Kendall A. Smith


The concept that interleukins mediate their effects on target cells by means of interacting with specific cellular receptors was first introduced with the discovery and characterization of the IL2 receptor (IL2R) in 1981[1]. Subsequently over the course of the 1980s the structure and function of the three IL2R chains were elucidated, and many additional interleukin receptors were discovered. Thereafter, the 1990s focused on experiments that discovered and characterized the intracellular signaling molecules and transcription factors activated by the interleukin-receptor interactions. These experiments transformed immunology from a descriptive science to one concerned with the molecular mechanisms of the control of the cells that mediate immune responses. The next decade promises to be focused on the discovery of the genes activated by the cytokine-receptor interaction at the cell surface, while the next century should see the evolution of this basic knowledge into a new understanding, which hopefully can be translated to new therapies for disorders of the immune system.


The Early Years

As mentioned in the description of the discovery of IL2, my idea that cytokines interacted with cytokine receptors germinated when I was in France, working with Torg Fredrickson on erythropoietin (EPO) in 1973. When confronted with the question as to how EPO mediated its effects, once we had developed the EPO bioassay using murine fetal liver cells, we performed experiments to try to adsorb EPO activity using increasing concentrations of EPO-responsive fetal liver cells. We found that there was a cell-concentration-, time- and temperature-dependent removal of EPO activity, thereby suggesting that there were EPO-specific receptors[2].

At the time, in the early 1970s, the only other receptors that had been described had been those for insulin, epidermal growth factor, and nerve growth factor. All of these ligands had been purified and characterized, and were known to be small proteins with molecular sizes of ~ 5-10,000 Da. The receptors for these ligands were first delineated by radiolabeled-ligand binding assays. However, this necessitated the availability of purified ligand, which could be radiolabeled. At the time of course, EPO was only an “activity” derived from the sera or urine of anemic individuals, and the molecular characteristics responsible for EPO activity had not yet been ascertained. Consequently, we had taken the EPO story as far as we could. It is noteworthy that at that time, protein purification was laborious and necessitated large amounts of starting material as well as a simple quantitative assay for the biological activity. About the only protein molecules that had been purified to homogeneity were enzymes that were purified from kilogram quantities of organs obtained from the slaughterhouses.


The Glucocorticoid Hormone Receptor Binding Assay

Soon after leaving France and returning to the United States and Dartmouth Medical School in 1974, I met Allan Munck, who was a Professor of Physiology at Dartmouth, and who had first described the radiolabeled glucocorticoid hormone-receptor binding assay in 1968[3]. Allan had been trained at MIT as an electrical engineer, and in endocrinology at Harvard, and had worked out the kinetics of binding of radiolabeled glucocorticoids to rat thymocytes. Since glucocorticoids were known to cause a rapid and marked involution of the thymus when administered to experimental animals, Allan had chosen the thymus as an experimental organ to identify glucocorticoid receptors.

Since Peter Nowell had shown that glucocorticoids markedly suppressed the proliferation of human lymphocytes in 1961[4], and since glucocorticoids were known to be immunosuppressive as well as potent anti-leukemia and lymphoma agents, I entered into a collaborative study with Allan to examine radiolabeled glucocorticoid binding in stimulated normal human lymphocytes and to try to correlate the number of receptors with the suppressive effects of the glucocorticoids. During the course of these studies, which we published in Nature in 1977[5], Allan taught me and Gerald (Gerry) Crabtree, who was a Pathology Resident at Dartmouth working with me, the principles and intricacies of radiolabeled ligand binding assays, both equilibrium binding and kinetic binding analyses. As well, he introduced us to the Scatchard Plot, which was first described by Scatchard in 1949 to graphically determine the number and affinity of receptor molecules[6].


The Radiolabeled Antibody Binding Assay

Subsequently, Gerry joined Allan’s lab as a postdoc and we collaborated on a project to determine the effects of glucocorticoids on the capacity of macrophages to bind radiolabeled antibody molecules. This project grew out of the well-known clinical observations that glucocorticoids suppress the uptake of antibody-coated RBCs in the spleen and is quite effective for the treatment of antibody-mediated anemias and thrombocytopenias.

