A STUDY OB' THE CHEMICAL CONSTITUTION AND BIOLOGICAL PROPERTIES
OF THE ENDO-ANTIGEN OF THE BRUCELLA GROUP OF MICRO-ORGANISMS

ProQuest Number: 10008404

All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.

uest,
ProQuest 10008404
Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106- 1346

A STUDY OF THE CHEMICAL CONSTITUTION AND BIOLOGICAL PROPERTIES
OP THE ENDO-ANTIGEN OP THE BRUCELLA GROUP OF MICRO-ORGANISMS

Thesis

Submitted to the Faculty of Michigan State College
of Agriculture and Applied Science in partial
fulfillment of the requirements for the
degree of Doctor of Philosophy

‘by
RY B/ gsnnell
July, 1937

a'
TABLE 0? CONTENTS,

Introduction *

.... ..........................

. 1

Isolation of the endo-antigen . - ........................ 6
Chemical nature of the endo-antigen

................. 8

Acetylation of the endo-antigen

................ 11

Acetone extraction of the endo-antigen................ 16
Ether extraction of the e n d o - a n t i g e n .................. 19
Tyrosine and tryptophane content of the endo-antigen.
The sugar a c i d ................

.23
28

Lead acetate p r e c i p i tation............................. 29
The remaining nitrogenous constituent.................. 31
Biological activity of the endo-antigen................ 35
Precipitation s t u d i e s .................................. 35
Toxicity

....................................... 44

A n t i g e n i c i t y ............................................ 50
Allergic a c t i v i t y ...................................... 55
Immunizing v a l u e ........................................60
Therapeutic v a l u e .................

.

61

Physiological s t u d i e s .................................. 63
D i s c u s s i o n .............................................. 66
S u m m a r y ................................................. *71
Acknowledgement .........................................74
B i b l i o g r a p h y ............................................ 75

110630

Ciiemical Consti tut ion and ^i o 1o,;ical
Properties of the Undo-antigen
of the Brucella '-roup of Micro­
organisms .*
chemical fractionation of the bacterial cell in order to
obtain pure

and suitable antigens has received considerable atten­

tion during the

last fev; years.

It has already been demonstrated

that chemically defined and reproducable antigens separated from
the bacterial cell serve as better immunizing agents, and furnish
more suitable materials for the study of antigen-antibody systems
and bacterial allergy than do the intact bacterial cells.
Considerable progress has already been made in the chemical
fractionation of brucella cells.

In previous studies (1 ), (2),

two significant highly antigenic fractions mere separated which
from preliminary investigations appeared to be of sufficient
i .rportance to warrant a further and more detailed chemical and
biological study.

The isolation of one of those, the C-substunce,

has recently been confir ed by Pchapira (b ).

AL though it is

neither carbohydrate nor protein, its antigenic) activity appears
♦This study was made possible by a grant in aid from
the Horace II. and Mary A. hftchham fund and the Hl(La
Sachs plotz foundation.

(2)

similar to that which is delegated to the specific polysaccharides
in many other organisms.

The albuminoid fraction from brucella

cells, a combination of the 3-substance with a protein-liice group,
exhibits in addition to the precipitation of immune sera, the
property of being

extremely toxic for normal guinea pigs upon

intraperitoneal injection.

This same toxic property upon injection

into normal guinea pigs is

also exhibited by a dense suspension

of

the intact bacterial organisms.
The syrup tons and signs shown by the guinea pig following the

intraperitoneal injection of a toxic dose of the albuminoid, of the
more highly purified endo-antigens or of a dense suspension of in­
tact cells begin to appear about two hours after injection.

The

animal remains quiet, its coat becomes roughened, and it hangs its
head in a state of

stupor.

signs of peritoneal

irritation characterized by rigidity and spasms

of the abdominal muscles.
bach.

At the fourth hour the animals show

The head is momentarily retracted on the

Later the spasms involve other muscles of the body, causing

the animal to jump upward when disturbed.

If the reaction continues

for 8 to 10 hours an opaque exudate containing neutrophiles collects
in the left eye.

The leukocyte count falls to 3,000, chiefly due

(3 )

to tile, disappearance of neutrophiles from the peripheral blood.
The administration of a dose which produces no lowering of the tem­
perature raises the leukocyte count, often to 20,000, the increase
in count being due chiefly to an increase in the neutrophiles.
The temperature begins to fall from 1 to 2 degrees the second
hour after injection, and continues downward until the animal's
death, preceding which a temperature of 34° to 32PC. is noted.
Death

occurs from 6 to 18 hours after injection depending upon

the toxicity of the material. The temperature never rises raore
than 2 degrees

above normal following the administration of a

non-lethal dose.

At autopsy, the only gross changes noted are

a small amount of clear viscous exudate containing lymphocytes and
neutrophiles in the peritoneal cavity, petechial hemo^hages in
the parietal peritoneum and

extensive he itOqo images in the

of the stomach and in the intestines.
pig are similar to if

not

mucosa

The reactions in the guinea

identical with those described by

Yaughan and Yaughan (4) in their studies of bacterial poisons.
The 3 -substance was reported to stimulate no anti-body pro­
duction, to elicit
specific serums.

non-specific skin reactions and to precipitate
The albuminoid, on the other hand, was found to

produce specific antibodies in normal animals and specific skin
reactions upon intradermal injection into sensitized subjects.
Gwatkin (5 ) has

described

the preparation of an alcohol

precipitate prepared from an aqueous extract of moist Brucella
organisms, as well as an extract involving autolysis of dried and
ground Brucella cells.
toxic for normal

These preparations were found extremely

guinea pigs.

The alcohol precipitate, character­

ized by qualitative tests only, was found to be polysaccharide in
nature.

In work reported previously from this laboratory (1) it

has been shown that precipitates and extracts prepared in a manner
similar to those used by G-watkin contain a variety of components
which may be separated by furrier chemical treatment.

It was

thus believed that some further purification of the toxic principle
should be possible.
The early work on the toxic fraction indicated that it was
liberated in a soluble form in the course of autolysis of the bac­
terial

cells.

Autolysates prepared at varying pH and under a variet

of conditions were subjected to fractionation by dialysis, alcohol
precipitation and by tri-chlor acetic acid precipitation.

Many of

the fractions obtained exhibited toxicity for normal guinea pigs.

( 5 )

The fractions were, however, chemically ill-defined and consti­
tuted no improvement of the previously obtained albuminoid.

This

work v/as further complicated by the fact that occasionally auto­
lysates were obtained which were non-toxic.

Extracts of ground

bacterial cells in lf/2 ilaQH, fractionated by various means', gave
results similar to

those obtained by autolysis.

Digestion of ground Brucella cells at pH 8 with trypsin and
erepsin and thorough centrifugation of the undigested residue fur­
nished extracts that were more constant in their toxicity than those
obtained by the first methods used,

fractionation of the digests

by alcohol precipitation provided little of interest.

Precipita­

tion with phosphotungstic acid, basic load acetate and with silver
nitrate,

and

decomposition of the precipitates yielded a biological­

ly significant oil-like constituent which in large doses produced a
typical toxic reaction ar.d'^addition produced an immediate but
temporary respiratory paralysis.

'his fraction was not antigenic.

In view of the close similarity between the toxic property of
Brucella

extracts

i
and that of Vaughan and VaughanJ (4 )

’‘soluble

protein poison’1 which v/as prepared from a variety of proteins,

(<5 )
their chemical separation procedure was carried out on -rncelln cells.
A fraction was obtained similar in nature to their preparations which
produced,

upon intraperitoneal injection into normal guinea pigs, a

toxic reaction entirely analogous to that given by injection of the
albuminoid or of T.tfhole organisms.

The fraction however, had no

antigenic properties whatsoever.
The

isolation of a

toxic antigenic fraction from Salmonella

aertrycke by Puistrich and Topley (6 ) and by Boivin, Tecrobeanu and
hesrobeanu (7 ) as well as from a number of other organisms by the
latter authors, surgested the application of their procedures to the
Brucella group.

A modification of the method of Boivin and collabo­

rators yielded a product which was found to be extremely toxic and
at the

sane tine a specific and highly active antigen.

This product,

moreover, was found similar to if not identical with the 3 -substance,
which

implies that there is a close connection between the new frac­

tion and the albuminoid previously discussed.
Isolation of the Bndo-antigen
The endo-antigen was xxrspared frcmi a number of strains of
Brucella a b o r t u s , Brucella inelitensis and Brucella suis in the

stock collection maintained in this laboratory.

The dried cells

"ere

(7 )

prepared in a manner
lication (1)*
7 days.

similar to that described in a previous pub­

'The dried organisms were ground in a ball mill Tor

The resulting Tine powder was suspended in sufficient

distilled water to make a suspension of 5 per cent by weight*
Toluene was added at the same tine with the water.

The suspension

was digested

under toluene by 0*5 per cent trypsin

I'lapCO^at 37°

C. Tor two and one-half days,

0*5 per cent

trypsin v/as again added and digestion was

proceed again for two and one-half days.

^.t the end

and0*5 per cent
of this time
allowed to

To the tryptic digest

suspension

v/as added tri-chlor acetic acid to a dilution

of Ij/4.

At the end

of three hours the mixture was neutralised and submitted

to dialysis in running tap water, followed by si: ilar treatment in
distilled

water until dialysis was complete.

The dialyzed suspen­

sion v/as concentrated to as small a. volume as was practicable by
distillation at 40° C. in vacuo.

The resulting concentrate was

precipitated with 10 volumes of alcohol containing a few drops of
glacial acetic acid*
suspended

The precipitate was

in a small volume of distilled

centrifuged and rewater and again precipita­

ted with alcohol, which procedure v/as repeated o times, or until
the supernatants contained no further extractable material-

(8)

The final aloohol precipitate represents the endo-antigen.
The yield in
15 to

a large number of such preparations has varied from

30 per cent of the dry weight of the bacterial cell; the

average

for all preparations being roughly 25 per cent.

The

endo-antigen forms unstable acidic suspensions in distilled water.
These suspensions may be converted into deeply opalescent solutions
by adding Il/l NaOH until no increase in clarity is observed.
liquid may then be

The

adjusted to pH 7 with no increase in turbidity.

