“Type 1 diabetes is particularly tragic because it usually starts in childhood and its effects worsen with time. ”
Matthias von Herrath, M.D.
Autoimmune Research, Diabetes
detailed lab report

cell-bullet2.jpg The main theme of my research program is to understand the regulation of autoimmune and anti-viral responses. Our work has shown that the amount of immunopathology or tissue injury is determined not only by the magnitude of a localized or systemic immune process, but also to a large extent by its components or the class(es) of responses it encompasses. Thus, each immune or auto-immune reaction has at least a more aggressive and a more regulatory component that balance each other and in this way have a strong effect on the duration or magnitude of the response and resulting tissue injury. In autoimmune diseases, it is possible to take therapeutic advantage of this paradigm and generate autoreactive regulatory cells (Tregs) by targeted immunization with self-antigens. We have shown that such cells can be induced by mucosal immunization and DNA vaccination. Antigen-induced autoreactive Tregs are able to site-specifically suppress ongoing autoimmune reactions, because they are preferentially retained in the draining lymph node closest to the target organ where they exert their regulatory function. Intriguingly, their induction can be successful in recent-onset diabetes when insulin immunization is combined with a short-term course of non-FC-binding anti-CD3.

Ongoing and future studies are focused on:

Honeymoon project: The lab is continuing a small human study evaluating T-cell responses during the partial remission phase of type 1 diabetes. Preliminary results indicate that this "honeymoon" phase is not associated with classical Treg markers (IL-10 and FoxP3) and that these two markers predict future glucose control in very different ways.

NOD-SCID IL-2R knockout mouse project: The lab is beginning studies to evaluate reconstituting this mouse with human peripheral white blood cells to better allow us to study the human immune system.

The role of key cytokines (TGF-beta, IL-4/IL-5 and IL-10), transcription factors (T-bet) and chemokines (IP-10) in regulating autoimmunity and virally-induced immunopathology. Pertinent issues are modulation of antigen presenting cells, in vivo trafficking of Tregs and their in vitro propagation and antigen specificities (Juedes J. Exp.Med. 2004, Homann Immunity 1999, 2002 and Bot JI 2001).

Dissecting how autoantigen-induced Tregs act in immune-competent hosts in vivo and how they can be used in the clinic to prevent diabetes. Combination of antigen-induced Tregs with systemic therapies such as anti-CD3 that favor a ‘Treg-friendly’ environment will be further developed (see von Herrath JI 2002 and Nature reviews from 2003 and 2004).

The Role of viral immunopathology in autoimmunity. Viral infections can enhance and abrogate autoimmune processes and we will continue to evaluate mechanistically how this occurs. Furthermore, persistent infection systems are being investigated to define approaches that ameliorate immunopathology and facilitate vaccine development (Christen JCI 2-2004 and 11-2004, Christen, Diabetes 2004, Bot J. Virol 2003).

Combination therapy of recent-onset diabetes with insulin immunization and anti-CD3. This new strategy that we pioneered is right now under revision at JCI. Since the murine experiments look very promising in showing strong synergy and life-long reversion after recent-onset diabetes in two animal models (NOD and RIP-LCMV), this strategy will likely be used in clinical trials in the US and Australia starting next year.

Paradigms developed from these studies will not only be useful in selectively suppressing autoimmune diseases, but can also be employed to lower immunopathology that accompanies viral infections.

Generation and characterization of insulin peptide-specific regulatory T cells

M. M. Martinic, J. Oldham, L. M. Togher, C. M. Filippi, D. Bresson, G. Fousteri, J. M. Jasinski, G. S. Eisenbarth and M. G. von Herrath.

Our goal is to generate insulin peptide-specific regulatory T cells (Tregs) in vitro, which suppress overt diabetes in prediabetic and reverse already ongoing disease in diabetic mice. Further we aim to analyze the suppressive properties and mechanisms of these Tregs in vitro and in vivo.

In order to generate insulin peptide-specific Tregs, we took advantage of insTCR transgenic mice, which express T cell receptor (TCR) transgenic CD4+ T cells specific for the insulinB9-23 peptide (insB) presented on MHC class II H-2IAg7/d molecules. The in vitro generation of insB-specific Tregs involved purification of either CD4+CD25+ or CD4+CD25- insTCR T cells, which were cultured for 1-2 weeks with the insB peptide, syngeneic antigen presenting cells and high doses of IL-2 yielding 25+ and 25- cultures, respectively. Using the classical in vitro suppression assay, only cells derived from the 25+ cultures were able to suppress while cells from the 25- cultures enhanced proliferation and cytokine secretion of CD8+ effector T cells. In vivo, however, cells from both cultures were unable to suppress lymphocytic choriomeningitis virus (LCMV)-induced diabetes. Interestingly, freshly isolated as well as IL-10-cultured CD4+CD25- but not freshly isolated CD4+CD25+ insTCR T cells suppressed spontaneous diabetes in NOD females. We are currently investigating the mechanisms underlying the in vivo suppressive potential of these CD4+CD25- T cells.

