Translating immunological tolerance into therapy


  • Berent J. Prakken,

    Corresponding author
    1. Department of Paediatric Immunology, University Medical Centre Utrecht, The Netherlands
    2. EUREKA Institute for Translational Medicine, Syracuse, Italy
    • Department of Paediatric Immunology / home mailbox KC.03.063.0, Lundlaan 3508 AB Utrecht, University Medical Centre Utrecht, The Netherlands, Fax: +31-30-2505350
    Search for more papers by this author
  • Salvatore Albani,

    1. EUREKA Institute for Translational Medicine, Syracuse, Italy
    2. Clinical Investigations Institute, University of California, San Diego
    Search for more papers by this author
  • Willem van Eden

    1. Department of Infectious Diseases and Immunology, Faculty of Veterinary Medicine, Utrecht, The Netherlands
    Search for more papers by this author


Thirty scientists from Japan, USA and Europe met from January 18th to 20th 2007 in a stormy Amsterdam for an expert meeting on translating immune tolerance into novel therapies. The meeting was held in the “Trippenhouse” a 16th century historical building, home of the Royal Netherlands Academy of Arts and Sciences, the host of the meeting.


juvenile idiopathic arthritis


rheumatoid arthritis


Experts from immunology, rheumatology and paediatrics joined together to discuss new insights into the maintenance of immune tolerance and how this relates to autoimmune diseases. The input of experienced clinicians mixed with the view of basic immunologists made the meeting lively, and clarified the challenges of translational medicine. The attendees discussed future steps towards the ultimate goal, treatment of human autoimmune diseases.

The challenge: Treating human autoimmune diseases

The clinical focus of the meeting was on diseases, such as juvenile idiopathic arthritis (JIA), rheumatoid arthritis (RA) and diabetes mellitus type 1, inflammatory bowel disease and multiple sclerosis. This medical perspective was provided by experienced clinicians (Wietse Kuis, Utrecht, The Netherlands; Patricia Woo, London, UK) who underlined the heavy burden of these diseases, not only for the patients and their families, but also for society as a whole.

The treatment of human autoimmune diseases has improved over the past decade due to the introduction of biological agents that intervene with specific cytokines or co-stimulatory molecules. The most successful to date are biological agents that target cytokines such as IL-1 (in auto-inflammatory diseases) and TNF-α (in RA, JIA and inflammatory bowel disease). So far none of the currently used therapies is able to re-establish self-tolerance and cure the disease. Still little is known of the immunological working mechanism of most drugs commonly used in the treatment of human autoimmune diseases. Only very recently it was determined that the blockade of the TNF-α pathway may act in part through restoration of Treg function in RA (Michael Ehrenstein, London, UK) 1, 2.

Good clinical and immunological definition and classification of patients is an important issue (Alberto Martini, Genova, Italy) as it improves the quality of mechanistic studies and may help to define new immune surrogate parameters. This can ultimately lead to more targeted, tailor-made therapies, as shown by the progress made in the field of autoinflammatory diseases. In these diseases, genetic defects predispose for an uncontrolled activation of the IL-1 pathway leading to severe clinical symptoms, such as fever, skin rash and organ dysfunction 3. The unravelling of the genetic and immunological basis of these autoinflammatory diseases, including the systemic form of JIA (Fabrizio de Benedetti, Rome, Italy), has led to a logical and effective treatment, namely through blockade of IL-1 with soluble IL-1 receptor (anakinra) [3]. Also, rare examples of self-remitting autoimmune diseases, most notably the self-limiting form of juvenile idiopathic arthritis, can provide clues for better targeted therapies when the immunological mechanisms of such disease remission are understood 4.

Defining immune tolerance

To define the field of immune tolerance, Irun Cohen (Rehovot, Israel) challenged the one dimensional concept of immune tolerance. He argued that a definition of tolerance is always dependent on the circumstances and the observer. The historical concept of tolerance, namely that the immune system needs to avoid responding to self antigens has to be redefined. Immunity against self-antigens is not only normal but probably also necessary according to the immunological homunculus, a concept Cohen already proposed in the early Nineties 5. He underlined this concept with new data with regard to the recognition of self-antigens before birth. Using an antigen array he showed that cord blood from neonates contains IgM and IgA that are specific for a defined set of self-antigens 6. This includes self-antigens that are also recognized later in life by IgG antibodies in the context of a well-defined autoimmune disease. Thus benign autoimmunity early in life may provide the basis for the development of autoimmune diseases later in life.

