19ZO
THE CHEMICAL NATURE OF THE ANTIGENIC
SUBSTANCES IN BACILLUS COLI
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy in the Graduate
School of the Ohio State University
BY
EDWARD EVERETT HALE BOYER, B. Sc., M. Sc.
The Ohio State University
1920
' ' :- '• '.:V\.: :. • ;•
THE CHEMICAL NATURE OF THE ANTIGENIC
SUBSTANCES IN BACILLUS COLI
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the
Degree of Doctor of Philosophy in the Graduate
School of the Ohio State University
BY
EDWARD EVERETT HALE BOYER, B. ScM M. Sc.
The Ohio State University
1920
I
-
EDWARDS BROTHERS
THE CHEMICAL NATURE OF THE 'A'WTTiGE^O, ^B^'AK
IN BACILLUS COLI
Introduction.
Certain substances when taken into the body paren-
terally will cause the production and appearance of
protective substances within the body fluids. The pro-
tective substances we call antibodies. That which calls
forth the production of antibodies is known as antigen.
When antigens are allowed to enter into contact with
their respective antibodies, either within the animal
body or in a test tube, definite changes are observed;
thus we have manifestations of these changes in the phe-
nomena of agglutination, precipitation, complement fix-
ation and other serological reactions.
But of what do these changes consist? Are they true
chemical reactions which follow the usual laws of chemis-
try, or are they due to alterations in surface energy and
molecular attraction in a physical or physico-chemical
sense? And what is the nature, chemically, of the react-
ing substances?
%
Most authorities state that all antigens are protein
substances. Such statements, as will be shown later, are
based to a considerable extent upon prejudice and precon-
ceived notions which are not entirely borne out by exact
experimental procedure. The substances most generally
used as antigens are body fluids (blood serum), body
cells, and bacterial bodies. These substances, in the
dried state, are composed largely of protein and mineral
salts, with a great predominance of the former. Fats and
lipoidal material are often present in such small amounts
as to escape detection unless large quantities of the
native material are used in the analysis. Since the pro-
teins seem to be the important constituent of substances
used as antigens, and since such substances may be used
with equal results after the salts have been removed by
dialysis or other method, it was but natural to consider
the proteins as antigenic principles.
A few investigators have endeavored to demonstrate
the value of fats and lipoids in serological work. Thus,
Jobling and Peterson (1) showed that when bacteria were
injected into the blood stream they absorbed lipoidal
material which, ordinarily acts as the anti-enzyme con-
stituent. Noguchi (2) showed that the tetanolysin frac-
tion of the tetanus toxin was neutralized by cholesterin
and alcoholic extracts of blood serum. Muller (3)
563911
i c I.* «'«: .v«
c ' <.' « V « ' ' ' «
,=4
.
T£^fc;.!f;?
r.
r. r r,
Vvr f s r
demonstrated that the action of tetanolysis (which is an
antigen) has to do with lipoid substances, and that
alcohol- soluble lipoids inhibit the action of tetano ly-
sis; thus confirming Noguchi's findings. Landsteiner (4)
found that ether extracts of red blood corpuscles were
capable of neutralizing tetanolysin. He also showed that
ether- soluble lipoidal substances were involved in serum
hemolysis. Bang and Forsmann (5) found that hemolytic
activity was due to lipoid substances in the stroma of
the red blood cells. These lipoid substances were ob-
tained by extraction of the cells with ether. Upon
analysis they found the extracted material to be composed
of lecithin, cholesterin, a phosphatid and a cerebroside.
This material, when used as an antigen and injected into
susceptible called forth the production of hernolytic
amboceptors. Working with various serological reactions
Kyes (6) (7) demonstrated that lecithin was the ingre-
dient which played the part of complement. Sachs (8)
studying hemolytic reactions, determined the importance
of lecithin in hemolysis. Pick (10) obtained a precipo-
tinogen by trypsin digestion of egg albumin, and he was
unable to demonstrate protein in such an antigen. Some
workers confirmed this finding, while others were
unsuccessful.
The results obtained by these workers have not gone
unchallenged. Nor does their work necessarily indicate
anything concerning the chemical nature of antigens. It
has been of value, however, in demonstrating the fact
that there is a relation between fats and lipoids, and
antigens. As to what this relation consists of has been
quite obscure, but the evidence is strongly in favor of
the theory that serological reactions are manifestations
of changes in surface tension and molecular attraction,
and not true chemical reactions according to Ehrlich's
idea.
The chief objection to the antigenic nature of fats
and lipoids has been that the investigators overlooked
the possibility of a protein constituent being present.
Thus the other extract, or other extracted material,
might have contained protein material unsuspected by the
workers. Furthermore, the antigens may have been not in
a pure lipoid state, nor in a pure protein state, but
rather in the condition of a complex conjugated lipo-
protein which perhaps was taken up by the solvent. Thus,
in their studies on* anaphylaxis, Pick and Yamanouchi (11)
used, as antigens, alcoholic extracts of horse serum.
This extract was evaporated and redis solved until it gave
a negative biuret test. They were successful in using
ad lo noxdOB srict dBfl
.-.a jSeoflBtfedire MoqzX rfdx'v of) od sBri
:o rroidoB orfd dxdirfnx ebloqxX alcftrXo
iicJ . sgnxxxill e'irfo^cH gnimiilnoo
Mcrioo booXd boi lo sdoflrrdxo isrfd©
:3 oeljs eH .nieYlonsctstf srrxsilBirJwert lo ol"^-
Kwlr -lew eooaiiuedjje iB.Moqil slcfi/Iog-it)
'omoil ;O'l (c) , Lnc gnr.H .r.ia^fo;
crro-irfe r ^somjrfeicfj^s : o^ eu,r» eiiw Y^^V^'
o-7ow -oad;fc f>ioqll o- .sllso hoolcf
noqU rf^ tw elloo srfct ' lo nolctO£a^xo
7:000 8 • iie^cm t .7x0 9/i^ bnuol ^e/f^ a^rYt
• fci jBrlqfeorfq « ,f .olorfo ^nirf^iooi
fain n^sidfiB HB BIS .-!£»rfw ^XjsiTO^Bn e
*/i lo noi^oi/boiq orf^ / o eldictqer
.oigoIoiDs supiifiv rfj .>(ioW .aio^qeo<
8flw nJtifcJtDfri ^»f£^ t- -aomt)f) (V) (d) <.-
:i . Jtrremslqmoo h9Yfi!q rfoirfw dr.-
'.t i)©r. lomori ^rtiyb
Fo (OJ) .;• f ni nirf^iool
'.i 339 lc rt$ ^cf
rtc HB rtetrs ni ^ o^
*•• '
•• .-.
•-390*ifl >faow ioPt ..JbonrroXlGfioniJ
::B lo eir/drn Xn?; K.l^rru
gnfdfiidenonrob . ev r>. nosri 8B/i
f'uB cdBl I. B 8^
.00 rioi
"3 ex : F ti
&ris enoxdoBtsi :
t ^OB lorn brtB : ni aagrterfo lo
•;.COOB O"''i.4 -'^';f:
arfi/d'Bn oxnssxctrtJB erf^ .j;rfo orfT
-lovo c-iodJBglJesjvnl abioqxl brra
- 3ni3ff dnotfcfi^8;to- xdxseoq1 ertt
odern f\9^0QTdxQ -terfdo lo t:^' i to odd
oeq&w2fnr iBii^dBm . niscr-o'
r no.fjd evB/l ^sra &f;o^i. ^
r ,edrde nie^oiq ortxjq I oiuq B
i-jedfiJjnoo X3lqraor.
qif n>: :
' te;
2 eatorf lo ados
rij bovloasiboi Jbr
•iL IiJleK^oow? s-xev/ v;ojiT
such an antigen to produce anaphylaxis. In a similar
manner Bogomolez (12) used alcohol and ether extracts of
egg yolks to produce anaphylactic phenomena. It is
known, furthermore, that the solubilities of proteins
and other substances may be profoundly altered by the
presence of lipoids. The existence of a protein- free
antigen, therefore, was not established until the recent
work of Warden (to be considered further on).
There is, however, one exception, namely the
protein- free antigen of Ford and Abel (9). These men,
studying the poisons of Amanita phalloides, found that
the toxic principle was an active glucoside. Their find-
ings have never been refuted.