Real-time imaging of the pancreas during development of diabetes

K. Coppieters, M. M. Martinic, J. Oldham, A. Althage, Zacarias Garcia, D. Bresson, L. M. Togher, C. Huber, and M. G. von Herrath.

Our aim is to visualize islet antigen-specific T cell trafficking through the pancreas to understand kinetics of diabetes development. We adoptively transferred GFP+ P14 T cells (specific for the lymphocytic choriomeningitis virus (LCMV) glycoprotein (GP)) into RIP-GP mice (express LCMV-GP in the beta-cells of pancreatic islets) followed by real-time imaging of the RIP-GP pancreas using an inverted confocal microscope. Injection of Dithiocarbazone (DTZ) allowed visualization of pancreatic islets in vivo.

One day after transfer, GFP+ cells were distributed evenly throughout the pancreas. Three days later, GFP+ cells started to accumulate specifically around DTZ+ islets. Seven days after transfer, GFP+ cells were not only surrounding but also infiltrating the islets. GFP+ cells located close to a blood vessel moved fast whereas cells within the pancreatic tissue moved slowly and were predominantly found as cell clusters. Thirteen days after transfer, only few GFP+ cells remained in the pancreas. In summary, our approach allowed us to visualize GFP+ cells and their movement in DTZ+ islets.

Currently, we are generating experimental setups inducing lymphocytic infiltration into pancreatic islets with or without complete beta-cell destruction. To ameliorate visualization of beta-cells, we will be using RIP-GPxMIP-GFP and P14xDsRed transgenic mice allowing us to visualize infiltration of DsRed+ P14 cells into GFP+GP+ pancreatic islets. Finally, to improve tissue penetration, we will perform future experiments with a multi-photon microscope, which has a much greater penetration depth than a confocal microscope.

Virally Induced Regulatory Mechanisms in the Prevention of Type 1 Diabetes

Christophe M Filippi, Janine Oldham, Lisa Togher, Tom Wolfe, Marianne Martinic, Evelyn Rodrigo, Urs Christen & Matthias von Herrath

Type 1 Diabetes (T1D) is an autoimmune disease that results from the selective destruction of insulin-producing beta cells in the pancreas. While viral infections may be capable of triggering autoimmunity in genetically susceptible individuals, accumulating evidence indicates that viruses can also prevent T1D. Furthermore, viruses have been suggested to be potent inducers of CD4+CD25+ regulatory T cells (Tregs), which are known to play a crucial role in the prevention of autoimmunity. However, how viral infections may protect from autoimmune diabetes is still under investigation. We found that activation of CD4+CD25+ Tregs by acute lymphocytic choriomeningitis virus (LCMV) infection provides these cells with the capacity to prevent both spontaneous and virally induced T1D in the mouse.

Our goal is to understand how regulatory mechanisms associated with anti-viral immunity can prevent T1D. We believe that polyclonal activation of regulatory responders during virally induced inflammation could serve on one hand to modulate anti-viral immunity and on the other hand to prevent or abort potentially associated autoimmunity. Activation of Tregs during viral infections may be a general feature accounting for the ability of viruses to prevent T1D as well as other autoimmune disorders.

The selective induction of immune modulation is of high interest for influencing the course of autoimmune disease. Our studies should facilitate utilization of Treg induction to prevent autoimmunity. Furthermore, our findings support a model explaining the differential ability of viral infections to modulate diabetes, and our work should help evaluate the potential risk of viral infections in genetically susceptible children, thereby addressing current discordance concerning the role of viruses in T1D induction.

IL-10 Receptor Blockade in the Resolution of Chronic Viral Infection

Christophe M Filippi, Mette Ejrnaes, Marianne Martinic, Janine Oldham, Lisa Togher & Matthias von Herrath

Persistent infections can pose severe health risks. In many cases, for example after hepatitis virus infection, only a fraction of infected individuals clear the virus. The resulting chronic infection can lead to liver damage or liver cancer, particularly after hepatitis C virus (HCV) infection. Other viruses such as Epstein Barr virus (EBV) can also cause cancer if systemic reactivation occurs. Numerous chronic infections, in particular HCV, have been associated with systemic increase in IL-10 production. We found that blockade of IL-10 signaling using anti-IL-10 receptor antibodies can completely resolve persistent lymphocytic choriomeningitis virus (LCMV) infection while reducing immunopathology in the mouse. The finding is unique because, to date, immunization with dendritic cells or attempts to directly augment anti-viral effectors with vaccines has mostly failed to show beneficial effects and in some cases was found to augment immunopathology.