Almost any inflammatory experimental model can now be very effectively prevented and even treated with various immune therapies. Despite this, the follow up in human diseases is still disappointing. Luciano Adorini (Milan, Italy) gave an overview of the attempts made at modulating the adaptive immune response in human diseases. Specific interventions have been attempted with the use of (auto)antigens and small molecules, such as vitamin D3 7. Several of these approaches, both polyclonal or more antigen specific, and their implications for immune tolerance were discussed as were immune competent cells with therapeutic potential such as NK-T cells (Dirk Elewaut, Gent, Belgium), mesenchymal stem cells (Nico Wulffraat, Utrecht, the Netherlands) and T-regulatory cells (see below).

T-regulatory cells: The key to dominant tolerance?

The field of immunology has been shaken up by the re-discovery of Tregs 8. The importance of CD4+CD25+ Tregs cells was first discovered by Shimon Sakaguchi (Kyoto, Japan) 9. At the meeting, he argued that studies both in mice and in IPEX patients point to a “dominant” tolerance for self-antigens that is broken whenever Tregs are removed from the immune repertoire or rendered functionally incompetent. Although IPEX is a very rare condition, it shows how a single gene defect can cause a wide spectrum of autoimmune diseases, ranging from dermatitis to diabetes mellitus type I and thyroiditis. Dominant tolerance for self-antigens, mediated by Tregs is illustrated by two other observations. Firstly, in IPEX patients, diabetes mellitus type I can have an onset already in utero. This indicates that an intrinsic abnormality of the immune system by itself, without contributions from environmental factors, can cause autoimmunity. Secondly, in IPEX patients “diabetes-protective” HLA alleles such as HLA-DR2 and DR6 do not prevent the onset of type I diabetes. Thus, the immune system carries a large repertoire of self-specific cells that is kept in balance by a system of dominant immune tolerance, mediated by – among others – Tregs. Consequently, if the immune balance is disturbed leading to autoimmunity, the logical approach would be to try to restore this network of tolerance.

From concept to treatment: How to target T regulatory cells in vivo

By far the most discussed question for the meeting was whether targeting Treg cells can be a treatment for human autoimmune diseases. If so, should this be done polyclonally and in a non-antigen-specific or antigen-specific fashion, or perhaps a combination of both?

T regulatory cell therapy

In most autoimmune diseases the critical autoantigens are not known. This could argue for a non-antigen-specific approach to re-establish tolerance. How can T regulatory cells be induced in a non-specific fashion? The most straight forward Treg therapy is the infusion of ex vivo expanded autologous or allogeneic Tregs. Indeed, in stem cell transplantation, Treg cell therapy is now being explored for life-threatening complications such as graft – versus – host disease. Matthias Edinger (Regensburg, Germany) showed that it is possible to expand under GMP conditions human CD4+CD25+ Tregs in vitro that maintain their suppressive effect despite the expansion. In contrast to other reports, in Edinger’s experience only CD45RA+ naïve CD4+CD25high Treg cells gave rise to stable Treg cell lines upon in vitro culture 10.

Polyclonal activation of T regulatory cells

Thomas Hünig (Würzburg, Germany) proposed transient polyclonal activation in vivo of the naturally occurring Treg repertoire as an attractive alternative to the ex vivo expansion of those cells. In particular, the ability of a subset of natural Treg cells to home to inflamed tissues and to fight immune pathology on site could be important. In rodents, one highly effective way to transiently expand the Treg compartment and treat autoimmunity is the stimulation with a novel class of CD28-specific monoclonal antibodies termed “CD28 superagonists” 11, which bypass the need for TCR ligation in vitro. Tragically, the first-in-man trial of a human CD28-specific superagonist led to a life-threatening cytokine storm and multi-organ failure in the healthy volunteers receiving the drug 12. Thomas Hünig argued that based on previous toxicity testing in cynomolgus monkeys (where the human CD28 superagonist TGN1412 binds with identical affinity and shows the same Fc receptor binding properties as in humans), this tragic incident was unexpected 13. A subsequent investigation by the scientific expert group installed by the British government confirmed that the preclinical data submitted by the sponsor was reproducible and correct. Accordingly, new safety standards for first-in-man trials were suggested including a new mode to calculate the entry dose based on a “minimum anticipated biological effect level” (MABEL).

Transient polyclonal Treg activation may be possible independently of the use of CD28 stimulating agents, as shown by the success of anti-CD3 antibodies in diabetes type 1. Jean-François Bach (Paris, France) discussed the intriguing mode of action of these antibodies and the translation from the mouse to humans 14, 15. In NOD mice anti-CD3 reversed diabetes by restoring immune tolerance through the promotion of Tregs that function in a TGF-β-dependent manner 16. The CD3 antibody is non-depleting and it can preserve beta cell function and decrease insulin needs in newly diagnosed type 1 diabetic patients. These results pave the way for the use of CD3 antibodies as tolerance-promoting agents in various T cell-mediated autoimmune diseases and in transplantation (Jean-François Bach).