Those who favor the protein theory of the nature of
antigens are prone to criticise all experiments which
attribute fats and lipoids a role in antibody production.
The one chief critical point lies in the face that the
lipoidal antigens have not been satisfactorily proven to
be protein- free. On the other side, however, is it not
fair to inquire if the so-called protein antigens are
fat- free? Take practically any native antigen for use in
antibody production, and we must concede that there is
as definite an amount of lipoid as there is protein; and
merely because of the proportionately greater quantity
of protein is no reason at all to assume the role of the
latter as antigen.
Numerous investigators, assuming that a given sub-
stance was an antigen, have endeavored to separate the
protein material and to determine what particular frac-
tion of the protein served in the capacity of antigen.
In the beginning they find that the substance under con-
sideration is a true antigen. They then, as a rule, pro-
ceed to separate the various protein constituents either
by fractional precipitation or by enzymotic digestion.
And finally they arrive at a point in the analysis beyond
which antigens are not found. In reviewing this v/ork one
is surprised to find in how few experiments the fat and
lipoid constituents have been eliminated. It is possible
that these lipoids were sufficiently bound to the pro-
teins as to be precipitated along with them; and due to
subsequent procedure, during the separation of the vari-
ous, fractions, the fats were liberated and discarded, or
remained in combination with higher fraction. If the
antigenic principle was lodged in the fatty portion, this
theory would, -of course, account for the results ob-
tained. Similarly, in analyses by means of enzymotic
reactions, the fats which are present may be digested by
. «
d ct n*I:
,1
oeioq orfd j'.i-rivf:
ns a«w oJqiontiq oixo?
,bOujjl3'T noocf iov:;n ovjsr? :
•J lo \Toeri^ nir?d-oiq arf^ rrovBri o/fw c>e-orfT
JI'jc? os o o;t onoiq 9i.e en^^-i
sfoi s -il £>ftB arfsl 8:ti;ol't
1.L- esli dn Ilio lairio eno
^£2 rtood . anogi^nB i.r.bi-'al V
•f* nO .ootl-rrie^oi ocf
bell; &rf^ li 0-f.f-jjpr:.' od
-too ^Bfjnt ew hn& ,noirfoi/J .
•* ££» i lo 3aJL'OfPB «£
1 'Jo 9 £' If-tf 0 9Cf T£ f 9'
o3 JXfl ?JE? nosf:f>T on <.-. j: niod'otq lo
•v Jji •> J- -f rf o r> (* tf t . a -T
.1
' ovsrf t -.& HB a
T8v1ef} otf bns Isiiodjs
rio siftf fti beviea nleJoiq arfd1 lo £io±
; 3 /lit ^flfl^ ball .raigsd 3r(J ril
fi sfl tn©rf^ vorfT .rregidnB owirf e ei
ni a do -iq e i/o -i IB v
• is Y:^ *^^ noi^B^x*
:.t drcioq A ^a c-;vif*UB \
3n^weivo'£ nl .fMttrol cfon eifi
e^ne*. v/orf ni haJtl
C .-t.1 ,.f -Ml/e 119 OCf O-VGrf :
lo rroi . .- 3/iol gniiifb to. ic ^nGj/pf>ccrr.fs
lodcl OT^W ectG'i srfct t2noict'orf'3:l BC/O
,fTOi^Ofll ri^
Y^ '-tcl >J" ."TBW °fq ! oino^j
v<r YTC-
lipase. Or the presence of aatabolic materials may even
produce a change in the configuration of the fat mole-
cule. There is no doubt that some fat -free proteins are
antigens, but before we eliminate fats and lipoids from
the class of antigens, we must first eliminate them from
the various protein substances used as antigens.
Although it would seem that some pure proteins are
antigens, it would also be equally apparent that other
pure proteins are not antigens. Starin (13), for
instance, working with a purified gelatin, was unable to
demonstrate antigenic function. It has been suggested
that antigenic properties of proteins rest with the aro-
matic radicals attached to the amino acids. Perhaps
these radicals have some affinity for fats and lipoids,
so that in the absence of such radicals there is also an
absence of fats; i.e., during the process of fractional
analysis those portions which bear the aromatic radicals
are" split off with fats, and the lower fractions, being
fat-free, are non- antigenic. The value and importance
of the fatty acids in treatment of diseases due to acid
fast bacteria have been demonstrated for several years.
Thus Chaulmoogra oil, sodium salts of the fatty acids of
cod- liver oil are used in treatment of tuberculosis and
leprosy (14). Walker and Sweeney (15) showed that the
fatty acids obtained from Chaulmoogra oil are specifi-
cally bactericidal for acid-fast micro-organisms, but
not for non-acid fast micro-organisms. Similarly,
Hollman and Dean (16) demonstrated the theraputic value
of esters of the fatty acids of Chaulmoogra oil in the
treatment of leprosy. These findings lend strength to
the theory that fats are intimately concerned in
immunological reactions.
The most interesting work which has been done on
the antigenic nature of fats is found in the recent
experimental data contributed by Warden. This investi-
gator has been studying the problem from the experimen-
tal standpoint during a period of several years. In
1915 he suggested, as a result of his studies on the
relation between bacterial fats and proteins, that some
fats are in a lipoidal combination with protein and are
not hydrolysed until the nitrogen portion is thoroughly
broken up (17). In the same year, reporting further
results, using the gonococcus as the source of his
material, he found that the organic nitrogen of the
gonococcus did not seem to be altogether available as
antigen. The fats of the gonococcus, chemically iso-
lated, possessed a much higher antigenic power (18),
Later, (19) (20), it was found that the fats, as
o ': t
&
n to B
^ ) X
•
•
1 1 i .-. t\
." tfi ^ene^i
3 .BiTOr^rfrfjs o4on QIC srtfo^o-ic ^
low ,sorr»c
•B i ^Prx;(r.frr- .
>£i;fn# d.
C1 /. •
•:ol Y IbB-- or
r( i^ fti 3&tit
Brt lo f*onc
. rftf riiE*'
ow.
;i;.
fctejoi
•
'
1o e-
,,
'
glycerol estus, did not cause as great a degree of
antigen reaction when used in serological tests as did
the usual gonococcus antigen; but by saponif ication of
the fats with subsequent .isolation of pure nitrogen- free
fatty acids, the serological tosts gave better results
when such fatty acids were used as antigens. Continuing
along this line of investigation, Warden (21) determined
that the neutral fats of the gonococcus were of little
value when used as antigens in complement-fixation tests,
but very excellent and specific results were obtained by
using the fatty acids, or still better, the alkaline
salts of the fatty acids. Due to such findings he con-
cluded that the important factors in such reactions were
not only the empirical chemical constitutions of the
respective antigens, but that the molecular configuration
was as significant and as specific. It has been shown,
furthermore, by the same author (22) that if the fats are
added to a "solution" of colloidal cholesterol the degree
of dispersion of the molecular aggregates of the antigens
is greatly increased, thus presenting more surface and
producing a more active and sensitive antigen. This
finding, of course, adds confirmation to our ideas con-
cerning the colloidal reactions in immunology. By simi-
lar methods the same author demonstrated the antigenic
nature of the fats from Bacillus typhosus, Bacterium
anthracis, pneumococci, streptococci, and red blood cells
(23) (24). These fats were used not only as antigens in
experiments "in vitio", but were injected into suitable
animals, with subsequent production of specific anti-
bodies. Moreover, the antibodies obtained were not only
specific, but were also protective against a dose of
homologous bacteria which was fatal for a control animal.
The careful work of Warden sheds a vast amount of
light on the question of chemical composition of antigens.
His results have: not yet been confirmed, nor have they
been disproved. No one will deny that certain fat- free
pure proteins cannot be used as antigens. But no one,
on the other hand, can maintain that all antigens are
protein.
The object of this work is to study the chemical
nature of the antigen or antigens, of Bacillus coli; to
determine whether such antigens are protein, fats, carbo-
hydrates, salts, or a mixture of two or more of these;
and to analyse, if possible, the antigen in order that we
may know of what it' is composed. The experimental
methods and data are set forth in the succeeding pages.
rrrdw
fog
si/oecfi/a rftfJw acte'
ni
nl b
i; u'fsw el^iofi ^ddal rfoue n
xi ^o 0nil 3irl^ Tif
srfct lo etel. Isirfi/on orftJ c<j.
ej IB SB £>oeu aorfw di/
.