Our goal is to understand the immunological mechanisms underlying the beneficial effects of anti-IL-10R treatment in chronic infection, and aid to translate our findings to clinical applications in other chronic viral infections. Mechanistic understanding of IL-10R blockade in persistent LCMV infection will indicate whether IL-10 affects the outcome of infection, the amount of immunopathology, the occurrence of further complications, or indeed whether it could be the actual cause of persistence. Furthermore, preclinical evaluation of combination therapies using anti-IL-10R with other antibodies and viral vaccines will facilitate the therapeutic use of IL-10 blockade as a therapy for the treatment of persistent infections in humans.

Conventional immunotherapy of persistent viral infections has been unsuccessful so far, and we propose that tackling the problem from a different angle may be a crucial step toward successful treatment of chronic infections in humans.

Preclinical models of immunomodulation with peptide epitopes – dose and route, mechanism and enhancement of efficacy.

Georgia Fousteri, Damien Bresson, Mark Peakman, Bart Roep and Matthias von Herrath.

One of the main goals in treatment of autoimmune diseases is the intervention with the immune system without causing immunosuppression. Therapeutic approaches for the treatment of type 1 diabetes which combine decreased doses of immune modulators such the non Fc binding anti-CD3 in conjunction with insulin epitopes (proinsulin) have become of great interest because of the efficacy of the treatment and the reduction of the anti-CD3 immunosuppression side effects. The usage of peptides alone and/or in combination (multi-peptide therapy) has now become of a great interest because of their immunomodulator effect. So, it is proposed that immunosuppression therapies will be conceded by peptide-based therapies due to their specificity, efficacy and risk for lesser side effects.

In our lab, the main goal is the identification of particular peptides and/or combinations of them which have therapeutic potential for the treatment of autoimmune diabetes. Insulin-derived peptides will be administered to prediabetic and recent onset diabetic mice and the disease incidence will be followed. In addition, peptides will be delivered to diabetic prone mice (NOD and Rip-LCMV), by using modulated APCs as carriers. The mechanistic understanding of the peptide based therapy is an important issue, which will be addressed mainly by evaluating the T cell responses ex vivo and the T regulatory cells activity in vitro and in vivo.

In silico modeling of immunological systems.

As part of a collaborative team we aim to investigate aspects of immunological systems using computational methods that allow for the manipulation of such systems beyond what is possible in vitro/vivo. The approach exploited uses a combination of a stochastic cellular automaton which uses rule sets to replicate immune cell behavior and a differential equation based system which models cell populations. The methods are used in address two different aspect of immunological study; the prediction of in vitro cellular interactions for the refinement of physical experimentation and the modeling of biological phenomena to reveal the mechanisms behind them.

Project 1. Combination of anti-CD3 systemic treatment with antigen-specific immunizations to treat type 1 diabetes after onset.

Damien Bresson, Lisa Togher and Matthias von Herrath

After several years of antigen-specific immuno-interventions to treat type 1 diabetes (T1D), the general conclusion is that such treatments have to be given early (during the pre-diabetic phase) in animal models. These interventions would be suited for a prevention trial but not a recent-onset trial (which is a safer time frame for initial human therapy). Thus, antigen-specific interventions will need 'help' to be used in human recent-onset diabetics. The rationale of this project is based on the fact that in vivo administration of non-mitogenic (NM) anti-CD3 antibody (Ab) promotes a milieu for activation/expansion of regulatory T cell (Tregs) both in mice and in humans. We thus hypothesized that co-administration of NM-anti-CD3 Ab and islet-specific autoantigen immunizations after new-onset will expand islet-specific Tregs that will maintain long-term tolerance in vivo. To address this issue we focused our study on two specific aims:

Aim 1: Does combination of anti-CD3 systemic therapy with antigen-specific immunotherapy exhibit a strong synergistic effect in treating recent onset T1D?

Aim 2: Does the combinatorial treatment induce regulatory T cells and /or systemic immune deviation or both?


