The discovery of thyroid autoimmunity

Abstract

In 1951 there were not many places where a newly minted PhD could go to learn immunology. Michael Heidelberger was at Columbia working on the immunochemistry of polysaccharides and Ernest Witebsky in Buffalo was engaged mostly in studying human blood groups and isolating the blood group substances. Having just completed a PhD in Stuart Mudd's department at the University of Pennsylvania, I decided to make my future in immunology and was seeking a suitable training position. Most of my colleagues thought that 1 was quite mad, since it seemed that the major problems in immunology had all been solved. The instructive template theories accounted for the formation of antibodies, and immunochemical studies by Landsteiner taught us all there was to know about antibody specificity. A number of useful diagnostic tests and a few vaccines had been developed. However, the biological basis of the immune response was entirely obscure and intrigued me sufficiently to want to learn more about immunology. After some discussion, Witebsky was kind enough to offer me an instructorship in his department together with the opportunity to complete my medical studies. Thus my wife and I packed all our belongings into the back of an ancient Oldsmobile and journeyed from Philadelphia to Buffalo in September 1951. A few months after I settled at the University of Buffalo, Witebsky suggested that I look into the problem that had intrigued him since his own student days, that is organ specificity. He considered that normal tissues and cancers contained alcohol-soluble constituents (called 'lipoids') that are unique and immunologically specific. Isolating and characterizing these organ-specific constituents or antigens would provide greater understanding of tissue differentiation and malignant disease. Witebsky (together with his research assistant, Ann Heide) was working hard on the 'lipoids' in the alcoholic extracts of brain and suggested that I might want to look at thyroglobulin, the principal antigen of the thyroid gland. Thyroglobulin, he thought, was a unique example of an organ-specific protein. Witebsky had studied thyroglobulin during his tenure at Heidelberg and had found that it is both thyroidspecific and crossreactive among thyroids of other mammalian species. However, he feared that these crossreactions were artifacts due to the inevitable denaturation of thyroglobulin during preparation. He challenged me, therefore, to prepare a native thyroglobulin to determine conclusively organ specificity and crossreactivity. Working with a skilled protein chemist, Sidney Shulman, 1 developed a simple method of stepwise ammonium sulfate precipitation for the preparation of a relatively pure thyroglobulin product from bovine and other thyroid glands. I then made the appropriate antisera in rabbits and showed a considerable degree of organ specificity and crossreactivity. Ah, ha! said Witebsky, you must have denatured the material. Greatly deflated by this reaction on the part of my new boss, I sought a way of proving that the material was still in its native state. I hit upon the idea of preparing thyroglobulin from rabbit thyroid glands and injecting the material into rabbits. It seemed obvious that a native protein would not induce a response in the same species, whereas a denatured product might well do so. I made a batch of rabbit thyroglobulin, injected it intravenously into rabbits and was gratified to find that no immune response followed. As was so often the case, Witebsky and I discussed these results in the Serology Laboratory of the Buffalo General Hospital while he was reviewing the day's Wassermann tests. We began to speculate about why the rabbit was incapable of producing antibodies to its own tbyroglobulin. Nothing in the instructive theories of antibody formation could explain it, unless we assumed that a surplus of antigen in the bloodstream was causing a kind of immunological paralysis. I then suggested the perfect experiment. Why not remove the thyroids from some rabbits, since we could then be certain that all of the organ-specific thyroglobulin would be eliminated? We would then immunize the rabbit with rabbit thyroglobulin and produce a response. Because of Witebsky's long experience with blood groups, he pointed out the possible importance of individual differences and suggested that I immunize each rabbit with an extract of its own gland. The appropriate control was the extract prepared from the single lobe of a hemithyroidectomized rabbit. With the help of John Paine (Professor of Surgery) and Richard Egan (Assistant Professor of Surgery), I performed the necessary thyroidectomies and hemi-thyroidectomies, and made up individual thyroid extracts. Since the available antigen was in such short supply, I needed the most cost-effective