^e to ,
I*. ,r ijel 9f{^ lo
• -t no
• I
.
t cfB/fd (S2) r-
-..sforfn .1
> E9C? lO
9icm 5/:J:J.i: ,-
JHB ovi .r . B -r'jiom
GJ rroJtc:.- ^ee 11/00
•: ni s
idfi ji-
ns ea Y-t*"10 3rfT
• rti"
jrfj
•
,:?
no
•
' JB
.
EXPERIMENTAL WORK.
The strain of Bacillus coli which was selected was
taken from a stock culture. To confirm the identity of
the organism it was subjected to the following tests;
a. The organism was a small rod- shaped organism,
with morphology, as regards the usual criteria, typical
of Bacillus coli.
b. It was negative to Gram's method of staining.
c. It produced abundant acid and gas when grown in
nutrient broth containing one per cent lactose. It pro-
duced acid and gas also in dextrose and sucrose broth.
d. It did not liquefy gelatin.
e. It was readily agg3.utinated by high dilution
(up to 1:2500) of B. coli antiserum.
In order that a sufficiently large quantity of
material be obtained, the cultures were grown in liter
flasks. About thirty flasks were used at one time. To
each flask was added ten grams of peptone, four grams of
sodium chloride and one liter of tap water. This titra-
table acidity of such medium varied for each lot, but
usually fell between 1.5 to 2.0$ acid, using phenolphtha-
lein as indicator. This comparatively strong acid re-
action was not adjusted, because, as the Bacillus coli
produces abundant ammonia in sugar- free protein media,
the ammonia gradually neutralizes the acids present.
This particular species usually thrives in a medium whose
initial reaction is not greater than 2.5$ acid. On the
other hand, its growth ceases when the reaction reaches
an alkalinity of 2.5$. It is obvious, therefore, that
with a comparatively strong initial acid reaction, the
longer will be the time interval before the maximum alka-
line reaction is reached; and presumably, therefore, the
greater will be the number of bacteria produced; and it
is a large number of bacteria that is necessary for the
work. The flasks were not plugged with cotton, but were
capped with three thicknesses of wrapping paper. They
were then sterilized in the autoclave for two hours at
fifteen pounds pressure. Then they were allowed to stand
at room temperature for a few days, at the end of which
time were discarded any flasks which revealed the
presence of bacterial growth. The paper caps were next
washed with bichloride of mercury solution 1:1000. The
flasks were then inoculated with a broth culture of the
oriv r'O cT . '0 ifootfe B fflO'rl ns:.
-.fiiwof... .'rf oo' &83o9t<fcr£' efilv ^
-01 IlBrae e BBW rceJinB'io SffT .
c.i- • r: s£ .Ygolorfqioni d(
.Jfoo aulllo
bo.^om e'teBiD o^ 0v.tct.c38.n esvv crl ,d
.bne blOB
"^^ri9o ^aq e . rs^noo rf^cid snori
../a Jbrt£, caoiJxob Xe cs Jbnfl MOB fcer>
T 3 on ttk dl ,b
-\jcf 1)3 ^.sni vl.cbfioi QBW ctl .9
.nurr>E.r .a 'lo (OOeiLJ:! 0,1
: -a B cJBrftf tabTO nl
••?: BO- • rJ- .banir^cfo scf lBi<r.*K
%
1 JB best' 9iow ejiasll • ifwocJA .a>fc:
1 ft&flO^qoq lo sr vact f>9bbB BBW ^eBll fl
4 lo sfiiioJHo mwi .
TOl -' 1C Y^i-i°fi 9J^
2 o3 6,1 no^w^ocf Ilo't Y^-^6t
n .lojrjoibni EB
SB ,8 B ^on BBW nor.:
•TirrB cfnBbnads eoour'
Hie s ' -B tfiflpntr.TB •
•'iirlif '^u sftlooqa ^BluuicriBq p. •'
nBrfrf ie>ctB9'i3 cton si no I
•
;uoivdb ei cH .^o.S lo \
I B 1 3 in 1 SHO i -i v i 3 B r
>lecf .
"urcoTq bnB \L. ei
9
3Bd" lo lacfrtxra sgiel .
?on ©tew ssfesil erfT ,>(
tt rfaiw £):.«••
:uorf ow^ riol avBl^Owi/B srf^ nl . rrarf^ 6 -
bewof/B 3iew Y»-rf^ n-. •*' utjq nsac'
^B ^«Y^ w^*i B. 10! f; 9d" moort
r>isv; on
.q 8<i?
. oa Y'j
^d B -
Bacillus coll. The inoculations were made by the use of
a hypodermic syringe and needle. The broth culture was
aspirated into a sterile syringe; then, after inserting
the needle diagonally through the paper caps, a few drops
were forced into the flasks. This method has, to commend
it, the advantage that the flasks are never opened for
inoculation after having been sterilized. The inoculated
flasks were then incubated at 37°C. for about ten days.
In order to separate the bacteria from the fluid,
the contents of the flasks were run through a Sharpies
laboratory "supercentrifuge". This machine is very simi-
lar to the ordinary cream separator in construction and
mechanism, but is run by steam or compressed air. Suffi-
cient force is obtained to completely separate out all
particles in suspension.
This material was washed with sterile physiological
sodium chloride solution and recentrifuged. The sediment
consists of bacteria with precipitated salts, sulphides,
and perhaps other amorphous material. Most of these
foreign substances are easily removed by filtration
through cotton.
The remaining bacteria were now subjected to saponi-
fication in potassium alcoholate, at a temperature of
100°C. for one hour. The resultant material was then
acidified with dilute hydrochloric acid. It was expected
that this procedure would yield a definite layer of fatty
acids which could be removed, but such proved not to be
the case. Instead, there was a mass of solid material,
including fatty acids, proteins and salts, in a state of
very fine suspension, which could not be separated by
gravity or by centrifugal force at three thousand revolu-
tions per minute. It was then decided to extract the
fatty acids with ether. After shaking up with ether it
was found that only about ten percent of the ether could
be recovered and that only when comparatively large
amounts, were used. Most of the ether became a part of
the suspensoid mass. The entire mass was then trans-
ferred to a ten- inch porcelain evaporating dish, and
evaporated at 37°C. for several days, until the moisture
content was very low (too low to support the previous
suspension). The residual mass was then shaken up with
ether and the latter separated, removed and evaporated
at room temperature. The residue consisted of a very
minute amount of crystalline and amorphous material,
presumably fatty acids. The amount was considered too
small to work with.
•
ahem aiew ancldalL'oonjt erfT . Ifoo otfli:
. slbssrr Jbne ?•;>.; ft ity 2 otmoboqv
;:j£alT£!? '.£>3e B oiai bectQil
s "B'. 9fi5 ffsj/c'Tfftf ^Uflao^cib olbsen
Si-ifT . e:.rcBll erf;* ocfr.i booiol c-
QO i
nai
'is^8 nsod gnJtVB/f loctls
0°?5 ctB b&ctBcfuonl norf^
flcf orfj 9cJB«iBq?s oj
eii
fq eliisct'E .dd±\v I "/ EBW
.bo3J/iJ:i^rr90a'r fc-rr*.
