Despite treatment advances, rheumatoid arthritis (RA) is still associated with significant disability, decreased work capacity, and reduced life expectancy. Effective immunotherapies to restore immune tolerance promise greater specificity, lower toxicity, and a longer-term solution to controlling and preventing RA. Design of effective therapies requires a fundamental understanding of the critical immunopathogenetic pathways in RA. This article reviews advances in the understanding of self-antigen-specific T cells in autoimmune diseases including RA and type 1 diabetes, which bring exciting insights to the mechanisms underpinning loss of tolerance and how tolerance could be restored for disease prevention in the preclinical or recent-onset period.
Key points
- •
pMHC tetramers specific for citrullinated autoantigens identify antigen-specific T cells in HLA-DRB1*0401 + rheumatoid arthritis (RA) patients and healthy controls. In type 1 diabetes, antigen-specific T cells are emerging as diagnostic biomarkers.
- •
In patients with RA, the frequency of self-antigen–specific CD4 + T cells is correlated with disease activity, similar to findings in collagen-induced arthritis and type 1 diabetes models.
- •
In RA patients, the proportion of self-antigen–specific CD4 + regulatory T cells is reduced relative to healthy controls, suggesting the hypothesis that HLA-DRB1 susceptibility alleles fail to promote citrullinated peptide–specific T cell regulation to prevent the development of RA in the face of genetic, environmental, or inflammatory pressures on regulatory T cells.
Introduction
Rheumatoid arthritis (RA) is a common and incurable systemic inflammatory autoimmune disease affecting 1% to 2% of the population. RA is characterized by inflammation of joint synovial tissues and extra-articular sites, with complications of erosive joint damage, lung fibrosis, atherosclerosis, and infection. Current treatments such as disease-modifying antirheumatic drugs and biologic inhibitors of tumor necrosis factor (TNF), interleukin 6 (IL-6), T cells, and B cells are nonspecific and are associated with side effects. Although current drugs block inflammation, they induce long-term remission in less than 50% of patients on treatment. Thus, despite treatment advances, RA is still associated with significant disability, decreased work capacity, and reduced life expectancy, leading to enormous social and economic burden.
Effective immunotherapies to restore immune tolerance are considered the “Holy Grail” in autoimmune diseases. Such strategies promise greater specificity, lower toxicity, and a longer-term solution to controlling and preventing RA. However, antigen-specific tolerizing immunotherapy has been more difficult to achieve than blocking inflammation. Design of effective therapies requires a fundamental understanding of the critical immunopathogenetic pathways in RA. The preclinical phase of RA represents a valuable period of time to understand the relationships between genetics, environment, and immune dysregulation in the development of RA and to evaluate preclinical interventions and novel immunomodulation including antigen-specific therapy. The capacity to visualize and characterize antigen-specific autoreactive T cells longitudinally during this period is central to understand dysregulation of adaptive immunity in autoimmune diseases such as RA.
In this article, the authors review advances in understanding self-antigen–specific autoreactive T cells in RA. These developments bring exciting insights to the mechanisms underpinning loss of tolerance and how tolerance could be restored to bring us closer to the goal of RA prevention in the preclinical or recent-onset period.
Introduction
Rheumatoid arthritis (RA) is a common and incurable systemic inflammatory autoimmune disease affecting 1% to 2% of the population. RA is characterized by inflammation of joint synovial tissues and extra-articular sites, with complications of erosive joint damage, lung fibrosis, atherosclerosis, and infection. Current treatments such as disease-modifying antirheumatic drugs and biologic inhibitors of tumor necrosis factor (TNF), interleukin 6 (IL-6), T cells, and B cells are nonspecific and are associated with side effects. Although current drugs block inflammation, they induce long-term remission in less than 50% of patients on treatment. Thus, despite treatment advances, RA is still associated with significant disability, decreased work capacity, and reduced life expectancy, leading to enormous social and economic burden.
