Special regulatory T cell review: How I became a T suppressor/regulatory cell maven

Abstract

I have briefly reviewed the factors that motivated me to change my views about the existence and importance of suppressor/regulatory T cells and to devote the majority of my laboratory efforts to this newly revitalized area of immunologic research. I am optimistic that manipulation of regulatory T-cell function will shortly be applicable to the clinic.

On the occasion of the 50th anniversary of Immunology, I would like to describe the factors that motivated me to become a ‘maven’ in the T suppressor/regulatory field. First, I believe it would help to define the origin of the term, ‘maven.’ Maven is derived from the Hebrew word, mevin, which in turn is derived from the Hebrew word, binah, meaning understanding. A maven is a trusted expert in a field and is one of the first to pick up a new trend and then pass the knowledge on to others in the field. I believe my experiences in the T regulatory cell field over the past 12 years qualify me a maven.

I had always been intrigued by the model of autoimmunity induced by thymectomy of mice on the third day of life (d3Tx) originally described in the late 1960s by two Japanese investigators.¹ In this seminal study, the authors demonstrated that thymectomy of mice on the third day of life, but not on day 1 or day 7, resulted in the development of organ-specific autoimmunity. These authors postulated that autoreactive T cells are selectively exported from the thymus during the first 3 days of life and somewhat later in ontogeny, between days 4 and 7, regulatory T cells are exported from the thymus that control the pathogenic potential of these autoreactive cells. In the absence of a thymus, the autoreactive effectors would induce organ-specific autoimmune disease. One of the major findings in this early study was that transplantation of a thymus back into the d3Tx mouse on day 10 of life provided a source of T regulatory cells and prevented the development of autoimmunity. I also closely followed the experiments from Shimon Sakaguchi’s laboratory in the mid-1980s² that attempted to demonstrate that depletion of T regulatory cells from adult mice also resulted in the development of autoimmune disease. These studies claimed that CD5 or Ly-1^low cells from adult mice were capable of inducing autoimmune disease when transferred to immunoincompetent mice and that co-transfer of CD5^hi populations prevented the development of disease. The major problem with this experiment and probably the reason it was not widely accepted in the field at the time of its original publication is that more than 90% of T cells were CD5^hi and it was therefore impossible to further define the characteristics of the suppressor population.

I have spent my entire scientific career in the community of immunologists at the NIH. As I think back, this microenvironment was certainly not supportive of the large body of data generated by other immunologists claiming the existence of suppressor cells and their cellular products, the T suppressor factors of the 1970s. My own interests at the time were devoted to studies of the genetics of macrophage T lymphocyte interactions, immune response gene function, and major histocompatibility complex (MHC) class II molecules in the guinea pig. I must confess that I had difficulties remembering the differences between the various recombinant and congenic mouse strains that were the models for the dissection of the various pathways of T suppressor function. I also distinctly remember participating in one of the Immune Response Gene Workshops, originally started by Professor Baruj Benacerraf, and listening to the first descriptions of the now non-existent I–J locus of the mouse MHC. I breathed a sigh of relief that I did not have to master this complex area. I was content studying the much simpler model of the guinea pig MHC using the two extant inbred strains. Yet, none of the major groups of NIH immunologists including any of my close colleagues in the Laboratory of Immunology jumped into the T suppressor field. We listened politely, asked questions, but I think even then all of us had major doubts as to how this field would develop. I abandoned the guinea pig as an experimental model in the early 1980s and my switch to the mouse closely coincided with the downfall of the suppressor field in 1983. I think most of the NIH immunology community felt relieved when the Davis lab cloned the T-cell receptor genes³ and showed that they did not encode immunoglobulin idiotypes and when Kronenberg et al.⁴ failed to identify the I–J locus.

One of the great pleasures of being a scientist is having the opportunity to interact with outstanding, bright, and creative postdoctoral fellows. Adriana Bonomo certainly fit this description when she joined my lab in the early 1990s. During our initial discussions regarding potential projects, she and I agreed that the d3Tx model merited further study. Because the concept of T-cell mediated suppression did not even exist at that time, she and I developed an alternative hypothesis for the pathogenesis of autoimmunity following d3Tx. The details of this model are spelled out in great detail in a hypothesis paper we published in 1995.⁵ Put simply, we proposed that the peripheral T-cell pool of the d3Tx mouse is limited in size and is enriched in autoreactive T cells. The ‘empty space’ in the d3Tx mouse facilitates T-cell activation by promoting interactions between the few T cells in the periphery and the large excess of professional antigen-presenting cells. These activated self-reactive T cells would then circulate and home to non-lymphoid organs inducing organ-specific autoimmune disease. In euthymic animals, there was no need to postulate the existence of a dedicated lineage of suppressor T cells, as the continuous export of T cells would fill the periphery and act as a barrier in limiting the activation, recirculation, and homing of autoreactive T cells.