.rBc; joncreilqioaiq rf.ltw
'. .iBi-ioctEm
nl
'
axrfT
efiJt-iofrfo mw.r
lo G.jgicrroo
aqBffioq
nsis1
i'
_
cts t IB m01aefijoq ni
,1-JJOrf 8HC toU ,0C''
o.- stfjulifc rfdiw t)3±lii
I ;? bS'iiy bluow aaubsoo'tq sxrict ;'
ad bli/oo /loirfw el
cena B e#w eiorf^ tbBe^eaI ,oeco
3 bru: ^oriq t&bioB Y^^^^ $fcibtfioni:
'oianeqet/E snil :
•T^nso Yd ^o Y^-^
rosb ns/iJ E .sctuftxiT! Tsq sn>
:'srfe •le^lA .leflite rf^lw ebioB Y-
9'rl3 " )0'ieq n^cf :; -;;I«o rfcrfd bftj/ol
jr Y-fs^-t^ 3 £>£TB boi^vooea
B 3C1L , .
j. trfeJcb &n' .^sleo -^ B o^ bc>r
*»rf^ lief ^B bs^B-ioqsve
?'iq exf^ -'2 o^ v.'ol ooct) wo I ^^V r:BW vine:
n! Ej
no
nerl eew
bovoms
8 .{noxeneq;
a acri;
8
Another thirty liter mass was prepared as in the
above experiment. The bacteria were washed and separated
by the same method, but instead of being saponified, they
were treated directly with ether. This ether-bacteria
mixture was allowed to stand, with frequent shaking, for
six days. The ether was then pipetted off. The residue
was again treated with ether in the same manner. These
two ether extracts were then evaporated. The residue
consisted of brownish-colored fats, some of which ap-
peared crystalline, other amorphous. This material was
washed with water, then taken up with ether, the ether
removed and evaporated. The weight of the resultant fats
amounted to approximately 0.4 gram. The fats were then
saponified with potassium alcoholate at 100°C. After
acidifying with hydrochloric acid, and cooling, a defin-
ite layer of fatty acids was obtained. The alcohol was
evaporated to nearly dryness and the fatty acids extract-
ed with ether. This ethereal extract was concentrated to
dryness and the fatty acids taken up with alcohol. This
was treated with sodium carbonate to convert the fatty
acids into the sodium salts. The sodium salts were then
drystallized out and redissolved in 100 cc. of alcohol.
This constitutes the fatty antigen.
It is desirable, at this point, to examine the
antigen to see if any protein material is present.
Several different tests were used as follows:
a. The biuret test. To three cubic centimers of
antigen was added an equal amount of strong potassium
hydroxide solution. The mixture was well shaken and
treated with a few drops of very dilute copper sulphate
solution. Absence of color change indicated absence of
protein.
b. The Kjeldahl method, 10 cc. of the antigen were
mixed with 20 cc. of concentrated sulphuric acid. 0.2
gram of copper sulphate was added and the material was
boiled gently for ninety minutes. It was then cooled and
diluted to 250 cc. with distilled water, then neutralized
with hydroxide, adding a slight excess of the alkali.
The material was then distilled, the distillate being
collected in 25 cc. of N/10 sulphuric acid. When one-
half the liquid had passed over, the process was stopped.
The acid solution was then neutralized with tenth-normal
sodium hydroxide, using congo red as an indicator. An
equal amount of alkali was necessary to neutralise the
acid, thus indicating the absence of nitrogen in the
antigen.
BJBW
lB8«XBffef9 cirfT
ct efcioe ^r^^al
ocfcfio fliifi£)on
fcns
Jbns f>ovc
B o3 bsrtntromjs
iw l>9Jllno<
lo
I/sctf-
rr*
i ej
1c
erfT
: ..: ^nlhftr, ,9
t: elb noif.1 BBW
Jblo£ oi'iirrlqli/E
eootq oxfcf ,i9Vo
T^t lo .00 6
csnq fcsrf
c. The ninhydrin test. To ten cubic centimeters of
antigen was added 0.2 cc. of a one per cent aqueous
solution of ninhydrin (triketophydridene hydrate). The
mixture was boiled for exactly one minute after the
appearance of the first bubbles on the side of the tube.
No color change could be detected, even after cooling
and standing for four hours. This indicates the absence
of alplia amino acids.
d. The cyanide test. About three cubic centimeters
of the antigen was evaporated, and the residue fused with
metallic sodium. The fused mass was placed in a small
amount of distilled water, boiled and filtered. To the
filtrate was added a few drops of fenous sulphate, fenie
chloride and hydrochloric acid. Absence of a blue colora«
tion indicated an absence of nitrogen.
It is evident that we are dealing with a nitrogen-
free substance.
The original bacterial cells, after being doubly ex-
tracted with ether, were then suspended in 500 cc. of
sterile physiological sodium chloride solution. For
preservation, sufficient carbolic acid was added to make
a O.5$ solution. The greater part of the solids in this
mass are supposedly bacterial proteins and for experi-
mental purposes the mass is called the protein antigen.
Obviously it is a bacterial antigen minus the fats.
For use as a control there was next prepared a
suspension in physiological saline, of a twenty-four hour
agar culture of the same strain of B. Coli, This was
made up with 0.5$ phenol and it constitutes the B. coli
of antigen.
We now have three distinct antigens:
1. Sodium salts of the fatty acids of B. coli.
2. B. coli after being extracted with ether.
3. A simple suspension of B. coli.
It is the purpose now to determine whether or not
these antigens are true antigens; i.e., whether or not
they give proper antigenic reactions "in vitro", "in
vivo", or in both. To this end the antigens were inject-
ed into suitable animals, at certain intervals, in order
to obtain antibodies.
The fatty antigen was diluted with physiological
salt solution, approximately one part of the alcoholic
solution to five parts of the saline. This gives a very
oi-duo fistf oT . J?;0d" nl'ii-Yrtniit erfT .0
>o leq eito a lo .0.0 ^.0 fcofcba sew nestc'ris ericf
.{ aae. i'qc;fre-/-: -/.'I;*) ntibyiinhi 'to no-tcfuXoe
>G ecfi.fnlm 9no ^Uouxo 10! b^Xiocf sew 'rvrif.ix.tai
Ic eb.t3 eritf no s-aldd^cf cfsifl orto lo esruaiB^
19.116 rrove tJbocto4cf'oJb ecf bli/oo a^asrfo ^oloo oH
% srf-? r:eviBO«tJbai atrfT .stnorf ii/o.l nol snibne^c bnfi
. ebJiOc oni.Tuo siIc.Cc lo
prf^ lo
Jt£jorr!2 i. Iq BBW ge^m besi/1 srfT .nxi/ibos o
rfct oT . ..:IJ:1 brtB bell ad ,Terfsw b»Il.tdEib
ao1 -i et/on»l lo sqoib WQ! B bobb£5 EJPV/ e^
loo 9i/Icf /? lo eorr^ecfA .bioe oiioIrfpoitY^ 'bns
lo soneede ne bs^B-oihrti nol
.! a ftetvr SnilBeb 9-10 aw 3fid3 ^nobive si ^1
, o o ne ^ a nut 9
:;ob
.00 OOfi ni bebneqejL/e n^rf^ 8*rew ,terf3e rf^lw br^-t
•1 .noi^i/foe sbliolrfo rajjtbos Isoi5.oIo.tFY^c
o^ bebba SBW bioB ^tlodieo vinoioiliys ,no.t j
: rrt BLiloe orfct lo 3ieq 13^^013 9dT .rroi^i/Joc 6.0 B
-iioqxe 10! bhB snJt^Joiq .Ifiiiocfoed Yibaeoqqcfa STB aRjsm
nisjcnq orfd bollso el eaem 8ii3 caeoq-Tiiq iB^neni
• '! ft-^^i i" ~~ flB ' ' Di?(J B K J! Ct E vX3ir— ;
a bsiflqsiq axun BBW oi^ri^' Xoi^noo B SB 931;
ijjol-y^n9Wv1 & lo tefiXXB8 iBOlgoXoie^rfq rtl
r etrfT *IIoQ .8 lo nXfiicte emi-.s orfd lo '
bnis Xonoriq ^6,0 rfrtxw
J3 10
sveii won
ebioB y33Rl orfct lo e^tXBs muiboQ .X
w be^ofiitfxd snldcl leftB IXoo .fl .2
.iXoo .8 lo ffoXafieqeire oXqmle A .5
w<>n aaoqiuq erfo
n TTO i&iiJ9ftw ,.a,l ^snejifctna emd aid Ens^t^ns ezhrtri^
'' ."oictlv hi" -enoi^aflei aXaftglinB laqoiq evlr* yod^
oiew en^^iJnfi siirf bne sirtt cT .xi^od ni 10 ,"ovjtv
mirre oXdu-llye o^ni bs
do oit
>^o'osTtiq 'w e^jjb aruv nsg^ns ^^^el a/fT
iXoiiooXf o rid. lo.rftaq sno Y^9^B^i^^^qcts tnoidi/Ioc
lo erfits evil o
10
opalescent suspension. The protein antigen was used
without further dilution, as was also the control B. coli
a'ntigen. The course of the immunation is given in the
following proctocals:
Table #1. Rabbit injected with fat antigen.
