which bears no necessary relation-
ship to the actual quantity or proportion of trypsin that is in-
activated by the serum. ‘Thus in the Gross-Fuld casein method
(1) the “antitryptic index’’ is expressed by the concentration of
trypsin that is requisite to overcome the inhibiting action of a
given quantity of serum to the extent of permitting the complete
hydrolysis (removal of substances precipitable by acetic acid)
of a given quantity of sodium caseinate in a given time. In
the Loeffler plate method (2) the ‘“antitryptic index’”’ is esti-
mated in terms of the concentration of trypsin that is required to
overcome the inhibiting action of a unit volume of serum to the
extent of permitting some digestion of heat-coagulated blood to
occur in a given interval of time. In the method advocated by
Mintz the antitryptic index is similarly estimated as the per
cent of a given trypsin solution that is Just able to overcome
the “paralyzing” action of an arbitrary unit of blood serum (3).
While each of these methods yields results of qualitative value,
they are not comparable with one another, nor do they yield any
quantitative measure of the relative trypsin-inactivating powers
of different sera. They would do so were (a) the quantity of
protein digested in a given time always directly proportional to
the concentration of trypsin acting upon it, and (b) the quantity
131
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2
IBY T, BRAILSFORD ROBERTSON AND SAMUEL HANSON
of trypsin bound by serum directly proportional to the quantity
of antitrypsin that it contains. Both of these assumptions
are, however, demonstrably invalid.
In the first place the relationship between time of digestion
and the quantity of protein digested is not one of simple pro-
portionality, and in fact in the case of sodium caseinate (employed
in the Gross-Fuld method of estimating the antitryptic index)
the relationship between time and degree of hydrolysis is very
accurately expressed by the logarithmic equation
a
a—-wZ
log = Kt
which is that which usually obtains in monomolecular chemical
reactions (4).
In the second place, the quantity of trypsin bound by different
samples or quantities of serum is not directly proportional to
their antitrypsin content, for in that event double the amount
of serum which suffices to halve the activity of a given concen-
tration of trypsin should reduce its action to zero. Now the
_ experimental fact is that although it is an easy matter to halve
the activity of trypsin by the addition of serum it requires a
relatively enormous excess of serum to abolish its proteolytic
activity altogether. Thus in rabbit 1 (table 1) the addition of
0.15 unit of serum neutralized one-half the specified quantity
of trypsin employed, but the addition of 0.33 unit of serum only
reduced the tryptic activity to one-third. Evidently in the inter-
action of trypsin and antitrypsin we have, asinso many reactions
involving biological antibodies (5) a “‘balanced”’ or incomplete
reaction in which the station of equilibrium is determined by the
relative masses of all of the reacting components.
Since the velocity-constant (K) of the hydrolysis of sodium
caseinate is, as usual in catalyzed reactions, directly proportional
to the concentration of trypsin (4) we have in the value of this
constant a quantitative measure of the concentration of free
trypsin in the digest. We have accordingly measured the value
of K in various mixtures of sodium caseinate, trypsin and anti-
tryptic sera with the object of ascertaining the actual quantita-
Pe
ANTITRYPTIC INDEX OF BLOOD SERUM 133
tive relationships which obtain between the trypsin and anti-
trypsin, on the one hand, and the quantity of trypsin inactivated
on the other. The following was the procedure employed.
II. PROCEDURE OF THE ESTIMATION
A number of glass tubes 12 to 14 em. in length, having an inside
diameter of 4 to 5 mm. and walls about 1 mm. thick are prepared
and sealed at one end. The open ends are sealed by means of
short lengths of sealed glass tubing of the same diameter as the
tubes and inserted into short lengths of rubber tubing.
The following are the reagents required:
SUCMUMM IV VCNORIME) 25.) 5 sie each behets ee ee 0.048 N
PRGCEUMCEA CIM 8th Ga ls 5 2 2 oe eaic So hou cs batieaonelo See eee eee 0.600 N
Soelavi CM OPIG OL A 2.5555). 8 dsr sicninceste AES A ae 6.0 per cent
SO CMUIMECAT ON AUC. bars ci. oyeu clsemisrs sane eae SEER oe ee acs 0.01 per cent
Ae HEIN ea ge once erie hcaicie, § vkeapapavades. > sae ts Ooh e Pee eee 6.0 per cent
Peaysin (GTUeblers)> 4s s.r). Was es setsa co shhgne dase aan 1.0 per cent
Serum diluted with 6.0 per cent sodium chloride
The acetic acid is prepared with sufficient accuracy by diluting
30 cc. of glacial acetic acid to one litre. This is kept as a stock
solution.
The casein solution is prepared from commercial casein (Himer
and Amend’s or Merck’s C. P. “Nach Hammarsten’’) specially
purified by washing in water, aleohol and ether (6). One and
a half grams of the casein powder is introduced into a 50 ce. flask
and 25 ce. of the 0.048 N sodium hydroxide solution is added.
The casein gradually dissolves with the aid of frequent shaking
and gentle warming. When clear and homogeneous it is filtered
to remove any particles which may not have completely dissolved.
The solution thus prepared is neutral to phenolphthalein ; it must
be freshly made up on the day on which it is to be used.
The trypsin solution was made by dissolving Gruebler’s trypsin
in 0.01 per cent sodium carbonate.
In preparing the diluted serum, fresh clear and only very
slightly haemolyzed serum was employed. The serum and the
requisite amount of 6 per cent sodium chloride to dilute it to the
desired concentration are introduced into a glass tube, a glass
bead dropped in and the mixture shaken thoroughly.
134 T. BRAILSFORD ROBERTSON AND SAMUEL HANSON
These various solutions having been prepared, the estimation
of the antitryptic power of a given dilution of serum jis carried
out as follows:
Three of the glass tubes are labelled A, B and C. A glass
bead is dropped into each of these. Into A is introduced 0.125
ec. of the diluted serum and into B and C 0.125 ce. of 6 per cent
sodium chloride solution. One cubic centimeter of the 6 per cent
casein solution is then added to each of the tubes and 0.125 ce.
of the 1 per cent trypsin solution. The tube labelled C is im-
mediately acidified by the addition of 0.125 cc. of the 0.6 N acetic
acid which stops all tryptic action and precipitates the casein.
The contents of all the tubes are mixed thoroughly by affixing
a stopper and shaking. The tubes A and B are now incubated
at 36°C. for one hour. After removal from the incubator these
tubes are acidified in the same way as C and again stoppered and
shaken. The liquid and precipitate in each tube are separated
by centrifugalization and the refractive index of each of the
fluidsis determined at the same time (i.e., at the same temperature)
in a Pulfrich Refractometer, the angle of total reflection being
read to within one minute, a sodium flame being the source of
light.
It has been previously shown that the refractive index of a
solution of the mixed products of the tryptic hydrolysis of casein
is identical with that of the original solution of casein from which
the products were derived (7). Hence the difference between
the refractivities of the fluids derived from the tubes B and C is
proportional to the amount of casein digested in B by the action
of the trypsin during the period of incubation. One gram of
casein or its hydrolysis products dissolved in 100 cc. of dilute
alkali changes the refractive index of the solvent by 0.00152.
Hence, dividing the difference between the refractive indices of
the fluids derived from B and C by the factor 0.0152 we obtain
the percentage of casein digested by the trypsin (7).
The difference between the refractivities of the fluids derived
from A and C similarly yields the percentage of casein digested
by the trypsin which has not been bound by the antitrypsin of
the serum. It is necessary to recollect, however, that the glob-
ANTITRYPTIC INDEX OF BLOOD SERUM 135
ulins of the serum are precipitated along with the casein by the
addition of the acetic acid.
It is for this reason that 6 per cent sodium chloride is employed
as the diluent of the serum, for it is found that on the average,
and within the experimental error of the determination, the
refractive index of serum from which the globulins have been
precipitated by acetic acid is equal to that of a 6 per cent solution
of sodium chloride. Hence by using this concentration of sodium
chloride as diluent the refractivities of the fluids derived from the
tube containing serum and from those containing 6 per cent so-
dium chloride in the place of serum are equally affected by the
respective additions.
The difference between the refractivities of the fluids from the
tubes B and C, therefore is a measure of the hydrolysis attribut-
able to the total amount of trypsin added, acting for one hour
at 36°C., while the difference between the refractivities of the
fluids from the tubes A and C is a measure of the hydrolysis due
to the proportion of the trypsin that remains unbound by the
antitrypsin of the serum, acting for the same period and at the
same temperature.
The order followed in mixing the reagents in the tube should
not be changed. If the trypsin and serum are mixed before the
casein is added, irregular and untrustworthy results are obtained.
We have thus measured the relative hydrolysis in the presence
and in the absence of the antitrypsin of the serum. We proceed
to ascertain from these measurements the actual proportion of
trypsin bound by the serum by computing the velocity constant
of hydrolysis in the presence and absence of serum by applying
the monomolecular formula
a
logy == eh
Oh RTA
where ¢ is the time in hours, a is the initial percentage of casein
and z is the percentage of casein that has been digested. The
value of K thus computed is directly proportional to the con-
centration of active (unbound) trypsin in the digest.
136 T. BRAILSFORD ROBERTSON AND SAMUEL HANSON
III. EXPERIMENTAL RESULTS
Illustrative results are shown in the accompanying tables in
which A denotes the dilution of serum employed, the undiluted
serum being taken as unity, and 7 denotes the proportion of the
trypsin computed to have been neutralized by the serum.
TABLE 1
Rabbit 1 Concentration of casein solution 4.36 per cent
= > 3 ti =
Xx K X 108 A L i=
1.26 148 0.33 0.68 4.00
1.52 186 0.25 0.60 6.00
1.68 211 0.20 0.54 5.87
1.78 228 0.15 0.51 6.93
2.04 274 0.10 0.41 6.95
2.55 383 0.05 * 0.17 5.00
2.79 444 0.03 0.04 4.83
2.86 463 0.00 0.00
Considering the number of factors involved in the determina-
tion, there is a very evident tendency towards constancy of the
ratio The significance of this fact will be clear from
fT
A ay
the following considerations:
If we consider the neutralization of trypsin by antitrypsin to
be due to the formation of a proteolytically inactive compound,
the simplest conception we can form of the process is that one
molecule of trypsin combines with one molecule of antitrypsin
to form the inactive compound. Applying the mass-law, there-
fore, we should expect the following relationship to hold good:
ka A= eT yar) =e | :
where a is the number of molecules of antitrypsin contained in
the specified proportion of wndiluted serum, 8 is the number of
molecules of trypsin-antitrypsin compound which would’ be
formed by the neutralization of the total quantity of trypsin
employed, and k is the equilibrium constant of the reaction.
ANTITRYPTIC INDEX OF BLOOD SERUM 137
Rearranging, this equation may be written:
De
=C
(C2 B54 Ge 1)
where C is a constant which is proportional to the molecular
concentration of antitrypsin in the sample of serum employed.
The experimental fact being that Ara = T) is constant, it is
evident that 67 is negligibly small in comparison with A, or
in other words the number of molecules of trypsin-antitrypsin
compound which would be formed by the neutralization of all
of the trypsin present would be negligible in proportion to the
total number of molecules of antitrypsin contained in the serum.
TABLE 2 TABLE 3 TABLE 4
Rabbit 1. Serum taken Rabbit 2 Rabbit 3
twenty-four hours
subsequently (this A c A 4
animal was recetv-
ing injections of 0.50 3.13 0.50 3.55
tr ° 0.33 3.32 0.33 3.25
ypsin)
0.17 3.29 0.25 3.14
A Cc 0.13 2.10
0.06 2.59
0.40 8.84
0.33 9.50
0.25 9.00
0.20 8.50
Evidently only a small fraction of the antitrypsin in serum is
actually engaged in binding the quantities of trypsin employed
and a very great excess of antitrypsin must be present in the
serum. We should thus be inclined to identify the antitryptic
fraction of serum with some quantitatively important fraction,
for example, as several authors have suggested, the serum
albumins (8). In the succeeding article, however, it is shown by
one of us (Hanson) that the antitryptic index of serum, as meas-
ured by the constant C, may be increased by immunization no
less than 300 per cent without any definite increase in the .per-
centage of albumins in the serum. It is evident, therefore, that
if the binding of trypsin is indeed accomplished by the albumin
138 T. BRAILSFORD ROBERTSON AND SAMUEL HANSON
fraction the process of immunization against trypsin accomplishes,
not an increase in the quantity of anti-trypsin, but an alteration
in its physical or chemical condition of such a character as to
enhance its affinity for trypsin. On the other hand it is possible
that the neutralization of trypsin is in reality accomplished by
‘some other and hitherto unidentified constituent of the serum,
which must, however, be present therein in disproportionate
quantity to the actual molecular concentration (which is of course
very small) of trypsin employed in these experiments.
SUMMARY
1. A simple and accurate method of measuring the antitryptie
indices of blood-sera is described.
2. It is shown that for varying proportions of antitrypsin
(= A) added to a specified amount of trypsin (regarded as unity)
the relation:
d fe
Age Ova
holds good for any given serum, 7’ being the proportion of the
trypsin neutralized by the serum and C a constant which is a
direct measure of the number of molecules of antitrypsin con-
tained in a specified volume of the serum.
3. The molecular concentration of antitrypsin in blood serum
is in great excess of the molecular concentration of proteolytically
active material in a 1 per cent solution of Gruebler’s trypsin.
REFERENCES
(1) R. Wertz: Arch. Intern. Med., 1910, 5, 109.
(2) BERGMANN AND Meyer: Berl. klin. woch, 1908, 45, 1673.
(3) S. Mintz: Folia Serologica, 1910, 6, 279.
(4) E. H. Watters: Journ. Biol. Chem., 1912, 11, 267.
(5) S. Arruentus: “Immunochemistry.’? New York, 1907.
(6) T. Brartsrorp Ropertson: Journ. of Physical Chem., 1910, 14, 528.
(7) T. BrartsrorD Rospertson: Journ. Biol. Chem., 1912, 12, 23.
(8) S. G. Hepin: Ergebnisse der Physiol., 1910, 9, 433.
E. RosEnTHAL: Folia Serologica, 1910, 6, 285.
THE NON-INFLUENCE OF INJECTIONS OF TRYPSIN
UPON THE PROTEIN QUOTIENT IN
BLOOD SERUM
SAMUEL HANSON
From the Department of Biochemistry and Pharmacology, Rudolph Spreckels
Physiological Laboratory, University of California
Received for publication February 18, 1918
INTRODUCTION
The relation of the serum proteins to infection and immunity
has engaged the attention of many investigators. The earlier
experimental evidence indicated that the production of antibodies
is always accompanied by an increase in the globulins. Later
work, however, has shown that immunity could be attained with-
out any change in the quantity of globulins.
Among the earlier workers may be mentioned, Hiss and At-
kinson (1), and Ledingham (2), who obtained a marked in-
crease in the globulins in the serum of horses immunized against
diphtheria toxin.
A great part of the recent work has been done by Hurwitz
and Meyer (3), and Schmidt and Schmidt (4). These investi-
gators found by extensive and varied experimentation that with
small doses of antigen carefully administered, a high degree of
immunity can be produced without a decided rise in the globu-
lins. The antigens employed by Hurwitz and Meyer (3), were
certain living bacteria, dead bacteria and fowl typhoid endotoxin.
The antibodies to all of these antigens are carried down with the
globulins on fractionating the immune sera. An antigen, how-
ever, whose antibody is known to be in the albumin fraction
has not been employed by Hurwitz and Meyer, Schmidt and
Schmidt, nor, to the writer’s knowledge, by any previous in-
vestigators. According to Kimmer and Mogulesko (5), pan-
139
140 SAMUEL HANSON
creas trypsin and yeast trypsin are antigens of this sort. It was
of interest to find whether or not a carefully produced immunity
to this rather unique type of antigens is associated with a
change in the protein quotient. This is the problem that incited
the undertaking of the present work.
METHODS AND MATERIALS
The experimental conditions were in general similar to those
described in a previous communication (6). Seven normal
rabbits were selected as the experimental animals. Determina-
tions of the serum proteins and of the antitryptic index were
made through a prolonged fore-period. At the end of the fore-
period the immunization of six of the animals was begun, while
one served as a control.
Solutions of Griibler’s pancreas trypsin puriss. sicc., freshly
made up and filtered was the particular trypsin employed.
The strength of the sample of this trypsin used was such, that
the velocity constant of digestion of an 0.38 per cent solution freshly
made up and filtered, acting in an equal volume of a 4.40 per
cent sodium caseinate solution at a temperature of 37°C., was
found to be 36 X 10-2 by Robertson’s refractometric method (7).
The same sample of trypsin unfiltered showed only a slightly
higher potency.
This enzyme was administered subcutaneously and intra-
venously in order that the slower and more rapid effects, re-
spectively, may be elicited. The quantities and frequency of
injection were also varied in the different rabbits so that the
optimum rapidity of immunization may be at least more or less
closely approached. In no case were the doses used so large
as to cause severe reactions or even a loss of weight in the animals.
The determinations of the serum proteins were made by
Robertson’s micro-refractometric method (8).
The estimations of the antitryptic index were carried out by
the method described in the preceding article (9). The prin-
ciples on which this method is based are briefly as follows:
Since it is a fact that the refractive index of a sodium caseinate
NON-INFLUENCE OF INJECTIONS OF TRYPSIN 141
solution is practically not altered by tryptic digestion, it is
possible to estimate the extent of hydrolysis in such a solution
containing trypsin by precipitating the undigested casein with a
definite volume and strength of acetic acid, centrifuging down the
precipitate and determining the refractive index of the super-
natant fluid, which is a solution of the products of hydrolysis (6).
The velocity constant (&) of the trypsin in question is then
calculated from the monomolecular formula,
a
a—-xz
= Kt
log
where a is the initial concentration of the substrate and x the
amount digested during a period of time f.
The velocity constant (k:) of the rate of hydrolysis is then
obtained in a similar manner for a casein-trypsin solution to
which a known quantity of serum has been added.
Since K is the velocity constant of the rate of normal hydrolysis
and K, is the velocity constant of the rapidity of digestion after
a part of the trypsin was neutralized by the antitrypsin of the
serum,
* a = fraction of trypsin unneutralized
x
It has been observed (9), that the trypsin-antitrypsin reaction
is reversible; hence it is practicable to calculate the antitryptic
index (C) from the equilibrium equation,
x —
ies
where x = the fraction of trypsin neutralized and s the amount
of serum employed.
DISCUSSION
The experimental results presented indicate that the normal
variation in the antitryptic index is on the average approximately
50 per cent in any particular rabbit and is considerably greater
in different animals. Rabbits 1 and 3, and to a lesser extent
142 SAMUEL HANSON
TABLE 1
Rabbit 1. Weight; July 17, 2862 grams; July 28, 2798 grams
z | Qo ids eco al
: Oe a pak ta
TREATMENT a g 3 ae 3 a ae
Pa) Z < ic} e ic} my ie)
per per per per per
cent | cent | cent | cent | cent
Normal: fae tis sos, 5d 2 eter eee (=) 3) 122), 4.03) 125, I> bale 2 falOneng
IN oye) |. SEe a ee Mee 2 cit f 511.3 | 4.2 | 1.8 |.6.0) 380 | 0:43))2%78
No rina all eresesy sisson SNe ote ere 7] 7 1-0 | 4.1 | 1.8] 5.9 | 30') 0-44)%3200
INOTMIAL Ass 2st bie s site se bce a oe C—O 27325 | 127 | 5.27) 33) OF48
Normal Seer oe on hse Cee 715) 125) 4:0: 1.2), 5:2 |) 23°) Os30aeas
0.2 gram trypsin subcutaneously
onstheml 6th’ S.24 sce. 8 seen CAT 123713287 | WAN 2 27) Okan led
0.3 gram trypsin subcutaneously
ongtherl Oth. Fee 4 eee 718) 1.4°)-4.1/] 1.5,|.5.6 | 23 .| 0-86) (3262
' 0.4 gram trypsin subcutaneously
Olgtteeae Gh dae- ee se ae eee 7-20) 1.67] 3:95] 1-35) 5.3] (26 0734) 78201
0.6 gram trypsin subcutaneously [| 7-23) 1.4] 3.5 | 2.8) 6.3 ke 0S
Aas OR SORE Bt {| 7-26] 1.3 | 3.2| 1.7] 4.9 | 35 | 0.50) 2.70
‘i, Sb. geet. {| 8- 2| 1.3) || 4:0) || 2.2 | 6.29) 5: Ors5ip2eso
TABLE 2 7
Rabbit 2. Weight: July 17, 2894 grams; July 28, 2724 grams
Zz 3 so | ¢6
5 4 a p
SL ce) BE? aloe aieS
TREATMENT = 3 5 ae A 2 alae Z
z Zz & 6 | s#16e8! oe
S z 2 5 |e *18 = om
per per per per per
| cent | cent cent cent cent
IN Oriiiaal 7.2 24.) 2 eee eee 7.511.3| 4.2] 1.8] 6.0] 30.0] 0.43) 2.78
Normals 36724 5 hes ten Cre eee VT 1-34) 420 | 125 | 5.5 | 270) 0s37|e2e22
Normal .:-7 22. 4has on ae ee ae 7-9 | 1.4] 4.1 | 1.8] 5.9 | 30.5) 0.44
Normal. 22 52'.6 sce Soe ee 7-15] 1.4 | 4.5 | 1.4 | 5.9 | 24.0] 0.31] 2.55
0.2 gram trypsin on the 16th )| 7-17| 1.4 | 4.7] 1.2] 5.9 | 20.0) 0.26] 1.90
subcutaneously 7-18] 1.6 | 4.1-| 1.3 | 5.4 | 24.0] 0.31) 2.14
0.3 gram trypsin on the 19th
(subeutaneously)................ 7-20) 1.4 | 3.4] 1.5 | 4.9 | 31.0) 0.44) 3.99
0.5 gram trypsin on the 22d
(subcutaneously,)\.c--- +20 oc: 7-23] 1.5 | 4.0 | 1.4 | 5.4 | 26.0) 0.35] 5.00
0.6 gram trypsin on the 25th \ 7-26) 1.3 | 3.6 | 1.7 | 5.3 | 32.0) 0.47] 3.06
(subcutaneously) | $2 | 1-2 478+] 1.2 |-6.0.| 20.0|.0-251) 185
NON-INFLUENCE OF INJECTIONS OF TRYPSIN 1438
TABLE 3
Rabbit 3. Weight: July 17, 2610 grams; July 28, 2383 grams
Z } 8d | 6
: a gee
di TREATMENT S a g fe 3 Ze
4 z, 3 elke ol chanel OE eels
per per per per per
cent cent cent cent cent
NOTIN ANIMAS Sls ie MA toktsh a cteeteno- 73) Vel 2n ASL el eOn i GuOn|Ro2 00846
IN| OTST IA LA a A iO (Adal Ae Oe liso Gaoole2o.0)) Oxssiecents
MSCOTTVA RP Vet sh, de Nas loi ds sestle te de C0 Whe 3s) 404 IS G2 22950) 041193200
INIGIETED GLA, Pe ee Oe CO PALA ASS ale 18h | Ooi |aeo ele O.42
DOIG se ae a 7-15, 1 4.| 4.5 | 1.25) 5.75). 22.0) 0.27) 3.25
0.1 gram trypsin on the 16th (sub-
EMPANeOUShY.). 0. 300.8. EL oe. 7-17 1.2 | 4.3 | 1.3.) 5:6 | 23.0) 0.32) 2.35
Same dose on the 17th and 18th. .| 7-18) 1.4 | 4.1 | 1.7 | 5.8 | 29.0) 0.41) 3.12
| 7-20] 1.4 | 4.3 | 2.0] 6.3 | 32.0) 0.46] 9.00
Daily on the 19th to 25th inclu- || 7-23) 1.2 | 3.9 |.1.9 | 5.8 | 33.0) 0.49] 8.61
sive 0.2 gram subcutaneously 7-26) 1.1.) 4.0%). 2.2) 1 6.24), 35.0) 0..55)'4580
|| 8-2 | 1.2 | 4.8 | 1.6 | 6.4 | 25.0].0.33) 2.81
TABLE 4
Rabbit 4. Weight: July 17, 2270 grams; July 28, 2383 grams
z oie aie
gq ° & q ro)
Sue eran Vet a deta OR
TREATMENT r FI : 5 3% Bz Bz
2-2 po lee [BE PR ee
per per per per per
cent | cent | cent | cent | cent
Normans: See 4 hee. Ps bee ed 7-3"|\ 1.3 | 4.25) 1°25) 5.5) | 24.01.0229
IN/@RTITGI |S Chol cle RO ORES eo ae Eee eae 7-5 | 1.8 | 4.4 | 1.38 | 5.7_| 23.0) 0.23) 2.50
NO UnI Dees eB te Se Bich ceed 7-7 | 1.3 | 4.0 | 2.1 | 6.1 |, 34.0) 0.52) 2.22
OTM SO eo Ras, hee es 7-9 | 1.05) 3.9 | 1.4 | 5.38.| 26.5) 0.36
IN[@ TTC Aa eee eee ere 7-15) 1.4 | 3.9 | 1.0 | 4.9 | 20.0) 0.26) 3.25
0.1 gram trypsin on the 16th (in-
UAVEMOUSIN) IS. Get ocitede ace sh. *.| 7-17) 1.4 |, 3.7 | 1.8 | 5.0 | 26.0) 0.85) 3.52
Same dose also on the 17th and 18th) 7-18) 1.3 | 3.8 | 1.2 | 5.0:) 24.0) 0.32) 2:93
Daily 19th to 23d inclusive 0.2 |
SRAM ee ae es. Scie: Bo tet ieee cies ol ae 7-20) 1.5 | 4.3 | 1.1 | 5.4 | 20.0} 0.26) 4.25
~ (| -¥-23)-4.3 | 4.0 | 1:7 5.7 |-30:0) 0.42) 4°48
eipavh and 25th 0.4 gram (1n~ )| 7 96|.1.2.|/4.3°) 1.87 | 610 | 30.0) 0.42) 3606
Beayenously) || $2 | 1.2] 4.6| 1.3 | 5.9 | 22.0] 0.28] 1.88
144
SAMUEL HANSON
TABLE 5
Rabbit 5. Weight: July 17, 2100 grams; July 28, 2270 grams
Zz
5 Zz
[e} Zz =
TREATMENT Fi 5 8
E Zz Bil 33
< ° re) n
a Z < oS
per per per
cent | cent | cent
INformr al ay seb fas: e255. icrottsys Soferstel te (=o (ALS | 3.94) 1.6
Norra eee, rs Po eee teeta 7-5 | 1.3 | 4.4) 1.7
BN OTT MUSES get ease 2 acc volt aie ake ie wi Rare (ei Nl 3 4)0430) 6
INGE ANE Ee Ws cleo «tists ais ths coger 0-9) 10) 455 | 123
Normale 2h. is onc ae weave eros 7-15) 1.3 | 4.4 | 1.2
INormraley g4.ti.5 00 toi ens sree es i ee: a UR 74
INorinale etry: acc cats seeiasaeeee 7-18) 1.5 | 3.8 | 1.3
INormMaltes -frs28\. «Rite. Sees b—201 155 450110
Normale Se 327. 0, ...tkelo sesiose eerie 7-23) 1.3 | 4.0 | 1.7
INOrmal ee. a.%50) oe tnise ka eee 7-26] 1.2 | 4.3 | 2.5
INOKNIAL. £6le sais «artnet se bea oe Safe) 2) | 465d
TABLE 6
Rabbit 6. Weight: July 17, 2383 grams; July 28,
s
BP ele
TREATMENT = = 3
& = p
Bee ah car |
< 2} | I
i=} Zz < S
per per per
cent | cent | cent
Normal. Ja% tn cere nee cis Gao tele Agel lees
Normal. (Sti: 7G wer ee mee oe aoe es 7-5 SN TAO)
INOrMal%, scenes cS Nae eee ee tN le2 | Asses Da
INommal tian aay fot aren pmo: 7-9 | 1.4 | 4.3 | 2.4
Normal 7>. etek fc eee seek ake eee 7-15} 1.3 | 4.8 | 2.0
0.1 gram trypsin on the 16th (in-
ETAVYENOUSLY) J .4.02 ee 12 een oe (Wh 12 VARS 15
Also same dose daily to the 23d || 8-18) 1.3'| 3.8 | 1.7
inclusive AZO) 5 ofc eich Lo
7-23] 1.3 | 4.4 | 2.2
On the 24th and 25th 0.2 gram... 7-26] 1.3 | 4.6 | 2.2
8-2) (°1.3, | 4.7 | 1.6
2 [53/8
2x |22z| BS
pF EFA BF | o
per per
cent cent
5.5 | 29 | 0.41
6.1 28 | 0.39
56 29 | 0.40} 3.99
5.8 22 | 0.29
5.6 21 | 0.27) 4:88
tise 23") (0) 29\fenoe
peal 25 | 0.34!) 2.93
5.0 25 | 0.25] 4:69
edl 30 | 0.42) 5.00
6.8 | 37 | 0.58] 2.58
nerf 20 | 0.26] 1.85
2440 grams
a4 Ze
ea chen ay)
per per
cent cent
6.5 28 | 0.38
(fell 28 | 0.39] 3.99
6.5 34 | 0.49) 3.32
6e7 36 | 0.56
6.8 29 | 0.42) 3.82
5.8 26° 0535| 158
5505) 31 | 0.45) 2.26
i; Tee See 5.86
6.6 33 | 0.50) 4.43
6.8 | 32 | 0.48) 2.70
6.3 25 | 0.34) 0.95
rs
NON-INFLUENCE OF INJECTIONS OF TRYPSIN 145
TABLE 7
Rabbit 7. Weight: July 17, 2610 grams; July 28, 2472.grams
z eae les
g io] & p
eB Zz i] oO o
TREATMENT 2 iS a ces] Ze
- mt EB BP | 3az|Psz| 84
é z ea 6 | #8 |oos8| Sz
BS z 2 oh SS 3 6)
per per per per per
cent cent cent cent cent
Sooner 7-8: | 1 J) Aeoelenes, eseGnl SO. | Oat
INC Ch TATU fers ciaci na, 5. stone Sse She sav om 7=5-| LeQuiv4acoailesaleono 28 | 0.40) 2.35
INL OMIA ly Sky See ee ea eee (HE Waterers OSU eects F 2222
load eee 7-9.| 4.0 13:9) | i730: |o144
“"oouid) «2a 7-15| 1.3 | 4.2] 1.6] 5.8] 28] 0.40] 1.80
ily 0.1 rypsi x
- see a aa 7-17| 1.4| 3.7| 2.1|5.8| 36 | 0.57] 2.35
a . 7-18| 1.6 | 4.0] 2.0| 6.0] 33] 0.50] 1.50
clusive |
| 7=20| 1:.4:| 3.7 | Le90le5:6 1) 84.) 0.51) 2255
7-23| 1.8 | 4.0 + 2.0 6.0] 33 | 0.50) 2.98
On the 24th and 25th 0.2 gram 7-261 1.3|3.9| 2.41 6.3| 38] 0.61] 2.05
[| $2] 1.2] 4.3] 2.1] 6.4] 33 | 0.49) 0.95
rabbits 2, 4 and 6, developed a certain degree of immunity to
trypsin as shown by the rise in the antitryptic index. Rabbit 5
(the control) showed also a rise in the antitryptic index. Al-
though this increse is considerably less than that shown by
rabbits 1 and 3, it is yet beyond the limits of normal variation.
Rabbit 7 failed to produce any immunity.
Very interesting is the phenomenon that soon after the first
one or two injections there is a marked fall in the antitryptic
index. This decrease in the antitrypsin was possibly due to a
neutralization of the antitrypsin by the trypsin injected, without
an accompanying equal regeneration of the former for a time.
The injections of trypsin in the doses used is not followed by
any decided change in the protein quotient. This fact is in
harmony with the negative results ofaimmunization against other
antigens, cited at the beginning of this article.
SUMMARY
1. The normal variation in the antitryptic index is very marked
in any particular rabbit, and is even greater in different animals.
146 SAMUEL HANSON
2. An unmistakable rise in the antitryptic index has been pro-
duced in at least two of the six rabbits immunized against trypsin.
3. The progress of immune body production against trypsin
is very peculiar, in that at first there is a marked fall in the
antitrypsin, followed soon, however, by rather a sudden rise which
is at best only 300 per cent above the normal variation. At this
level the antibody content remains approximately stationary
for a comparatively brief period and is then supervened by a
rapid return to the normal, in spite of continued periodical in-
jections of trypsin.
4, Immunity to pancreas trypsin appears to cause no change in
the protein quotient. This fact may serve as additional evidence in
favor of the recently emphasized view that immunity 1s non-depend-
ent on the concentration of the serum proteins.
REFERENCES
(1). Hiss, P. H., anp ATKrinson, J. P.: Jour. Exper. Med. 1900-1901, 5, 47.
(2) Lepinenay, J. C. G.: Jour. Hyg., 1907, 7, 65.
(3) Hurwitz, S. H., anp Meyer, K. F.: Jour. Exper. Med., 1916, 24, 515.
(4) Scumipt, E. S., anp Scumipt, Cart L. A.: Jour. of Immunology, 1917, 2, 343.
(5) Kamer, H., anp Mocutesxo, J.: Z. f. Immunititsf., 12, 16-28.
(6) Hanson, S., anp McQuarriz, I.: Jour. Pharm. and Exper. Therap., 1917, 10, .
261.
(7) Rospertson, T. Braitsrorp: Jour. Biol. Chem., 1912, 12, 23.
(8) Rospertson, T. Brattsrorp: Jour. Biol. Chem., 1915, 22, 233.
(9) Ropertson, T. BrarsrorD, AND HANson, SAMUEL: Journal of Immunology,
1918. 3. 131. ;
EFFECTS OF INTRAVENOUS INJECTIONS OF A
COLLOID (GELATIN), UPON RABBIT SERA
GUY W. CLARK
From the Department of Biochemistry and Pharmacology, Rudolph Spreckels
Physiological Laboratory, University of California
Received for publication February 28, 1918
The general purpose of the following experiments was to
study the effects of intravenously injected gelatin upon the non-
proteins and proteins in normal rabbit sera. The special phases
considered were:
1. The effects of the injected gelatin upon the ratio of the
serum albumins and globulins and upon the total amounts of
proteins and non-proteins.
2. The time required for the gelatin to disappear from the
blood stream.
3. The effects due to the gelatin attracting fluids from the
body tissues, resulting in an excessive dilution of the blood
(hydremic plethora).
Four medium sized female rabbits (2.0—2.5 kilo), were used as
experimental animals. For several days before and throughout
the experimental period these animals were under uniform con-
ditions as to feed, water, etc.
The micro-refractometric method of Robertson (1) was used
for the determination of the albumins, globulins and non-pro-
teins. This method makes it possible to obtain accurate results
with small amounts of serum, which is necessary when using
the smaller experimental animals. The blood was removed
from the vein of one ear and the injections made into a vein of
the opposite ear. The analysis of the serum was made on alter-
nate days, injections being given every day during the experi-
mental period. The analytical work was carried out under
147
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 2
148 GUY W. CLARK
uniform conditions, about six hours being the average time for
completing an entire set of determinations.
Gelatin was selected as a colloid suitable for injection be-
cause it is non-toxic and is not precipitated by N/50 acetic acid
and subsequent boiling. It therefore appears quantitatively
among the non-protein constituents. The chief disadvantage in
using gelatin arises from the fact that the term ‘“‘gelatin” has
only a relative meaning. Nelson’s ‘‘Gold Label” gelatin was
selected for this work but before making up the solutions it was
placed in a desiccator over CaCl, and allowed to dry several days.
The gelatin then contained 7.5 per cent water and 1.5 per cent
ash. From this product two solutions were made; a 10 per cent
and a 20 per cent (10 grams of gelatin to 90 grams of water
and 20 grams of gelatin to 80 grams of water, respectively).
Not all of the water was added as such, since the solutions were
acid and it was necessary to add several cubic centimeters of
N/10 alkali to make the solutions neutral to phenolphthalein.
After neutralizing and diluting to weight the solutions were
sterilized in an Arnold steam sterilizer on three successive days
to insure killing any spore-forming organisms. For injection the
solutions were warmed on a water bath to 37 to 40°C.
The accompany tables on pages 149 to 152, show that the serum
proteins of an individual rabbit vary from day to day through
quite a wide range and that this variability applies to a greater
extent between different rabbits. The variability of the serum
proteins has been observed by Schmidt (2) and others employ-
ing the same methods of experimentation. From the same
tables it is also evident that the injected gelatin showed no
marked effect in altering either the ratio or the total amount of
the serum albumins and globulins, nor the amount of non-
proteins.
Moll (8), in the course of investigations on the effects of im-
munization upon the globulin content of their blood sera, in-
jected gelatin into rabbits and obtained a very marked increase
of the globulins in every case, in one instance his figures indicate
an increase of 81 per cent. His method (4) for estimating the
globulins depended upon a separation by salting out, followed
EFFECTS OF INTRAVENOUS INJECTIONS OF A COLLOID 149
by successive washings with water, alcohol, ether, then drying
at 110°C. and weighing. This method could hardly be expected
to give quantitative results since the salt used to coagulate the
proteins would be carried down in considerable amounts with
TABLE 1
Ratio of globulins to total proteins
ANIMAL NO. FORE-PERIOD PERIOD AFTER-PERIOD
per cent per cent per cent
( 18.4 Ace 22.0
| 19.8 29.7 22.1
I J 18.1 14.6
BES es ye eres ots 143 4.1
18.2 22.2
21.4
IAVCLAR Ctra... «amos: 17.8 21.5 22.0 «
18.0 19.9 20.3
20.9 22.8
24.7 28.0
19.7
Lin S360 are 19.5
17.4
17.8
[ 19.3
PANVERAGCW.
ss
pg
7 a]
Caan
@
iad »
“<)
-
ion
a
F
7 7 "
:
.
oon 2
-, * Pacis
=a
t
Py
_
4
[bad
coe” ©
S -
a,
oe =
- if
_—
THE INFLUENCE OF ACTIVE NORMAL SERUM
(COMPLEMENT) UPON MENINGOCOCCE
Il. THE BACTERICIDAL AND PROTECTIVE VALUE OF FRESH
NORMAL SERUM ALONE AND IN COMBINATION WITH
ANTIMENINGITIS SERUM FOR MENINGOCOCCI
TOITSU MATSUNAMI anno JOHN A. KOLMER
From the McManes Laboratory of Experimental Pathology of the Unwersity of
Pennsylvania
Received for publication March 29, 1918
In addition to specific opsonins and toxin neutralizing anti-
bodies, potent antimeningococcus sera are generally regarded as
possessing a distinct meningococcidal activity; the latter is
ascribed to the presence of specific bacteriolysin requiring the
presence of complement for lytic activity although this does not
explain the total bactericidal activity inasmuch as the early ex-
periments of Flexner (1) demonstrated that heated serum pos-
sesses bactericidal activity and 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 ren-
dered active and destructive for the microérganisms.
The defibrinated bloods and sera of normal persons have been
found by Davis (2), Flexner, (1), McKenzie and Martin (3) and
others to possess meningococcidal properties; this activity was
found most marked with fresh active sera and diminished but
not totally removed by heating at 60°C. for thirty minutes. In-
asmuch as the antimeningococcus sera ordinarily administered
are free of complement and the cerebrospinal fluid in meningococ-
cus meningitis apt to be free or contain but traces of this con-
stituent of serum, we have sought to determine the bactericidal
activity of the fresh serum of persons and guinea-pigs alone and
in conjunction with anti-meningococcus serum for virulent men-
1 Aided by a grant from the Pediatric Society of Philadelphia.
177
178 TOITSU MATSUNAMI AND JOHN A. KOLMER
ingococci, in an experimental study of the probable value of
complemental bacteriolysis in resistance to meningococci.
In our first communication concerning the influence of active
normal serum (complement) upon meningococci (4) we have
shown that the opsonic activity of fresh active antimeningitis
serum is greater than the opsonic activity of the same serum
after heating or the addition of tricresol, followed by exposure to
room temperature for several days; also that the addition of
active normal serum to commercial antimeningitis serum in-
creases opsonic activity. It would appear, therefore, that the
opsonic activity of antimeningitis serum is maximal when the
serum is fresh and active, that is, contains the labile opsonin or
complement, and that this constituent may be restored and the
commercial immune serum reactivated to a certain degree, by
the addition of fresh human or guinea-pig serum.
In this investigation we have followed three main lines of study:
1. To determine the bactericidal activity of normal human
and guinea-pig sera and various anti-meningococcus sera alone,
and in combination, for virulent meningococci, by in vitro,
plating and pipet methods.
2. A study of the agglutination of meningococci by normal
and immune sera in relation to the bactericidal activity of these
sera.
3. A study of the mouse test of Hitchens and Robinson (5)
for determining the protective value of antimeningitis serum and
the influence of normal sera alone and in conjunction with anti-
meningitis serum upon virulent meningococci, as determined
with this technic.
The results of a large number of experiments are briefly sum-
marized in this communication; with plating methods the bac-
tericidal activity of various antimeningococcus sera alone as well
as the sera of normal persons and guinea-pigs alone and in com-
bination with antisera, has been found surprisingly low although
these sera and particularly the whole blood of certain persons and
guinea-pigs, have been found definitely bactericidal for menin-
gococci with a delicate technic. The protection test of Hitchens
and Robinson has, we believe, a definite and practical value al-
INFLUENCE OF SERUM UPON MENINGOCOCCI 179
though not sufficiently delicate for the purposes of our experi-
ments, namely, to determine the influence of complemental
bacteriolysis alone as a factor in protection against virulent
meningococci. Inasmuch as any animal protective test calls
into consideration the influence of immunity principles in the
body fluids of the experimental animal and thereby masking the
possible influence of the addition of fresh normal serum (comple-
ment) upon the bactericidal activity of the immune serum, we
have given most attention to test tube experiments.
THE BACTERICIDAL ACTIVITY OF FRESH NORMAL SERUM ALONE
AND IN COMBINATION WITH ANTIMENINGITIS SERUM
a. The bactericidal aciivity of anti-meningococcus and normal
sera alone and in combination, as determined by plating methods.
All experiments were conducted with strain 124 of normal men-
ingococei? kindly furnished by Dr. George Robinson. Anti-
meningococcus sera were obtained from various laboratories;
normal human sera were obtained from ourselves and others of
the laboratory staff by vein puncture, and from normal adult
guinea-pigs by heart puncture.
Bactericidal tests in vitro, and particularly with plating meth-
ods, are generally unsatisfactory; strict attention to the reaction
of the culture medium and numerous preliminary tests to deter-
mine the proper dose of culture to employ in order to avoid too
few or too many colonies in a plate with frequent repetitions of
the main experiments, were necessary in this study. The pro-
tocols of several experiments are given in tables 1, 2, 3 and 4
and the results may be summarized as follows:
1. As shown in table 1, active or fresh antimeningitis sera are
more bactericidal than the same sera after inactivation by heat-
ing at 60°C. for thirty minutes. The addition of 0.3 per cent
tricresol to the sera increased bactericidal activity in dilutions up
to about 1:20, but this preservative did not appear to appreci-
ably injure or reduce the bactericidal activity of the serum itself
as tested within one week after the addition of tricresol, although
it reduces to some degree the opsonic activity of the serum.
* This strain was isolated on June 2, 1917, on the hospital ship Solace.
180 TOITSU MATSUNAMI AND JOHN A. KOLMER
2. Complement free antimeningitis sera possess bactericidal
activity as previously shown by Flexner and it is highly probable
that complemental bacteriolysis exerts but a minor réle in the
total bactericidal activity of immune serum.
TABLE 1
The influence of heat and tricresol upon the bactericidal activity of antimeningococcus
serum
FINAL DILUTIONS OF ANTISERUM
ANTIMENINGITIS SERUM
(No. 2279) | |
1:80 1:160 1:320 | 1:640 | 1:1280 | 1:2560
180, 30) Ster-! Ster-| Ster-| Few
ile ile ile
1:2 1:10 1:20 | 1:40
Fresh active....... 1500*; 900 |} 400 | 200
Five days after
addition of 0.25
per cent tricresol.| Ster-| Ster-| Ster-| 20} 40) 240) 20] Ster-| Ster-| Few
ile ile ile ile ile
Fresh serum after
heating at 60°C.
for one-half hour eth 1600 | 1800 | 300) 180; 20) 30 10 | Ster-| Few
ile
Heated serum +
active human se-
EL 4 meee ee 1800 | 1600 | 1200 LUT LY 1800 | 1400 | 1200 | 1800
* Colonies per plate.
In these experiments the culture (no. 124) was used in amounts of 0.5 cc. of a
1:500,000 emulsion of a twenty-four hour serum agar culture; 0.5 cc. of this emul-
sion mixed wih 0.5 ce. sheep serum broth and plated in amounts of 0.5 ce.
after two hours incubation, showed 600 colonies per plate.
Dilutions of serum in amounts of 0.5 ec. of culture were incubated at 37°C.
(water bath) for two hours and plates prepared with 0.5 ec. from each tube in
10 cc. of sheep serum dextrose agar.
0.5 cc. of human serum (1:5) plus 0.5 ec. emulsion of cocci and plated in amounts
of 0.5 cc. after two hours incubation, showed 900 colonies per plate.
3. The bactericidal activity of practically all the antisera
tested was low according to the results observed with the tech-
nic employed although some bactericidal activity was generally
apparent as compared with the controls; not infrequently the
higher dilutions of serum were more bactericidal than the lower,
these results being similar to those published by Jochmann (6).
4, As shown in table 4 active human and guinea-pig sera gen-
INFLUENCE OF SERUM UPON MENINGOCOCCI 181
erally possess some bactericidal activity for meningococci; the
bactericidal activity of these sera vary in different persons and
animals and, as a general rule, human sera are more bactericidal
than pig sera.
5. In practically all experiments however, the addition of equal
parts of 1:5 or 1: 10 dilutions of fresh active human or guinea-
pig sera to varying dilutions of different anti-meningococcus sera,
was without appreciable increase of bactericidal activity; indeed
TABLE 2
Bactericidal action of anti-meningococcus serum alone and in combination with
equal parts of a 1:5 dilution of active guinea-pig serum
PIG
FINAL DILUTIONS OF ANTISERUM SERUM] CUL-
SUBSTANCES AND | TURE
CUL- |ALONE
BP) 1;4 1:8 1:12) 1:16 | 1:24] 1:32 |] 1:48] 1:64 | 1:96 | ruRE
Antiserum
(com-
mercial) A
alone....| Ster-| Ster-| 540*/2160) Unc.| Unc.| Une.| Une.|Une. |Une. |Une. | Une.
ile ile
Antiserum
Sr yeahs é
serum....| Unc. | Unc. |6480 |9720} 4800} 8000) Unc.) Unc.| Une. |Unc.
* Colonies per plate; Une. = too many colonies for counting.
In this experiment a series of dilutions of the immune serum was prepared in
sterile test tubes with sheep serum broth, in amounts of 1 ecc.; each tube was
seeded with 0.1 cc. of a twenty-four hour serum dextrose broth culture of menin-
gococci and incubated in a water bath at 37°C. for two hours, shaken and 0.5
ec. from each tube plated in 10 ce. of sheep serum dextrose agar. In the second
series a 1:5 dilution of fresh sterile pig serum was used instead of sheep serum
broth in making the dilutions, thereby rendering the final dilutions of immune
serum the same in bothseries. Controls on the sterility of the immune and nor-
mal sera, culture, etce., were included. The plates were counted after forty-eight
hours incubation. ;
the colonies of meningococci were generally more numerous than
with the antiserum alone, suggesting that the addition of normal
serum enriched the culture medium and thereby favored the
more rapid multiplication of the surviving cocci.