This model for post-d3Tx autoimmunity was certainly consistent with many of the experimental observations. Yet, it was difficult to test experimentally. Adriana established the d3Tx model in the lab and her outstanding surgical skills allowed us to also confirm that the transplantation of a thymus to the d3Tx mouse always prevented the development of disease. At that time, we also chose to study autoimmune gastritis (AIG), as it was the most prevalent disease that developed in BALB/c mice following d3Tx. Elisabeth Suri-Payer joined the lab at that time and her project was to fully characterize the fine specificity of CD4⁺ effector T cells that initiated AIG.⁶ Her identification of the gastric parietal cell H/K ATPase as the target for the CD4⁺ effectors provided us a major tool in dissecting the pathogenesis of organ-specific autoimmunity.

Shortly after publication of our ‘open space’ model, Shimon Sakaguchi first published his study on the use of the CD25 antigen as a marker for suppressor T cells.⁷ I read his paper shortly after it was published and was puzzled by the expression of CD25 on up to 10% of CD4⁺ T cells in a normal mouse. I am compulsive about methods used by others and noted that Shimon had used a monoclonal antibody to mouse CD25, 7D4, that had been generated in my lab⁸ in the early 1980s by Tom Malek who was then a postdoctoral fellow. When I went back and read Tom’s paper, I was surprised that we had shown that 8% of CD4⁺ T cells in normal spleen expressed CD25. We interpreted this finding as secondary to activation of the CD4⁺ T cells in vivo by environmental antigens in our non-specific pathogen-free animal facility and did not pursue the observation further. Nevertheless, I became more intrigued by the Sakaguchi paper and sensed that the model he proposed was indeed correct and our ‘open space’ hypothesis was unlikely to be viable. At that point in time, Angela Thornton, joined the lab as a postdoctoral fellow and I handed her the Sakaguchi paper and told her simply to repeat his studies. We had all the tools in hand, the anti-CD25 antibody, and the AIG model. She rapidly repeated all the findings made by Sakaguchi’s group and I rapidly became a convert to believing in suppressor T cells.

Every 4 years, the scientific programmes of the intramural laboratories in the NIAID are reviewed by an outside panel of referees. During the review of my programme in 1996, I received a negative review by a very senior immunologist of my proposal to begin the study of CD4⁺CD25⁺ T suppressor cells. I honestly felt like a schlemiel (a dolt) rather than a maven at that time. Nevertheless, I persisted in the characterization of the function of CD4⁺CD25⁺ T cells in both in vitro⁹ and in vivo¹⁰ models. The development of a simple, rapid, and reliable assay for the in vitro measurement of suppressor activity⁹ served as a firm foundation for all future studies in the field. Yet, skeptics remained, and it took 3 more years before the studies we had done in the mouse were repeated with human CD4⁺CD25⁺ T cells.¹¹

There is little doubt that in 2008 the study of suppressor/regulatory T cells represents one of the major areas of research in immunology and almost every published paper contains some reference or experiment to the involvement of these cells in every model system. I hope this brief discourse provides some insight to beginning investigators as to how one enters a new field and contributes to its development. My advice to a young investigator would be to keep an open mind and be ready to challenge prevailing dogma, but to always be critical. Changing direction in research is always difficult, but one advantage of the Intramural Program of the NIH is that one is not required to write grants and that monetary support for a programme is usually stable. It was certainly easier for me, in spite of the critical review I received, to pursue what I recognized as important area for future study.

One function of a maven is to evaluate the state of the art and to predict the future in his field of expertise. In my view, the T regulatory cell field is progressing at an incredibly rapid pace and the complete elucidation of the biochemical/molecular/cellular basis for T regulatory mediated suppression should be accomplished in the next 5 years. Both small molecules and biologics should then be available to manipulate the function of these cells in vivo. Transient diminution of T regulatory function will then be adopted as an ‘adjuvant’ for the administration of tumour vaccines and weak vaccines against infectious agents or for treatment of chronic infections. Conversely, transient or even continuous augmentation of T regulatory cell function will become part of the routine therapeutic management of autoimmune and inflammatory diseases.