Antigen-specific targeting of Treg cells

Approaches, mostly preventive, aimed at antigen-specific tolerance induction are effective in controlling autoimmunity in experimental models. Steve Anderton (Edinburgh, UK) showed that peptide tolerance can effectively switch off the ongoing autoimmune response in EAE. He showed that Tregs work at several levels in EAE. In local lymph nodes, Treg cells control the risk of autoaggression as a result of molecular mimicry. Tregs also facilitate recovery in the target organ by accumulating in the CNS. Steve Anderton argued that if we are to develop the best antigen-based therapies, we need to understand what antigen Tregs recognize. HSP are among the candidate antigens, as they are immunodominant and expressed at sites of inflammation (Willem van Eden, Utrecht, The Netherlands) 17. Salvatore Albani (San Diego, USA) proposed that antigens such as HSP may act as “dimmers” of the immune response, and he gave the proof of principle that this can be the basis for immune therapy in patients with RA 18. In addition, António Coutinho (Oeiras, Portugal) pointed out that HSP might be examples of endogenous ligands for innate receptors, such as TLR4, challenging herewith the earlier hypothesis that only non-self recognition was determined by innate receptors. The presence of TLR on Treg cells would again indicate the contribution of innate receptors to self-nonself discrimination and maintenance of self-tolerance 19, 20.

Lost in translation?

Many are the hurdles that still impede translation from basic immunology to therapy of human diseases (see Table1). Often successful interventions in basic models are preventive, and cannot be directly translated into therapies. As the ongoing pre-existing autoreactive resp­onse can negatively affect the ability to induce Tregs, disease modifiers might be needed, such as anti-CD3 or anti-TNF-α 21 (Matthias von Herrath, La Jolla, USA). Also, suitable assays need to be developed that predict potential efficacy in vivo and can be used as biomarkers of therapy (Vicky Seiffert, Bethesda, USA; Johannes Roth, Munster, Germany). The Immune Tolerance Network (ITN, USA) has taken up this task for several assays and trials, but no such initiative has yet been started in Europe. Another major challenge is to define biomarkers for Tregs in humans. It is difficult to distinguish activated T cells from true Tregs, which is underscored by the recent reports that effector T cells can also transiently express FOXP3 22. During the meeting, in addition to the already described markers such as GITR, CD62L, CTLA-4, CD127(low), data were presented on Folate receptor 4 (Shimon Sakaguchi) and CD30 (Yvonne Vercoulen, Utrecht, The Netherlands) as markers on adaptive FOXP3+ Treg cells. Lastly, preclinical work (in vivo in animals or in vitro in humans) does not always predict what will happen in vivo in humans. Traditional pharmokinetics does not apply for tolerance induction, and dosages used are difficult to compare between animal models of disease and humans. This is especially the case in children, as age-specific details are very difficult to predict from data obtained in animals.

Table 1. Hurdles to be taken for translation, and possible solutions suggested at the Amsterdam meeting.
ChallengesPossible solutions suggested
How to target Treg cells in vivoPolyclonal, antigen-specific or a combination of both?
Which antigen can be used in human autoimmunityBystander antigens such as HSP
How to influence ongoing autoimmunityDisease modifiers (anti-CD3, anti-TNF-α)
Dose finding in mucosal therapyTest wide dose ranges, be aware of bell-shaped curves
Suitable markers for defining Tregs in humanCD62L, CD127, CD30
Biomarkers for immune toleranceDevelop “bioinformatics” for immune tolerance parameters
Age-specific differencesStudy and develop age specific immune parameters
Interaction between basic science and clinicsInternational consortia that combine efforts and different specialists through the Immune Tolerance Network (US) or Integrated Project/Networks of Excellence (European Union, Framework Programme 7)

The challenge from bench to bedside, and back again

‘Translation’ needs to go back and forth between model systems and patients and back again. The meeting underlined that this is the true challenge: bringing together basic scientists, biotechnologists and clinicians. Translation takes time and has costs. Traditional differences in culture, training and interests need to be set aside in order to bridge the gap. Therefore, a different training – one that combines the essential parts of those different worlds – is needed for young scientists. Such a training, proposed as a “Masters Program for Translational Medicine” (Salvatore Albani) was strongly supported by the attendees. To reach these goals, the EU could play an essential role, as the 7th Framework program offers the possibilities for setting up such far-reaching collaborations (Fatiha Sadallah, Brussels, Belgium). The meeting showed the challenges, but also the inspiration that emanates from translational medicine. Stepping out of the comfort zone of one’s own expertise opens up a whole new world of opportunities.


The authors would like to thank Wietse Kuis (UMC Utrecht), Nico Wulffraat (UMC Utrecht) and Lucy Wedderburn (UC London) for their intellectual support, and Martine Wagenaar (Royal Netherlands Academy of Arts and Sciences) for expert administrative assistance.

Meeting URL: _id=975