Exploiting T cell crosstalk as a vaccination strategy for rheumatoid arthritis

T cells are among the key players in the pathogenesis of rheumatoid arthritis (RA) (1). This is illustrated by many observations, ranging from the identification of an association between disease susceptibility and certain class II major histocompatibility complex (MHC) alleles in patients with RA, to more recent findings suggesting a role of Tregs in RA (2, 3). As a consequence, over time, many attempts have been made to interfere in the disease process by targeting potential pathogenic T cells. Most of these attempts, ranging from the depletion of CD4+ T cells to induction of mucosal tolerance with crude antigens, have shown great promise in animal models of arthritis, but only moderate efficacy in patients (4). So far, none of these strategies have been incorporated into the current mainstream therapy for RA.

Ideally, therapy directed against T cells should target only those T cells that are involved in the perpetuation of the inflammation, for 2 reasons. First, nonspecific targeting of T cells will inevitably lead to undesirable side effects, such as highly increased susceptibility to infections. Second, T cells not only are among the culprits involved in synovial inflammation, but a subset of T cells, the so-called Tregs, is involved in control of the inflammatory response (3, 5). Eliminating these cells could arguably lead to more severe inflammation.

Thus, T cell–directed therapy has to be specific, and indeed, several novel approaches for specific targeted T cell therapies, including epitope-specific therapy, are now being developed for RA and other autoimmune diseases (4). T cell vaccination is one of the earliest technologies developed to enable specific interference with T cell function in autoimmunity.

T cell crosstalk and T cell vaccination

T cell vaccination conceptually relies on the basic principles of T cell activation. CD4+ T cells are activated after specifically recognizing, with their T cell receptor (TCR), a processed peptide antigen in the context of a class II MHC molecule on an antigen-presenting cell. Not only “professional” antigen-presenting cells, such as dendritic cells and B cells, but also T cells can express class II MHC molecules on their cell surface and present antigens to other T cells. Class II MHC expression on T cells occurs after cell activation and is a characteristic marker for activated CD4+ T cells (6). The resulting “T–T interaction” (a T cell presenting an antigen to another T cell) helps to down-regulate the activated T cell.

This T cell crosstalk has functional consequences, as is underscored by observations made in the late 1980s in experimental autoimmune encephalomyelitis (EAE), a well-studied experimental model of multiple sclerosis. Lohse and coworkers studied the effect of vaccination with the attenuated autoimmunity-inducing T cell clone on EAE (7). Vaccination with an attenuated autoimmune T cell clone led to the induction of a specific T cell response directed against the autoimmune T cells, and a subsequent clearance of the disease. This T–T crosstalk is the guiding principle of T cell vaccination. In animal models of autoimmunity, T cell vaccination is an effective treatment, not only of EAE but also of experimental diabetes mellitus, myasthenia gravis, and arthritis, e.g., collagen-induced arthritis and adjuvant-induced arthritis (8, 9). Pivotal studies of both multiple sclerosis and RA, including the study by Chen et al in this issue of Arthritis & Rheumatism, have underscored the great potential of T cell vaccination, and T cell–targeted therapy in general, for the treatment of human autoimmune diseases (10–12).

Antiidiotypic and antiergotypic T cells

To fully understand the value of the study by Chen et al, some further clarification is needed. Studies in animal models have shown that at least 2 types of T cells are induced by T cell vaccination: antiidiotypic and antiergotypic T cells (10) (Table 1). The antiidiotypic response is a specific T cell response directed against the autoimmune T cells targeted in the vaccine. Antiidiotypic T cells respond to peptides derived mainly from the CD3 region of the TCR of the autoimmune T cells. The antiidiotypic response consists of a CD4 T cell response against TCR peptides presented by autoimmune T cells in the context of class II MHC, followed by a CD8 T cell response against TCR peptides presented by class I MHC (HLA–E in humans, Qa-1 in mice) on autoimmune T cells (10, 13).

Table 1. Main characteristics of 2 types of T cells induced by T cell vaccination in autoimmune diseases

	Antiidiotypic T cells	Antiergotypic T cells
* IL-10 = interleukin-10; TGFβ = transforming growth factor β.
Specificity	T cell receptor of autoimmune T cells targeted in the vaccine	Ergotopes: activation molecules on activated T cells
Type	CD4 and CD8	CD4 and CD8
HLA restriction	Class II and class I (HLA–E in humans, Qa-1 in mice)	Class II and class I
Mechanism of action	Mixed	IL-10, TGFβ*
Function	Targeting activated autoimmune cells	Targeting activated cells and bystander suppression

However, this mechanism, which is specific for the autoimmune T cells driving the disease, does not account for all the effects of T cell vaccination observed in animal models. By the late 1980s, investigators had found that T cell vaccination is successful only if T cells are activated in vitro prior to injection; resting T cells are not effective. A second observation was that (again) attenuated activated T cells with specificities other than those of autoimmune T cells are also effective in EAE (7). This was remarkable, because, as discussed above, antiidiotypic T cells induced by T cell vaccination are directed against peptides derived from the variable region of the TCR of autoimmune T cells, and thus are highly specific for their target cells.

Cohen et al (10) called this second type of Treg induced by T cell vaccination an antiergotypic T cell. The word “ergotype” is derived from the ancient Greek word “ergos,” which can be translated as “work” or even “working together.” This is meant to express the fact that those T cells are directed against molecular markers (“ergotopes”) on activated T cells. Examples of ergotopes are interleukin-2 receptor α (IL-2Rα) (CD25) and stress proteins such as Hsp60; many other molecules may also fulfill these criteria (9). Antiergotypic cells exert strong regulatory effects on activated effector T cells and are of importance for the natural regulation of the immune response (10). Antiergotypic T cells increase the therapeutic potential of T cell vaccination in human diseases (such as RA) without known disease-inducing autoantigens or T cell clones.

