Structural Chemistry in Relation to

Biology and Medicine

Second Bicentennial Science Lecture of
The City College Chemistry Alumni Association

New York, December 7, 1949

Linus PAULING
The California Institute of Technology

HE progress made in medicine during the

past half century has been due in a large
part to the development of the science of chem-
istry. Much of the effective medical research
that has been carried out has involved chem-
istry in a significant way.

The years have seen a steady change in the
way in which medical discoveries are made.
Chemotherapeutic agents were at first discov-
ered by accident. Then there came a period
during which a general investigation of the
possible usefulness of chemical substances in
the treatment of disease was carried out. In re-
cent years great progress has been made by the
synthesis and testing of very large numbers of
new substances more or less closely related in
structure to a substance known to have useful
physiological activity; this method of attack
has been very valuable.

For example, quinine, like many other use-
ful drugs, was found to be useful in the treat-
ment of disease by the people living in the
region in which the plant containing this alka-
loid grows. Salvarsan was found by Ehrlich to
be valuable for the treatment of syphilis as a
result of the planned investigation of the activ-
ity of a great many chemical substances in the
treatment of this disease, in the simple hope
that one of them would be found to have
chemotherapeutic action. After the discovery
of the bacteriostatic activity of sulfanilamide, a
great many analogs of the substance were syn-
thesized, and some of these were found to be
stil] more valuable than sulfanilamide itself in
the control of infectious diseases. The action of
penicillin was discovered by accident; the
knowledge that an extremely valuable anti-
biotic agent is produced by a mould then led
at once to the vigorous investigation of the
substances produced by other moulds, and to
the discovery of many other effective antibio-
tics, including chloromycetin and aureomycin,

the first chemotherapeutic agents which are ef-
fective against rickettsial and viral diseases.

All of this work has been carried on in scien-
tific darkness — Or, at best, in a dim twilight.
Except for the idea that sulfa drugs operate by
inhibiting the capacity of the bacteria to
utilize an essential substance, p-aminobenzoic
acid, we do not understand the mechanism of
action of any chemotherapeutic substance. In
the absence of this understanding, research is
necessarily in large part haphazard, and pro-
gress is slow. We can be sure that if a detailed
understanding of the molecular basis of chemo-
therapeutic activity were to be obtained, the
advance of medicine would be greatly accel-
erated.

It is my belief that we are just on the verge
of taking this great step forward, that during
the next ten or twenty years very significant
progress will be made in the development of a
real science of pharmacology, and that there
will be great consequent progress in medical
practice.

In order to achieve this end we must learn
three things: the detailed molecular structure
of the chemotherapeutic substances, the de-
tailed molecular structure of at least some con-
stituents of the organisms against which they
are directed (the pacteria, rickettsial bodies,
viruses) and of the human organism, with
which the agents come in contact, and the na-
ture of the forces involved in the intermolec-
ular interactions between the substances and
the organisms.

During the past 25 years very significant
progress has been made in solving the first
problem. The chemists of the last century had
formulated the science of organic chemistry,
and had succeeded in finding the structural
formulas of many organic compounds. Since
1913 the X-ray diffraction method and other

methods of investigation of the structure of
baskerville chemical journal

molecules have made it possible to find out
exactly how atoms are attached to one another
in molecules such as those of the sulfa drugs
and penicillin. Chloromycetin, whose discov-
ery was announced only last year, has gone
through only the first stage of investigation —
its structural formula is known, and it has been
synthesized in the laboratory, but it has not so
far been subjected to a detailed molecular
structure examination, so that the exact shape
and size of the molecule have not yet been
determined. Aureomycin has not been put
through even the first stage; the methods of
organic chemistry seem not yet to have led to
the determination of its structural formula.
Nevertheless, it can be said that at the present
time this problem has in principle been solved
for the relatively simple chemotherapeutic
agents, containing less than 40 or 50 atoms in
the molecule.

The methods of molecular structure determi-
nation have sometimes shown the older struc-
tural formulas to be incorrect. An example is
provided by the arsenobenzenes, such as salvar-
san. The organic chemist has always written
the formula for these substances with a double
bond between two arsenic atoms, correspond-
ing to the double bond between the two nitro-
gen atoms in azobenzene; recent X-ray investi-
gation has shown, however, that the molecule
contains a ring of six arsenic atoms, each
bonded to two others by single bonds.