So far, we tested seven islet-aAg injected by different means, in combination with NM-anti-CD3:
- Human proinsulin II peptide (B24-C36) administered intranasally.
- GAD65 protein injected subcutaneously.
- GAD65 and bovine insulin B-chain injected intramuscularly by DNA vaccines.
- The insulin B(9-23) peptide as well as its Altered Peptide Ligand from Neurocrine administered intranasally.
- Human insulin injected intranasally or orally as a full protein.

1- The combinatorial therapy was tested in the NOD and RIP-LCMV models showing various efficacies. The data clearly underlined that intranasal human proinsulin II peptide (B24-C36) strongly synergized with NM-anti-CD3 and showed the most promising efficacy in treating T1D after recent onset. This efficacy was observed in both animal models and the combinatorial therapy revealed was always more efficient than the monotherapies (anti-CD3 or peptide alone). Mechanistically, we confirmed the induction of CD4+CD25+ proinsulin peptide-specific Tregs. Those cells also express the forkhead box transcription factor (FoxP3) and the glucocorticoid-induced tumor necrosis factor receptor (GITR) Treg markers. Most interestingly, the proinsulin-induced Tregs secreted "regulatory" cytokines Th3 (TGF-ß) and Tr1 (IL-10) after in vitro antigen-specific stimulation. In vivo, after adoptive transfers the proinsulin-specific Tregs divided rapidly in the pancreatic lymph nodes and not in the spleen (showing a site-specific activity). Importantly, they are capable of dampening self-antigen specific CD8+ immune responses by bystander suppression. This is of particular importance since after new-onset diabetes when the primary aAg has been targeted, the epitope spreading phenomenon may have redirected the responses against several 'secondary' aAgs. These data were published in the Journal of Clinical Investigation (Bresson D et al. J. Clin. Invest. 2006).

2- Co-administration of NM-anti-CD3 with a human GAD65-expressing plasmid (pCMV/hGAD65) showed also a strong efficacy in the RIP-LCMV-GP model. However, the same protocol did not synergize in the NOD model. This discrepancy underscores the importance of the genetic background in inducing efficient antigen-specific Tregs. Experiments are currently undertaken to understand mechanistically such a difference.

3- Based on the DPT-1 clinical trial, we learned that human insulin does not efficiently delay or prevent autoimmune diabetes when administered to at-risk patients in the prediabetic phase. However, it was worth trying to combine this aAg with NM-anti-CD3 to evaluate whether synergy is observed in treating T1D. The data obtained so far look promising and the protection increased from 48% to 56% from anti-CD3 alone to combination therapy, when human insulin is administered orally. More striking is the difference observed for the youngest animals (before 19 week of age) where 70% of NOD mice were protected with the CT as compared with 54% after anti-CD3 alone.

Project 2. To evaluate the pathogenetic role for epitopic and antigenic spreading in autoimmune diabetes.

Damien Bresson, Lisa Togher, Urs Christen and Matthias von Herrath.


Type 1 diabetes (T1D) is a T-cell mediated autoimmune disease, where insulin producing pancreatic beta cells are destroyed by autoaggressive CD4+ and CD8+ T lymphocytes. Importantly, autoreactive T cells are found in the peripheral blood of most individuals but are usually not activated. Thus, true tolerance, in the sense of complete absence of autoreactive lymphocytes by thymic negative selection or peripheral deletion, appears to be uncommon for certain autoantigens (aAg). The number and avidity of autoreactive cells that escape the thymic selection process are likely to be reduced under physiological conditions. It is still unclear how spontaneous loss of unresponsiveness or ignorance to non tolerant self-antigens can occur in the periphery under physiological conditions and what role the initiating putative aAg(s) plays during the different phases of the disease process.


The aim of our project is to evaluate the pathogenetic role of epitopic and antigenic spreading in T1D and to determine whether an initiating aAg needs to be expressed throughout the autoimmune process to maintain activated, autoaggressive T cells. Therapeutically, if we want to tolerize the response to one or a few antigens, we have to be sure that they are the drivers of the pathology. Because if not, the effect on the autoimmune process at a later stage (which is the realistic time for human therapy) remains iffy and is probably low to none, since autoaggressive responses to other antigens will have taken over. Thus, our basic question about antigenic spreading will be crucial for future immune interventions.

Short Description of Project:

We will use the RIP-LCMV model to assess the pathogenetic role for epitopic and antigenic spreading in T1D. H2-b RIP-LCMV-NP mice (slow-onset diabetes model) will be tolerized specifically against different autoaggressive CD4 (GP61-80 and NP309-328) and CD8 (GP33-41, NP396-404 and NP205-212) epitopes at different times before or after LCMV infection and diabetes incidence will be followed.

Project 3. RIP-LACK-NP mice: a new model to study bystander suppression and peptide/MHC class II dimer immuno-intervention in type 1 diabetes.

