THE INFLUENCE OF MOLECULAR STRUCTURE ON BIOLOGICAL ACTIVITY

By Linus Pauling

Gates and Crellin Laboratories of Chemistry

California Institute of Technology

Pasadena, California

*Presidential Address, Pacific Division of American Association for the Advancement of Science, delivered at the 27th Annual Meeting, University of Nevada, Reno, June 19, 1946.

Our effort to understand nature consists of a search for structure, the spatial interrelations of the constituents of a material system, and for function, which is the mode of action of structure, or the change of structure with time. And we shall not have achieved an understanding of biological phenomena until we have obtained complete knowledge of the structure of organisms in terms of their fundamental constituents, atomic nuclei and electrons.

In the field of chemistry the search for structure has led to more and more precise knowledge of the way in which atoms interact with one another to form molecules and crystals. The qualitative concepts of valence bonds between atoms, structural formulas of molecules, the tetrahedral carbon atom, cis and trans isomers have been refined by a clarifying theory of chemical combination and by precise experimental information about interatomic distances and spatial configuration for hundreds of molecules and crystals. The detailed structural mechanisms of many chemical reactions have been discovered — we understand the process of the approach of a hydrogen molecule and an iodine molecule to form an activated complex with trapezoidal configuration, which may then dissociate into two molecules of hydrogen iodide; and we understand the process of the Walden inversion of the configuration of a carton atom, as a hydroxyl ion attacks the tetrahedral molecule at the center of one of its faces, and forms with the groups at the corners of this face a new tetrahedron to the center of which the carbon atom moves, as the fourth group becomes a free anion.

There has been developed during the last twenty-five years a good general understanding of the structure of atoms in terms of nuclei and electrons, and investigators in this field seem to agree that there remain no important general principles to be discovered. A great deal is known also about the electronic structure of molecules and crystals, but there is still much to be discovered: in particular, the empirical and theoretical knowledge about intermetallic compounds and other metallic systems is fragmentary, and a great amount of progress will have to be made before it can be claimed that an understanding of the structure of these systems has been attained.

The structure of atomic nuclei remains to be discovered. The constituents of nuclei are presumably protons and neutrons — but what they are doing in the nuclei, how they interact with one another, we do not know. The physicists are attacking this problem with such vigor that we may expect that within perhaps two decades the present empirical generalizations of limited validity will be replaced by a complete theory of nuclear structure. Moreover, progress in structural biology will not have to await the development of this theory, since there is no evidence that the properties of biological systems depend in any way on the detailed internal structures of nuclei.

The search for structure in biology — the structure of organisms in terms of cells, of cells in terms of membranes, cytoplasmic framework, protoplasm, chromosomes — has been thoroughly prosecuted throughout the dimensional region down to the limit of the resolving power of the visible microscope, about 10000 A., and during the last few years some structural information about the region between 10000 A. and 100 A. has been obtained by the use of the electron microscope. The structures of many simple molecules, of size up to about 10 A. in linear dimensions, have been accurately determined by the diffraction of x-rays and electrons. There remains the region between 10 A. and 100 A. This region must be thoroughly investigated in order that we may obtain complete information about the structure of living organisms: the basis for great progress in biology and medicine will be laid only by our obtaining a real understanding of biological function, and this understanding of function can be obtained only from detailed knowledge of the structure of the organisms in terms of the atoms which compose them.

What the experimental methods will be by which structure in the region between 10 A. and 100 A. will be determined cannot be predicted. The techniques of electron microscopy, which without doubt will be further refined, will probably continue to be most important in the study of structures with linear dimensions between 100 A. and 10000 A., and may well also be of great value below 100 A. The application of diffraction methods to determining the detailed atomic structure of very large molecules will he extremely difficult, hut these methods are very powerful, and the importance of the problems of biology and medicine will justify the great amount of work involved in their application to the complex materials of biological significance. Some help can be expected from the use of ultracentrifuges and other instruments, and the measurement of magnetic, and electrical properties; but let us hope that there will be discovered a new and powerful experimental method especially suited to the determination of structure in the 10 Å. to 100 Å. range.