We hypothesized that glucocorticoids mediate their effects by inhibiting the expression of antibody-receptors on macrophages. During the course of these studies, Gerry developed a binding assay to quantify the binding of radiolabeled-antibody to macrophages. Accordingly, the methods we used for this assay and those that Allan had developed for the glucocorticoid lymphocyte-binding assay became the for-runners of the radiolabeled-IL2 binding assay.


IL2 Adsorption Experiments


As soon as we had devised the bioassay for T cell Growth Factor (TCGF) in 1978, which was an adaptation of the EPO assay that Torg Fredrickson and I had developed previously in France, we performed TCGF adsorption experiments in the same way that we had done for EPO adsorption. We found that only mitogen-activated T cells adsorbed TCGF activity[7]. As well, just as with the EPO experiments, the removal of TCGF activity was cell concentration-, time- and temperature-dependent. From these experiments we could conclude that cells could remove TCGF activity, but we did not know whether this removal of a bioactivity resulted from its binding to a receptor, especially a cell surface receptor. Another plausible alternative possibility suggested by Allan included cellular metabolism of the molecule(s) responsible for the activity. Therefore, it was especially important when we found that formalin-fixed cells also adsorbed the TCGF bioactivity, because it suggested that the cells were not simply metabolizing the TCGF activity.

Beyond activated T cells, no other cells adsorbed TCGF activity. In particular, resting T cells, resting or activated B cells, as well as many leukemia cell lines were negative for TCGF binding. Therefore, there was appropriate target cell specificity to the TCGF adsorption: only the cells that could proliferate in response to TCGF also adsorbed TCGF activity.


Radiolabeled-IL2 Binding Experiments

To proceed beyond simple adsorption experiments it was clear that we had to purify the molecule(s) responsible for the TCGF activity. At the time, other investigators suggested that several molecules made up the TCGF activity in the lymphocyte-conditioned media. Therefore, we were especially hopeful that a single molecule was responsible when Richard Robb, a postdoc who joined the lab in 1979, demonstrated that a molecule that migrated as if it had only a single size and charge[8]. These findings strongly suggested that only a single molecule was most likely responsible for all of the TCGF activity. However, we still had to perform additional experiments to actually prove this idea.

Since it was difficult to generate enough purified TCGF that could be externally radiolabeled with 125I, we tried to produce the TCGF in medium containing radiolabeled amino acids. Thus, by biosynthetically radiolabeling the molecule, we also hoped that we could obtain data that would further demonstrate that a single molecule was responsible for the TCGF activity. Using biochemical techniques, Rich was able to show that a single radiolabeled molecule was bioactive in the TCGF bioassay. To construct the 1st radiolabeled cytokine-binding assay, we borrowed on our experiences with the radiolabeled glucocorticoid and immunoglobulin assays that we had developed and used previously. In particular, Gerry Crabtree had developed a method to separate the cell-bound radiolabeled antibody from free unbound radiolabeled antibody by using an oil layer, upon which the labeled cells were layered. Then upon centrifugation at 10,000-x g in a microfuge, one could obtain an instantaneous separation of the radiolabeled-antibody bound to cells, which were precipitated through the oil, from the unbound radiolabeled ligand in the supernatant, which remained above the oil layer. This assay proved readily adaptable for radiolabeled TCGF binding.

At this point, Allan Munck’s advice was crucial. In the design of our very first binding experiments, Allan advised us to perform the assay at 37 oC rather than at 4 oC, so that we could compare the concentrations that bound with those that promoted T cell proliferation in the TCGF bioassay. As well, he helped us to determine how best to show that the binding was saturable, using competition with unlabeled (“cold”) TCGF, thereby showing that the number of binding sites/cell were finite. Accordingly, in series of experiments we were able to show that radiolabeled TCGF binding satisfied all of the characteristics of true hormone receptors. Thus, radiolabeled TCGF bound to saturable binding sites with a high affinity (KD = 10-11 M), which was equivalent to the physiologic concentrations of the TCGF activity (EC50 = 10-11 M) that promoted T cell proliferation. As well, there was target cell specificity to the binding, in that only TCGF-responsive cells had detectable TCGF binding sites. Moreover there was ligand specificity to the binding, in that other cytokines and hormones could not compete for radiolabeled TCGF binding. Thus, the first cytokine receptor was discovered and characterized. In my search for the origins of the receptor concept, I found in my reading that Langely first introduced the concept of a receptor in 1878, by making the distinction between a binding site and a true receptor[9]. A receptor not only binds a ligand with high affinity and specificity, it also conveys a signal to the tissue that results in the recognized physiologic effect of the ligand.