The preparations from Br. abortus, Br. melitensis and Br. suis are
grossly similar.
The similarity of this product to the earlier mentioned albumin­
oid is shown in that suspensions of the finely ground albuminoid in
distilled water, digested with trypsin and treated as above, yield
preparations exactly analogous to those obtained from the pulverized
whole cell.
Chemical Nature of the TSndo-antigen.
The endo-antigen gives a positive, though very slow reaction in
the Kolisch test for carbohydrates; a positive reaction with Hial’s
test for pentoses; a negative reaction to the di-phenylamiiie test
for ketoses; positive reactions to the biuret test for the peptide-

(9 )
linkage and Millonfs test for tyrosine; a very faint reaction
in the Rosenheim test for tryptophane*
The nitrogen content of the endo-antigen.(micro-Ejeldahl
method) is from

5 to 10 per cent of the total './eight.

nitrogen (method of Van

Slyke ) is absent.

ducing sugars present as such.

Amino

There are no re­

After hydrolysis with dilute hy­

drochloric acid, however, reducing sugars calculated as glucosd
(Schaffer-Hartmann procedure ) have been found to comprise from
4 to 12 per cent of the fraction.
not present.

Phosphorus and sulphur are

Determinations of acetyl groups by the method of

Kuhn and Roth (8) give an average value of 6 per cent.
acid has

Acetic

not been isolated in this procedure and, because of

data to be considered later, it has
value may

seemed possible that this

represent some other volatile acid produced by the

hydrolysis employed in the procedure.
A typical analysis of endo-antigen is given.

This parti­

cular preparation was obtained from cells of Dr. melitensis

strain S0&*

(10)

Total N: 9*403 per cent
Amino M: absent
Reducing sugars before hydrolysis: Absent
Reducing sugars

after hydrolysis: 5*93 per cent

Sulphur: Absent
Phosphorus: Absent
Acetyl groups: 3.48 per cent.
At this point it is

perhaps well to consider a similar examin­

ation of a sample of S-substance prepared from the same strain of
organisms as was the above endo-antigen.

The 3-substance gives a

weekly positive Holisch reaction upon standing; the biuret reaction
is a

light pink; the Rosenheim test is negative; with Rial's re­

agent the reaction is positive but much weaker than that given by
the endo-antigen.

is
Milion's reaction was weakly positive.

analysis of the 3-substance is- as follows:
Total N: 5.86 per cent
,Amino N: Absent
Reducing sugars before hydrolysis: Absent
Reducing sugars after hydrolysis: 8.13 per cent.

Sulphur: Absent

The

(1 1 )
Phosphorus: Absent
Acetyl groups: 48.43 per cent.
It will be seen that the data for the endo-antigen and
those for the S-substance correlate closely with the exception
of the acetyl content.

The probably meaning of the high figure

for the acetyl content of the S-substance will be considered
later.

The analytical data for the S-substance differ from those

previously presented in that reducing sugars are found present
after hydrolysis and the biuret test is found weakly positive.
This latter reaction, checked several times to assure its authen­
ticity, is very weak and might be easily overlooked.

The reduc­

ing sugars upon hydrolysis are detectable only by the use of a
quantitative reagent and do not afford reduction sufficient for
visible detection.
Acetylation of the Endo-antigen
Since the endo-antigen presumably contained determinable
acetyl groups, the importance attributed to the acetyl groups of
the specific polysaccharide of gneumococcus Type 1 (9) suggested
a study of the effect of acetylation upon this substance.

A

(13)

portion of endo-antigen prepared from cells
strain

01

dr. aborts

1151, was precipitated with acid alcohol and the pre­

cipitate packed tightly by centrifugation.

The precipitate,

still

moist with alcohol, was suspended in acetic anhydride containing
a small amount of freshly fused sodium acetate.
warmed to start the reaction and
water bath for one-half hour.

The mixture was

7/as then heated in a boiling

_tt the end of this tine the mixture

was neutralized, then dialyzed in running tap water and in dis­
tilled water until dialysis was complete.

The acetyl content of

the endo-antigen before acetylation was 2.23 per cent, after
acetylation, 7.74 x^er cent.
Because of the interesting effect of the above treatment
on the biological properties of the endo-antigen, to be discussed
later, a second preparation which showed little biological activity
from the same strain of organisms, was acetylated.

Acetyl content

was 3.17 and 3.57 per cent before and after acetylation respectively
The method used in determining acetyl groups involved sapon­
ification in methyl alcoholic !TaOft«

A number of the solutions left

after distillation of the acid from the saponification mixtures were

{13 )

collected, reduced in volume and precipitated with alcohol con­
taining glacial acetic acid.

This precipitate was tested bio­

logically and was then acetylated and retested.

The acetyl

content, which was assumed to be zero after saponification,
increased to 29.03 per cent after acetylation.
For comparison with the endo-antigen a sample of S-sub­
stance was also acetylated.

Its apparent acetyl content in­

creased from 48.43 per cent before, to 68.19 per cent after
acetylat ion.
Felton and Prescott (10 ) have shown it possible to restore
completely antigenicity to the S-substance of Pneumococcus Type

1 after removal of the acetyl groups, by allowing it to stand
at 4°G. in 25 per cent UH 40H.

In view of their remarkable

findings, a portion of endo-antigen from Br. melitensis strain
748 was heated for l/2 hour in k/2 NaOH at 86° 0.

Upon cooling,

the suspension was precipitated with alcohol without neutraliza­
tion.

One portion of this precipitate was acetylated and one

portion was suspended in 25 per cent BH 40H and allowed to stand
overnight in the icebox.

(14 )

It was then precipitated with alcohol and ether, resuspended
in distilled water and reprecipitated.

In Table 1 is presented

the chemical data obtained for the three preparations.
be noticed

It may

that treatment with EH OH had no influence upon the
4

Ti content of the material*

It is also apparent that heating with

N/S NaOH for one-half hour reduced

both the nitrogen content

and the reducing sugars determinable upon hydrolysis.

The acetyl

content on the other hand shows an unexpected increase by this
treatment.

The effects of those procedures on the biological

properties

of the endo-antigen are considered in a later portion

of the paper.

(15 )
Table 1
Chemical Properties of an Endo-antigen
Treated with N/2 NaOH, Acetylation
and i7H4QH
Per cent of
total N

Endo-antigen from
3r. melitensis strain 748

per cent of
Reducing Sugars
After Hydrolysis

ler cent of
Distillable
Acid Formed
in Acetyl
Determination

9.35

4.53

3.34

7.00

3.37

6.05

Endo-antigen acetylated after
heating in N/2 NaOH

5 •84

3.22

14.85

Endo-antigen treated with HH OH
after heating in M/2 NaOH

6.88

2.52

6.84

Endo-antigen heated
1/2 hr. at 80°C. in

it/2

Ha.OH

Table 3
Acetyl Content of Several Preparations
Before and After Acetylation
Per cent of Distillable Acid Formed
in Acetyl Determination
After Acetylation
Before Acetylation
Endo-antigen from
Br. abortus str. 1151

2.28

7.74

Inactive endo-antigen
from Br. abortus str.
1151

3.17

3.54

Material recovered after
saponification in the
acetyl determination

0 *00

29.03

48.43

68.19

3

-substance from Br.
melitensis str. B06

(1 6 )
Data showing the effect of acetylation on the acetyl content
of several preparations are presented in Table 2.

That the acetyla­

tion procedure may have some effect on the endo-antigen other than
*

addition of acetyl groups is suggested by the one experiment in
which the acetyl content of the endo-antigen was unchanged by the
process although its biological properties were found to have in­
creased.

The acetyl content of the re-acetylated product, which

can hardly entirely represent acetyl groups, suggests that acetyla­
tion has brought about an hydrolysis, forming some distillable acid
other than that representing acetyl groups.

The product is known

to have been partially broken down already by saponification. This
explanation is borne out by the acetyl contents of the products
obtained when the endo-antigen was heated in n /2 NaOH &nd treated
with NH 4OH, both of which should have been acetyl-free.

The extreme­

ly high figures obtained for the S-substance would, by analogy, sug­
gest the S-substance to be a partially hydrolyzed endo-antigen. Final
analysis of the effect of acetylation must await identification of
the products obtained in the acetyl distillation.
Acetone Extraction
When the endo-antigen was precipitated with alcohol containing

(17 )
a few drops of glacial acetic acid, and the resultant centrifuged
precipitate was thoroughly extracted with acetone containing a few
drops of 5N HC1, an intensly yellow extract was obtained.

No pig­

ment was extracted by acetone except in the presence of HG1 which
seemed to act as a catalyst.

After distillation of the acetone from

the extract there remained an orange-brown, semi-solid mass with a
peculiar sweetish-pungent odor.

Nitrogen was absent.

When submit­

ted to saponification by refluxing for four hours in 4 per cent
alcoholic KOH, the material formed a deep reddish-brown solution.
Upon cooling and diluting the saponification mixture with three
volumes of distilled water, ether extraction removed much of the
pigment.

After distillation of the ether from this extract there

remained a yellow oily material having a spicy fragrance.

The

small quantities of this material which have been obtained in a
pure state have precluded its adequate characterization.

The com­

pound gives the characteristic red color given by ketones and alde­
hydes with diazo-benzene sulphonie acid.

This, together with its

physical properties and its spicy odor, suggest it to be a di-ketone.
It was shown to be responsible for the brilliant red ring sometimes

encountered in testing the endo-antigen with Molisch reagent.

(13)

shown to be responsible for the brilliant red ring somiti. .as en­
countered in testing the endo-antigen with holisch roagent.
acidification of the aqueous soap solutions an 3. extract ion with
ether yielded, upon distillation of the ether, a slightly pigmented
fatt;' acid.

The fatty acid was a ..Led to that obtained, by subsequent

ether extraction of the mido-antigen after preiininary exanination
had shown their similarity.
In a single experiment the aqueous solutions remaining after the
removal of the unsaponifiable and the
examined.

From these

fatty acid fractions cere

aero obtained a s\.v gestion of the presence of

polysaccharides consisting of a slowly developing Holisch reaction
and 2.3 3 per c e n

of re due ing substance

calculated, as silicons.

po i ysaccnaridas were isolated and no osazonos

go

could be obtained.

