a

146
Hl

Report of Dr. Goebel.

On the nature of pneumococcus antibody and its union with hapten.

The older concopts regarding tie non-protein nature of antibocies, and the
hypothesis of Buchner that antigen itself formed a part of the antibody com-
plox mst now be regarded merely as evolutionary steps in an attempt to
understand the true nature of these biologically active substances. During
the past decade those older theories have gradually given way to accumlated
chemical evidence which at the present tine cloarly points toward the fact
that antibodies are in reality modified serum globulins.

The ideas of the French School of Immunologists sponsored by Bordet
accounted for the interaction of antigen anc antibocy by the phenomenon of
adsorbtion. This concept, however, failod to explain the specificity of
immunological reactions. More recently the specificity of antibody pro-
duction has been explained inéependently by Mudd in this country and by
Breinl and Horowitz in Germuny. In their opinion, the normal course of
serum globulin synthesis is altered in tne animal boy when antigen reaches
the site at which the synthesis takes place, Under the influence of the
antigen, the serum glomlin ip altered in a way. characteristic for the
foreign stimlus. When the nodified globulin eventually encounters the
foreigh autigen either in the circulation, or in vitro, interaction of the
two is possible,

From the following interesting analogy - the results of certain
experiments carried out in our laboratories - it seems not unlikely that
the synthesizing function of a cell may be specifically oriented by a given
stinmlus. When the "R" form of Pncumococcus is grown in the presence of a
"snecific activator" derived from a cell-free filtrate of the "5S" organism,

the former is induced to resynthesize the capsular polysaccharide. If the

orn: SREY o .cgebeme arate
pereanlaern kaa 5 an Sone he TTT

Pe a ee ee

I nee ee

 

cre srt a aa

3

 

 

 

 

 

 
147

Hea

nature of the activator i: changed, however, the "R" organism can be made
to synthesize a polysaccharide quite different in its type-specificity
from that which it »roduced when stimlated by the original activator.
Thus it is scen that bactcrial cells, under the same general environmental
conditions may, by uicans of specific stimli, be directed to synthesize
products which, though chemically related, are biologically specific.

In regard to the non-protein nature of antibodies there are cven
today certain immnologists who maintain that solutions of antibodies may
°ve socured which give no chemical tests for proteins. It must be pointed
out, however, that the sonsitivity of immunological reactions far excoeds
the limit of sensitivity of chemical tests for proteins, Until weighable
quantities of non-protoin antibodies can be secured, the evidence for
this point of view remains inconclusive.

An analogous situation has oxisted in t..e attempts to define the
chemical nature of enzymes. It has boen the method of the Huropean
investigators to purify enzymes by adsorbtion on colloids. Solutions have
thus been secured which become incroasingly lower in nitrogen content -
without loss of enzymatic activity. This group of investigators has
failed, however, to secure by. these mothods protein-free enzymes,. nor have
they obtained ‘substances’ tho: chemical‘ composition and biological potoncy’ -
of which bear any’ constant relationship to one another, However, a new
approach.to the problem of enzyme chemistry has made: possible the prepara.
tion of crystalline proteins bearing the biologically active component of
crude enzyne nixturcs. These crystalline proteins show both constant
physical properties and biological activity on subsequent crystallization.

The success of Northrop and his associates in isolating from

nixtures of proteins a crystalline derivative of unique biological and

 

 

 
148

H 3

chemical properties has served as a stimulus to us in attempting to isolate
an immmne protein from antipneunococeus scrum. It is conceivable that this
protoin night have certain properties which woul¢ differentiate it from the
accompanying inert serum proteins. The nature of serun globulins, however,
is a problem of groat complexity, Fortunately tne investigations of the
Carlsberg laboratories have afforded an insight into the chemical nature of
globulin conplexos as they occur in normal horse serun.