A most immediate problem is that of the determination of the structure of proteins. These substances may well be described as the most important of all constituents of living matter, occurring in all cells, as framework material, constituents of protoplasm, enzymes, hormones, oxygen carriers, antibodies — and we have astoundingly little knowledge of their structure. It is known from the work of Fischer that the main structural feature of proteins is the polypeptide chain — but the sequence of amino acid residues in the polypeptide chain is not known for any protein, nor even for any good-sized fragment of a protein. The exact numbers of the different amino acid residues comprising the molecule are not known for any protein; the amino-acid analysis which approaches this goal most closely is that recently reported for p-lactoglobulin by Brand and his collaborators. The approximate general configuration of the extended polypeptide chain has been determined by Astbury by the x-ray study of B-keratin and other proteins; but the configuration of folded chains is not known even roughly for any protein. The precise configuration of all of the natural amino acids and of the simpler peptides could be determined (with the application of a good measure of hard work) by the x-ray techniques now at hand; but so far only four such structure determinations (of two amino acids, glycine and alanine, and two peptides, glycylglycine and diketopiperazine) have been reported, all by Corey and other workers in the Gates and Crellin laboratories. One of the first jobs which must be done in the attack on the general atomic structural problem in biology is the determination of the structure of all the natural amino acids, of many peptides, and of some proteins.

Even though we do not know the detailed atomic structures of proteins and other cell constituents, it has become clear that these structures are of the greatest importance in biological phenomena, and that biological activity is determined by the details of the structure of molecules, even to within 1 Å. or less, and by intermolecular interactions over distances of a few Angstroms, rather than by direct long-range forces. The evidence for this knowledge comes mainly from the specificity of biological phenomena — the substance thiamine does its job as a vitamin, and many organisms will accept no substitute for it; many other vitamins, hormones, enzymes, drugs show similar specificity — the bactericidal or bacteriostatic action of the sulfa drugs and of penicillin depends on precise details of chemical structure — and it has now become clear that this extraordinary specificity, which is far more pronounced than that shown by substances engaging in ordinary chemical reactions, is determined mainly by the exact sizes and shapes of the molecules and by the nature of the weak intermolecular interactions acting over very small distances.

The field of biology with which I am most familiar is serology, and it is this field which shows the phenomenon of specificity most strikingly. It was shown by Landsteiner and his collaborators that anti-sera have the power to distinguish not only between proteins so closely related that no other method differentiates between them, but also between similar simple haptenic groups of known structure, such as the m-azobenzoic acid group and thep-azobenzoic acid group. Landsteiner's evidence that serological specificity is determined by the detailed atomic structure of the haptenic groups has been supplemented by many experiments carried out in our laboratories in Pasadena during the past six years. We have found that antibodies detect the replacement of a single atom in a hapten by a chemically similar atom one-quarter of an Angstrom larger in radius, that the combining groups of antibodies have a surface configuration which reflects that of the haptenic group of the homologous antigen to within less than one Angstrom, that antibodies contain proton-donating or proton-accepting groups so situated as to form hydrogen bonds with complementary groups in the antigen, and that antibodies to an electrically charged haptenic group contain an electric charge of opposite sign which can be brought near to the charge of the haptenic group. A multitude of experiments on serological precipitation and its inhibition by haptens has made the conclusion inescapable that the specific attraction of antibody and antigen is due to forces acting at distances not of hundreds or tens of Angstroms but of one or two Angstroms, and that the specificity of the attraction is due to a detailed complementariness of structure on an atomic scale.

There is a strong presumption that in many other biological systems the specificity of interaction is due to the same detailed complementariness of structure. The malonate ion acts as an inhibitor of the enzyme succinic dehydrogenase, and many other enzyme systems are known in which substances closely related in atomic structure to the substrate molecules act as competitive inhibitors. Woods pointed out in 1$^Q that sulfanilamide exerts its bactericidal action by blocking the activity of the closely related essential substance p-aminobenzoic acid. The functioning of thiamine as a vitamin is inhibited by pyrithiamin, the corresponding substance with a pyridine ring in place of the thiazole ring; and many other examples of competitive biological activity of substances closely related in atomic structure are now known.

As our understanding of the structure of biological systems increases it will become possible to attack the problems of biology and medicine in a more and more straightforward and logical way. Thimann has pointed out that there is some similarity in structure between local anesthetics and acetylcholine, and there is also a structural relation between histamine and the anti-histamine substances investigated by Loew; but in neither case has our understanding of the phenomenon yet become complete enough to permit the confident prediction of more effective substances. The time will soon come, however, when this can be done; and when this time does come we shall enter upon a period of increasingly rapid progress in medicine as well as in biology.