Time
Dose
Weight in
Remarks
interval
in cc.
grams
0.25
1770
Prompt recovery from injection
5 days
0.5
1800
» M i» M
6 days
0.75
1960
Mild signs of shock
6 days
1.00
2040
Prompt recovery
12 days
0.00
2100
Animal bled
Table #2. .Rabbit injected with protein antigen
Time
Dose
Weight in
Remarks
interval
in cc .
grams
0.25
1630
Moderate shock.
5 days
0.50
1635
" "
6 days
0.50
1680
Severe
6 days
0.50
1695
Moderate
12 days
0.00
1730
Animal bled
This animal was the third to be used for protein
injections; the other two died immediately after the
initial injection, thus demonstrating the possibility of
a toxic fraction present in the protein mass.
Table
Rabbit injected with B. coli antigen
Time
Dose
Weight in
Remarks
interval
in cc.
grams
0.5
1690
Prompt recovery
5 days
1.0
1700
ti ti
6 days
1.0
1700
ii it
6 days
1.0
1850
it it
12 days
0.0
1910
Animal bled
The injections were all given by the intravenous
method, using the prominent veins in the ears as the
sites of injection.
n s
rt:
eo
cs . u
a.o
ev.o
00
00
.,L_
00 nl
es.o
oa.o
00.0 E
O r
1 .00
OJ
11
Twelve days after the last Injection each rabbit was
anesthetized with ether and by the use of a sterile
syringe and needle about forty cubic centimeters of blood
were aspirated from the heart. The blood was placed in
sterile glass tubes, which were then placed in the ice-
box until the serum separated from the clot. The clear
serum was then pipetted off into sterile tubes and an
equal amount of glycerol was added for preservative.
Each of the above antisera was then tested against
the three different antigens. The tests used were as
follows:
a. Complement fixation
b. Agglutination
c. Precipitin
d. Anaphylaxis
a. The complement fixation test. In testing the
antigens and antibodies by the method of fixation of
complement, guinea-pig serum diluted with ten parts of
physiological saline was used for complement. The anti-
human-rabbit hemolytic system was used; i.e., the blood
serum of rabbits which were immunized against human red
blood corpuscles was used as hemolytic antibody. The
following table sets forth the procedure for obtaining
the hemolytic antibody:
Table #4. Rabbit injected with hemolytic antigen.
Time interval
Dose in cc.
Weight
Remarks
2.0 cc.
1680
Prompt recovery
4 days
3.0 cc.
1725
U 11
4 days
4.0 cc.
1755
it it
5 days
5.0 cc .
1790
ti H
10 days
0.0 cc.
1835
Animal bled
The animal was bled and the serum was obtained and
preserved in a manner similar to that previously des-
cribed. The hemolytic system was titrated as follows:
the antigens consisted of a five per cent suspension of
washed blood cells; the complement was guinea pig serum
diluted with ten parts of salt solution. The hemolytic
antibody was then diluted with twenty parts of salt
solution and titrated according to the following table:
.jor^ni dsfil orftf
1o 9J. Y<* £ttfl
.::3rcer» Y^**0^ duode, slb&w bnB agnl-'
nl <sj3W boofcf erfT . cfriBsrf srftf mot'i boctet ".fiqeu otow
- ni bf^OBlq aerfrf eiew'rfolrfw teocfjLrd csfilg ellisrfs
T .cJolo e/ij raoi": bectaiaqse nrjjiea 9.r£J lictnir xocf
. eod0^ 2 o^nl: llo be^cfaqlq rro/frf aav/ rrrx;ioa
'•lJj?v. ^'1 £>0bbe esw IOT^OY^S ^° ^m/omB Ii?:
o .ro-ss
r> j H8/f^ GBV; j.-.'jgsi^aB jjvodr.
* -;^c9d arfT .
no x vt
. ci IT! . dssd noiduxi'i
i^Bxil lo f>ortf&fii srfct Y^
lo -y nod rldlw 'bedulib mm
.ctnsmolqmoo -'-Tfis
t t. a. i ibQKu
• fiSs :
•
oii/bs
erfT «B
.:
s
BTOW riojtrfw E^ic'dBt io mtrise
SB bsa0 EBW eslocuioo fcoold
Tol eJee
.
• y[,j
oo
'XCKJP:
f3r;od
g;*i iv;o f
SlrfBT
UT
0
H it
5SVI
.00 0 . S
8YC*'1 ^
eevx
.00 0.^.
CYB!- 1^
oevr
.00 0.6
Titr of* /•*
<j V *> J..1 w
asei
.00 0.0
p.Yab 01
o fjBw iiimoa srfd boe bold sow IfiirlnB
b Y-tci/oivs- -f^ o^ ri£limic rtenriBm B :
~^?>
>c glq sonijj^ BBW ^rtf? " o 9/i^ ;a.Cj
i
•
Table #5.
Tube #1
- . #2
Complement 0.2cc.
n tt
antigen 0.2 cc.
ii
•H Antibody
0.5.cc.+
" 0.10
No henolysis
it n
#3
ii n
n
" 0.15
Partial
#4
n n
M
11 0.20
Hemolysis
Complete "
#5
n ii
u
11 0.25
11 U
#6
n n
M
" 0.30
n n
#7
n n
II
11 0.35
n u
#8
it n
It
" 0.40
M u
t
C r( #9
o o(
n i(#10
s
M it
n u
M
It
n _______
" 0.4
No hemolysis
u n
The unit of antibody, therefore, was 0.2 cc. Using
this unit of antibody and varying the amount of comple-
ment, the unit of complement was determined. Two units
of complement were used in running the complement-fixation
test, also two units of hemolytic antibody. Each bacter-
ial antigen was nob tested against each of the three anti-
sera. The results of such tests are shown in the follow-
ing proctocols: (The preliminary titrations of antigens
and antibodies are purposely omitted because they vary
with each lot and merely tend to confuse the reader.
Table #6. Protein Antigen and Homologous Antiserum.
Tube
56° for 30'
37° for 30'
37° for 30'
Result
1
2
3
Antiserum
IT-" H
Antigen & Complement
•fhemolytic system
Partial
hemolysis
Complete
u u
Antigen &
it M
s%
l»
t, ..
V^L
It
•i n
H
0 w
..
ii if
3%
P,.o "
n
.
n n
•^ !
v*
^0 "
t»
I*
„
Jt
.,VL \ r>
^•Tv / T J
»
kO '
"
n M
.-
iplx <*
; . uselqmoo erf^
5 .xbodiun
3fi3 lO ff.0i39
'lo
brus xboc
sew -Jn^fnolqraoo 'lo ;.?i.nt/ ori? t
ni beet/ 9ri°«t7 ^fismalqr^oo
orl lo Svtinjj owrf osZfl Xci
bod sec1 joii ee-\v negl-lriP
rioi/c lo i.iltrr.O'i eriT .B
arfT) :elcon^ooiq
'
.
o
'Oo 10! YS
imjnalqrftoO ^ n^si^rfA
.j
13
Table #7. Antigen and Anti-coli Serum.
Tube
56° for 30 »
37° for 30'
37° for 30*
Result
1
2
3
Antiserum
it
+Antigen & Complement
ii
+Hemolytic system
M n
Complete
hemolysis
n u ""
Table #8. Protein Antigen and Ant i -fat Serum.
Tube
56° for 30 »
37° for 30 «
37° for 30'
Result
1
2
Antiserum
ii
+ Antigen & Complement
it
+Hemolytic system
n n
Hemolysis
n
Table #9. Protein Antigen and Normal Serum
Tube
56° for 30'
37° for 30'
37° for 30'
Result
1
Serum
+ Antigen & Complement
+Hemolytic system
Hemolysis
2
ii
it
n it
n
Table #10. B.Coli Antigen and Homologous Antiserum.
Tube
56° for 30'
37° for 30'
37° for 30'
Result
1
2
3
Antiserum
ti
+ Antigen & Complement
ii
+Hemolytic system
ii M
n n
No hemolysis
Complete "
n n
it ir
Table #12. B Coli Antigen and Anti-protein Serum
Tube
56° for 30'
37° for 30»
37° for 30'
Result
1
2
3
Antiserum
ii
+ Antigen & Complement
it
Hemolytic system
ii ii
n u
Complete
hemolysis
n M
n »
ii ii
•t<-»?j tctnA
'OS
.8%-
ivTornsh-f rfrt • ;?•: >1 TrcO v ns'
«0« 10! °V5
A ft.: :
•06 1.