Damien Bresson, Lisa Togher, and Matthias von Herrath.

Collaboration: Valerie Verhasselt, Julie Cazareth and Nicolas Glaichenhaus (INSERM E03-44, Nice, France).


Type 1 diabetes (T1D) is one of the most common autoimmune diseases affecting almost twenty million people worldwide. During pathogenesis, insulin-producing pancreatic beta-cells are progressively destroyed by autoreactive CD4+ and CD8+ T cells. A destruction of approximately 80% of beta-cells occurs before T1D become symptomatic. In the past two decades, immune-modulatory approaches to prevent or cure T1D have been developed and tested with some encouraging recent results. Our group as well as others developed autoantigen-based immune interventions that have the ability to 'induce/expand' antigen-specific regulatory T cells (Tregs). These 'auto-reactive' Tregs can act as 'bystander suppressors' and block site-specifically (mainly in the pancreatic lymph nodes) heterologous autoreactive immune responses.

Recognition of a peptide-MHC complex by the T cell receptor (TCR) is a key interaction that initiates T lymphocyte activation or silencing during an immune response. Recombinant MHC class II-peptide reagents function as soluble mimetics of this interaction, bind to their specific TCR, and allow for detection/activation of antigen-specific CD4+ T cells. These cells are key effectors of autoimmunity, and those MHC class II peptide reagents, through their signaling properties, might also provide therapeutics to block the autoimmune process at its onset, analogous to the use of OKT3gamma1 (Ala-Ala) anti-CD3 antibody but in an antigen-specific fashion. The aim of such therapeutics is to potentiate different physiological control mechanisms to restore immune tolerance. Mechanisms initiated by this pathway may be capable of triggering elimination of pathogenic T cells through antigen-specific apoptosis and anergy, combined with the induction of regulatory T cells with broad suppressive function.


First, this project will aim to provide new insights into the mechanisms involved in the bystander suppression of autoimmune diabetes in vivo. Second, we will use recombinant MHC class II-LACK peptide dimers and evaluate their therapeutic capacity when injected into prediabetic or newly diabetic RIP-LACK-NP mice. All together, these data will help us in designing a new class of safe and efficient antigen-specific immune modulators for the treatment of T1D.

Short Description of Project:

We will use the transgenic balb/c RIP-LACK-NP mice, expressing both LACK (Leishmania homolog of the receptor for Activated C Kinase) and NP (nucleoprotein of the lymphocytic choriomeningitis virus [LCMV]) antigens specifically in the islets of Langerhans, to address both aims of the project. In this model, the NP constitutes the primary target for the immune system after infection with the LCMV. Generation of auto-reactive NP-specific CD8+ (and possibly CD4+) T cells lead to diabetes through a destruction of the NP-expressing beta-cells. We hypothesize that destruction of the pancreatic islets might result in the presentation of several islets antigens (among them the LACK protein) by the immune system in the pancreatic lymph nodes (PLNs). This should lead to the generation of LACK(156-173)-specific CD4+ T cells (in the balb/c background an immunodominant CD4 epitope encompassing the amino acid residues 156 to 173 is recognized after immunization with the LACK antigen). By using purified MHC class II-LACK(156-173) peptide dimers, we will try to modulate the LACK-specific CD4+ response and polarize these antigen-specific T cells toward a regulatory phenotype. The therapeutic potential of such MHC class II-peptide reagents will be evaluated and if successful this approach will enable us to study mechanistically the bystander suppression of autoaggressive NP-specific T cells by the LACK(156-173)-specific CD4+ Tregs.

Relevance to Type I Diabetes:

Development of potent antigen-specific therapeutics is the 'holy grale' in T1D as well as other autoimmune diseases. The attractive aspect of generating/expanding islet antigen-specific Tregs is that such cells can act as bystander suppressors, circumventing the need to identify all autoantigenic targets, and operate locally in the PLNs and islets, where they recognize their cognate antigens, without systemically affecting the immune system and compromising host defense. Antigenic vaccinations have been relatively unsuccessful when used alone to cure T1D after onset (which would be safer and require a shorter time frame for initial human therapy) and have to be combined with systemic immune modulators to become fully potent (Bresson D et al. J. Clin. Invest. 2006). The immune modulator calms the immune system and opens a 'window of opportunity' to expand vaccine-derived islet-specific Tregs. The HLA class II-peptide molecules combining an antigen specificity (such as antigenic vaccinations) and an immediate bio-availability (such as systemic immune modulators) could be an attractive antigen-specific therapy for the future.