TOITSU MATSUNAMI AND JOHN A. KOLMER
182
OO00L | OOSF
0O00L | 00cE
O0F9 | O0OL
OO8O0T} 0009
0006 | 09&
ANOTV | ANOTV
log: any)
-100
WOULS
TVNYHON
0022
0061
0002
00GT
0008
OFS
‘oul)
ou)
OVE
“UURY UD)
Ser :T
49:1
GET
OOTZ
0001
OOF
OOOT
OcOL
00€
‘oun
OO8OT
0082
o]L104g
i)
O0SG
006
006
006
OGL6
00€
‘oun
OFS89
O9GG
O[L10}g
ou)
‘ouy)
8:1
WOUASILINVY JO SNOILATIAG TVNIA
91949
By ieais
tandos UBUINTY +
Spm etn a: oUuoTe ([BIotoWUIOD)
umes uetmny +
SAONVISHOS
wuInAOsTyUV
WN.LOSsT}UW
wn.LOsTyUy
wnJostyuy
wndostyUy
umMJestyUy
wINn.LOsT}UV
wndestyuy
UINnAosIyUy
wndosTyUy
WIN.IOST}UV
S229 aS SSS
INDW
“TNH d Xa
€ QATaVL
unas Bid-paumb pun UDwUny ]DUWLOU YIN UOUDULQUOI UL PUD BUOID DLASYUD YIN 8989} JOPIDLLa}IDG fo fiunwmwngs
af
kK
183
INFLUENCE OF SERUM UPON MENINGOCOCCI
° ‘s}WoUUTIOdxe [][v UT popnfOUT o10M BIOS PUB 9IN4[Nd UT s[oayUOD sSNoeuNy
“UOT}EQNOUT SINOY FYITO-A4IOJ 10} JB OpVUI 919M S}UNOD {.1v3B og0I}xep
wnsos dooys Jo 90 OT YIM poye[d yowo wody 090 GQ puB UdyeYS ‘sinoy 9014} OF “H2E 9B poyeqnour osoM soqny [TV
TWUNTOS UBUINY [EUIOU OY} YIM UOTPVUIGUIOD UL IO OUOTE WUNIOS OUNWUTT YIM Solios You UT OUIeS OY} SABATY OOM UNIES
OUNTMUUT JO SUOT}NIIp [BUY oY} ‘TooooOSUrTUDUE Jo SoINy[Nd YYoIG oSOIXOp WMMAVS MOY aNoj-£4U0My JO SuOTyn]Ip SutArea jo
09 G°0 YFTA popoes O19M Sogn} oy, “UNIS UBUINY 911048 YSoIJ JO SUOT}N]IP OL: T 10 ¢:] YITA JO YZoIq UMAds doays jo sjavd yenbe
YIM popniip puv soynurur AyA1Y} OF “Hg 4B YWVq 19yVA B UT pozvoYy ATSnotAord ora BIOSTJUB IY} S}UIUTTIOdX9 Vseyy UT
‘GZ WNJOS SI}ISUTUOUL
“que f¢:] jo°oo GQ umes Std fyyMoIS IvBe umaoas jo uorsuedsns (00‘00¢:1T JO 99 GQ UT pesn amjy[ng *g yueumtsedxyT
‘% WINANS STPISUTUOUI
“que $G¢:] Jo 99 GQ) WNIOs UBUINY Y}MOIZ IvBe UNMIS Jo UOTSUdsNsS (Y0N‘00E: I jO *09 G'°Q UL posn oiNyND *¢ yueuTTIedx|
"p uInJes
SIPIBUIUOUNTZUB !G:] Jo 00 GQ Whies URUMY ‘oIny]ND YAorq UMaaS OOST:T JO “99 GQ UL posh oiny[Ny ‘fF queuTedxA
5 "e UNIS
SyIsulueWMTUe fe: Jo ‘00 GQ Wnses UvUMY forng]ND YOIq WNIOS QOOT:T jo ‘00 GQ UI posn oiny[nyg *¢ yuoUTTIEdxA
*Z wnies
SIPISUIUOUTTYUB fG:] Jo ‘09 GQ UMIJOS UBUMY foINy[NO YoIq UMS CNET] Jo *09 GQ UI pasn onyynD °*z yuoUTTIedx]
*[ uinies
SIFIZUIMOUIT} Ue SOT: [ Jo ‘09 gg UNIeS UBUNY {ony[No YZorq UNAS OOOT:T JO “99 GQ UL pasn eIngND *T yuauTIEdxny
184 TOITSU MATSUNAMI AND JOHN A. KOLMER
b. The bactericidal activity of anti-meningococcus and normal sera
alone and in combination, as determined with a pipet method. In
this technic which was modified after that of Wright (7) by Dr.
George D. Heist and Prof. Lacey, the number of microérganisms
exposed to the action of the serum for many hours was reduced
to a minimum thereby rendering the test for bactericidal activity
quite delicate.
From five to eight dilutions of a culture of normal meningo-
cocci (strain 124) prepared with sheep serum broth in a series of
tubules, were employed with a many stemmed pipet; the various
dilutions of culture were taken up in the respective sterile pipets
and then expelled, leaving meningococci clinging to the interior
TABLE 4
Bactericidal activity of active human and guinea-pig sera for meningococci*
FINAL DILUTIONS OF SERA
SERA
13:2 1:4 1:8 1:16 1:32 1: 64
Dy Mee aces eee |e 200) 1200 1400 1600 1800 3000
1D ee eee ene | 2700 900 1000 3700 3300 5000
1 Baa end A 2700 7500 4500 3700 2700 5500
PAG TD! a eee eee 7500 9000 9000 8000 5000 8000
* In these experiments 1 ec. of each dilution of serum was mixed with 1 ce. of
culture and incubated for two hours on a water bath when plates were prepared.
The culture alone showed 1200 colonies per plate before incubation for two
hours in the water bath and 7000 per plate after this period of incubation.
of the pipets.. Each pipet was now filled with serum, sealed and
incubated for twenty-four hours when smears were prepared,
stained and examined for meningococci. Numerous controls of
each dilution of culture were included in which sheep serum water
broth was substituted for serum. In these pipets in which all
meningococci were killed, microérganisms were not found in the
smears, whereas in those in which killing was incomplete or en-
tirely absent, few or many cocci were to be found in the stained
smears. After practice in the manipulation of the pipets, the
results become fairly consistent and the method is well adapted as
a ready and simple test for measuring the bactericidal activity of
whole blood inasmuch as but a drop or two is required and this
amount is easily obtained from a finger.
INFLUENCE OF SERUM UPON MENINGOCOCCI 185
With this technic fresh normal human and guinea-pig blood
were found distinctly meningococcidal, as shown in the protocols
given in table 5; the blood of different persons and guinea-pigs
were found to vary in bactericidal activity. As a rule human
« TABLE 5
The bactericidal activity of normal human and guinea-pig blood for meningococct
measured with a pipet method
DILUTIONS OF CULTURE*
FRESH WHOLE BLOOD ————
Undiluted 13/5 1:25 12/125. 1: 625
LIER DTN Snr aii a = = =z
Tekno ae ee OSE ee ar ar ar ae ae
1d hWieNPiy ise CHR SLA RE eee — — — = =
Elm any Assersye carlo e ee -- — _ = =
le tremavnn Gio 25 eae ee = = = = =
Errante Nel ay, che ces ove glen ciese a + + = =
EUTTAD ATID He vey desde ed a eats _ — = = =
nimea-pigide ye. eek ee as ais a ar ai
SONS Nh 2 sr + = 7 =
ORDINTROLET Bis noe eS I re ee + oF + ar ar
* A twenty-four hour serum broth culture of normal meningococci (strain
124).
+ + = growth; — = no growth (sterile).
TABLE 6
Whole blood more bactericidal for meningococci than serum
i}
RESULTS WITH BLOOD AND DILUTIONS RESULTS WITH SERUM AND DILUTIONS
OF CULTURE* OF CULTURE
SOURCE
Undi- ' i 5 Undi- , , c F
fated 1:5 1:25 1: 125 1625 ineeal 1:5 1:25 1:125 | 1:625
Homan.1.....) -+-T _ — _ = aL 4 = 2: ae
imam 3.6225 \ — — _ — — ate ae sd = a
Haman 6:....| - + = - _ a8 chs ae ails us
* A twenty-four hour serum broth culture of normal meningococci (strain 124).
++ = growth; — = no growth (sterile).
blood was found to possess a similar or slightly higher bacteri-
cidal activity than guinea-pig blood.
Whole blood of persons was found somewhat more bactericidal
than the corresponding fresh active sera, as shown in the proto-
cols given in table 6.
186 TOITSU MATSUNAMI AND JOHN A. KOLMER
Fresh active human and guinea-pig sera were more bacteri-
cidal than after heating at 56° C. for thirty to sixty minutes,
although the latter are bactericidal to a degree, and showign
thereby that the total bactericidal activity of normal serum is
not dependent entirely upon complemental bacteriolysis.
TABLE 7
The bactericidal activity of antimengitis sera alone andin combination with normal
serum
DILUTIONS OF CULTURE*
COMPLEMENT FREE ANTIMENINGITIS SERA 5
Undi | 1:5 | 1:25 | 1:125 | 1:250 | 1:500
Antimeningococcus serum (A) undiluted| + — _ — = _
Antimeningococcus serum (A) 1:10......) + a = = _ --
Antimeningococcus serum (A) 1:100....) + _ ~ _ _
Antimeningococcus serum (A) 1:100 +
human serum complementf........... — — = — ~ _
Antimeningitis serum (B) undiluted.....) + = = - _ _
Antimeningitis serum (B) 1:10.......... + - _ _ = =
Antimeningitis serum (B) 1:100.........) + — - _ —
Antimeningitis serum (B) 1: 100 + guin-
ea=pic-complement: 2.20: 2M .eeee ee + - + + + os
@ontrols idence eee ee + 7 + + — +
Plater os ete ee ae ee one Une. | 2700 | 1620 |} 960} 900} 120
* Culture prepared by washing off twenty-four hour serum dextrose agar
slant growths with 2 cc. of sheep serum broth, emulsifying and diluting with
sheep serum water.
+ Plates were prepared by drawing culture into the capillary tubes employed,
expelling the culture, and then washing the cocci adhering to the inner wall into
petri dishes with twelve changes of sheep serum broth; the number of 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 various sera.
t Four parts of antiserum 1:100 plus one part of undiluted normal serum.
With this technic the bactericidal activity of antimeningitis
sera Was more pronounced; in order to measure this activity, it
was found necessary to use a larger number of meningococci than
could be obtained in a twenty-four hour serum dextrose broth
culture. The results of several experiments with antisera are
shown in table 7; also an idea of the number of cocci exposed to
the bactericidal substances of the sera.
As shown in table 7 dilutions of antimeningitis sera up to 1: 100
INFLUENCE OF SERUM UPON MENINGOCOCCI 187
were found bactericidal and this activity was sometimes, but
not always, increased by the addition of fresh normal serum and
particularly human serum.
AGGLUTINATION OF MENINGOCOCCI IN RELATION TO THE BACTERI-
CIDAL ACTIVITY OF NORMAL AND IMMUNE SERA
Inasmuch as agglutinins are regarded by many as aiding bac-
teriolysis and phagocytosis, we have studied the agglutinin con-
tent of horse antimeningitis and normal human and guinea-pig
sera, for normal meningococci; furthermore laboratories engaged
TABLE 8
Agglutination of meningococci* by normal human and guinea pig sera
&
ie 4 FINAL DILUTIONS
ap ee
BAB) 1:2 | 1:3 | 1:4 1:5 | 1:6 18 | 1:10
deg. C
uM am SERUM: 4.5.5: sss + 55 + Se = = — ~ =
Guinea-pig serum......... 55 a =F | Se | = = = =
* Normal meningococci strain 1.
} For sixteen to eighteen hours.
in the production of antimeningitis serum commonly employ
the agglutination test as a measure of antibody response and we
have sought to determine as a by study the relation between the
agglutinin, opsonin and bactericidan content of the immune and
normal sera.
The agglutinin content of all the various immune and many
normal human and guinea-pig sera employed in this study, was
determined with a macroscopic technic ;? likewise after the addi-
tion of fresh normal human and guinea-pig sera to the immune
sera to determine if these increased the agglutinating activity of
the latter.
Normal human and guinea-pig sera were found to contain
traces of agglutinin for normal meningococci as shown in table 8
with one serum of each.
3 We are indebted to Dr. Shigeki Sekiguchi for assistance in conducting the
agglutination tests.
188 TOITSU MATSUNAMI AND JOHN A. KOLMER
The addition of fresh normal human serum to antimeningitis
serum in amounts of 0.5 cc. of 1:5 dilution appeared, but not
uniformly, slightly to influence agglutination as shown in table 9.
The addition of tricresol to fresh antimeningitis sera to the
point of 0.25 per cent, followed by exposure of the sera to room
temperature for a week, was usually followed by slight decrease
in agglutinating activity, which may be ascribed to the loss of
labile substances or to slight decrease in agglutinin or both;
heating antimeningitis sera at 60°C. for thirty minutes was
TABLE 9
The influence of normal serum upon the agglutination of meningococci by antimen-
ingococcus serum*
ac FINAL DILUTIONS 1:
SERA ‘gp
als 60 | 80 | 100 | 120 | 160 | 200 | 240 | 320 | 400 | 480 | 640
deg. C Ree
Antiserum alone): 22... +457: 37 | +) +) +) +) el tl -| -| -| -| -
DO.) ae) ep ay aa] ae AP S| Se
Antiserum+ human serum 0.5 ce. 37) +) +) +) +) +) 4+) 4+) -1 -| -| -
(1: 5) See oS ier bere edparhariss| sr) =, =
Autiserum + pig serum 0.5 ce. {| 37 | +) +) +) +) +) +) +) -| -| -| -
(1: 5) a feel ea Val fe fen Mia eurf== [hh |
* Strain of normal meningococci 124 employed.
t Incubated for sixteen to eighteen hours.
found to markedly decrease the agglutinating activity and this
was not restored by the addition of normal serum. The results
observed with a few of the sera tested are shown in table 10.
Incubation of the mixtures of serum and culture in the macro-
scopic tests at a temperature of 55°C. for at least sixteen hours
was found to produce the most marked agglutination and usually
higher than observed when the mixtures were incubated at
37°C. for the same period of time (the latter tests being conducted
under sterile conditions to avoid the growth of contaminating
bacteria which mask the reactions. .
Despite a large amount of work with numerous repetitions of
experiments, it is a difficult matter to draw deductions upon the
i a
INFLUENCE OF SERUM UPON MENINGOCOCCI 189
TABLE 10
~The influence of tricresol and heat upon the agglutinating activity of antimeningocccus
sera*
FINAL DILUTIONS
SERUM
1:10| 1:20! 1:40! 1:80 {1: 100|1: 320)1: 640|1: 1280
Peootresh and: active... ..2..s.....--- Se SES Se Se tee | Eb _
Gao yarvenwricresOlane ssh) Pica Soe +yey4+i]y4+)4+;,4+)]—- =
Boo atten ME AbIN Ga c8 wi. so cole ole saa +/+)4]—-|]-|]-|]- =
eT EReSI ANG ACUIVES. .. i. 5. os ene te +y/+)/+)])+;+!4+)]/2 ea
PG merahuer GMeCLeSOli a... 2 <2 cs ccecsce wees + yey tye l tye] =
Muborranten NeauINe: yest Sle eek ences 3s Sa er le || | | =
Bios mesh sand aeulyers. 22 obra. Sone bias se - t+rt}y+)/+y+]}4+)- =
Pas ater NEabln esse... si. x qpenicic ies. bacpels +/+ )])-|- | -- | —|— =
* All tests were conducted with strain of normal meningococci 124 and at
55°C. for twenty hours.
+ Heating at 60°C. for thirty minutes.
TABLE 11
A comparison of the agglutinin, opsonin and bactericidan content of various anti-
meningitis and normal sera*
Sit} ‘ CULTURE
bse 5 : : EUSGOCTENC, RESULTS OF BACTERICIDAL TESTS CONTROLS
Bea eons | 1:4 | 1:10] 1:20) 1:40| 1:80 |1:160|1: 320, 7CHPAL
oe [SHHeS TESTS
(20:01 Died in 17 hours | 13 | 0.01 | Died in 17 hours
12 | 0.005 Died in 17 hours | 12 | 0.005 | Died in 18 hours
12 | 0.0025 | Diedin17hours | 11 | 0.0025| Died in 18 hours
10 | 0.00125 Survived 6 days 11 | 0.00125, Died in 4 days
10 | 0.0006 Survived 6 days 11 | 0.0006 Died in 5 days
uniform growths in fluid culture media have so far proven
unsuccessful.
5. Highly virulent strains of meningococci yielded more regu-
lar results than strains which had undergone some loss in viru-
lence by reason of prolonged cultivation without animal passage
(table 12); the protective tests were more regular in their results
when highly virulent strains were employed.
6. For preparing the suspensions and dilutions of cocci it was
not found necessary to use active guinea-pig serum; sheep serum
broth was found satisfactory for this purpose. Comparative
tests with heated and unheated (fresh) guinea-pig serum diluted
INFLUENCE OF SERUM UPON MENINGOCOCCI 193
1:4 showed occasionally different results, the minimal lethal
dose being less with the fresh active serum (table 13) and indi-
cating the probable bactericidal effect of the active serum.
7. Our experiments have not been sufficiently numerous to
warrant an opinion upon the relation between protection test in
mice and such reactions in vitro as agglutination, bactericidal and
opsonic tests for measuring the potency of antimeningitis serum,
but we have tested different antimeningitis sera with the general
result that those possessing the highest protective value also
contained most agglutinin and opsonin. As previously stated
the results of bactericidal tests in vitro were too irregular to
permit comparison.
8. Even with a highly virulent strain or strains of cocci the
results of protective tests on different days with the same anti-
meningitis serum varied to some extent. This was expected
because animal tests cannot yield strictly regular and mathe-
matically exact results, owing to variations in susceptibility in
addition to other factors. On the basis of our experience with the
protective test of Hitchens and Robinson we believe the follow-
ing conditions important:
a. A highly virulent strain or strains of normal or para meningo-
cocci or both should be employed—the virulence being maintained
by passage through mice.
b. A uniform suspension of cocci should be secured by wash ng
off eight hour growths from twelve or more slants of serum dex-
trose agar with sheep serum broth and mixing thoroughly. For
measuring the polyvalency of a serum twelve or more different
cultures of meningococci and parameningococci may be mixed
and employed.
c. Mice of approximately the same weight should be employed.
d. In conducting a test for the protective value of a serum, one
or two sets of mice should be injected with the suspension of
cocci alone, the doses being arranged according to the approxi-
mate virulence on the basis of former tests in order to determine
the minimal lethal dose of the particular suspension being used in
the protective tests. Owing to the likelihood of variation in re-
sults with different suspensions, it is not advisable to determine
194 TOITSU MATSUNAMI AND JOHN A. KOLMER
the minimal lethal dose with one suspension and conduct the
protective tests with asecond. In other words the minimal lethal
dose of each suspension used in the protective test should be de-
termined at the same time and the results expressed according to
the protection afforded by 1 cc. of serum over a period of 48
hours against multiples of the minimal lethal dose of culture.
TABLE 14
The protection activity of antimeningitis serum alone and in conjunction with
normal guinea-pig serum for virulent meningococct
GUINEA-PIG
WEIGHT CULTURE* ANTISERUM Saree RESULTS
grams ce. cc cc
25 0.5 0 0 Died in 36 hours
23 0.25 0 0 Died in 2 hours
2 0.125 0 0 Died in 13 hours
10 0.0625 0 0 Died in 20 hours
20 0.5 0.5 0 Died in 24 hours
10 0.25 0.5 0 Died in 36 hours
22 0.125 0.5 0 Survived over 5 days
2a 0.0625 0.5 0 Died in 15 hours
24 0.5 0 OS Died in 22 hours
23 0.25 0 0.5 Died in 20 hours
23 0.125 0 0.5 Survived over 5 days
22 0 0525 . 0 0.5 Died in 20 hours
25 0.5 0.5 0.5 Survived over 5 days
20 0.25 0.5 0.5 Died in 24 hours
16 0..125 0 0.5 Survived over 5 days
16 0.0625 0.5 0.5 Survived over 5 days
* Strain 124 normal meningococci suspended in heated guinea-pig serum diluted
U4:
9. While this protection test in its present state has not
yielded us results as close and delicate as may be obtained with a
satisfactory toxin in determining the antitoxic value of diphtheria
antitoxin, we believe the test to be one of value as a means for
determining and measuring the approximate protective and
curative value of antimeningitis serum.
INFLUENCE OF SERUM UPON MENINGOCOCCI 195
b. The protective value of normal human and guinea-pig sera
alone and in combination with antimeningitis serum. ‘The results
of numerous experiments to determine the protective value of nor-
mal active human and guinea-pig sera alone and in combination
TABLE 15
The protective activity of antimeningitis serum alone and in conjunction with normal
guinea-pig serum for virulent meningococct
NORMAL :
WEIGHT CULTURE* ANTISERUM GUINEA-PIG RESULTS
SERUM
grams cc. cc
15 0.01 0 0 Died in 14 hours
14 0.005 0 0 Died in 24 hours
12 0.0025 0 0 Survived over 4 days
12 0.00125 0 0 Died in 15 hours
16 0.0006 0 0 Survived over 4 days
16 0.01 0.5 0 Survived over 4 days
15 0.005 0.5 0 Survived over 4 days
13 0.0025 0.5 0 Survived over 4 days
12 0.00125 0.5 0 Survived over 4 days
12 0.0006 0.5 0 Survived over 4 days
16 - 0.01 0 0.5 Died in 15 hours
14 0.005 0 0.5 Died in 20 hours
14 0.0025 0 0.5 Survived over 4 days
138 0.00125 0 0.5 Died in 21 hours.
12 0.0006 0 0.5 Died in 19 hours r
18 0.01 0.5 0.5 Survived over 4 days
14 0.005 0.5 0.5 Survived over 4 days
14 0.0025 0.5 0.5 Survived over 4 days
13 0.00125 0.5 0.5 Survived over 4 days
13 0.006 0.5 0.5 Survived over 4 days
* Strain 124 normal meningococci suspended in heated guinea-pig serum diluted
1:4.
with different antimeningitis sera may be summarized as fol-
lows; the protocols of several experiments are shown in tables
14, 15, 16, 17 and 18. In an effort to render the protection
tests more delicate the technic of Hitchens and Robinson was
modified in some experiments by using smaller doses of immune
and normal serum.
196 TOITSU MATSUNAMI AND JOHN A. KOLMER
1. Normal active human and guinea-pig sera were practically
without demonstrable protective value in these tests employing
white mice.
TABLE 16
The protective activity of antimeningitis serum alone and in conjunction with normal
human serum for virulent meningococci
ANTIMENINGITIS
WEIGHT CULTURE’ SERUM HUMAN SERUM RESULTS
grams ce. ce. cc
16 0.1 0 0 Died in 13 hours
17 0.1 ) 0 Died in 18 hours
19 0.1 0 0 Died in 19 hours
14 0.1 0 0 Died in 48 hours
18 0.1 0 0 Died in 13 hours
18 0.1 0.2 0 Died in 30 hours
23 0.1 0.1 0 Died in 24 hours
20 0.1 0.05 0 Died in 13 hours
19 (Qual 0.025 0 Died in 24 hours
16 0.1 0.0125 0 Died in 3 days
16 0.1 0 0.2 Died in 13 hours
18 0.1 0 0.2 Died in 24 hours
20 0.1 0 0.2 Died in 24 hours
23 0.1 0 0.2 Died in 3 days
20 0.1 0 0.2 Died in 13 hours
18° 0.1 0.2 0.2 Died in 24 hours
16 Opt 0.1 0:2 Died in 24 hours
19 0.1 0.05 0.2 Died in 3 days
20 0.1 0.025 0.2 Died in 13 hours
20 0.1 0.0125 0.2 Died in 24 hours
* Strain 124 normal meningococci suspended in heated guinea-pig serum
diluted 1: 4.
2. Occasionally the addition of normal human and guinea-pig
serum to antimeningitis serum appeared to afford increased pro-
tection as shown in the protocol of one experiment with pig
serum in table 14, but a general review of all experiments showed
that no protective influence could be attributed to the addition
of the normal to immune sera.
INFLUENCE OF SERUM UPON MENINGOCOGCI 197
As previously stated it is scarcely to be expected that an ani-
mal test would bring out the increased protective value following
the complementing of an immune serum if such actually does
occur, because the body fluids of the test animal contains
complement.
TABLE 17
The protective value of active human serum (complement) alone and in combination
with antimeningitis serum
WEIGHT eth acacia Ft HUMAN SERUM CULTURE* RESULTS
grams cc. cc. cc.
21 0.1 0 0.01 Survived.
24 0.1 0) 0.05 Survived.
26 0.1 0 0.1 Survived.
16 0.1 0 OR2 Survived.
20 Onl 0 0.4 Died in 24 hours
24 0.1 0.1 0.01 Survived
24 (Oyal 0.1 0.05 Survived
24 0.1 0.1 0.1 Survived
19 0.1 0.1 OFZ Died in 12 hours
21 0.1 0.1 0.4 Died in 12 hours
19 0 0 0.01 Died in 12 hours
17 0 0 0.01 Died in 12 hours
18 0 0.1 0.01 Died in 12 hours
16 0 0.1 0.01 Died in 12 hours
* Strain 124 normal meningococci suspended in heated guinea-pig serum
diluted 1: 4.
We have also conducted combined in vitro-vivo tests by incu-
bating in a water bath for one hour varying amounts of a sus-
pension of virulent meningococci alone and in combination with
antimeningitis serum and with antimeningitis and normal serum,
followed by tests for virulence and bactericidal activity by in-
traperitoneal injection into white mice; the results of these ex-
periments were quite irregular, but the addition of normal serum
did not appear to influence the results.
198 TOITSU MATSUNAMI AND JOHN A. KOLMER
TABLE 18
The protective value of active guinea-pig serum (complement) alone and in combina-
tion with antimeningitis serum
WEIGHT at a freee CULTURE* RESULTS
grams cc. cc. cc.
15 0 0 hyp sOXOL Died* in 15 hours
14 0 0 0.005 Died in 15 hours
12 0 0 0.0025 Survived*
12 0 0 0.00125 Died in 15 hours
16 0 0 | 0.0006 Survived
15 0.1 0 0.01 Survived
25 0.1 0 0.005 Survived
De 0.1 0 0.0025 Survived
23 0.1 0 0.00125 Survived
19 0.1 0 0.0006 Survived
20 0.1 OL 0.01 Survived
23 0.1 0.1 0.005 Survived
26 0.1 0.1 0.0025 Survived
20 On 0.1 | 0.00125 | Survived
20 0.1 0.1 0.0006 Survived
19 0.1 0.1 0.01 Died in 12 hours
17 0.1 0.1 0.01 Died in 18 hours
* Strain of normal meningococci 124 in heated pig serum 1: 4.
SUMMARY
1. The bactericidal activity in vitro of different antimeningitis
sera was found to be quite low. Fresh or active antimeningitis
sera was somewhat more bactericidal than the same sera after
inactivation by heating at 60°C. for thirty minutes.
2. Active normal human and guinea-pig sera are generally
slightly bactericidal for meningococci.
3. The bactericidal activity of horse-antimeningitis and nor-
mal human and guinea-pig serum, is largely independent of
complemental bacteriolysis.
4. The bactericidal activity in vitro of normal and immune
sera was best shown by a pipet method employing small num-
bers of cocci and relatively large volumes of serum.
b
INFLUENCE OF SERUM UPON MENINGOCOCCI 199
5. The addition of active normal human and guinea-pig serum
to antimeningitis serum sometimes increased the bactericidal
activity of the latter.
_6. Whole human and guinea-pig blood was found slightly more
bactericidal than the sera alone.
7. Normal human and guinea-pig sera frequently agglutinates
meningococci in final dilutions up to 1:4, but not in higher
dilutions.
8. Antimeningitis sera containing the largest amounts of ag-
elutinin were found to possess most opsonin and apt to prove
most bactericidal in vitro.
9. The mouse test of Hitchens and Robinson was found under
certain technical conditions to be of value as a means for deter-
mining and measuring approximately the protective and cura-
tive value of antimeningitis serum.
10. Normal active human and guinea-pig sera were practically
without demonstrable protective value in mice infected with
virulent meningococci, although the addition of these normal
sera to antimeningitis serum appeared in some experiments to
slightly increase the protective power of the latter.
11. The addition of active normal human or guinea-pig serum
(complement) to antimeningitis serum cannot be expected on ihe
basis of our experiments, to greatly augment bactericidal activ-
ity because complemental bacteriolysis exerts but a minor réle in
the relatively feeble bactericidal activity of antimeningitis serum,
but the addition of normal sera definitely increases opsonic ac-
tivity (10) and since it would appear that a large part of the
curative properties of antimeningitis serum is to be ascribed to
the presence of opsonin, it is suggested as worthy of clinical trial
to complement the antimeningitis serum by the addition of active
human or guinea-pig serum prior to intraspinal injection and
particularly in the treatment of severe and serum-resistant
infections.
For this purpose human serum is superior to guinea-pig serum
and may be obtained from the patient or a volunteer with the
usual precautions for asepsis.
200 TOITSU MATSUNAMI AND JOHN A. KOLMER
REFERENCES
(1) Frexner, S.: Contributions to the biology of diplococcus intracellularis.
Jour. Exp. Med., 1907, 9, 105-141.
(2) Davis, D. J.: Studies in meningococcus infections. Jour. Infect. Dis.,
1905, 2, 602-619.
(3) Farruey, N. H., anp Stewart, C. A.: Cerebrospinal fever, 1916, Service
publication no. 9, Commonwealth of Australia, 161-162.
(4) Kormer, J. A., Toyama, I., anp Matsunamti, T.: The opsonic activity of
fresh normal serum alone and in combination with antimeningitis
serum for meningococci. Jour. of Immunology, 1918, 3, 157.
(5) Hitcuens, A. P., anp Roprnson, G. H.: Standardization of Antimeningitis
Serum. Jour. of Immunology, 1916, 2, 345-353.
(6) Jocumann, G.: Versuche zur Serodiagnostik und Serumtherapie der epi-
demischen Genickstasse. Deut. med. Wehnschr., 1906, 32, 788-793.
(7) Wricut, A. E.: Handbook of the technique of the test and capillary glass
tube. Constable and Company, London, 1912. 119.
(8) Amoss, H., anp Woxtstern, M.: A method for the rapid preparation of an-
timeningitis serum. Jour. Exp. Med., 1916, 23, 403.
(9) Exser, W. J., AND Huntoon, F. M.: Studies on meningitis. Jour. Med.
Research, 1909, 20, 372-541.
(10) von LinGELSHEIM AND Leucus: Tierversuche mit dem Diplococcus intra-
cellularis. Klin. Jahrbuch., 1906, 15, 489.
(11) Kormer, J. A., Toyama, I., anp Matsunamt, T.: The opsonic activity of
fresh normal serum alone and in combination with antimeningitis
serum for meningococci. Loc. cit.
THE RELATION OF THE MENINGOCOCCIDAL ACTIV-
ITY OF THE BLOOD TO RESISTANCE TO
VIRULENT MENINGOCOCCI
TOITSU MATSUNAMI anp JOHN A. KOLMER
From the McManes Laboratory of Experimental Pathology of the University of
Pennsylvania
Received for publication April 3, 1918
In a previous communication (1) we have shown the greater
bactericidal activity of whole blood for the meningococcus as
compared with serum and briefly described a simple and con-
venient method for measuring the bactericidal activity of whole
blood devised by Dr. Heist employing the many stemmed capil-
lary pipet of Wright. Further experiences with this technic
have demonstrated its simplicity; but a few drops of undefibri-
nated blood are required and the results are usually quite sharp
and convincing. As is true with all bactericidal tests, however,
the results observed with the blood of one person or lower animal
tested on different days not infrequently yields shghtly varying
results, owing not only to probable fluctuations in the bacteri-
cidan content of the blood but more particularly to fluctuations
in the numbers of viable meningococci in the culture employed;
with careful attention to technic however, the latter error can be
reduced to a minimum.
The object of the present investigation was to determine
whether or not a relation exists between the meningococcidal
activity of the blood and resistance to infection with virulent
meningococci; furthermore whether or not the high natural im-
munity or resistance of certain of the lower animals to the menin-
gococcus is to be ascribed in part to a higher meningococcidal
activity of their blood.
Young and old guinea-pigs, rabbits and white mice were used
in these experiments. Flexner (2) has found that guinea-pigs
201
1 Aided by a grant from the Pediatric Society.
202 TOITSU MATSUNAMI AND JOHN A. KOLMER
weighing from 175 to 200 grams are highly susceptible to virulent
meningococci as compared with pigs weighing 350 to 400 grams
or more and that young pigs are more susceptible than mice;
Von Lingelsheim and Leuchs (3) and Kolle and Wassermann (4)
have also found young guinea-pigs highly susceptible. Betten-
court and Franca (5) Elser and Huntoon (6) and Hitchens and
Robinson (7) claim, however, that white mice are more highly sus-
ceptible and the latter employ these animals in a protection test
with antimeningitis serum. All investigators who have worked
with rabbits found them highly resistant; likewise all appear to
agree that young guinea-pigs are more susceptible than older and
heavier animals.
EXPERIMENTAL
Our experiments were conducted with a single strain of men-
ingococcus, the virulence and cultural characteristics of which
was quite familiar to us; bactericidal tests were made with the
whole blood of different series of young and old pigs, mice and
rabbits followed by the intraperitoneal injection of the same
culture in graded doses and according to the body weight of cach
animal into each series of animals; all animals were kept under
observation for two to four days and after death the blood of the
heart examined for meningococci. The strain of meningococcus
employed was highly virulent and all animals succumbing within
four days invariably showed the presence of meningococci in the
blood of the heart. In order to render the virulence tests with
different animals strictly comparable, all were injected on the
same day with the same emulsion of meningococci and, as stated
above, according to body weight. By experiments of this nature
we have sought to determine whether the resistance or non-
resistance of a certain species of animal bore a relation to the
meningococcidal activity of the whole blood of this species and
whether variation in resistance of different species of animals
bore a direct relation to variation in bactericidal activity of the
whole blood.
cali
MENINGOCOCCIDAL ACTIVITY OF THE ELOOD 203
Bactéricidal tests
In the bactericidal blood tests the many-stemmed pipet devised
by Wright for measuring the coagulation-time and anti-tryptic
power of the blood (8) was employed after the method devised
by Dr. Heist. Each pipet measures about 9 em. in length and
about 1 mm. or somewhat less in thickness and six are used at
one time. In conducting the tests we have employed twenty-
four hour serum broth cultures of a strain of meningococcus un-
diluted and in five dilutions prepared with sheep serum broth,
namely, 1:5; 1:25; 1:250; 1:500 and 1:1000. These cultures are
arranged in sterile tubules and allowed to run by eapillary attrac-
tion into the six sterile and numbered pipets respectively; the
pipets are now blown out and each loaded to the saine level
with blood secured by pricking the skin after cleansing with
alcohol.. The pipets are now sealed by dipping the top in paraf-
fin and incubated for twenty-four hours when a smear is made
of each and stained for meningococci. With care in the technic
contaminations are rare and the results quite sharp and regular.
The numbers of microdrganisms exposed to the germicidal action
of the whole blood are quite small being those which have ad-
hered to the inner wall of each capillary tube. Controls were
always included with each dilution of culture in which sheep
serum broth was substituted for blood.
Not infrequently the blood of persons and lower animals was
able to destroy all cocci in the undiluted culture and in order to
measure the bactericidal activity, denser cultures were prepared
by washing off twenty-four hour serum dextrose agar cultures
with 2 cc. sheep serum broth and preparing further dilutions of
this emulsion.
In order to estimate the number of viable meningococci in
each dilution of culture adhering to the walls of the capillary
tubes, plates were prepared by washing out the cocci with several
changes of serum dextrose broth into sterile petri dishes followed
by the addition of serum dextrose agar and counts at the end of
twenty-four hours incubation. & | Number| onrqrvan
Z
E
© | sup-cuurunzs
a GROWTH
:
oO
+|+ |+ |+ |+|-
+) +) +) +]4-
+| +} +) +|-[-
+} +/+ |+ |+|-
+|+ |-*
+|+ [+ |+ |-|-
+)|—*
+|+ |+ |+ |-|-
+|+ |+ |+ |+/-
+|+ |-*
+/+ |+ |+ |4]-
+/+ |+ |+ |4|-
+\+ |+ |+ |4/+
+|+ |+ |-
+/+ |+ |+ 4+
+\+ |+ |+ |4]+
+/+ [4 |+ |4]+
+|+ |+ |-
+/+ [4 |+ |4)+
+)/-— |- |-
+)/— |— j—
+/— |— |-
+)/— |— |—
+/— |-— |-—
In the original cultures of control B cultures 20 to 24 a medium
of normal plasma 1 part + 0.9 per cent NaCl solution was used.
The fragments from these cultures were transplanted in the same
medium used in the subcultures of the experiment, cultures 1 to
12.
Meo he
je ae
TISSUE CULTURE AS A MEANS FOR ESTIMATING TOXIN 227
No growth was seen in any of the subcultures of control B.
No growth was seen in any of the fifth series of subcultures of
the experiment. The cells on the other hand were growing ac-
tively in the fifth subculture of control A, cultures 13 to 19.
Burrows and Neymann (2) find that cells grow actively from
fragments of heart muscle of twelve-day chick embryos through
from 6 to 8 transplants in a medium of normal plasma 1 part +
0.9 per cent NaCl solution when the transplants are made seventy-
two hours apart. Growth ceases after this time.
Table 8 gives the results of a series of experiments very similar
to those given in table 7. The same normal and immune plas-
mata were used. The experiment differs in that different quan-
tities of diphtheria toxin were added to the different cultures. The
growth of the subcultures is alone indicated. Controls A and B
were prepared in exactly the same manner as those in experi-
ment 7, table 7.
By these experiments it is possible to show that the tissue cells
of chick-embryos are able to resist otherwise lethal doses of diph-
theria toxin after they have remained a short time in the plasma
of a passively immunized chicken. It was interesting that the
cells ceased to grow in the cultures as quickly when small amounts
of toxin are added to the medium as when larger amounts are
added.
In the next series of experiments it became of interest to ascer-
tain whether it is possible to study by this method the distribu-
tion of the antitoxin substances in the animal.
Two young chickens five weeks old were selected. They were
pullets of one hatching. They were selected from twenty of the
hatching and were as nearly alike as one could obtain chickens.
Into the wing vein of one of them 2 ce. of diphtheria antitoxin
was injected. The other was used as control. The injected
chicken is designated A. The control B. After forty-eight
hours blood was taken from each and plasmata A and B re-
spectively were prepared. A piece of artery was also taken
from each of the chickens, tissue A and B respectively.
About fragments of the arteries of this age embryo the cells
commence to grow only after a considerable latent period. In
228 MONTROSE T. BURROWS AND YOSHIO SUZUKI
TABLE 8
Tissue: Fragments of the heart muscle of an eleven day chick-embryo
Medium: Original culture not indicated in the table. Plasma obtained from an
adult chicken which has received forty-eight hours previously 2 cc. of diphtheria
antitoxin into its wing vein. Plasma, 1 part; 0.9 per cent NaCl solution, 1 part.
In subcultures indicated in the table: normal chicken plasma, 1 part; various di-
lutions of diphtheria toxin in 0.9 per cent; NaCl solution, 1 part
Medium: Control A. Original culture not indicated in the table. Plasma pre-
pared from the blood of the passively immunized chicken, 1 part; 0.9 NaCl solution,
1 part; subcultures: Normal chicken plasma, 1 part; 0.9 per cent NaCl solution, 1
part
Medium: Control B. Original culture not indicated in the table, normal plasma, 1
part; 0.9 per cent NaCl solution, 1 part. Subcultures: Normal plasma, 1 part;
1 per cent diphtheria toxin in 0.9 per cent NaCl solution, 1 part
GROWTH OF SUB-CULTURES
EXPERIMENT Number of times toxin is diluted
2 i Re
100. so 1 APstot aliecdag cl ited ale
aan ss dle Phe at elt ar lie
400 { i i x ri E:
600 { ee i i A =
1000 { u cl iz i a
Control A. No toxin { 25 7 i i Fi
Control B. 100 { ‘i = a
the table (table 9) the growth was recorded each day for a
period of three days.
In the first series of cultures, no. 1, table 9, fragments of tissue
A were planted in a medium consisting of 1 part of plasma A +
a>
TISSUE CULTURE AS A MEANS FOR ESTIMATING TOXIN 229
1 part of 0.9 per cent NaCl solution control, or 1 part of 0.9
per cent NaCl solution containing varying amounts of diph-
theria toxin. On the third day the cells were growing actively
in all these cultures; 2 cultures were made in each of the media
tested.
In the second series no. 2 fragments of tissue A were tested in
a medium containing 1 part of plasma B + 1 part of 0.9 per cent
NaCl solution or 1 part of a 0.9 per cent NaCl solution contain-
TABLE 9
as NUMBER OF TIMES DILUTED
MEDIUM z 5 CONTROL
& 100 500 1000 3000 | 5000
days
(1) Toxin is added to il ee ome nes Rat ot il
ee A and ae in es ey fe Spe ra bia hight A
culture conta n
eB) og Jt ttt tlt44 4 [4 sie ale4 44
tissue A
(2) Toxin is added to
plasma B and used in
whe: 2 |- — ao FE Fa 4] OE
culture containing Ce: a! Kia
tissue A : a a oa aie jepive ween
(3) Toxin is added to { an up a Sodas ui) 28s De
plasma B and used in be. Thee’, Sy |
culture containing pila «= 71goes a (ees Pe ese amy en 7
tissue B |
* The cells from the fragments of the arteries of this aged chicken grow slowly.
ing varying amounts of diphtheria toxin. The growth of the
cells of tissue A was not active in the presence of plasma B and
toxin. No growth is seen where as much as 1 per cent of toxin
is contained in the medium.
In the third series of experiments no. 3 tissue B was tested in
the same manner in plasma B. No growth was noted except in
the control cultures.
A second series of experiments of this kind was also performed,
table 10. In this series two young chickens similar to those used
in the other were obtained. Into the wing vein of one of them
230 MONTROSE T. BURROWS AND YOSHIO SUZUKI
0.5 ec. of diphtheria antitoxin was injected. This chicken was
designated as C. The other chicken D was kept for control.
_Blood together with a piece of artery was taken from each after
forty-eight hours and 4 series of cultures were prepared. In the
first series no. 1 the tissues C were tested in plasma C, 1 part
+ 0.9 per cent NaCl solution 1 part or 0.9 per cent NaCl solu-
tion which contain varying amounts of diphtheria toxin as indi-
cated in the table. In the second series no. 2 fragments of tissue
Dweretested in a medium consisting of 1 part of plasmaC 1 part of
TABLE 10
NUMBER OF TIMES DILUTED | 5
GROWTH a
AFTER SS Le E
100 500 | 1000 | 3000 | 5000 S
days |
(1) Toxin is added to plasma ‘‘C”’ 1 =) ee eee
and used in cultures containing 2 1+ jt —|4+ +)4+ 4]/4 4)/4+ 4+
tissue C Bo a Eeestlae Sala Gel alee
(2) Toxin is added to plasma ‘C”’ 1 = al s/o al saa
and used in cultures containing 2 + 4+)— 4+)/— — +/+ 4/4 +
tissue D () 38 J+ H+I— 4]— 4/— +/+ 4/4 4+
(3) Toxin is added to plasma “‘D”’ 1 = aoa). 4/2" ee
and used in cultures containing 2 — —|— +/— +)— —|— 4+)4+ +
tissue C 3 = =|— $)— 4H = Be
(4) Toxin is added to plasma ‘‘D”’ 1 eli lala , S eSeaeS e
and used in cultures containing 2 SS SS SS ae
tissue D 3 — —|j- -|- -|- -l- -]4+ +
+, Only a few cells are seen to grow.
0.9 percent NaCl solution or 1 part of 0.9 per cent NaCl solution
which contained varying amounts of diphtheria toxin. In the
third series of experiments no. 3 tissue C was tested in a medium
similar to the others but which contained plasma D. In the
fourth series no. 4, tissue D was tested in a media containing
plasma D.
Active growth is seen in no. 1 and good growth is seen in
no. 2. In no. 3 active growth is seen only when very small
amounts of toxin are added to the medium. No growth is seen
inno. 4. The growth is active in the controls of all 4 series.
TISSUE CULTURE AS A MEANS FOR ESTIMATING TOXIN 231
After the results of these experiments had been obtained the
question arose, ‘Do the cells of passively immunized animals
have any increase resistances to the toxin.” Is not the slight
increased resistance which they show due to the plasma which is
TABLE 11
Tissue: Fragments of the artery of chickens A and B cut into very small pieces and
teased apart with needles until single cells are liberated. These are washed three
times in 0.9 per cent NaCl solution
Medium: Plasmata prepared from the blood of chickens A and B. One of the plas-
mata, 1 part; various dilutions of toxin in 0.9.per cent NaCl solution and cell emul-
ston, 1 part
Control medium: The corresponding plasma, 1 part; 0.9 per cent NaCl solution and
cell emulsion, 1 part
TOXIN NUMBER OF TIMES DILUTED
TISSUE |MEDIUM TROL REMARKS
20 100 500 1000 2000 6000 10000
a ie os
an | au ait. oy Ae = Wish ag } Fragments
= a ais = ae a :
pms gee ve Ta ae i } Single cells
is zis os at
= sah nT, if eee sua \ Fragments
ed ie ee te
\ Single cells
= = |= aig ar = ais
aPaRM |= ORES ¢ =e = a \ racer eres
B A Be 56 Geer Sr. Ne =
a Aa ak Pale rt \ Single cells
= ary yoo = Sica lal + j
= Bar yas is 2s ee Bi
z 7 = ae |e us eeeshays mata fet a \ Fragments
= = = =) = ap wSe
\ Single cells
= seal = = \5= Site late
possibly retained in the fragments. To prove this it was neces-
sary to study the reaction of carefully washed single cells and
very small fragments.
In experiment 11 two chickens similar to those used in 9 and
which had been treated in similar manner were selected. The -
232 MONTROSE T. BURROWS AND YOSHIO SUZUKI
experiment was performed in the same manner except that single
cells or very small fragments which had been washed once or as
many as three times in 0.9 per cent NaCl solution were tested.
The method of washing described in the preceding paper
(Suzuki) (1).
The chickens are designated as A and B respectively. They
were one and one-half months old. A weighed 340 grams, B
weighed 345 grams. A received forty-eight hours previous to the
taking of plasma and tissue 1 cc. of diphtheria antitoxin into a
wing vein. :
The results of this experiment is given in table 11. As indi-
cated in the table it has ben demonstrated that both the plasma
and the cells of animals injected previously with diphtheria anti-
toxin have a greater resistance to diphtheria toxin than normal
chicken tissue and plasma.
CONCLUSIONS
Thus it has been possible to show that the tissue culture has a
very definite value for the study of toxic and antitoxic substances.
It broadens the possibility for the investigation of these sub-
stances and their properties. One toxin has alone been inves-
tigated. A study of this one has indicated, however, the value
of the method for determining the neutralizing power of a plasma
for any given toxin. Chicken tissues have been used in these
experiments. The tissues of other animals may be similarly
investigated.
REFERENCES
(1) Suzux1, Yosuro: 1918 This journal.
(2) Burrows, M. T. anp Neymann, C. A.: Studies on the metabolism of cells
in vitro. The toxicity of amino-acids for embryonic chicken cells.
Jour. of Exp. Med., 1917, 25, 93-108.
THE STUDY OF PROBLEMS OF IMMUNITY BY THE
TISSUE CULTURE METHOD
I. A STUDY OF THE CELLS AND BLOOD PLASMA OF ANIMALS
WHICH ARE NATURALLY RESISTANT AND OTHERS WHICH
ARE SUSCEPTIBLE TO DIPHTHERIA AND TETANUS TOXINS.