There ./as no trace of glycerol or of any other alcohol.
hydrolysis of f.m acetone

axtrao'-

of -03CH in ..:.lcohol ras a ttempted in
&

*X

brown semi-solid m s

/if.- a 1 per cent solution

sia-le experiment.

'he oranse

recovered unohmy ml.

ghat tils acetone extract mas not entirely extraneous matter
inerely accompanying the endo -antigen is indite., ted by the fact that
intraperitoneal injection into

normal guinea pigs prove., if to be

(19 )

mrdly antigenic.

Doses of from one to five millf-rans were found to

elicit agglutinins and opsonins in low titers

for Brucella cells.

Ether Extraction
Further material

was removed from the acetone extracted endo-

antigen by thorough extraction with ether.
nitrogen-free.

The ether extract was

To these ether extracts, in many instances, were

added the ether soluble fatty acid obtained upon saponification of
the acetone extract.

Upon removal of the. ether by distillation

there remained a straw-colored waxy residue.

This residue was

saponified by refluxing for 1 hours in 4 per cent alcoholic HOHThe unsaponifiable fraction consisted of a minute amount of the
fragrant yellow oil obtained in the acetone extract.
soap solutions were
The
acid.

The aqueous

acidified and thoroughly extracted with ether.

ether extract was washed with distilled water until free from
It was then drier over Dehydrite ana reduced to a small

volume by distillation.

quantities of white stocky njedles were

obtained having an odor r e s ending tar soap.

afrer crysoo.llization

they were found to nave bee one less soluble in eth er b wt
in acetone•

.ere soluble

The fatty acid was sa buratea anu quickly became li quid

at 25° 0. and atmospheric pressure,

ho farther stud*' of this compound

(20 )
has been made.

Its resemblance to the acetone soluble, saturated,

liquid fatty acids first noted by Anderson (11) in his studies of
the lipides of the tubercle bacillus is obvious.
Examination of the aqueous solutions remaining from saponifi­
cation was undertaken in a single experiment.

There was a sugges­

tion of polysaccharide in a weakly positive Molisch reaction.
There was no trace of glycerol.

There was obtained, however, a

small amount of crystalline material which was benzoylated by the
Schotten-Baumann reaction.

Saponification of the benzoylated pro­

duct and back titration suggested an alcoholic nature for the
crystals.
Intraperitoneal injection of the ethereal fraction had no anti­
genic or toxic effects.

This fraction may represent extraneous

material accompanying the endo-antigen.

It is present, however,

even when the endo-antigen is prepared from previously defatted
cells.
The acetone and fcther fractions together, represent from 10
to 15 per cent of the total weight of the endo-antigen.

In the

particular separation just discussed, these two extractions re­
moved 13.4 per cent of the total weight of the endo-antigen.

(2 1 )
Acetone-Ether Extracted Endo-antigen
The endo-antigen, which was usually an amber color in solution
or a gray powder when dried, beeame white after the two extractions.
Upon suspension in distilled water it became a salmon-pink in color.
The resuspended material was found to be acidic and quite insoluble.
After neutralization, it became partially soluble and suffered no
loss in its biological properties.

Further addition of N/l NaOH

brought about further solution of the material, but was found, in
contrast to the findings before extraction, to reduce the biological
properties of the fraction.
The extracted endo-antigen still gave similar reactions to all
the qualitative tests to which it had previously been subjected.

The

Millon*s test for tyrosine was much enhanced, while the Rosenheim re­
action for tryptophane had become extremely faint and even absent in
some cases.

In comparison to the unextracted endo-antigen, the nitro­

gen content was slightly higher, running from 8 to 12 per cent, while
sugars determinable as glucose upon hydrolysis were lower, 3 to 8 per
cent,

values for acetyl group content varied from 5 to 10 per cent
The chemical constitution of the acetone-ether extracted endo-

antigen from Br. melitensis strain 606 is as follows:

(2 2 )
Total N: 12.0 per cent
Amino N: Absent
Reducing sugars before hydrolysis: Absent
Reducing sugars after hydrolysis: 4.0 per cent
Acetyl groups: 5.42 per cent
After lipid extraction, a series of tests was employed to
identify the nitrogenous compounds.

The analysis indicated the

presence of amino acids or amino-sugars and of pyrrol rings only,
boivin and co-workers (1 2 ) reported the splitting of the
antigen obtained from Salmonella aertrvcke by hydrolysis in dilute
acetic acid into toxic non-specific and non-toxic specific fractions.
Portions of extracted endo-antigen were hydrolyzed by refluxing for
one hour in w/5 acetic acid, one hour in n /2 acetic acid, one hour
in n/2 HOI, one hour in u/l HC1, eight hours in jm/2 sulfuric aeid,
one hour in $/4 RaOH and one hour in ti/2 HaOH.
very unsatisfactory.

The results were

A number of fractions were obtained, each of

which was an amorphus powder containing polysaccharide and nitrogen,
and many of which retained some part or all of their biological acti­
vity.

from these experiments it was learned that only hydrolysis for

one hour in iM/5 acetic acid had no material effect on the biological

properties of the endo-antigen.

There was noted a gradual destruction

(23 )
Of these properties

as the strength of the acid and the length of

hydrolysis was increased.

After eight hours in N/l sulphuric acid

the hydrolytic products showed only slight precipitation with
specific serum.
h

On the other hand, hydrolysis with

h

/4 or with

/3 Naoty for one hour completely destroyed the biological activity

of the fraction.
An attempt was made to isolate the polysaccharide constituent
of the endo-antigen

by the method used by frankel and Jellinek (13 )

in the isolation of

the polysaccharide from egg albumin.

Hydrolysis

for three hours in 10 per cent barium hydroxide in flowing steam
failed to liberate any free polysaceharide.
osazones failed.

Attempts to produce

The modification method of Levene and Mori (14)

entailing longer hydrolysis produced similar results.
The repeated production of a pink to light lavender color upon
the addition of acids, particularly of sulphuric aeid, to the extracted
endo-antigen suggested that tryptophane might be present in the substance
even though the Rosenheim reaction was but faintly positive.

The method

evolved by Gortner and Holm (15 ) for the quantitative determination of
tryptophane and the minimum determination of tyrosine in the course of
their studies on humin formation was decided upon, in that it would

■(8 4 )
allow a complete nitrogen partition to be run on each sample*

These

workers found that hydrolysis of a protein with 20 per cent HC1 in
the presence of in c r e a s i n g amounts of formaldehyde produced a
gradual increase and finally a decrease in the humin nitrogen formed.
At the peak of this curve the humin nitrogen represented a quantita­
tive determination of tryptophane in the protein*

Repeating this

procedure with benzaldehyde in the place of the formaldehyde, the
amount of humin nitrogen formed rose to a still higher value.

The

difference between the peak value of the formaldehyde curve and
that of the benzaldehyde curve represented a minimum determination
of the tyrosine of the protein.

The method of Tan Slyke (16) was

used in the further partition of the nitrogen of the fraction after
removal of the humin.
From the curves shown in Graph 1, it is apparent that the
/

percentages of tryptophane and tyrosine in the three species of
Brucella vary one from the other,

while the percentages obtained

from Br. melitensis and Br. suis form curves approximately similar
to those expected from the data of Gortner and Holm, the values for
the benzaldehyde curve of Br. abortus differ widely from the expected.

14*.
3
2
3

1

«1

.15
A

«25

•3

.2

.25

3

.15
.2
.25
C
CUBIC CENTIMETERS OF ALDEHYDE

3

£0

0

*2

1

K

PER CENT HUMIN

NITROGEN

•05

2
•05

.1

.15
B

3

2
2
i

1

i
)

1

*C>5

.1

Graph 1. Curve 1. - Humin formation in the presence
of formaldehydeCurve 2 - Humin formation in the
presence of benzoldehyde.
A. - Br- melitensis; B - Br. abortus; C - Br. suis

(25 )
According to the method, this could only mean that tyrosine -is not
present.

The percentages of the two amino acids for the three

species of Brucella are given in Table 3.

(26 )

Table 3

Distribution of Tryptophane and Tyrosine in the
■Kndo-ant. 1gflns of the Three Species of Brucella
Per cent of
Tryptophane
Nitrogen

Per cent of
Tyros ine
Nitrogen

Per cent of
Tryptophane

Per cent of
Tyrosine

Br. melitensis

2.46

0.60

18.92

8.45

Br. abortus

1.49

none

11.46

none
........... ............ .— *=----------------------

Br. suis

0.91

0.66

7.00

9.29

(2 7 )
The results tabulated in Table 3, show that the endo-antigens
from the three species of Brucella vary widely in their tryptophane
and tyrosine content.

That the differences in percentages bear

no relation to the t5hl nitrogen content of the endo-antigen is
shown by the fact that the total nitrogen of the three endo-antigens
was: Br. melitensis 8.11 per cent, Br. abortus 9.08 per cent, B r .suis
10.07 per cent.

The surprising feature of these data is that while

the Rosenheim test for tryptophane was extremely faint for all three
endo-antigens,
three oases.

the Millon’s test for tyrosine was brilliant in all
The possible significance of this fact will be discussed

later.
It is apparent that the tyrosine and tryptophane nitrogen account
for but a small portion of the total nitrogen.

In every instance the

remainder of the nitrogen fell into two classes, namely, acid-amide
nitrogen and mono-amino acid nitrogen.

The acid-amide nitrogen values

remained almost constant throughout, averaging 1.57 per cent.

This

nitrogen is generally regarded as being derived from acid-amide link­
ages -CONHg*

Gortner and Holm have shown, however, that in a 24 hour

hydrolysis there may be an appreciable amount of deamination of the

amino acids, which nitrogen occurs in this fraction.

The remainder

(28 )
of the nitrogen of the endo-antigen fell into the class containing
the mono-amino mono-car boxy lie, and the mono-amino dfccarboxylic acids.
There was no indication of the presence of any of the basic amino
acids.
The Sugar Acid
The solutions remaining from the nitrogen determinations of
each endo-antigen were united.

These solutions were concentrated

to a small volume in vacuo at 40® C.