[••rrrtoO
• f
«0£ ic'l Ovio
•A 4
mw: ':
n£»-:> / ^ riA i I oO . 9 ,2 1-^ o I cr/?T
iol c
Ian?'
«?.J
14
Table #13. B Coli Antigen and Normal Serum
Tube
56° for 30»
i
37° for 30 «
37° for 30 »
Result
1
2
3
Serum
ii
+ Antigen & Complement
u
•fHemolytis system
n n
n n
Complete
hemolysis
u n
n u
ii n
Table $14. Fat Antigen and Hemologous Ant i serum
Tube
56° for 30'
37° for 30*
37° for 30 »
Result
1
2
3
Antiserum
it
•f Antigen & Complement
u
+Hemolytic system
n u
n n
No Hemolysis
Complete "
n n
n n
Table $15. Fat Antigen and B. Coli Antiserum
Tube
56° for 30'
37° for 30'
37° for 30'
Result
1
Antiserum
+ Antigen & Complement
+Hemolytic system
No hemolysis
2
3
ii
n
u u
n n
Complete "
it it
ti n
Table $16. Fat Antigen and Protein Antiserum
Tube
56° for 30»
37° for 30*
37° for 30'
Result
1
2
3
Antiserum
it
+Antigen & Complement
+ "
fHemolytic system
+
u n
Complete
hemolysis
n it
u ti
ii ii
Table $17. Fat Antigen and Normal Serum
Tube
56° for 30 »
37° for 30'
37° for 30»
Result
1
2
3
Serum
ii
Antigen & Complement
n
Hemolytic system
n u
n u
Complete
hemolysis
It II
n ii
, u fi
H brrjs nsgr.htA jfi'?
'
:
<.' 1 i T
J • r I J. - ' -1. V 1 I '*. 1
H
1 1 I F. * I W U •*•
If
.I" • - ~
Jt-
II II
^ - -
c
tc't °3d
•OS icl °V?:
IffnoO A ctoai
-.
»0£
»0e -rol °0? odjjT
^TkfljS I
15
In the preceding (16-17) protocols all tubes marked
#2 are the serum controls, used to show that the serum
itself will not fix complement. Similarly the #o tubes
are antigen controls.
The Bacillus coli antigen, when used with a hemolo-
gous antiserum, (table 10) gives complete fixation of
complement. The same antigen when used with the ant i- fat
serum, (table 11) gives similar results; but when used
with the anti-protein serum, (table 12) one fails to
obtain a positive reaction. These findings confirm the
idea that fats may act as antigens and lend proof to the
specificity of the reaction. The protein antigen, when
used with its homologous antiserum (table 6) gave results
which were nearly completely positive, but it is worthyof
note that the reaction was not as clear cut or as satis-
factory as the previous tests. When used with the other
antisera (tables 7-9) there was no complete fixation.
Thus there is evidence that the fats are true specific
antigens and that the protein material plays a minor role,
if any.
B. Agglutination tests. Although the presence of
agglutinins probably does not indicate protective ability,
as Bordet (26) showed that horse serum clumps tetanus
bacilli; yet the horse is very susceptible to tetanus,
nevertheless, the experimental production of agglutinins
in the serum of immunized animals is very indicative of
a specific reaction.
The sera obtained from the immunized animals were
tested against the B. coli antigen. Ten drops of serum,
in various dilutions, were added to ten drops of antigen.
The tubes were well shaken and incubated at 37°C. for one
hour. The results are recorded as follows:
Complete
agglutination
Partial
agglutination
No
agglutination
Table #18. B. Coli Antigen with Heraologous Antiserum
Dilution of
Serum
1:100
1:200
1:300
1:400
1:500
1:600
1:700
1:800 1:1200
Result
*
•j-
-;-
+
+
+
+
/j n -r'v ^at. i^;i* 1:1 on ~.i:If IsaO -jrfT
• • •-• •> £»-.•• '- v ( C\ r •* fff <"• -I \ "ff if^ip r "'TT R p i
U-J O ?7 V - ^ \ v/ -. v L U .•> •- / ^.UW i •JCi -U ». II o*w-'^5
oajj rrerfw #$$l:3rrB tMnnc oriT .dnr<m9lq
^1 UP. 91 isllffiis af/vl; (II efcted) ,;
91 olcfect) trrri,"ief- nto^oTq-i^n^ e.rfj rfrtiw
.'-' v B
i'Koiq arfT .fioi.-t^fisi -arfd. to Y
c -/•:!£ 3) tnurf.9£l^nB cuogoloriori acfl -rf^iw boeu
rcf tovioieoq ^Idtfsl^moo Y^'1^0^ Qi^w rfolrfw
o Tfislo SB ^on ssw rioictojsot erfo cfsrW o^on
boe.t/ noriV' .ecfseJ L-woiveiq sifcf cs Y'10^0-6^
alqrnoo or: BBW oporto (G-V seldB^) BioBJt^fiB
o^£ set i?": 9ii-i d-crfJ eonsbive r,i e'larf^ s
C^T I#iiPGr-.rn nieaoiq s>rf^ Jsffrf bnB snesictna
1 Jt
jia ^^Br»ii-:ij cforr
aqmulo n:ir'ij«! ©sior! J
GJ elo't.vqect^fs if-^ov
• -JB ->o ao.':otufcoTtq" ^fsml'ieqxo )fi'^ ,F
icJ rroil bsnic^co Bioe.e
tfnfc iloo .3 erfj .laniBgB
i/£if> ai/oiiBv ni
s'iow eecfi;--
0ee-: .1
"•!,;«H d^l-v.' iir/ji^nA ilcO .
16
Table #19. B. Coli Antigen with Antifat Serum
Dilution of Serum
Result
1:100
4
1:200
4
1:300
•f
1:400
4
1:500
+
1:600
i
1:700
1:800
Table #20. B. Coli Antigen with Antiprotein Serum
Dilution of Serum
Result
1:100
1:200
1:300
1:400
1:500
1:600
1:700
Table #21. B. Coli Antigen with Normal Serum
Dilution of Serum ;
Result
1:50
1:100
1:200
1:300
1:400
1 : 500
1:600
A control suspension of the B. coli antigen did not show
agglutination*
Comparison of the results by use of the complement
fixation and agglutination tests give a very striking
picture. We find that the fat antigen (table 19) gives
rise to specific antibodies and that the protein antigen
(table 20) is of negligible importance in this respect.
C. The Precipitin Reaction. This test has to do
with the precipitation of solid matter out of solution of
invisible "colloidal suspension". Obviously the B. coli
and protein antigens cannot be used, but the fat antigen
is suitable for the reaction. The most desirable antigen
is one which is perfectly clear and transparent. Clarity,
in a fat suspension, depends upon the state of the fat
particles, so that the greater the degree of dispersion,
the greater the clarity. A perfectly clear antigen was
not obtained, but a sufficient degree of dispersion was
made to run the tests. Ten drops of serum was layered
under ten drops of antigen, and after the incubation at
37°C for fifteen minutes, the following results were
obtained.
Table #22. Fat Antigen and Homologous Ant i serum
Dilution of Serum
1:25
1:50
1:75
1:100
1:125
1:150
1:175
Result
+
+
4-
+
±
-
-
!_L *
r, f ,-•
\ J50/19S Ifc ffCK
. '"> •«• rJ • I
:
00£:I
c. 'to riold
r |r.~ .\ . r 1 '•.< F . r
- I V '•- V-. J I VJ M . -*
DOS: I
001:1
06:1,
ffi/^eg 1o
I
1 - i - i :
**
v i
^nB l/oi> .8 srfd- lo nol^
?. lo
^to noaarrio
dl^ilgsn
;: lo
^rfit lo
•
airfT .ftoidofiefl
ts.tfrfsiffl £)iloe lo
lvdO ."ffoicneqej
ircf b^ct scf ^oon
e*! orfj 10!
el
:f> lo ©ei^eb tfneiollljfl
fitirrse lo r.qcib rroT
L Oi1^ i^dle fcne ^e^i^fl^
ss*! snivrolfcl ©xlrf t-sodunlm
03 elcfj
v
•
:
17
Table #23. Pat Antigen and Anti-coli Serum.