YOSHIO SUZUKI
From the Pathological Laboratory, Washington University Medical School, St. Louis,
Missouri!
Received for publication April 6, 1918
Although it has been appreciated for a long time that the tis-
sue culture method may be used for solving many problems in
immunity and infection only a few of those problems for which it is
particularly suited have so far been studied. A few years ago
Drs. Coca and Burrows undertook to ascertain by this method
whether the natural immunity of the rat for diphtheria toxin is
due to peculiarities of the cells of these animals or whether it is
due to the existence of substances present in the serum, blood
plasma or generally distributed throughout the body of the animal.
Owing to circumstances which arose this work was not com-
pleted. Dr. Burrows asked me to undertake this problem and I
am indebted to him for the general method of experimentation
used and the preparation of the manuscript.
If we are to come to a definite and comprehensive understand-
ing of the true nature of toxins and antitoxic substances we must
accumulate more careful quantitative data concerning the physi-
cal-chemical properties of these substances. Consequently, a
study of the distribution of protective substances in the organism
will aid in this direction. The tissue culture method allows such
studies. In the present paper the author wishes to describe a
method for determining the presence and studying the properties
1 This work was commenced in the Pathological Laboratory, Johns Hopkins
University and completed in this laboratory.
233
234 YOSHIO SUZUKI
of the protective substance in animals that are naturally immune
and in those that are susceptible to bacterial and other toxins.
Rats are resistant even to large doses of diphtheria toxin while
chickens and guinea-pigs are susceptible. The connective tissue
cells of rats grow readily in their own plasma and in the plasma
of guinea-pigs and chickens. In turn the connective tissue cells
of chickens and guinea-pigs grow readily in their own plasma and
in the plasma of either of the other two animals respectively.
It was thus possible, by the addition of varying quantities of
diphtheria toxin to the various plasmata, to test the relative re-
sistance of the tissue cells of each of these three animals in each
of the three plasmata.
It has been shown that the growth of cells in the tissue culture
is proportional to the size of fragments provided the fragments
are not greater than 1 mm. in diameter. The growth is also pro-
portional to the cellular content of the fragments. This pro-
portion to the cellular content means not only the number of
cells per unit area in the fragment but also the kinds of cells
present. Again the cells of embryos, foetuses and young ani-
mals grow more readily than those of adults. The growth not
varies only with these factors but also with the amount of fibri-
nogen present in the plamatic medium. Diluting the plasma
with a liquid substance causes variations in the growth (Burrows
(1)))
For this reason the control cultures in these experiments were
plasma diluted one-half with isotonic sodium chloride solution.
The toxin was diluted in isotonic NaCl solution and one part of
diluted toxin added to one part of plasma was the medium in the
experiments. In each case a tissue has been selected from which
a large number of fragments of similar cellular content can be
obtained. .
The toxin used in these experiments was obtained from the
Board of Health of New York City. A few of the samples con-
tained preservative and others did not. All toxins were tested
for the presence of bacteria. Fresh toxin was used and it was
filtered when it arrived in the laboratory and it was handled at
all times most carefully so that it did not become contaminated
CELLS AND BLOOD PLASMA RESISTANT TO TOXINS 235
with bacteria. A large number of the various bacteria that
might easily gain entrance, grow readily in the cultures and pre-
vent the growth of the cells. These when present can, however,
be detected and need not lead to error. The cultures were
made by placing small fragments of tissue 1 mm. in diameter in
a drop of liquid medium which was spread over a small part of
TABLE 1
Tissue: Fragments of the heart muscle of rats, one to five days old
Medium: Plasmata prepared from the blood of adult rats, guinea-pigs, and chickens.
One of the plasmata, 1 part; diphtheria toxin pure or diluted in 0.9 per cent NaCl
solution, 1 part
Control medium: The plasma corresponding to the experiment, 1 part; 0.9 per cent
NaCl solution, 1 part
TOXIN—NUMBER OF TIMES DILUTED*
TROL
0 5 10 20 50 80 100 | 150 | 200 | 250 | 300
MEDIUM a oe ae | ——— |
‘
, ;
oF
> ~
/
‘
'
A NOTE ON THE RELATION BETWEEN PROTEOLYSINS
AND HAEMOLYSINS
ARCHIBALD McNEIL anp REUBEN L. KAHN
From the Department of Bacteriology and Hygiene, New York University, New York
City
Received for publication May 23, 1918
This note is a report of studies undertaken with a view of
determining whether specific proteolysins are produced in animals
on protein injections, if a procedure simulating the production of
specific haemolysins be adhered to.
The question of the presence of specific proteolytic substances
in the blood of animals injected with proteins, is an old one and
no attempt will be made here to go over the range of literature.
Suffice it to recall that the so-called Abderhalden reaction for
eancer and pregnancy, which was based on the assumption of the
presence of such proteolysins in the blood (2), is now discarded.
More recently, Taylor and Hulton (1) also, studied this problem
from a biochemical viewpoint. They injected various proteins
into rabbits with a view of detecting the presence of proteolytic
ferments in the blood of these animals, obtaining negative results.
Although various attempts have been made from time to time
to attack this problem from an immunological viewpoint (2),
there is, to our knowledge, no work on record where a procedure
was observed similar to that recorded in this paper. Generally
speaking the procedure adapted for calling forth proteolysins
in the animal organism was similar to that observed in the
laboratories of the Department of Health of New York City,
when attempting to produce haemolytic antibodies, except that
some purified protein was substituted for red cells. After proper
immunization with a given protein, the serum was tested for
specifie proteolysins by digesting it with its protein antigen and
guinea-pig complement for a definite interval at 37°C. and sub-
sequently determining the increase in amino acid nitrogen.
295
296 ARCHIBALD McNEIL AND REUBEN L. KAHN
Two purified proteins, edestin from hempseed and phaseolin
from kidney bean (kindly furnished by Dr. Thomas B. Osborne)
were used in these experiments. Two rabbits weighing approxi-
mately 5 pounds each, served as the experimental animals. The
injections were made intravenously.
The following is a table of the quantity of protein injected and
time of injections.
RABBIT A RABBIT B
DATE, 1915 SS. |= =e
Quantity of Quantity of
phaseolin injected | edestin injected
mgm mgm
October ™ (Qs 215i MT Aes ee ee eee 50 50
October: «7 /28s226 6 & See eet eee Stee ee 75 75
October SOs an 4 scenes Soe eee eee 100 100
INGVemben= elses see eee eee eae era: 125 125
INO VEMDer Cae cree oe ee Oe eet. oa ee 150 150
Both animals were bled under anesthesia, on November 9,
six days after last injection. The sera were separated from the
clots with much care to prevent hemolysis, and were divided
into two portions, one-half being inactivated at 56°C. for a half
hour, the remaining half being used in an unheated form.
Our measure of ferment action consisted in the amino acid
increase after incubation for six hours at 37°C. of mixtures of the
rabbits’ serum with guinea-pig’s complement and the specific
proteins.
The amino acids were determined by means of the Van Slyke
micro amino apparatus. The advantage of this procedure over
other measurements of proteolytic action are well pomted out
by Van Slyke and his co-workers in their recent paper on
the Abderhalden reaction (3). ‘First, it is quantitative, and
permits accurate results with the small amounts of material
available. Secondly it is specific for proteolysis; it permits one
to follow the chemical change which is characteristic of protein
hydrolysis.”
The principle of this gasometric method (4) for the determina-
tion of amino-nitrogen, is based on the fact that aliphatic amino
groups react with nitrous acid with the liberation of nitrogen gas.
RELATION BETWEEN PROTEOLYSINS AND HAEMOLYSINS 297
Thus R.NH, + HNO, = ROH + H,O + N:.
The quantity of nitrogen gas liberated serves as the measure
of the amount of amino nitrogen present in the unknown solution.
Another factor was essential for the accurate determinations
of free amino nitrogen, namely, the removal of the proteins from
the serum. The method adapted for this purpose was first sug-
gested by Rona and Michaelis (5) and it has since been success-
fully employed by other investigators (6) (7).
Briefly stated the procedure was as follows: 2 ec. of serum
were diluted to 20 ec. in a beaker and heated to boiling. 1.5
ee. of colloidal ferric hydrate were added drop by drop the
mixture being shaken with each addition. Precipitation of pro-
teins was then complete. 1.0 cc. of a 20 per cent MgSO, solution
was added to coagulate the excess of iron.
The solution was then filtered through a hardened paper into
a 100 ec. evaporating dish, the filtrate being water clear. After
the filtration was completed the precipitate was washed by means
of a hot water wash bottle into the original beaker, about 20 ce.
of water being used in the process.
The mixture was again heated to boiling and the contents of
the beaker were filtered into the first filtrate the same filter
paper being used. Finally, the filtrate was evaporated nearly to
dryness and it was redissolved in 0.5 ec. of water just previous to
the amino nitrogen determination.
The digestive mixture consisted uniformly of 2 cc. each 1 ce.
of rabbit’s serum, 0.5 ec. of complement and 0.5 cc. of protein
suspension containing 0.005 gram of protein. This quantity of
protein was found to be sufficient in view of the fact that by means
of the micro-amino apparatus one can measure accurately small
fractions of a milligram of amino nitrogen.
The following tabulation gives the procedure in detail with the
results obtained.
Tube A contained 1-cec. serum rabbit A (phaseolin), 0.5 ce. guinea-
pig’s complement, 0.5 cc. phaseolin suspension in saline. Placed in
the incubator for six hours, after which the protein was precipitated
THE JOURNAL OF IMMUNOLOGY. VOL. II. NO. 4
298 ARCHIBALD MCNEIL AND REUBEN L. KAHN
and the amino nitrogen determined. Quantity of amino nitrogen gas
found = 0.310 ce.
Tube B (control) contained same constituents of tube A, except that
the amino nitrogen was determined immediately, without incubation.
Quantity of amino nitrogen gas found = 0.310 cc.
Tube C contained same constituents as tube A except that the serum
was first inactivated for one-half hour at 56°C. The complement and
phaseolin suspension were then added and placed in the incubator for
six hours. Quantity of amino nitrogen gas found = 0.320 ce.
Tube A’ contained 1.0 ce. serum rabbit B, 0.5 ec. complement, 0.5 ce.
edestin suspension in saline. Placed in incubator for six hours, after
which the protein was precipitated and amino nitrogen determined.
Quantity of amino nitrogen gas found = 0.360 cc.
Tube B’ (control) contained the same constituents as tube A’ except
that the amino nitrogen was determined immediately, without incuba-
tion. Quantity of amino nitrogen gas found = 0.345 ce.
Tube C’ contained the same constituents as tube A’ except that the
serum was first inactivated for one-half hour at 56°C. The complement
and phaseolin suspension were then added and placed in the incubator
for six hours. Quantity of amino nitrogen gas found = 0.360 cc.
These findings indicate that the serum of rabbits immunized
against protein, possesses no greater proteolytic activity than
normal serum.
CONCLUSIONS
An attempt was made to find whether proteolytic substances
are produced in rabbits on protein injections, if a procedure simu-
lating the production of haemolytic substances in these animals,
be resorted to. Proteolysis was determined by observing the
increase in amino acid nitrogen after digesting mixtures of the
immune serum, the specific protem and complement for a given
period. The results gave no evidence of any increase in amino
acids under these conditions, which would indicate that haemoly-
sis and proteolysis are probably two distinct phenomena.
RELATION BETWEEN PROTEOLYSINS AND HAEMOLYSINS 299
REFERENCES
(1) TayLor anp Hutton: Jour. Biol. Chemi., 22, 59, 1915.
(2) BRONFENBRENNER: Journ. Lab. and Clin. Med., 1915, 1, 79. (Reviews the
literature. )
(3) Van Styke, VinoGrap-VILLCHUR AND Losers: Jour. Biol. Chem., 1915, 23
377.
(4) Van Styrxe: This method is fully described in the following numbers of the
Journal of Biological Chemistry: 1911, 9, 185; 1912, 12, 275; 1913-1914,
16, 121; 1915, 23, 407.
(5) Rona anp Micuae is: Biochem. Zeitsch., 13, 121.
(6) C. G. L. Wotr: Jour. Physiol., 1914, 69, 89.
7) Van SLYKE: VINOGRAD-VILLCHUR AND LOSEE: Loe. cit.
THE INFLUENCE OF ARSPHENAMINE AND MERCURIC
CHLORID UPON COMPLEMENT AND ANTIBODY
PRODUCTION
IKUZO TOYAMA anp JOHN A. KOLMER
From the Dermatological Research Laboratories of the Philadelphia Polyclinic
Received for publication May 31, 1918
Numerous investigations within recent years have indicated
that certain drugs may induce a state of temporary immunity to
trypanosome infections by stimulating the antibody producing
tissues, the leucocytic mechanism or both, or, combine with
antibodies and render the latter more active.
Ehrlich and Shiga (1) have shown that mice infected with caderas
and treated with one or more injections of trypan-red, developed a
temporary immunity which could not be ascribed to an antibody
response following infections with the parasites alone or to the presence
of unexcreted dye, but rather to the presence of antibodies in response
to the stimulating influence of the drug; later Ehrlich (2) demonstrated
the same phenomenon with 7. brucez and Halberstaedter (3) in similar
studies found the immunity highly specific, that is, mice infected with
dourine and treated with trypan-red developed an immunity to dourine
alone and not to other trypanosomes as 7’. brucet or vice versa. Cor-
roborative evidence of the apparent effect of this and other drugs upon
antibody production was given later by the extensive work of Terry (4)
who found that a strong immunity against surra of India was obtained
by injecting mice with dyes either alone or in combination with acetyl-
atoxyl. That the action of the drugs is indirect rather than wholly
trypanocidal, was seemingly shown by the fact that large intraperitoneal
injections of surra and caderas were capable of infecting mice when
introduced as early as twenty-four hours after the drug and before the
latter had been wholly excreted.
Further indications of the possible important relation of drugs to
immunity is shown in the reports of severai homeopathic physicians as
in that Watters (5), who claimed that the administration of calcium
301
302 IKUZO TOYAMA AND JOHN A. KOLMER
sulphide increased the opsonic index to staphylococci; of Mellon (6)
who found that the administration of baptisia influences favorably the
production of group agglutinins for typhoid and other closely related
bacteria and that veratrum viride increased the opsonic index to pneu-
mococci; of Wheeler (7) who claims that phosphorus increases the op-
sonic index of human serum to the tubercle bacillus; of Wesselhoeft (8)
whose experiments were interpreted as indicating the curative effects
of quinine in malaria, could not be ascribed entirely to its parasiticidal
activity but probably in part to a favorable influence upon the pro-
duction of anti-plasmodial antibodies; and of Hooker (9) who showed
that the administration of phosphoric acid, arsenious anhydrid and
mercuric chlorid homeopathically to normal persons, resulted in the
elaboration of agglutinins and complement fixing antibodies for B.
typhosus, B. paratyphosus A and B and B. dysenteriae. In several of
these investigations the drugs alone were administered to healthy per-
sons and the appearance of an increase of certain group antibodies in
the blood serum was interpreted as an increase of normal or natural
antibody and an indication of the possible stimulating influence of these
drugs upon antibody producing tissues and a means of their curative
value in certain diseases. Likewise the investigations of Arkin (10)
concerning the influence of drugs upon phagocytosis may be mentioned
in this connection; medicaments which have an inhibitory action upon
oxidative processes as chloral, morphine and ether were found to depress
phagocytosis while mercuric and other chlorids, colloidal metals, strych-
nine, arsenic and others, were found to stimulate phagocytosis in vitro
and 7n vivo.
Following the introduction and encouraging results of arsenical
compounds in the experimental chemotherapeusis of protozoan in-
fections, several investigators have a studied their possible influence
upon antibody production and particularly the influence of dioxydi-
amidoarsenobenzol (salvarsan), with the result that a general impression
exists that part of the curative influence of dioxydiamidoarsenobenzol
in spirochaetic and trypanosome infections, is to be ascribed to the
influence of the drug in stimulating the production of protective and
curative antibodies in addition to its powerful parasiticidal activity.
Aggazzi (11) found that arsenious acid, atoxyl and arsenophenylglycin
increased the output of typhoid agglutinin; Friedberger and Masuda
(12) claim that salvarsan increases the content of normal agglutinins
and hemolysins in the serum; Boehneke (13) found that the adminis-
tration of salvarsan may be followed by an increase of diphtheria anti-
COMPLEMENT AND ANTIBODY PRODUCTION 303
toxin and of various bacteriolysins, opsonins and precipitins, but not of
complement binding substances; Weisbach (14) also claims that the
administration of salvarsan results in an increase of agglutinin and hemo-
lysin, while Reiter (15) was unable to note any such influence, his experi-
ments indicating that large doses of the drug lowets resistance to various
bacteria.
As further indications of the probable important relation of certain
drugs to immunity, are several reports indicating that their administra-
tion may be followed by an increase of complement in the serum. Weil
and Duport (16) have reported that the intravenous administration of
sodium bicarbonate to rabbits resulted in an increase of serum com-
plement; Fenyvessy and Freund (17) claim similar results with the
intravenous administration of calcium chlorid and Ciuea (18) found
that the injection of tartar emetic and salvarsan was followed by an
increase of serum complement in normal and trypanosome-infected
animals, while the administration of atoxyl caused a decrease of com-
plement in the serum of normal animals and in a proportion of try-
panosome-infected animals.
EXPERIMENTAL
Since the results of these investigations have indicated that
salvarsan may exert an important effect upon complement and
spirocheticidal antibodies, we have conducted a series of experi-
ments for the purpose of a further study of the probable influence
of arsphenamine! (arsenobenzol) and mercury upon pales
and complement production, as follows:
1. A study of their probate influence upon the production of
immune antisheep and antihuman hemolysins and agglutinins for
sheep and human erythrocytes in rabbits.
2. Upon the production of immune typhoid agglutinin in
rabbits.
3. Upon hemolytic complement and normal antisheep hemoly-
sin in rabbit serum.
1 Arsphenamine is the trade name proposed by the Federal Trade Commission
for salvarsan and its substitutes. Throughout this study, the arsphenamine
prepared by Dr. J. F. Schamberg, Dr. Geo. W. Raiziss, and Dr. John A. Kolmer
in the Dermatological Research Laboratories of the Polyclinic and known as
arsenobenzol, was employed in alkaline solution.
304 IKUZO TOYAMA AND JOHN A. KOLMER
4. Upon normal typhoid agglutinin and hemolytic complement
in human serum.
We have included a study with mercury bichlorid, because
salts of mercury do not appear to have been previously employed
in experiments of this nature, while their curative influence in
syphilis is well known.
Additional experiments similar to those recorded in this paper
but with the employment of animals infected with trypanosomes
are being conducted, inasmuch as the results may vary according
to the nature of the stimulant (erythrocytes, typhoid bacilli or
trypanosomes) used, but we have considered it advisable to record
the work finished under the above plan.
I. The influence of arsphenamine and mercury upon the pro-
duction of immune hemolysins and hemagglutinins
Experiment 1. In this experiment the sera of six large healthy
rabbits were given preliminary tests for the presence of antisheep
hemolysin and sheep agglutinin and 1 ec. of a 5 per cent suspension
of washed sheep cells per kilogram of body weight injected intra-
venously every three days followed two hours later by the intra-
venous administration to rabbits 1 and 2 of arsphenamine in
dose of 0.01 gram per kilo (equivalent to 0.6 per 60 kilos) and to
rabbits 3 and 4, of bichlorid of mercury in dose of 0.0001 gram
per kilo (equivalent to 0.006 gram or about j grain per 60 kilos);
rabbits 5 and 6 were controls and received injections of cells only.
All animals were bled from an ear three days after each injection
of cells and drugs, the sera separated and inactivated and the
titer of hemolysin determined in the presence of 1 cc. of 1:20
dilution of guinea-pig’s complement and 1 cc. of 2.5 per cent
suspension of washed sheep cells; the agglutinin titer was de-
termined with 1 cc. of 2.5 per cent suspension of sheep cells
alone; both readings were made after incubation of 38°C. for two
hours.
The results of this experiment are shown in tables 1 and 2 and
may be summarized as follows:
305
COMPLEMENT AND ANTIBODY PRODUCTION
oe eS eee ee ee ee
!
Peld | 91} OT -L} OF: ssey | OL:T sseyT!] OT:T
8897 | OL ‘Tf Ssey | OL:] sseT (Dade lira i pai ee rae JO1}U0D | 9g
991+) $6:T} OS:T} OL:T OT-T 8897 | OL: 1 S897 | OL: 1 S807 | OL:T sso'T A el ce en ease ‘Jor}UOD | ¢
991 -T( OST.) OF :T] OOT :T 0S :T OL 1 S8e'T | OL :T Sser'T | OT: ssey | O00’ |°****Arnosou TO THOS Sy
OS *T) OG :T [SZ ‘1 | OOT :T OOT :T OL 1 SST | OL:T SsoT | OT 1 ssey | T000'0 |°***-Amnosom OTH OTE iets
SGT ‘1 (SCT :T SZ :1 | 99T :1 OOT :T OL -T S8oT | OL:T sso | OT :T sse7y LOUOT ae as “jozueqouesiy | Z
99T -T /99T =T S20 :T} OG :T O¢ :T OTT S8o'T | OL: Sse | OT ‘7 sse7 10°0 |" """"*"Jozueqouesry | |
men Ot a worzoo fut worqo eur uorqoeful uworqzoelfur uorqoofur
maceisa ane re WWnoy Pay puoveg FSI Are urunty arg Ox uaa shee ‘ON
asoa TVWINY
S9OuUd GNV STTGO JO NOILOACNI uaLay UdLIL NINILQIODVNaAH
umn bbouay daosysyup aunwun fo uoyonposd ayy uodn fiunasau pun aurupuaydsiv fo aauanyfuy
6 WIV
PFC |OOO9T :T |OOOOT | OOF :T | 991: |OOT:T 0 OT -T ssoT aie Sede WAP aos te te fetju0H | 9
OO0LE :T \OOSZT :T |O000T :T [00SZ:T jOOOL:1 | O¢:T 0 OT -T ssoT SE | alate Sic Pd 3 [OtMOO 5) 2S
Q00S -T | 00S :T| 000¢ :T jO00E :T [ooOT :T SG ‘T 0S :T OL :T SST | 1000°0 |" ****Amnorow ‘sopyorg | F
OOSS +I | 000S :T |OO00T :T JOOS: |O00T :1 SG ‘T OT -T S8eT | OL 1 ssey | 1000'0 | °° ****Arno sour OT Gore &
OO0S -T | 000¢ :T | O00 :T jo0Sz:T [o00T :T OL :T 889T | OL 1 S8eT | OT :T ssorT LOMO es ees “lO ZueqouesIy | Z
000¢ :T | 000S :T |O000T : 1 |000¢ :t [ooOT :1 OT :T G3 :T OT -T ssory T0°0 [°° ""*""***Jozueqoussry |
morjoef ar} uoroefut uoryefur | uorjoefut mor}o0fut mor49e fut aorjoef ul worqoofur
Yueaeg | qIXIg DHE | Wanog | pays, puoveg a AISULUNPOIT || oar ama cave
asoa pean TVWINY
S9Oud GNV STIG0 JO NOILOGINI uaLavy UGLIL WALIONGH
ee ee
urshjomoy dasysyun ounuur fo uwononpoid ayy uodn fhaunasam pup auiunuaydsip fo aruanpzfuy
I
ATaAVL
306 IKUZO TOYAMA AND JOHN A. KOLMER
1. In the production of hemolysin neither arsphenamine nor
bichlorid of mercury appeared to exert a stimulating influence,
inasmuch as the amount of hemolysin produced in the drug
treated animals was not higher than that in the contrdls; on the
contrary they appeared to produce less hemolysin. It is possible
that the doses of arsphenamine and mercury were too large for a
three day interval of administration, but all animals appeared
to stand the injections well and the majority gained slightly in
weight during the course of the experiment.
2. Both drugs however, appeared to increase the output of
hemagglutinin to a slight extent, as shown in table 2 by the more
rapid production of this antibody in the drug treated animals.
Experiment 2. In this experiment a series of six rabbits were
bled and the inactivated sera were tested for normal antisheep
hemolysin and hemagglutinin. Each animal was then given a
daily intravenous injection of 1 ec. of a 1 per cent suspension of
washed sheep cells per kilogram of body weight followed immedi-
ately with an intravenous injection of the following to the first
four animals:
Rabbit 27. 0.004 gram arsphenamine per kilo (0.24 gram per
60 kilos).
Rabbit 28. 0.001 gram arsphenamine per kilo (0.06 gram per
60 kilos).
Rabbit 29. 0.00004 gram bichlorid mercury per kilo (0.0024
gram or about 4% grain per 60 kilos).
Rabbit 30. 0.00002 gram bichlorid mercury per kilo (0.0012
gram or about g grain per 60 kilos).
All animals were bled from the ear twenty-four hours after the
injection of cells and drugs and the hemolysin content of the in-
activated sera was determined by titration in the presence of 1 ce.
of 1:20 guinea-pig’s complement and 1 cc. of 2.5 per cent sus-
spension of sheep cells; the results of hemaggtutination tests were
read after incubation at 55°C. for twelve to sixteen hours.
The results of this experiment are shown in the tables 3 and 4
and may be summarized as follows:
1. Generally considered the daily administration of cells and
drugs did not appear to result in a greater production of hemolysin
COMPLEMENT AND ANTIBODY PRODUCTION
Influence of arsphenanine a
307
TABLE 3
nd mercury upon the production of immune antisheep
hemolysin
HEMOLYTIC TITER AFTER INJECTION OF CELLS AND DRUGS
9 | DOSE Pe r= A a ra re | s
3 HALEN PER KILO 2.8 2 2 2 2 8 2 at
3 fe) 8 Sec) coc) Somes ae
5 oa | #8| 88) 38 ja | 28 $8 al
< Ay cm 7) al Fe & 1) mM
27| Arseno-
benzol. .|/0.004 1: 10!) 1: 20/1; 40) 1: 40 | 1: 40 |1 40 1: 500 | 1: 1000
28| Arseno-
benzol. .|0.001 1: 20} 1: 40) 1: 40) 1: 100) 1: 125] 1: 250 | 1: 10000) 1: 25000
29| Bichlor. Less
mercury|0.00004) 1: 10| 1: 20) 1: 20) 1: 100) 1: 125] 1: 500 | 1: 5000 | 1: 16000
30} Bichlor.
mercury/|0.00002} 1: 20} 1: 40) 1: 40! 1: 166} 1: 250} 1: 2500} 1: 50000) 1: 50000
Less
Sti Control.../ 0 1:10] 1: 10) 1:20) 1:40 | 1:40 | 1: 500 | 1: 16000) Died
32) Control... 0 1: 10) 1: 20) 1: 20) 1: 100) 1: 40 | 1: 500 | 1: 16000} 1: 50000
TABLE 4
Influence of arsphenanine and mercury upon the production of immune antisheep
agglutinin
AGGLUTININ TITER AFTER INJECTION OF CELLS AND DRUGS
¢ Borate |AOEM she sme lie oe akabeBBIS ABS) od al MAbs
2 28 23 33 2 ae ee ae a8
z £4 a 37 aA ie BA ue BF
Less
27; Arseno-benzol |0.004 Gene Uses Sy eles SON eis i) ibe 220i ( ile 0)
Less | Less Less
28} Arseno-benzol \0.001 Ibe eee Ie ai lorsye| Sey ale Sie alo sy | ier) are akg)
29| Bichlor. Less | Less
mercury... .|0.00004) 1:3 | 1:3 | 1:3 | 1:10] 1:10] 1: 40) 1: 125) 1: 125
30} Bichlor. Less | Less | Less | Less
mercury... -|0.00002) 1:3. | 1:3. | 153 | 0:3) 1220) 1:40) 1: 166) 1250
Less | Less
oll) Control... .... 0 PS hes) P38) | LSP Oe 20) 14 OR Died
Less | Less | Less
Bel Control... 5... 0 Se LS ssi) ero elie O) eel,
140) ts.
{ i
308 IKUZO TOYAMA AND JOHN A. KOLMER
than was observed in the controls receiving cells alone; rabbits 27
and 29 receiving the larger doses of arsphenamine and mercury
produced less hemolysin than the animals receiving the smaller
doses, indicating that the larger doses tended to depress anti-
body production even though the general health and body weight
of all animals was maintained at a normal level.
2. Likewise both drugs did not appear appreciably to increase
agglutinin production; on the other hand the larger doses of both
arsphenamine and mercuric chlorid appeared to limit agglutinin
production, as was true of hemolysin production just described.
Experiment 3. In this experiment a series of six large healthy
rabbits were bled and the inactivated sera tested for normal
antihuman hemolysin and hemagglutinin. Each animal was
then given daily intravenous injections of washed human cells
in dose of 1 ec. of a 5 per cent suspension per kilogram of body
weight; immediately after these injections the animals were in-
jected intravenously with the following:
Rabbits 13 and 14. 0.001 gram arsphenamine (0.06 gram per
60 kilos).
Rabbits 15 and 16. 0.00001 gram bichlorid mercury (0.0006
gram or about i35 grain per 60 kilos).
Each animal was bled every three days and the inactivated
sera titrated for hemolysin with 1 ce. of 1: 20 guinea-pig’s comple-
ment and 1 cc. of a 2.5 per cent suspension of human cells; also
for hemagglutinin employing a 1 per cent suspension of cells and
reading the results after incubation at 55°C. for twelve to sixteen
hours.
The results are shown in tables 5 and 6 and may be summarized
as follows:
1. The production of antihuman hemolysin was slight with all
animals including the controls; neither drug appeared to influence
hemolysin production.
2. Both drugs appeared slightly to increase the output of
hemagglutinin, as shown by a somewhat quicker response on the
part of the drug treated animals and the final higher titer of the
serum of two (nos. 14 and 15).
oa
_
4
|
7
:
>
‘
_.
POT Ie t
COMPLEMENT AND ANTIBODY PRODUCTION 309
TABLE 5
Influence of daily injection of arsphenanine and mercury upon the production of
NO,
ANIMAL
immune antihuman hemolysin
HEMOLYTIC TITER AFTER INJECTION OF CELLS AND DRUGS
DRUG DOSE jl
EE peers) Preliminary After two After five After nine
injection injections injections injections
Arsenobenzol.. ./0.001 Less 1:10 | Less 1:10 | Less 1:10 | Died
Arsenobenzol....|0.001 Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10
Bichlor.
mercury...... 0.00001) Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10
Bichlor. j
mereury...... 0.00001; Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10
Wontroll)...oc.< 2: 0) Less 1:10 | Less 1:10 | Less 1:10 | Less 1: 10
Controle. «oe 0 Less 1:10 | Less 1:10 | Less 1:10 | Died
TABLE 6
Influence of daily injections of arsphenanine and mercury upon the production of
— tet | ANIMAL BO.
Or he Ww
ee
on SD
immune antihuman agglutinin
aT.
AGGLUTININ TITER AFTER INJECTION OF CELLS AND DRUGS
DRUG DOSE
PER KILO! Preliminary After two After five [After nine
injection injections injections injections
Arsenobenzol......../0.001 Less 1: 10 | Less 1: 10 1:50 | Died
Arsenobenzol......../0.001 Less 1:10: | Less 1: 10 aes ele 250
Bichlor. mercury... .|0.00001; Less 1:10 | Less 1: 10 LO} 500
Bichlor. mercury... .|0.00001| Less 1:10 | Less 1: 10 L216.) 1 166
Gontrolete.c; cic > 0 Less 1:10 | Less 1:10 | Less 1:10 | 1: 166
Monroe os =. han. - 0 Less 1:10 | Less 1:10 | Less 1:10 | Died
The influence of arsphenamine upon normal antisheep hemolysin
Experiment 4. In this experiment we have studied the influence
of a single large dose of arsphenamine administered intravenously
to rabbits, upon the total hemolytic activity (normal hemolysin
and complement) of their sera for sheep cells and upon the normal
hemolysin alone; as shown by Kolmer and Williams (19) the
sera of about 70 per cent of normal rabbits contain antisheep
hemolysin and about the same percentage (63 per cent) was
found in the present series.
310 IKUZO TOYAMA AND JOHN A. KOLMER
Each rabbit was bled immediately before, three hours and
again eighteen hours after the intravenous injection of arsphena-
mine in dose of 0.06 gram per kilo (equivalent to 3.6 grams
per 60 kilos of body weight). The serum was separated and
its hemolytic activity determined by titrating varying amounts
of active serum with a constant dose of 1 cc. of 2.5 per cent sus-
pension of washed sheep cells. Each serum was then heated at
55°C. for thirty minutes and its content of normal antisheep
hemolysin determined in a titration employing 1 ce. of 1:20
dilution of guinea-pig’s complement and 1 ec. of 2.5 per cent sheep
TABLE 7
The influence of arsphenanine upon the total hemolytic activity of the sera of
normal rabbits
= s RESULTS WITH SERA RESULTS WITH SERA
peal anon ci tet ae ee pag AFTER
£ } DOSE
at PER KILO . . . . - . . .
elal2i=(8l8lelslslsisislelsls/si8is
o o o Oo So o o o o o o o Oo o o o o o
gram |
606 | 0.06 |C|C|C|M| S| S|C/M/M] S| S| N| C|/M|M|M| S|N
607 | 0.06 |C|/M|M|M/M|S|C|M|M|M|M|s|C/M|M/|M|S]S
608 | 0.06 |C/M|M| S/N|N/|S|S/|N|N|N|N|M/M| S|N|N/N
611 | 0.066 |C/M|M/S|S|S/|M/M|S|S|S/S|C|M!S|S|S]S
618 | 0.066 |C|M|M/S|S/S/S|S|S|S/S|S/S|S|S|S}S|8
622 | 0.06 |M|/S/S|S/S|/S|S|S|S/S| S/S] S]S|S|S|S|S
C = complete hemolysis; M = marked hemolysis; S = slight hemolysis; N =
no hemolysis.
*
cells. The results in both tests were read after incubation in a
water bath at 38°C. for an hour. The results of a few of these
tests are shown in tables 7 and 8 and the whole has been sum-
marized as follows: -
a. The total hemolytic activity of the sera of 25 rabbits tested
three hours after the preliminary tests and injection of arsphena-
mine showed: No change with the sera of 10; a decrease with the
sera of 15.
b. Tested eighteen hours after the injection of arsphenamine,
the sera of 22 of these rabbits surviving showed: No change with
COMPLEMENT AND ANTIBODY PRODUCTION Salle
the sera of 9; an increase of hemolytic activity with the sera of 2;
a decrease of hemolytic activity with the sera of 11.
c. The content of normal antisheep hemolysin in the sera of 30
rabbits tested three hours after the preliminary tests and injection
of arsphenamine showed: No change with the sera of 27; a slight
decrease with the sera of 3.
d. Tested eighteen hours after the injection of arsphenamine,
the sera of 25 of these rabbits surviving showed: No change
with the sera of 22; a slight decrease with the sera of 3.
From these experiments it would appear that a single large dose
of arsphenamine tends to decrease the content of hemolytic
TABLE 8
The influence of arsphenanine upon the antisheep hemolysin in the sera of normal
rabbits
RESULTS BEFORE RESULTS THREE HOURS RESULTS EIGHTEEN HOURS
INJECTION AFTER INJECTION AFTER INJECTION
DOSE —
NO, PER
KILO . S) 3 ) : ) SS) cS) . : ) ) 9
S 3 o o o 3 8 Co o Co S S o o o
Saat ae |) Soles bP Sel rey | a |S Se S ke Wis |eSeS! |S
| o o o o Co o o o o Oo o Oo Oo o So
gram Fx,
Pee or ss. | IS. | aol N | No SSeS | NUNS Sis |) Nahe IN;
cee, 0.06 | S| N| N|N{|N/ S|NIN| NIN] S|] S|] Si NIN
mo006! S|/S;)s|S/S|/S| sis] sr s|sis|s| sis
feos | ON No N | NN | ON | ON CNS ON CN | NN | WN ON EN
meaowe | S| N | oN | NNN [NE NO) ANUIONG), NUN ON | NaN
my @.06; S| NIN; N|NIN| NI -N| NI NI N|N/ N| NN
S = slight hemolysis; N = no hemolysis.
complement, normal hemolysin or both and particularly the
complement, for a period of at least eighteen hours following the
administration of drug. Smaller doses of arsphenamine however,
administered to persons, may produce an increase of hemolytic
complement following a primary depression, as shown in the
following experiment.
II. The influence of arsphenamine wpon hemolytic complement of
persons
Experiment 5. The fresh active sera of 20 persons suffering
with syphilis were titrated in varying doses to determine their
ile IKUZO TOYAMA AND JOHN A. KOLMER
content in hemolytic complement? with a constant unit of anti-
sheep hemolysin and 1 cc. of 2.5 per cent suspension of washed
sheep cells; each person was then given an intravenous injection
of 0.4 to 0.6 gram arsphenamine in the clinic of Dr. Jay F.
Schamberg in the course of their regular treatment. Blood was
drawn one and eighteen hours later and the complement content
was determined with each fresh serum and the same unit of
hemolysin and dose of cells. The results observed with the sera
of 19 persons secured one hour after the administration of arsphen-
amine were as follows (a few being given in table 9 as examples) :
TABLE 9
The influence of arsphenanine upon the hemolytic complement of adult syphilitic
persons
RESULTS WITH SERA RESULTS WITH SERA RESULTS WITH SERA
Serons mgaonon | OFM EOS artax | mcneney roa oo
a1 nN =| =~|/8 3 om 1A = - Ss Silal|ta = SE || S
Oo o o Oo o o =) o Oo o —) o o o o o o o
gram
BB: 0.6 |C|M|M!|S|S|N/]|CIM/IM/| S| S|N|C|C|M] S| S/N
8.8. 0.6 |CiIM|S|S|N/N|C{/M|S/N|N/|-N/C|C!}C) C|M/S
G. G. 0.6 |C|C|M!| S| Ni N| C)LCIM| S| SiN] C] CC} C) per
Ja eise 0.6 1Ci GiC| S| 8S) S| CrRC|Mi S| S| S)C) CC) Sitsits
PC. 0.6 | Cl Ci] Cc] C|M| S| C} C| €|CiM/| SC) CC) CiMiaiets
S: HE: 0.6 C|C}C}M/ S| S| C}/C|C)M) S/N) C)C!Cc;} Cc) C|M
C = complete hemolysis; M = marked hemolysis; S = slight hemolysis;
N = no hemolysis.
No change with the sera of 15; a decrease of hemolytic activity
(complement) with the sera of 4.
Seventeen of these persons returned to the clinic on the fol-
lowing day, and their sera secured about eighteen hours after the
administration of arsphenamine showed: No change with the
sera of 5; an increase of hemolytic activity (complement) with the
sera of 9; a decrease of hemolytic activity (complement) with the
sera of 3.
2 As the majority of these sera also contained normal antisheep hemolysin and
as this was not removed, the results indicate the influence of the drug upon the
total hemolytic activity of the sera rather than upon complement alone.
COMPLEMENT AND ANTIBODY PRODUCTION Bills:
IV. The influence of arsphenamine and mercury upon immune
and normal typhoid agglutinin
Experiment 6. In this experiment the sera of a series of six
rabbits were titrated for typhoid agglutmin and then each
animal was given intravenously 1 ce. of a heat-killed monovalent
typhoid vaccine per kilogram of body weight; each cubic centi-
meter of vaccine contained two billion bacilli. The vaccine was
administered every three days and it was followed two hours
TABLE 10
Influence of arsphenanine and mercury upon the production of immune typhoid
agglutinin
|
| AGGLUTININ TITER AFTER INJECTIONS OF VACCINE AND DRUGS
DOSE |
fe} De I S eI = i 5
4 DRUG PER gS a) i) 3 iS EB 3
z aq | ee | Bao| ae a eae les
4 & = BR a & iS RD
gram
Less | Less
7 | Arsenobenzol..../0.01 L744) 14) | Dred 0 0 0
Less | Less | Less | Less
8 | Arsenobenzol..../0.01 4S 4a Aa ae S64 all ST Oai 048
9 | Bichlor. Less | Less
MOETCULY.. 46.. .: 0.0001} 1:4 | 1:4 | Died 0 0 0 0
10 | Bichlor. Less | Less | Less
Mercury... .. 0.0001} 1:4 | 1:4 | 1:4 | 1: 768 | 1: 1024] 1: 6144] 1: 8192
Less | Less
Mie Gontrol. 2... .. is 20 1:4 1:4 | 1:24 | 1: 1024) 1: 1024) 1: 6144; 1:12000
Less | Less
Reet ontrol......... 0 Lira ler aS | Ue DAS ae 768 leios. [els 2043 hla 3072
later by the intravenous administration to the first two rabbits
of arsphenamine in dose of 0.01 gram per kilo (equivalent to
0.6 per 60 kilos) and to the third and fourth rabbits, of bichlorid
of mercury in dose of 0.0001 gram per kilo (equivalent to 0.006
gram or about 7 grain per 60 kilos). Each animal was bled
three days after the injections and the inactivated serum was
titrated for typhoid agglutinin in a macroscopic test, the results
being read after incubation at 55°C. for twelve to sixteen hours.
The results of this experiment are shown in table 10 and indi-
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 4
314 . IKUZO TOYAMA AND JOHN A. KOLMER
cate that the influence of both drugs was an inhibition on agglu-
tinin production; possibly this effect was due to the use of two
large doses of both arsphenamine and bichlorid.
Experiment 7. The sera of 24 persons suffering with syphilis
were titrated for their content in normal typhoid agglutinin and
the tests repeated one and again eighteen hours after the first dose
of 0.6 gram of arsphenamine received in the treatment of their
infection. The results of these tests showed no increase of normal
agglutinin after the one injection of arsphenamine.
SUMMARY
1. Numerous studies indicate that several drugs and partic-
ularly arsenic compounds, may stimulate antibody production
and that their curative effect in the treatment of disease and
particularly syphilis and experimental trypanosomiasis, may be
ascribed to this influence upon antibody production in addition
to their direct parasiticidal influence.
2. In our experiments, the intravenous administration of
arsphenamine to normal rabbits in doses varying from 0.01
gram to 0.004 gram per kilogram of body weight (equivalent to
0.6 to 0.24 gram per 60 kilos), did not result in an increased out-
put of immune hemolysin to sheep cells, but rather suppressed
hemolysin production; smaller doses did not appear to retard
hemolysin production to sheep’s and human cells but likewise
they did not result in an increase. The smaller doses, however,
generally produced a slight increase of agglutinin for sheep’s
and human cells.
3. Similar results were observed with mercuric chloride; large
doses appeared to suppress hemolysin and agglutinin production
while smaller doses tended to increase the production of hemag-
glutinins but not that of the hemolysins.
4. A single large dose of arsphenamine (0.06 gram per kilo)
administered intravenously to rabbits reduced the hemolytic
activity of their sera within a period of twenty-four hours after
injection, probably by an influence upon the hemolytic comple-
ment; the administration of a smaller dose (0.6 gram) to syphilitic
COMPLEMENT AND ANTIBODY PRODUCTION 315
persons as part of the treatment of their infection, was found
generally to produce a depression in the hemolytic activity of the
serum as tested one hour after injection followed by a general
increase within eighteen hours.
5. Large doses of arsphenamine and mercuric chlorid to rabbits
tended to limit agglutinin production for typhoid bacilli ; single
doses of arsphenamine (0.6 gram) in adults did not influence the
amount of normal typhoid agglutinin in their sera.
6. The general result of these experiments indicates that while
massive doses of arsphenamine and mercuric chlorid tend to
suppress antibody production and cause a decrease in hemolytic
complement, smaller doses tend to increase the production of
antibody (agglutinins) and augment the complement content after
a primary decrease. It is probable that both drugs administered
in the treatment of syphilis, owe part of their therapeutic efficacy
to their influence upon increasing antibody production and com-
plement.*
REFERENCES
(1) Enrutcu, P., anp Suiaa, K: Farben therapeutische Versuche bei Trypano
somenkrankung. Berl. klin. Wochn., 1904, 41: 329, 362.
(2) Enruicu, P.: Chemotherapeutische Trypanosomen Studien. Berl. klin.
Wochn., 1907, 44: 233, 280, 310, 341.
(3) HaLeerstarpter, W.: Untersuchungen bei expeirmentellen Trypanosomen
erkrankungen. Centralb. f. Bakt., orig., 1915, 38: 525.
(4) Terry, B. T.: Chemo-therapeutic trypanosome studies with special reference
to the immunity following cure. Monograph No. 3, Rockefeller
Institute.
5) Watters, W. H.: Homeopathy and immunity. North Amer. Journ. Home-
op., 1909, 24: 460-472.
(6) MELLO, R. R.: The effect of baptisia in the production of typhoid agglu-
te «© tinins. Med. Century, 1913, 20: 261-166.
(7) WurEetmr, C. E.: Recent experiments in the field of BOMmeC Hay. Brit.
Homeopath. Jour., 1914, 4, 243.
(8) WessELHorrt, C.: Sireioe a in regard to the action of quinine on the malarial
plasmodia. New England Med. Gaz., 1913, 48, 64; 637.
(9) Hooxmr, S. B.: The relation of drugs to immunity. New England Med.
Gaz., 1914, August.
3 An exception to this general statement is the influence of both drugs upon the
syphilis antibody or reagin concerned in the Wassermann reaction which tends to
decrease during active treatment; the reagin however, may not belong to the
category of antibodies.
.
316 IKUZO TOYAMA AND JOHN A. KOLMER
(10) Arkin, A.: The influence of strychnine, caffein, chloral, antipyrin, choles-
sterol and lactic acid in phagocytosis. Jour. Infect. Dis., 1913, 13,
408-424. .
(11) AcGazzi, B.: Ueber den Hinfluss einiger Arsenpraparate auf die Intensitat
der Bildung von bakteriellen Antikorpern (agglutininen) beim Kan-
inchen. Ztsch. f. Immunitatsf., orig., 1909, 1, 736-740.
(12) FRIEDBERGER, E., anp Masupa, V.: Ueber den Einfluss des Salvarsans
auf die Intensitiit der Antikérperbildung beim Kaninchen. Therap.
Monatschr., 1911, 25, 288-291.
(13) Boruncxe, K. E.: Die Beeinflussing der Intensitit der Immunkérperbildung
durch das Salvarsan. Ztsch. f. Chemotherapie, 1912, orig. 1, 136-155.
(14) Weisspacu, W.: Zur theorie der Salvarsani virkung. Ztsch. f. Immunititsf.,
1914, orig, 21, 187-192.
(15) Rerrer, H.: Beeinflusst das Salvarsan die Intensitiit der Antikorperbildung.
Ztsch. f. Immunititsf., orig., 1912, 15, 116-144.
(16) Wurtz, B., anp Dupvort: Effets des injections de bicarbonate de soude sur
la teneur en alexine du milieu sanguin. Compt. rend. Soe. Biol., 1913,
74, 802-803.
(17) Fenyvessy, B., anp Freunp, J.: Ueber kiinstliche Beeinflussung und Mes-
sung der Komple rentwirkung im lebenden Tiere. Ztsch. f. Immuni-
titsf , 1913, orig , 13, 666-681.
(18) Cruca M.: L’action de quelque substances médica nenteuses sur le pouvoir
alexique du serum. Bull. d.1. soe Path. Exot, 1914, 7, 626-632.
(19) Koumer, J. A., AND Winurams, W. W.: Concerning natural hemolysins in
rabbit serum Jour. Infect. Dis , 1913, 12, 96-102.
PROCEEDINGS OF THE AMERICAN ASSOCIATION OF
IMMUNOLOGISTS
FirraH ANNUAL MEETING, HELD at THE New Mepicau LApor-
ATORIES AND IN THE HYGIENE LABORATORY OF THE UNI-
VERSITY OF PENNSYLVANIA, PHILADELPHIA
March 29-30, 1918
The President, Dr. John A. Kolmer, in the Chair.