As the data from the first binding experiments accumulated, it seemed evident by Scatchard analysis that there were at least 2 distinct binding sites, one with a high affinity (KD = 10-11 M), and another with a much lower affinity. Because the biological dose-response curve (i.e the recognized physiological effect of TCGF) coincided with the binding to the high affinity site, Allan favored concentrating on this site, and he suggested introducing a wash step after the binding reaction had come to steady state (10 minutes) using an excess of unlabeled TCGF to promote dissociation of the labeled TCGF from the lower affinity binding site. He felt that the lower affinity binding would be removed by this step, while the higher affinity binding would be maintained, which proved to be the case. In addition, it was obvious that the lower affinity binding sites were much more numerous than the high affinity sites, and as well, the affinity constant was > 100-fold lower, so that a much higher concentration of both hot and cold TCGF would be necessary to fully investigate this site. Since we had to produce and purify all of the TCGF, both the labeled and unlabeled molecules, we felt that we must wait to characterize this 2nd, lower affinity binding site.

All of these experiments required us to produce and purify both cold and hot TCGF from lymphocyte-conditioned media, which was a tedious and time-consuming task. To generate as much lymphocyte conditioned media as possible, we obtained tonsils from surgery with the help of Nat Gurkink, an ear nose & throat specialist at the Hitchcock Hospital. Both Margaret (Maggie) Favata and I, with the help of a high school student, minced the tonsils to make single cell suspensions so that we could culture many liters of TCGF-containing conditioned media, which Rich then concentrated and purified. Needless to say, the TCGF bioassay was crucial to detect and quantify the purified fractions.


The Structure of the IL2 Receptor and How it Binds IL2

In the early 1980s, as these binding experiments were being done, no one had identified the molecular structure of any hormone or cytokine receptors. The insulin receptor, EGF receptor and NGF receptor were perhaps the receptors that had received the most attention, but there were still no data available as to their molecular composition. About this time I had to apply for a competitive renewal to my grant that supported all of our studies. We had just completed our work showing that TCGF receptors could be detected using the radiolabeled TCGF binding assay, and I proposed to use this assay to screen for monoclonal antibodies that would block the binding activity. Of 10 specific aims that I proposed for the renewal application, the review committee liked 9 of them. The only one they didn’t think would work was the 10th, to look for TCGF-receptor antibodies. On this basis their “enthusiasm was dampened” and they gave me a priority score that was below the funding line. Therefore, I had to resubmit a revised version of the grant, which eventually was funded. Such are the vagaries of the so-called “peer review” process. Over the course of the next 25 years, each 5-year cycle that I reapplied for renewal of my very 1st grant to study “The Regulation of T cell Proliferation”, the grant review committee had questions and problems with the application, so that I had to answer their questions and resubmit the application for re-review each time. Of course these efforts compounded the time and energy necessary to keep the laboratory afloat and funded. However, I am proud to say that I kept my very 1st grant active for 25 years, and I just retired it in 2000.

Soon after we published our paper that described the discovery and characterization of TCGF receptors in 1981, I received a letter from three investigators at the NIH. The letter, co-signed by Warren Leonard, Warner Greene, and Tom Waldman, said that Takashi Uchiyama, a postdoc from their lab, had generated a monoclonal antibody that appeared to only react with activated T cells[10]. Moreover, it blocked T cell proliferation. They wondered whether it might react with the TCGF receptor that we had described, and they asked me to send some radiolabeled TCGF to them so that they could test this hypothesis. I called them, and suggested that I come down to the NIH and meet them, to discuss how best to do the experiments.

When I arrived at their labs, we had a brief meeting, which I found to be a little strange. Although Tom was clearly the senior scientist, both Warren and Warner seemed to be competing with Tom for the designation as to whose project or experiments these were to become. In retrospect, this competition amongst these three individuals accounted for the co-signatures to the letter that they had sent to me. Anyway, at the meeting I suggested that instead of me supplying radiolabeled TCGF to them and teaching them how to do the binding assay, it would work better if they gave some of their antibody to me and I would see whether it competed for radiolabeled TCGF binding, in the same way the unlabeled TCGF competed for the binding. After much discussion, Tom Waldman disappeared for a time and reappeared with a Styrofoam box containing the antibody, thereby demonstrating who was really responsible for the antibody that they had named Tac, for “activated T cell”.