Dilution of Serum
Result
1:25
+
1:50
+
1:75
+
1:100
±
1:125
1:150
1:175
Table #24. Pat Antigen and Antiprotein Serum
Dilution of Serum
Result
1:25
1:50
1:75
1:100
1:125
1:150
1:175
Table #25. Fat Antigen and Normal Serum
Dilution of Serum
Result
1:25
1:50
1:75
1 : 100
1:125
1:150
1:175
+ = Complete precipitation
* = Partial "
- = No "
The results are very indicative of a specific pre-
cipitatin reaction, in which the fats play the most
important part. Too much weight, however, should not be
attached to these reactions of precipitins because in
some cases (preliminary titrations) spontaneous reaction
occurred. Such reactions also frequently occurred after
the fifteen minute incubation period. This was probably
due to the unstable dispersoid phase of the antigen, so
that it was very susceptible to the reactions of the
various sera.
D, Anaphylactic Reactions. It was found by trial
that 1.0 cc. of a 1:10 dilution of the fatty antigen was
usually fatal to guinea pigs when injected directly into
the heart, but the pigs could withstand the same amount
of a 1:20 dilution. This amount, therefore, was used as
the "toxic" dose. The animals were sensitized by intra-
cardiac injections of 1.0 cc. of a 1:500 dilution. They
promptly recovered from the effects of the injections and
gradually gained in weight. Tv/elve days after the sensi-
tizing dose the pigs received the "toxic" dose. The
results are recorded as follows:
I 001:1
3?:£
KLCtL
T L!
ttf-osqictrrA bct& ne^.t^'nA etc1! .£2% elcteT
£JOOI
DC : £
32:1
£>
Ir;TTrioH bnjji ne^l^flA 3R'*L ,62% olrfBT
;:l|OOr:I e?:l[03:J:|e- :"ia
o
t't3vov*o
arrl^li
1o
.boiisq
orfcf lo e.sarfq ft
fw nl
rCount ooT
saerfd
e'!': 1oi;8
9,-tjjatm
fTi^crfi-
^nfi^i^
beiis
erfrt
oct
fcrrxrol
Y
BV/ ctl .snoictoc^H oi^ojs
arfct io oi^iflifc 0£:I e lo .00 O.I c>.
t nsrfw si jesnii/3 oct I
SB?/ j
OOC:I B
eri^ .to R
ctrft .noitfwlib OS: I B lo
8ffT .eeob Msixo^" si
0 O.I lo enoi-instni o.eli
orfo moit ^eidvoosi x-C^^"
.rM^iev/ nl fca/Tjta^ yllBi/I:.-
a^lq orfcf ee^h ftj
as
18
Table #26. Anaphylactic Reactions.
Temperature
Symptoms
Before injection
102. 2°F
5 rain, after "
10 " " ft
15 " " "
100.4
100.0
' 99.6
Slight motor paralysis of
hind legs
Increased motor paralysis of
hind legs
Dyspnoea
20 " " "
99.0
25 " " "
99.8
Paralysis decreasing
30 " " "
100.4
Dyspnoea disapeared
35 " " "
101.0
Paralysis disappeared
A normal, non- sensitized animal was then injected
with the "toxic" dose. The results were as follows:
Table #27
Temperature
Symptoms
Before injection
100.8
None
5 min. after "
99.2
n
10 " " "
99.4
:
15 " «
99.8
M
20 " " "
100.0
II
25 " " "
100.4
"
30 " M "
100.4
; » • ' ' ::.
35 " " "
100,4
M
These experiments were repeated three times and the
results were practically the same with each set. The
sensitized animals always showed a marked sub-normal
temperature, with paralysis of the hind legs. The acted
as if they were trying to "huddle up" and the hair "stood
out". The underlying principles of anaphylaxis are as
yet very obscure, but it is very doubtful that proteins
rr-)G.
SOI
eeiortl
bnlrf
•
?if> i
o.ee
0.101
rue bo;
non
b "
3,
n<
11 5i
rrxoft A
Jtf orfrf if •
(V v> I l o ''— X
19
are essential for sensitization. Either the fats may
split off a molecule that is toxic, thus giving rise to
the reaction according to Vaughn1 s idea, or the reaction
may be due to variations in surface energy and molecular
attraction.
So far, then, we have prepared an antigen from the
fats of Bacillus coli and have demonstrated the ability
of this antigen to cause the production of various sero-
logical reactions when injected into suitable animals.
Our interest now turns toward the chemical composition of
the antigen and the attempt is made to analyse the fatty
mass .
In the previous preparation of the fatty antigen
about 0.4 gram of fat was obtained from thirty litres of
broth culture. In order to obtain a larger amount of
material two hundred litres of broth were made and inocu-
lated as before. The bacilli were collected in the same
manner. The bacteria mass was then transferred to a
litre flask and covered with ether. The flask was then
fitted with a reflux condenser and heated to 50°C on an
automatic electric water-bath. The heat was applied con-
tinuously for a period of forty-eight hours. At the end
of this time the flask and contents were cooled and the
ether was pipetted off into another flask. This latter
flask was tightly stopped and placed in the icebox. The
bacterial residue was extracted again in the same manner
and the ether extract was added to the first lot. A
total of five extractions was made. The mixed ether from
the five extractions was then evaporated under reduced
pressure (obtained by attaching an ordinary filter pump),
at a temperature of 25°C. The remaining fatty mass was
then treated with potassium alcoholate. The flask was
fitted with a reflux condenser and the fats were saponi-
fied at a temperature of 90°C. The flask and contents
were cooled and acidified with hydrochloric acid. The
flask was then attached to the reflux condenser and the
fluid was evaporated at 70°C under reduced pressure. The
concentrated mass of fatty acids was then taken up with
ether. The ethereal solution, containing the fatty
acids, was pipetted off and the ether evaporated at room
temperature under reduced pressure. The mass of residual
fatty acids weighed 2.873 grams. A portion of these
acids was fluid at 30°C, but the remainder were solid up
to about 65°C. The acids were then subjected to distil-
lation in steam. The distillate was treated with solid
sodium chloride ,and then shaken up with ether. The ether
layer was removed and evaporated in vacuo at room temper-
ature. There was a very slight trace of residue, too
ixvicr arto
E r ^ (?rr
Jj ' V .JO
CIO O3 18
to 2
orfT .
Jbrrc -rfsfill
•3leQl^ TeiWcxtfi ocffti lie -J
£ 9rf^ ax beoclq fiw- baqqotfe
T!£g erfd ni nxBc bodo^icfxo sew
•8
r M r
20
small to weigh with any degree of accuracy. This repre-
sents the volatile constituent.
The remaining mass was redissolved in alcohol and
made slightly neutral by adding potassium alcoholate,
using phenolphthalein as indicator. The so3.ution was
diluted with distilled water to about 100 cc. 30 cc. of
a ten per cent solution of lead acetate was diluted to
150 cc. with water and boiled. The hot solution was run
into the soap solution, constantly shaking the latter so
that the lead soaps would adhere to the sides of the
flask. The flask was then filled with hot water and then
allowed to cool. The lead salts all adhered to the
glass, leaving a clear supernatant fluid which was de-
canted off. The soap was then shaken up with ether at
37°C. The ethereal solution was then cooled and filtered.
The filtrate contains the lead salts of the liquid fatty
acids. This filtrate was then shaken with twenty per
cent hydrochloric acid, to decompose the lead salts. As
the fatty acids are set free from the lead salts, they
are taken up by the ether. The ethereal layer was re-
moved; washed with water until the washings were free
from acid; and the ether then evaporated. The remaining
liquid fatty acid weighed 1.021 gram. This acid solidi-
fied when cooled on ice and melted at 13°C. The neutra-
lization value was then found as follows: 1.021 grams
were neutralized by 34.9 cc. of N/10 KOH. This is equiva-
lent to 34.1 cc. per gram of acid, or 191 grams KOH. The
neutralized acids were now acidified with 20% hydro-
chloric acid and shaken out with ether. The ether was
then evaporated and the iodine value was obtained as
follows (Hubb's process):
The iodine solution is prepared by dissolving 13.5
grams iodine in 250 cc. of 95$ alcohol, and by dissolving
15 grams of mercuric chloride in 250 cc. of 95$ alcohol
and mixing these two solutions. A standardized solution
of sodium thiosulphate was prepared by dissolving 24
grams of the salt in one liter of water. 0.2 gram of
iodine and 1.0 gram of potassium iodide are dissolved in
about 50 cc. of water. This solution is titrated to
neutrality by the thiosulphate solution, using starch
solution as indicator. It was found that 15,4 cc . of thi-
sulphate solution was equivalent to 0.2 gram of iodine.