1. Report oF THE CouNcIL. EXECUTIVE SESSION.
2. THE ROLE oF IMMUNITY IN THE ConpuUCT OF THE PRESENT WAR
John A. Kolmer (President’s address, see this volume, page 371).
3. EXPERIMENTS ON THE PRODUCTION OF ANTIPOLIOMYELITIC SERUM
IN RABBITS
Edgar H. Tsen (see this volume, page 213).
4. ActivE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS
Harry L. Abramson (see this volume, page 437).
DISCUSSION
Arthur F. Coca: It is important to emphasize the concluding remarks
of Dr. Abramson. The conditions under which Dr. Abramson was
working to demonstrate immunity were highly artificial and the im-
munity attained was possibly more than was necessary to constitute
an absolute resistance to the natural infection as occurring in human
beings. In the artificial experiments the mechanism of resistance at
the natural atria of infection was circumvented. It would have been
hard to foresee, from the guinea-pig experiments, how little antitoxin is
necessary in order effectively to prevent the natural infection with
diphtheria in human beings.
Wiliam H. Park: I wish to say a word in line with Dr. Coca’s re-
marks. In test animals, one species might react, and one might not
react. In regard to protective sera: The rabbit gives very feeble
protective serum; the monkey gives better serum, but the horse gives
as good, or even a better serum than the monkey. Perhaps Dr.
Abramson will say a word on that question. In my own experiments
317
318 PROCEEDINGS
I first used virus from the brain, digested by trypsin, as suggested by
Dr. Neustadter and then later untreated virus. Different species of
animals show different periods of development of antibodies and also
a different degree.
John A. Kolmer: Has the serum been used in the treatment of human
infections?
Harry L. Abramson: 1 wish to add a word about rabbit poliomyelitis.
I spent about six months inoculating rabbits with material from human
cases and experimental poliomyelitis with rather discouraging results.
Only a few of a large number of young animals inoculated intracere-
brally, perineurally and intraperitoneally exhibited flaccid paralyses.
The presence of paralysis was considered the only reliable evidence of a
possible poliomyelitic infection. The pathologic findings in the cords
of these animals were not like that of human poliomyelitis.
The sera of these animals were not tested for the presence of immune
substances. However, an attempt was made to determine whether
injections into monkeys of emulsions of cords from rabbits that exhibited
flaccid paralyses would prevent the experimental disease. In one in-
stance an animal so treated was paralyzed after an incubation of four-
teen days whereas the control animal was paralyzed in six days. It is
problematical whether this prolongation of the incubation period in the
treated animal was due to any degree of immunity conferred by the
injections of rabbit’s eord or whether it was simply an instance of
variability in incubation period. The test virus used was that of the
1916 epidemic, which is quite variable.
As regards an anti-poliomyelitic horse serum, two animals have
been subjected to injections of polio-virus emulsions over a period of
three to six months. Both have neutralized in vitro a 5 per cent emul-
sion of the highly virulent Rockefeller strain in the proportion of one
part of serum to one part of virus.
I have had no personal experience in the use of this anti-poliomyelitic
horse serum in the treatment of the human disease. Dr. Neustadter
of New York treated three cases, in all of which there were rather severe
reactions. It is very difficult to draw conclusions as to the value of
serum therapy in poliomyelitis because of the great difficulty in prognosis
in this disease. Large numbers of unselected cases must be treated and
the mortality figures compared with that of an equal number of un-
selected cases in the same outbreak before one can make any positive
statements on this matter.
Rosenow’s anti-poliomyelitic horse serum was subjected to the im
vitro neutralization test and was found to neutralize the Rockefeller
virus in proportion of one part of serum to one part of virus.
5. Types or Mernrtncococct CoNCERNED IN EPIDEMIC INFECTIONS
A Parker Hitchens and C. H. Robinson: The troops in this country
are now experiencing the same conditions that the British troops
experienced at the beginning of the war. The work of English investi-
PROCEEDINGS 319
_ gators is here of value, especially that of Gordon on the meningococci.
Gordon typed the different strains of these organisms as has been done
with the pneumococci, into four different groups. This work has been
followed in the present experiments, and 100 cultures, received from the
first of the year, have been typed according to the Gordon method.
The strains have been received from several concentration camps in
this country. The cultures, from spinal fluids or from naso-pharyngeal
swabs—were Gram negative cocci. They corresponded to a greater or
less degree to Gordon’s types; namely, I, parameningococcus; Ue
normal meningococcus: III and IV, C omparatively rarely found in this
country. The grouping of the strains varied with different camps.
Camp Jackson had 1 strain of type I; 9 strains of type II; 1 strain of
type III. Great Lakes camp showed: type I, 1 strain; ‘type II, 28
strains; 2 Flavus; 2 undermined. It was shown that each camp had a
predominating type, but whether this was a true epidemic was not
determined. The work was undertaken with the hope that specific
sera might be evolved to counteract the types. Gordon has had
excellent results with his specific strains. He found that if specifie
serum were given before the third day of the disease the mortality was
only 9 per cent: if on the seventh day the mortality was 50 per cent.
In testing the types the precipitin test has given valuable results,
which were confirmed by cultural methods. Agglutination of strains
seems to give distinct differences which follow the types as laid down by
specific sera, but what the relationship is, is not yet conclusive. A rapid
method of typing the meninogococci was sought, so as to expedite the
application of specific therapy.
*
6. THe INFLUENCE oF NorMAL HuMAN AND GUINEA-PIG’S Sut
(COMPLEMENT) ALONE AND IN COMBINATION WITH ANTIMENINGITIS
SERA OPON VIRULENT MENINGOCOCCI.
F
John A. Kolmer, Toitsu Matsunami and Ikuzo Toyama (see this
volume, page 157).
7. THE RELATION OF THE MEnNINcococcwaL Acrrvenat OF THE Boop
TO RESISTANCE TO VIRULENT MeniNGococer
Toitsu Matsunami and John A. Kolmer (see this, volume, page 201).
DISCUSSION.
William H. Park: I would like to ask\a question, as to. the use of
“Specific therapy” in the treatment of |cerebro-spinal meningitis
Could one get a monovalent serum for.one type which would be appreci-
ably stronger than the polyvalent serum: for the same. type?,, Lf,one or
two types could be injected into the herse at. once and a nearly equally
good antiserum produced for each, the advantage, would, be much greater
than if one used a different horse. for.each|type. ,. Is there evidence that
| 320 PROCEEDINGS
a polyvalent serum had been produced? If not, it would be advantage-
ous to produce a serum for one or two types. In the pneumococcus
serum it is advisable to make antisera against distinct types, because
the microdrganisms are not closely related; but in this case a polyvalent
serum against two or more types would seem to be equally potent.
Dr. Hitchens: Dr. Park’s point is very well taken and I agree up
to a certain point. There might, however, be a greater possibility of
reducing the death rate through the use of specific sera. It is possible
that there may be immunological variations of meningococci within
the serological types. It is well known that with the best baianced
sera obtainable one finds cases of meningitis to resist their action. The
meningococci isolated from spinal fluids in these cases are resistant in
culture to the action of the polyvalent serum, although belonging to
one of the well known fixed types. These may be serum-fast strains or
not agglutinable strains. The Rockefeller Institute agglutination work
would show that meningococci can not be accurately and sharply
grouped within certain types, but follow the order of a chromatic
scale. There are an almost innnumerable number of variations, al-
though the variations are slight. The problem of getting a polyvalent
serum is therefore difficult. It is felt, however, that work should fol-
low the lines of making a polyvalent type serum and find out whether
this would not confer more resistance in the refractory cases. It will
be found necessary for this purpose to use many strains of meningo-
cocci. One can not select single strains of the four types and immu-
nize a horse to these four strains because there are an infinite number
of variants within the types. In one case where the spinal fluid gave
a specific reaction with the precipitin test, the patient did not do well
with the polyvalent serum. With a specific serum the case recovered.
In regard to Dr. Matsunami’s paper, those who are interested in
active immunization against cerebro-spinal fever will welcome the
technic of Dr. Heist. If even a low grade of immunity can be demon-
strated, this technic will be of considerable assistance.
William H. Park: Dr. Hitchens has not exactly answered my ques-
tion. Can not one horse be injected with two or three types, as well
as with one type alone? In New York we are using the combined serum,
and we have obtained a reduction of mortality to 22 per cent. It will
not do to take the time to type individual strains asin pneumonia. Has
it been proved that one can not produce a serum from a single horse
that would answer to several strains? If not, there must be a horse for
each strain. This would be very confusing from the laboratory stand-
point.
H. Parker Hitchens: I thought I had replied to Dr. Park. If the
number of antigens could be limited to four or five there would be no
reason to think of specific sera.. It is not possible to produce sera
equivalent to each four strains, as high as for a single strain. In an
antiserum for all strains it may be as high as against a single strain.
lf there were only five variants it would not be necessary to think of
specific sera; but, if, as has been suggested, there isa chromatic scale of
PROCEEDINGS aol
-variation—forty to sixty different varieties, one can not obtain sera as
high for individual variants, as if they were split into groups of four or
five. There is, however, as yet, no positive evidence in support of this
theory.
Charles Krumwiede: I have been working on the precipitin test for the
diagnosis of type. From my experience I anticipated considerable cross
reaction between groups one and three and between groups two and four.
Before progress can be made, an easier way of differentiation eliminat-
ing the necessity of absorption to obtain an accurate determination of
type will have to be found. However, the test might differentiate, with
some degree of accuracy, groups one and three from groups two and four.
Nevertheless, there is danger of a misleading reaction due to group
antibodies even though the strain does not belong to any of the four
groups. I have found, as has Dr. Kolmer, that horse serum is rich in
normal opsonins and it is difficult for this reason to standadize ther-
apeutic sera by determining the opsonic content. Attempts to dif-
ferentiate between “normal” and “immune”’ opsonins by heating and
reactivation have failed.
8. THe NATURE OF ANTIANAPHYLAXIS
J. Bronfenbrenner and M. J. Schlesinger: The convulsion in eclampsia
and epilepsy as well as the characteristic symptoms of asthma, hay
fever and certain dermatoses have been interpreted by a number of
investigators as anaphylactic phenomena. Our earlier experimental
studies of anaphylaxis (1) led us to the conclusion that the latter are
due to the liberation of proteolytic enzymes in the blood of the individ-
uals and experimental animals affected. We found that the activation
of these proteolytic enzymes is controlled by the mechanism based
upon the balance of the proteolytic ferments and antiferments in the
blood (2). The actual measurements of antitryptic index in the
individuals during eclamptic and epileptic seizures, confirmed our
view that a decrease of antitryptic powers of the blood leads to pro-
teolysis 7n vivo, and vice versa, that recovery from anaphylaxis is ac-
companied by a marked rise of antitryptic power of the blood (3).
Since it was shown by several investigations that sensitized animals can
be protected against anaphylaxis by special treatment, we started
out to study the mechanism of this protection with the view of finding,
if possible, a theoretical basis for the methods of prevention of clinical
anaphylaxis in man. It is an established fact that the injection of a
sublethal dose of antigen preliminary to the test injection prevents shock
in sensitized animals. The explanation of this phenomenon offered by
Friedberger was that the sublethal dose of antigen exhausts the serum
of its specific antibody by combining with it, and thus prevents the
formation of anaphylatoxin when a subsequent test injection is made.
There are, however, some phenomena in antianaphylaxis which cannot
be adequately explained on the basis of exhaustion of antibody. Thus,
for instance, it was noticed that already a few days after the injection
a22 PROCEEDINGS
of a sublethal dose of antigen into hypersensitive animal, its hypersen-
sitiveness partly returns and that, in general, the length of the anti-
anaphylactic state depends upon the size of the vaccinating injection.
Moreover, the experiments of Anderson and Frost suggested that anti-
body is present in the blood of antianaphylactic animals long before
they return to the state of hypersensitiveness. Our own experiments
conducted with the view of determining whether the exhaustion of anti-
body was the underlying mechanism in the experiments of Friedberger
led us to the conclusion that such was not the case. We found that
animals sensitized simultaneously against two proteins and receiving a
vaccinating injection of one of them become resistant to the test in-
jection of the second protein, though there could be no question of
exhaustion of the second antibody. We found further that guinea-
pigs receiving a large sublethal dose of any anaphylatoxin prepared
in vitro become resistant to a subsequent test injection of several lethal
doses of the same or any other anaphylatoxin. Moreover, direct
experiments 7m vitro by means of the Abderhalden reaction show
definitely that the antibody is not exhausted from the blood of sensi-
tized animals receiving a sublethal dose of antigen. That the state
of antianaphylaxis in experimental animals is not due to the changes in
antibody concentration in their blood is also suggested by the fact that
animals can be rendered resistant to anaphylactic shock by a number of
nonspecific methods. ~Our analysis of the mechanism of the anti-
anaphylaxis produced in experimental animals by a number of such
nonspecific methods brought out the following conclusions. Intro-
duction of certain poisonous substances, which may cause destructive
changes in the tissues, in quantities not sufficiently large to kill the
animal outright, is followed by the death of tissues immediately
affected by poison. In this process the intracellular ferments are
set free, and together with the ferments thrown out from the sur-
rounding fixed cells as well as from the blood serum and leucocytes
digest the dead material. The split products of such digestion exert
antagonistic action, and retard or stop further activity of the proteolytie
ferments. The time of appearance, the rate of their increase and the
length of time during which these split products remain in the blood,
determine the antitryptic titer of such blood. We have tested with this
point in view the effects of ether, chloroform, alcohol, choral, morphine
and scopolamine, and we found that the power of these substances
to protect the animals against anaphylaxis is strikingly parallel with
their power to increase the antitryptic titer in the blood of these animals.
We found that the animals were protected against anaphylaxis only so
long as the antitryptic titer resulting, from the treatment remained
above the normal. The same relation between the power of substances
administered to protect against anaphylaxis and to increase the anti-
tryptic titer of the blood of treated animals was found also in the cases
of treatment with BaCle, CaCl: as well as with lecithin and cholesterin.
It was suggested by several investigators that starvation or exposure
to low temperature may also protect the sensitized animals against
PROCEEDINGS 323
anaphylaxis. Chemical studies by various investigators suggest very
strongly that there is a certain amount of analogy between the changes
in metabolism of animals during starvation and anaphylaxis. We
found accordingly that the antitryptic index of the blood or starving
guinea-pigs is above normal. As for the protective influence of low
temperature, our measurements have shown that it 1s not connected
with the antitryptic balance, but is merely due to the fact that the
whole process of anaphylaxis becomes slow under the influence of cold
and the symptoms become, accordingly, less acute.
9. StuDIES ON So-CALLED CELLULAR ANAPHYLAXIS
W. P. Larson and E. T. Bell.
DISCUSSION
John F. Anderson: I wish to ask Dr. Bronfenbrenner whether my
impression is correct that if the animal were sensitized to more than
one protein there would be an inhibitory effect from the injection of a
sublethal dose.
J. Bronfenbrenner: In the experiments with animals sensitized to
two antigens in which one of the antigens is injected in order to vaccin-
ate against the effect of subsequent injection of the second antigen, one
has to be very careful properly to select the vaccinating dose. If the
amount of antigen injected is but a small fraction of the lethal dose, it
may not protect sufficiently, but if the amount inoculated is sufficiently
close to the minimum lethal dose, so that 1t may even cause slight ana-
phylaxis, the animal is temporarily protected against at least one and a
half or two lethal doses of the second antigen.
Arthur F. Coca: Dr. Larson’s results contradict those of some pre-
vious work of my own; that is, they seem to show that my conclusions,
as far as they pertain to the percentage of residual blood after per-
fusion, were wrong. This, however, does not invalidate my chief con-
clusion, which was that the site of the anaphylactic reaction is in the
tissue cells, because in the same paper evidence is presented proving
that if as little as 50 per cent of the actively sensitized guinea-pig’s
blood is removed by perfusion the residual 50 per cent will not, in half
of the animals, contain an amount of the sensitizing antibodies sufficient
to sensitize a single guinea-pig. The experiments in perfusion of pas-
sively sensitized guinea-pigs were still more conclusive inasmuch as
Weil had shown that the blood of such animals contains no demonstra-
ble antibodies forty-eight hours after the sensitizing injection.
Dr. Bronfenbrenner’s experiments illustrate what has been called
nonspecific antianaphylaxis. For the purpose of Dr. Bronfenbrenner’s
argument it is necessary to show that the desensitization observed is
specific; otherwise the experiments can throw no light on the nature of
specific “antianaphylaxis.”” In general it should be borne in mind
that any humoral theory of anaphylaxis is at a disadvantage in not
being able to explain the latent period in passive sensitization.
324 PROCEEDINGS
G. H. A. Clowes: I can substantiate Dr. Anderson’s position. Six
years ago he had tried desensitization against hay fever in patients
sensitive to spring and autumn fever. Timothy and ragweed were
used and it was found that the skin test was strongly specific. Patients
desensitized to timothy were not altered in their reaction to ragweed.
Immunity developed only against the agent employed. The matter of
antitryptic reaction is related to the matter of surface tension. The
soaps, lecithins, calcium chloride and barium chloride play a part in
this question. The nonspecific interference is a matter of surface tension
I myself have obtained a marked desensitization to hay fever with a
marked antitryptic index. This was followed up with care on ac-
count of its relation to the cancer question. The high antitryptic
index was coincident with a maximum desenaitisation following a slight
anaphylactic shock which fell appreciably a few months after the hay
fever period. Anaphylactic effects appear to be due to increased per-
meability of the protoplasn ic film.
J. Bronfenbrenner: Of course the the time which is allowed to elapse
between the vaccinating and the test injections in my experiments is a
very essential element. The test injection must follow closely enough,
so that the antitryptie index of the vaccinated animal is still above the
normal at the time of the test injection. Another equally important
factor is the method of injection. If the vaccinating injection is given
int-aperitoneally or subeutaneously—the results can never be as sharp
as in the case of intravenous inoculations.
Dr. Coea is quite right in making a distinction between specific and
nonspecific phenomena in antianaphylaxis, but in my experiments
I did not intend to study this question. My problem was to see whether
the mechanism of antianaphylaxis in the experiments of Friedeberger
was that of exhaustion of antibody, and if not whether the mechanism
in this case as well as in other instances of antianaphylaxis is the
same. I suspected that there must be only one essential process (or
set of processes) underlying all the phenomena of antianaphylaxis, be-
cause it was found empirically that the specific anaphylaxis can be
regularly checked by such nonspecific treatment as the administration
of ether or the injection of BaCl. While the experiments of Bron-
fenbrenner do not prove that the entire mechanism of the phenomena
ean be reduced to the question of the control of ferment action in the
blood, they show, nevertheless, conclusively that in all the cases of anti-
anaphylaxis studied by him, the high antitryptic index of the blood was
a part of the symptom-complex of antianaphylaxis in guinea-pigs, and
vice versa; whenever the antitryptic index of the blood in the guinea-
pigs pa raised (by whatever method) the animals were refractory to
shock.
Dr. Larson’s experiments seem to me exceedingly valuable. It was
very interesting to see so clearly demonstrated that the perfusion of
organs is very irregular. Dr. Larson himself has drawn conclusions as
to the importance of his observation in relation to the question of cel-
lular anaphylaxis as contrasted with humoral anaphylaxis. There is _
PROCEEDINGS 325
very little doubt but that the cells of the body take part in what we
observe as final symptoms of anaphylactic reaction. What is the
exact part played by the cells is not yet well understood, but it seems
that in all of the experiments in which the behavior of the tissue was
studied during the anaphylactic shock the presence of the serum in
such tissues was not definitely excluded. While such was the feeling
of several investigators all along, Dr. Larson’s experiments for the
first time demonstrated that such was the case. I would like to ask
Dr. Larson whether he can offer any suggestions as to the reason why
the fluid with which the organ is perfused does not reach certain parts
of the tissues.
10. EXPERIMENTAL POLLINOSIS IN GUINEA-PIGS
Henry L. Ulrich.
DISCUSSION
G. H. A. Clowes: I do not feel competent to discuss Dr. Ulrich’s
paper as a whole, but I am interested in the results. In experiments
made with human material, numerous attempts have been made with
a variety of means to induce sensitization in individuals that were not
previously sensitive, but without result. This might give a clue to the
work upon this mysterious condition.
John F, Anderson: A good many years ago it was demonstrated that
animals could be sensitized by instillations in the nose so as to respond
with symptoms of anaphylaxis. Has Dr. Ulrich confused the sen-
sitization to protein with the sensitization to pollen. These are two
entirely different propositions and opened a field for large experimenta-
tion.
Arthur F. Coca: No doubt we all recognize this work of Dr. Ulrich
as revolutionary. Despite the innumerable injection into human
beings of various kinds of foreign protein no instance of the experi-
mental production of a condition corresponding to hay fever has yet
been reported. In a series of experiments with Dr. Cooke and Dr.
Flood I was unable to produce the usual condition of experimental
anaphylaxis in guinea-pigs with strong extracts of poilen.
A. Parker Hitchens: I would like to ask Dr. Ulrich: what pollen was
used and what method of extraction? All the pollen I have used was
very badly contaminated with bacteria of different kinds. The bac-
terial contamination might play a large part in the phenomena which
resulted.
G. H. A. Clowes: I found contamination by bacteria and minute
insects. Reactions to these might be introduced. I never succeeded
in getting reactions in normal individuals.
Geo. H. Smith: We instilled pollen into the nose and mucous mem-
brane of normal animals and also we tried to injure the mucous
membranes to see whether they would respond more after injury.
The animals were shut up in glass cages in which pollen was kept cir-
culating by means of the electric fan. No sensitization was obtained.
326 PROCEEDINGS
Henry L. Ulrich: The pollen was washed in acetone and the sterile
pollen was injected into the peritoneal cavity. It never showed any
growth. It was used dry in powder form.
11. A Skin ReaActTION TO PNEUMOTOXIN
Charles Wess: This study has been instituted with the idea of using
the endocellular toxin of the pneumococcus (hemotoxin) rather than
the bacterial emulsion. The protein free pneumococci were dissolved
in sodium cholate. A specific reaction was obtainable in guinea-pigs
and also in persons suffering from lobar pneumonia. The guinea-pigs
were sensitized with a sublethal dose of pneumotoxin. The animals
survived the injection and there was a true pneumotoxin reaction, which
was different from the reaction to pneumococcus protein or autolysates
of it. A vaccine of dead pneumococci was tried and three animals
reacted to it. It was found that heating pneumococci at 56°C. for one-
half hour was not sufficient to destroy the endocellular pneumotoxin.
The skin test would appear to be a true test for the presence of pneu-
motoxin in the body of the animal. By various chemical tests it is
found that the pneumotoxin is a true protein, and in human cases with
lobar pneumonia there was a positive reaction demonstrable before the
crisis and a negative reaction after the crisis. In children the reaction
was most characteristic. Other persons tested, suffering with tuber-
culosis, appendicitis, skin broncho-pneumonia or with acute or chronic
infections not of pneumococcic origin, as well as healthy adults and chil-
dren did not react.
The reaction is regarded as similar to the tuberculin reaction and is
indicative of a state of allergy to pneumotoxin. Sensitization to the
toxin presumably takes place with its liberation (by the action of nor-
mal body enzymes upon pneumococci normally localized in the lung
alveoli) at the time of the prolonged chilling due to exposure. Failure
to elicit the reaction during convalescence indicates the establishment
of a temporary immunity or the disappearance of excess of toxin. This
skin test does not seem to be of value as a method of serological type
diagnosis but may aid in differential diagnosis between appendicitis or
tuberculosis and pneumonia (especially in children). It is also of inter-
est because of its bearing on the mechanism of the crisis.
12. THE INFLUENCE OF ARSENOBENZOL AND MERCURY UPON ANTIBODY
PRODUCTION
I[kuzo Toyama and John A. Kolmer: The possibility of certain drugs
acting as antigens have been the theme of several studies by different
workers. The antigenic action of drugs may account for acquired drug
tolerance and also aid in resistance to infection by stimulating antibody
production against microparasite apart from direct action of the drug
on the parasite. The latter phase of the subject is the one particularly
dealt with by the authors. Considerable evidence would point to the
PROCEEDINGS 327
conclusion that many drugs exert a stimulating action on antibody
production by the tissues. Such drugs as arsenous anhydrid, phosphoric
acid and mercuric chlorid, administered by mouth, are found to act in
this manner. Salvarsan appears to stimulate agglutinin: production,
according to some workers. The present experiments were conducted
on rabbits, to determine whether small daily doses of arsenobenzol and
mercuric chlorid tend to increase antibody response to alien erythrocytes
or to typhoid bacilli. After a series of experiments, five in all, on a
large number of rabbits, and control animals, it was found that no
increase of antibody production was shown, after injections of arseno-
benzol or mercuric chlorid. On the contrary, it would seem that there
is a lowering of antibody production, probably due to lessening of re-
sistance by toxic effects. Further experiments, however, are in prog-
ress, upon the action of these drugs in experimental trypanosome in-
fections. It is felt that such work should be done upon human sera, as
work upon the sera of lower animals may not be a true index in human
eases. The importance of the subject demands careful experimental
work in this direction.
JOINT SESSION WITH THE AMERICAN ASSOCIATION OF
PATHOLOGISTS AND BACTERIOLOGISTS
1. A CoNTRIBUTION TO THE BACTERIOLOGY OF B. FUSIFORMIS; ITS
. MorpHoLocic PHASES AND THEIR SIGNIFICANCE
Ralph R. Mellon.
2. THE VARIOUS IMMUNOLOGICAL REACTIONS IN GLANDERS
G. Benjamin White: During an epidemic of glanders in a herd of ninety-
five horses an excellent opportunity was afforded for observations on
the results of various immunological tests. The following observa-
tions may be reported.
In all cases where glanders lesions were found at autopsies the horses
had given positive subcutaneous mallein tests whereas the eye test
and the complement fixation and agglutination tests were in some
cases positive and in some cases negative. Horses giving positive eye
tests always gave a positive subcutaneous test but were either positive
or negative by complement fixation or agglutination. Complement
fixation and agglutination tests were sometimes positive with a negative
ophthalmic, negative subcutaneous and no lesions at autopsy.
It was found that the subcutaneous injection of mallein never sen-
sitizes a horse to the extent that it will react positively to a second
subcutaneous test nor does such an injection produce sensitization of the
mucous membranes. The eye test, therefore, in such cases cannot be
changed from negative to positive with the subcutaneous injection of
mallein even when the injection is repeated several times. Such in-
jections, however, do cause complement fixation and agglutination to
328 PROCEEDINGS
change from negative to positive. The injection of dead mallei bacilli —
produces no general or eye sensitiveness but does change a negative
complement fixation and agglutination to positive. It is felt, there-
fore, that a positive subcutaneous mallein test shows the presence of a
glanders lesion. The lesion, however, may be a healed, or inactive, one.
The subcutaneous injection of mallein renders glandered animals
insensitive to subsequent injections for a period of about four weeks.
No horses were found refractory when tested five weeks after the first
injection.
The similarity between immunity reactions in glanders and those in
tuberculosis is pointed out.
DISCUSSION
William H. Park: There is no absolute relation between the comple-
ment fixation, agglutination and mallein tests in horses. Very few New
York horses would pass all three tests. This has been a serious difficulty
in determining what horses should be used for food. It has been found
necessary to require both serum reactions to be positive or a definite
mallein reaction, before reporting the horse as infected. This has been
done and it has been found to admit sufficient horse meat to alleviate
the food shortage.
A. Parker Hitchens: It is my feeling that many more horses react to
the subcutaneous test than those having glanders. I apply the sub-
cutaneous test, as well as eye test and complement fixation, and I
depend upon complement fixation. I have found that we must accept
some horses that give positive subcutaneous tests. We have had no
trouble with those animals. Reading of the subcutaneous test requirés
considerable skill; the symptoms vary. Just what is a positive sub-
cutaneous reaction and what is not, is open to question and depends
largely upon the individual making the examination.
G. Benjamin White: I quite agree with Dr. Hitchens that a horse
may give a positive subcutaneous test and yet have no active glanders.
A healed lesion may be present and it has been my experience that in
such a case the horse develops no active glanders unless subjected to
some particularly severe strain.
3. PERSISTENCE OF ACTIVE IMMUNITY IN THOSE IMMUNIZED AGAINST
DIPHTHERIA
William H. Park: In testing different species of animals, some have
been found to have no immunity; some are entirely immune and again
there is a group in which some are immune while others are not. Guinea-
pigs have no natural immunity whereas horses almost always possess
antitoxic immunity. The guinea-pig requires from four to eight weeks
to develop immunity after toxin-antitoxin injections while the horse
can be immunized very rapidly and nearly always produced consider-
able antitoxin. The guinea-pig loses its immunity in about nine months;
PROCEEDINGS 329
in the horse at the end of twelve months the developed immunity drops
to the original level, i.e., 45 to } unit. Human beings differ from both
horses and guinea-pigs in that some are immune while others are not.
Dr. Zingher has made interesting studies in immunizing groups of
babies. One group was of immune babies with immune mothers;
these remained immune after the period of passive immunity from the
mother had passed. Another group was from mothers that were not
immune; these became immune after the injections.
Dr. Zingher tested children in institutions and found that fourteen
months after immunization, those that had received but one injection
of antitoxin were as immune as those that had had received two or three
injections. The development of the immunity, however, was slower.
Results observed by Dr. Rosenberg are very encouraging; the very
great majority of children remained immune. It seems as though the
acquired immunity to diphtheria will prove to be as permanent as
natural immunity. Of 404 immunized children tested, 396 remained
immune. Two of these had never become immune.
From these results it seems practical to immunize infants; this will
give protection at the period of life when there is the greatest danger.
DISCUSSION
Alfred F. Hess: The use of toxin-antitoxin mixtures has made a
great difference in the Hebrew infant asylum in New York City. For-
merly there were about 15 cases of diphtheria a year in the institution
with three or four deaths, but since the use of toxin-antitoxin was
introduced there has been no diphtheria. Of 150 cases injected and
followed for from one to two years, 98.75 per cent have remained im-
mune. As regards the best time for immunization, early injection
should be made because, while 80 per cent of infants have derived a
passive immunity from the mother and are thus immune during the
first few months of life, yet this immunity is for the most part lost in the
later months of the first year. Immunization with toxin-antitoxin
injections will, I believe, entirely do away with diphtheria in institutions.
John A. Kolmer: What is the nature of the reaction following the
injection of toxin-antitoxin? Is a single dose advocated or is it wise to
give a second one?
William H. Park: Children receiving only one injection develop
immunity more slowly than those receiving two or three injections;
90 per cent of the former become immune by the end of six months
whereas an equal percentage of the latter are immune by the end of the
third month. For immediate protection antitoxin alone must be given.
The toxin-antitoxin method is being used in the Navy as diphtheria
is not always recognized at once and furthermore if the seamen are
immunized one does not need to worry about carriers. In a few cases
the injections cause a slight toxic effect. If the inert protein in the
mixture can be controlled the procedure will become entirely harmless.
John F. Anderson: Is there a quick deterioration in the mixture?
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 4
330 PROCEEDINGS
William H. Park: So-called “‘standard”’ toxin that has become rel-
atively stable through long standing, is used for the toxin-antitoxin
mixtures; the mixtures can, therefore, be kept for months without
appreciable change.
4, A SmmpLt—E Meruop For BLoop CULTURES
John G. Wurts and S. W. Sappington.
5. A BACTERIOLOGICAL StuDY OF PostT-OPERATIVE PNEUMONIA
Miriam P. Olmstead: One hundred and thirty cases of post-operative
pneumonia have been studied and a pathogenic organism has been
recovered from at least thirty-one, a percentage of 23.8. Two cases
had a Pneumococcus I in the blood stream, one of these Pneumococcus I
in pre- and post-operative sputum. One had a Pneumococcus II in
the sputum, pre- and post-operative, and the urine was precipitated
by Pneumococcus II serum. In two cases with an atypical II in pre-
and post-operative sputum, the etiology was established, in one by a
positive agglutination test of the patient’s serum, and in the other by a
urine precipitin reaction with Type II serum. Pneumococcus III
was established as the inciting factor in five cases; one had a positive
blood culture, from one the organism was obtained by lung cultures at
autopsy, the blood of two agglutinated the Pneumococcus III re-
covered from the post-operative sputum, the urine of these two and one
other, who had Pneumococcus III in the pre-operative sputum speci-_
men, gave a precipitin reaction with Pneumococcus III serum In
eighteen cases a Pneumococcus IV was found to be the inciting factor.
It was recovered from the blood once and seventeen cases gave a posi-
tive agglutination reaction with strains from the sputum. In one of
these cases the urine precipitated with Pneumococcus IV serum. A
hemolytic streptococcus was found in the blood stream of one case,
and a hemolytic streptococcus in the sputum of another gave a positive
thread reaction with the patient’s serum. A mucoid streptococcus was
found in sputum and chest fluid of one case. Some of these organisms
were undoubtedly the inciting factors of other post-operative pneumonia
cases, but in the absence of a positive blood culture or an immunological
reaction, the significance of their presence in the sputum is uncertain.
The most common inciting factor, as Whipple (Surgery, Gynecology
and Obstetrics, 1918, xxvi, 29) has stated, is a Pneumococcus IV.
It is at least suggestive that a pneumococcus, usually a IV, was recovered
from either a pre-operative or a post-operative sputum specimen in
ninety-seven of the one hundred and thirty cases studied, i.e., 74.6
per cent, while a pneumococcus was recovered from the sputum of only
32.2 per cent of all the surgical cases examined before operation, and of
50 per cent of the cases that subsequently developed post-operative
pneumonia.
PROCEEDINGS 331
DISCUSSION
Alfred F. Hess: Did pneumonia develop among cases where the pneu-
mococcus was absent from the pre-operative sputum; if so, how many
were there? In what percentage of cases was there a discrepancy in
the type of pneumococcus where pneumonia developed?
John A. Kolmer: How many patients, with pneumococci in the blood
stream, died? How many showing a precipitin reaction died? I am
under the impression that persons showing the precipitin reaction have
rather a bad prognosis.
Miriam P. Olmstead: In reply to Dr. Hess I would say that pneu-
monia developed in 44 cases from whose pre-operative sputum a
pneumococcus was not recovered. In 14.9 per cent of the cases studied,
there was discrepancy in the pre- and post-operative sputum findings.
Type IV was present in 8 of these but the strains recovered were found
to belong to different groups. It is probable that in at least some cases
both strains were present in the sputum at the same time, but that,
owing to the similarity of the colonies, only one strain was isolated.
In reply to Dr. Kolmer, I would say that 2 patients with*pneumococci
in the blood stream died; one had a type I, the other type III organism.
There were no fatal cases among those whose urine gave a precipitin
reaction.
Augustus B. Wadsworth: I am greatly interested in this work. The
fact that pneumonia and not a general pneumococcemia without local
lesions developed in man is an indication of some degree of insusceptibil-
ity or immunity against pneumococcus infection. In my experiments
on rabbits pneumonic lesions developed only when the animal was
partially immunized. Forthermore, the prevalence of pneumococcus
infection of the upper respiratory tract suggests that different individuals
acquire varying degrees of immunity from time to time. It is thus
generally assumed that man possesses varying degrees of pneumococcus
immunity, and the demonstration of the specific reactions of pneumococ-
cus immunity in the serum of the normal healthy human subject is of
interest.
6. Ture Active IMMUNIZATION AGAINST PNEUMONIA
R. Kohn.
7. PrRopucTIOoN oF PNEUMOCOCCUS ANTISERUM AND THE CORRE-
SPONDING CURVES OBTAINED BY PROTECTION AND AGGLU-
TINATION TESTS
G. Benjamin White: On account of the unusual demand for anti-
pneumococcus serum every effort had been put forth to speed up and
increase the production of this serum. The problem had three phases;
first, to produce a potent serum; second, to bring the horses to produc-
tion in the shortest possible time; and, third, to carry out the immuniza-
tions with the least detriment to the horses. The basic plan of immuni-
332 PROCEEDINGS
zation originated by Cole and his co-workers was followed. The first
two phases have been studied with success and experiments on the third
phase are now under way.
The best type of horse is the medium heavy animal of the draft type
with considerable spirit.
A single strain of the type I organism appears to be sufficient. It is
doubtful whether it is necessary to have a strain of high virulence.
The stock cultures are carried on blood agar while the cultures used for
injections are grown for sixteen to eighteen hours in meat infusion broth
with a reaction of about plus 0.8. The cultures are centrifuged, the
sediments are not washed but are emulsified in physiological salt solu-
tion and injected into the horses intravenously immediately after prep-
aration. Where killed cultures are to be given, the emulsions are
devitalized by heating for three-quarters of an hour at 56°C.
Various plans of immunization were tried and that suggested by Cole
was found to give excellent results. A slight modification of this
brought a horse to production in the record time of twenty-eight days.
The dosage is regulated by the temperature reaction following the
injection of the day before. When the horses have reached the stage of
production, injections of live pneumococci are given once each week
and the horses are bled six days after every second injection, 7500 ce.
being taken at a time.
No parallelism was found between the agglutinative and the pro-
tective titers and the former is found to be less stable than the latter.
All serum produced is released for distribution on its protective titer
regardless of its agglutinative power.
DISCUSSION ,.
A. Parker Hitchens: I congratulate Dr. White on the results obtained
in the rapid immunization of the horses. This is an especially impor-
tant point at this time. In regard to bleeding, what was the plan of
bleeding as regards the time of injections? What was the interval
between the injections?
John F. Anderson: We have been using a method of six daily injec-
tions, in three courses, with seven-day intervals. The horse can be
bled in thirty-two days. A standard has been made to protect against
0.1 cc., 0.2 ec. and 0.3 cc. of culture.
George W. McCoy: I am glad to hear Dr. White bring out the point of
the unreliability of the agglutinating power as compared with the
protective value. We have had sera sent back marked N.G. because
they did not agglutinate type I pneumococci although they possessed
a high protective power. All commercial sera, before being sent out
are tested by the hygienic laboratory.
Augustus B. Wadsworth: Had the high protective serum any agglu-
tinating power at all?
George W. McCoy: It had no agglutinating power at all.
Augustus B. Wadsworth, Was it controlled?
George W. McCoy: Yes.
PROCEEDINGS 300
_ William H. Park: In the South of France there was a great deal of
pneumonia among the African troops. At one camp the men were
vaccinated by Borrel with cultures from casesof pneumonia in the camp.
In 18 days the pneumonia cases ceased to develop. The men in a
distant camp were given the same vaccine, but the results were nega-
tive. The supposition was that the pneumonias in the second camp
was due to a different type. The cessation of cases may have been a
coincidence, but it is striking that the cessation of the epidemic in the
camp followed the use of the vaccine.
G. Benjamin White: The horses are bled six days after the second
injection.
We have had four interesting instances of horses dying from pneu-
mococcus infection during immunization. Three horses died from lobar
pneumonia due to type I and at the time of death their serum showed
both good protective and agglutinative power. Another horse died of
a pneumococcus (type I) endocarditis with pneumococcus bacteremia
during the course of immunization. The organism isolated from the
blood was found to be serum-fast.
A similar paradox was observed in the case of a horse immunized
against paratyphoid B. Three weeks after the end of the course of
immunization the animal was found to have a bacteremia and from the
blood paratyphoid bacilli, the B type, were isolated. The organism
was found to have acquired no serum-fastness. The serum of this
horse at the time of death agglutinated both the hemotogenous strain
and the stock strain in a dilution of 1: 5,000.
8. A Rapip StmuepLeE METHOD FOR THE DETERMINATION OF TYPE OF
PNEtmMococcus IN SputuM OF LOBAR PNEUMONIA
Charles Krumwiede, Jr.
DISCUSSION
O. W. H. Mitchell: I have gotten the same results by the method of
extracting with sand. I am convinced as to the specificity of the
reaction. The rapidity and simplicity of the method makes it an
excellent one. Our sera are furnished by the N. Y. State laboratory.
The tests have been checked by Dr. Wadsworth and invariably when the
sputum has reacted definitely my report has been corroborated.
Young rats have been found easier to procure than mice for the
pneumonia work. Children keep them as pets and a few inquiries
generally result in getting as many as are necessary. Half grown rats
are preferable to mice. The peritoneal cavity is larger.
9. ON THE INFLUENCE EXERTED BY SALTS ON THE ELECTRICAL
RESISTANCE AND PERMEABILITY OF TISSUFS
B. H. A. Clowes: I have recently demonstrated by electrical conduc-
tivity experiments that tumor tissues are more permeable than normal
334 PROCEEDINGS
tissues in both plants and animals. I have also produced artificial
membranes made by saturating filter paper with emulsions of oil in
soap which exhibited variations in electrical conductivity when exposed
to various antagonistic salts similar to those exhibited by laminaria and
other marine organisms experimented with by Osterhout.
Both laminaria tissue and emulsion membranes when exposed to the
influence of NaCl exhibit a rise in permeability. If subsequently
transferred to CaCk for a short period they exhibit a fall in permeability,
and alternating variations in permeability within certain well defined
limits can be effected in both cases by alternating treatments with
NaCl and CaCh. These experiments are paralleled by experiments on
the surface tension of soap films previously reported. If an aqueous
soap or NaOH solution is allowed to flow from a capillary pipette
through olive oil a given number of drops is obtained. If NaCl is
added to the solution the number of drops is greatly increased.. If
CaCl, is added to the solution the number of drops is diminished but if
NaCl is mixed with CaCk in certain balanced ratios in which they oecur
in the blood, sea-water, etc. the number of drops approximates that
given by the original soap solution. These effects are all attributable
to the influence of electrolytes on the state of dispersion and conse-
quently the permeability of interfacial soap films. A practical demon-
stration as to how the electrolytes in question may control the per-
meability of emulsions is obtained by shaking a suitable emulsion of oil
dispersed in water by means of soap with increasing proportions of
CaCl. The conductivity which serves as an index of permeability
remains approximately constant up to a critical point at which the
emulsion of oil in water is converted into one of water in oil. At this
point the resistance rises to an enormous extent owing to the transforma-
tion from an emulsion which is permeable to water to one which is
impermeable. NaCl, alkalis, etc. exert an effect the reverse of that of
CaCl, promoting the permeability of emulsions and also of tissues.
A further proof that these experiments on conductivity soap films
and emulsions actually afford an index of the permeability to water
and water-borne substances is afforded by introducing layers of a suitably
constituted emulsion into long glass tubes, supported by filter paper and
tightly fitting rubber tubes and passing various solutions through this
emulsion diaphragm and determining the.rate of flow. Distilled water,
sea-water and properly balanced mixtures of NaC] and CaCk flow
through at nearly the same rate of speed. NaCl flows a great deal
faster than the balanced solution and CaCl, considerably slower, the
relative rate of flow corresponding remarkably with the number of
drops obtained in the surface tension experiments.
From the above experiments it appears probable that the mechanism
controlling the permeability of the protoplasm is dependent upon an
extremely delicately balanced emulsion of soaps, lipoids and fats and
that proteins simply afford the mesh or net-work in the capillary spaces
of which the influence exerted by electrolytes on the permeability of
emulsions would be accentuated. Pathological changes are frequently
PROCEEDINGS 335
'
attributable to disturbances in the balance of soap and fats in emulsions.
For example, fatty degeneration appears to be simply the aggregation
of fatty globules under the influence of surface tension changes to a
point at which they become readily visible.
Anaphylaxis and sensitization similarly appear to depend upon
changes in permeability. A point particularly to emphasize is that the
permeability of any given structure is not dependent simply upon the
size of the pores of the filter but upon substances like soaps which lower
the surface tension in the capillary spaces exerting an effect analogous
to that of a lubricant.
DISCUSSION
James Ewing. Yesterday after listening to the philosophical pres-
entation of Dr. Clowes I suggested that it would be well for the speaker
to attempt an interpretation of his observations in terms intelligible to
the pathologist and the immunologist. It has seemed to me that we
were on the verge of seeing a great light, which never dawned. One
wishes to know how physical systems have a direct bearing on patho-
logical conditions such as fatty degeneration. Iam not prepared to admit
that the fatty changes are as simple as the interactions of these systems
of emulsions. This line of work corresponds to that done by Novy.
The phenomena of anaphylaxis will probably be found to fall in line with
all of these observations. I feel that I can congratulate Dr. Clowes
on having made his experiments intelligible—to the pathologist.
Henry Ulrich: I feel that surface tension’ is a vital factor in the
formation of spores. B. subtilis will not grow on media with lowered
surface tension. B. anthracis entirely loses its pathogenicity when
grown on media with lowered surface tension. It will then produce
no symptoms whatever. When blood serum is added to culture media,
the tension is lowered. The tension has to be suited to the growth of
the organism. This work of Dr. Clowes is very suggestive.
J. Bronfenbrenner: In my opinion one can not emphasize too much
the usefulness of physical methods in the study of biologic phenomena.
Already several years ago Ascoli noticed the changes of surface tension
in the mixtures of antibody-containing sera with suitable antigens, and
he proposed to use the measurements of surface tension of such mixtures
for diagnostic purposes. Though his results were not very sharp, there
is no doubt that the application of more recent methods of measurement
of surface tension will corroborate his conclusions. In my own experi-
ence in the study of various questions in immunity, I was amply
convinced of the necessity of applying physico-chemical methods to the
study of these problems. In my study of the Abderhalden reaction, I
noticed that the surface tension of the serum undergoing autodigestion
always decreased, and vice versa, whenever the surface tension of serum
was diminished, it underwent autodigestion. In collaboration with
Dr. Fleisher I also noticed that the refractive index of serum increased
during the process of autodigestion, thus indicating that the dispersion
of the colloidal particles of the serum increases during this process. It
336 PROCEEDINGS
is more than likely that the mechanism of the anaphylactic shock,
or rather the nature of the cellular response to the humoral reaction in
anaphylaxis, may be largely a surface reaction, affecting the permeability
of vital cells, or of their electro-conductivity. There is also a great deal
of evidence that the whole question of ferment action may be bound up
with the question of colloidal dispersion. In general, the phenomena to
which Dr. Clowes has called our attention are of fundamental impor-
tance in the study of various biological problems.
G. H. A. Clowes: I feel immensely interested in the observation
about anthrax spores. That corresponds with my own findings. Any-
one who has studied amoeboid movement, spore formation karyokinesis,
budding, ete., side by side with emulsions will be convinced as to the
importance of surface tension. I have produced from emulsions objects
that looked exactly like leucocytes and I have shown them to pathol-
ogists who thought they were leucocytes. Dr. Bronfenbrenner has
mentioned the myostagmin reaction, which was Ascoli’s work. He
let drops fall into the air instead of into oil. Ferment action appears
to be definitely due to surface tension changes. Perfect contact must
first be effected, then dispersion or aggregation may occur. ;
10. EXPERIMENTS UPon THE CHEMOTHERAPY AND CHEMOSEROTHERAPY
or PNEumMococcus INFECTION
John A. Kolmer, Edward Steinfield and Charles Weiss.
11. SruDIES ON THE ToxIcITy oF PNEUMONIC LUNGS
John A. Kolmer, Charles Weiss and Edward Steinfield.
DISCUSSION
A. Parker Hitchens: Has Dr. Kolmer taken into consideration the
great increase of toxicity in tissues that undergo autolysis? A piece
of liver digested in sterile salt solution for twenty-four hours becomes
extremely toxic.
Ralph R. Mellon: Do these extracts of empyema fluid bear any re-
lation to the aggressins of the disease?
Charles Weiss: It is very difficult to decide whether or not the toxi-
city of the pneumonic exudate is due to the fact that the tissue was
undergoing autolysis. The hemolytic properties of the pneumonic
exudate may not have been specific for pneumonia. They have been
attributed to various fatty acids. Specific anaphylactic results were
obtained by sensitizing guinea-pigs to normal and pneumonic exudates.
