THE AMERICAN NATURALIST Vou. LXXVIT May-June, 1943 No. 770 IMMUNOLOGY AS A TOOL IN BIOLOGICAL RESEARCH! IMMUNOCHEMICAL APPROACHES TO BIO- LOGICAL PROBLEMS DR. MICHAEL HEIDELBERGER ASSOCIATE PROFESSOR OF BIOCHEMISTRY, COLLEGE OF PHYSICIANS AND Surcrons, CoLUMBIA UNIVERSITY; CHEMIST TO PRESBYTERIAN HospPitan, New Yor« City I venture to address this gathering of geneticists and zoologists with an exhilaration engendered by a sense of the daring involved in an excursion into well-explored fields of knowledge remote from those into which my own work has extended. I trust, however, you will forgive this excursion or incursion, as it is intended more to re- mind you of progress already made in your fields along immunochemical lines, rather than to suggest the adop- tion of wholly foreign techniques and ideas. However, before reviewing these applications, it might be well to describe again in modern chemical terms some of the concepts fundamental to immunology. ; Knowledge of antigens, or the substances stimulating immune responses in animals, has been greatly extended in recent years. Thanks to chemical fractionations, the ultracentrifuge, the Tiselius electrophoresis apparatus and other powerful tools, one may no longer consider horse serum, for example, or an animal or bacterial cell, as ‘‘an antigen,’’ but must recognize it as a collection of antigens, 1 Four papers from a symposium scheduled to be presented by the Genetics Society of America at the annual meeting of the Ameriean Association for the Advancement of Science, which was cancelled at the request of the Office of Defense Transportation, December, 1942. 193 194 THE AMERICAN NATURALIST [Vou. LXXVII each with distinct properties and potencies. Many in- munological observations were and are difficult to inter- pret because this complexity was not taken into account. It is also apparent that many antigens are proteins and that most proteins are antigenic. Much work has been done showing that denaturation as well as introduction of the most varied chemical groupings at almost any point of substitution results in a definite change in immunologi- cal specificity. Now most of you will remember that some time ago, in Avery’s laboratory, we found that type specificity among the encapsulated bacteria depended upon another kind of antigen. This type specificity was due to a peculiar group of polysaccharides resistant to the usual sugar- splitting enzymes. The specific polysaccharide of each pneumococcus type, for example, was different from those of other types, and could be characterized by its distinc- tive physical and chemical properties. The sugars from types II and III pneumococcus were obtained free from nitrogen, and were the first instances in which immune specificity had been rigorously demonstrated in a class of substances other than proteins. With this brief discussion of specificity as a basis, what can be said about the requisite conditions for antigenic- ity? Itis obvious that we must have a complex structure and large molecules, and one of the important things seems to be the repetition of structural units. This is a highly probable consequence of the modern views of pro- tein structure. We also know that the specific carbo-- hydrate of type III pneumococcus, for instance, is made up of many cellobiuronic acid units. Some multiple of this unit must function as the immunologically reactive grouping, for when the carbohydrate is partially broken down by mild hydrolysis the fragments of two or more units still react in anti-pneumococcus type III horse serum. Therefore we may assume that in order to func- tion fully as an antigen a substance of large molecular size must be of such nature as to allow repetition of cer- No. 770] IMMUNOCHEMICAL APPROACH 195 tain structural units. Possibly for this reason ordinary lipids do not appear to have a clear-cut antigenic function. I think Dr. Landsteiner would add that any simple chemical substance may also function as an antigen espe- cially if the chemical properties are such as to allow its combination with protein to form new antigens. Com- plex structure is not necessary, therefore, if a nuinber of molecules of a smaller entity ean combine to form part of a larger structure. With regard to antibodies, the immune substances en- gendered in animals as a result of the antigenic stimulus, we are in a position to be equally definite. Use of new quantitative chemical microanalytical methods made it possible to measure antibodies in sera in actual weight units. One could, for the first time, express antibodies in terms of specific nitrogen per cubie centimeter of serum, because after precipitation with a slight excess of antigen non-specific material could be washed out. Since the amount of nitrogen in the added antigen is known this may be subtracted and the residual nitrogen in the washed precipitate is due to the antibodies. Highly purified anti- body solutions obtained as a consequence of information gained by these new methods were examined in the ultra- centrifuge and electrophoresis apparatus and were shown to have the properties of typical serum proteins. Buchner’s hypothesis that antibody contained frag- ments of antigen was proposed at a time when the actual nature of antibodies was not understood. This hypothe- sis never appealed to the chemist because in a number of instances like repels like, rather than attracts. In 1932 Breinl] and Haurowitz proposed a theory that antibodies are formed by a modification of the normal process of sertn globulin synthesis as a result of penetration of antigen or specific portions of the antigen to the site of globulin synthesis. The disturbance so brought about in- fluences the course of that synthesis in a sense character- istic of the antigen so that when the modified globulin ap- pears in the circulation and again encounters the antigen L9G THE AMERICAN NATURALIST — [Vou LXXVIT interaction igs possible. This not very clear picture was later expressed in somewhat more definite form by Mudd. An extension of this hypothesis has recently been made by Pauling which is even more graphic and reasonable but as devoid of experimental