Progress in the solution of the problem of
the molecular structure of living organisms has
been slow, necessarily so because of the com-
plexity of the substances involved. Probably
the most important step that could be taken
would be the determination in detail of the
structure of some protein molecule. Professor
Robert B. Corey and many other workers in our
laboratories have been attacking this problem
during the past fifteen years. This attack has
included the investigation of the crystal struc-
ture of amino acids and simple peptides. The
results that have been obtained are self-con-
sistent, and make it possible to predict the
interatomic distances in the polypeptide chains
that are believed to be present in proteins with
confidence that the values are correct to ap-
proximately 0.02 A. In addition, the work on
the amino acids and simple peptides shows
clearly that the hydrogen bond is an extremely
important structural feature, and that hydro-
gen bond forces must constitute an important
part of the forces operating between segments

5

of a polypeptide chain in one protein molecule
and between adjacent molecules. It seems not
unlikely that during the next few years the
order of amino acid residues in the polypep-
tide chain of some simple protein will be de-
termined, and that in addition the detailed
configuration of the chain in the protein
molecule will be discovered,

It is likely that some drugs operate by under-
going chemical reaction with a constitutent of
the living organism, and that others operate by
the formation of complexes involving only
forces that are usually called intermolecular
forces. It is my opinion that, in general, the
intermolecular forces that are involved are
short-range forces, operating effectively through
distances of only a few Angstroms, and that, in
order to be effective, a chemotherapeutic agent
must come into atomic contact with a molecule
of the organism against which its activity is
directed. The evidence about the nature of
these forces that we have obtained in Pasadena
has been provided by studies of the interaction
of antibodies and simple chemical substances.
I became interested in the general problem of
the structure of antibodies and the nature of
serological reactions through conversations
with Dr. Karl Landsteiner, of the Rockefeller
Institute for Medical Research, during the
years 1936 to 1939. Experimental work in this
field has been carried on in our laboratories by
Professor Dan H. Campbell, Dr. David Press-
man, and many other investigators. It was
found that quantitative data about the inhib-
iting effect of simple haptens on the precipita-
tion of precipitating antigens and antibodies
with combining regions directed against simple
chemical groups could be interpreted by the
use of a theory of heterogeneous antibody, in
such a way as to give values of the average
equilibrium constant of combination of the
hapten with the antibody, and hence of the
standard free energy of combination of haptens
and antibodies. It was then found that the
values of the standard free energy of com-
bination could be interpreted in a rough
quantitative way in terms of the structure of
the hapten. The first requirement for the
formation of a good bond between the hap-
ten and the antibody is that the hapten
possess the proper size and shape. A devia-
tion in size or shape corresponding to a
change in dimension by as much as one
Angstrom, a quarter of the diameter of a single
atom, may lead to a great decrease in the power
6

of combination with the antibody. This effect
is attributed to steric hindrance — it is believed
that in the process of manufacture of the anti-
body a surface configuration of the antibody is
produced that is complementary to that of the
haptenic groups of the antigen, and that the
degree of complementariness is such that a fit
between hapten and antibody is not achieved
unless the haptens correspond closely in size
and shape to the original haptenic group. A
second factor of importance is the presence in
the hapten of a positive or negative charge cor-
responding to that in the original haptenic
group. Presumably there is a complementary
negative or positive charge built into the anti-
body. A determination of the magnitude of the
interaction energy between the haptenic
charge and the complementary charge of the
antibody led to the conclusion that the com-
plementary charge was within 2 Angstroms of
the minimum possible distance from the hap-
tenic charge. A third structural feature opera-
tive in the system is the hydrogen bond. If the
original haptenic group contains a hydrogen-
bond-forming group, haptens with a similar
group are found to combine more strongly
with the antibody, by about one or two kilo-
calories per mole in standard free energy, than
closely similar haptens in which the hydrogen-
bond-forming group is not present. The fourth
important type of intermolecular force is the
electronic van der Waals attraction that oper-
ates between all molecules, corresponding to a
potential energy inversely proportional to the
sixth power of the distance between atomic
centers. The magnitude of this interaction
energy is proportional to the product of the
polarizabilities of the interacting atoms, and it
has been found that the change in standard
free energy of combination with change in
polarizability of the haptens is in the predicted
direction, and has the magnitude correspond-
ing to contact between hapten and antibody,
with the atomic surfaces in juxtaposition to
within better than one Angstrom.

It seems to me likely that the interaction be-
tween haptens and antibodies can be taken as
the model for the interaction between mole-
cules of a chemotherapeutic agent and the
protein molecules of the organism that is af-
fected by the agent, at least in many cases. How-
ever, direct evidence for this hypothesis is lack-
ing, except in the case of the sulfa drugs, wherc
the competition with p-aminobenzoic acid, a
molecule very closely similar in structure to
those of the sulfa drugs, is indicated by the

baskerville chemical journal

structural similarity of the molecules to operate
through this mechanism.