The latter are rich in fibrin and the react’on may have been specific
to the fibrin. But the experiments indicate that at least part of the
toxicity and hemolytic activity of the exudate was due to the presence
of toxins liberated from the pneumococci.
PROCEEDINGS Bor
James Ewing: This work is an important step in the right direction.
I do not feel competent to offer any criticism, but I wonder why the work
of immunologists turn always to specific toxins and away from products
of tissue changes which the pathologist is interested in. I believe that
immunologists have been held back by exclusive attention to Ehrlich’s
theories.
G. H. A. Clowes: Were these substances soluble in fats and lipoids
or in water? This is of vital importance in determining théir char-
acteristics.
John A. Kolmer: We have no means of absolutely controlling this
work in relation to aggressins. We tried to ascertain whether this toxic
substance would retard phagocytosis of pneumococci in vitro and we
found that it does so to a slight extent.
Charles Weiss: I have isolated albumin, globulin, uric acid and lipoid
substances from the extracts, but this work is still under way.
H.G. Wells: There is an error in the last remark. Uric acid is formed
only in the liver. Xanthin is probably what the speaker meant.
John A. Kolmer: Mr. Weiss referred to the literature on the subject.
12. THr PROPERTIES OF PNEUMOTOXIN AND ITS PROBABLE ROLE IN THE
PaTHOLOGY oF LOBAR PNEUMONIA
Charles Weiss and John A. Kolmer.
FINAL SESSION
1. THe EXAMINATION OF THE BLOOD PRELIMINARY TO THE OPERATION
OF BLoop TRANSFUSION
Arthur F. Coca (see this volume, page 93).
DISCUSSION
John A. Kolmer: I desire to describe briefly a microscopic method
which I have employed during the past year with very satisfactory
results and which is fashioned after a microscopic technic described by
Lee.
A small amount of blood is obtained from the finger of the patient
and each of the donors in small and separate test tubes to supply a few
drop of serum; also a few drops from each in small test tubes containing
1 ce. of a 1 per cent sodium citrate salt solution to supply a suspension of
cells. The sera are separated and the cells washed once with the centri-
fuge, although the latter is not absolutely necessary. In setting up the
tests, hanging drop slides are employed as in the Widal reaction. Ona
series or cover glasses, two loopsful of the recipient’s serum is mixed with
one loopful of corpuscle suspension from each of the donors; in a second
series, one loopful of the recipient’s suspension of cells is mixed with two
loopsful of serum from each donor. The preparations are examined
microscopically with the low power objective, fifteen minutes later.
~
338 PROCEEDINGS
Controls are included with the cell suspension of the recipient and each
donor. Agglutination is well marked when it occurs and easily read.
With this method no attempt has been made to group the bloods but
it is extremely simple and it has been found very practical. I wish to
ask Dr. Coca’s opinion of the practical value of these tests inasmuch
as surgeons occasionally express themselves as believing that the tests
are not essential to successful transfusion, although he personally does
not share this view according to experience.
Arthur F. Coca: The choice of methods may depend upon what ap-
paratus is at hand. The procedure that I have described is quite as
easy as that of blood counting and all of the apparatus required for it
is available in any laboratory in which blood counting is done. Dr.
Kolmer’s method requires much more time, much more blood and more
apparatus than the one that I have described. His second series of
mixtures is unnecessary to the purpose in view.
So far as I am aware the deaths that have occurred as a result of
transfusion have happened when the compatibility test had not been
made.
2. THe. ISOLATION, PURIFICATION AND CONCENTRATION OF IMMUNE
HEMOLYSIN
M. Kosakai (see this volume, page 109).
DISCUSSION
William H. Park: Has Dr. Kosakai been able to separate antibodies
from the bacteria?
John A. Kolmer: I should like to ask Dr. Kosakai whether it is pos-
sible, with his method, to separate the hemagglutinins from the hemoly-
sins. It is desirable, particularly, where the anti-human hemolytic
system is to be used, to produce a serum preparation that is free from
agglutinins. Was Dr. Kosakai’s final solution free from agglutmin?
M. Kosakai: I have not attempted to separate anti-bacterial anti-
bodies from the bacteria with my method. I have not been able to
separate the hemagglutinins from the hemolysins.
3. A Rapip StwetE METHOD FOR THE EXTRACTION OF PRECIPITIN
ANTIGEN FROM BACTERIA
Charles Krumwiede, Jr. (see this volume, page 1).
DISCUSSION
George H. Smith: Can this “precipitin antigen” be employed also for
immunization?
Charles Krumwiede: I have not tried to produce an immune serum
with an antigen prepared in this way from bacteria. An extract made
in this manner from meat did not stimulate the production of antibodies.
PROCEEDINGS 339
4. A Metruop oF PREPARING BACTERIAL ANTIGENS
J.C. Small (see this volume, page 413).
DISCUSSION
John A. Kolmer: Several years ago I experimented with bacterial
antigen prepared by different methods, working with the typhoid
colon group. ‘The results of the complement-fixation tests were similar
to those of Dr. Small but with that method I was not able to differenti-
ate between paratyphoid A and paratyphoid B. Were rabbits used
for the production of the immune sera and if so was the serum used
active or was it heated? In previous work I found that the serum of
some rabbits when heated at 56°C. for one-half hour develops the prop-
erty of fixing complement in a nonspecific manner; for this reason
animals should be tested in a preliminary way before immunization is
begun and those that do not show this phenomenon may be selected.
Otherwise the nonspecific fixation may be avoided by heating the serum
at 62°C. instead of 56°C. for thirty minutes as was done by Meyer and
Boermer with the serum of mules.
Charles Weiss: Has the speaker found that the results are different
after the removal of the lipoids from what they were without this pro-
cedure? In regard to the heating of antigen it has been found in work-
ing with streptococcus antigen that heating destroys the anticom-
plementary qualities but at the same time it weakens the antigenic
properties of the preparation.
G. H. A. Clowes: In reference to the question of heating is it the expe-
rience of the members present that the quantitative activity of the serum
is diminished by heating at 62°C.?
John A. Kolmer: The heating of immune sera at 62°C. may cause
slight deterioration of the specific antibodies present. The deterio-
ration is, however, difficult to measure. Investigations have shown that
nonspecific inhibition of complement is especially likely with bacterial
antigens.
M.A. Wilson: Our experience has been that there is no deterioration
of the antigenic preparation if it is heated.
Hassow von Wedel: I have used tubercle bacilli antigen that has been
heated at intervals five or six times without noting any difference in
its antigenic value.
5. A CONTRIBUTION TO THE StruDy. OF COMPLEMENT FIXATION IN
TUBERCULOSIS
M. A. Wilson (see this volume, page 345).
6. A CONTRIBUTION TO THE StTuDY OF COMPLEMENT FIXATION IN
TUBERCULOSIS
Hassow von Wedel (see this volume, page 35).
340 PROCEEDINGS
DISCUSSION
Paul Lewis: I have been much interested and instructed by these
papers. The results may be valuable from a diagnostic point of view.
I ean add some data from my own experience from work done within
the last year. At first I used bacterial suspensions or the autolysate
which latter I found, lost rapidly in anticomplementary effect. Test-
ing with the same serum day after day, I found that one frequently
does not get the same result on two days running. I think this may have
been because the reaction between antiserum and antigen in those
cases was a weak one and that the readings were only permanently
positive with the strongest sera. Experiment showed that by prolong-
ing the fixation period more durable reactions could be obtained. Four
hours proved to be the maximum period which it was practicable to use
owing to irregular deterioration of the complement in longer exposures.
Using a four-hour incubation period I have titrated numbers of sera
to determine the amount of complement fixed. In certain instances
this may reach 2 cc. of the usual 1 + 9 dilution of guinea-pig’s com-
plement. It has been found that the tuberculous sera can be well pre-
served by mixing with an equal quantity of neutral glycerme. Pooled
serum from a number of cases giving a strong deviation, thus preserved
has been used over a period of six months to study the properties of
various antigens.
G. H. A. Clowes: In working on cancer cases and also on syphiltic
cases eight years ago I attributed the development of an increase in
complement fixing power in serum to changes in the colloidal particles,
due to their being held in the ice-box for some time. I wish to ask
what was the temperature of the ice-box used. Was the serum frozen?
Hassow von Wedel: I am not able to answer Dr. Clowes’s question.
The ice-box probably had a fluctuating temperature.
G. H. A. Clowes: I have seen complement deviation variations as
high as 10 to 1 It is easy to vary the surface of a particle by changes
in physical conditions or by variations of the hydrogen ion content.
This can frequently be determined by the ultramicroscope which shows
variations in the size of the particles.
William H. Park: I feel that, at present, this test should be used for
primary cases; later we may be in position to use it on a diagnostic
basis.
John A. Kolmer: I wish to ask whether the peculiar property of
human serum of developing fixation powers after standing was found
also for the Wassermann and gonococcus-fixation tests.
Miriam Olmstead: Did the guinea-pigs that gave a fixable comple-
ment for the tuberculosis antigen also give a fixable complement for the
gonococcus antigen? How many specimens giving positive gonococcus
fixation were used for the tuberculosis tests?
M.A. Wilson: We killed, at one time, 10 guinea-pigs and we found
that -of the sera all were fixable in the meningococcus fixation; four
were fixable in the gonococcus fixation and only one was fixable in the
tuberculosis fixation.
PROCEEDINGS 341
A Srupy oF CONTROLLED POSTMORTEM WASSERMANN REACTIONS:
A SUPPLEMENTARY REPORT ON 400 CASES
Stuart Graves: 1. Post mortem Wassermann tests confirmed ante-
mortem Wassermann tests in 97 per cent of 68 controlled cases. A four
plus positive reaction in a specimen obtained sixty hours post
oe as a four plus reaction obtained ante mortem in the
same one with anatomie and clinical evidence of syphilis. A
a eee in a specimen taken twenty-two hours post mortem
confirmed a negative ante mortem reaction.
2. In 91.2 per cent of the cases showing anatomic lesions of syphilis
and presenting evidence of syphilis in their histories the sera post
mortem gave positive Wassermann reactions.
3. The fact that only 2.5 per cent of the sera were anticomplementary
or otherwise unfit for use compares favorably with the fact that 1.14
per cent of 6000 ante mortem specimens were similarly unfit.
4. Only 2.6 per cent of 378 cases showing anatomic evidence of syphi-
lis gave negative reactions.
5. The reactions conformed to the anatomic and historic evidence in
304 of 378 cases or 80.4 per cent, which is considerably lower than it
would have been if satisfactory histories and physical examinations had
= recorded in Class V.
There is no logical reason for supposing that acute infections or
get tumors cause positive Wassermann reactions.
7. The positive reaction appeared in 2.7 times as Many negroes as
whites, in 1.7 times as many males as females and in only 11 white fe-
males or 6.5 per cent.
8. The Wassermann reaction, carried out on post-mortem blood
according to the methods followed in this investigation, is practically
as reliable a test for syphilis as when performed with ante-mortem
specimens and is of great value in pathological anatomy and in medico-
legal cases.
DISCUSSION
John A. Kolmer: I wish to endorse all that Dr. Graves has said in-
asmuch as my own results coincide entirely with his experience. I am
familiar with the published reports just described tending to discredit
the practical value of the Wassermann reaction and I think it is parti-
cularly unfortunate that reports published with this object in view should
reach the practitioner of medicine. No one that has dealt with patho-
logical findings would deny that a pathologist cannot exclude syphilis on
the basis of autopsy findings alone. Spirochaetes might not produce
much tissue reaction and yet suffice to produce antibodies, which would
be indicated by the Wassermann reaction. In the last four or five
years much has been learned of the technic of the Wassermann reaction
rendering it a test of great diagnostic value. While we must welcome
attempts to point out sources of error, criticisms should be carefully
controlled.
342 PROCEEDINGS
8. OBSERVATIONS ON THE INTRASPINOUS AUTO-SALVARSANIZED SERUM
THERAPY OF CEREBROSPINAL SYPHILIS
Benjamin A. Thomas.
9. EXPERIMENTS UPON THE PASSIVE TRANSFER OF ANTIBODIES TO
THE CEREBROSPINAL FLUID
John A. Kolmer and Shigeki Sekiguch (see this volume, page 101.)
DISCUSSION
A. Parker Hitchens: The question whether the antibodies find their
way from the blood to the spinal fluid is mportant, especially in cerebro-
spinal fever. In severe cases of epidemic meningitis in the army
camps the antimeningitis serum has been given by intravenous as well
as by intraspinous injection. The results of this treatment are en-
couraging. Dr. Kolmer’s work would show that specific treatment
by the intravenous route should be of value in meningeal infection.
J. Bronfenbrenner: It is still not settled definitely whether the anti-
bodies normally pass directly from the blood into the spinal fluid. This
passage is even more questionable in such pathological conditions in
which the pressure in the spinal canal is much greater than in the blood
vessels. The fact that the introduction of antimeningitis serum into
the circulation seems to produce clinical improvement could possibly be
due rather to another mechanism than the direct passage of antibodies
from the blood into the spinal canal. Recent investigations of English
workers lead them to the conclusion that the meningococcus produces a
soluble toxin. Such a toxin may be forced out of the spinal canal into
the circulation on account of the pressure in the spinal canal, and so
cause a certain amount of general toxemia aggravating the specific
symptoms of meningitis. The therapeutic serum may contain a cer-
tain amount of antitoxin and to that extent may improve the clinical
condition of the patient when introduced intravenously.
A. Parker Hitchens: About every three hours during the severe
stage of the disease the spinal fluid is removed, and with the rapid
filling up of the canal again there may be some effect.
John A. Kolmer: Our paper simply deals with the passage of anti-
body from the blood to the spinal fluid under normal conditions. In
the experimental animals it was found necessary to have a large
amount of antibody in the circulation.
10. VacctnE DosaGE
Joseph Head.
11. Tae Vaccine TREATMENT OF ACNE, WITH SPECIAL REFERENCE TO
THE R6LE OF BACILLUS COLI
Albert Strickler and Jay F. Schamberg: Thousands of cases of this
disease never come to the physician at all; it is only when suffering from
PROCEEDINGS 343
severe forms with great disfigurement, that the patients seek advice.
The lesions are chiefly in the sebaceous glands, which are very active at
puberty. Puberty is, in fact, the primary predisposing cause. Anemia
and constipation are found to be pretty constant accompaniments of
acne. The condition often develops after typhoid fever, and intestinal
intoxication evidently plays a réle in the etiology. In the treatment of
acne by vaccines, it was found that there is a complement fixation in
63 per cent of the cases with an antigen prepared from a colon bacillus
isolated from the intestinal tract of the patient. The fixation is higher
when the antigen is from the patient. It was resolved to treat 50 cases
solely with vaccines prepared from an autogenous colon bacillus. These
cases were controlled by cases treated with other methods such as
vaccines from other germs, therapeutic and hygienic measures. The
B. coli vaccines were found to possess better curative effects than any
other mode of treatment.
DISCUSSiON
Jay F. Schamberg: I would like to emphasize one or two points.
Vaccines prepared from B. acne and from staphylococci have been in
use for many years, and In some cases they have given brilliant results;
in others they have failed to respond. The present experiments were
carried out in a hospital clinic upon a large number of patients. 63
per cent gave positive complement fixation with the strains of B. coli
used for the vaccines. A large percentage of patients responded to a
remarkable degree. It is likely that at puberty, when there is great
developmental activity, there is liability to infection from the intestinal
tract. The activity of the intestinal organisms may then produce
noxious effects. The complement-fixation tests would seem to merimin-
ate especially the colon bacillus.
A. Parker Hitchins: It is animportant point to find out which strain
of B. colt is the chief factor in producing the disease. Has any work
been done on the sera in this respect?
Albert Strickler: In regard to the present work, I personally welcome
an active case as I feel that a good deal could be done for the patient.
Formerly I rather dreaded to see such patient come, as the results of
treatment were so often disappointing.
Jay F. Schamberg: No.attempt has been made as yet to differentiate
strains. The complement-fixation test merely showed that there is
greater specificity to the B. coli from active cases.
12. Lipo-VACCINES
Eugene R. Whitmore and E. Fennel
DISCUSSION
A. Parker Hitchens: This plan for the preparation of bacterial vac-
cines is likely to be of immense value. If one can give the entire treat-
344 PROCEEDINGS
ment in one day, the economic value of the procedure is obvious. At
Camp Beauregard I saw a detachment injected with meningitis lipo-
vaccines; the reactions were very severe, as were also those from triple
typhoid vaccine. The meningitis vaccine caused severe headache and
insomnia. All the symptoms disappeared the next day. I feel great
confidence in the outcome of the work of Colonel Whitmore and Lieuten-
ant Fennel.
Lieutenant Fennel: There was a very severe reaction to the first
vaccine we used. Since the improvement in technic the severity of the
reaction was much lessened.
13. A Srupy OF THE IMMUNIZING PROPERTIES OF BACTERIAL VACCINES
PREPARED AFTER VARIOUS METHODS
M.W. Perry and Sara Levy.
DISCUSSION
Lieutenant Fennel: I wish to ask Mr. Perry whether he titrated the
agelutinable culture by the standard set by Oxford. If so, the results
will be of considerable value for the army work.
John A. Kolmer: The culture was secured from a German university
sixteen years ago and was standardized for agglutination work. It is
highly susceptible to agglutination. It is not standardized by the
Oxford method.
A CONTRIBUTION TO THE STUDY OF THE COMPLE-
MENT FIXATION REACTION IN TUBERCULOSIS.
M. A. WILSON
From the Research Laboratory of the Department of Health, New York City
Received for publication May 14, 1918
I. ON THE STANDARDIZATION OF COMPLEMENT
In our study of complement fixation in tuberculosis we have
found a point of technique that has increased the efficiency of
the test. It has to do with the standardization of the guinea-
pig’s serum to determine the value of the complement. In this
preliminary report, we shall describe the method or standardiz-
ing the complement, the preparation of our tuberculosis antigen,
and the diagnostic test, with such results as we have obtained
thus far in our study.
Technique
All reagents are used in one tenth the classic Wassermann vol-
umes. Fixation period, one hour, 37°C.
Serum. The patient’s serum is inactivated for thirty minutes
at 56°C. 0.02 ce. and 0.01 ec. of the undiluted serum are used
in the test, and 0.04 cc. for control of anticomplementary action.
Antigen. Two antigens have been used. One was made
from 12 stock cultures of human tubercle bacilli, the other from
strain 305 (used for tuberculin production.) Those antigens
are suspensions in 0.9 per cent saline of dried bacilli, from which
all constituents soluble in aleohol and ether have been removed.
The bacilli were grown in glycerin-broth. The polyvalent anti-
gen cultures were grown for three weeks, and the monovaient
ones for three months. The broth cultures were killed by heating
them in the Arnold sterilizer for one hour. The cultures were
1 Preliminary report.
345
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 5
346 M. A. WILSON
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 10 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 a 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 ce.
of saline. This gave a concentrated emulsion convenient for
storing as a stock antigen. The emulsion was heated for one
hour at 80°C. The antigen was now ready for use, and it was
standardized to be used in such a dilution that 0.1 ec. contained
two standard fixation units and one fourth, or less, of the anti-
complementary 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 complement
known to be potent for tuberculosis fixation. The standard
dilution of the two antigens employed is 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. The tests
reported will show that they are specific and stable. They were
made ten months ago and are perfectly efficient today.
2 As determined with the use of one-tenth of the standard amount (i.e., 0.1 ce.
instead of 1 ec.) of 5 per cent sensitized sheep’s corpuscles.
3 In some instances a sample of antigen that has been standing in the ice-box
for some time has been found to have increased in its anticomplementary action.
This change is not accompanied by any deterioration of the ‘‘antigenic’’ property
of the preparation and, as we have found, it can be removed by heating the
diluted preparation for one-half hour at 56°C. The change is not of frequent
occurrence; however, as a routine precaution, we heat all of our diluted antigen
before using it in the tests.
COMPLEMENT FIXATION REACTION IN TUBERCULOSIS 347
Complement. Guinea-pig’s serum, twenty-four or forty-eight
hours old; pooled from six to ten pigs. Before the pooling, the
serum from each pig is tested for its hemolytic strength, for anti-
sheep amboceptor, for anticomplementary reaction with a hete-
rologous serum and for fixability with the combination of tuber-
culosis antigen and tuberculosis serum. This last test we
emphasize as an essential, if uniform results are to be obtained
with different lots of complement; it has proven beyond a doubt
that although a guinea-pig’s serum may react perfectly in all other
respects it may fail to be fixed by tuberculosis antigen and serum.
Of the pigs we have tested 64 per cent failed in fixability, while
they were perfectly good in other respects. If in a pool of six or
ten complements, there are several strongly fixable, the presence
of the negative complements in the pool may not appreciably
affect the test; on the other hand, if most of the complements
are negative the fixability of the pool will not serve to give a true
reaction with the patient’s serum. In such a case a four plus
reaction may drop to a two or three plus reaction and a two or
one plus reaction may become negative.
Tables 1 and 2 show the variation in fixability of the guinea-
pig’s serum.
Indicator for the fixation reaction. 0.1 ec. of a 5 per cent
suspension of sheep cells sensitized with three standard units of
antisheep amboceptor.
Controls for diagnostic fixation reaction:
Antigen—for anticomplementary reaction.
for specificity.
for potency.
Serum—for anticomplementary reaction.
for specificity.
for natural antisheep amboceptor.
Complement—for stability (system control).
Sensitized cells—for stability (reading control).
348
M.
A. WILSON
TABLE 1
OONooPr wW ND | pra No.
DATE BLED
November
November
November
November
November
November
November
November
November
November
November
November
13
13
13
13
13
13
13
13
13
13
13
13
COMPLEMENT FIXATION REACTION
Tuberculosis
No fixation
No fixation
Complete fixation
No fixation
Weak fixation
No fixation
No fixation
Weak fixation
No fixation
Weak fixation
Complete fixation
Complete fixation
Meningococcus
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Gonococcus
Complete fixation
Complete fixation
Complete fixation
Complete fixation
Complete fixation
No fixation
No fixation
Complete fixation
Complete fixation
No fixation
Complete fixation
Weak fixation —
TABLE 2
Showing the number of guinea-pigs serums efficient for tuberculosis complement
fixation
NUMBER OF PIGS BLED
10
3
11
8
129
NUMBER EFFICIENT FOR COMPLEMENT FIXATION
Tuberculosis
mPmeowW or POW WH WwW OF WwW re
o>
(or)
Meningococcus
Gonococcus
WHDDOPADPORP KF wWOAAWwWHFE
(o/)
ie)
aimboceptor.
The pooled serum is titrated with cells sensitized with 3 standard units of
The reaction is read at the end of fifteen minutes.
units of the pooled complement are used.
Two hemolytic |
COMPLEMENT FIXATION REACTION IN TUBERCULOSIS 349
Results
The results of tests of serums from 344 cases are given in table 3.
TABLE 3
CASES
: 5 __ | POLIOMY- :
| PULMONARY BONE AND TUBERCU ELITIS HAVING NO
ACTIVE JOINT LAR SPINAL SYMPTOMS
DISEASES GLANDS FLUIDS OF TUBER-
CULOSIS
per cent per cent per cent per cent per cent
Strongly positive.............. 67 12 35 0 0
Wed POSItIVe......56n0.5-5 4 25 10 23 0 0
"EDU Oe Ss oe GU ane ree 8 78 42 100 100
Conclusions
1. Not all guinea-pigs’ serums are efficient for tuberculosis
complement fixation.
2. The serum from each guinea-pig should be tested for
fixability with tuberculosis antigen plus tuberculosis serum before
pooling the complement for diagnostic tests.
II. OBSERVATION OF THE VON WEDEL REACTION
Dr. von Wedel, working independently in our laboratory,
found that some serums from active tuberculosis cases gave a
negative complement fixation reaction when the test was made
on the first day after bleeding, and when the same specimens were
tested a week later, having stood in the ice-box during the interval,
they gave a positive reaction. This occurred repeatedly on later
bleedings from the same patient. The complement used for all
tests had been previously tested for fixability with tuberculosis
antigen plus tuberculosis serum: therefore, the negative reaction
in the first test was not due to an inefficient complement. Serums
from non-tubercular cases gave no fixation at any period after
bleeding, and this fact rules out the question of non-specificity
of the later positive reaction following the early negative phase.
The controls for anticomplementary reaction in the patient’s
serum and in the antigen were all completely hemolyzed.
350 M. A. WILSON
This early negative phase was not demonstrated in the serums
from all tubercular cases; but the percentage was so large as to
be significant.
Having personally observed the accuracy of Dr. von Wedel’s
technique, and the many repeated tests he made to discover the
possibility of an error, I was convinced of the verity of the re-
action and of the necessity for making the later test before the
tuberculosis antibody content of all serums can be determined.
All tubercular serums tested, throughout the remainder of
my study, will be given the early and later tests.
A CONTRIBUTION TQ THE STUDY OF THE COMPLE-
MENT FIXATION REACTION FOR TUBERCULOSIS
HASSOW VON WEDEL
From the Bacteriological Department of the New York University and the Bellevue
Medical College
Received for publication May 14, 1918
The complement fixation reaction for tuberculosis has occupied
the attention of many investigators for the past seventeen years,
and it has been studied mainly from the standpoint of its pos-
sible value to clinical medicine in the diagnosis and prognosis of
this disease. The methods commonly employed for diagnosing
tuberculosis leave much to be desired and there is little doubt
that many cases escape detection until marked damage has been
done. A test, therefore, that will give a sure and early diagnosis,
is of the utmost value both to the patient and to the general
public.
The results of the early work with this complement fixation
test were of such a contradictory nature that they were of little
practical value. However, the reports of recent investigators
seem to promise that this test will be of marked value to the
clinician.
The following is a brief review of some of the more important
investigations made during the past few years.
Widal and LeSourd (1) 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 (2) in 1903 demon-
strated the presence of antibody capable of uniting with tubercle bacilli
and fixing complement in the sera of tuberculous animals. Wasser-
mann and Bruck (3) in 1906 also demonstrated the presence of an
antibody to tuberculin in patients treated with tuberculin. Caulifield
(4), Laird (5), Hammer (6), Calmette and Massol (7) using various
forms of bacillary emulsions as antigens, obtained results which were
351
BSA HASSOW VON WEDEL
very inconclusive, ranging from 33 per cent to about 97 per cent of
positive results in cases of tuberculosis. Much (8), using various acid-
fast bacteria as antigens, with sera from tuberculous and healthy per-
‘sons, obtained fixation in 77 per cent of the healthy cases, in other words,
a large number of non-specific fixations. Frazer (9), usimg 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. Dudgeon, Meek and
Weir (10) state that all of their cases that had been treated with tuber-
culin gave positive results. Products of the bacilli were found to be very
satisfactory as antigens. With an alcoholic antigen, (11) prepared from
tubercle bacilli they obtained 89.3 per cent of positive results. Bron-
fenbrenner (12) using an antigen of tubercle bacilli grown by Besredka
on egg broth, obtained a very high percentage of positive results, but
also obtained quite a large percentage of non-specific positive results
and 24 positive reactions with syphilitie sera. McIntosh, Fildes and
Radcliffe (13) criticized Besredka’s (14) antigen and concluded after
testing many antigens, that the living bacillary emulsion was best.
Inman (15) and Kuss, Leredde, and Rubenstein (16) found the antigen
non-specific. Stimson (17), using a variety of antigens, reported a
small number of cases with but fair results. Corper (18), in 1916
using an autolysate as antigen and also a bacillary emulsion antigen,
concludes that the complement fixation test for tuberculosis is not
absolute, being positive in only about 30 per cent of all clinically definite
cases of tuberculosis, both active and inactive. In 1917, Corper and
Sweany (19) comparing their autolysate antigen with the bacillary anti-
gen of Miller, (20) concluded that there is not a great deal of difference
between the results obtained with these antigens and that it is impossible
by means of this test absolutely to differentiate active from inactive
tuberculosis. They prefer the autolysate to the bacillary antigen.
They obtained 65 non-specific cross fixations out of 92 specimens of sera
that gave positive Wassermann reactions. Slack, Burns, Castleman
and Bailey (21) state that it appears from their observations that the
complement fixation reactions are specific for tuberculosis, and that
they obtain no cross fixation with positive Wassermann sera.
Of recent investigators, Craig (22) and Miller (23) obtained the best
results. Craig reports the results of 166 examinations on cases of pul-
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 353
monary tuberculosis, in which he used an alcoholic extract antigen of
human tubercle bacilli. He obtained 92.6 per cent of positive results
in active cases, and 66.1 per cent in inactive cases. Out of 150 cases of
syphilis, which were free from lesions in the lungs, none gave positive
reactions. All of 100 other examinations (various diseases) gave nega-
tive results. However, Lucke (24) and other investigators report that
alcoholic extract antigens, as used in Craig’s work, have many times
proven inpracticable and worthless, and as his description of the prep-
aration of his antigen was wanting in detail, the described preparation
was frequently found impossible. Miller reports 100 per cent of positive
results in active cases and 100 per cent of negative of results in non-
tubercular cases.
A great number of various antigens have been used and a
great diversity of results has been obtained, reaching from 30
per cent efficiency up to Miller’s claim of 100 per cent. There
seems to be no uniformity in the findings and conclusions of any
two observers; even the most promising of all these investigations,
that is, Craig’s and Miller’s are of limited practical value, as
serologists have had great difficulty in using their antigens, and
therefore have been unable to reproduce their results.
The main object of this preliminary study of the complement
fixation reaction for tuberculosis was, therefore, with the use of
the perfected Wilson antigen, which is easy to prepare and
which can be kept for a long period of time without becoming
anticomplementary, to find a method that would give specific
positive results in active tubercular cases and which would not
give non-specific results in negative cases.
I have made 1078 complement fixation tests on 200 specimens
of blood serum taken from cases with no clinical history of tuber-
culosis and from patients with active, inactive and primary
pulmonary tuberculosis. This study was made in conjunction
with Miss Wilson in the New York City Board of Health Re-
search Serological Laboratory. The cases were practically all
from the Westchester County Hospital and I have the complete
clinical data on all the cases. The clinical data consist of age;
past and present temperature, pulse and respiration records;
sputum reports and physical symptoms with clinical diagnoses.
354 “ HASSOW VON WEDEL
TECHNIQUE
The technique 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 I found no differ-
ence in the results, I have 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 complement fixa-
tion value. All these complements were titrated with 2.5 per
cent sheep cells, sensitized with three units of anti-sheep ambo-
ceptor; the unit was recorded at the end of fifteen mimutes.
Exactly 2 units were used in the regular test. i
Antigen. The Wilson antigen, which was used throughout
this study, is simply a suspension of tubercle bacilli killed with
heat, extracted with alcohol and ether, and dried. The com-
plete technique of preparing this antigen is given by Miss Wilson
in her paper on the study of the complement fixation reaction for
tuberculosis in the current issue of this journal.
Sheep’s 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 amboceptor in the water
bath for half an hour.
Amboceptor. Three units were used in the tests.
Fixation period. After the patients’ serum, complement, anti-
gen 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.
Six series of tests have been made for the purpose of comparing
the results after the following different methods of incubation:
one hour in the water bath; two hours in the water bath; two
hours in the water bath followed by two hours in the ice-box;
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 355
and four hours in the ice-box. Apparently the one hour water
bath incubation gives the most uniform results. The ice-box
incubation gives by far the weakest fixation. The two-hours
in the water bath followed by two hours in the ice-box gives
almost the same results as the two hour water bath incubation
alone which gave quite a number of anti-complementary reac-
tions in the tests where we doubled the regular Wassermann
amount of patients’ sera. These comparisons, of course, are too
few to allow a positive statement as to which is the best method of
incubation. A large number of comparisons will be made in the
near future to determine this question.
Results were reported as plus minus if any degree of fixation
was observed; 1 plus if marked fixation in the first antigen tube;
2 plus if complete fixation in first tube; 3 plus if complete fixation
in first tube and marked fixation in second tube, 4 plus if com-
plete fixation in both tubes.
During 1913, Dr. 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. Discarding
all those cases that were anticomplementary in the regular
Wassermann amounts, and considering only those which ordi-
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 investi-
gators have made favorable reports on the use of larger quanti-
ties of patients’ sera, I have made all my tests since January 1
with the regular Wassermann amount and with double that
356 HASSOW VON WEDEL
amount of patient’s serum; that is, 0.04 cc. of serum in the first
antigen tube with 0.08 cc. of serum in its control tube; 0.02 ee.
of serum in the second antigen tube with 0.04 ec. of serum in its
control tube, and 0.01 cc. of serum in the third antigen tube.
All the specimens of sterile sera from 95 known non-tubercular
cases, gave negative results with 0.04 cc. of patient’s serum in the
antigen tubes. Four contaminated specimens of serum from
known non-tubercular cases, gave non-specific weakly positive
reactions. Forty-nine sera from cases with active tuberculosis -
gave strong positive reactions with double the amount of serum
and only 43 gave strong positive reactions with the regular
amount of serum. Ten sera gave positive reactions of 2 plus
to 4 plus when double the amount of serum was used, and only
doubtful or negative results with the single amount. Of the type
II cases, 46 gave positive results with double the amount and only
41 with the single amount. Of the type IV cases, 6 gave positive
results with double the amount and only 2 with the regular
amount. We, therefore, had an appreciable number of cases
which gave definite positive results with the double amount and
negative or doubtful results with the single amount of patient’s
serum.
The patients were bled Thursday afternoons and the sera were
separated from the clots Thursday evenings or Friday mornings
and inactivated Friday mornings. Each specimen of serum was
tested the morning after it was removed from the patient and this
serum was retested week after week for from four to five weeks
with an interval of seven days between each test, together with
specimens from new patients each week. We, therefore, have a
record in many instances of eight weekly complement fixation
tests on the same specimens of sera kept under as nearly sterile
precautions as was possible in the ice chest at about 8°C.
By this proceedure, I made a very interesting observation
which may possibly account for some of the wide discrepancies in
the various complement fixation results reported by the different
workers. The complement fixation results on sera from positive
cases made the first day sfter taking the specimens from the
patients were in a very large percentage of cases negative or
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 357
weakly positive; while in most instances, seven days later these
same sera gave a strongly positive reaction and continued to give
this strongly positive reaction week after week with unvarying
regularity. None of the non-tubercular sera gave a positive
reaction the second, third, or fourth week after the specimen had
been obtained from the patient.
Of all positive sera in our series, only 12 gave a 3 or a 4 plus
reaction the first day tested. After being preserved in the ice-
box under sterile precautions for seven days, 49 sera gave 3 or 4
plus reactions. Of 82 sera from tubercular cases of all types that
gave negative or doubtful reactions the first day, the results on
the seventh day were positive in 37. The reactions apparently
did not change after the sixth day.
TABLE 1
Results obtained with the same specimens of serum one day, seven days, fourteen days,
and twenty-one days after the blood sera were removed from the patients
SERUM NO. FIRST DAY* SEVENTH DAY FOURTEENTH DAY TWENTY-FIRST DAY
13 = 4+ 4+ 4+
25 = 4+ Pee 3+
14 = 4+ 2+ -
61 + 4+ 4+ 4+
67t | = = = =
|
* This refers to the day after the blood was removed from the patient.
+ The serum was obtained from a normal non-tubercular individual.
I have attempted to find cut on just what day the positive
result first appears in most of these delayed reactions, but hereto-
fore I have been able to carry out only two series of daily titra-
tions and from these I could not draw any positive conclusions.
This work, however, will be continued and when our final results
are published, I hope that they will include information upon this
important question. The phenomenon just described apparently
bears no relationship to the type of case, as I find it occurring in
the old active cases, in the primary cases and in the inactive cases.
In order to see whether the phenomenon was possibly due to
anticomplementary reaction, I have used four times the usual
amount of patient’s serum in the serum controls without, however,
358 HASSOW VON WEDEL
obtaining any more anticomplementary results with this large
amount of serum at the end of the first week than on the first day.
This phenomenon may be due to the presence of some inhibitory
substance which disappears from the serum upon standing for
several days.
Stimson mentions the role possibly played by the inhibiting
substances in the patient’s serum, which are stated by Caul-
field, Calmette and his co-workers to occur in certain tubercle
sera of tubercular individuals and which have the effect of pro-.
ducing negative reactions in sera that contain anti-bodies. They,
however, did not state whether these inhibitory substances
disappeared from the sera upon standing. This is a point that
will require rigid investigaition.
Various specimens of serum taken from the same patient at
different times gave complement fixation results which were
comparable. I took duplicate specimens from one to six weeks
apart on 31 cases. The results on 24 were identical. The re-
maining 7 gave results that were closely alike; the slight differ-
ences being, perhaps, explained by the variations in the physical
condition of the patients.
CLASSIFICATION OF THE CASES
The types of tuberculosis cases have been variously classified
by many investigators. Noted classifications were made by
Williams, of Brompton Hospital, Cornet, L. Bard, Koeniger,
Turban, (25-26), Meissen (27), and Walther L. Rathbun (28).
Rathbun classifies tuberculosis as incipient, moderately advanced
and far advanced. The following broad classification seems to
be necessary both from the clinical and from the laboratory stand-
point. Types II, III and IV must be separated, for they give
entirely different clinical pictures and also entirely different com-
plement fixation results.
Type I. Primary cases; very few physical symptoms present;
no tubercle bacilli found in the sputum or found only after the
examination 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.
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 359
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. 7
TABLE 2
Comparison of results in the various types of cases
% a ! Ah & 2 Oe. AA AA 1 ed 1!
Behera olor Se |) Se. eH aleeec| Beane Lone
eee) |. ee | eee) emer E :
Q Z DZ DZ
S 5 Allen gE Gey WiGrrs eS SS» | ta 3
i=] a a & =) a a Mon Mong a
n fe k& A QP hes gaRO eis) ae E, = aS
< vr o op Oo 2 as) oP ° 4 5 a OZ OoZ<
o ps a Z <8 Z & zat Aa we C) 3a
& 2 eg ee a I aise Aap: Bp % 3
a r=} 2 2 D
ro) R Ap a Za ae 27a Bm & Ansa Be aad
a om ee] as D Bae Aes Ag > gam aye
= = = =P Om RO cele! 20s saa aps BDz
S 3) be pe ae gA Bae pm & pan p@A paw
B Z ZA a ry cy ry Z Za Z Z
1 6 6 0 100 0 83 1 6 1 5
2 47* | 46 1 98 2 90 15 46 10 41
3 11 3 3 27 27 100 1 3 1 3
4 19 6 5 31 26 0 1 6 0 0
5 12 3 0 25 0 0 0 3 0 0
6 4 0 0 0 0 0 0 0 0 0
7 99 0 0 0 0 0 0 0 0 0
|
* Two sera anticomplementary.
In table 2 I have grouped the results of the complement-
fixation tests according to the clinical type of the cases examined.
Of 6 type I cases, all gave positive complement fixation re-
actions, or 100 per cent of positive results. Of 49 type IT cases,
46 were positive, 1 was doubtful and 2 were anticomplementary,
or omitting the 2 anticomplementary sera 98 per cent were posi-
tive and 2 per cent were doubtful.
Of 11 type III cases 3 were positive, 3 were doubtful and 5 were
negative; or 27 per cent were positive and 27 per cent were
doubtful. Of 19 type IV cases, 6 were positive, 5 were doubtful
and 8 were negative; or 31 per cent were positive and 26 per cent
were doubtful.
360 HASSOW VON WEDEL
Of 12 type V cases, 3 were positive, 9 were negative; or 25 per
cent were positive. Of 4 type VI cases, all were negative. Of
92 sterile type VII sera, all were negative. The contaminated
type VII cases gave weak non-specific reactions.
The clinical diagnoses were all made either by Dr. Rosenberg,
who was formerly diagnostician for tuberculosis at the West-
chester County Hospital, or by Dr. Slade, diagnostician for tuber-
culosis for the New York City Board of Health.
As Dr. Rosenberg left the Westchester County Hospital before
my work was completed, some of the latter cases were diagnosed
and classified by the internes in the hospital. All of these latter
cases were examined and reclassified by Dr. Slade thus the above
diagnoses were all made by expert diagnosticians. Dr. Slade’s
diagnoses were made after my results had already been recorded.
In attempting to determine whether the temperature, pulse
and respiration records of the patients bore any direct relation-
ship to the complement fixation reactions, I have compared my
laboratory findings with the record charts, and I find that the
cases giving a 4 plus reaction had an average temperature of
99.4; pulse, 94; and respiration, 27. The patients giving a 2
plus reaction had an average temperature of 99; pulse, 85;
respiration, 24. Those giving a doubtful or negative reaction
had an average temperature of 99; pulse, 95; and respiration, 25.
40 per cent of the 4 plus eases and only 10 per cent of the doubt-
ful and negative cases had a temperature of over a hundred. In
other words a large percentage of cases having a high temperature
and respiration records gave strongly positive reactions; vice-
versa, a large percentage of cases having low temperature and
respiration records gave negative reactions. Forty per cent of the
4 plus, 10 per cent of the 2 plus and only 4 per cent of the doubt-
ful and negative cases had a respiration record of 30 or over. I
also attempted to see whether, possibly, the age of the patient
had any effect on our reactions and found that the average age
of the patients giving a 4 plus reaction was 36; the average age of
those giving a 2 plus reaction was 43; the average age of those giv-
ing a 1 plus reaction was 35; the average age of those giving a
plus-minus reaction was 40. As all of the groups contained pa-
tients both young and old, no conclusions could be made.
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 361
_ Miss M. A. Wilson has observed that the serum of a large per-
centage of guinea-pigs is unsuitable for use in the complement
fixation test for tuberculosis because the complement of these
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 Bron-
fenbrenner (29) in their article on the ‘Variation of the comple-
ment activity and fixability of guinea-pig’s sera in Wassermann
work.” 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
specimen of guinea-pig’s serum. However, the irregularity
observed by the latter authors was relatively infrequent as com-
pared with that reported by Miss Wilson, who found that for the
complement-fixation test in meningococcus and gonococcus
infection and in tuberculosis, respectively about one-tenth,
one-third and two-thirds of all guinea-pigs supply inefficient
complement.
With the purpose of further studying this phenomenon I
carried out a number of tests with sera from actively tubercular
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 or eight guinea-pigs. These parallel tests gave practi-
cally the same results. In a few instances I obtained a 2 plus
instead of a 1 plus reaction, or a 3 plus reaction with the tested
complement and a 2 plus reaction with the untested complement;
but this difference was not regularly encountered and in a few
instances, in fact, better results were obtained with the pooled
untested complement than with the specially tested complement.
This experiment being inconclusive I then carried out a series
of tests with sera from seven frank tubercular cases (six of which
I had already examined) testing each guinea-pig’s complement
*
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 3
362 HASSOW VON WEDEL
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 taking
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 ec. of patient’s serum and a 4 plus reaction
with double that amount. Tests made on the ninth and eleventh
days with complement from guinea-pig 3, gave negative results
TABLE 3
Showing variations in flexbility of the complement of different guinea-pigs’ sera in
the complement-fixation test in tuberculosis
+
+ if:
COMPLEMENT|COMPLEMENT|COMPLEMENT COMPLEMENT COMPLEMENT |COMPLEMENT
no. 1, no. l, No. 2, No. 2, No. 3, No. 3,
FIRST TEST | LATER TEST | FIRST TEST | LATER TEST | FIRST TEST | LATER TEST
POOLED
TESTED COM-
PLEMENT
FIRST TEST
PATIENT’S
SERUM
NO.
Say || 9 ce eS eee | ee eee
1
2} — — | — — |44 34 /4+ 34+]14 -— | — -— | 44+ 2+
3} — — | — — {44 24/34 2+] — -— | —- -— |14+ =
4} — — | — — [44+ 44]4+ 44 ]}/14 - 44 3+
5}2+ — |384+ 14 /4+ 44/44 44/44 -— | 44 -— | 4t 44
6} — — | — — |4+ 14/384 14] —- -— | —- -— ] = =
7) — — |2+ — [44 44/44 44] —- — | — — 184 14
* The results in the first column were obtained with 0.04 cc. of patient’s serum.
1 The results in the second column were obtained with 0.02 ec. of patients’
serum.
t Complement was preserved in the interim with an equal amount of 18 per
eent salt solution and kept in contact with ice.
with all amounts of patient’s serum. On the fourteenth day, the
pooled complement, made from sera of six tested guinea-pigs,
gave results which were practically the same as those obtained
when complement from guinea-pig 2 was used, that is, a 2 plus
reaction with 0.01 ce. and 0.02 ec. of patient’s serum and a 4
plus reaction with 0.02 cc. and 0.04 ce. of patient’s serum. On
the sixteenth day, complement from an additional guinea-pig
which may be designated 4 gave aplus-minusand a 1 plus reaction.
Patients’ sera 2, 3, 4, 6 and 7 gave almost identical results.
Serum 5 gave a 1 plus and a negative result with complement
ela a ieee
a
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 363
-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.
The results of this limited experiment confirm those reported by
Miss Wilson, inasmuch as only one of the three individual guinea-
pigs’ sera was found to be suitable for use in the test. In view
of the fact that with this one serum (guinea-pig 2) the different
patients’ sera reacted differently and since the otherwise un-
satisfactory guinea-pig’s serum 3 was fixed with the double
amount of one of the patients’ sera (serum 5) it would seem advis-
able in testing guinea-pigs’ sera as to fixability, to test them with a
strongly reacting patient’s serum, or, better, with two such sera
and to use the latter in the usual quantity, not with the double
quantity.
H. J. Corper (19) states 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,
concludes that in the presence of a positive Wassermann reaction,
the presence of a positive complement fixation test for tuberculosis
is of no practical value. He also states (18) that, while the most
reliable investigators 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 bacillary emulsion are the small
leeway between the antigenic and the anticomplementary dose,
the turbidity produced in the tubes and the fairly high percent-
age of non-specific reactions.
My own experiences are in disagreement with all of Corper’s
conclusions. First, because in my series of cases there were
26 specimens from patients that gave positive Wassermann
reactions but offered no physical symptoms of tuberculosis. None
of these gave any cross fixation with the Wilson tubercu-
losis antigen. Secondly, in no instance did I obtain a positive
reaction using double the regular amount of sterile patient’s
serum in my entire series of known non-tubercular cases. Thirdly
I have never found the Wilson antigen to be anticomplementary
in four times the dose used in the test, if the antigen is heated at
364 HASSOW VON WEDEL
55°C. for one half hour just before being used. Fourthly, my
tests in the active tubercular cases gave 98 per cent of positive
reactions. Fifthly, the very slight turbidity produced by the
antigen in no way interferes with the reading of the results.
CONCLUSIONS
1. The tubercle bacillus antigen of Miss Wilson is not anti-
complementary in four times the amount capable of producing
positive complement fixation with sera from the great majority
of cases of active tuberculosis.
2. Pooled complement from at least six guinea-pigs should be
used in making the tests, or the complement from single pigs
should be tested for its complement fixation value with known
positive sera.
3. Double the original Wassermann amount of patients’
serum should be used.
4. No report should be made until the sera has been tested,
after having been kept under sterile conditions in the ice chest
for from four to six days, preferably six days.
5. My results seem to indicate that if the afore-mentioned
modifications of the original complement fixation tests are used,
100 per cent of non-tubercular cases will give absolutely negative
results; nearly 100 per cent of the primary and active cases will
give positive results with the exception of the dying cases; and
about 25 per cent of the partially inactive and inactive cases will
give only weak positive results.
Before final percentage results can be arrived at, it will be
necessary to make many more tests on a large number of sera
from active, inactive and incipient pulmonary tubercular cases
and a large number of control sera from non-tubercular cases.
I hope to report on the results of about three thousand tests in
the fall.
I wish to take this opportunity to thank Dr. Wm. H. Park
and Dr. Russell, chief of the staff of the Westchester County
Hospital for their invaluable assistance and advice, without
which this study could not have been made.
