Development of the Insulin Radioimmunoassay, the Watershed Moment in Diabetes Research: Revisiting 1960 Diabetes Classics by Berson and Yalow

In the summer of 1960, Soloman A. Berson and Rosalyn S. Yalow (Dr. Yalow is depicted on this month’s cover in celebration of Women’s History Month) published three papers that described a novel method for measuring insulin in the plasma of humans and demonstrated the application of this new tool for the study of diabetes (1–3). These studies addressed, and started to resolve, important questions on the pathogenesis and treatment of diabetes, dramatically reshaping how the field would be researched and understood thereafter. Moreover, their novel method, the radioimmunoassay (RIA), would revolutionize the science of endocrinology specifically and biomedicine more broadly. Of these three papers, a publication in the July issue of the Journal of Clinical Investigation provided a detailed and comprehensive description of the RIA—a culmination of 5 years of intensive, logical, and careful research (1). The other papers, in the July and September issues of Diabetes (2,3), described the results of studies focused on discrete hypotheses tied to ongoing debates in the diabetes community. These two papers are submitted here as Classics in Diabetes and are arguably the papers with the greatest overall impact to appear in our journal.

When Drs. Berson and Yalow started their now-legendary collaboration in 1950, their initial goal was not to develop a specific and sensitive method for measuring hormones and other circulating compounds. They were part of a movement in American medicine to try to harness radioisotopes and radiation physics for diagnostic and therapeutic uses. Dr. Yalow, a young PhD physicist, was given charge of the Radioisotope Service at the Bronx Veterans Administration Hospital, and Dr. Berson, who had recently finished his medical residency, was hired to provide her a clinical associate (4). Their initial studies focused on the use of radioactive iodine (¹³¹I) to assess thyroid function, radioiodinated albumin to measure its kinetics in healthy and diseased states, and ¹³¹I-globin to determine blood volume. In 1956, they published a seminal paper (5) on the kinetics of radiolabeled insulin to test the hypothesis that diabetes might be the result of increased rates of insulin degradation (4,6). In fact, their results were quite the opposite and indicated that individuals who had previous insulin treatment of at least several months showed delayed clearance of the insulin tracer. Using chromatographic methods, Berson and Yalow demonstrated that in people who had been exposed to insulins derived from porcine or bovine sources, the insulin tracer migrated with the globulin fraction of serum proteins. Using complementary physicochemical techniques, they established that insulin-treated individuals made antibodies to beef-pork insulin that circulated in low titer but could be detected with a pure, high-specific-activity ligand. Moreover, using serum from individuals with diabetes, they determined that the binding of radiolabeled insulin by circulating antibodies could be competitively displaced by the addition of unlabeled insulin. With these observations, they had uncovered the critical components of the RIA (5,7).

Over the next several years, Berson and Yalow performed detailed studies of insulin-antibody interactions, cross-species differences in this reaction, and refined methods for generating ¹³¹I-insulin (8,9). A key step was the generation of specific antisera for use as the stage on which competition between labeled and unlabeled insulin could play out; serum from guinea pigs, immunized with crystalline beef insulin, “serve[d] admirably” (2). By 1959, they had a sufficiently developed assay on which to present early results at the American Diabetes Association meetings as well as in other settings (10). Looking back, the trail of publications, detailing the problem-solving, interpretive strides, and sheer volume of output, is remarkable and is an impeccable example of the scientific process. Over a very short period of time, these two investigators generated the underpinnings for how endocrinologic, and arguably immunologic (11), research would be conducted thereafter. Recognition of the importance of the work was immediate, with Berson receiving the Lilly Award from the American Diabetes Association in 1957 (7) and Yalow receiving it in 1961 (12).

In the June 1960 paper in Diabetes, Berson and Yalow compared insulin secretion during oral glucose tolerance tests in individuals with no diabetes and a group who had recently been diagnosed with “maturity-onset diabetes” (2). This paper includes some validation of the insulin RIA but refers readers to their other publications for most of the details and highlights primarily physiological outcomes. Here, they reported that while fasting insulin concentrations were comparable in people with and without diabetes (by current standards most of the cohort with diabetes would now be termed impaired glucose tolerant), these comparable concentrations were seen at slightly different levels of blood glucose. In response to a 100-g glucose load, both groups had significant elevations of plasma insulin, but the response was delayed in the group with diabetes—a pattern that has become the hallmark of early type 2 diabetes (Fig. 1, left). While exhibiting their usual caution and open-mindedness in interpreting results, Berson and Yalow raise the likelihood that type 2 diabetes is associated with decreased β-cell sensitivity to glucose and insulin resistance, suggestions that turned out to be prophetic. The second article (3), published in the September 1960 issue of Diabetes, took up the question of the mechanism by which sulfonylureas reduce blood glucose—a topic generating discordant opinions at the time (13). In groups of participants with and without mild diabetes, similar to those in their previous study, they observed that tolbutamide caused acute elevations of plasma insulin in both cohorts, although this elevation did not match the rise seen after a drink of oral glucose (Fig. 1, right). In this study, they also demonstrated the insulinotropic properties of glucagon, presaging the role of G-protein–coupled receptors in β-cell regulation. These two articles are brief, limited in scope, and include only small numbers of study participants. Nonetheless, they point to the new horizon in diabetes research, namely, that to study diseases that are fundamentally insulin deficiency states, the ability to make direct, accurate, and facile measurements of insulin is enormously powerful.

The concept of the RIA, the careful and nuanced descriptions of how to perform it (1), and the power of the technique quickly took hold and was widely promulgated after the initial publications. Berson and Yalow made assays for gastrin, growth hormone, ACTH, and parathyroid hormone, among others. Unger et al. (14,15) developed a widely applied assay for glucagon, with assistance from Berson and Yalow. Other laboratories developed their own insulin assays as well as RIA for a large number of peptide and other hormones; in the 1960s, development of a new RIA was a reliable stepping stone to a career in academic science. With RIA, the secretion of hormones, the active regulators of endocrine physiology, could be determined directly, providing new depth of understanding to a wide range of disease states; the importance of this to the field cannot be overstated. Moreover, the generation and directed use of antisera in the development of RIA added to the field of immunology, helping advance the application of antibodies to a range of methodologies, including immunocytochemistry, other immunometric assays, and immunoneutralization for loss-of-function experiments.

The importance of the work commented on here was ultimately honored with the awarding of the Nobel Prize for Physiology or Medicine to Dr. Yalow in 1977; she recognized her deceased research partner Berson in her acceptance speech (11). The Berson and Yalow collaboration was remarkably productive and suggests intellectual and tactical compatibility. Their work was based on clear thought and a high standard of experimentation but also included a bold and convincing translational component: their early studies applied their assay across a range of conditions with abnormal insulin secretion (1). That the story of the RIA, and the characterizations of diabetes physiology it allowed, played out in part in the pages of Diabetes remains a milestone for the journal.