Addition of 5 volumes of

alcohol to each of those produced a precipitate, which upon stand­
ing formed a gummy residue.

This gummy precipitate was thoroughly

examined from only the endo-antigen of Br. melitensis.

it was

re-dissolved in distilled water, decolorized by boiling with
Norite.

Re-precipitat ion wi th alcohol produced an almost colloidal

white precipitate, which upon collecting became a colorless sirupy
material.

This material gives a slowly positive Molisch reaction;

nitrogen is absent; Bial’s test is negative; there is no reduction
of Benedict’s reagent; with FeCla a deep yellow color is given
which is produced by many alpha-hydroxy acids.
acidic, titrating slowly.

The material is

For complete titration it is necessary

to heat in a sealed tube in N/l alcoholic KOH.

The fraction

(39 )
appears to be an anhydride or a lactone of a sugar acid.

Because

of its sirupy nature, equivalent weights obtained are meaningless.
The compound is optically inactive.

Attempts to prepare saccharic

and mucic acids from it have been futile.
That this compound is a lactone of a sugar acid seems highly
probable.

Since it is optically inactive, this suggests that it is

produced in the course of the hydrolysis.

It is difficult to con­

ceive that jfts precursor could have been the usual type of reducing
sugar as those are completely destroyed by prolonged heating in
acid solution.

There is a possibility that it is formed from a

sugar acid present as such in the original endo-antigen, or that
it is fonned by oxidation and deamination of* an-amino sugar.
Lead Acetate Precipitation
The alcoholic supernatants from the above precipitation of the
sugar acid of each endo-antigen were combined and freed from alcohol
bjr distillation in vacuo at 40° C.

The aqueous concentrates were

thoroughly precipitated with neutral lead acetate, the bulky pre­
cipitate consisting largely of lead chloride.
tates were decomposed with

The lead precipi­

The supernatants from the Sulphide

precipitates, upon evaporation on the steam bath, deposited crystals

(3 0 )
of lead chloride.

Occluded to the lead chloride in a manner which

resisted separation during four recrystallizations from hot water
were small quantities of the fatty acid, the yellow fragrant oil
and the sugar acid previously mentioned.
products.

There were no nitrogenous

These materials were detected because of the persistent

antigenicity of the crystals.
saponification.

They were eventually liberated by

This surprising ability of an inorganic salt to

carry with it through repeated crystallization traces of an anti­
genic material, is worthy of further investigation.
The behavior of the three endo-antigens differed somewhat,
however.

With the endo-antigen from Br. suis needle-like crystals

separated before and during concentration for the precipitation of
the sugar acid and upon concentration for precipitation with lead
acetate.

This crystalline deposit was obtained only from Br. suis

and has not yet been characterized chemically.

The Br. melitensis

endo-antigen differed from the other two, in that while the lead
sulphide precipitates from the Br. suis and Br. abortus preparations
eould be decomposed with HC1 with the formation of PbCl*, the lead
sulphide precipitate from the Br. melitensis endo-antigen could not
be decomposed,

it was found to contain organic matter.

(31 )
It is significant to note that injection into normal rabbits of
the combined products obtained from lead acetate precipitation (although
they were nitrogen-free and had been hydrolyzed in strong acid ) pro­
duced agglutinins and opsonins for the intact Brucella cells.

This is

possibly due to the presence of small quantities of the diketone frac­
tion which has been previously noted as antigenic,

m

an earlier

preliminary report (17 ) mention has been made of a hydrocarbon, a
probaol® isomer of b-anthraquinone carboxylic acid, obtained from the
lead acetate precipitates of Br. melitensis and Br. abortus but not
from Br. suis *

Subsequent work has shown this product to be impure*

The anthraquinone derivative is present only in traces and not in
significant quantities as was previously supposed.

It is possibly

an artifact? The antigenicity previously attributed to it seems due
to traces of the diketone fraction.
The Remaining Nitrogenous Constituent
The supernatants from the lead precipitation yielded upon con­
centration, an orange sirupy material which became partially crys­
talline in the ice box, but a sirup again at room temperature.

This

sirup was thoroughly examined only in the case of the Br. melitensis
endo-antigen.

Upon saponification in 4 per cent alcoholic KOH an

(3 2 )
insoluble precipitate of inorganic matter is formed.
trace of lipide present.

There is no

The aqueous solutions from this saponifica­

tion contain nitrogen and give a positive Molisch reaction.

They

do not reduce Benedict’s reagent; no osazone is produced; and Bial’s
test is negative.

The nin-hydrin reaction is positive.

Although

the color developed in the Elson-Morgan test for glucose-amine
and chondrose-amine (18 ) deepens in proportion to the nitrogen con­
tent of the samples being tested, neither of these compounds was
isolated.
In Table 4 are presented, as complete as possible, summarized
data of the chemical constitution of the endo-antigens.

In con­

sideration of the chemical examinations of the endo-antigens, suf­
ficient data have been presented to demonstrate their dissimilarity
to any previously described bacterial products.

The differences

in the distribution of the tyrosine and tryptophane in the endoantigens from the three species of Brucella, as well as the differing
behavior of the hydrolytic products, demonstrate that the three
endo-antigens are not chemically identical.

Although quantitative

methods of determining the sugar acid and the remaining unidenti­
fied nitrogenous fraction have not yet been developed, the amounts

(3 3 )
of each obtained preclude the presence of any other constituents
in the endo-antigen.

In fact, assuming the sugar acid to arise

from the reducing sugars demonstrable in the endo-antigen, and
assuming the nitrogenous constituent to represent even a very
small mono-carboxylic amino acid, the percentages already add
up to more than 100 per cent.

This can only indicate that some

of the fractions are identical and are being counted twice as
would be the case if the residual nitrogen, the reducing sugars
and the sugar acid all had their origin in an amino-sugar.
An attractive but at present purely speculative correlation
of the chemical data is possible if one considers the proposition
of the endo-antigen being a combination of an amino-sugar with a
diketone.

That amino-sugars may form a variety of hetero-cyclic

compounds, including those containing a pyrrol ring has been d e ­
monstrated (19).

Their combination with diketones to form such

compounds is the basis of their colorimetric determination (18 ).
Precedence for such a combination is thus established.

The fact

that the tryptophane and tyrosine determination do not correlate
with the qualitative tests for these compounds might indicate
their presence not as actual entities, but as formed by the com-

(34 )
Table 4
Summary of the Chemical Constitution of
the Sndo-antisens of Brucella
Per cent

;al N.

:Br. melitensis
endo-antigen
strain 606

9.40

Lipide-extracted hr .abortus Br.suis
Br.melitensis
endo -antigen endo-an­
strain 1151 tigen
endo-antigen
strain
strain 606
1635

12.00

Lno N

-

lucing sugars
>fore hydrolysis

*

lucing sugars
‘ter hydrolysis

5.93

4.00

6.08
-

6.49

.fur

-

-

-

isphoriaus

-

-

-

8.86

5.86

-

-

4.4

8.13

-

-

5.82

10.03

8.53

6.00

9.00

ty acid

4.87

9.00

5.57

ptophane N

2.13

2.46

1.26

0.77

osine N

0.51

0.60

0.00

0.52

16.38

18.92

9.74

5.98

7.31

8.45

0.00

7.93

ityl groups

3.48

itone soluble
H o w oil

ptophane
osine

5.42

S-sub stance
strain
606

48.43

ar acid

Percent
not
determined

Percent
not
determined

Percent
not
determined

percent
not
dete m i n e d '

dentified
itrogenous
omponent

Percent
not
determined

Percent
not
de termine d

Percent
not
determined

Percent
not
determined

isch test

+ Slow

+ Very
slow ..
+

+ Slow

♦Slow

♦Slow

+

+

Weak

lfs test

+

ret test

+

+

+

Lon’s test

+

2+

+

snheim test

-t-'aint

Faint
negative

.Faint

Very
faint
+
Faint

+
!:®gat-?

135 )
bination of Other compounds,

The data disclose that a diketone is

present in, or closely associated with, the endo-antigen.

There is

also evidence that the final nitrogenous residue gives a color re­
action for an amino-sugar.

All of these data could be correlated

with such an hypothesis.
It is hoped that further work may complete the identification
of each of the constituents of the endo-antigen, and reveal the
manner in which they are combined.
Biological Activity of the iundo-antigen
The studies of the biological properties of various prepara­
tions of the endo-antigen were undertaken with the following objectives
in mind:

precipitability; antigenicity; toxicity and nature of the

toxic reaction in the guinea pig; sensitizing activity; immunizing
value and therapeutic property.

Three types of material, viz. endo-

antigen, lipide-extracted endo-antigen and acetylated endo-antigen,
were used in these studies.
Precipitation studies: The precipitation studies were made with
serums prepared for each of the three species of Brucella.

Normal

goats were injected intravenously with suspended living bacterial
cells from one agar slant of growth.

When the serums of the goats

(36

)

reached an agglutination titer of 1 -10 ,000 , the goats were bled to
obtain a large amount of serum.

The collected serums were diluted

one-half and sterilized by passing through a Seitz filter.

All pre­

cipitation tests were made in small glass vials using 0*2 cc. of
serum layered with 0.2 cc of the antigen solution.

The tubes of

serum and antigen dilutions were incubated at 37 ° 0 . for 2 hours
and then allowed to stand in the cold room at 10°C. over night.
The antigen dilutions were made on a dry weight basis.

The results

are recorded by giving the highest dilution of antigen showing pre­
cipitation.
in Table 5 are recorded the results of an experiment to determine
the

effect of pH, heating and aging on an endo-antigen prepared from

ar#

a bortus strain 1151.

one

of which was adjusted to pH 6 , one to pH 7 and the other to pH 8 .

The antigen was divided into three parts,

These were then divided into equal parts, one of which was left u n ­
heated and the other heated at 80°G. for 3 hours in a water bath.
The results indicate that the precipitability of the endo-antigen is
only slightly affected by heat, age and alkalinity.

