Two Perspectives: On René Dubos, and on Antibiotic Actions

TWO PERSPECTIVES: ON RENÉ DUBOS, AND ON ANTIBIOTIC ACTIONS BERNARD D. DAVIS* Because René Dubos had an enormously beneficial influence on my own career, I particularly appreciate this opportunity to pay homage to him. In doing so I will pursue both suggestions in the invitation: to discuss my own experiences in working with him, but also to write under a broader title: Perspectives on Antibiotics. In pursing the first topic I will also emphasize the difference in research style between Dubos and another pioneer in antibiotic research, his teacher Selman Waksman. And in discussing antibiotic action I will focus on my own research, which has led to a general conclusion that I know would have interested Dubos: The search for a single key action, which has been taken foi granted in antibiotic research, can be misleading, as it has turned out to be for one major group, the aminoglycosides. In considering the qualities that made Dubos so influential, among scientists and later to a wider public, I will also compare critically his talents as a critic and as an experimenter. I am sure this approach is what he would expect of me. For during the year that I spent in his laboratory we argued incessantly and intensely about all kinds of problems , scientific and social. I was the naive, idealistic young American, seeking absolute truths and social perfection, while he was the worldly European, seeing subtle complexities in every problem. We were rather like Herr Settembrini and Herr Naphta in The Magic Mountain. First: how I got to Dubos's laboratory. During my four years as a medical student at Harvard I had done research part-time with E. J. Cohn, in one of the few laboratories in the world then devoted to protein chemistry. Hence when I served in the United States Public Health Service (USPHS) during World War II I was assigned to war projects under two outstanding immunologists, Elvin Kabat and Jules Freund. *Bacterial Physiology Unit, Harvard Medical School, Boston, Massachusetts 02115. Reprinted with permission from Launching the Antibiotic Era, The Rockefeller University Press, 1990. Perspectives in Biology andMedicine, 35, 1 ¦ Autumn 1991 | 37 At the end of the war I accepted the invitation from the newly formed Tuberculosis Control Division of the USPHS to set up a laboratory to explore whatever aspects of tuberculosis seemed interesting. In view of the subsequently expanded role of the USPHS in biomedical science it may be ofinterest to note Cohn's annoyance that I would remain in such an obscure research organization, rather than returning to a university! Immunology had not yet reached the stage where its complex phenomena could be translated into mechanisms, and I did not feel much attraction to the field. I decided instead to pursue microbiological aspects of the problem of tuberculosis—though in medical school I had also been bored by bacteriology, for the same reason. I now needed training in research in this field, and Dubos's laboratory looked to be the most exciting place. It did indeed prove to be very exciting. Dubos had just devised a simple medium in which the waxy tubercle bacillus grew in a somewhat dispersed manner rather than in the usual large clumps. The medium contained the non-ionic detergent Tween 80, and also bovine serum albumin. With my background in protein chemistry I naturally studied the function of the albumin. I had already found, while an intern at Johns Hopkins Hospital, that it binds sulfonamide drugs; and I now found that it bound traces of fatty acids, coming not only from residual soap on the glassware but also from slow hydrolysis of the Tween during the weeks of incubation; the fatty acids were strongly inhibitory to the tubercle bacillus [I]. This finding emphasized that a growth factor for bacteria might promote growth by providing protection, without serving as a nutrient. Another implication was that the accepted role of albumin in mammalian blood, in supporting colloid osmotic pressure, was only part of its function [2]. In fact, I also found traces of free fatty acids in human serum [3], which were evidently rendered nontoxic by the albumin. But I later realized that my publication of such an isolated...