Fron extensive studies on the proteins of normal horse serun,
Sorensen cane to the conclusion that the globulins occur not as nixturcs
but ae labile compounds of eu- and pseudo globulin. He found that these
conpounds are easily dissociable and that by means of fractional precipita-
tion and dialysis the constituents can bo partially separated as protein
fractions which become, with each subsequent fractionation, increasingly
richer either in eu-~ or pseudo ¢,lobulin, as the case may bee Although
globulins have never been obtained in a state of purity, it has been
possible to dissociate the cormlexes of serun globulins to such an extent
that fractions have been obtained of which oither cu- or psoudo globulin
is the predominant constituont. = = °° °° ee

.We have rade“uso of this princivle of the: dissociation of the serun
globulins in the separation of the antibodies in Type I Pneunococcus anti-
serum fron the accorpanying inert globulins with which the immune bodies
are associated. The water insoluble globulin fraction of antipnounococcus
scrun, described by Felton and commonly used in the therapeutic treatnent
of lobar pnewaonia, has boen separated into an inert evuglobulin fraction,
and a pseudo globulin fraction containing practically all of the anti-
bodies present in the original nixturo of the two serun proteins.

When finally dissociated and separated from inort euglobulin, the

 

 

 

 

 

 

 
; 149
HY

water soluble immnoe globulin is considerably more potent in antibody

content than the parent substance. Half a centimeter of a solution of the

   

purified antibody containing as little as 0.005 me. of protein is effective
in protecting mice against 500,000 minimal lethal doses of virulent
pneumococci of the homologous typo.

In adaition to increased biological potency, the inmunoe globulin has
certain chemical properties which distinguish it from normal serum globulin.
Determinations of the basic amino acid distribution of normal and immne
Globulin reveal no striking aifference in the chemical units which con-
stitute the two proteins. However, the immne globulin hae a very alkaline
isoelectric point (pH 7.6) whereas that of normal serum globulin lies in
the more acid range (pH 5.2). Thus the nost distinguishing property of the
{maine protein, its alkaline isoelectric point, is one which can be
accounted for only by a relative increase in the number of basic amino
groups in the protein molecule.

Efforts have boen made to crystallize the immune globulin, but
thus far without convincing results. Solubility measuremonts have
indicated that the antibody in its present state of purity ie not a chemical
entity, despito many ropeated fractional precipitations, It may be for
this reason that attempts to crystallize the antibody have thus far been
unsuccessful.

In view of its umsual basic properties it scems not unlikely that
the amino groups of the antibody protein may play some role in the union
of the antibody molecule with its type-specific haptene - the capsular
polysaccharide of Typo I Pneumococcus, In collaboration with Dr. Chow,
it was subsequently found that when the amino groups of the antibody were

acetylated with ketene vas, under conditions which climinated any

 

 

 

 
150
H 5

possibility of protein denaturation, the resultant acetylated antibody.
protein failed not only to precipitate the type-specific capsular poly-
saccharide, but to agglutinate suspensions of Type I pneumococci, and to
protect mice against infection with homologous virulent Type I pneumococci
as well,

The basic amino groups of the antihody protoin have also been
covered by the grouping = CHy, with a resulting loss in serological activity.
When the grouping = CHg is replaced by two hydrogen atoms, howover, the
omino groups are restored and the immnological proportics of the antibody
protcin are fully regained. The chemical reactions which are involved are
as follows:

(1) .- NH3 + HCHO _m6+»+» - N+ CHa + H20 (serologically inert)

(2) ~ N= CHa + H20 _» ~ NH, + HCHO (serologically active).
Reaction 1 goos to completion if an aqueous solution of pneumococcus anti-
body is treated with formaldehyde at 0°C. and at pH 8.5. The excess
formaldehyde is removed by dialysis, and the solution of the resulting
"formalized" protein is adjusted to pH 7,5. The protein derivative is
perfoctly stable under these conditions and shows no serological activity
whatsoever. If the aqueous solution of the formalized antibody is
adjusted to pH 5.0, however, reaction 2 tekes place rapidly. When the
reaction mixture is dialysed and the pH again adjusted to.7.5, the immno-
logical properties of the protein are fully restored. From these experi-
ments it appears that the presence of basic amino groups is of utmost
importance in dotermining the biological activity of the antibody protein.

