Approximately 25 years ago, the introduction of methotrexate revolutionized the treatment of rheumatoid arthritis (RA) and juvenile idiopathic arthritis (JIA). In the late 1990s, this was followed by a second revolution: the introduction of biologic agents. Since the introduction of these treatment approaches, progress has not stopped as new targets for biologic therapies emerge and second-generation biologic agents are being produced. Nonetheless, this development represents progress mostly in terms of diversification and optimization rather than a new revolution. The next challenge in the treatment of RA and JIA is to reestablish self-tolerance with the goal of achieving long-term medication-free remission; in other words, a cure (1).
The discovery of Treg cells, first in animal models and later in humans, offered a completely new perspective regarding the establishment and maintenance of self-tolerance (2). Treg cells regulate immune responses through their ability to suppress other immune cells, such as Teff cells, natural killer cells, and antigen-presenting cells (APCs). In humans, Treg cells mostly reside within the population of CD4+CD25+CD127low T cells. They are characterized by expression of the lineage-specific transcription factor FoxP3. In various experimental models, Treg cells indeed make the difference between health and autoimmunity and were shown to be crucial for the maintenance of self-tolerance and immune homeostasis. The subsequent discovery that a gene mutation in FoxP3 leads to uncontrolled inflammation and autoimmunity in children with immune dysregulation, polyendocrinopathy, enteropathy, or X-linked syndrome raised expectations that in human autoimmune diseases, chronic inflammation may be caused by a deficiency in the Treg cell population. However, closer inspection revealed that the answer to the question of the role of Treg cells in humans was more complex.
Data on peripheral FoxP3+ Treg cell numbers and function in human autoimmune diseases are contradictory and remain subject to debate (3). To complicate this point further, it has become clear that FoxP3-expressing T cells are not a homogenous population (Table 1). First, FoxP3 is not a definitive marker for human Treg cells: conventional T cells can temporarily up-regulate FoxP3 expression upon activation without displaying a suppressive function (4), which hampers identification and isolation of human Treg cells. Moreover, it has been demonstrated that FoxP3-positive Treg cells may have a certain degree of plasticity (5) or instability (6, 7), resulting in differentiation into Teff-like subtypes with a reduced or complete loss of suppressive function, especially under inflammatory conditions. To add to this conundrum, new data have emerged indicating that Treg cells are not one of a kind: various subtypes can be determined, including specific populations that share Th cell characteristics but retain their suppressive capacities (8).
|1. “Classic” stable Treg cells with normal suppressive function|
|2. Plastic Treg cells that start to produce proinflammatory cytokines and have reduced suppressive function|
|3. Unstable Treg cells that lose FoxP3, differentiate into Teff cells, and do not have a suppressive function|
|4. Treg cells that share cytokine production with Th cell subsets but retain FoxP3 expression and suppressive function|
Various groups of investigators have reported an increased number of Treg cells in the synovial fluid of inflamed joints in patients with RA or JIA (3, 9–11). This inevitably leads to the question of why these Treg cells in the synovium are not capable of controlling inflammation. Is the proinflammatory environment simply too strong for Treg cells to overcome, or are Treg cells in inflamed synovial fluid not true Treg cells but effector cells in disguise that have temporarily up-regulated FoxP3?
In this issue of Arthritis & Rheumatism, Walter et al (12) present important new data that shed light on this puzzle. In elegantly performed in vitro studies, Walter and colleagues show convincingly that activated monocytes induce expression of not only antiinflammatory cytokines (interleukin-10 [IL-10]) but also proinflammatory cytokines (IL-17, interferon-γ, tumor necrosis factor α) by human CD4+CD45RO+CD25+CD127low Treg cells. Importantly, despite their capacity to express proinflammatory cytokines, these Treg cells maintain a regulatory phenotype in vitro and can still effectively suppress T cell proliferation and cytokine production. The authors conclude that these cytokine-expressing FoxP3+ T cells are true Treg cells and thus may still be potent suppressors at sites of inflammation. This is especially intriguing, because Th1- and Th17-like Treg cells produce exactly those cytokines that are most closely associated with the detrimental proinflammatory immune response in RA and are being targeted for immune intervention with biologic agents.
It must be noted that Walter et al made use of an in vitro system that mimics the activation of Treg cells in synovial fluid, using APCs that were preincubated with cytokines that are dominantly present in synovial fluid. Directly stimulating such cells with synovial fluid from joints with active arthritis is not feasible practically, while cells directly isolated from the synovial fluid of patients with RA are scarcely available. This is an important drawback, because the phenotype and function of T cells and APCs from the synovial fluid compartment likely differ from their peripheral blood counterparts because of their residence in an environment of chronic inflammation. Moreover, although in vitro suppression assays, as used by Walter et al, are the state-of- the-art technique for testing Treg cell function, such assays have important limitations. They reflect the suppressive effects on Teff cell function, while obviously many more cells contribute to joint inflammation in a real-life situation. Last, in vitro suppression may not adequately reflect in vivo suppression (Figure 1).
Despite these obvious limitations, the study by Walter and associates represents an important step forward in unraveling the role of Treg cells in arthritis. The results support recent observations suggesting that Treg cells present in inflamed synovial fluid are bona fide, stable Treg cells (9, 11). Furthermore, by showing that Treg cells cultured in an inflammatory environment not only retain a suppressive function but also acquire specific proinflammatory cytokine–suppressive capacities, these investigators add to the concept of phenotypic and functional specialization of Treg cells (8, 13). The emerging picture contains a broad pallet of activated T cells that appears to include not only “true” Th17 and Th1 cells, but also Th1-like and Th17-like Treg cells that may develop in parallel, colocalize, and act as a counterbalance (8).
Other important questions remain, e.g., regarding the role of effector cells. Various studies have implied that rather than a deficiency in Treg cell function or number, resistance of Teff cells may be crucial (3, 14). Interestingly, it seems that APCs play a pivotal role in priming both Treg cells and Teff cells. It is likely that cells other than APCs, including neutrophils, will both affect and be affected by Treg cell function.
Treg cells are crucial for restoring immune tolerance in chronic inflammation, and the study by Walter et al underscores that these cells also are likely to play a regulatory role in RA. Future studies are needed to reveal the full picture of the different subtypes of Treg cells, their function, and their relationship to other immune cells at the site of inflammation. This is a clear challenge but is worth the efforts, because true comprehension of the role of Treg cells in joint inflammation may hold a clue for a new and truly innovative manner of treating RA and JIA.