Effective immunotherapies to restore immune tolerance are considered the “Holy Grail” in autoimmune diseases. Such strategies promise greater specificity, lower toxicity, and a longer-term solution to controlling and preventing RA. However, antigen-specific tolerizing immunotherapy has been more difficult to achieve than blocking inflammation. Design of effective therapies requires a fundamental understanding of the critical immunopathogenetic pathways in RA. The preclinical phase of RA represents a valuable period of time to understand the relationships between genetics, environment, and immune dysregulation in the development of RA and to evaluate preclinical interventions and novel immunomodulation including antigen-specific therapy. The capacity to visualize and characterize antigen-specific autoreactive T cells longitudinally during this period is central to understand dysregulation of adaptive immunity in autoimmune diseases such as RA.
In this article, the authors review advances in understanding self-antigen–specific autoreactive T cells in RA. These developments bring exciting insights to the mechanisms underpinning loss of tolerance and how tolerance could be restored to bring us closer to the goal of RA prevention in the preclinical or recent-onset period.
Preclinical Rheumatoid Arthritis
The preclinical phase of RA encompasses an initial phase of risk secondary to genetic and environmental factors, characterized by asymptomatic autoimmunity. Multiple genetic and environmental factors influence RA pathogenesis and progression. First-degree relatives (FDR) of patients with RA are a known, high-risk population with a 5- to 7-fold increased risk for incident RA. FDR as well as cohorts of unrelated individuals with high genetic and environmental risk factors for RA represent valuable, enriched populations in which to further investigate asymptomatic autoimmunity and preclinical RA and to examine the factors influencing the development and progression of autoimmunity, using appropriate biomarkers. Further, immunotherapeutic interventions aimed at disease prevention can be explored in such populations before the onset of disease.
The molecular mechanism of shared epitope and Anticitrullinated Protein Antibodies+ Rheumatoid Arthritis
Historically, alleles associated with seropositive RA such as HLA-DRB1*0401 , *0404 , *0101 molecules in Caucasians, HLA-DRB1*0405 in Asians, and HLA-DRB1*1402 in North American Natives (NAN) were proposed to share a conserved epitope around amino acids 70 to 74. This epitope is commonly known as the “shared susceptibility epitope” (SE) and includes a positively charged residue at position 71 (see entries in bold in Table 1 ). The positively charged residue at position 71 has been proposed to dictate the nature of amino acid that can be accommodated in the P4 pocket. In line with this observation, 70% of RA patients have the disease-specific anticitrullinated protein antibodies (ACPA). Several citrullinated (cit) autoantigens, including fibrinogen, aggrecan, vimentin, collagen type II, and α-enolase have been described in RA. Citrulline is a post-translational modification of arginine, which occurs during inflammation, endoplasmic reticulum stress, and autophagy. Presence of ACPA in RA serum implies the presentation of cit-antigens to T cells by dendritic cells (DC) and B cells to support ACPA production by B cells.
HLA-DRB1 Amino Acid at Position | Classical HLA-DRB1 Alleles | Odds Ratio | |||
---|---|---|---|---|---|
11 | 13 | 71 | 74 | ACPA+ RA | |
Val | His | Lys | Ala | *0401 | 3.3 |
Val | His/Phe | Arg | Ala | *0408, *0405, *0404, *1001 | 2.8–10.3 |
Leu | Phe | Arg | Ala | *0102, *0101 | 1.5 |
Pro | Arg | Arg | Ala | *1601 | 1.2 |
Val | His | Arg | Glu | *0403, *0407 | 0.5–0.7 |
Asp | Phe | Arg | Glu | *0901 | 2.1 |
Ser | Ser | Arg | Ala | *1402 | >1 |
Val | His | Glu | Ala | *0402 | 1.2 |
Ser | Ser | Lys | Ala | *1303 | 0.6 |
Pro | Arg | Ala | Ala | *1501, *1502 | 0.9 |
Gly | Tyr | Arg | Gln | *0701 | 0.7 |
Ser | Ser/Gly | Arg | Ala | *1101, *1104, *1201 | 0.5–0.7 |
Ser | Ser | Arg | Glu | *1401 | 0.6 |
Leu | Phe | Glu | Ala | *0103 | <1 |
Ser | Gly | Arg | Leu | *0801, *0804 | 0.4 |
Ser | Ser | Lys | Arg | *0301 | 0.5 |
Ser | Ser | Glu | Ala | *1102, *1103, *1301, *1302 | 0.5–0.6 |
In 2012, a large haplotype association study reported by Raychaudhuri and colleagues attributed most of the DR-associated risk of RA to positions 11, 13, 71, and 74 of DRβ, effectively extending the SE to additional amino acids in the floor of the antigen-binding groove surrounding pocket 4. Unexpectedly, positions 11 and 13 had the highest statistical association with RA risk; the reasons for this are not yet clear.