Studies carried out in this laboratory on the chemical nature of
the type-specific polysaccharides derived from encapsulated microorganisms

have revealed the interesting fact that all the specific carbohydrates

 

 

 

 

 
Ap

thus far investigated contain uronic acids as constituents of the poly-
saccharide molecule. The invariable occurrence in these bacterial products
of glucuronic acid or its isomers sugzests that the highly polar carboxyl
group of the uronic acide and its stereochemical relationship to other groupe
in the molecule determines the biological specificity of these substances
and actually enters into chemical conubination with tho basic amino group
of the homologous antibody when these two substances are brought together
either in vivo or in vitro. Thus, if the carboxyl group of tho Type I
specific carbohydrate could be covered by an oster grouping, the derivative
ghould no longer react with its homologous antibody, The polysaccharide
was therefore treated with an ethereal solution of diazo methane, and the
methyl estor separated from the reaction mixture. Solutions of the methyl
ester of the polysaccharide gave no serological reaction when added to
homologous antibody. When the methyl groups were removed by gentle
alkaline hydrolysis the carbohydrate again reacted specifically with anti-
body in dilutions as high as 1 part in 4 million. Quantitative determina-
tions of the methoxyl content of the carbohydrate before and after
hydrolysis showed that the methoxyl derivative contained one methoxyl group
per carboxyl group. The reversible changes in immunological specificity
brought about by reversible alterations in chemical structure indicate
that at least two of the molecular groupings involvod in antigen-antibody
union are the clectro~positive amino group of the immune protein and the
electro-negative carboxyl group of the specific polysaccharide.

This concept alone does not explain the specificity of the reaction.
There is, however, certain evidence to support the view that the
specificity of such serological reactions is governed by the arrangement

in space of the polar groups of the reactive carbohydrate. It has been

 

 

 

 

 

 

 

 

 
‘6
452

shown that antigens prepared from the diazophenol élycosides of ¢:lucose
and galactose react only in their homolo;:ous antisera. In respect to the
number and nature of their polar groups, these tio ¢:lycosides are identical.
Since the reactive groups are in each instance the same, the nechaniem
underlying tho union of both glycosices with their homologous antibodies
mist likewise bo the samo. The specificity of this reaction therefore can
bo accounted for only by known differences in the spatial arrangencnt of
the polar groups of the fourth carbon atom of each glycoside.

Since the spatial arrangement of idontical polar groups in a carbo-
hydrate suffices to determine specificity, it therefore seens justifiable
to assune in the case of the antibody that the spatial arrangement of polar
Groups in inmmune protein nay likewise detcrnine its apecific capacity to
react with a given hapten,

In view of these considerations it is believed that the feneral
mechanism underlying the union of antibody and carbohydrate involves the
interaction of polar groups of opposite charge. In the case of the Pneuro-
coccus Type I hapten and its specific antibody, it is a certain amino group
of the latter, anc the carboxyl group of the polysaccharide which interact
to form tho immune precipitate, It is sug: ested, furthermore, that the
specificity of this reaction -is detemined »y the stereochenical relation-
ship of the doninant polar groups in the reacting molecules, whether they
be antigen or antibody, If the spatial pattern of the polar groups of
Doth antigen and anti»ody is of exactly the correct order, then union occurs.
If, however, this relationship is disturbed by artificial neans, as has been
experinentally cernonstrated by covering the dominant polar froup of either
antigen or antibody with a chemical radical, the pattern is destroyed, and

union between antigen and antibody fails to take place. When the original

constitution of the acting substances is restored, however, serological
epecificity is once more rogained.

 

 

 

 

 

 

 
45

Publications.

Goebel, Walther F, and Zabers, Frank H., Derivatives of glucuronic acid.
I. The preparation of glucuronic acid.from glucuron and a
comparison of their reducing values. J.Biol.Chem.,

1933, 100, 573.

Goebel, Walther F. and Sabcrs, Frank H., Derivatives of glucuronic acid.
II. The acetylation of flucuron. J.Biol.Chem., 1933,
100, 734,

Goebel, Walthcr F. and Sabers, Frank H., Derivatives of glucuronic acid.
III. The synthesis of diacetylchloroglucuron,
J.Biol.Chem., 1933, 101, 173.

Avery, Oswald T. and Goebel, Walther F., Chemo-immunological studies on
the soluble specific substance of Pneunococcus. Isolation
and properties of the acetyl polysaccharide of Pneuno-
coccus Type I, J.Exp.Med., 1933, 58, 731.

Goebel, Walther F, and Avery, Oswald T., Chemo-immunological studies on .
conjugated carbohydrate-protein antigens. VIII. Influence
o£ the acetyl group on specificity of hexoside-protein
antigens. J.EBxp.Med., 1934, -