An example of one way in which medical
research may be carried out in the future is
provided by some studies that have been car-
ried out in our laboratories during the past
four years on the disease sickle cell anemia.
Drs. Harvey A. Itano, S. J. Singer, Ibert C.
Wells, Walter A. Schroeder, and Robert B.
Corey have collaborated in this work. This
very serious disease, which afflicts about 1/400
of all American Negroes, is characterized by
the presence in their venous blood of a large
fraction of the red cells in abnormal shape.
The change in shape occurs in response to
change in the partial pressure of oxygen. When
the oxygen pressure is lowered, the cells change
from the normal biconcave disk to crescent,
holly wreath, and other forms. ‘This process is
known as sickling. It is prevented not only by
oxygenation of the hemoglobin, but also by
combination with carbon monoxide, to forn.
carbonmonoxyhemoglobin.

We have discovered * that the hemoglobin
present in the red cells of these individuals is
different from normal adult human hemo-
globin. This is the first time that any adult
human beings have been found to possess
hemoglobin differing from adult hemoglobin.

The difference between sickle cell anemia
hemoglobin and normal hemoglobin is evident
in a difference in the electrophoretic mobility
of the two substances. This difference corre-
sponds to a shift in isoelectric point by 0.23 pH
units, with the isoelectric point of sickle cell
anemia hemoglobin in the more basic region.
The difference corresponds to a difference in
electrical charge of about three electronic
charges per molecule.

The characterization of dimethyl esters of
porphyrins made from normal hemoglobin and
sickle cell anemia hemoglobin by determina-
tion of melting points and by their X-ray dif-
fraction patterns has given strong evidence
that the heme of sickle cell anemia hemoglobin
is identical with that of normal hemoglobin,
and that the difference in structure of the two
hemoglobin molecules is accordingly to be at-
tributed to a difference in the globins. Amino
acid analyses of hydrolysates of the two hemo-
globins are now under way for the purpose of
identifying the groups that are responsible for
the difference of charge and other properties
of the molecule.

 

* L, Pauling, H. A. Itano, S. J. Singer, and I. C. Wells,
Science, 70, 543 (1949).
baskerville chemical journal

We may picture the mechanism of the
sickling process in the following way. Let us
suppose that there is a surface region on the
globin of the sickle cell anemia hemoglobin
molecule that has a configuration comple-
mentary to that of a different region of the
hemoglobin molecule, and that this structural
feature is not possessed by normal hemoglobin.
The situation would then be somewhat ana-
logous to that which exists in antigen-antibody
reactions. Under appropriate conditions these
hemoglobin molecules might interact with one
another at these sites in such a way as to cause
at least a partial alignment of the molecules
within the cell, resulting in the erythrocytes
becoming birefringent, as is observed, and re-
sulting also in a distortion of the cell mem-
brane to accommodate the relatively rigid
structure within its confines. The addition of
oxygen or carbon monoxide to the cell might
reverse the effect by disrupting the bonds, pre-
sumably through a mechanism of steric hin-
drance. The pathological effects of the disease
are in general the results of the change in
shape of the cells, which causes the flow of
blood in the capillaries to be restricted, and
leads to anoxia in the affected tissues.

We can accordingly describe sickle cell
anemia as a molecular disease. Moreover, we
may feel some confidence that we possess a de-
tailed understanding of the nature of the
disease —a more penetrating understanding
than is possessed for any other disease. With
this knowledge, the task of finding a chemo-
therapeutic agent no longer seems without
hope; instead there appears the possibility of
planning out, on the drawing board of the

7

molecular architect, the specifications of a sim-
ple molecule that would have the power of
combining with sickle cell anemia hemoglobin
in such a way as to interfere with its pseudo-
crystallization, and thus to prevent the crisis of
the disease. If such an agent were to be found,
its discovery would represent the first time that
a chemotherapeutic agent had been developed
purely through the application of logical sci-
entific argument, without the significant inter-
position of the element of chance.

Progress in a field of science is often sig-
nificantly accelerated by the explicit prerecog-
nition of its possibility, and the provision of
suitable physical facilities for it. It seems to me
that the time is now ripe for the foundation of
a new institute of medical chemistry, in which
basic medical research would be carried out by
application of the most modern methods, and
in which also young doctors and young chem-
ists would receive training in the special fields
necessary to give them well-rounded prepara-
tion for the job of attacking medical problems
in this new way. I think that such an institute
should be set up in a university, in close juxta-
position to strong departments of the basic
sciences, rather than as part of a medical
school, or as an independent institute. Progress
in the field of structural medical chemistry, as
described in the preceding paragraphs, is sure
to be slow, because of the great difficulty of the
problems; but the possible importance of the
attack on this problem, with its promise for
greatly increased health and happiness for
humanity, is such as to justify our making a
great effort to achieve it.