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 365
NAME
May Cooper: s...c%- i...
LORE BG) ia eee
O70 fd i eae |
SPUTUM REPORT*
essoltreulie
PH+tt++t+et+tteeteet+segs+ste
TYPE OF DIAGNOSIS
NNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNKH HBR Yee
CHART 4
Clinical and laboratory records of all our tuberculosis patients
TEMPERATURE
OO
I 0S
98
130
135
100
136
RESPIRATIONT
COMPLEMENT-FIXATION RESULTS
ONE DAYy* |One week*| LATER®
gia uh oi
ama eaeal| resi «|v | ae] ee
0) ea es er
44 0 0
—| |4t] s+
+] - B+ [44 |2+
+ | = [4+ [2+ [3+ | +
+] — |4+ [2+
es ea ees lee
2+ | + [4+ |4+
Oa ae dee |e
24+ | — 4+ |4+ [3+ | +
2+ | — 4+ |4+
4+ 4+ [a+ + 4+ 2+
4+ \24 [4+ [3+
4+ [2+ a+ [3+
4+ l2+ [4+ [3+
= Aa
38+ 4+ 44
A+ 2+ [4+ |4+ [44 [4+
2+ | — |4+ |/4+
—| [4+] f+
—| B+ ~
oe
+) +] [8+ 4+ [2+
4+ 4+ |4+ |4+
4+ |4+ |44 |44 /44 /4+
44+] + 2+ | + 4+ [4+
— | — [4+ [4+ [44 [3+
4+ /2+ /44 |4+
4+ 3+ | 0} 0 |4+ |4+
—| 4+] [4+
—/| f+] [4+
—| fat] 4+
+] — [4+ |4+
2. 4+
—| [4+] 4+
—| [4+] [4+
—| jab 4+ jo+
4+} 4+ -
366 HASSOW VON WEDEL
CHART 4—Continued
2 COMPLEMENT-FIXATION RESULTS
B18 i |
o iy 2 % | One day* |One week*| Later®
NAME a A & 2
pte ele lela|@ o | alae
= :
: + 2/25 | 99] 76) 24) — | — 2+ |=
DHICIOS = cee ne ae =e: { ri Ge (eM el nb uD) ee i ne Pcl Far
Mannix samara: + ole + 2 | 39 | 99 |108 | 24 4+ 4+
a ae { + | 2] 50] 98 | 84} 26} — | — |84+ |] + /2+ ] —
ey hone a) eR alae o7 |72 |} 93,| — | — lap || eee
Smit hieee cheers cece — 2] 35 | 98 | 84} 20] — |} — j8+ ] +
May Hide {3}, ayer tue) 28 |
RU MEVOSS cat tie e-werers ei { Tally eae ee z ee oa
+o 2 haa aS dS = ac
7 2 | 34 |101 {122 | 28 == 2+ 4+
Mc wenieee eee see , in 9 102 {104 | 29 | | = lat 1b
IGUSTIZO Sot tan cetct +} 21] 53] 98 | 86} 32 2+ 38+ 4+
Kiernanh erties nee + 2 Ne52 | 974-76) 20 — — |4+ /2+
Taylor { — | 2] 39 | 98 | 88 | 24 _ = (2+ | —
SAN. aa ees ins 67-76 )\oo tee leet] eee
Moeia { + 2;}40| 98 | 76} 22;—/}—}—|—-/]+] =
Pete ee eRe roa ae 100 | s0|20| —| — lor | +
: +] 3] 32] 99) 86} 29| +] = |44 /38+
Mirsainicheeremccccess { Get og ea] a) ge ge
AaigetPerilla-ckg.c cee oe + 3 | 19-|102 {132 | 30 4+ 4+ 4+
Wood titre ee see + 3 | 44 100 }100 | 28; =} —}|—|]—]—] =
Chas: Doyle=:.=:.4.'.8: { gid (ial coe se 8 eer bea etal a
3 SF | ea ee
sly ae BANAT Feteig | au ee
Kouwarte.. eee _ 3 | 34 | 99 |102 | 30 = — —
st 3 97 | 76} 22} —|} —|]—]|—
+ 3 | 20 |104 |142 | 28 _ _—
Rogneracancseane se one { ao ge ae
+ 4) 55} 98 | 72} 22) —|—/]}+] —
Monahanesenne sneer +) 4 97 | 72} 20} —| —| —]| —
are aE |e | i
Wipe Ahrens. nc: ie bree orme + 4} 53 | 98 |112 | 26 — — _
+] 4] 48 | 98 | 90 | 28 _ —|+]-—-
© Bennett#ea.-.. so. { a 4 Oeil ea leoanl =a lcs |-ae| oe is
; se 4 |} 31 |101 |110 | 24 _ _ _
PiraROussO-eee ese { moles pee ils aes teed i$ ee
Deets eet es tae + | 4] 44] 98] 80 | 22 + 2. | aan
Pohsamisarers too sscce + | 41] 26 | 99 |104 | 28 o 0] ac] ae
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 367
CHART 4—Concluded
HI COMPLEMENT-FIXATION RESULTS
°
4
o = % | One day* |One week*| Later®?
NAME a = 2
* < & ++ tor
a n a mM P
+ 4} 30 |100 |116 | 29; —} —}] —] —
SS { ill pe ee ilhet
i — 4 | 30 |100 | 88 | 28] ac} ac] — ] —
EVO a eae 59 2) a aH
Mena Stokes..i. 0.0... 5) 4) 35] 99 | 86] 22}; =|] —]—] —
i Spr cla) I Gil 86.1822) eee eed lee
ge burturmmack.<........ { rd 4 eee ie he es le
N — 4 | 41 | 98 | 92 | 24 — j2- )e) =] =
SDIN7o eee AY pale | ee ee Pn
— 5 | 32 | 99 {120 | 26 — —|—|-—
Jeo |S) V0 | er { S BU BD 597 >. 76 al Seales Fs Ee cea ieee |
WaerBennett: i... .... 2.0. _ 5 | 54 | 98 | 96 | 24 — —|/—-|-—-
LEO Cae 5 ens Sale ee 5 | 44 | 97 |104 | 26 —/—-/;/—]|-]-—-
12-0700 Re aa 5 | 43 | 98 | 68 | 22; —| —|] —] —
MMOPOUIE SSG. coc es — | 5 | 65'| 981 92 | 28 — a 25
The =e | 47 1-97 1-76: 122 — —/-|]-
MNDOU AGS So lacest. creas = |) 5) 120) 99" || 725° 20 - —|—|—
LSGT Sees oa nee - 5 | 49 | 98 1100 | 26 — —|/—-—|-—-
Leno Fiorella......... { iz : BEEN NUEO |, Ze a si z % yer
May Sedden............. — 5 | 50'| 98 | 88 |) 22 j2-- | =
Wine Webbe 22h ese... - 6 | 34 | 97 | 80} 22 _ — —
MMe =|) 6 |y35 | 98 | 90 | 24 = a Pees
eet =" | = 6: | 402) 97 1-80) || 22 —;}/-;|}—-—-]-]-
“SEG 0 ESS sO ae ea _ 6 |.22 | 99 | 70 | 24°) — | — |= =
* Sputum report obtained within two weeks of the time the blood specimen
was removed.
{ Temperature, pulse and respiration were taken the same afternoon the blood
was removed.
{ Refers to the use of double the regular Wassermann amount of patient’s
sera.
§ Refers to the use of the regular Wassermann amount of patient’s sera.
“ Refers to the time the reaction was made after the specimen was removed
from the patient.
° These results were obtained from two to six weeks after the specimens were
removed from the patjents.
368 HASSOW VON WEDEL
REFERENCES
(1) Wipat anv LeESourp: Cited by Shennan and Miller. Edenburgh Med.
Jour.; 1918, 10, 81.
(2) Borpet AND GENGoU: Compt. rend. Acad. de sc., 1903, 187, 351.
(3) WASSERMANN AND Bruck: Deutsch. med. Wehnschr., 1906, 32, 449.
(4) CAULFIELD: Jour. Med. Research, 1911, 24, 122.
(5) Larrp: Jour. Med. Research, 1912, 27, 163.
(6) Hammer: Munchen. med. Wehnschr., 1912, 59, 1750.
(7) CALMETTE AND Massou: Compt. rend. Soc. de biol., 1912, 73, 120.
(8) Mucu: Miinch, med. Woch., 1912, 59, 685.
(9) Frazer: Ztschr. f. Immunititsf., 1913, 20, 291.
(10) Dupcron, Meek AnD WetrR: Lancet, 1913, 184, 19.
(11) DupGron, MEEK AND WerR: Jour. Hyg., 1914, 14, 52, 72.
(12) BRONFENBRENNER: Arch. Int. Med., 1914, 14, 786.
(13) McInrosu, FILDES AND RapcuirFe: Lancet, 1914, 185, 485.
(14) BesrepKa: Compt. rend. Acad. d. sc., 1913, 156, 1633.
(15) Inman: Compt. rend. Soc. de biol., 1914, 76, 251.
(16) Kuss, LAREDDE AND RUBENSTEIN: Compt. rend. Soc. de biol., 1914, 76, 244.
(17) Strnson: Bull. Hyg. Lab. U.S. P. H. and M. H.S., 1915, no. 101, 7.
(18) Corprr: Jour. Inf. Diseases, 1916, 19, 315.
(19) CorpER AND Sweany: Jour. A. M. A. 1917, 68, 1598.
(20) Miuuer, H. R., anp Zinsser, J.: Proc. Soc. Exper. Biol. and Med. 1916,
13, 134.
(21) Stack, Burns, CASTLEMAN AND Battery: Jour. Amer. Med. Ass., 1917, 68,
1386.
(22) Craic: Am. Jour. Med. Sce., 1915, 150, 781.
(23) Mriuuer, H.R.: The clinical value of complement fixation in tuberculosis, The
Jour. A. M. A., Nov. 18, 1916, p. 1519.
(24) Luckre, Batpwin: Jour. Immunol., 1916, 457.
(25) Turpan: Diagnosis of tuberculosis of the iung, Edition 1906, 42-49.
(26) Turpan: Jour. of Amer. Med. Ass., Jan. 30, 1909.
(27) Merssen: Handbuch der Tuberculose, Band 1, 743, published by Barth,
Leipsig, 1914. ;
(28) Ratrusun, W. I.: Tuberculosis Monograph, no. 4, March, 1917, N. Y. C.,
Dept. of Health.
(29) Nocucut AND BRONFENBRENNER: J. Exp. Med., 1911, 13, 69.
(30) Rupee. AND RickMANN: Zeit. f. Immunititsf., 1910, 6, 344.
(31) Laus: Zeit. f. Immunititsf., 1911, 9, 126.
(32) Scuutrz: Zeit. f. Immunititsf., 1911, 9, 709.
(33) BorissJAK, SIEBER AND METALNIKOW: Zeit. f. Immunititsf., 1911, 12, 65.
(34) Porter: Jour. Hygiene, 1911, 11, 112.
(35) CauLFrireLp: Arch. Int. Med., 1911, 8, 440.
(386) CAULFIELD: Arch. Int. Med., 1911, 8, 128.
(37) DEBRE AND ParaF: Compt. Rend. Soe. Biol., 1911, 228 et seq.
(88) Dr1LMANN: Zeit. f. Immunitatsf., 1911, 10, 421.
(39) Larrpb: Jour. Med. Res., 1912, 27, 163.
(40) Mouuers: Cent. Bact. ref., 1912, 54, Beiheft, 202.
COMPLEMENT FIXATION REACTION FOR TUBERCULOSIS 369
(41) Hammer: Deutsch. Tierarzt. Woch., 1912, 20, 593.
(42) Zweta: Berl. klin. Woch., 1912, 49, 1845.
(43) CALMETTE AND Massou: Compt. Rend. Soc. Biol., 1912, 73, 122.
(44) Meyer, Kurt: Zeit. f. Immunititsf., 1912, 14, 359.
(45) LetuLue: These Fac. Med. Paris, 1912.
(46) Aoxr: Zeit. f. Immunititsf., 1912, 13, 192.
(47) Fraser: Zeit. f. Immunititsf., 1913, 20, 291.
(48) DupGcron, Merek anp WEIR: Lancet, 1913, 184, 19.
(49) Harris AND Lanrorp: Jour. Inf. Dis., 1913, 13, 301.
(50) Bana anp ANDERSON: Cent. Bact. Orig. 69, 517.
(51) AnMAND, DELILLE, Rist AnD VAaucHER: Compt. Rend. Soc. Biol., 1913, 74,
791.
(52) WyscHELLESKyY: Zeit. f. Tuberk., 1913, 19, 209.
(53) BunpscuuH: Zeit. Hyg. 1913, 73, 427.
(54) FRANCESCELLI: Zeit. f. Immunitiitsf., 1913, 20, 309.
(55) KINGHORN AND TwITCHELL: Zeit. f. Tuberk., 1913, 20, 11.
(56) CALMETTE AND Masson: Compt. Rend. Soc. Biol. 1913, no. 28, 160.
(57) Rorue AND BrerBaum: Deutch. Med. Woch., 1913, no. 14, 544.
(58) Momose: Deutch. Med. Woch., 1913, no. 22, 1029.
(59) BesrepKa: Zeit. f. Immunititsf., 1914, 21, 77.
(60) Wwepensky: Revised by Wulffius., Zeit. f. Immunititsf. referat, 1914, 8,
931.
(61) DEBAINS AND JUPILLE: Comp. Rend. Soc. Biol., 1914, 76, 199.
(62) Inman: Comp. Rend. Soe. Biol., 1914, 76, 251.
(63) McInvrosH AND Fiupss: Lancet, 1914, 11, 485.
(64) Rapciirre: Lancet, 1914, 11, 488.
(65) Duparon, MEEK AND WetrR: Lancet, 1914, 11, 72.
(66) BRONFENBRENNER: Jour. Exp. Med., 1912, 15, 598.
(67) BRONFENBRENNER: Soc. for Exp. Biol. and Med., 1914, 11, 92-93.
(68) E1caHorN AND BuuMBERG: Jour. Agricultural Research, 8, no. 1.
‘J - ray we ie ot | bi 7 ale ' 7 i ros
rth + . ; i
mh ’ ip r a
i . i : “ : a
: a
/
«
THE ROLE OF IMMUNITY IN THE CONDUCT OF THE
PRESENT WAR!
JOHN A. KOLMER
Philadelphia
When the history of the present great war is written a notable
victory over the common enemy, disease, will be recorded as one
of the greatest triumphs in this greatest of all conflicts. In all
probability this triumphs over disease will also be recorded as the
most important single factor in explanation of the stamina and
long-sustained man-power of the involved nations; never before
in the history of the world have so many men been engaged in
combat with such freedom from internal deterioration due to
disease not only among the warriors in preparation and at the
line of battle, but also among the supporting civilian population;
history records many instances of cessation of wars and sieges
due to disease among offenders or defenders or both and a re-
markable freedom from pestilence in the present conflict has
undoubtedly played a prominent réle in permitting it to reach
the dimensions of the greatest of all wars.
This triumph over disease is due in most part to prevention
by sanitary measures, specific immunization and improved meth-
ods of treatment of the inevitable and unavoidable sick and
injured. With the exception of small-pox, in which disease the
science of immunity long ago contributed the most important
and one essential means of prevention in the form of cow-pox
vaccination, sanitary measures embracing the proper disposal
of infectious material and the prevention of the spread of infec-
tious diseases by the processes of isolation and quarantine and
including the maintenance of individual resistance by proper
1 Presidential address at the Fifth Annual Meeting of the American Asso-
ciation of Immunologists held at the University of Pennsylvania on March
29 and 30, 1918.
371
372 JOHN A. KOLMER
food, work, rest and play, has played the most important rdle,
with the science of micro-parasitology and immunity embracing
a knowledge of the parasitic causes of so many of the acute in-
fectious diseases and specific immunization of several by means
of vaccines and sera, a close second worthy of the division of
honor and credit.
Mention has just been made of cow-pox vaccination in the
prevention of small-pox; history shows that without this im-
munological discovery and process great wars would be impossible
and particularly one of the present dimensions involving so many
countries and millions of men and offering splendid facilities
for the rapid dissemination of the virus; the prevention of ty-
phoid and paratyphoid fever by means of active immunization
with vaccines while not as successful as cow-pox vaccination,
must be credited with a great measure of success in the pre-
vention of these diseases formerly so widely prevalent among
armies; certain measures of success which in some instances are
quitemarked, have also attended the prevention of bubonic plague,
bacillary dysentery, cholera and rabies by means of active
immunization.
The prevention and treatment of tetanus and diphtheria with
their respective antitoxic sera have proven most valuable im-
munological procedures and particularly so in the prevention of
tetanus at a time when the modern earth digging methods of
war have widely distributed the bacillus and rendered practically
every wound regardless of severity and location a real danger and
menace to life; likewise in the treatment of epidemic cerebrospinal
meningitis a potent antiserum has proven conclusively that it is
the best means of treatment, its free and intelligent use resulting
in a considerable reduction in the percentages of death and the
disabling sequelae. In the treatment of that dreaded disease,
pneumonia, ‘‘The Captain of the Men of Death,” the science
of immunity has contributed a means for the serologic diagnosis
of the type of pneumococcus present and prepared a serum for
the treatment of type 1 infections which has proven its worth
and right to a prominent place in the modern treatment of this
disease. Still more recently the science of immunity has pro-
THE ROLE OF IMMUNITY IN THE PRESENT WAR ite
duced for the toxins of the gas-producing bacillus which has
played havoc among so many of our wounded heroes in the past
and present, a serum that bids fair to prove of value in the
prevention and treatment of this dangerous infection.
Immunological reactions are also proving of practical value
at the present time in the diagnosis of several diseases and
particularly the serological reaction in the diagnosis of syphilis,
which disease menaces all peoples at present and particularly
in the future, by reason of its wide dissemination and insidious
nature rendering all persons regardless of age and sex vulnerable
and liable to its attack. Furthermore in the treatment of this
“Third Great Plague,” the newly developed branch of chem-
otherapy in the field of immunity, has contributed a most
remarkable remedy in the form of dioxydiaminoarsenobenzol
or the popularly known ‘606,’ and our hopes for the present
prevention of syphilis and protection of the future and unborn
peoples, resides in large part in the treatment of the infected
until they are rendered less infectious even if not completely
cured, by the widespread and more free employment of this and
other anti-syphilitic remedies. To this end all efforts made to
lower its cost and thereby facilitate its use in the treatment of
the poor and of large numbers of persons, are to be welcomed as
commendable and a work of first rank importance.
Therefore, while the science of immunity has contributed
considerable that is of practical value in the diagnosis and treat-
ment of various diseases of particular importance in relation to
the present war, much and indeed more, remains to be accom-
plished of which mention may be made of but a few of the more
pressing problems as follows: The discovery of a test of effective
natural immunity to pneumonia and meningococcus meningitis,
if such immunity exists, comparable to the Schick test for im-
munity to diphtheria, as a means of encouraging and facilitating
active immunization with vaccines in the prevention of these
diseases; a test for natural immunity to tetanus, which may be
developed along the lines of the Schick test if some means can
be devised for removing the danger of the spore; a means of
specific immunization against measles, acute anterior polio-
374 JOHN A. KOLMER
myelitis, syphilis and gonorrhea and an improvement of our
means for active immunization against cholera, plague,dysentery
and typhoid fever, not to forget that problem of problems, namely,
the discovery of a means of specific immunization and treatment
for tuberculosis.
At out meeting last year the Association officially passed resolu-
tions offermg to our federal government the services of our
members and laboratories in the conduct of our great war; before
and since then not a few of our members have enlisted for active
duty in the federal service and at least one has given up his life
as a sacrifice to duty; many and probably all members of our
Association are more or less intimately associated in some work
having a direct bearing upon the problems of health and disease
and particularly those menacing or likely to menace the health
of our armies abroad and at home; to all the Association would
hold up in pardonable pride the accomplishments of the science
of immunity in the past and wish all God-speed in their work for
the present and future for the health and happiness of mankind
for all time and everywhere.
en heh: MODE OF* ACTION IN) VITRO AND THE
PREPARATION OF HEMOLYTIC
ANTIBODIES
A. K. BALLS anp JOHN H. KORNS
From the Department of Bacteriology, College of Physicians and Surgeons, Columbia
University, New York
Received for publication June 13, 1918
This work was undertaken with the idea, first, of studying the
mechanism of amboceptor action in vitro, and secondly, of as-
certaining, if possible. what part of the red blood cells is respon-
sible for their antigenic property. Inasmuch as our work has
now been brought to a necessary close, we are bringing together
in this article our findings even though our conclusions at this
stage cannot be far-reaching.
Many of the general principles applying to the mode of action
of hemolytic antibodies have been worked out by Muir (1) who
showed that the combination of amboceptor and cells is a weak one,
easily dissociable, and occurs according to a law apparently re-
sembling the law of mass action. He showed, further, that
amboceptor is not destroyed by hemolysis but remains bound to
the receptors of the hemolyzed cells from which it can dissociate
in the same manner as from the whole cells, such dissociation
being more marked at incubator temperature.
By analogy with the law of mass action, if the amount of un-
hemolyzed cells is very large in proportion to the amount of
hemolyzed receptors the dissociation from the hemolyzed por-
tion will be practically complete, and almost the entire amount
of amboceptor at any one time will be resident on the unhemo-
lyzed cells. This was found to be so in experiments of which the
following is an example.
A relatively large quantity (1 ec.) of washed sheep cells, con-
centrated by rapid centrifugation for ten minutes, was diluted
275
376 A. K. BALLS AND JOHN H. KORNS
to 40 ec., and 80 units! of amboceptor added, followed by three
units of complement. After incubating for thirty minutes a
hemolysis of about 10 per cent of the cells had occurred. The
liquid was centrifuged again at the same speed and for the same
period as before, the volume of concentrated cells was measured,
and the sediment was then made up to a 5 per cent suspension.
A quantity of untreated cells was then taken, equal to the amount
of cells remaining after the partial hemolysis just described, and
to this 80 units of amboceptor and enough salt solution to secure
a 5 per cent suspension were added. The amount of complement
necessary completely to hemolyze 0.5 ec. of each of these two sus-
pensions was determined and was found to be the same. We thus
concluded that both samples of cells contained the same amount
of amboceptor, showing first, that under these conditions, dis-
sociation, as might be expected from chemical reasoning, is negli-
gible from the unhemolyzed cells, and secondly, that amboceptor
is not only set free during hemolysis, but is quantitatively
unaltered. The results are noted in table 1.
TABLE 1
‘
Showing the action of amboceptor to be “ progressive ”’
COMPLEMENT 1:10
0.10 ce. 011 ce. 0.12 ce. 0.13 ce. 0.14 ee. 0
iireatedsecellsce...e2s- ose. +++ )+-++ sie +4+4+4+/4+4+4++4+ =
Wntreatedscells' sss. c0 es +++ |)+++ siete ++++)++++ -
++-+-+ indicates complete hemolysis.
-— Indicates no hemolysis.
1 Throughout this paper the ‘‘unit’’ of amboceptor is taken as the smallest
amount of inactivated immune rabbit serum which hemolyzed completely 0.5 ce.
of 5 per cent red cell suspension, with the addition of 0.05 ce. of complement, after
one hour’s incubation at 37°C. and in a total volume of 2.5 cc. The “‘unit’’ of
complement is the smallest amount of guinea pig serum which completely hemo-
lyzes in one hour at 37°C. 0.5 ec. of 5 per cent red cell suspension in the presence
of two (2) units of amboceptor, in a total volume of 2.5 cc. The 5 per cent red
cell suspension is prepared by centrifuging washed sheep cells rapidly at a definite
speed for ten minutes and then bringing up the measured sediment to the proper
dilution with 0.85 per cent NaCl solution.
PREPARATION OF HEMOLYTIC ANTIBODIES 377
- Should the amount of hemolyzed cells be large or should sey-
eral washings with salt solution be made, the amboceptor on. the
cells so treated will be smaller in amount than on the control
series, showing a considerable loss by dissociation.
If the hemolyzed receptors are heavily loaded with amboceptor,
it is easy to conceive that this should in part dissociate, and undis-
solved red cells being still present, that the dissociated ambo-
ceptor should attach itself to these, causing their hemolysis.
The action of amboceptor would therefore not cease with the
hemolyzing of one cell, but would be continuous, and as the
amount of hemolyzed receptors increased and the amount of
unhemolyzed receptors decreased, the hemolyzed receptors
would become less saturated with immune body, and conse-
quently would split off less, thus causing the velocity of the reac-
tion to decrease, a phenomenon well recognized in hemolytic
work, for the last traces of unhemolyzed cells disappear very
slowly. That this is not chiefly due to the deterioration of
complement at incubation temperature can be shown by the
fact that the system still contains that component, usually in
considerable amounts. Furthermore, the velocity of deteriora-
tion of complement in the presence of amboceptor is practically
the same as in pure salt solution (as far as our rough immunologi-
cal methods will permit us to measure). The products of comple-
ment deterioration, so called ‘“‘ecomplementoid,”’ likewise do not
inhibit hemolysis when used in quantities comparable to the
amounts of complement ordinarily used in hemolytic work. On
the other hand, laked cells, whether dissolved by amboceptor and
complement, or by distiled water and subsequently made isotonic
are capable of greatly HOIST the reaction.
In table 2, A is guinea-pig’s serum heated to 56°C. for thirty
minutes and represents “complementoid.” B represents 100
per cent cells laked with distilled water, made isotonic, and diluted
with salt solution to ten times the original volume. Each tube
contains 0.5 cc. of 5 per cent cells, two units of amboceptor and
one unit of complement, the total volume being 2.5 ec. An addi-
tional control for A not tabulated and containing 1 ce. of Bee.
serum but no complement showed no hemolysis.
.
THE JOURNAL OF IMMUNOLOGY, VOL. II, NO. 5
378 A. K. BALLS AND JOHN H. KORNS
This observation under B of table 2 coincides with the work of
Muir (2), who showed that the receptors of the red cells are not
destroyed by hypotonicity. They are, therefore, capable of
binding amboceptor just as the whole cells do, and their addition
to the system will cause a diminution in the amount of ambo-
ceptor available for hemolytic purposes.
Bordet (3) found that the stroma of hemolyzed red cells is
capable of fixing antibody, and he was able, by injecting them
into an animal, to produce antibodies hemolytic for the whole
cells. Stewart (4) confirmed this but found the hemolytic prop-
TABLE 2
Showing the absence of inhibiting effect on hemolysis by heated guinea-pig serum,
and the presence of such effect by isotonic laked cell products
AMOUNTS OF A OR A ADDED TO HEMOLYTIC SYSTEM (cc.)
0.1 0.2 0.3 0.4 0.5 0.6
A. Heated serum 1-10...... a ae es On Gat ef | lI cL LL el
B. Laked cell products 10
DETRCON Greet txt eer =- + + + ae
AMOUNTS OF B OR B ADDED TO HEMOLYTIC SYSTEM (cc.)
0.7 0.8 0.9 00 0
A. Heated serum 1-10.............. ++++)/++4+4]++44+]/4+4+4+4+]44++
B. Laked cell products 10 per cent..| Tr. ar’ irs AU. igre
+ Signifies perceptible hemolysis.
Tr. a trace of hemolysis.
++-+4+ signifies complete hemolysis.
erty of the antiserum less marked than the agglutinative prop-
erty. By filtration of the laked cells through a Berkefeld filter,
Muir (5) was able to retain all of the stroma and with it practically
all of the receptors.
7
.
>
i.
PREPARATION OF HEMOLYTIC ANTIBODIES 379
pended in salt solution, it possesses remarkably active ambo-
ceptor-binding properties. As proof of this latter statement we
briefly review one experiment. To 1 cc. of diluted stroma, rep-
resenting not more than the equivalent of 4 cc. of 5 per cent
cells,s 8 units of amboceptor were added and salt solution up to
8 ec. This mixture was kept at room temperature for thirty
minutes and it was then centrifuged at high speed. To 2 ce. of
the supernatant fluid, 0.5 ec. of 5 per cent cells and 2 units of com-
plement were added. This mixture was incubated at 37°C. for
thirty minutes. There was no hemolysis. The mixture was
again centrifuged at high speed and to the supernatant fluid
0.5 ec. of sensitized cells were added. On incubating at 37°C.
complete hemolysis promptly occurred.
The amboceptor-binding properties of the stroma were found
by us to be partially destroyed by heating to 65°C. for thirty
minutes, markedly diminished at 70° and completely destroyed
at 80°. These results differ from Muir’s (6) in that he found
evidence of amboceptor-binding even after heating to 100° for
one hour. Our results point to the protein nature of the anti-
genic substance.
Dried stroma on being resuspended in salt solution act in the
same manner as the moist freshly-prepared material, while ox
cell stroma is without effect, showing the species specificity of
the reaction.
Rabbits were then injected intravenously with the suspensions
of stroma from sheep cells and it was observed that the produc-
tion of both hemolysins and agglutinins was marked. Since
the animals so treated receive very little protein in comparison
with those injected with whole cells, it was thought possible to
increase the quantity of injected material. Accordingly the
stroma of approximately 25 ec. of concentrated sheep cells, sus-
pended in about 10 ce. of salt solution, was injected each time,
and apparently with little if any ill effect. Table 3 shows the
details in the case of two of the rabbits.
3 The calculation was made on the assumption that all the stroma had been
secured by our process. We intentionally overestimate the amount.
K. BALLS AND JOHN H. KORNS
A.
380
‘S891 IO »
008- 009-T | 00&-1 O0Z-T | 006-1 00Z-T | OOT-T | OT-T | OI-T [0°07 °° ** "1097 oATYVUTQNISSV
OO8-T |OSZI-T |OOSE-T |000Z-T OOTI-1 |O0&E-1 OOEE-T |OOSE-T |xOOT-T pale cee ~ ie Geel "19}1} OTPATOWO FY
ye, 9 G ; p (2 Z T SORA IISI 00) tatey tan jo roquinN
QZ6T | O96T | OL8T | SPGT | O88T | OOST | OSGT | OGAT | O96T | OL6T J SUUvIS UT FUSION
: Cael
002-T 006-1 QOC=T OOS=T | 00G=1= |, O0T=E || OTSh) OTE 101} OATPVUTINIS
. 009-1 |OOOT-T |OO9T-T |000Z-T |OOSZ-T |OOOS-T jOOSE-T |OOOS-T |OOSZ-T jOOOT-T |x0OT-T yn Joy} o1yATOUO TY
g v & Gj T [ccc ttt uoroofur yo roquiny
OPST | OSFT | SZST | O&FT | OSET | SZET | OFVET | OCFT J SUIvIS UT J YSIOM
[pase Ke kSta
oe Avy | 22 Av | st Sew | er Auw | 6 dew | 9 = z= = ells Seale D0, ONO) ORO Oras CG (OL: I) A UM
= = = = =e = = = = = = = =a Stas Mallee raeser sages (Gi) touxe[y YITM
dnois A190} UesA(T
4
4
3 Se es alhces eect orl ears less ies steel aie ae (21:1) proydsy, WHIM
ni ies | ea me ome feat oe ee ice tects > ean) navy Cae ce
3 — |— |= |= J+ J++] ++) ft [44 | tt pr) & eed WM
- ps — ms, —-+ Dy olverie) \s) vice) le) buleule: (Oeme\e o°R@ ‘UIn..as a BIBI
3 Ste sales el ete et acter tee |) aoa atect [estes Aaa (21:1) proydsy, WIM
a —~ |-— |= |= |- |= | -4+] 44 +] 44 fOr) 2 eed WIM
i Se ae eect cp el neem iaeeugepcaaler |e os (OL?) V Mt WHIM
a = —-+ + CIO.O “YeOeOecM sys Ct o°@ ‘cWdIn.tes V BIB
ceeealccatee ee| eae iea ca ate l ciecte alictonty ai cleat (OL :T) & PBT YIM
SS a ee ee paras OL 1) ered aE
Se ie cee Neee (ree cis erat aae|| oasis elas GT: De PIOHT AS EM
= a = Seis as “o°e ‘undies proyds J,
a eae a se th es eee
dnois proyda y,
90 | v0 | £0 | 20 rol 500061000200 0|ez000| e100 | no 0 | 900'0 c10'°0 | e200 | ¢0°0
[o-@) LSaL ‘0O°V ‘NADIINV AGOL udd WOUdAS AO SUALAWIINGD 018090
penn era eee Scr a ee ee ee ath os ee ee
st
suoyniy wabyuy
T HIAViL
“qsoq Arejuoue|duI00I}UB OY} UL paesn UOTN{Ip usesTjuB Jo syunomy |
419
PREPARING BACTERIAL ANTIGENS
|
|
— |- |-+| +
— |-+] + | ++
— |-+] + | ++
— |+ |-++} ++
— |-+] + | ++
dnoid snovo003urueyy
‘pasn uorlsuedsns uss1yUe JO UOTINIIpP SozvoTpUy ,
OFZOT-T| OCTS- 1) O9SS:T| O8ZT: 1) OFD=T | OZE:T | OOT*T 08:1
pee
STE aro eee cor aati
ae er eee? STR
Seiten: egeeretie sa tial
Be fer Sha reece eager tr
FEE rata eae et
i ee en ee ae
fearalca eee (08:1) Td WM
sara eee ee (4:1) Id WHIM
Erg Bee traces ee re Sane een
ee ||| i oe ete verte ohegp hese Neairs se ‘cIn..es T ‘ouInoUg
Pee Rogers ete. acai
ord eae comets acon
dina Parle ten Maat ey,
Td ghee ae ae cag eae
SP (ae ROO OT) Pa a 2) ‘uIMn.1es O9IN
rahe erase cS aoe ae ie
TO eee seme TGeanehl
eat Sees ee ere sen
Hed. ae eceen tenaa ve oueah
es [leeches oiteices oetelre tebe la elem o°@ ‘crn..es Of IN
“geutal| walle te ake Tecan
fete oe eran ae Saracen
ra, Were Mr awe erect
Talis amochret cae? ee accai
| Oras Learn Ore On o°@ ‘canes OLIN
are me sacs (ST :1) O9IN WHIM
++ fo OL ET) Os WEA
+H fo (ST ST) OLIN WHT
ee ee
Se). Gh Rian Pen 9 te Ce ear over ae o°R ‘cOn.1es IW
06:1 suoK NIC, UNseg
420 JAMES C. SMALL
antigen was in this way determined. With this serum dilution
as a constant in the next titration, the smallest amount of antigen
giving complete fixation in this serum dilution was determined.
This amount doubled became the antigen unit in subsequent
comparative serum titrations.
The above tables show the results obtained with these anti-
gen preparations in direct and cross titrations of immune rabbit
sera.
TECHNIQUE OF SERUM TITRATIONS
Here, as in determining the antigen unit, the anti-goat hemo-
lytic system was used. The technique followed is more or less
standard and will not be discussed in detail. The immune sera
used were shown to contain natural amboceptor for goat’s cells
so that it became necessary to remove this. Sensitized goat
cells were used throughout the titrations. These were prepared
by adding the previously determined amboceptor unit to a 5
per cent suspension of cells and incubating the mixture at 37.5°C.
for one hour. After the cells were thrown down by centrifugali-
zation the supernatant liquid was drawn off and replaced with an
equal amount of fresh salt solution in order to resuspend the
cells. The complement used was a 40 per cent solution of pooled
guinea-pigs’ sera.
In order to obtain the serum dilutions as represented in the
tables, a series of tubes was set up with | cc. of normal salt solu-
tion in each tube except the first, which contained 1.9cc. To the
first tube 0.1 cc. of immune serum was added. After being
thoroughly mixed 1 cc. of the dilution was carried across to the
second tube and this process was continued down through the
series to the last tube where 1 cc. of the dilution was discarded.
Serum, complement, and antigen were added in order and the
tubes were incubated for a half hour at 37.5°C. oD Sa Nn
10 c c c c Ww c c c tr c ce c
0.8 c c c c tr c c ce — | vst c c
. 0.6 c c c c — c c 9) - Ati a avst iene
0.4 ale c c c — c c Cc = tr st c
On st c c c a tr c c — w | vst
0.1 w | ale c c — -- w | ale |] — _ tr st
0.08 tr | ale c c — — st | vst | — — (He |) yyy
0.06 — | ale c c — -- tr st — — _ tr
0.04 — | vst c c - — _ Ww — = — —
9.02 _ st | vst c — _ _ tr _ _ --
TABLE 3
.
von
tonic saccharose solut
.
Uso
nce of different salts on saponin hemolysis in a medium of
>
2
Showing the influc
PSPS PEL CRU RTE is AP aie cme:
TOR
mro[ ddd tri ti rr rss
coat’ | (28 O12) 2) ©. B98 es Sasa
pooET ||P)? 948 9 a EERE LE et
ale
ale
ale
st
st
st
WwW
WwW
‘Convaxk | on 00 9 0 0
nan | OR By rere ae
OO! Ss mH fh
rostn| © 10S © 2 ead pote
edilinaed
Hw 9 — »
ostCHN) | +2 Mita: oR a ESE
coo o oo OL ae &
FOSteN Fa eee a ial
anne ees i
oHonn | (FS 5 me gi 2° Si eseee i el
cicye icy =
S10008H | tae hoa Metacare uae GW el |
Sy i ER 4 F
ex000'HD | OOF QNOROR Oe RO sae ese
Oe
greet Sto etree ea | | al
OINLIO BN QS SOS Soar ||| | | ie |
(>) mm
‘OcHM SRS) S80) OTC Oe ee eee
= ° = Fa
Xosy | O} OF O50 10; 0 9 OL ie Eas
J a
SON 2g EN SENS Si) Spo 3 SS Sh ‘ical
5 Oo +> w
2ONtN Ol BO OROFO Ose oe ine
s coo oe Oo OL aH et
7ONM ep a eS eal
‘ © © 45 Pa
ZONtN ON SS. oS St ie ern i of
> > OF =>) B
IM 810) OO. 8 OF 91010 Se ee ea
i
igey OOOO) oO TONON Oa = ipl
OS > ‘
Eee Rl ECR CR Nr 2
oH tH
IOM OPS KONO! OF O10 10. aaa
coo ooo OL aH ae
[O8N BAP eaPRFese |
anon
SLTVS SGADHNDMAOAANHODCOCSO
#0 |NOMLOIOS "TVIEHON 9/0) 9 ees lh
A sO
ASOUVHOOVS AO } SI S0''O0) ON CN HCO 9B. G22 S31 C3
NOILAIOS LNAO udd g Sse oe SS oe
00¢‘29:T ASOUVHOOVS Serna eS ea Ss) Ss Soo) SOS oes Ss Ss“
4O NOILQIOS INGO Sy Ns SNARE oe aT 5 RIEL ae
udd g NI NINOdYsSoDar Ty tae’
ws
TABLE 4
TORE
novo | | alma sl aE he ale al
eee
Se |
S ‘oom | FP SO ee es SS tt
‘~
S
|
= woman | FEET Prt rit b it ti
iva)
go mH he
2 Tan | SSS ie Pete) It I
=)
Ss oovovoD Oo De Fah te? bette
> ‘OS3W PaeFe eee S|
©
Ss o » |
S 'OS*CHN) | = 2) 2) hn eee
3
. >} 1S) HY
= Oa CRO TO OOM Orome = OF
= ose | aia ee |
oo oO :
TTT CAO Oo LOMO oon ~*~ Fat
§ onor™ | atria ee |
“>
a ooo0van 00S ZRH HOH
= MOOO®HO a eq mae 2: ||
= 5 : ooo Sfolone ooo sss be
& NOOODPHO aoa a8 ==
2 OAC Same et
S mma || © 8b Soe SiS EES pelt |
=
= OSS SIS tS > ft
s O1ULIO BN aie ile lel
<
SIS) a
& rattan | 02" PSe Poe Le Ola eae
=
S Ow mH OF
= Nagst || C7 1S Pee Os Sheela a eee eke
ny
= Op
S roNE PS Ses Sree ease LP le ile
”
RS Saat
5 OheN|) ooo e Soe eae ON at ane
SS o oo oO oO © oO 4 : ee
= ‘ONS aBPEBBSS
=
aS fONPN |e 'e eee vere lela,
> Sree iy Gr Oo SS 2 >
S$
oo i
2 IgM OO EN ORS ita Bahia eas
>
s OO!) OL OMOon Omens sa
© Igen aaa ai -sarll yy |
em
Om, ra fH
= IOM 909 99 910 So pe Pe
3
= mH
iS ION SOO ONS Oo (OA OwOGr Ri Wh IE |
R Bras
aonta
ishran As SCmMaOHtNO lo ES i — a — a)
AOVNOWMATOUTVNON | eats et ere SS
Nt oo
aso001D 40 pNHOWDONAHOHODRAAAASG
NOILOTIOS ING) Wad 9G Ocqoooo nn Hse A etnA AA
000‘0¢ :1 #809015 40 Oo ©. So oro) So orosere SSS
NOILOIOS INGO uad SAT aye ape : Menges Pct aS
GoM trGsySooNr: |e nS ek ee ee Te
427
THE JOURNAL OF IMMUNOLOGY, VOL, Il, NO. 5
TABLE 5
Showing the influence of different salts on sapotoxin hemolysis in a medium of isotonic saccharose solution
add g NI NIXOLOdyvs
428
108d Ge teat ede ie Hb ite ell) Abe ok
007M D9) OO: COO OP OTONOROm ermine
£OO78N o.G 6 8 Oo. wo oo ose eum
Sie rete la reyevce
TOS | > b> ae H HS | I eam ab
Sareea
7(ON)SH | 2 PSS Sasa we Eee
See a Ae
rOStN |) 2) Sire SNe Bess a3 0) 1
re ra ee 74
*OS*(FH.N) Pome ae ee [| {||
SF :
costs | COCO SC CeCe eon a Ee YT |
EO Om a> od
smn | 9° OS SER ERE SY eae |
es?
ene | EE OREN SS
= om + |
omoon| O° SSR GEESE III I
© 45 =e
mmm | 9889 2.2 8 oo a Fee
¢ Sy mom Om
soomno| COS SSEE EEE ET II
enooxHo| coo oo REEERI III I
musy| oo ooo ooo OO REE! |
ume | 2 °° © 8 29 608 Eee
nosxy| coco ooR8REBEEI II
Ram ee rey ee
onn| POC SSOKERESEEE II II
Gat e
one | ©.° °° Pa rire Ehime
Sow a |
onn| CC OCS RE SESS I 111
r=)
Sony mw
consen| °° SCR BREE E111 1S
¢ ‘ Onags 4
m| Po fe SS ears (iam
~~
So D7 a
ion | °° 9 oe Eee | ee
Che Saar
mon | S99 OSB RSS 1 ee
—_—
nm ~~ wo
in| eyo oe ee SS i i
~~
“ ©} (0) 1O MOMS) 12) ie {a
TORN a > @ > |
oot aA
xo ieamacdao ipa eee WE Sea eS
SOS a St et oS Oo SO Sor oaames
at Oo
asouvmoovsgo | | NHODONTOHRAAASS
NOILQIOS INGO Hdd Q oooocornn nin wAwA nnn
0S2'9:1 ASOUVHOOVS Soo CO Sooo oO; OS SO Clore
40 NOILQIO0S INGO RiEia ie ee SORA Oren e
Se ee oe oe
TABLE 6
« ‘J
uence of di
infl
Showing the
~
S » 1
‘OdERN RE NS SF |
oOo ~~
oo 0 8
‘aH | PS SS sel al |
—
hay AI ES SL a ee
: rgonE | CRON pre Sree Sia e
COON eovv09s ooo FD OO One
Dw Gn SUP eee 4) Ot OER) SP BE oe Sn
won| PPR BEES | hs oe ee
aera oo ran
(CONDI | SUS VENOUS OV Rs eee esse |
fferent salts on sapotoxin hemolysis in a medium of isotonic glucose so!ution
oor Spam hoo
rosa | eisesecanenes (i a ie lea
FANQ2(F Hepp moots t
OS*(FHN) o Em > 2 a nea
2 Sse ee Ee
tOS*tN ee i a ei =) (lene TA
tOstEN | SON OO 9 SO Ol OLonoroemin il s
op wow
soooHo | °° °° OZ REESE | eal < (PER
ex0008H | SRS aoe lee bal
OIULID Y
Coo gle ge a acl |
ad o CU Hy 1 fas
poet | “°° SSS SS. Te
4 3) 42
tony | a eis ce ee eal) alive
uh fe
a
tONSN | BONS at ees cea n Mle ly ip alata ail
5 a) fet
con | °° RS eS vie ui ele ie slieal
< ~~ ~ me
onn| 2 O@FEBI LIT Vi ill
Rae =
1 | SAP Perse ee inh cy bya elt
> Sr = B
game Oe SENS ae a a Wel ety ul
oOo oO Se
Jq@N = Pe ere wr lea aie |
Sw |
om pie? tts seers Fea can
Sor moO
on|oeger SSS 511111 |
DOoDHA
Deen ST See eee Se
e 5 a Se eS CLS SS SuS Srers
N st Oo CO
aSo0021D Ao [OR Sty SOOO ECAH ASCO RCS 12 Se Ce
NOILI TOS INA) ¥ad 9 ¢ oOo © COs Str rt et oe
a te hrs! dc ah i ee
000‘0T:7] as09aT9 oooco sec © Co Soyo toto oC CFO
HOUNO MD TOS SND) aS at aot eter tt
udd 9 ¢ NI NIXOLOdVS
429
430 TANEMOTO FURUHATA
seem to be indifferent. Ammonium sulphate is peculiar in its
action inasmuch as in the larger quantities it alters and pre-
cipitates the red blood cells and at the same time it inhibits the
hemolysis by the saponin and sapotoxin. Furthermore, we see
that salts of the same cation with different anions exert a differ-
ent influence on saponin and sapotoxin hemolysis and that the
same is true of the different cation salts with the same anion.
This fact is interesting in view of the experiments of Hdéber (6)
and also of Port, who pointed out differences between anion and
eation in the influence of salts upon hemolysis.
From the standpoint of the tonus of the medium, since the
sugar solutions were isotonic, it is also important to study saponin
hemolysis in isotonic solutions of the salts instead of in normal
solutions. I, therefore, prepared the following isotonic solutions:
per cent
IN a NOB se ass ow diss weave savas etelw Bare ievese wtonhe ale later ae aide eyesore tere oias eres eae 1.30
|S Un eee rep en emma hater cain o 2.35
INa2S OS eek aoe cle oh crate ca ie ormeeve Bieistouel Sieieekaie Glepecaeiete nakest Lore totetete eto 1.63
RB GIs cis ecco espe ake ee tases Oasis e Balestier oan gene Caer eee eet 1.85
IN AIBEE :o: ass sase toc. Sree is Que tonera ths mers ois eels austere actors aie ee eee rotor 1.58
NaSON« 5 ceeds ese nos eines See ese ooo Ooe eh ea Dee EE eee 1.24
TE ki sig aces hie ete Ore: ein Discs Siecae ae wc wee eos ae Obs OF Cee een 1.14
In the medium of these solutions the saponin hemolysis was
tested as in the previous experiment, and it was found that in
all of them saponin can act more rapidly than in sugar solution.
According to experiments of Poyarkoff (7), the action of sperma-
toxin, studied in an electrolyte and a nonelectrolyte medium is
much influenced by the viscosity of the environment. Poyarkoff
found that when the viscosity is very much increased or de-
creased by suitable mixtures of electrolyte and nonelectrolyte,
the action of the spermatoxin, which was measured by the
time taken to cause the motion of the spermatozoa to cease, is
very much reduced. There is, accordingly, an optimal concen-
tration of both substances. I have investigated the question
whether this fact is also applicable to the action of saponin or
sapotoxin. In this experiment the red blood cells of the rabbit
were used. The results of the experiment are shown in table 7.
es it ae
A STUDY OF SAPONIN HEMOLYSIS 431
In table 7 the absolute quantity of saponin is the same in each
mixture the variations affecting only the quantity of glucose and
of salt. An increase in the quantity of sugar causes an increase
of the viscosity of the medium. We see thus that as the viscosity
of the medium increases there is an increased inhibition on of
the hemolysis by saponin. I have not been able, however, to
find any optimal viscosity for the saponin hemolysis such as
that found by Poyarkoff for spermatoxin. The hemolytic power
of saponin appears to be inversely proportional to the viscosity
of the medium. The increase of viscosity may interfere with
the velocity of diffusion of saponin into the red blood cells.