(37 )

Table 5
Effect of Heat, tH and Age on Br# Abortus
Endo-antigen Strain 1151

Age of
Antigen
One week

Three
months

pH

Precipitation Titer of Antigen
Heated
TJnheated

6

409,600

409,600

7

409,600

409,600

8

409,600

409,600

6

409 ,600

304,800

7

409,600

304,800

8

409,600

103,400

(3 8 )
Since much emphasis has been placed on the importance of
acetyl groups in antigens, several experiments were made to
determine the effect of acetylation on the precipitability of
several endo-antigens.

The results of one experiment comparing

the precipitability of 4 antigens before and after acetylation
are recorded in Table 6 .

It is quite evident from the results

that acetylation enhances the precipitating property of the
endo-antigens.

The S-substanee is unaffected.

(39 )

Table 6
Effect of Acetylation on the Ability of
Endo-antigens to precipitate
Specific Serums
Precipitation Titer of Antigen
Before Acetylation

i^fter Acetylation

345,000

1,500,000

Inactive endo-antigen
from Br* abortus
strain 1X5!

57,600

921,000

Material recovered
after saponification
in the acetyl deter­
mination

1,800

921,000

409,600

409,600

Endo-antigens from
Br* abortus strain .|
1151

S-substance

(40 )

The effect of acetone and ether extraction on the pre­
cipitating ability of the endo-antigen was studied.
tive results are presented in Table 7.

Representa­

The data indicate that

extraction of the endo-antigen with acetone containing HC1 and
with ether causes some loss of ability to precipitate specific
serum.
ever.

The precipitating ability is by no means destroyed, how­

(4 1 )

Table 7
Effect of Lipide Extraction on the Ability
of endo-antigens to Precipitate Specific Serum
Precipitation Titer of Antigen
Before Extraction

After Extraction

Endo-antigen from
Br- abortus strain
TT51

573,440

192,560

Endo-antigen from
3r- melitensis
sTrain 748

437,400

128,000

Endo-antigen from
Br- suis strain
1635

349,830

316,200

(4 2 )

Extremely Significant results were obtained in the study of
precipitating ability of the endo-antigen after treatment with
N/2 NaOH at 80®C. for one-half hour and with subsequent acetyla­
tion or treatment with HH*OH of the product thus obtained.
results of one experiment are presented in Table 8 .

The

While heat­

ing the endo-antigen in n /2 NaOH destroys the ability of the
material to precipitate specific serum, this ability is restored
either by acetylation or by allowing the material to stand in
NH^OH, thus demonstrating that the beneficial effect which acetyla­
tion has been shown to have on the precipitating ability of the
endo-antigen is probably not due to the addition of acetyl groups,
but to some other influence.

These results correlate with those

presented by Felton and Prescott (10) in their work on the specific
substance of ^neumococcus Type I.

The amount of precipitate pro­

duced by the acetylated and by the NH^OH treated material was not so
heavy as that produced by the original esndo-antigen, indicating
possibly a partial depolymerization.

(4 3 )

Table 8
Precipitation of Specific Serum by an Endo-antigen
After Chemical Treatment
Precipitation Titer
of Antigen
Endo-antigen from
Br_. melitensis Strain
748
Endo-antigen heated
l/2 hour at 80°C. in
N/2 NaOH

512,000

None

Endo-antigen ace ty la ted
after heating in N /2
NaOH

182,000

Endo-antigen treated with
NH 4OH after heating in
N/2 NaOH

512,000

(4 4 )
Toxicity: In the preliminary studies of the toxicity of the
endo-antigen it was found that various preparations differed con­
siderably in their effect on the guinea pig*

Prom those early

experiments it was learned that the differences were due chiefly
to the size of dose used and the pH of the preparation*

The admin­

istration of too large a dose as well as of too small a dose of a
particular antigen failed to elicit a reaction in the guinea pig.
The toxic activity of each preparation of endo-antigen was arrived
at by injecting guinea pigs, weighing from 350 to 450 grams, intraperitoneally with one to two cubic centimeters of the preparations,
containing definite amounts measured on a dry weight basis*

The

criteria of their toxicity for the guinea pig are as follows:
Upon intraperitoneal injection into normal guinea pigs, the
first symptoms are rarely noticeable before the end of the first
hour and much more frequently at the end of the second or third
hours*

The temperature of the animal begins to fall and may con­

tinue to do so until death occurs, or the temperature may rise
again to normalcy if the injection is not fatal*

As the tem­

perature drops the animal becomes irritable and nervous.
on the back; becomes roughened.

The hair

The abdomen may become very tense

(4 5 )
and distended and a white purulent exudate may appear in the left
eye*

Toward the end of a fatal injection the animal is frequently

seized with periodic spasms*

Upon post-mortem examination the

most usual findings are an intense congestion of the blood vessels
of the mesentery, hemorrhagic areas in the mucosa of the stomach
wall,

the intestines, the greater omentum and the genital organs*

Of these, the mucosa of the stomach is most often and most severely
affected.

The spleen is often congested with blood*

In Table 9 are set forth the results of one of several exper­
iments which show the effect of pH, heat and aging on the toxicity
of an endo-antigen prepared from cells of Br* abortus strain 1151.
It was found that the antigens, shortly after preparation show their
greatest toxicity in 20 mg. doses.

Slight increasing or decreasing

of the dose from the optimums, results in little or no reaction in
the guinea pig.

In determining the toxicity of a particular anti­

gen at intervals after preparation, the usual procedure followed
was to inject guinea pigs with doses of 1,5,10,15,20, and 25 mg.
The results would reveal any decrease or increase in toxicity.
The results in Table 9 are tabula ted only for the amount of antigen

(4 6 )
which produced the maximum reaction or death of the guinea pig.
Preparations that were allowed to age at 10°C., especially those
made at or below pH 7 and heated, increased in toxicity.

The

previous size doses then failed to elicit reactions in the
guinea pig while smaller ones were effective.

Preparations

made at a pH above 7 tend to decrease slightly in toxieity with
age.

{47 )

Table 9
The Effect of pH, Heat and Aging on the
Toxicity of the Endo-antigen
Age of
Antigen
One week

Toxic Dose

PH
Unheated

Heated

6

20 mg* Died 24 hrs.

10 mg. Died 4 days

7

20 mg* Died 24 hrs*

20 mg. Died 7 days

8

20 ng* Temp, fell to
35.5®0. Recovered

Two months

8
jrour months

8

10 mg. Temp, fell to
37*C. Recovered

20 n g . Die d 10 hr s .

5 mg. Temp, fell to
37.5®c. Recovered.

1 mg. Temp, fell to
37.5®<J. Recovered

20 mg. Died 10 hrs.

25 mg. Died 7 hrs.

20 mg. Died 9 hrs

20 mg. Died 10 hrs

5 mg. Temp, fell to
33.2®C- Recovered

1 mg. Temp, fell to
36®0. Recovered

20 mg. Died 6 hrs.

25 mg. Died 7 hrs.

20 mg. Died 4 hrs.

(4 8 )

The effects of acetone and ether extraction on the toxic
property of the endo-antigen were quite significant.
tive results are recorded in Table 10*

Representa­

In each case it is seen

that the toxic activity of the endo-antigen is markedly increased
by the extractions.

(49

)

Table 10
.Effects of Lipide Extraction on the Toxicity
of the Endo-antigen
Toxicity
Before Extraction

After Extraction

20 mg. Temp, fell
to 36.t>°G. Re­
covered

1 mg. Temp, fell
to 37.1°0. Re­
covered.

Endo-antigen from
Br. melitensis
strain ?48

20 mg • Die d 4 br s •

5 ng • Temp, fell
to 37. 2 °0 . re­
covered

Endo-antigen from
Br. suis strain
1635

10 mg. Temp, fell
to 37 .4 °C. Recov­
ered.

5 mg. Temp, fell
to 34.4°C. Re­
covered.

Endo-antigen from
Jt$r. abortus strain
1151.

1

(5 0 )
The effect of acetylation on the toxicity of the endo-antigen
is less clear.

While the toxicity was never destroyed it appeared

to he considerably reduced.

In endo-antigens where a lethal dosage

was found before acetylation, none could be discovered afterwards
although a variety of dosages showed mild

toxicity.

The same

findings occurred when the endo-antigen was heated with n/2 NaOH
at 80° C.

for one-half hour.

The original toxicity was not r e ­

stored in the latter by treating with IIH4OH or by acetylation.
One may summarize that the toxic property of the endo-antigen is
not easily destroyed, but once impaired, its original toxicity
cannot be restored.
Antigenicity:

As a rule guinea pigs were used to study the

ability of the endo-antigen to provoke antibodies.

The rapid and

test-tube agglutination tests were used to measure agglutinin
production; the phagocytic system previously described
used to measure opsonin production.

(lo ) was

Those pigs in which a given

intraperitoneal* dose of antigen was non-toxic or which failed to
succumb following a systemic reaction were bled on the tenth day

(5 1 )
after injection and examined for antibodies.

Preliminary

examinations showed that the antibodies reached their peak by
the tenth day.
In Table 11, are presented representative data from one of
several experiments which illustrate the antibody response to
varying doses of heated and unheated antigen.

It is interest­

ing to note that antigenicity and toxicity correlate very closely.
When large doses of the antigen fail to provoke a toxic reaction
in the guinea pig, they also stimulate little if any antibodies.
A one milligram dose of the antigen, if toxic, appears to be
just as capable of stimulating antibodies as larger doses.

The

phagocytic system was found to be much more delicate than an
agglutinin or precipitin system for measuring the antigenicity
of the endo-antigens.

Many times opsonins were produced when

the presence of agglutinins could hardly be detected.
There was found to be little difference between the anti­
genicity of heated and unheated antigens.

The storing of anti­

gens in a cold room at 10 °C. for four months does not impair
their antigenicity.

152 )

Table 11
The Antigenic Activity of the .Sndo-Anti gen

Dose

Antigen

Unheated

Heated (b)

Toxicity

20 mg.

Agglutinin
Titer

Negative

0

1

8

16

5 mg.

Slight

+1/25

18

5

2

0

1 mg.

Slight

+1/500

18

7

0

0

20 mg.

Negative

6

9

1 10

0

10 mg.

Slight

+ 1/100

20

5

0

0

5 mg.

Slight

+ 1/100

21

4

0

0

1 mg.