The very 1st experiment that I performed using anti-Tac was absolutely definitive. Anti-Tac competed for radiolabeled TCGF binding and the concentrations that inhibited TCGF binding were identical to those concentrations that inhibited TCGF-dependent T cell proliferation. So much for the wisdom of NIH grant review committees.

The anti-Tac then was used by the NIH group to precipitate a molecule from the cell surface that had a molecular size of ~ 55 kDa. Thus, the 1st structural information indicated that the TCGF receptor was a single chain[11].

Subsequently, Warren Leonard and Warner Greene left Tom Waldman’s lab, and formed their own research groups. Then, independently all 4 groups reported the discovery of the second TCGF receptor chain[12-15], which had a molecular size of ~ 75 kDa, and became known as the β-chain, while the p55 chain was named the α-chain. Then, 6-7 years later, Kazuo Sugamura’s group from Sendai, Japan reported the discovery of the 3rd chain of the IL2 receptor[16], which is now known as the γ-chain, or “common γ-chain” (γc), in that it is also used by several other interleukins as one of their receptor chains[17].

While these events were evolving, Huey-Mei Wang, a graduate student in the lab, took on the project to try to unravel how each receptor chain contributed to the very high affinity of IL2 binding[18]. In a series of delicate experiments, she was able to show that the -chain binds IL2 very rapidly, with an association rate constant of 107 M-1s-1, while the chain appeared to contribute by a relatively slow dissociation rate (k’ = 10-4 s-1). Since the equilibrium dissociation constant = k’/k = 10-11 M, the kinetic binding experimental constants confirmed the equilibrium binding experiments.

Subsequently, Tom Ciardelli’s group at Dartmouth performed a series of binding experiments using purified recombinant receptor chain proteins and Surface Plasmon Resonance. They confirmed and extended Huey-Mei’s results, and showed how the -chain contributed to the slow ‘off-rate’ by cooperating with the -chain[19].

The complete 3-dimensional structure of the trimeric IL2 receptor (IL2R) has just now been solved by two groups[20, 21], so that we now have a complete picture of how IL2 binds to each of the chains and how the very high affinity of the IL2R is created (see “Find Out” section of this web site).


IL2 Signaling of T cell Proliferation

By 1982 we had available all of the reagents necessary to begin to investigate exactly how IL2 binding to its receptor triggers T cells to proliferate. These reagents included homogeneous, purified IL2, radiolabeled IL2, monoclonal antibodies reactive with IL2 and its receptor, and cloned IL2-responsive T cells. Accordingly, the key to these reagents was that we knew them to be absolutely pure and individual molecules. Consequently, we could detect and quantify molecular events that promoted changes in individual cells for the 1st time.

Doreen Cantrell, from Nottingham England, began a postdoctoral fellowship in the lab just at this time, and Doreen brought to the lab an expertise in the use of the flow cytometer, an instrument introduced to immunology by Len Herzenberg and his colleagues at Stanford a few years earlier[22]. The flow cytometer uses laser technology to detect and analyze individual cells amongst populations of thousands and millions of cells. Thus, we had the molecular reagents and the instrument to use them uniquely that enabled us for the 1st time to perform experiments that had been impossible previously.

Over the previous 50 years experiments had been conducted with various cell populations by many investigators, who attempted to determine the variables responsible for the control of cellular division. The 1st experiments were reported in 1932 in studies of bacterial cell growth, and subsequently, all living cells, including yeasts, unicellular parasites, as well as cells from all invertebrate and vertebrate species were observed to proliferate in exactly the same manner[23]. The cell population would proliferate with the same average time form one division to the next for many generations, apparently indefinitely. However, individual cells within a population would divide with highly variable rates, some very rapidly, while others were much slower. Even so, the most rapidly dividing cells did not overgrow the slower ones, so that the time interval from one division cycle to the next did not appear to be determined genetically, passed on from one generation to the next. Instead, the cell cycle time of individual cells seemed to be determined mysteriously by some unknown forces at each cell generation.