TABLE 7
bAZA La eae baz? &
Deo oe ae aS ees a HEMOLYSIS (1 HOUR OF 37°C.)
BARE Zan 12 BAD 4
BOSD Pa a EVOrS i
2 oe ZAz Zoa A708 By
Q i) =< 0 ° a m iD
ofaga| zee zen | °2ad 1 9 4 4 bel 4
aeons ZOR maR BASDe BD S S = S
Agana ons ong Aaaae BA = = = =
GAMA w Oa hoa MAMQOn yo = Ss S oS
19 n n Ye) on a oo a wD
1.0 1.0 0 0 110) c c e c
10 0.8 0-2 0 9:1 c ce c ale
1.0 0.6 0.4 0 8:12 c c ec ale
1.0 0.4 0.6 0 US 33 c c ale st
1.0 0.2 0.8 0 6:4 c Cc vst tr
10 0 120 0 IS 5) c c st --
0 0.8 0.2 1.0 4:6 c c Ww —
0 0.6 0.4 © One c c UE --
0 0.4 0.6 1.0 2:8 c ale —_ —
0 On2 ls (ORS 1.0 13:9 c ale _
I have also investigated the question whether the relations
described above with respect to rabbit’s blood hold as well for the
blood of other animals.
The results of these experiments are shown in tables 8, 9, 10 and
Pt
With red blood cells of the pigeon and of the sparrow hemolysis
could not be produced with even a 1: 500 solution of jegosaponin.
In these experiments, with all of the species of blood corpuscles
that are susceptible to the hemolytic action of saponin and
sapotoxin the adjuvant effect of electrolytes is apparent. We see,
432 TANEMOTO FURUHATA
furthermore, differences in the resistance of the different species
of corpuscles to saponin. Such differences have been noted by
various workers. Thus, according to Meyer (8), the order of
increasing resistance in corpuscles of the different species is:
TABLE 8
Saponin and sapotoxin hemolysis with the red blood cells of the horse
HEMOLYSIS (AT 37°C)
Jegosaponin Sapotoxin
1: 5000
Salt medium | Saccharose medium Salt medium Saccharose medium
30 minutes; lhour |30minutes}) 1 hour (30 minutes}; lhour (30 minutes} 1 hour
ey) c c c c c c tr tr
0.8 c c c ec c c — =
0.6 © e c ce c c _ _
0.4 c c c c c c = =
0.2 E c c zac tr Ww — —
0.1 c c c c = tr _ _
0.08 c c c c _ tr — _
0.06 Cc c st c - tr _ —
0.04 c & tr st = — _— =
0.02 w c _ — — = _ =
TABLE 9
Saponin and sapotoxin hemolysis with the red blood cells of the pig
(HEMOLYSIS (AT 37°C.)
Jegosaponin Sapotoxin
1: 5000
Salt medium Saccharose medium Salt medium Saccharose medium
30 ey lhour /30 minutes| lhour |30minutes) 1lhour |30minutes) 1 hour
120 c ec ce c c c c c
0.8 c c c C c c c c
0.6 en 5 c ce ec Cc c c c
0.4 c ec ce ce c c st st
OR? c c c ce st vst tr tr
0.1 c c c ce Ww st = =
0.08 c ce c ec Ww Ww = =
0.06 c c st vst tr Ww = =
0.04 ale c _ - tr tr =
0.02 st ale — — {; —
A STUDY OF SAPONIN HEMOLYSIS 433
horse < rabbit < pig < dog < sheep < ox. Meyer states,
also, that this represents the order of the lecithin-cholesterin
coefficients in the red blood cells of those species. However,
according to Abderhalden, the order of the lecithin-cholesterin
TABLE 10
Saponin and sapotoxin hemolysis with the red blood cells of the sheep.
HEMOLYSIS (AT 37°C.)
Jegosaponin Sapotoxin
1: 5000
“3 Salt medium Saccharose medium Salt medium u Saccharose medium
30 minutes) 1 hour (|30minutes| lhour /30 minutes; 1 hour |30 minutes 1 hour
IO c c c c Cc @ c c
0.8 ec c c e c © c Cc
0.6 c c Cc c ale c tr. Ww
0.4 c c c c st c —- _
On? c é c c Ww ale — —
0.1 c c c Cc tr WwW — -
0.08 c c ale c — tr _ =
0.06 c c tr Ww — — — =
0.04 tr WwW — _ - — —_ —
0.02 tr tr _ _ | — - -— _
TABLE 11
Saponin hemolysis wiih the red blood cells of the guinea-pig
HEMOLYSIS (AT 37°C.)
Jegosaponin
1: 5000
Salt medium Saccharose medium
30 minutes 1 hour 30 minutes 1 hour
1.0 c c c c
0.8 c @. c c
0.6 c c c c
0.4 c c c c
0.2 c c ale ec
0.1 c e st ale
0.08 ale ye Ae tr st
0.06 vst c = tr
0.04 st c = =
o
=)
i)
|
ee
S|
|
|
434 TANEMOTO FURUHATA
coefficients in the corpuscles is: horse < pig < rabbit < dog <
sheep < ox.
The results obtained by other*workers with respect to the
resistance to the saponin hemolysis are as follows:
Rywosch (10): Guinea-pig < rabbit < dog < pig < cat < ox
< goat < sheep.
Schauzenbach (11): Guinea-pig < man < horse < pig < ox
< goat < sheep.
Port (4): Rabbit < man < dog < pig < ox < sheep.
According to my own experiments the order of resistance of
the red blood cells against jegosaponin is as follows: Horse <
guinea-pig < rabbit < pig < sheep < pigeon.
SUMMARY
1. The hemolytic action of saponin or sapotoxin is, to 2 certain
extent, inhibited in a nonelectrolyte medium. This phenomenon
is, perhaps, attributable to the increase of viscosity of the medium,
which makes the diffusion of saponin into red blood cells the
more difficult.
2. Tons of various salts favor saponin hemolysis even in higher
concentration, except (NH4):SO., which can alter the red blood
cells in higher concentration. BaCl, and CaCl, are indifferent.
3. The resistance of red blood cells against saponin is different
in the different species of animal.
I desire to express my indebtedness to Professor S. Mita for
his kind direction and encouragement during my experiments.
REFERENCES
(1) E1rsurerR: -Zs. f. Imm., 1909, 2, p. 159.
(2) Ransom: Deutsche med. Woch., 1901, no. 3.
(3) PorGEs AND NEUBAUER: Biochem. Zs., 1907, 7, 152.
(4) Port: Deutsch. Arch. klin. Med., 99, 259.
(5) Asantna AnD Momoya: Mitt. a. d. pharmazeut. Inst. d. Univers. Tokio, 395.
(6) H6sErR: Biochem. Zs. 1908, 14, 209.
(7) PoyarKorr: Compt. Rend. d. 1. Soe. Biolog., 1916, 79, 1150.
(8) Meyer: Beitr. z. Chem. u. Physiol., 1908, 11.
(9) ABDERHALDEN: Zs. f. physiol. Chemie, 1898, 25.
(10) Rywoscu: Pfliiger’s Archiv, 1907, 116.
(11) ScHAvuzENBACH: See work of Rywosch.
ACTIVE IMMUNITY IN EXPERIMENTAL
POLIOMYELITIS!
H. L. ABRAMSON ann HERMAN GERBER
From the Bureau of Laboratories, Department of Health, New York
Received for publication March 29, 1918
The search for a method of specific inoculation against acute
poliomyelitis is not a new one. Shortly after the announcement
by Landsteiner and Popper (1), in 1909, of the successful trans-
mission of this disease to members of the monkey family, work-
ers in this country and Europe began to study the immunity
problems of this disease.
Flexner and Lewis (2), in 1910, demonstrated that monkeys
could be immunized against poliomyelitis by repeated subcu-
taneous injections with increasing amounts of a saline suspension
of the crude unmodified virus. They injected animals over a
period of two and one-half months. About ten days after the
completion of the course of inoculations, the animals were in-
jected intracerebrally with 2 cc. of a filtrate of a very potent
virus, of which 0.05 to 0.1 ce. would prove fatal. The animal
tested presented no sign of infection, whereas the control died
of poliomyelitis. In a later communication (3), they state that
artificial active immunity either by the injection of a single large
dose or by series of increasing small doses over a period of time
is not uniformly successful. In the former method, some of the
animals would develop poliomyelitis as result of the subcutaneous
injection, and in both, some animals: so inoculated would not
resist the test intracerebral inoculations of rather large doses of
a highly potent virus. In the blood serum of animals so im-
munized, the presence of neutralizing principles for the potent
virus was demonstrated in good concentration.
‘Read before the American Association of Immunologists at Philadelphia,
March 29, 1918.
435
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 6
436 H. L. ABRAMSON AND HERMAN GERBER
Landsteiner and Levaditi (4) attempted to devise a method for
the prevention of poliomyelitis analogous to the Pasteur method
for the prevention of rabies. They dried cords for as long a
period as twenty-four days. However, while some animals so
treated developed immunity without any ill effects as a result of
the treatment, other animals similarly treated developed the
clinical picture of experimental poliomyelitis.
Romer and Joseph (5) thought they had produced immunity
in monkeys by the intracerebral injection of a mixture of virus
and serum that contained neutralizing substances for the virus.
They found that a monkey so inoculated resisted apparently a
subsequent suitable intracerebral injection of straight virus.
However, that this is not invariably true is evidenced by the
experience of Flexner and Lewis and Landsteiner and Levaditi,
who had no difficulty in infecting animals that had been previ-
ously intracerebrally injected with a neutralized mixture of
serum and virus. In fact, Flexner and Lewis have had no dif-
ficulty in infecting an animal that had previously resisted a suf-
ficient intracerebral dose of straight highly potent virus.
We can confirm these findings concerning neutralized and
straight virus, and further than that we have re-infected, by
suitable intracerebral inoculations, two animals that had been
paralyzed over a year ago, and which presented residual palsies
at the time of injection. These animals succumbed as promptly
as control animals, which eceived a third of the dose. This
fact will be pointed out later.
Flexner (6) stated in 1910 that while the results in artificial
active immunity thus far achieved were encouraging, our knowl-
edge at that time was not sufficient to render those results of
practical value. This statement was the stimulus for the work
which has engrossed our attention for the past year and a half.
The necessary requisites of any suitable method of artificial
immunization are:
1. That the method shall protect against any reasonable ex-
posure to the disease for which the individual is immunized.
2. That the inoculations in themselves shall be absolutely
harmless.
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 437
3. That the method itself shall not be too cumbersome or
prolonged, so as to render the production of immunity tco slow
to be of practical value.
These have been the conditions with which we have attempted
to comply in our efforts to arrive at a suitable method.
After an extended and rather discouraging experience with
young rabbits, we turned our attention to the use of monkeys of
the rhesus variety in our efforts to devise a practical method.
The striking similarity in the characteristics of the virus of
rabies and of that of poliomyelitis impressed us so as to cause us
to work with the idea of attenuating a highly potent Rocke-
feller strain. Attempts were made to this end with two methods,
one with chemical means and the other by subjecting the virus
to the action of heat.
MODIFICATION OF POLIO VIRUS BY CONTACT WITH 0.5 PER CENT
FORMALDEHYDE
Cummings devised a method of anti-rabic treatment in which
he put a 2 per cent emulsion of fixed rabic virus in contact with
0.5 per cent formaldehyde for four hours in the ice-box. At the
end of this time, he dialyzed the formalin from the mixture
through collodion sacs into distilled water until the cord emulsion
failed to give test for formalin. The material was then inocu-
lated daily into the rabbits to be protected in increasing doses.
We applied this method in our attempt to chemically modify
polio virus. We used a 10 per cent emulsion of the cords and
brains of monkeys dead of highly virulent poliomyelitis virus and
made it up fresh for each injection. This 10 per cent emulsion
was kept in contact with 0.5 per cent formaldehyde for four
hours. It was assumed that this contact would kill the polio
virus. However, our experience proved to us without any chance
for doubt, that it did not kill the virus. The protocol follows:
Experiment. Macacus rhesi nos. 89 and 90 were injected subcuta-
neously as follows:
438 H. L. ABRAMSON AND HERMAN GERBER
March 28, 1917, 10 cc.
March 30, 1917, 15 cc.
Aprilt 2 O17, 5 ce:
April, 5, 1917, 5" ee,
On April 7, 1917, no. 90 appeared ill; refused food; seemed to be
tremulous; showed no sign of paralysis.
April 8, 1917, no. 90 died during the night. Postmortem examina-
tion showed nothing of note in the abdominal or thoracic viscera.
Section of cord showed some swelling and reddening of the gray matter.
Microscopic examination of the cord showed slight perivascular cell
infiltration; moderate diffuse cell infiltration; considerable nerve
cell degeneration and neurophagocytosis; only slight changes in the
meninges; that is, all the classical pathologic findings of poliomyelitis.
Macacus rhesus 89 remained well until April 26, 1917, when it ap-
peared to be ill. It would not feed well but it showed no paresis of
any kind. The animal died during the night. Microscopic section
of the cord showed the lesins of poliomyelitis.
These animals presented a rather unusual type of infection,
one dying within ten days and the other within one month after
the institution of the subcutaneous injections. The type re-
sembles very much that described by Flexner as the ‘“‘marantic”’
type of monkey poliomyelitis. In this type, animals may be
sick over a longer period of time than the animals in the above
experiment, yet they do not present flaccid paralysis.
Our results with this particular method were decidedly not
encouraging, and as monkeys were very scarce, we turned our
attention to the possibility of attenuating the virus by heat.
MODIFICATION OF VIRUS BY HEAT
It has been demonstrated that poliomyelitis virus is rendered
inert by exposing it to a temperature of 50° to 55°C. for one-half
hour. With this fact as a basis, it was decided to expose a highly
potent virus obtained from the laboratories of the Rockefeller
Institute to heat in two ways. One method aims at a graded
increase in the virulence of the material inoculated, paralleling
aiter a fashion the use of increasingly more virulent cords in the-
production of immunity to rabies. This method consists of the
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 439
subcutaneous injection on four successive days of 5 ce. of a 10
per cent emulsion in saline of brains and cords of monkeys that
have recently been paralyzed by intracerebral injections of highly
virulent virus, heated as follows:
_PTCATIR CLES 28 22 Sa mi > ae ae PT 55°C. for one-half hour
PASCLCLE CLR Ver vyo cla ahs. nits eittcape ie isiaicietanes Pelco 55°C. for one-half hour
PACING As ers see bends eh MN tee 45°C. for one-half hour
BURGER CAVE Nete's << suisis eee a> ee bidlasle cape aentee 37°C. for one-half hour
On the fifth day, the animals received 5 cc. of 10 per cent emul-
sion of the virus unmodified by heat.
The other method of injection used consists in the subcutane-
ous injection on ten successive days of 5 ec. of 10 per cent emulsion
in saline of brains and cords from monkeys recently paralyzed,
which had been heated uniformly to 55°C. for one-half hour.’
This method corresponds to the use by Semple of subcutaneous
injection of killed rabic virus in the Pasteur treatment. Such
a method as this immediately removes the dangerous possibility
of producing poliomyelitis in the treated animal, inasmuch as
the heated material is unable to infect a monkey when injected
in large amounts by the intracerebral route. In other words,
the virus is apparently killed.
In testing the protective value of these two methods of vac-
cination, we wished to determine two facts:
1. Are the animals so treated capable of resisting a multiple
lethal dose of the virus intracerebrally?
2. Does the blood of animals so treated contain neutralizing
principles for a highly potent polio virus?
Three weeks after the completion of the series of subcutaneous
injections, three animals of each of these series were bled from
the heart. The serum obtained from these bleedings were put
into contact with a 5 per cent emulsion of highly potent Rocke-
feller virus in the proportion of one to one for two hours at 37°C.
and twenty-two hours in the ice-box. Then 0.6 ce. of each mix-
ture were inoculated intracerebrally into each of six normal ani-
mals. At the time of these injections, the vaccinated animals
were tested with the intracerebral injection of 0.15 to 0.3 cc. of
440 H. L. ABRAMSON AND HERMAN GERBER
the same virus emulsion used in the neutralization test. Control
animals were also inoculated intracerebrally with the same amount
of the same emulsions used in testing the vaccinated animals and
in the neutralization tests.
Charts 1, 2 and 3 give in detail the results obtained by these
methods of modifying the virus. It will be noted that Macacus
rhesus 81 of the 5-injection series had received a similar series of
injections, and that Macacus rhesus 80 of the 10-injection series
had received 5 injections of killed virus two and one-half months
previously. At this time we were unable to obtain new monkeys
for controlling our test inoculations and for our neutralizing ex-
periments. We thought that if immunity in poliomyelitis lasts
no longer than that of rabies in rabbits, it would be well to sub-
ject them to another series of injections.
Chart -1 shows that of five animals treated by the 5-injection
method, three survived the test intracerebral injection of 0.3 ce.
of an emulsion of a highly potent virus, 0.05 to 0.1 cc of the fil-
trate of which is fatal tomonkeys. The two animals, nos. 91 and
97, that succumbed to the injection exhibited paralysis on the
sixth day after the test injection. The normal control animals
nos. 19 and 23 showed paralysis on the fourth and fifth day
repectively.
Macacus rhesus 24, an animal that had survived an infection
from virus of the 1916 epidemic, about one year ago, and which
at this time presented a residual diplegia, was also injected in-
tracerebrally with 1 cc. of the same emulsion. This animal ex-
hibited paralysis of the arms six days after the inoculation and
died on the seventh day. Animals that have survived an attack
of the experimental disease are supposed to have a very high
degree of immunity, yet here was such an animal that suecumbed
almost as readily as a control to only three and one-third times
the dose used in the controls. This result argues well for the
virulence of the material used in the test.
The virus neutralization table of this series shows that of the
blood of the three animals bled, all show the presence of neutrali-
zation substances, but in varying degrees of concentration. The
serum of no. 81 neutralized completely in the proportion of one
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 441
aA0g® SV SI6I ‘OT Azenuer pot
‘ST6I ‘g Arenuvp pozAyered | LT6T ‘FI tequieseqd | 26
aAOqe SB SI6I ‘OT Atenuer potq
90 | LZI6T ‘Ef Jequieseq | 26
ez puv GT OU sforjzuoo snatA | SII ‘fF ArenuEP pozdyeivg | LIGI ‘FI 1equiecsq | 86 | 9°0 | LTGT ‘ST Jequieced | 96
DAOGB S¥ GY ‘OU [O1}ZUOD SNATA
‘Tjom—oatly | LTT ‘2z tequuaydeg | —¢ | 90 | LTT ‘9% Tequieydeg | 18
mr)
z g
on °} od
ECSERE Bes
SUUVNaU aWOOLDO NOILVIOQOONI ao giva | 5 | e q 2 aagta ig Pic
Ha ® Hx addy
Bins Heo ZO
ie) Z 0 e a &
snwia 139)9 fayooy
quao sad ¢ yand [ 02 wnuas und J ‘uoysodorg *8ar1as 2a0QD ay} fo shayuou fo wnwas oY} fgg ‘uoynzypujnau snivA
‘soisyed jenpis
-01 YPM Ory
-oajut snoraoid
WOIf patoaooay |/TET ‘TZ oquiedseq ped | LIGT ‘0% 1equiedeq_ | 00'T | ZT6T ‘pL Joquie.eq VG
* jorquog |ZTeT ‘Tz toquIa0eq polq | LIGI ‘61 Lequiessq | ¢§°0 | LIGI ‘PI oquisseq &%
[o1}UOD |LTGT ‘TZ Tequieseq ped | LIGI ‘Sf tequesoq | €°'0 | AI6T ‘pT Joquieseq 61
LIGI ‘1% Loquiasaq ped | LIGI ‘OZ tequieseq | ¢€°0 | LI6I “PI Jequteo0q | LT6T ‘EZ JoqMOAON | 16
TT24k— 9ATTYV auoN | €'0 | ZIGT ‘FT Jequisoeq | LT6I “EZ 1equIOAON | 96
LIGL ‘TZ equieseq ped | LIGT ‘OZ Toquiesaqy | €'0 | LIGT ‘FL Joqutsoeq | AT6T ‘ES JoqUIEAON | 16
[eS eA ouoN | $0 | LIGT ‘FI toquIoo0q | LT6T “EZ JoquIOAON | 06
TS “OU IJ [OLPUOL) LIGI ‘G 1040990 patd | LT6I ‘E 10q0990 | 0 | LTET “Lz raquiezdag GL
[]24—eATLV auoN | 2°0 | LI6T ‘2% tequiezdag | LT6T ‘ST Jlaquieydeg | Tg
‘92
; :
SUUVNaH aWOOLAO SISAIVUVd 3 OM Sema?
-QNOONI ISG JO ALVA |NOILVTIOOONI GAILINGAGUd
sosaHnu
soovovn
(worspnwa qua sad g U2 sn.na L9210fay90RT) P19 pu UWD.g pajpj0..a9h76 ‘pasn
ipuaony “PayDOYUN PUD “DOLE ““OoSh sede 999 0] unoy fyDYy-aUo payway snirA *sarsas UoYseluy-g *squawmrsad xa U01I9}01q
1 LU VHO
442 H. L. ABRAMSON AND HERMAN GERBER
toone. The sera of nos. 96 and 97 extended the incubation period
‘of four and five days of the controls nos. 19 and 23 to twenty-one
and twenty-five days respectively. This prolongation of the
incubation period of highly potent virus seems to us to be an indi-
cation of the presence of neutralizing substances. We have been
compelled to apply a rather severe test on account of lack of ani-
mals. The proportion of 10 of serum to 1 of virus has been con-
sidered sufficient for the determination of neutralizing principles.
How great a quantity of these substances is developed by this
process of injection, remains to be determined in experiments
already planned, and which will be carried out as soon as monkeys
are obtainable. No normal monkey serum plus virus control
was used in these experiments, as it has been quite firmly estab-
lished that the blood of normal monkeys does not contain neu-
tralizing substances for poliomyelitis.
Another series of three monkeys were subjected to the 5-injec-
tion method, except that in these the inoculations were made up
from glycerolated cords only, and that the intracerebral test in-
oculation was 0.15 cc. of the 5 per cent Rockefeller virus. This
was done with the idea that possibly cord material might produce
a higher degree of immunity. The test dose was reduced to one-
half, because 0.3 cc. appeared to be much more than enough to
bring down the controls.
Chart 2 tabulates the results of this series. This shows that
two of three animals survived the intracerebral test dose of
0.15 ec. The third, no. 27, showed the first symptoms on the
sixth day after the test inoculation. Both controls nos. 73 and
45 showed the first symptoms on the fourth day after the test
inoculation.
An additional indication of the potency of the testing virus
is shown in the fact that monkey 10, which had survived an ex-
perimental infection from the virus of the 1916 epidemic, with
a residual paralysis of the right lower, was reinfected fatally with
0.5 cc. of 5 per cent Rockefeller virus. Despite the high degree
of immunity conferred on the animal by recovery from the dis-
ease, this animal succumbed readily to re-infection with cnly
three times the dose of the control animals in this series, and to
less than twice the dose used on the controls in the first series.
*
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 443
Gp puUv EJ “SOU S[O1}WOD SATA SI6I ‘Z Yue pod
‘S161 ‘I youwpy pozAywavg | St6r ‘Et Axenaqeg | gf | ¢°0 | ST6I ‘gt Sxenaqag | 99
CH puB Gy “SOU S[O1JUOD SNATA SIGI ‘F Yorvyy porq .
“SI6I “€ Youvyy pozAywreg | ST6T ‘6T Arensqog | TL | 90 | SI6T ‘ST Areniqag | 22%
Cf PUB EJ “OU STOI}UOD SATA To“ pus oalTy | SI6T ‘6T Areniqaq | F | GO | STIGT ‘ST Azeniqaq | ge
*99
one ox zee
des | dq BAS
SHUVNAY GNOOLAO NOILVTOOONI dO Gilva % q qd = q 4 aala a is
2 | BBO es
oy oo °
: Zz
Ssnsia sanjafayooar
quaa sad g qund J 07 wnusas qund J ‘uoysodotg 891108 aaoqn ay} fo shayuow fo vias ayy ig “woynzyvsjnau snirA
uot}
-dayur snotaoid
WOIf paldA0d0yT SIGT “OZ WIV PP | SI6T ‘ST youvyy G0 | S161 ‘IT yore Or
JOLIWOD | ST6T ‘9z Arvnagey porq | STI ‘gz Arenaqeg | CTO | SIGT ‘6T Arenaqoag CF
JO1}UOD |SI6I ‘9g Areniqay potq | giT6r ‘gz Areniqay | STO | SI6I ‘61 Arenaqay el
]]94 pues salty suON | STO | SI6T ‘61 Avenaqog SI6I ‘6c Aisnuee 99
SI6I
‘9g Arenigeay poztioyyg | ST6I ‘¢z Arenaqgeq | STO | ST6T ‘“6t Arwnaqaq | SI6r ‘6g Axenuee | 1z
]]94 pue sar[y ouON | ST'O | ST6I “61 Arenaqoq | ST6T ‘6g Arenuve | g¢
‘99
a ee es a ae x
ere ee hala rank ing el Be | onoGhe aicn ao imnvace Noamvanao.te aiisennme E :
oq dd
VAD no
(uoisjnwma quad sad gG UL SNura saqjafayI0y) pod pajyn}0190h76 “pasn yorsaqoyy
“PaIDIYUN PUD "DLE “Oo9% “Oo9F “OoGG 07 noy fypy-au0 paynay snsvA ‘saruas Uuoyoalur-g ‘*sjuamrsadxa W014NI9}01g
@ LUVHO
444 H. L. ABRAMSON AND HERMAN GERBER
Of the sera of the three animals of this series, one, no. 36,
completely neutralized the virus in the proportion of one of serum
to one of virus. The sera from nos. 27 and 66 prolonged the
four day incubation period of the controls nos. 73 and 45 to°
fifteen and twelve days respectively. As in the first series, the
sera from the treated animals contain neutralizing substances,
but in varying degrees of concentration.
The controls, nos. 73 and 45, showed the first symptoms of the
fatal disease in four days, though the intracerebral inoculation
was only 0.15 cc. of a 5 per cent emulsion, or one-half the dose
used in the first series. |
Three other monkeys were subjected to the injection of in-
creasingly virulent material. In these, the tests were not com-
pleted, and therefore they were not included in the tables. The
protocols follow:
Experiment 1. Macacus rhesus 57 received subcutaneously injec-
tions of 5 per cent Rockefeller virus as follows:
February 13, 1917, 10 cc. heated te 455°C. for one-half hour.
February 16, 1917, 10 cc. heated to 45°C. for one-half hour.
February 20, 1917, 5 cc. heated to 37°C. for one-half hour.
February 24, 1917, 5 cc. unheated 5 per cent virus.
This animal was not bled for the determination of neutralizing sub-
stances, and remained well till time of intracerebral test injection.
March 20, 1917. Injected intracerebrally with 0.6 cc. of a 5 per
cent emulsion of the Rockefeller virus. This was twice the dose used
in the series shown on chart 2.
March 26, 1917. Left hind leg is weak. Does not attempt to get
up. Has tremors.
March 27, 1917. Completely paralyzed. Abdominal respiration.
Etherized to death. Postmortem examination shows typical changes
of poliomyelitis in the cord.
Comment: This animal was not protected against a very large dose
of a virus of high potency. However, the vaccination in itself had no
harmful effects.
Experiment 2. Macacus rhesus 94, an animal that had some months
previously been injected with an emulsion of rat fleas obtained from
houses in which had occurred eases of poliomyelitis and that had shown
no symptoms as result of such injection, was inoculated with a 10 per
cent emulsion of Rockefeller virus as follows:
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 445
April 20, 1917, 5 ec. heated to 55°C. for one-half hour.
April 23, 1917, 5 cc. heated to 45°C. for one-half hour.
April 25, 1917, 5 ec. heated to 37°C. for one-half hour.
April 30, 1917, 5 ec. unheated.
This animal had not been well for some time. Had been coughing
and was somewhat emaciated.
On May 6, 1917, animal died.
Postmortem showed generalized pulmonary tuberculosis.
Microscopic examination of cord showed no changes indicative of
poliomyelitis.
Experiment 3. Macacus rhesus 100 was injected subcutaneously
with a 10 per cent Rockefeller virus as follows:
May 10, 1917, 5 ce. heated to 55°C. for one-half hour.
May 11, 1917, 5 ec. heated to 55°C. for one-half hour.
May 12, 1917, 5 ec. heated to 45°C. for one-half hour.
May 14, 1917, 5 cc. heated to 37°C. for one-half hour..
May 15, 1917, 5 cc. unheated.
May 29, 1917. Animal is well. As result of cardiac puncture to
obtain blood for the purpose of testing for the presence of neutralizing
substances, the animal died. Postmortem showed hemopericardium.
Microscopic examination of cord showed no lesions of poliomyelitis. %
June 1, 1917. A mixture of 0.2 cc. of the serum of Macacus rhesus
100 and 0.2 cc. of 5 per cent Rockefeller virus that had been in contact
for two hours at 37°C. and twenty-two hours in the ice box, was in-
jected intracerebrally into Macacus rhesus 51. This animal survived
the inoculation and is alive and well.
Comment: The vaccination in itself was harmless and the serum of
vaccinated animal contained neutralizing substances.
SUMMARY OF RESULTS OF 5-INJECTION METHOD
Eleven animals were subjected to injections of the highly
potent Rockefeller strain modified by heat as follows:
First heated to 55°C. for one-half hour; second heated to 55°C.
for one-half hour; third heated to 45°C. for one-half hour; fourth
heated to 37°C. for one-half hour. The fifth injection was made
up from glycerolated material from recently paralyzed animals
and injected without previously being subjected to heat. These
injections were administered subcutaneously on successive days.
446 H. L. ABRAMSON AND HERMAN GERBER
Not one of these animals exhibited any untoward symptoms that
were recognizable by careful observation as a result of the course
of injections.
Five out of eight of these animals tested intracerebrally re-
sisted successfully the very reasonable test injection of from three
to six fatal doses. The potency of the testing virus in 1910 was
such that 0.05 to 0.1 cc. of a Berkfeld filtrate was sufficient to
produce the fatal disease in monkeys. It has been passed
through many additional generations since and it is safe to assume
that the virulence has mounted considerably in the past eight
years. ;
The sera of seven of these animals tested in the proportion of
1 part serum to 1 part 5 per cent virus contained neutralizing
substances. Three completely neutralized the virus in this pro-
portion and four prolonged the short four to five day incubation
period as shown in the controls to from twelve to twenty-five
days.
10-INJECTION SERIES WITH VIRUS KILLED BY HEAT
Chart 3 shows that of three animals vaccinated subcutaneously
by daily injections of 5 cc. of 10 per cent emulsion of Rockefeller
virus heated to 55°C. for one-half hour, on ten successive days,
only one Macacus rhesus 1 survived the intracerebral test in-
jection. This animal recéived 0.15 ce. of 5 per cent Rockefeller
virus intracerebrally three weeks after the completion of the
process of vaccination.
On the fourteenth day after the test inoculation, this animal
began to show weakness in the legs, which progressed slowly, so
that on January 5, 1918, both arms and legs were completely
paralyzed. After this, progress of paralysis ceased. On Janu-
ary 10, animal was as lively as he could be under the circum-
stances. ‘The eyes were bright and he eagerly ate fruit that was
held up to his jaws. The animal, however, became so infected
with body lice that he was etherized to death on January 20,
1918.
Two others, Macacus rhesus 80, died one week after the test
intracerebral inoculation of 0.3 cc. of 5 per cent Rockefeller
Sa
- —
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 447
ie ae ek ee eS ee
LIGT
ez sns_ | ‘Tg requioo0qq perp ‘LT6T
-oyI SnNovoByT [O1}UOO SNATA ‘Te soquieoeq pozkyeied | LT6T ‘1g lequisv9q | €8 9°0 LIGBL {0% 1equIeD0q_ |Z
ez sns SIOMOT [VNPTSOL TFTA OATTB
-oyi snovovyyy [O1}UOD SNATA | ‘SI6T ‘6 Arenuve pozApeieg | LIGT ‘ZG toquieseq | OOT | 90 LIGBL ‘1% Lequieoeq, 1
g sns
-oyt snovovyy [OL}UOO SNATA JOM pue oarTy | LIGT ‘L I8q0}90 | 66 9'0 LI6BI ‘9 19q039Q | 08
*09
g > 2
Sree: | Bree FS
ane | ee Bes
SUUVNAL WWOOLNO NOILVTOOONI dO Aiva 3 wield q z aa1a a be
Ban (EB ads )
= i} ONG)
UG Spe Zz
y i}
sna sajjafeyooy yuao sad g jund J 07 wn.«as quod J ‘UuoyLiodoig ‘sartas aaogn fo shiayuou fownsas hg “uouvzypsnau sn A
“OAISN[OUL
TEL ‘Og euNL-9z oUNL “MOY J[Vy-oUO LOZ “HCG OF popwod SNITA Usd od QT Jo SUOTPOLUT OA Jo Solos poATooOI PVH t
Gapas
[ ‘SOU IOJ [OIJWOD |ZTGT ‘8s Joquiovoq ped | LT6T ‘yz taquieoeqy | STO | LT6T ‘TS T9qures9q] GG
LT@L ‘9z roquroseq, perl | LIGT ‘2Z Loquiovoq | GTO | LIGT ‘Te toqutoood | LIT ‘0g ToquIBAON | @
ST61
‘og Arenueep pozt S39]
-19Y}9 {patoaoovoy | puw SUI fo sishjeivg | S16 ‘g Aaenuee | GTO | LIGT ‘Tz raquiaoeq | LT6T ‘Og ToquIeAON | T
08
“ON OF OrZUO)Y FI 10q0990 pod | LIGT ‘ZI 1090790 | €'0 | ZT6T ‘L 1040990 #8
FI 1940990 peld | LIGI ‘ET 1090790 | € 0 | LTT ‘4 4040990 | LT6T ‘FI aoquiaydeg | Ost
“00
ss c
eo a NOILV'T agtLaTawod ae
SSL V FLUE. Coron ene) SISK IV ad 8 ° " -OOONI ISaL 40 ALYa NOILVIQOONI DAILNGAGUd is La
gq dq
24 aD
(uorsynwma yuao dad GUL sn.tqa Layafayooy) P4oo pup UWAG paqnj0.1a0h yb “pasn 4011090
ynoy fyny-euo Lof "Do9G OF payway sniva pUd0 wad Of yun shop aarssavons uo suoyoalur Ua, "s7UWIUWIdLI WO2499}0Lq
€ LUVHO
448 H. L. ABRAMSON AND HERMAN GERBER
virus and Macacus rhesus 2 died one week after an intracerebral
inoculation of 0.15 ce. of 5 per cent Rockefeller virus. Macacus
rhesus 84, the control for no. 80 also died 1 week after the intra-
cerebral injection of 0.3 ec. of 5 per cent Rockefeller virus.
Macacus rhesus 25, the control for nos. 1 and 2, died one week
after the intracerebral inoculation of 0.15 ee. of 5 per cent
Rockefeller virus. The blood of all three of the animals vacci-
nated with killed virus contained neutralizing substances but in
varying degree of concentration as virus neutralization table on
chart 3 indicates.
Macacus rhesi 99, 100 and 83 were each injected intracere-
brally with 0.6 cc. of a mixture of equal parts of 5 per cent emul-
sion of Rockefeller virus and serum obtained from Macacus rhesi
80, 1 and 2 respectively. Rhesus 99 is alive and well. Rhesus
100 began to show first symptoms on the eighteenth day after the
inoculation. The next day complete paralysis of both legs. ‘This
did not progress. The animal is alive with residual palsies of
both lowers. Rhesus 83 showed first symptoms on the tenth
day after the inoculation and died the same day of respiratory
paralysis.
The control animals in this series, Rhesi 84 and 25, which
received 0.3 ec and 0.15 ce. of 5 per cent Rockefeller virus in-
tracerebrally respectively, died of the infection with an incuba-
tion of five and six days respectively.
SUMMARY OF RESULTS OF 10-INJECTION SERIES
Of three animals so treated, two succumbed to the intracere-
bral test dose as promptly as the controls. The third exhibited
paralysis after an incubation of fourteen days. This animal,
Rhesus 1, recovered, but with complete paralysis of arms and
legs. The infection of the control animal in this case had an
incubation period of six days.
Of the three sera tested, one, that of Rhesus 80, neutralized
completely in the proportion of 1 part serum to 1 part 5 per cent
Rockefeller virus. Another, that from Rhesus 1, delayed the
incubation period to eighteen days with recovery, but residual
paralysis. The serum from Rhesus 2 prolonged the incubation
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 449
period to ten days. None of these animals showed any ill effects
from the method of injection.
It would seem from the above results that the animals sub-
jected to ten subcutaneous injections of virus killed by heating
at 55°C. for one-half hour, did not offer as great a resistance to
the intracerebral test injection of from three to six fatal doses,
as those that received the 5-injection series. We, therefore,
endeavored to increase this resistance by injecting virus heated
to 50°C. for one-half hour for seven injections and finishing off
with three injections of virus heated to 45°C. for one-half hour.
Two animals were so injected. The protocols follow:
Macacus rhesus 46. January 18, 1918. Received daily subcu-
taneous injections of 5 cc. of 10 per cent Rockefeller virus in saline
heated to 50°C. for one-half hour for seven days.
January 26, 1918. Received daily subcutaneous injections of 5 ec.
of 10 per cent Rockefeller virus of saline, heated to 45°C. for one-half
hour for three days.
January 29, 1918. Series of injections completed.
January 30, 1918. Animal is quiet. Does not feed well. Seems
tremulous on moving.
January 31, 1918. Completely paralyzed.
February 1, 1918. Autopsied. Sections show poliomyelitis.
Macacus rhesus 7. January 18, 1918. Received a course of injec-
tions similar to that of Rhesus 46. .
January 29, 1918. Course of injections completed.
February 6, 1918. Animal appeared sick. Did not eat his food.
Was quiet. No sign of muscular weakness.
« February 7, 1918. Was found dead in the cage. No sign of pa-
ralysis had been observed. Autopsy showed pneumonic consolidation
of lower right lobe. Section of cord showed slight but unmistakable
signs of poliomyelitis.
Comment: From this experiment, it is apparent that heating virus
to 50°C. for one-half hour does not sufficiently attenuate to exclude
the dangerous possibility of infection as a direct result of the injections
per se. One animal developed frank symptoms of paralysis on the
eleventh day after the beginning of the series of injections. The other
died 19 days after the first injection of lobar pneumonia with lesions of
poliomyelitis in the cord.
450 H. L. ABRAMSON AND HERMAN GERBER
DISCUSSION
We feel that this work opens up a field for the practical appli-
cation of specific preventive measures in poliomyelitis. We have -
subjected eleven monkeys to the 5-injection series without any
ill effects to the animals as a result of the injections. The pos-
sibility of harmful effects from this method, if applied to human
beings, is, of necessity, less than when applied to monkeys, inas-
much as this virus has been adapted to monkeys for the past nine
years and by continuous passage through this animal, has become
highly virulent for the monkey.
The period of time required for this series of injections is a
short one, only five days, which, of course, would render it highly
practicable in time of epidemic, when a rapid method is to be
desired.
This method confers a substantial degree of immunity as shown
by the resistance to multiple intracerebral doses of this highly
potent virus and by the presence of neutralizing substances in
the serum.
it is a well established fact that persons who recover from
poliomyelitis have in their blood neutralizing substances forthe
virus of poliomyelitis and that in all probability, the presence of
these substances is an indication of immunity to re-infection.
The same can be said of monkeys that recover from the experi-
mental disease. If this is so, then why should not animals that
have been artificially made to produce such substances and some
of which have been made to resist multiple intracerebral injec-
tions of most virulent material, be considered immune? The
natural disease is far less serious than is the intracerebral infec-
tion of experimental poliomyelitis. In the former condition, the
body fluids have an opportunity to combat the virus of the infec-
tion at the portal of invasion. In the experimental disease, the
defensive forces of the body are circumvented. It would, there-
fore, be true that an animal body, which contained a large reser-
voir of these anti-poliomyelitis substances artificially produced,
would be able easily to take care of the comparatively mild infec-
tion of human poliomyelitis.
ACTIVE IMMUNITY IN EXPERIMENTAL POLIOMYELITIS 451
REFERENCES
(1) LANDSTEINER AND Popper: Zeitschrift f. Immunititsf. 1909, 2, 377.
(2) FLEXNER AND Lewis: Jr. A. M. A., 64, May 28, 1910.
(3) FLEXNER AND Lewis: Jr. A. M. A., 55, August 20, 1910.
(4) LANDSTEINER AND LevapiTI: Compt. rend. Soc. de Biol., 1910, 68, 311.
(5) R6mer anp JosepH: Miinchener Medizinische Wochenschrift, March 8,
1910, 505.
(6) Fuexner, S.: Jr. A. M. A., 1910, 55, 1105.
THE JOURNAL OF IMMUNOLOGY, VOL. II, NO.6
o
“6
aes
EXPERIMENTAL POLLINOSIS!
PRELIMINARY REPORT
HENRY L. ULRICH
From Division A (University of Minnesota Medicine) City Hospital, Minneapolis
Received for publication August 28, 1918
That hay fever is a form of hypersensibility to pollen proteins
is conceded by everyone. The mechanism of this form of hyper-
sensibility, likewise the phenomenon of its desensitization with
pollen extract (phylactically or prophylactically administered)
are still mooted questions.
Cooke, Flood, Coea’s (1) definition of hay fever as a ‘‘clinical
symptomatic expression of local hypersensitiveness” cannot hold.
The skin and conjunctional reactions in hay fever subjects are
certainly extranasal. Sewall’s (2) criticism that no local mani-
festation of hypersensibility can occur without the background
of a general hypersensibility is amply justified.
Up to now animal experiments to produce clinical hay fever
have not been successful. Koessler (3) reports passive anaphyl-
axis in guinea-pigs. Cooke, Flood and Coca (4) and myself?
have failed to produce active or passive anaphylaxis with pollen
extracts. My experiments to produce a passive anaphylaxis
were made with patients’ blood out of season. It is possible that
Koessler’s experiments were made in season. At any rate the
assumption (5) that there are no circulating antibodies in hay
fever patients should be made with some reservation.
In an effort to produce hay fever in animals, an entirely new
method was used which proved satisfactory enough to Justify
its publication. A group of animals were sensitized to pollen pro-
_ 1Read before the annual meeting of the American Association of Immunolo-
gists, Philadelphia, March 29, 1918.
2Unpublished experiments.
453
HENRY L. ULRICH
454
QYNIT- SAV UNS ONY PO StRUSMAS O
SLL/ LOWOOLG T7709 QWIAT + SHVTKHIENY QV O
eo NT TH PALYOIN SIS NIUAOF
MLLFO -SINY TAI NY +
atl
Ly vy,
L-YUOWS (§) wf
Lunes (5) FJ
Lon
os 3 | Muang ()
HR/
Lhe UndA
3 ee ES i MUYY(S) O
es Slhnio ‘ a ae weve Ze
L- UMS (§) J
LOSS 1
alot “ae L-LYf7-(6) Y
SISONIT10d IV.LNSEWIaSdxS
EXPERIMENTAL POLLINOSIS 455
tein by the injection of a suspension of pollen in salt solution,
intraperitoneally. One series received 1 mgm. of pollen in 1.0 ce.
of salt solution respectively; the other series received 25 mgm.
suspended in 1.0 ce. of salt solution respectively. Ten to twelve
days later insufflation of pollen in the nares was begun. This was
repeated every other day for a long period of days. After each
insufflation notes were made of the clinical behavior of the animals.
The animals were watched usually for one hour after the in-
suffation. Notes made were on each individual animal. The
curves on the chart portray the indices of reactions on successive
days of the groups. The curves, therefore, represent cage reac-
tion and are merely charts of clinical symptoms measured graphi-
cally. The pollen used in these experiments was collected during
the 1917 season. It was cleansed and washed in pure acetone,
dried and stored in sterile vials of two grams capacity. Another
group of animals were sensitized to horse serum—l ce. intra-
peritoneally. Ten days later instillation of horse serum in the
nares was begun and repeated every other day for a long period.
This group virtually was a control to the pollen series. Some
experimental data with this method had already been published
by Sewall (6).
Another group of animals was not previously sensitized but was
subjected merely to pollen insufflation at two days intervals.
This group again was controlled by a horse serum group, the
serum being instilled in the nares at two day intervals without
previous sensitization by the intraperitoneal route. Lastly, a
group of animals was subjected to insufflation of pure starch.
This group was to act as a control to the pollen group in order to
throw out the presumption that sneezing was caused by the local
irritation of dust particles (7). Microscopically, the surface of
starch granules and pollen granules are not at all alike. The
question of irritation may be settled at this point by the fact that
in these experiments, the pollen animals, which reacted by sneez-
ing, always exhibited a latent period after insufflation. In the
starch animals there were no evidences of sneezing or disagree- ‘
able sensations at any time.
456 HENRY L. ULRICH
It may be also added, that a similar series of experiments were
started at the same time with rabbits. The rabbit groups were
conducted for six weeks. At the end of that time the experiments
were stopped because at no time was there any evidence of hyper-
sensibility obtained by the methods employed. In other words the
nasal route as a method to demonstrate clinical hay fever (hyper-
sensibility) in rabbits is not possible.
The symptoms in guinea-pigs obtained by these methods were
those closely resembling clinical hay fever inman. After a latent
period of five to twenty minutes following exposure to pollen,
sneezing, lachrymation, itching of the nares and the body were
obtained, this being often followed by excitability, or depression,
increased respiratory rate, deefication and urination. As the days
progressed, nasal and bronchial stenosis were observed with rales
in the chest, moist and rhonchus in type, but at no time was
there prolonged expiratory breathing, so typical of asthmatic
attacks. In the serum animals, similar clinical symptoms were
observed but never so severe nor so consistent as in the pollen
groups. During the period of the experiments, the animals
thrived and grew in size and weight. No controls in growth or
weight were made. Several of the females became pregnant and
gave birth to healthy young.
Experiment 1, chart A, pollenI. On November 24, 1917, three guinea-
pigs average weight (575 grams) received 1 mgm. of pollen suspended in
1 cc. of salt solution, respectively. December 6, 1917, each animal re-
ceived a nasal insufflation of pollen by means of a medicine dropper.
The animal was held gently on its back by an assistant. One nostril?
was held closed, the other mostril received the tip of the dropper, which
contained a column of dried pollen. The animal thus inhaled the dust
aided by a gentle compression of the head of the dropper. The other
nostril was treated in the same way. This method was repeated every
other day. Fifty-three exposures to pollen were made, covering a period
of one hundred and six days. Including the period of sensitization the
whole time involved was one hundred and sixteen days. The first
insufflation called forth reactions resembling mild hay fever, sneezing,
’Guinea-pigs will not breathe through their mouths except in the last gasps of
anaphylactic shock.