The prediction that RA risk allomorphs permit binding and presentation of citrullinated peptides was confirmed by solution of the crystal structure of several cit-self-peptides bound to SE + HLA-DRB1*0401 and 0404 . Citrulline occupies the P4 pocket and interacts with Lys71β through a hydrogen-bonding network. In contrast, arginine could not be accommodated at P4. His13β forms van der Waals contacts with the P4 citrulline aliphatic moiety. Moreover, a His13βSer polymorphism, present in RA-resistant allomorphs such as HLA-DRB1*1301 , is predicted to affect the packing of citrulline in this location. Curiously, the NAN RA risk allele HLA-DRB1*1402 contains the SE, as well as Ser11β and Ser13β. Individuals carrying the non-SE allele HLA-DRB1*0402 have been previously proposed to be protected from RA development. In contrast to HLA-DRB1*0401 or *0404, which did not accommodate Arg, HLA-DRB1*0402 was able to accommodate both citrullinated and native vimentin due to the presence of negatively charged aspartic acid and glutamic acid at positions 70 and 71, respectively. The knowledge from understanding the molecular association of SE and RA warrants further investigation on how the differential presentation of peptides influence the T-cell response.
Immune tolerance and the development of autoimmunity
Thymic deletion of immature self-reactive T cells with high-affinity T cell receptors (TCRs) constitutes a major mechanism of self-tolerance during neonatal life. Such overtly autoreactive T cells can also be silenced or anergised, either as immature thymocytes or as mature T cells in the periphery. These “recessive” tolerance mechanisms reduce the likelihood of autoimmunity resulting from high-affinity autoreactive T cells. However, low-affinity autoreactive T cells escape the deletion threshold and persist, requiring “dominant” tolerance mechanisms involving regulatory T cells (Treg) for their control. Foxp3 + Treg may be selected intrathymically or in the periphery in response to antigen presentation by specialized subsets of DC. Autoimmune diseases, including RA and type 1 diabetes (T1D), are associated with genetic variants such as IL2, ILl2RA, and CTLA4 , which perturb Treg generation, proliferation, or function. Moreover, RA is associated with genetic variants such as TNFAIP3 that enhance the proinflammatory environment and may promote Treg dysfunction. However, despite the evidence that HLA is the strongest genetic association with RA and that the susceptibility allomorphs preferentially bind cit-autoantigens, it is not yet clear how the recognition of self-antigens presented by susceptibility or protective HLA molecules influences Treg development or function. The dominant protection associated with the inheritance of protective HLA alleles in RA suggests that these alleles promote cit-peptide–specific T-cell regulation to prevent the development of RA more effectively than alleles associated with RA susceptibility.