Various theories were proposed to explain the extreme variability of the cell cycle times of individual cells within a given population, and eventually these theories coalesced into 2 opposing views: The Deterministic Theory, which stated that the differences of cell cycle times between individual cells comprising a cell population (even a cloned, genetically identical cell population), were attributable to many small differences between the cells, so many differences that they could not easily be determined and accounted for at a given point in time[24]. This theory predicted that it was going to be very difficult to determine all of the variables involved. By comparison, the Probabilistic Theory stated that the differences in cell cycle times between individual cells within a cell population were determined stochastically, so that they were due simply to chance[25]. Accordingly, this theory predicted that it would be impossible ever to fully account for the variables involved.

In experiments that Doreen performed over the next 2 years, we gained evidence that neither of these theories was correct. Instead, the variables determining cell cycle times were definite, and could be readily identified, provided the molecules known to be responsible for T cell proliferation were quantified carefully and at the individual cell level. Thus, T cell cycle progression, DNA replication and mitosis could be explained by only 3 variables: 1) the IL2 concentration available, 2) the IL2 receptor density on the cell surface, and 3) the duration that these 2 molecules had to interact with one another[26].

Accordingly, if the IL2 concentration was limiting, the time required to traverse the cell cycle was longer than if the IL2 concentration was high enough to be receptor saturating. Also, the cells in a population with a low density of surface IL2 receptors would require a longer time to traverse the cell cycle than cells with a higher IL2 receptor density. Finally, there was a finite time interval necessary for the IL2-IL2 receptor interaction, which had to be surpassed before any of the cells within the population would make the critical decision to replicate DNA and divide. In other words, at the level of the single cell, cell cycle progression is a quantal (i.e. all-or-none) process. Previous investigators had had to deal with cell populations instead of individual cells, and they had not first identified the critical molecules responsible for determining cell cycle progression for the cell populations they were studying.

These experiments established that T cell proliferation is dependent upon only 2 critical molecules that determine cell cycle progression at the cell surface, and that there is no rate-limiting biochemical step on the inside of the cells that comes into play. Therefore, these findings were extremely important for the concepts of how normal cell growth and division is regulated, and for our understanding of abnormalities of the regulatory system that might lead to uncontrolled cell growth that could manifest itself as cancer. They also provided the impetus to discover the molecules on the inside of the cell, triggered by the IL2-IL2 receptor interaction at the cell surface that conveyed the message to the nucleus, so that there would be a coordinated duplication of DNA, and a subsequent organized mitotic process culminating in cellular division.



The discovery that IL2 promotes its growth stimulating effects on T cells by means of binding to classic cell surface specific high affinity receptors provided for a new mechanistic framework for understanding how the immune system operates. Prior to this discovery, the immune system had been considered as distinct from other physiologic organ systems, such as the cardiovascular system or the endocrine system, disciplines already well known to be regulated by neurotransmitters and hormones. In contrast, the immune system was viewed as regulated only from without via the invasion by environmental molecules that were recognized as foreign and that then drove the immune response.

With the advent of the IL2-receptor concept, it became evident that the immune system is regulated in much the same way as the other organ systems, by a set of unique hormones and receptors, which had never been described and characterized before. Thus, like the central nervous system, which also recognizes signals from the environment, the immune system directs its response to the invasion by external molecules via the secretion of interleukins, which orchestrate the tempo, magnitude and duration of the immune response by interacting with specific receptors.




1. Robb RJ, Munck A, Smith KA: T cell growth factor receptors. Quantitation, specificity, and biological relevance. J Exp Med 1981, 154:1455-1474.

2. Fredrickson TN, Smith KA, Cornell CJ, Jasmin C, McIntyre OR: The interaction of erythropoietin with fetal liver cells I. Measurement of proliferation by tritiated thymidine incorporation. Exp Hematol 1977, 5:254-265.

3. Munck A, Brinck-Johnson T: Specific and nonspecific physicochemical interactions of glucocorticoids and related steroids with rat thymus cells in vitro. J Biol Chem 1968, 243:5556-5565.

4. Nowell PC: Inhibition of human leukocyte mitosis by prednisolone in vitro. Cancer Research 1961, 21:1518-1523.

5. Smith KA, Crabtree GR, Kennedy SJ, Munck AU: Glucocorticoid receptors and glucocorticoid sensitivity of mitogen stimulated and unstimulated human lymphocytes. Nature 1977, 267:523-526.