T cell vaccination in arthritis

Indeed, one of the most compelling problems associated with T cell vaccination in RA has been the selection of cells or TCR peptides for vaccination. In RA, “universal” disease-inducing T cell clones are not known, so which cells should be targeted? The first controlled study of T cell vaccination in RA made elegant use of the finding revealed in some (but not all) studies, of a skewing of the T cell repertoire toward certain TCRV_α and TCRV_β genes in synovial fluid and tissue of RA patients (2). Based on this and other preclinical studies, 3 peptides derived from 3 TCRV_β chains (V_β3, V_β14, and V_β17) were selected for a multicenter, placebo-controlled, phase II clinical trial in RA patients. Patients enrolled in that study, which was published in this journal in 1998, received a total of 4 intramuscular injections of a mixture containing these 3 peptides in Freund's incomplete adjuvant (14). The vaccine was safe, well tolerated, and seemed to induce some clinical improvement in patients with active RA.

Although that study was an important step forward, some questions remained (15). The TCR peptide vaccine was not strongly immunogenic. Therefore, it was not clear whether the vaccination resulted in the desired immune deviation, and, if so, whether such immune deviation was related to clinical efficacy. This approach was aimed at inducing antiidiotypic T cells, directed at a certain subpopulation of activated autoimmune T cells. In RA, many overlapping effector mechanisms are in play, and it seems unlikely that only the T cell clones targeted in the vaccine are responsible for the perpetuation of the disease. Ideally, to increase efficacy, it would be preferable if the vaccine could also influence other (antiergotypic) regulatory cells.

In this issue of Arthritis & Rheumatism, Chen and coworkers describe an approach for T cell vaccination that bypasses the need for defined TCR peptides from autoimmune T cell clones (used in the classic T cell vaccination studies) and activates a broader repertoire of regulatory cells: both antiidiotypic and antiergotypic T cells (11). Their approach was essentially simple and resembled the first attempt with T cell vaccination in RA (16). Based on the assumption that most relevant autoimmune T cells in RA patients can be found in the synovial fluid, Chen et al isolated synovial cells from inflamed joints of patients with active RA. They characterized those cells for their TCRV_β repertoire, and expanded them in vitro using a straightforward expansion protocol with IL-2 and phytohemagglutinin. After expansion, the cells were reanalyzed to determine whether their phenotype was changed significantly during the expansion period (which was not the case), and were frozen in aliquots for the vaccination.

Typically, the vaccine T cells were mixed CD8 (67%) and CD4 (26%), skewed toward V_β14 (41%), V_β17 (18%), and V_β13 (27%), expressed no Foxp3 (the transcription factor associated with T regulatory function), and produced tumor necrosis factor α (TNFα) and interferon-γ, but no IL-10, upon activation. Thus, they fulfilled a phenotype of activated effector cells, as would be expected with synovial fluid–derived mononuclear cells. Prior to vaccination, the cells were activated in vitro and irradiated. They were then administered subcutaneously without an adjuvant, 6 times over a period of 9 months.

An important aspect of this study is the thorough analysis of the immunologic changes induced by the vaccine over 9 months. The authors observed that T cell vaccination induced a mix of both antiidiotypic and antiergotypic T cells. The antiidiotypic T cells consisted of cytotoxic CD8+ granzyme B–producing cells and CD4+ (mainly V_β2+), IL-10–producing T cells, both cell types being specific for the vaccine. In addition, changes in the immunologic repertoire of the patients that went beyond changes in the antiidiotypic (vaccine) T cells were found. The authors observed an increase in CD4+,CD25+ Foxp3+ T cells with an increased in vitro suppressive capacity. Moreover, the vaccine induced CD4+ cells that responded to activated but not resting T cells. Interestingly, some of these vaccine-induced cells responded to peptides derived from IL-2Rα (CD25), making them typical antiergotypic T cells.

The importance of this study lies in 3 areas. First, the study underscores the importance of developing surrogate immunologic markers of efficacy in trials of novel immunotherapies. Because these approaches aim to affect specific pathways of the immune response, conventional measures of therapeutic efficacy are not sufficient to understand their in vivo biologic effects. This is also illustrated by other recent studies that showed convincingly that the effect of TNFα inhibitors goes well beyond blockade of the TNFα pathway; they also influence Treg function (3). Second, this study provides the proof of principle that T cell vaccination can induce specific Tregs, both antiidiotypic and antiergotypic, in patients with RA, even when an autoantigen or an autoimmune T cell clone is unknown. Third, the study shows that the progress in basic knowledge of T cell regulation is now finally reaching a level at which it can be translated to patient care. This may open many new opportunities, ranging from T cell vaccination to other strategies directed at specific T cell deviation.

Obviously, because this was an open-label, phase I trial, not much can be said of the clinical effects of T cell vaccination in this study. The data provided by the authors do show that there were no side effects during the course of the study, and treatment did not worsen the disease. This is consistent with the excellent safety profile observed in other studies of T cell vaccination or other specific T cell therapies in RA. Will the approach reported by Chen et al develop into a treatment used in daily practice of rheumatology? This seems unlikely, since the procedure (isolating and expanding cells in vitro) requires extensive in vitro manipulation, which makes it expensive and time-consuming. However, this does not detract from the importance of this investigation, which shows that it is possible to specifically influence the T cell immune repertoire in RA patients in a beneficial way. Future steps could include the use of specific ergotopes such as CD25 and heat-shock proteins for the induction of Tregs (9). The study by Chen and colleagues demonstrates the safety and immunologic strength, and thus the attractiveness, of such an approach.