The 1.021 gram of fatty acid is dissolved in 10 cc. of
chloroform and 50 cc. of iodine solution added. The
solution was allowed to stand for fifteen hours. 20 cc.
of 10$ potassium iodide solution was added, and the
volume made up to 500 cc. with water. This solution was
then titrated with the thiosulphate solution. It was
rfrttw 139 3 tt
tforf rtcfiw fcoIXil rrsrtt SBW rtpfill sriT .*?
>TO<1&0 II»' edJtsi: bfial erfT .Xooo oct I^owcClB
[v/ fbiirf't 3ftfi Jani^qua icelo B guivcwl tce.
:w q;j fio>l«ffa nerf^ SJBW
>Ioe:> rjorftf cew fiol^wloa
d^ io p.cJjBa b&ot aiid a
;d bne tlorioo
Q lo .05 06fl
;'.'1 '<••-•>;•' $ p. f\rr-\ ' 'i'XDi
.( Jg Jboctlon Irtfi 9oi #<"> folooc* rt^rf
V"T»rXof; £! f • '" 'iri/1'!1''* flOfljt C'^ViT £>JjJf-V
'»! OX\H lo .no e,^£ vrf "jDasirni^
C 1C ,bXDfi Ir*- flifilg. T'^Q ,D3 f , J^£ O
iw &9 ft Ibi'jfl won ©TOW ebto£ bsstf
.Torf^v^ rictiw tfwo nojfB'le bni5 hloe o
1 V
L ', H
21
found that 0.863 gram of iodine was absorbed. 100 grams
of the fatty acid, therefore, would absorb 84.6 grams of
iodine, the iodine value, then, being 84.6. The neutra-
lization value of oleic acid is 198; its melting point is
14°C; and its iodine value 90. We are dealing with an
unsaturated liquid fatty acid, whose melting point is
13°C; neutralization value 191; and iodine value 84.6.
Oleic acid is an unsaturated liquid fatty acid with the
above given constants. The unknown approaches very
closely to oleic acid in these three values, and it is
permissible, under the circumstances, to consider such
unknown as oleic acid.
The remaining insoluble lead soaps were now decom-
posed with hydrochloric acid and the liberated fatty
acids were extracted with ether. The ethereal portion
was pipetted off, washed with water, and the ether
evaporated. The residue weighed 1.84 grams. The iodine
value was 2.3, due, probably, to admixture of the un-
saturated acid previously described. We are dealing,
therefore, with two or more saturated fatty acids, since
it was found that about one-half of the mass melted at a
temperature of about 32°C, while the remaining portion
did not melt until heated to a temperature of 68°C. The
saturated fatty acids which have melting points between
30°C and 35°C are confined to one member, capric acid.
The saturated acids which melt between 65°C and 70°C are
confined, likewise, to one member, stearic acid. To sep-
arate these acids, when one has such a very small amount
of material, is hardly a feasible procedure. Nor can one
determine the analysis by the melting points, since such
acids form entectic mixtures which often have a melting
point higher or lower than any of the individual con-
stituents (27) .
SUMMARY
The Bacillus coli has been divided chemically into
two parts, one of a protein nature, the other fatty. The
protein portion alone is not an antigen in the sense that
it will give rise to the specific antibodies of the
bacillus coli.
The fatty portion, on the contrary, contains the
substances essential for the production of specific anti-
genie reactions.
9ft£box 1o mr,
ft rr . •' •
. v< 1 17 i".,' j.y iy no ^ j
" r blofi oxolo 1o 91'!
noicii J
grri.r
T .pelf IB.
n.t MOB oxslo o^
orfcJ "is?"-
i)Jtofl : OB ;••
.-.•.Til ^ri'J bfiB biOJS o -rid
isff^o sriT .tarf^o rfrfiw
rf^ bflB 419JtBW rf^CW bS:
f?V6
.
.joaaxiosgo YJ
°^ %c
. .!Bri-9n
irfw t'
J3 O^ f
. ,1 rioi rt bsjair
• ^lecfmsm one o^ be-.
3 0°6d n99w^sd ^I«3fli rfotrfw
as oxTBS^a ^ledfioom an
.: visv B riouc 8Bii.-3nr ^is 9/
• eiifbeootq oldflseal t'jodsra lo
^adnioq gnx-tlsm orf^t t '^ ^nlrni9d9b
n B 9vor( rt9^1o rfo/rfw a- :>1 abxoa
-noo iBt/bxvrbfix Bffd lo x^s ^fi^ rf ^n
BCtnSIf^ :
YflAMMUS
tvifc flOSCf 0£fi if?
,ect'iJBq.
nlsfcti
a 111.
.Tfloo t^*iB'rc;nc'0 arf^ no
't er{;j
22
It has been analysed to show the following approxi-
mate constitution:
a. Volatile fats, trace
b. Oleic acid
c. Capric acid
d. Stearic acid
CONCLUSIONS.
The specific antigens of the Bacillus coli are
chemical entities bound up in the fats of the organism.
The proteins of the organism are not concerned in
the specificity of the antigens.
The fats of the antigens consist of volatile,
saturated and unsaturated acids.
&HOI3IUQHQ
orfo lo 3
a»1o1 orid rri qv Lnwco c '.e -Ijeolff:
erf? 1o a:
e'- -
23
BIBLIOGRAPHY.
1. Jobling & Peterson, Quotes from Zinsser, "Infection
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
22.
23.
24.
25.
26.
27.
and Resistance'1, 1918, p. 39.
Noguchi, H., Univ. Pa. Med. Bull., Nov. 1902.
Muller, Centrabbl. f. Bakt., 34, 1903, p. 562.
Landsteiner, " , 33, 1905, p. 318.
Bang & Porsmann " , 40, 1906, p. 151.
Kyes, Berl, klin. Wchnschi, 1902, pp. 856 & 918.
1904, p". 494.
Sachs, Wien, 1905, p. 901.
Ford oc Abel, Jour. Biol. Chem., 1902, 2, p. 273.
Pick, E., Kolle u. Wassermann Handbuch Vol. 2.
u. Yamanouchi, Zeitche. f. Gumuntatschforch,
I, 1909.
Bogomolez,
5, 1910.
Star in, W., Jour. Infec. Dis., 1918, 25, p. 139.
Editorial, Jour, Am. Med. Assoc. 1919, 73, p. 609.
Walker & Sweeney, Jour. Infec. Dis., 1920, p. 256.
Hollmann oc Dean, Jour. Cut. Dis., 1919, 57, p. 367.
Warden, C. C. , Jour. Infec. Dis., 1915, 16, p. 426.
Jour. Am. Med. Assoc., 19T5, 65, p. 2080.
1917, 53, p. 432.
& Schmidt, L.E., Jour. Lab. & C3.£n. Med.,
1916, 1 p. 333.
Jour. Infec. Dis., 1918, 22, p. 133.
1918, 33, p. 504.
1919, 3?, p. 285.
1919, 35, p. 399.
Vaughn, V., Protein Split Products, Lea & Pebiger,
1912.
Bordet, J., Am. de 1'inst. Pasteur, 1896, 10, p. 193.
Lewkowitsch, Chemical Technology of Oils, Fats &
Waxes. 3rd ed. vol. 1.
it u
AUTOBIOGRAPHY
I, Edward Everett Hale Boyer, was born in Lynn,
Massachusetts, February 15, 1893. I received all of my
secondary school education in the public and high schools
of the City of Lynn; my undergraduate education at the
Massachusetts Agricultural College, from which I obtained
the Degree of Bachelor of Scisnce in 1916. I pursued
graduate studies at the Ohio State University, from which
I obtained the Degree of Master of Science in 1917 and
the Degree of Doctor of Philosophy in 1920.
. ooarsA , JbeM .mA
• w " .
6J9I
6X91
OE^A . LoM '.caA . i
ootnl .TCUOL ,
!G.. cJuD .iijoi
'
ni
J 8 ft X ' A 9 '
:oT XBo-iir
.1 .lov .
V-
Binder
Gaylord Bros.
Makers
Stockton, Calif.
PAT. JAN. 21. 1908
56394 1
UNIVERSITY OF CALIFORNIA LIBRARY