Slight

Pl/25

16

9

0

0

-

(a) Ma-Cells showing marked phagocytosis
Mo-Cells showirg moderate phagocytosis
S

Opsonic Activity
Cells (a)
S
N
Mo
Ma

Cells showing slight phagocytosis

N — Cells showing no phagocytosis
(b ) At 80°C. for 2 hrs*

(5 3 )
When the endo-antigen is freed of its lipide content, the
antigenicity is unimpaired,

it still correlates with the toxi­

city of the material, however.
In Table 12 are recorded the effects of acetylation on the
antigenicity of the endo-antigen.

It is seen that antigenicity

is unquestionably enhanced by the process.

This is deomonstrated

most clearly in the case of the product recovered from acetyl
determination and in the case of the S-substance.

Although the

toxicity of the endo-antigen is changed by acetylation so that
there is no lethal dose determinable and so that there is little
gradation with dosage, the antigenicity is increased regardless
of the size of the dose.

The correlation between toxicity and

antigenicity is,therefore, no longer present.

(5 4 )

Table 12
Effects of Acetylation on the Antigenicity
of the JUndo-antigen and of the S-substance
Opsonic
Activity (a )

Agglutinin
Titer

Dose
Before After
Acetyla­ Acetyla­
tion
tion

Befor e
Acetyla tion

Before
Acetyla­
tion

After
Acetyla­
tion

After
Acetyla­
tion

Ma MO S N Ma Mo s N
Endo-antigen from
Br* abortus strain
U51
’
-

20 ng.

20 mg*

P 1/50

P i/loo 22

3 0 0 24

1 0 0

Inactive endo-anti­
gen from Br*abortus
strain llfTT

20 mg.

20 mg.

P 1/50

+ 1/500

7

6 6 6 25

0 0 0

Material recovered
after saponifica­
tion in the acetyl
determination.
y— -----------------3-substance from
Br. melitensis
strain 606

20 mg.

15 mg.

Negative

+ 1/50

16

5 4 0 25

0 0 0

15 mg.

15 mg.

Negat ive

+ l/50

8

13 4 0 25

0 0 0

[a) Ma - Oells showing marked phagocytosis
Mo - Cells showing moderate phagocytosis
S —

Cells showing slight phagocytosis

N —

Cells showing no phagocytosis

..

.

(5 5 )
In the experiment in which the endo-antigen was heated in

n /2

NaOH for one-half hour at 80°Q. antigenicity was completely de­
stroyed*

Treatment with NH^OH failed to restore the antigenicity

while acetylation brought about but a very slight restoration.

As

it was noticed in the study of this treatment on precipitating
ability of the endo-antigen that partial hydrolysis had probably
occurred, it may be supposed that this explains the failure of acetyla­
tion to restore complete antigenicity in this case.
Allergic Activity of Endo-antigens;

The comparative allergic

activities of endo-antigens prepared from the three species of Brucella
were determined on sensitized rabbits.

The rabbits were sensitized by

one intravenous injection of 1 cc, of living cells suspended in sterile
saline solution.

The turbidity of the suspension was made to compare

with tube 2 of the McFarland Nephelometer.

This procedure will pro­

duce a high degree of skin sensitiveness to Brucella in 30 days for
measuring quantitatively the allergic activities of various prepara­
tions.

When antigens prepared from each of the three species of

Brucella were compared as to their allergic activities, successive
dilutions ranging from 1-5000 to 1-1,280,000 were made in sterile

(56 )
saline solution.

An injection of Q.lcc of each dilution of the

three antigens was then made intraderraally into a rabbit sensitized
££* abortus, one to Br. suis and one to Br. melitensis.

The

intraderraal tests were observed at 24 hour intervals for 5 days.
The reactions were recorded at the 48th hour, although they often
persisted for as long as 5 days.
In Tables 13 and 14 are recorded the results of two of several
experiments designed to determine quantitative differences in the
allergic activity of acetylated and lipide extracted endo-antigen
in rabbits sensitized to Br. abortus, to Br. suis and to Br. melitensis.

Rabbits sensitized with Br. abortus were found to show

the same degree of allergic response to both Br. abortus and Br.
suis, acetylated and lipide extracted endo-antigens.

The allergic

titers of the two melitensis antigens were considerably lower than
those from Br. abortus or Br. suis.

The allergic titer of Br. suis

acetylated antigen was considerable higher than the 3r. abortus or
Br. melitensis antigens in B r . suis sensitized rabbits.

Lipide

extracted antigens induce little or no allergic response in dilu­
tions of 1-5000 and above in Br. suis sensitized rabbits.

Rabbits

sensitized to Br. melitensis showed about the same allergic response

(57 }
to dilutions of the acetylated antigens and to dilutions of the
lipide extracted antigens*

The allergic titer of the melitensis

antigens was either low or negative in the Br. melitensis sensitized
rabbits•
It is apparent from the results that rabbits, when sensitized
to the different species of Brucella show a considerable difference
in their allergic response to varying dilutions of endo-antigens pre­
pared from each of the three speeies*

If one employs the proper

dilutions of acetylated and lipide extracted endo-antigens it should
be possible to detect the species of Brucella responsible for an
j

existing state of allergy.

It is believed that in these antigens

may be found the means for detecting the infecting species of Brucella
in human beings.

sCD

o

o
p
cf

co

K*
H*
cf

4

O

p

P

C
O

W
P
o'
o'

CO

CD

o'
0
4

H*

P

01

cf
p

CD

P

01

CO
00

N

0

H*

P*

cf
O

I
<D

f-*
H*
cf
CD

P
0

G
O
p

0

p

o'

o
4

Cf
p

CQ

&
0

f-*
H*
Cf
0
J3
0
H*
0

0
P
H*
0

P
O'
O
4
cf
P
0

g
C
D
H*
H*

(0
P
H*

C
O

c+

p

o'

o
4

cf
p
C
O

CD

P

01

0
H
H*

0
P
H*
0

cf
0

P

0
H*
0

H*

C
O

P
O'
o
4
cf
P
0

00

ft
P
Pi
O
1
P
P
cf
H*
(ft
0
P

f3
P*
0
O
O
0
hd
p

I

Ol

4

P

cf

H*

<
0

Oi

Ol

0)

H
I‘
p
O

i>
f—1
1—i

t?3 0
P 4

co

CO
7
T
p.

K
I
CO

CO

P

1-3
P*

W

o

0
P

O

co

CO

CO

Ol

Ol

cf

H
t

H*
O
P
0

O

o

£0

M
I

03

00

O

i-3
P*

P

O

03

CO

4)

cf

H*

o

M
I
M

p

0

o»
o
f3

O

P*

fctf
P
P
0
1

H
I
03
CO

C
O

&

O

0*)

9

0

P
0

CO

1 I

I
o>

o

•-3
P*

CO

CD

Pi 0*3 i-3

O H* P
1 O o' «»—.
f-* Ol
oo
P > 0

cf O
H* cf f-»
n h * 03
0 <4
P H*
0 cf
*«
o

t>
0
O

cf

<<
M
P
cf
0
Pi

U— rn
it
ii
ii

n
n
it

H
II
II
II

n

O
O
P
c+
4
O
H

ii
ii

&
0

0
0
H*
0

P

ti

n

m10
1
IIH
II H*
II c+

110
110
II H*
110

IIP

«

4
ct
P
0

ii
ii
ii
ii
ii

ii
ti

_
ii_
n

o

0

it
ti
it
ii
n

C
D
0
P
0
Hc+
M
*
N
0

0
cr

P
H*

ii
ii

H*
H*
et

n

0

n

0
p
H*
0

0
O'

o

4
ct
P
0

.

0
P
H*
0

0

Cf
o
ct
P

0

.

II&
II0
IIM

II H-

Hct

0

II P

03
P
H*

P
O'

0

4

ct*

3
0
J
—
I
Hc+

p

0

0

P

o

0
P.

0

tel
P

0

4
ct
P
0

o

H*

O'
o

i

0

p
ct

0

II 0
II H II 0

H-

H-

0*3
0
P

0

It

H

'it--

I—1
I
CJ1

fl

U
Hi

n

III

HS

II§
II•

-«-----

H

II

II

n

HI

ti i
ii

n

II

H
IL_
H
H
II

H
II
RI
H

II
II
fe­
ll

II
II
ItI
H
II
II
II
II
II I

II
n
ii

IT"
It

-H-----

II I

H
H

II

-»—

I)
II
II
II I

II
II
II
-II—

II
II
II
II I
II
It

n
u—
n
n
ii

ni
ii
if
n
n—

II

ii
ii

n

it

III
ii

n
N

n
it

R

H
I
A
0
►
*3
{3*

II
It
II I
It
II
II

It
U
—
II
H

ti­
ll

C
O

ii
ii

i

H

II
II

fl

n
j_

II
II
II
II
II
II

"P1
00

I

o

-IIM----II
II
II
II
II
II

I

_
J
J
__
It
II
II
II I
II
II
II
dt-----It
II
II
II I
II
II
II
I t -----It
II
It I
II
II
(I
II
It
II
II
II__

£0

M
I
^iiriR

Ol

0

•-3
P*

"P*
1
03
CO

o

0

O
P
O
t-b

S'
sH*
O

w

f*r

o

n
-H-----

Hi
H

II
II
II

t

P
4
H*

Jj>
M

l
I—
i
O
t3
P*

II
II 03

S
3

ct

-H
-----

II

O

o

P*
0

i-3

H
H
II

n
III

H*
ct

ct
O

R

II

o'
o'

Pi

n

II
110
110
II H
IIH*
II ct
110
HP
110
IIH*
110
II

W
0

}>
O
c+

cf

0

0
O
c+

H-

o
0
o

e+

<<
O
4>

p

hd
H*
po

t*
H*
P
ct

c+

4

0
O
e+

H*
O
P

0

0

Pi

o

tel

4>
s

pi

P
P*
0
1

0

P

{3*

0
1

I
01

&
c+
H*

0*3

0*3

0

0

$
H1

H-'

CO

CD

O

t-3

0

0
P

0

ct

H0

P

M
0

p

Ol

to

(60)

value of the endo-antigen:

All attempts to im­

munize guinea pigs with various preparations of the endo-antigen
in varying size and number of injections against experimental
Brucella infection have thus far failed*

Guinea pigs in groups

of five have been injected intradermally, subcutaneously and intraperitoneally with from one to nine successive doses of the antigens
in amounts varying from 2 mg. to 20 mg.