. Rea os
EXPERIMENTAL POLLINOSIS 457
itching of the nares and the body, and lachrymation occurred. A
latent period varying from five to twenty minutes was noticed after each
insufflation. After twenty minutes the symptoms usually were at their
height. Spasmodic sneezing, besides the itching, lachrymation and
nasal stenosis were most striking. At no time were true asthmatic
symptoms obtained. Rales in the chest were made out at the height
of some of the attacks. Usually after an hour’s time all evidence of
clinical manifestations had ceased. Occasional sneezing was heard
some hours after the exposure. A definite rise and fall of nasal irrita-
bility was noticed during these fifty-three exposures. The height of the
first rise occurred approximately at the thirteenth exposure and depth of
the fall at about the twenty-eighth exposure. Another rise reached its
height at the thirty-second exposure and dropped to its lowest at about
the forty-third exposure. Another high tide was at the close of the
experiments March 26, 1918. At each successive rise and fall there was
a definite impression of waning sensibility (see chart). A refractory
stage was impending but wasnever entirely established (?). On March
23, the blood of one animal was tested for precitins with negative results.
Another animal on the same day received 1 ce. of pollen extract (0.5
per cent protein, refractometer method) intracardially with negative
results.. The third animal on March 20 received a stronger extract
(1.77 per cent of protein, refractometer method). This animal reacted
with symptoms of acute anaphylaxis with recovery.
Experiment 2, chart B, serum I. Inthe meanwhile a control group of
animals was under observation. On November 13,1917, three guinea-
pigs had received 1 cc. of Lederle’s normal horse serum intraperitoneally.
This was repeated at two day intervals until forty-seven instillations had
been given. The serum was dropped into nares off the needle that
usually accompanies the Lederle’s package. No definite amount was
given. Sometimes one, sometimes two or three drops were used in
each nostril. Enough was instilled to insure bathing of the turbinates.
The characteristic welling up of salivary secretion was noticed in these
animals after each instillation. It was not until the third or fourth
instillation that any semblance of hay fever symptoms were manifested.
On the whole these animals exhibited the same complex of symptoms
that were obtained in the pollen animals, but never so marked. At no
time was asthma obtained. The group also showed the curves of in-
creasing and decreasing reactivity. In fact at the forty-second in-
stillation complete refractivity was obtained. This continued there-
after and therefore after the forty-seventh instillation the animals were
458 HENRY L. ULRICH
no longer subjected to horse serum by the nasal route. On March
6, one hundred and sixty-five days after the initial intraperitoneal
exposure, two‘ of the animals received 1 cc. of horse serum intracardially.
Acute anaphylaxis occurred with recovery in both animals. This
group supported or illustrated more than any others the possible hy-
pothesis put forth by several observers (8) that there may be a pro-
tective mechanism as a part of the function of the nasal mucosa.
Experiment 3, chart C, pollen II. Five animals, average weight 418
grams, were subjected intrapertioneally to 25 mgm. of pollen, suspended
in 1 ce. of salt solution respectively. On January 8, 1918, insufflation
of pollen in nares was begun and continued at two days intervals. This
experiment was a repetition of experiment 1, with younger animals, a
larger sensitizing dose and a larger number of animals. The first in-
suffation provoked the characteristic symptoms. The reaction in
these animals was more intense than those in experiment 1. Nasal
and bronchial stenosis was obtained as early as the seventh insufflation.
These animals also differed from the first in that no curves of refractivity
were obtained. On the other hand, a steady rise of nasal susceptibility
was noticed. On March 1, 1918, one of the highly susceptible animals
was bled. The blood was tested for precipitin with negative results.
March 31, 1918, one of the animals was injected intracardially with 1
ee. pollen extract (0.5 per cent protein by refractometer method). There
were no symptoms of anaphylaxis. The blood serum of this animal
previous to the intracardiac reaction was tested for precipitins with
negative results. On March 20, 1918, two animals in this series were
injected by the cardiac route with 1 cc. of a 1.77 per cent protein pollen
extract. Both animals reacted with acute stormy anaphylactic symp-
toms with recovery.
Experiment 4, chart D, pollen III. January 13, 1918, three animals
with an average weight of 604 grams were subjected to insufflation of
pollen without previous sensitization of any kind. The insufflation was
conducted at two day intervals just as in the other experiments. Janu-
ary 19, 1918, after the fourth insufflation mild symptoms of sneezing,
itching, and lachrymation was obtained. These animals were subjected
to thirty-four treatments. No curves of refractivity were obtained but a
steady rise of susceptibility was noticed. Precipitin tests were made on
one animal on March 1, 1918, likewise another on March 21, 1918, and
‘The third animal had disappeared from its cage on February 15, and was not
found again.
EXPERIMENTAL POLLINOSIS 459
on March 26, 1918; intracardiac injections of 1 cc. of pollen extract
in one animal for each date resulted in no symptoms.
Experiment 5, chart E, serum II. These animals, five in number of an
average weight of 400 grams, were used as controls for experiment 4.
Horse serum was instilled intranasally without previous sensitization.
The instillation was begun on January 23, 1918. No effect was noticed
until after the fourth instillation. For nearly one month indefinite,
individual reactions occurred, resembling the symptoms of the pollen
animals but not so uniformly or clearly defined. During the next
month, however, more definite symptoms were obtained. On January
6, 1918, three of the animals received 1 cc. of horse serum intracardially.
Two responded with marked anaphylaxis with recovery, the third died
in typical anaphylactic shock. On January 17, 1918, when the curve of
sensibility was most marked up to this time, the two remaining animals
were subjected to 1 ec. of horse serum intracardially and both died in
typical anaphylactic shock. The series conclusively proved that hy-
persensibility can be produced by the nasal route.
Experiment 6, chart F, starch I. January 25, 1918; three animals, of
an average weight of 390 grams, were subjected to insufflation similar
to the method used in the pollen animals. The animals were used as a
control to the pollen group in order to answer the question which might
be raised whether the nasal irritation and symptoms were due to mere
mechanical irritation of dust. The animals were exposed to 29 insula-
tions on the same days as the pollen animals and at similar intervals.
At no time did the animals manifest any symptoms of discomfort or
annoyance by the treatments.
CONCLUSIONS
1. Clinical manifestations of hay fever, pollinosis, can be pro-
duced in laboratory animals.
2. Horse serum produces similar manifestations in the guinea
pig but not as clearly nor as uniformly. At no time were true
asthmatic symptoms obtained.
3. Evidence of refractive phenomena was obtained by rhyth-
mic exposure of the nasal mucous membrane to foreign proteid.
4. Sensitization by the nasal route was established; first, by
the manifestation of clinical signs of pollinosis after the fourth
exposure; secondly, by the anaphylactic phenomena in the serum
animals when serum subsequently was injected into the blood.
5. Non-protein dust gave no symptoms of hypersensibility.
460 HENRY L. ULRICH
DISCUSSION
That we can sensitize the guinea-pig to pollen by mere exposure
of the nasal mucous membrane is of great importance to those
who are interested in the membrane from a physiological point of
view. Cooke and Van der Veer (9) have shown the influence of
heredity on protein sensitization. Their. cases were those
occurring entirely ‘‘spontaneously,” from exposure either to
pollen or to food. It would be well to consider, in reflecting on
the mechanism of hay fever, whether we are not dealing with a
congenital or acquired defect of function of the nasal mucosa. It
may be possible that the difference between the person not sen-
sitive to pollen and the one sensitive lies in the rate of the diges-
tion of the pollen proteids by the respective nasal mucosa. The
sensitive membrane may have lost the faculty of rapid conversion
of the proteins to innocous amino acids or may never have had
it. This may be the explanation of the food types of sensitization
through an intact mucosa. The rabbit is insensitive to nasal
exposure to pollen. The guinea-pig is quite sensitive. The
difference in reactions of the mucous membrane in the two species
is suggestive of a difference in function phylogenetically.
The animals in chart D, pollen 3, which had not been pre-
viously sensitized, at no time showed precipitin reactions with the
blood, nor anaphylactic symptoms after pollen protein had been
introduced into the blood. This experiment approaches in some
details what has been found in the human sufferer. Clowes
(10) claims he has found precipitin and Koessler (11) reports
passive transfer of immune bodies from the patient to the guinea.
pig. Cooke, Flood and Coca (12) and I myself have never been
able to reaffirm these claims. It would be of interest to attempt
passive transference during the height of the hay fever season.
Apropos of this: one animal highly sensitive, in chart C, pollen
2, was bled and its serum was injected peritoneally into a normal
pig. Forty-eight hours later pollen extract introduced intra-
cardially elicited no response. Again a group of (3) animals were
injected with 3 ec. of human serum respectively, from a hay fever
sufferer. Two days later these animals were subjected to insuf-
Se
EXPERIMENTAL POLLINOSIS 461
flation of pollen into the nares. Symptoms such as sneezing,
itching and lachrymation occurred. The experiment was re-
peated in another group (3), with another patient’s blood with
entirely negative results. These observations will be repeated
and elaborated at some future time. The possibility of demon-
strating a passive transfer of immune bodies by a reaction such
as that in the skin or mucous membrane, is, as far as I know, a
new method of approaching this problem. Lastly, the fact that
we can inject pollen grains intraperitoneally in the guinea-pig
without untoward results suggests the idea of the possible use of
pollen grains direct as a phylactic or prophylactic measure in the
treatment of the disease.
REFERENCES
(1) Cooks, R.A., Fuoop, E. P., Coca, A. F.: Jour. of Immunology, 1917, 2, 217.
(2) Sewatt, H.: The Journal of Laboratory and Clinical Medicine, 1917, 2, 875.
(3) Korss er, K. K.: Forscheimer’s Therapeutics, 1914, 5, 685.
(4) Vide (1).
(5) Vide (1).
(6) Vide (2).
(7) ScHEPPEGRELL, W.: Archives of Internal Medicine, 1917, 14, 959.
(8) Pacer, O.: Medical Record, 1915, 88, 470.
Pacet, O.: Medical Record, 1917, 92, 668.
Hermann, Cuas., N. Y. State Journal of Medicine, 15, 233.
Kurca, A. C.: J. H. Bulletin, 1913, 24, 69.
Sewatu, H: Arch. Int. Med., 1914, 13, 856.
(9) Cooks, R. A., VAN DER Veer, J. A.: Jour. Immunology, 1916, 1, 201.
(10) Crowes, G. H. A.: Proceed. Soc. Exp. Biol. and Med., 1912-13, 10, 69.
(11) Vide (3).
(12) Vide (1).
PROMPT MACROSCOPIC AGGLUTINATION IN THE
DIAGNOSIS OF GLANDERS
OLGA R. POVITZKY
From the Bureau of Laboratories of the Department of Health of the City of New
York
Received for publication September 16, 1918
The agglutination test for the diagnosis of acute cases of
glanders is conceded by all workers to be of great value. Miess-
ner (1) and others have shown that agglutinins in glanders are
generally at their height between the fifth and eleventh day of
the infection, after which time they begin to decline. At the
end of two months, they may fall even below 1: 500. This test,
if used alone, would fail to detect chronic cases and it should,
therefore, be used as an adjunct to the complement fixation! and
ophthlamic mallein? tests. No one test can be depended on alone
as each one has its peculiar value in certain stages of the disease.
If all these tests are combined, very few cases of glander can
escape detection. Even though a case may appear to be nega-
tive by one or two of the methods, it may be found positive by
the third.
The great objection to the employment of the macroscopic ag-
glutination method used heretofore in the routine diagnosis of
glanders, is the length of time required for the appearance of a
reaction. While some tests can be read at the end of twenty-
four hours, others do not show definite agglutination within
forty-eight or seventy-two hours or even longer. This drawback
led to the centrifuge method for which Muller (2) claims priority.
1 Specific amboceptors for the complement fixation test may be demonstrated
seven to ten days after infection and they remain during the course of the dis-
ease (6).
2 The ophthalmic mallein test is reliable twenty-one days after infection while
the subcutaneous mallein test may be relied upon for diagnosis fifteen days after
infection (6).
463
464 OLGA R. POVITZKY
By it the process of the agglutination reaction is shortened con-
siderably. Miessner (1) also Pfeiler (8) employed this method
claiming it to be a great success.
According to this method, the test fluid is prepared in the
same way as in the old method—namely, a forty-eight hour
growth of a suitable strain of B. mallei is suspended in salt solu-
tion and heated at 60°C. for two hours. It is then diluted to a
certain density with 0.5 per cent carbolic acid. The tubes are
set up in the ordinary way, each tube containing 2 to 3 ce. of the
titrated test fluid with varying quantities of serum to make
the final dilution, 1: 400, 1: 500, 1: 800, 1: 1000, ete. After one
hour incubation at 37°C. the tubes are centrifuged for ten minutes,
after which they are allowed to stand at room temperature for
one and one-half hours. The tests are read at the end of that
time. The appearance of an irregular veil like clumping at the
bottom of the tube with clearing of the upper portion is consid-
ered an agglutination, while a dense white precipitate with a
cloudiness of the upper part is considered a negative reaction.
Though the centrifuge method possesses great advantages
over the old method in point of time, it is difficult always to
separate an agglutination from a sedimentation and the result
is not always readable. Thus, Anthony and Grund (4) found
this test unreadable in 12 per cent of their cases tested, even
after an incubation of twenty-four hours.
In order to eliminate the shortcomings of the former methods,
I have worked out a modification by which an agglutination re-
action, clear cut in appearance, can be obtained in less than two
hours with positive sera in dilutions up to 1: 2000 or higher.
The strain of B. mallei used was one called B. M. 5, isolated
at this laboratory about five years ago from a case of human
glanders. Fourteen other strains from various sources (chiefly
horses) were tested also, but none have given invariably the
prompt and clear cut reaction that-B. M.5 does. The speci-
ficity of this strain was ascertained by testing it with a number
of syphilitic, typhoid, streptococcus and pneumococcus sera; also
with a number of sera from horses with equisepticus. One of
the latter sera gave a positive reaction but the possibility of
AGGLUTINATION IN DIAGNOSIS OF GLANDERS 465
glanders could not be ruled out. When passed through guinea-
pigs, B. M. 5 became more virulent but, after successive passages,
did not agglutinate so well; in fact, I found that it was not ad-
visable to pass this strain through a guinea-pig, as most writers
advocate to keep the strain agglutinable. Other strains, how-
ever, have not proved so constant as this one.
Next in importance to the native agglutinability and con-
stancy of the strain, is the medium on which it is grown. The
medium which has given the most satisfactory results is potato-
glycerin-veal agar that is 2.5 per cent acid to phenolphthalein.
It is prepared as follows:
Wealeintisrons (UNAGTUStEG) sce... acts Acie eteeeeeenet 1000 ce.
LM ERY Po Nee aE eae eS ARR ER ns ORS Lis 7 0d 30 grams
INFO ES 68 Bee Cen an ae MERE SI: FCoE A pn 5 grams
Re pporm (acl) so Aas teas cayetoue 3 sagcoye4 51.9.9 eyesore near 10 grams
* Veal infusion: Chopped lean beef 10 pounds, water 10 liters. Soak over
night. Heat for one hour at 45°C. then boil for one-half hour. Strain through
cheesecloth. This infusion may be sterilized and stored for stock or used at
once.
Put in autoclave or Arnold to melt agar.
Titrate at room temperature and lower the natural acidity, if
necessary, to 2.5 acid (phenolphthalein) by adding normal sodium
hydroxid. Do not add acid under any circumstances.
Clear with egg and filter. Titrate again and if necessary, ad-
just to 2.5 acid. Add glycerin (C. P.) 5 per cent and potato
juice 5 per cent. Sterilize in autoclave for one-half hour at 15
pounds pressure. .
The potato juice used in this medium is prepared by adding
one pound of unpealed potatoes, sliced thin, to 1 liter of distilled
water. Autoclave one-half hour at 15 pounds, strain through
cheese-cloth and a thin layer of cotton. Bottle for stock and
sterilize in autoclave one half hour at 15 pounds. This stock
must be filtered each time before adding to medium.
That too much stress cannot be laid on the careful preparation
of the medium is shown by the following experience. At one
time in the beginning of my work, the suspension of B. M. 5
grown on a freshly prepared medium did not give the proper
reactions with the control sera. Keeping in mind that the strain
466 OLGA R. POVITZKY
was old and in need of rejuvenation, I passed it through a guinea-
pig and recovered the organism from the heart’s blood. Mean-
while, a second fresh medium was made upon which the newly
recovered organism was planted. This suspension reacted very
successfully and I should have believed that the guinea-pig pas-
sage was responsible for the success had I not planted the old
organism on the second medium also. The suspension of this
growth worked as well as the one just isolated from the guinea-
pig. Therefore, the old culture is still used.
Careful attention should be given to all the glassware used in
. connection with cultures as well as the tests. It should always
be neutralized. Bichlorid of mercury should be avoided as a dis-
infectant for the pipettes, bottles, tubes, etc.
The stock culture of B. mallet should be transplanted every
ten to twelve days and incubated at 37°C. for two or three days.
It should then be kept in the ice-box at 10°C.
The suspension for the tests is prepared in the following man-
ner: Forty-eight hour cultures are used to inoculate agar slants
of the potato-glycerin-veal agar described above. A good growth
should be insured by inoculating the surface generously. After
forty-eight hours incubation at 37°C. the growth is washed off
with 0.85 per cent salt solution and killed by heating at 60°C.
for two hours. No earbolic acid is added to this stock suspension
which is of considerable density. It can be kept in the ice-box
for two months or more in 100 cc. bottles, corked and capped,
if handled with aseptic precautions. The tests are carried out
with a fresh dilution of the stock suspension made by adding
0.85 per cent saline solution.
A sample dilution, which has been tested with known negative
and positive sera, should always be kept in the ice-box and the
fresh dilutions of the suspension compared with it on printed
matter. The fresh suspension (dilution) is not filtered as the
filtermg seems to hold back something essential to the reaction;
moreover, this is not necessary as it can be shaken up to a per-
fectly homogeneous fluid.
The tests are carried out in the following manner: A primary
dilution of the serum 1:40 is made. Each tube receives varying
AGGLUTINATION IN DIAGNOSIS OF GLANDERS 467
quantities of this dilution to which 3 ce. of the bacterial suspen-
sion (prepared as directed above) is added to make a final serum
dilution of 1: 500, 1: 800, 1: 1000, 1: 1200, 1: 1600, 1: 2000. A
known positive and a known negative serum are always used as
controls with each test; also a control tube of the bacterial sus-
pension without serum. The tubes, in copper racks are placed
in a water-bath (37° to 42°C.) for two hours.
With this technic a reaction up to 1000 or more may be ob-
tained in ten to twenty minutes; a positive reaction always ap-
pears in two hours. The tests may be kept over night in the
ice-box and read again in the morning. The reaction, if com-
plete is designated by two plus signs, incomplete by one plus,
slight by plus minus, negative by adash. The reaction is so clear
cut there is no difficulty in interpreting it. One can watch the
bacilli clump through the tube, then fall to the bottom in a
white granular mass, leaving the supernatant fluid crystal clear in
a complete reaction. There is no necessity to look for “‘but-
tons,” “‘veils’ or “films.” It is either an agglutination as we
see it with other organisms or it is not considered a reaction.
In reading the tests, a reaction is considered as positive which
has double plus through 1: 1600; as suspicious with double plus
in 1: 1200; as doubtful with double plus in 1:1000. Any reac-
tions below 1: 1000 are considered as negative.
It is essential to titrate the dilute bacterial suspension often
with known positive and negative sera to be sure it is working.
Tests should never be one with a suspension that has not been
titrated the previous day.
The Department of Health of the City of New York has been
employing the agglutination test for the diagnosis of glanders
tegether with the complement fixation and ophthalmic mallein
tests. During the last year, over 2000 sera were examined,
which might have given us very rich material for arriving at
definite conclusions as to the comparative value of the sero-
diagnostic and ophthalmic tests, checked by autopsy findings,
had there been closer codperation between the city and state
veterinarians and the laboratory. Such codperation should be
based on a systematic method of recording full data in connection
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 6
. 468 OLGA R. POVITZKY
with each horse in a central file. For a thorough study of glan-
ders in any region, each horse should be marked and registered
under an identifying number, as are automobiles, only in this
case the number should be a permanent one which should pass
with the horse when it is sold. The traditional methods of
horse dealers might make this a difficult matter, yet, for economic
reasons a law to that effect rigidly enforced, would be thoroughly
worth while.
On only 180 horses was it possible to obtain complete data
which I have endeavored to analyze. These horses fall into three
groups as shown in tables 1, 2, 3.
Street. railway horses: «.. oc. fare.2 de oe fo lok coe eeoke. © sv e0e snes Oe ne 111
Horses from ‘an infected stable.-)'.-0---e se tee ee eee 14
Horses for ‘slaughter: <5... Beet eee eee eee 55
"EOtalls coc jieciceon She tin neo ee oe ne ee eee eee 180
The street railway horses and those from a badly infected
stable (125 in all) are considered together as at autopsy they all
proved to have glanders with the exception of two. These two
were in the infected stable and they were killed with the others
because of their marked exposure to the disease. One of these
horses, number 562, table 2, had been given vaccine; thus the
number of 125 is reduced to 123 on which the percentages are
based.
The 55 horses for slaughter were among those that received
careful tests to determine their fitness to be used for food. These
horses came under the supervision of Dr. L. Price, veterinarian
of the Department of Health and the data are the most satisfac-
tory of all.
Lately, we have been trying to confirm autopsy findings by
pathological sections and this promises to be of great advantage.
Many lesions which are suspicious of glanders do not prove so
microscopically.
The bacteriological inoculation of suspicious material into
guinea-pigs for the Strauss reaction does not prove of much aid,
as, according to Miessner (3), it is successful in only 25 per cent
of the cases. My own experience is in accordance with this
ee aa ee he ee a
AGGLUTINATION IN DIAGNOSIS OF GLANDERS 469
estimate. It was only in acute cases when B. mallet was isolated
from the blood (human) or in horses with an acute infection of
glanders that the Strauss reaction was obtained in the guinea-
pig after moculation with emulsified material from the lesions.
All other inoculation tests resulted negative’y.
In considering the horses, in tables 1 and 2, the nece sity of
applying all three tests stands out prominently. Of the 123
horses which proved at autopsy to have glanders, 122 were shown
to be positive, suspicious or doubtful by one, two or three of the
tests. Only one case (horse 123) was negative with all three
tests—a failure of about 0.8 per cent.
The agglutination test showed 31 horses to be positive, 15 sus-
picious and 32 doubtful, making a total of 78 horses or 64.2 per
cent. Thus it failed in 35.8 per cent of the cases, probably be-
cause most of the horses in this series were chronic cases. Three
horses? were picked out by this test alone—the complement fixa-
tion and eye mallein tests being negative. These may have
been acute cases.
The complement fixation test showed 74 horses as positive, 4
suspicious and 17 doubtful making a total of 93 horses or 75.6
per cent. This test failed wholly in 24.4 per cent of the cases.
Four horses! were picked out by this test alone.
On the other hand the mallein test (ophthalmic except for 9
horses in table 2 on which subcutaneous mallein was used) gave
39 horses as positive, 65 suspicious and 2 doubtful making a
total of 106 horses or 87.8 per cent. It failed in 12.2 per cent,
but 15 horses® were picked out only by this test.
The grouping of suspicious and doubtful reactions with the
positive ones is based on the interpretation that any serodiagnos-
tic reaction not negative, should place an animal in the suspicious
category to be subjected to close observation and frequent re-
tests. That such a method is of great value is fully shown in
tables 1 and 2.
3 Table 1, horses 76, 1218, 1504.
4 Table 1, horses 570, 1281, 1389, 2047.
5 Table 1, horses 19, 449, 462, 494, 612, 621, 1143, 1146, 1310, 1874 and 1390,
Table 2, horses 730, 737, 751, 763.
470 OLGA R. POVITZKY
Further analysis
Positive by vallithree’ tests... 022.2 .22412 26a e ee 9 horses*
Positive by agglutination and negative complement fixation. 6 horsest
LALO) 0 Ive Ba Oe a ee eI mere ccm ke i el Cas 23 horses
Positive by mallein and negative by complement fixation and
SAP CRI ARION: ©. 5.2 .10.. dap et Dee eee eee ee oe eee 15 horsest
* Table 1, horses 418, 506, 1776; table 2, 645, 683, 738, 746, 750, 757.
{t Table 1, horses 163, 181, 605, 1218, 1360, 1504.
t Table 1, horses 19, 449, 462, 494, 612, 621, 1143, 1146, 1310, 1374, 1390; table 2,
horses 730, 737, 751, 763. 5
Of the 55 horses tested before slaughter for food purposes
(table 3), 4 horses® proved to have glanders. The complement
fixation test failed to detect this fact in all 4 cases while the ag-
glutination test failed in 3 of them and the ophthalmic mallein
in only 1 (horse 2001) which the agglutination reaction picked
out. The agglutination test gave 3 false “suspicious’’ reactions,?
the complement fixation none, and the eye mallein 9.
In a set of horses such as are listed in tables 1 and 2, a large
number of chronic cases may be expected and the agglutination
tests positive in the least number of instances. On the other
hand in a large number of sera examined during a whole year
where different stables are tested there may be a great many
cases of beginning infection or cases in the acute stages of the
disease. .
There were tested 1890 sera from various sources and among
these the agglutination test gave positive and suspicious reac-
tions in a larger proportion than the complement fixation. Un-
fortunately the eye mallein and final outcome of the cases could
not be obtained.
Thus, with the complement fixation, about 14.5: per cent of the
cases were either positive or suspicious while the agglutination
registered 28.2 per cent. The doubtful reactions occurring with
two tests were almost the same in number (complement fixation
300 and agglutination 291).
® Table 3, horses 2001, 2006, 3044, 3064.
7 Table 3, horses 3027, 3040, 3053.
AGGLUTINATION IN DIAGNOSIS OF GLANDERS 471
With complement fixation test
2 TST NG ae ee ea AL erat) Ice A nd 254
SrCIOUB ER chi io. oi rheetley Ais FR AU pe pel pd boy 4. oes 21
“JOULE GTS SS BA a ene pO dW OC Rca Se td pd i a 300
1 EV LET 720 A SIN GOR Bc ae a 1315
With agglutination test
iP DELICE Go a pee ee Ye TE OR 262
IMPICEOUS EE 2 2ec's fess at eee Cha bo neo ee ee 171
Metraet PEEEAM AEP. « o's GD ee rss oo oh! | 6 ee ee ae pete ate 291
Be eR TCO as os av SBYAS Gato R ABI o's 5.6 Vig EG Oe Cees 1166
With both tests
2 DELESTGL 25 Se Ree, 3 See ee eo eae ek ee 138
PSUS TOLCLO US menses ar A ier hack oS ch emacs As cla sarc vce tala oe 5
12 Ti) DEAE OSS See OR mt ers eek ls) RI 66
LS GISIES TES Cag GAR A ian eR INES, D806, oe aha ot 800
With either one of the two tests
RSM Car 5A. M4, Sida Ss a 1h lerctlanitcy decetss Sehakcs \ alpape ee eee ele eae 378
SOTES DOE DUE a oe I ee ri dom ets tats 187
PNPUIMRRES CERES NO coc farce feave.S iasg atone pies Uke Sue w ad ORR ee 525
SRSA Ce Sse te IRE s Saal Ae oRenes 3 ath ORS a eee 800
These 1890 sera did not represent 1890 horses for about 15
per cent were retests. If it were not for these retests,a greater
percentage of agglutination reactions would have been obtained
for, after a lapse of time, the agglutinins disappeared in the
horses retested, while the complement fixation antibodies re-
mained, or appeared for the first time.
Another factor which influences the sero-diagnostic tests is the
subcutaneous injection of mallein. It is generally held that six
to eight weeks should elapse before the blood is taken for tests
after mallein has been injected subcutaneously. Recently, in a
large number of horses that had been given subcutaneous mal-
lein, many showed four plus complement fixation and positive
agglutination reactions, when retested about six weeks later.
At the time of the first test these same horses had shown either
doubtful or negative results. These horses are being kept under
observation and time will show whether the high titers in both
tests were due to the subcutaneous injection of mallein or to a
fresh infection.
472 OLGA R. POVITZKY
SUMMARY
1. A method has been devised whereby a prompt clear cut
macroscopic agglutination for the diagnosis of glanders can be
obtained in two hours.
2. Fifteen strains of B. mallet were tested and one obtained
which is constant in its agglutinability with positive sera.
3. The best culture medium is glycerin-potato-agar (2.5 acid
to phenolphthalein) carefully prepared (see text).
4, Great care must be used in the preparation of the stock
suspension of B. mallez (see text). For the tests, fresh dilutions
in 0.85 per cent salt solution are made from the stock suspension.
No carbolic acid is added to the stock nor to the dilutions.
5. The agglutination test is valuable in the routine examina-
tion of horses. With this test nearly all early and acute cases
can be detected and thus the spread of the disease can be
prevented.
6. A negative reaction by a single agglutination test if not
confirmed by the ophthalmic and the complement fixation tests
does not prove the case negative; nor does a single negative re-
sult either by the complement fixation or ophthalmic tests. -ar= slo eee 67
——. The non-influence of injections of trypsin upon the protein quotient in
blood serum: foie se ck ss so s yc sible ee ise S Dele he oe uae Pe Or een 139
—— and Robertson, T. Brailsford. Anew method of estimating the antitryp-
tie index of bloodserum:, :..52. «oa l.sser tncrs fe nee eee > ea eee 131
Hartman, C. C., and Lacy, G. R. Specific reactions of the body fluids in
PNEUMOCOCCIE IMFECHION. 2.0. 2c) so 5s sins wise ie os Ves oe ee See 43
Heist, George D., Solis-Cohen, Solomon, and Solis-Cohen, Myer. The bac-
tericidal action of whole blood, with a technique for its determination. 261
Hemblysis, A. study Of saponin: «20665 2.60 = .4 Sos 5S ace ee eee 423
Immune bodies, The isolation, purification and concentration of........... 109
—— hemolysin, A studyouol:s> eee, om eases owe Seek eae eke ee 109
Immunity, Active, in experimental poliomyelitis................-..-...45+- 435
——., The rdle of, in the conduct of the present war...........---++++++++:- 371
——. The study of problems of, by the tissue culture method.......... 219, 233
Immunizing properties, A study of the, of bacterial vaccines prepared after
various, methods... 62242 wate och bes > Rae Bale hic ee 247
Immunologic properties of uveal pigment, The..............-----...+--+-+ 75
Indigo test; On Von Dungern’s; forjsyphilis . asajeecer -.---1a<41205-- aera 11
Intensive digestion, The constancy of the protein quotient during, and pro-
longed) starvation... .....). 01. 2's gmc Agaie mie ee =n a se eee eee 67
INDEX 483
Intracutaneous absorption, The specificity of.................... Sheps tits
Intravenous injections, Effects of, of a colloid (gelatin) upon rabbit sera.. .
Isolation, purification and concentration of immune bodies: a study of im-
PAGE REM O Ly SUIT Netty ees eae ey Ne AL. BALL Latha ote ee
Kahn, Reuben L., and MeNeil, Archibald. A note on the relation between
proteolysinssand haemolysinst.g+..ssus ee tess sete etio dels noel See ntees ss ?
——-——. Archibald. Complement fixation with protein substances.......
Kolmer, John A. The réle of immunity in the conduct of the present war...
—— and Matsunami, Toitsu. The relation of the meningococcidal activity
of the blood to resistance to virulent meningococei.....................
—— and Perry, M. W. A study of the immunizing properties of bacterial
vaccines prepared after varlous methods. 2)... os 2 vee cite e ew. oe ks
——and Sekiguchi, Shigeki. Experiments upon the passive transfer of
antibodies from the blood to the cerebrospinal fluid....................
—— and Toyama, Ikuso. ‘The influence of arsphenamine and mercuric
chlorid upon complement and antibody production. .
—— and Weiss, Charles. Studies in pneumonia. VIII. “Abslaian eeictiow: to
A SIVE URINE UO RAT RERMEO TREE NUS Ses 1K GR 1/08 (ek eeee CRE sive BRE en oUt enal i eta Ly SAP Re Rae PAM
——, Toyama, Ikuzo, and Matsunami, Toitsu. The influence of active nor-
mal serum (complement) upon meningococci. I. The opsonic activity of
fresh normal serum alone and in combination with antimeningitis serum
OTIC IVETN EO COC Cina Say aeia takes ioosesocuat eines Faycici ae eae ort ei es Ge
— and Matsunami, Toitsu. The influence of active normal serum (com-
plement) upon meningococci. II. The bacterial and protective value
of fresh normal serum alone and in combination with antimeningitis
APMMELY KOR MINE IN ZOCOCEL J. hep Aaya A gets wes oectegs MUD Ane 4 cle
Korns, John H., and Balls, A. K. On the mode of action in vitro and ‘fie
preparation of hemolytie antibodiess./ 4/512. MN oe one. anna
Kosakai, M. The isolation, purification and concentration of immune
bodies-va study of immune: hemolysin) .)) 8. led a ee Pe one
Krumwiede, Jr., Charles, and Noble, W. Carey. A rapid method for the
production of precipitin antigen from bacteria: an attempt to apply it
to the determination of the type of pneumococcusin sputum............
Lacy, G. R., and Hartman, C.C. Specific reactions of the body fluids in pneu-
HAO COC CLE MMLC CLONAL Cpa e nak Repeal LhGbs eettats, ae cine erolonete) akere Se Retey eke t=
MeNeil, Archibald, and Kahn, Reuben L. A note on the relation between
PLOUCOKVSIN Sean GUMACIMOlY SING Hats co -are cietee ta ey~ ctael« eles tats MRepebouaPle) teers
Complement fixation with protein substances................+--
Macroscopic agglutination, Prompt, in the diagnosis of glanders...........
Matsunami, Toitsu, Kolmer, John A., and Toyama, Ikuzo. The influence
of active normal serum (complement) upon meningococci. I. The opso-
nic activity of fresh normal serum alone and in combination with anti-
MEHINPIFIS, BELUM LOF MENINGOCOCCL: ...... 5.052 sti oe derssle a alae tee neee oles
—and Kolmer, JohnA. The influence of active normal serum (complement)
upon meningococci. II. The bacterial and protective value of fresh
normal serum alone and in combination with antimeningitis serum for
BUCH O COCO eke segs is isha ave vb SRI Sho ate « woe oh reph aphad Ws lee exh MRS Neer
—— and Kolmer, John A. The relation of the meningococcidal activity of
the blood to resistance to virulent meningococcl............--+-+++.++-
THE JOURNAL OF IMMUNOLOGY, VOL. III, NO. 6
. 301
395
157
157
177
484 INDEX
Meningococci, The bacterial and protective value of fresh normal serum
alone and in combination with antimeningitis serum for................ 177
——, The influence of active normal serum (complement) upon........ 157, 177
——, The opsonic activity of fresh normal serum alone and in combination
with -antimeningitis serum for. . : 32:32. aSacs 4. been eee oe ee 157
——, The relation of the meningococcidal activity of the blood to resistance
tonvirul enibaseekyse ne ce eae 413
Mode of action in vitro and the preparation of hemolytic antibodies, On the 375
Noble, W. Carey, and Krumwiede, Jr., Charles. A rapid method for the
production of precipitin antigen from bacteria: an attempt to apply it
to the determination of the type of pneumococcus in sputum........... 1
Opsonie activity of fresh normal serum alone and in combination with anti-
meningitis serum for meningococel, The... .:: : .......25.2225-7-=-e eee 157
Perry, M. W., and Kolmer, John A. A study of the immunizing properties
of bacterial vaccines prepared after various methods................... 247
Pneumococcic infection, Specific reactions of the body fluids in............ 43
Pneumococcus in sputum, A rapid method for the production of precipitin
antigen from bacteria: an attempt to apply it to the determination of the
RY POOL «0. ses cise nade gol a doe a eas Bs Risiere ee eee ee 1
Pneumonia, Studies-ins : sce eebeotk jae ae okt Os eC eee eee ie ee eee 395
Pneumotoxin, A skin reaction t0:..c-5--e->--4-ss eee ee eee eee 395
Poliomyelitis, Active immunity in experimental..........................-- 435
Pollinosis,, Experimental: i022 22th: 30... «3:1... eee ieee ee eee 453
Povitzky, Olga R. Prompt macroscopic agglutination in the diagnosis of
glanders: «icicle ois coed sea nt ae die deh = eee eels Reno eee ae 463
Precipitin antigen, A rapid method for the production of, from bacteria.... 1
Proceedings of the American Association of immunologists................-. 317
Active immunity in experimental poliomyelitis. Harry L. Abramson.. 317
— immunization against pneumonia, The. R. Kohn................. 331
Bacteriological study of post-operative pneumonia, A. Miriam P.
Olmstead 4.5 HBA iis ross s:d wists Skee 4 sae ee ee eee 330
Contribution to the bacteriology of B. fusiformis; its morphologic
phases and their significance, A. Ralph R. Mellon.................. 327
— to the study of complement fixation in tuberculosis, A. Hassow
von: Wedel! pe ki cts Ste.x Sehr Ae SPO Se eee 339
to the study of complement fixation in tuberculosis, A. M. A.
Wilson... | .<..)...k4bepe at eetten 25824. sabes dt cate ee eee 339
Examination of the blood preliminary to the operation of blood trans-
fusion, The. Arthurvle Cotas iss conocer ot See eee eee 337
Experimental pollinosis in guinea-pigs. Henry L. Ulrich.............. 325.
Experiments on the production of antipoliomyelitis serum in rabbits.
Bidgar {H. ‘Tsent: 20.26: 2e 2s Shae S82 Se ee ie * AS Re. eee 317
upon the chemotherapy and chemoserotherapy of pneumococcus
infection. John A. Kolmer, Edward Steinfield and Charles Weiss.... 336
upon the passive transfer of antibodies to the cerebrospinal fluid.
John A. Kolmer and Shigeki Sekiguchi... . =... .<-.- Oy be. ss) rie AO lOWEs tt tee ete tt ce. aise ee oe 333
—— of arsenobenzol and mercury upon antibody production, The.
Kuo; Loyama-and John Asolmer:. :2h0: stent ees eda eect ls Jecee 326
—— of normal human and guinea-pig serum (complement) alone and in
combination with antimeningitis sera upon virulent meningococci,
The. John A. Kolmer, Toitsu Matsunami and Ikuzo Toyama........ 319
Isolation, purification and concentration of immune hemolysin, The.
iN. LGR ie Agape gehen a it St Seem RE le aga we RU Aled Mga 338
Lipo-vaccines. Eugene R. Whitmore and E. Fennel.................. 343
Method for preparing bacterial antigens, A. J. C. Small.............. 339
Nature of antianaphylaxis, The. J. Bronfenbrenner and M. J. Schlesin-
CIE reese ois hte tetas es Eee she a tes rie alot ceenejenerens bite eters eine bon case 321
Observations on the intraspinous autosalvarsanized serum therapy of
cerebrospinal syphilis. Benjamin A. Thomas........................ 342
Persistence of active immunity in those immunized against diphtheria.
Wwvriliieyam ISL, TReay a See is 2s Fae a aie aN eo eae aA ar smn BER A nA a ae ee 328
Production of pneumococcus antiserum and the corresponding curves
obtained by protection and agglutination tests. G. Benjamin White.. 331
Properties of pneumotoxin and its probable réle in the pathology of lobar
pneumonia. Charles Weiss and John A. Kolmer.................... 337
Rapid simple method for the determination of type of pneumococcus in
sputum of lobar pneumonia, A. Charles Krumwiede, Jr............. 333
—— —— —— for the extraction of precipitin antigen from bacteria, A.
tren csmiGrunnwiede mmrsae yor eters fad Site ea as cme es eer ete 338
Relation of the meningococcidal activity of the blood to resistance to
virulent meningococci, The. Toitsu Matsunami and John A. Kolmer 319
Réle of immunity in the conduct of the present war, The. John A.
TACCOUITO ET a ‘ath gies a bee pch i Scusth Cin cae RAE IE aad: Oc RRS as 8 Av, 317
Simple method for blood culture, A. John G. Wurts and S. W. Sap-
DIAN SO LIM ETS Se eae Ne et ke ete He al nic tate se cle aa eis Mech cigs Saas a os tte 330
Skin reaction to pneumotoxin, A. Charles Weiss...................... 326
Studies on so-called cellular anaphylaxis. W. P. Larsonand E. T. Bell 323
— on the toxicity of pneumonic lungs. John A. Kolmer, Charles
Wielssranaalidwardgocelnteldr. ss. s.tti. ceca: os ccs e cists cee Hae sai emer 336
Study of controlled postmortem Wassermann reactions: a supplemen-
tary report on 400 cases, A... ‘Stuart Graves... ........ 0.05000 ecenue: 341
— of the immunizing properties of bacterial vaccines prepared after
various methods, A. M. W. Perry and Sara Levy................... 344
Types of meningococci concerned in epidemic infections. A. Parker
Hiatchenssands@. Les obinsOns yeee tte eek cise nee one ee eee 318
Waceme dosage.” doseph Head! 3) e602 is awa ontlneeee ss cence 342
— treatment of acne, with special reference to the réle of Bacillus coli,
The. Albert Strickler and Jay F. Schamberg........ . 342
Various immunological reactions in glanders, The. G. Benjamin White 327
Protein quotient, The constancy of the, during intensive digestion and pro-
One edesbAnVvatlOnrey emanate mkt. rine cee entre eae ooeitieioreus eyemsienn se 67
—— ——,, The non-influence of injections of trypsin upon the, in blood serum 139
—— substances, Complement fixation with................-2..-++2+e+esee 277
486 INDEX
Proteolysins and haemolysins, A note on the relation between.............. 295
Rabbit sera, Effects of intravenous injections of a colloid (gelatin) upon.... 147
Rabbits, Experiments on the production of antipoliomyelitic serum in...... 213
Relation between proteolysin and haemolysins, note on the................. 295
Robinson, T. Brailsford, and Hanson, Samuel. A new method of estimating
the antitryptic index of blood serum... 326: sa ooer np cia ee 131
Saponin-hemoelysis; A study: of. ....<42./02h sens eee eee nee eee eee 423
Sekiguchi, Shigeki, and Kolmer, John A. Experiments upon the passive
transfer of antibodies from the blood to the cerebrospinal fluid.......... 101
Serum (complement), The influence of active normal, upon meningococci, 157, 177
Skinvreaction to pneumotoxin, As). seeat ac 2 -ccloae = deheenee eek Cane 395
Small, James C. A method of preparing bacterial antigens................ 413
Smith, G. H., and Cook, M. W. The specificity of intracutaneous absorption 35
Solis-Cohen, Myer, Heist, George D., and Solis-Cohen, Solomon. The bac-
terial action of whole blood, with a technique for its determination .... 261
Solis-Cohen, Solomon, Heist, George D., and Solis-Cohen, Myer. The
bacterial action of whole blood, with a technique for its determination 261
Specificity of intracutaneous absorption, The....................---eceeees 35
Starvation, The constancy of the protein quotient during intensive digestion
and! prolonged as. .¢22 sakes eee ers a aan Ce eee 67
Studies:in ‘pneumonia: ...,452 7860.4 fe sale gone sce Oe ee eee eee 395
=——— onthe vantitry psink Ofserums.2j-620.cc 1s orice DO eee oe oan 51
Suzuki, Yoshio. The study of problems of immunity by the tissue culture
method. I. A study of the cells and blood plasma of animals which are
naturally resistant and others which are susceptible to diphtheria and
tetantis toxins 3.ire.t os lade Bronkiece ShisdoSoters se cee ene eee 233
— and Burrows, Montrose T. The study of problems of immunity by the tis-
sue culture method. II. The tissue culture as a means for quantitatively
estimating toxin and antitoxin and determining the distribution of anti-
toxin in‘passively ammunized ianimals....5.:5, .41,.4n¢- 260 320 4260 ee 219
Syphilis; On: Von Dungern’s indigo test for..: 2% ...)..9.4-5-2- - -- ae 11
Tissue culture as a means for quantitatively estimating toxin and antoxin
and determining the distribution of antitoxin in passively immunized
ANIMAl Ss: ney ewe pach lon oratectehece yeu ae tegen ISR eee ee 219
— —— method, The study of problems of immunity by the......... 219, 233
Toxin and antitoxin, The tissue culture as a means for quantitatively esti-
; mating, and determining the distribution of antitoxin in passively im-
Muinizéd. -animalses ty Ger..es-al dies 4 5.che-21) Sa ee aes A xa Oe 219
Toyama, Ikuzo, and Kolmer, John A. The influence of arsphenamine and
mercuric chlorid upon complement and antibody production............ 301
— —, Kolmer, John A., and Matsunami, Toitsu. The influence of active
normal serum (complement) upon meningococci. I. The opsonic ac-
tivity of fresh normal serum alone and in combination with antimenin-
Sibis, Serum! Lor IMENnINgOCOCCIERe en. a CeO eeeeee eeee 157
Trypsin, The non-influence of injections of, upon the protein quotient in
blood ‘serumits 3» 5053). ds.Go see eae ee tee a 139
Tsen, Edgar T.H. Experiments on the production of antipoliomyelitic serum
In: TAbbUS ech col acs shearek SAS eae Mes eer aerate cio ee eee 213
INDEX 487
Tuberculosis, A contribution to the study of the complement fixation reac-
BTEURETL yee Pe Se hee At LAN Go Giah ae nn 0 fad ws SENG re alalel vig SI 345,
Typhoid-antityphoid serum complex, Extracts of antibodies obtained from
specific precipitates Of............ 6... s eee eee teeter tenet eens
Ulrich, Henry L. Experimental pollinosis. Preliminary report
Uveal pigment, The immunologic properties of...........-.--.--.+...50e05-
Von Wedel, Hassow. A contribution to the study of the complement fixation
Bete eTMEGTA UO GLCUI OSES: 05/55) to tents tafe bie oetale sual ietaene wie Gain. etsle in ysiel er
War, The réle of immunity in the conduct Of thepresenite fae. Sec oa ate se
Weil, M.O.R.C., Major Richard. .........-..-- 2... sees eee eee eee eens
Weinstein, Israel. Extracts of antibodies obtained from specific precipitates
of typhoid-antityphoid serum complex.........-..- + +065 0+e essere eees
Weiss, Charles, and Kolmer, John A. Studies in pneumonia. VIII. A
skin reaction to PMEWMO CORUM ous vas awe) s sic + ree eames os she aie «
Wenner, John J. A note on bleeding guinea-pigs and on preserving sheep’s
EEyMAEGCVES ss. asi nicaie(s Ue amc ood ass | alate ne # cle eeya Fagus o> ieinrs vin
Whole blood, The bacterial action of, with a new technique for its determina-
PATIO esa feat te re UNE sch thon oseleltytel chet steve nieteue at oravelaie Ml # mime
Wilson, M. A. A contribution to the study of the complement fixation reac-
PAM MIDE CULOSIS » «62's... Saline s sleaie oo es cle ee leininiels = Bliye sc mie ese aie spel
Woods, Alan C. The immunologic properties of uveal pigment
oc
-*
Shee
tie
AT yu Shag te
Weta ms si
CR The Journal of immunology
180
J6
Vv 23
cop.2
Biological
& Medical
Serials
PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET
UNIVERSITY OF TORONTO LIBRARY
Te Laneks batabes
Ready
Lett ears
wr .
sete rsnina® bad an aiodiad
*
aru est see 1e Sr" Litt
“ree SAcaaaed
++ mae aay ee
eo
4
.
whe Bi dm a te
Cede aes
Coase Surin
*
eee
ae
4m
oFeteaelt cr)
haa
rep Ce ee
+ 2 sn eat tee se en
esr erarye
+
fi ; Alaa.
. Sime
22
OO. or waurrssys
2 er Ee a SOIT
5 -
7 TASedes
2 ‘ Tata
ona . oe
- io
Nota ata : .
ated
pie.
*
eet
: see
“htt sees state?
-_
9 me au
sha ae
y
-
Sen
aa sets? oe
? Sete wha’ : coi
BP aes 3 | | | |
a
=
wat
Tepieseysotety oe
: ne
Cee y ‘ fet,
ese
+e
Wot
Corners . ae stake 5 .
res
fe
ne
at 5 Ste
aatie! yet Wah eet
ha tatae,
Ferhe8 oes.