Measuring antigen-specific T-cell responses
Antigen-specific CD4 + and CD8 + T cells are key players in autoimmune diseases such as RA and T1D; however, the frequencies and roles of antigen-specific T cells in disease pathogenesis have been difficult to elucidate. T cell responses have typically been measured in bulk T cell populations using indirect methods such as enzyme-linked immunospot assay (ELISpot), flow cytometry–based intracellular cytokine, and activation markers (eg, CD69, CD154, and CD25). However, these methods require in vitro stimulation, which may alter their phenotype and functionality. Furthermore, these methods can only detect antigen-experienced T cells that are capable of producing cytokines on restimulation, ignoring T cells that produce little cytokine on stimulation and overestimating cells stimulated nonspecifically in vitro. Therefore, relying on these techniques to enumerate and characterize the T cell response may be inadequate. The development of multimeric peptide-bound major histocompatibility complex (pMHC) complexes (tetramers) has revolutionized the detection of antigen-specific T cells in mice and humans. Altman and colleagues first described tetramers in 1996, which allowed ex vivo detection of virus-specific T cells by flow cytometry. Since then both class I and class II pMHC tetramers in combination with other staining techniques have been used to study CD8 + and CD4 + T cells respectively in viral infections, tumor responses, and autoimmune diseases.
Tetramers: tools of the trade
αβ TCR recognize peptides derived from pathogens, tumors, and self that are presented in the context of MHC molecules on antigen-presenting cells. This recognition is very specific for the MHC molecule and peptide. It is therefore possible to detect antigen-specific T cells using soluble pMHC labeled with fluorochromes directly by flow cytometry. The interaction between TCR and pMHC complex is characterized by low affinity and fast off-rate, which are necessary for each TCR to make serial contacts with multiple pMHC complexes. However, these features represent a challenge for staining antigen-specific T cells with monomeric pMHC complexes. Higher-order multimer construction resulted in increased pMHC-binding affinity, decreased off-rate, and successful staining of T cells. One of the most commonly used pMHC multimers is a tetramer, consisting of 4 identical biotinylated MHC molecules, each loaded with a single peptide linked to a central streptavidin molecule and coupled to a fluorochrome ( Fig. 1 ).
Tetramers have become an essential tool in flow cytometric studies, but use in histologic sections has been limited in human studies due to technical challenges. The advantages are that they have high specificity and sensitivity, discriminating extremely rare antigen-specific T cells when used sequentially with cell enrichment (10 antigen-specific T cells in 2.5 × 10 8 mixed populations of cells). Therefore, tetramer assays can be used to enumerate and characterize antigen-specific T cells ex vivo without in vitro stimulation. For example, a recent study found that greater than 50% viral antigen–specific CD4 + T cells displayed a memory phenotype in adults naïve to those viral infections. This phenomenon was attributed to the strong propensity of αβ TCRs to cross-react with different pMHC complexes due to their flexible binding sites. In support of this, the study found an extensive cross-recognition of viral antigen-specific T cells with other microbial peptides.
The major limitation of tetramer assays in human studies is that prior knowledge of specific epitopes for each HLA allele is required to construct the tetramers. This limitation is not so critical when an immunodominant epitope is known, for example, hemagglutinin protein of influenza (HA 306-318 ) restricted by the SE allomorph DRB1*0401. CD4 + T cells specific for HA 306-318 acquired through natural exposure or annual flu vaccination can be readily detected in peripheral blood (PB) of HLA-DRB1*0401 + individuals. However, when there is a set of complex epitopes recognized by T cells, or when the epitope is unknown, alternative strategies are needed to identify binding epitopes. Identification of MHC-binding epitopes has traditionally involved screening large libraries of peptides with functional assays such as interferon gamma ELISpot, which is laborious and prone to overestimation in diseases such as RA, where background cytokine production is high. An unbiased approach known as the tetramer-guided epitope mapping (TGEM) has been designed to identify T cell epitopes. Briefly, a panel of overlapping peptides is divided into pools and each peptide pool is then loaded onto soluble MHC molecules to generate pooled peptide tetramers. The pooled peptide tetramers are then used to stain PB mononuclear cells (PBMC) that have been stimulated with the corresponding whole antigen; this allows the identification of peptide tetramers with positive staining within the pool. Each peptide in the pool is then loaded individually onto MHC molecules and the PBMC staining is repeated to identify the antigenic epitopes ( Fig. 2 ). However, TGEM lacks the ability to identify post-translational modified epitopes. A recent strategy that combined TCR selection of highly diverse yeast-displayed peptide-MHC libraries with deep sequencing identified activating microbial and self-ligands for human autoimmune diseases.