After an interval of 30

days the pigs were exposed to infection by placing one drop of a
dense suspension of a newly isolated anaerobic strain of Br. abortus
in the conjunctiva.

At the end of five weeks after exposure, the

treated pigs as well as an equal number of untreated controls were
killed and the organs examined for gross changes and cultured.
All pigs treated with varying size doses.of the endo-antigens as
well as the untreated controls, showed evidences of infection,
grossly and culturally.

Since guinea pigs are not easily immunized

against Brucella infection and since those used in these experi­
ments were exposed to a massive dose of the virulent culture, the
results do not necessarily mean that the antigens are without
immunizing value.

(6 1 )
Studies are being conducted to determine the immunizing value
of the antigens against Brucella infection in cattle and human
beings.

While the studies are far from being complete, the results

thus far show that if the proper size dose of the lipide extracted
endo-antigen is injected intradermally, cattle and humans develop
a high Brucella opsonin
Therapeutic value

titer which persists for more than 6 months.
of the endo-antigen: Several experiments

were carried out on small groups of guinea pigs 30 days after
exposure to a virulent culture of Br. abortus, to determine whether
the course of infection could be altered by the administration of
endo-antigen.

The endo-antigen was used intradermally and sub-

cutaneously at intervals of 3 days and 7 days in 5 and 10 mg. doses.
One group of pigs received one injection, another 3 injections and
another 9 injections.

Infected guinea pigs are very allergic to

Brucella antigens, and consequently many pigs died within 24 hours
following the injection

of the larger amountof endo-antigen.

When

the surviving pigs were

killed 5 weeks after the last injection of

antigen in each series of experiments, all showed gross changes in
the organs characteristic of Brucella infections, The organism was
cultured from all pigs exposed to infectioh.

(6 2 )

Biological Activity of the Constituents of the 3ndo-antigen
Uach of the constituents obtained from the endo-antigens, has
been examined for ability to precipitate specific serum, for toxicity
and for antigenicity.

None have been found to be toxic and none have

been found to precipitate immune serum, although the lipide fractions
could not be examined for precipitating ability because of their
insolubility in water or in saline.

The portion of the endo-antigen

removed by acetone extraction has, however, been found to produce
agglutinins and opsonins for Brucella upon injection into normal
animals.

This fraction is nitrogen free.

Likewise, the material

precipitated by lead acetate, after hydrolysis in 20 per cent HC1,
has been found antigenic.

This activity may also be due to the

diketone fraction, as a small amount of this material was removed
from the lead acetate precipitate.

(63 )
Physiological Studies
Blood sugar: It was noticed by Delafield (21) that the toxic
antigenic fraction isolated by Baistrick and Topley (6 ) from
Salmonella aertryoke had a hyperglycemic effect when injected into
animals#

The endo-antigen from Brucella was examined for this

activity.
In a single experiment a guinea pig was bled from the heart
immediately before, 3 hours after and 24 hours after intraperitoneal
injection of 5 mgs. of endo-antigen.
Somogyi-Shaffer-Hartmann method (22).

Blood sugar was run by the
The following results were

obtained:
Blood sugar in mgs.
per 100 cc.
Before injection:

137 mgs.

3 hours after injection:

174 mgs.

24 hours after injection

97 mgs.

The endo-antigen is thus seen to bring about in injected animals
a hyperglycemia which is followed by a hypoglycemic condition.
Basal metabolism: A study was made of the effect of injections
of the endo-antigen on the basal metabolism of the guinea pig.

A

glass metabolism chamber was arranged so that a regulated amount of

(6 3 )
dried, COg free air could be drawn through it.

The moisture and

C0e of the air leaving the chamber were removed by passage through
Dehydrite and ftscarite respectively.

A normal metabolism was run

on each animal before its use in the experiment.
varied greatly in size and age.

The animals used

The variation in basal metabolism

of the normal animals is explainable largely by this factor.

Young

guinea pigs invariably had a higher metabolic rate than did the
older animals.
body weight.

The amounts of endo-antigen injected varied with the
The first tests were run for a period of 8 hours.

second set of tests were run for 4 hours.
are recorded in Table 15.

A

The results of these tests

(6 4 )

Table 15
The Effect of Endo-antigens on Basal Metabolism

Antigen

Calories per kilogram per hour
8 hour period
4 hour period
Before
After
After
Before
injection
Injection
Injection
injection

Br. abortus

10 mgs. per kg.

3.35

4.03

8.6

8.8

20 mgs. per kg.

2.56

3.80

11.8

4.7

30 mgs. per kg.

2.56

4.15

3.7

0 t65

8.3

4.1

2.8

2.56

4.5

3.2

10 mgs. per kg.

3.1

3.0

20 mgs. per kg.

10.5

7.7

30 mgs. per kg.

2.5

Br. suis

10 mgs. per kg.
20 mgs. per kg.
30 mgs. per kg.

*

Br. melitensis

0.97

(6 5 )
In most cases, if the animal is observed over a period of 8
hours, a definite decrease in the basal metabolism is noted.

If,

however, the time of observation is limited to 4 hours the data
show a definite increase in the calories consumed per kilogram
per hour.

In other words, upon intraperitoneal injection of the

endo-antigen, the basal metabolism of the animal is at first stimula­
ted, and is then depressed, the depression usually being more than
sufficient to counteract the increase.

(66 )
DISCUSSION
The endo-antigens prepared from the three species of the
Brucella differ from any previously described bacterial fractions.
Although their structures have not yet been fully elucidated it
appears from the data presented that they may be described as
neither protein nor carbohydrate in nature, but a combination of
certain units of both of these classifications with a lipide frac­
tion.

The endo-antigens are thus somewhat similar to the prepara­

tions described by Boivin and coworkers (7, 12), but differ from
these in their stability, their incapability of being separated
into specific non-toxic and toxic non-specific fractions and in
their containing no phosphorus or sulfur.
In previous work on the chemistry of the Brucella group a
soluble specific substance, the .S-substance, was found to exist in
Br. melitensis but could not be prepared from the other two species
(1, 2, 3).

A very similar substance was obtained, however, by

chemical treatment of a protein-like fraction, the albuminoid, occur­
ring in all three species.

The present work has shown the identity

or close similarity of the endo-antigen with the 3-substance, and
has shown that the endo-antigen may be prepared from the albuminoid

of all three species.

The implication is, therefore, that the

endo-antigen is united with a protein in the bacterial organism,
occurring in a free state only in Br. melitensis.
The quantitative determination of tyrosine and tryptophane in
the endo-antigens of B r . melitensis, B r . suis and Br. abortus demon­
strates clearly that the three endo-antigens are not chemically iden­
tical.

There has been, as yet, no clear demonstration of species

variation in the other constituents of the endo-antigen, but this
must obviously appear upon further investigation.
The relationship of acetyl groups to the biological properties
of the endo-antigens does not emerge clearly from the work accom­
plished.

Although distiliable acid, v/hich is believed to be acetic

acid, is obtained in the determination of acetyl groups, the bio­
logical properties of the endo-antigen do not correlate closely
with the acetyl content, as would be expected if the groups possessed
the importance delegated to the acetyl content of the specific poly­
saccharide of the Type 1 Pneumococcus (9 )•

On the other hand,

acetylation of the endo-antigen brings about a remarkable increase
in its biological properties, although, here too the increase in

antigenicity does not always correlate with the increase in acetyl

(68 )
groups*

It has been shown further that when the ability to pre-

cipitate specific serum has been lost by the endo-antigen, this
property may be restored either by acetylation or by allowing the
endo-antigen to stand in 25 per cent NH*0H at 10°C., in a manner
similar to that demonstrated for the restoration of the specific
polysaccharide of Pneumococcus Type 1 (10).

One must conclude,

therefore, that the presence of acetyl groups in the endo-antigens
is not requisite to their biological activity, but that their pres­
ence induces a state which is necessary for biological activity and
which can be reproduced in their absence by other means.
The role of the lipides in the endo-antigen is not clear from
the present work.

Whereas the lipide-extracted endo-antigen is

increased in toxicity and antigenicity it shows a decline in the
i

ability to precipitate specific serum.

It has also been noticed

that the lipide-extracted endo-antigen is much more sensitive to
the presence of alkali than is the unextracted antigen.

It seems,

thus, that the lipides play a part in the antigenic pattern of the
endo-antigen but do not contribute to toxicity or antigenicity.
The surprising dependence of the toxicity and antigenicity of
the endo-antigen upon the optimum dosage, and the disappearance of

this dependence after acetylatioh, do not admit ready explanation*
Similar evidence of antigenicity in small doees but not in large
doses has been noticed with the specific polysaccharide of Type
1 Pneumococcus (9 ) and with the Porssman antigen (23)* The sig­
nificance of this observation on the application of the endoantigen to other animals and in attempted immunization is being
investigated.
Although the experimental data are not yet complete, the endoantigens show some promise of value as immunizing, therapeutic, and
diagnostic agents.

Attempts to experimentally immunize guinea pigs

have, it is true, been unsuccessful although the immunizing injedtions stimulated excellent antibody production in the animals.
Because of work done in this laboratory on guinea pigs with agents
known to immunize other animals, the use of guinea pigs in immunizing
experiments seems questionable.

Similar results have been obtained

Gwatkin (24) in his work with Brucella fractions*

by

Incomplete work on

the immunizing value of the endo-antigens for cattle and humans appears
very promising.
The therapeutic value of the endo-antigen for infected guinea pigs

(7 0 )
has been found to be negligible.

Here, as with experimental immuni­

zation, the use of guinea pigs as experimental animals seems to be
of little value.

There is excellent evidence that at least a part

of the activity of Brucellin, a therapeutic agent developed in this
laboratory and used very successfully in humans (25, 26), is due
to the presence of endo-antigen in that product.
has been isolated from this agent.

The endo-antigen

Experimental therapy in humans

with the endo-antigen is being investigated.
As diagnostic agents the endo-antigens have been investigated
only in regard to their ability to elicit shin reactions in sensitiz­
ed animals.