basis as the Breinl and Haurowitz theory. The Pauling hypothesis carried a second idea—that if one could take normal globulin, dena- ture it and fold it up again in the presence of antigen, artificial production of antibodies might be accomplished. Pauling now believes he has been successful in this, but such details of lis experiments as have been published do not inelnde complete controls. Burnet has recently pro- posed the origin of antibodies through modification by antigen of intracellular proteases which provide the framework for synthesis of partial replicas of themselves (globulins or antibodies). This would provide for anti- body formation after destruction of antigen and for pro- evessive changes in antibodies with successive immuniza- tions. These theories of antibody formation have been given a physiological basis in recent years by Dr. Florence Sabin as a result of experimental work with a red protein dye. Dy. Sabin has observed macrophages in the omen- tum and cells of the reticulo-endothelial system and found that, at a-certain stage of development, surface layers which form folds waving back and forth finally disap- peared as if they were being extruded from these cells. She believes this to be the source of serum globulins. and that the presence of an antigen (for example, the red pro- tein dye) results in the specific modification of these glo- bulins into the appropriate antibody. Now for a few applications to genetics and biology: Nuttall’s pioneer work on the mapping of biological re- latiouships through the study of the interaction of aninal sera with antibodies formed when these sera are injected into a standard animal such as the rabbit was most. fruit- ful and has been extended by numerous workers. The inununochemist has shown that interpretation of the com- No. 770] IMMUNOCHEMICAL APPROACH 197 plex findings is often simplified if a single protein is used, vather than serum, which we know to be a complex mix- ture of albumin, at least three globulins, complement with its four components, and other minor substances, all or most of which may function as antigens and cause over- lapping or zone effects in reactions with antisera. Nor is it certain that precipitation in different antisera is always due to the same antigen when such a mixture is used. An extreme instance of the simplification wrought by the use of pure, crystalline proteins was the demonstration by Landsteiner and myself of the non-identity or identity of the oxyhemoglobins of various species by a physical- chemical (solubility) method as well as by the serological technique. By use of the quantitative precipitin method, in which the amount of antibody nitrogen precipitated by a single purified antigen is measured, information as to species relationships may be gained that is unobtainable by qualitative measurements. In this way Stokinger and T were able to show the close relationship, but lack of identity, of sheep and bovine thyroglobulins, and _ to demoustrate that even this organ-specific globulin hor- mone possessed a species-specificity entirely independent of that of the corresponding serum globulins. With the sane quantitative method Treffers, Moore and I were able to give a plausible explanation for the differences shown by normal horse y-globulin and antipneumococcus horse y-globulin in rabbit antisera to the antibody (horse). Tt is not always necessary, however, nor is it neces- sarily au advantage, to study the immunological behavior of single antigens, as Irwin and his collaborators have shown in their intricate but clearly defined studies of the numerous gene-linked antigens of avian and mammalian red cells. Another fruitful inmunological approach to genetic problems has been made by Tyler in his studies of the agglutination of sperm by egg substances of Arbacia. The discovery of the specific polysaccharides of pneu- mococcus in Avery’s laboratory and the recognition that these sugar derivatives are the deternunants of type- 198 THE AMERICAN NATURALIST [Vou. UXXVIL specificity in this and other groups of pathogenic micro- organisms have led to far-reaching results, most of which are beyond the scope of this lecture. However, Griffith’s initially almost unbelievable discovery that one pneumo- coceus type could be converted into another has impor- tant iniplications, not only in carbohydrate chemistry and bacteriology, but in general biology and genetics as well. As many of you know, pneumococci of Type I, for example, may be degraded to a form devoid of type- specificity and then converted, theoretically, at least, back to the original type, or into any one of some forty-odd other types. This was originally accomplished by grow- ing the degraded cells in the presence of a heat-killed suspension of pneumococci of the type into which the liv- ing cells were to be converted. Studies by Avery, Daw- son and Alloway showed, however, that certain extracts of type-specific pnemnococci contained a substance or sub- stances responsible for this conversion, and that the type- specific carbohydrates themselves were not the determin- ing factors. Thus, any pneumococeus cell is potentially able either to synthesize, one at a time, nearly fifty differ- ent specific polysaccharides or may be so influenced by a series of substances that such varied syntheses become possible. The immunochemist nust leave it to the geneti- cist to decide whether or not these processes are true mutations, but I am happy to say that Avery is continuing the study of the transforming principle, and the eventual elucidation of its nature is certain to throw light on this and mnany other questions. These are merely a few of the instances in which im- munologieal and immunochemical nethods have provided an insight into biological mechanisms. To give a more complete summary would carry me far beyoud the allotted time, but I hope you will recall other examples which I would have liked to mention. More important, however, I hope that those of you who may have required this re- minder of the possibilities of these powerful tools will consider them as aids in the solution of present and future biological problems.