6. Scatchard G: The attractions of proteins for small molecules and ions. Ann NY Acad Sci 1949, 51:660-673.

7. Smith KA: T-cell growth factor. Immunol Rev 1980, 51:337-357.

8. Robb RJ, Smith KA: Heterogeneity of human T-cell growth factor(s) due to variable glycosylation. Mol Immunol 1981, 18:1087-1094.

9. Langley J: On the physiology of the salivary secretion. J Physiol (Lond) 1878, 1:339-369.

10. Uchiyama T, Broder S, Waldmann TA: A monoclonal antibody (anti-Tac) reactive with activated and functionally mature human T cells. I. Production of anti-Tac monoclonal antibody and distribution of Tac (+) cells. Journal of Immunology 1981, 126:1393-1397.

11. Leonard WJ, Depper JM, Uchiyama T, Smith KA, Waldmann TA, Greene WC: A monoclonal antibody that appears to recognize the receptor for human T-cell growth factor; partial characterization of the receptor. Nature 1982, 300:267-269.

12. Sharon M, Klausner RD, Cullen BR, Chizzonite R, Leonard WJ: Novel interleukin-2 receptor subunit detected by cross-linking under high affinity conditions. Science 1986, 234:859-863.

13. Teshigawara K, Wang HM, Kato K, Smith KA: Interleukin 2 high-affinity receptor expression requires two distinct binding proteins. J Exp Med 1987, 165:223-238.

14. Tsudo M, Kozak RW, Goldman CK, Waldmann TA: Demonstration of a non-Tac peptide that binds interleukin 2: A potential participant in a multichain interleukin 2 receptor complex. Proceedings of the National Academy of Sciences USA 1986, 83:9694-9698.

15. Dukovich M, Wano Y, thi Rich Thuy L, Katz P, Cullen BR, Kehrl JH, Greene WC: A second human interleukin-2 binding protein that may be a component of high-affinity interleukin-2 receptors. Nature 1987, 327:518-522.

16. Takeshita T, Ohtani K, Asao H, Kumaki S, Nakamura M, Sugamura K: An associated molecule, p64, with IL-2 receptor beta chain. Its possible involvement in the formation of the functional intermediate- affinity IL-2 receptor complex. J Immunol 1992, 148:2154-2158.

17. Leonard WJ, Shores EW, Love PE: Role of the common cytokine receptor gamma chain in signaling and lymphoid development. Immunol Rev 1995, 148:97-114.

18. Wang HM, Smith KA: The interleukin 2 receptor. Functional consequences of its bimolecular structure. J Exp Med 1987, 166:1055-1069.

19. Liparoto SF, Myszka DG, Wu Z, Goldstein B, Laue TM, Ciardelli TL: Analysis of the Role of the Interleukin-2 Receptor Gamma Chain in Ligand Binding. Biochemistry 2002, 41:2543-2551.

20. Wang X, Rickert M, Garcia KC: Structure of the Quaternary Complex of Interleukin-2 with Its {alpha}, {beta}, and {gamma}c Receptors. Science 2005, 310:1159-1163.

21. Stauber D, Debler E, Horton P, Smith K, Wilson I: Crystal structure of the interleukin-2 signaling complex: paradigm for a heterotrimeric cytokine receptor. Proc Natl Acad Sci U S A 2005, In press.

22. Parks DR, Bryan VM, Oi VT, Herzenberg LA: Antigen-Specific Identification and Cloning of Hybridomas with a Fluorescence-Activated Cell Sorter. Proc Natl Acad Sci U S A 1979, 76:1962-1966.

23. Pardee A, Shilo B, Koch A: Variability of the cell cycle. In Hormones and Cell Culture. Edited by Sato G, Ross R. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory; 1979: 373-392

24. Koch A, Schaecter M: A model for statistics of the cell division process. J Gen Microbiol 1962, 29:435-444.

25. Burns V, Tannock I: On the existance of a G0 phase in the cell cycle. Cell Tissue Kinetics 1970, 3:321-333.

26. Cantrell DA, Smith KA: The interleukin-2 T-cell system: a new cell growth model. Science 1984, 224:1312-1316.