As shown in the experimental part, the acetylated and

lipide-extracted endo-antigens show some species differentiation,
which gives hope for the development of an agent capable of differ­
entiating infections due to Br. abortus, Br. suis and Br. melitensis.

(7 1 )
SUMMARY
A method is described for the preparation from cells of B r *
abortus, B r . melitensis and Br* suis of a highly antigenic fraction,
the endo-antigen, which is toxic for normal guinea pigs, and which
precipitates immune serum in dilutions of from 1-500,000 to
1-5,000,000*
Endo-antigens obtained from the three species of Brucella are
grossly similar.

The endo-antigen comprises roughly 25 per cent of

the bacterial cell.

While containing the same or similar constituents,

however, the endo-antigens from the three organisms have been shown
to differ markedly in the distribution of some of these constituents.
Positive reactions are given to the Molisch test, the biuret
test, Millon’s test, Bial*s test and a very slight reaction to the
Rosenheim test.
to 8 per cent.

The nitrogen content of the fraction varies from 6
Reducing sugars are absent before hydrolysis.

Cal­

culated as glucose after hydrolysis reducing sugars represent from
4 to 12 per cent of the endo-antigen.
and sulfur are absent.

Amino-nitrogen, phosphorus

In the determination of acetyl groups, dis-

tillable acid representing an average of 6 per cent of the endoantigen is obtained.

This acid is presumably acetic acid although

that product has not been isolated#
From the endo-antigen there may be extracted by acetone and
ether a compound having the properties of a di-ketone, and an
acetone soluble, saturated liquid fatty acid.

These two compounds

represent from 10 to 15 per cent of the fraction.
The acetone-ether extracted product still reacts positively
to the above mentioned qualitative tests.

Tryptophane and tyrosine

have been found to represent 18.92 per cent and 8.45 per cent, 11.46
per cent and 0 per cent, 7 per cent and 9.29 per cent of the ex­
tracted endo-antigens of Br. melitensis, Br. abortus and Br. suis
respectively.
From the remaining 65 to 70 per cent of the original fraction
there has been obtained an unidentified nitrogenous fraction and an
optically inactive sugar acid.

These are obtained in quantities such

as to preclude the occurrence of any further compounds in a signifi­
cant amount.
The endo-antigen is shown to be relatively stable in the presence
of dilute acid, and dilute alkali, upon heating, and upon long stand­
ing.

Its activity is not completely destroyed by hydrolysis with

dilute acids, but is destroyed by similar treatment with dilute alkali.

The ability to precipitate specific serum is lessened by extraction
with acetone and ether, but is enhanced by acetylation or by treat­
ment with 25 per cent K H 4OH*

The toxicity and antigenicity of the

endo-antigen are shown to be dependent upon proper dosage, an overdosage as well as an under-dosage giving poor results*

The toxicity

and antigenicity are increased by lipide extraction of the endoantigen*

Acetylation causes a distinct decline in toxicity but a

marked increase in antigenicity*
The endo-antigen elicits specific skin reactions in sensitized
animals, the lipide-extracted and the acetylated endo-antigens show­
ing some species specificity in this reaction.

Injections of endo-

antigen have failed to protect normal guinea pigs from subsequent
exposure to virulent Brucella organisms and have failed to alter the
course of the disease in infected pigs.

Experiments with cattle and

humans, however, seem to promise immunizing and therapeutic value to
the endo-antigen.
Injection of the endo-antigen causes a hyperglycemia followed by
a hypoglycemia in experimental animals.

The basal metabolism of in­

jected animals is at first stimulated and then depressed.

A leuko­

penia, chiefly due to the disappearance of neutrophiles from the

(74 )
peripheral blood, follows injection of the endo-antigen into
normal guinea pigs.
The endo-antigen may be produced from the previously des­
cribed albuminoid fraction of the Brucella, thus accounting for
the toxicity of that fraction and suggesting that this albuminoid
is a combination of the endo-antigen with a protein-like group.
The endo-antigen is shown to be similar to or possibly identical
with the previously described S-substance, the latter being pro­
bably a partially hydrolyzed endo-antigen.
The Author wishes to express his grateful appreciation to
Dr. I. Forest Huddleson for his guidance and cooperation in this
problem, to thank Mrs. Myrtle Munger for conducting the blood
examinations mentioned in the study, and to thank Drs. G. D. Ball
and C. A. Hoppert of the Department of Chemistry for their helpful
suggestions and advice.

(7 5 )
BIBLIOGRAPHY
1.

Huston, R. C., Huddleeon, I.F. and Hershey, A.D. - 1934.
"The Chemical Separation of Some Cellular Constituents of
the Brucella Group of Micro-organisms." Technical Bulletin
No. 137, Agr. Exp. Station, Michigan State College.

2.

Hershey, A.D., Huddleson, I.F. and Pennell, R.B. - 1935
"The Chemical Separation and Biological Activity of the
Polysaccharide Constituent in Brucella Cells•" Journal of
Infectious Diseases, 57, 183.

3.

Schapira, G. Bianealani - 1936 "Ulterior! indagini biochimiche
intorno ai polisaccaridi ricavati wdai batteri del gruppo
brucella (varietSi paramelitensis )." Lo Sperimentale 90, 450.

4.

Vaughan, V. C. and Vaughan, J.W. - 1913 "Protein Split Products
in Relation to Immunity and Disease.".
Lea and Febiger, Phila.

5.

Gvsratkin, R. - 1935
" A Study of Various Fractions of Brucella
Abortus.
I. Preparation, Toxicity and Biochemical Nature of
Alcoholic Precipitate".
Canadian Journal of Research, 12, 115.

6.

Raistrick, H. and Topley, W. W. C. - 1934 "Immunizing Fractions
Isolated from Bact. Aertrycke.".
British Journal of Experi­
mental Pathology, 15, 113.

7.

Boivin, A. and Mesrobeanu, L. - 1934 "Remarques concernant la
technique d Textraction du Complexe Polysaccharidique Antige*—
nique renferm^ dans le Bacille d fAertrycke." Comptes Rendues
Societe* de Biologie, 115, 304.

8.

Kuhn, R. and Roth, H. - 1933 "Mikro-Bestinmtt*ng von Acetyl-,
Benzoyljund O-Methylgruppen." Berichte der Deutsche Chemische
Gesellschaft 66B, 1274.

9.

Avery, 0*T. and Goebel, W.F. - 1933 "Chemoimmunological Studies
on the Soluble Specific Substance of Pneumococcus. I. The
Isolation and Properties of the Acetyl Polysaccharide of Pneumococcus Type I." Journal of Experimental Medicine 58, 731.

10. Felton, L.D. and Prescott, B. - 1936 "Studies of the Immunizing
Antigen of Pneumococcus Types I and II. "Journal of Bacteriology,
31, 69
11.

Anderson, R.J. - 1929 "The Chemistry of the Lipoids of Tubercle
Bacilli.
III.
Concerning Phthioic Acid. Preparation and Pro­
perties of Phthioic Acid." Journal of Biological Chemistry,
83,169.

12.

Boivin, A., Mesrobeanu, I., Mesrobenau, L. and Nestorescu, B.-1934
"Extraction d'un Complexe polysaccharidique toxique at antige*nique, a partir de diverses Bacte*ries autres que le Bacille
d*Aertrycke." Comptes Rendues Societ/ de Biologie 115, 306.

(76 )
«

13.

Frankel, S. and Jellimtek, C. - 1937 "Uber die sogenannte
Kohlenhydratgruppe in Eiweiss (Darstellung der Glucosamino—
Mannose )". Biochemische Zeitschrift 135, 392.

14.

Levene, P.A. and More, T.
- 1929 "The CarbohydrateGroup
Ovomucoid." Journal of Biological Chemistry 84, 49.

15.

15.

of

Gortner, R. A. and Holm, G. E. - 1920 "The Origin of the
Humin Formed by the Acid Hyrolysis of Proteins, v.*Journal
of the American Chemical Society 42, 821.
Van Slyke, D. D. - 1915 "Improvements in the Method ofAnalysis
of Proteins by Determination of the Chemical Groups Character­
istic of the Different Amino Acids." Journal of Biological
Chemistry 22, 281.

17.

Pennell, R. B. and Huddleson, I.F. - 1937 "The Chemical Con­
stitution of the Endo-antigen of Brucella Cells." Journal
of Bacteriology, 33, 42.

18.

Els on, L.A. and Morgan, W.T.J. - 1933 "A colorimetric Method
for the Determination of Glucosamine and Chondrosamine."
Biochemical Journal 27, 1824.

19.

Pauly, H. and Ludwig, E. - 1922 "Glucosamin als Bildner heterocyclische Verbindungen*" Zeitschrift fur Physiologische Chemie 121,
170.

20.

Huddleson, I.F. - 1934 "Brucella Infections in Animals and Man;
Methods of Laboratory Diagnosis." Commonwealth Fund, New York.

21.

Delafield, M.E. - 1934 "Blood-Sugar Changes and Toxic Effects
Produced in Rabbits by Certain Fractions Derived from Bact.
Aertrycke." British Journal of Experimental pathology 15,130.

22.

Somogyi, M. - 1930) "A Method for the Preparation of Blood Fil­
trates for the Determination of Sugar." Journal of Biological
Chemistry 86,655.

23.

Levene, P.A., Landsteiner, K. and van der Scheer, J. - 1927
"Immunization Experiments with Lecithin." Journal of Experi­
mental Medicin* 46,204.

24.

Gwatkin, R. - 1935 "A Study of Various Fractions of Brucella
Abortus. III.
Immunity Experiments with Alcoholic Precipitate."
Canadian Journal of Research 12, 133.

25.

Huddleson, I.F. and Johnson, H.W.- 1933 "Brucellin, a Possible
Specific for Undulant Fever in Man." American Journal of Tropical
Medicine, 13, 485.

26.

Huddleson, I.F., Johnson, H.W. and Beattie, C.P. - 1936 "Undulant
Fever: A Report of 100 cases treated with Brucellin". Technical
Bulletin Ho. 149, Agricultural Experiment Station, Michigan
State College.