Autoimmune hepatitis (AIH) is a chronic, progressive necroinflammatory disease putatively caused by loss of tolerance to hepatic autoantigens.1 The current concept of pathogenesis involves initiation by environmental triggers in persons with immunogenetic predisposition reflected by strong human leukocyte antigen (HLA) class II haplotype associations, loss of immunological tolerance to liver autoantigens, generation of an unregulated T-cell–mediated immune attack against those autoantigens, and production of non–species-specific autoantibodies.1-3 In the absence of biomarkers, the diagnosis of AIH is based on characteristic histological features of interface hepatitis, clinical and laboratory findings, and the presence of autoantibodies that permit subclassification of AIH into type 1 (anti-nuclear and/or smooth muscle autoantibodies) and type 2 (anti-liver-kidney-microsomal-1 [LKM-1] autoantibodies).1
Prevention of autoimmunity is achieved through the interaction of professional antigen-presenting cells (APCs), T effector, and T regulatory (Treg) cells (Fig. 1).4-6 Autoimmunity primarily results from failure of natural Treg generated in the thymus to prevent initial reactions against autoantigens. Thereafter, organ-specific autoimmunity is driven by interplay between T effector and antigen-specific inducible Treg (iTreg) that determine the duration, extent, and distribution of inflammation within the organ. Complete understanding of the interplay between T effector cells and Treg cells in AIH may lead to novel strategies for therapeutic regulation of the autoimmune response in this and other diseases.7
The pathogenesis of organ-specific autoimmune diseases involves interplay of CD4 T helper (Th) 1 cells promoting immunopathology through proinflammatory cytokines, CD8 cytotoxic T lymphocyte (CTL) cytotoxicity, CD4 Th2 cells promoting antibody production, Th17 and Th9 cells intensifying inflammation, and the immunoregulatory functions of antigen-specific, inducible Treg cells (Fig. 1). Collectively, they determine the extent of immunopathology through their direct effector functions and activation and recruitment of macrophages, neutrophils, eosinophils, and natural killer (NK) cells.5 The adaptive immune response of T cells requires processing of antigens into peptides that fit the antigen-binding grooves of major histocompatibility complex (MHC; termed HLA in humans) class I and class II molecules and presentation of the MHC-peptide antigen complexes to T cell receptors (TCR). The TCR of CD8 T cells recognize only class I MHC-antigen complexes, while those of CD4 T cells recognize only class II MHC-antigen complexes. T cell activation is optimal when MHC-antigen complexes are presented by professional APC, comprised of activated B cells and macrophages and mature dendritic cells (DCs), because these APC coexpress costimulatory molecules CD80 and CD86. Binding of CD80 and CD86 to the CD28 receptor on both naïve CD4 and CD8 T cells delivers a costimulatory signal required for functional differentiation into CD4 T cell subsets and CD8 CTL (Fig. 1).
Treg cells are indispensable for maintenance of self-tolerance and regulation of the extent and duration of normal immune responses. T cell precursors traffic to the thymus, where those with TCR with high affinity for autoantigens are deleted.8 During thymic development of CD4 T cells, a variable proportion express repressor forkhead winged helix transcription factor box (FoxP3) and later differentiate into natural CD4 Treg cells with the phenotype CD4+CD25+highCD45RO+highCD62L+highCD127lowFoxP3+high, where “high” refers to the degree of expression, CD25 is the receptor for the α-chain of the T cell mitogenic cytokine IL-2, CD62L is a lymph node homing receptor, and CD45RO is an activation marker in humans.4, 6 While this phenotype is similar for natural and iTreg, iTreg are functionally less stable. Of potential importance, FoxP3 expression is subject to epigenetic control, indicating that a functionally altered gene can be inherited by progeny cells.9 Furthermore, it is now clear that T effector cells also express low levels of FoxP3; thus, FoxP3 expression alone cannot be used to define Treg cells.6 In addition to classical CD4+CD25+FoxP3+ Treg cells, suppression of autoimmunity and regulation of normal immune responses are also mediated by CD4 interleukin (IL)-10-secreting T regulatory 1 cells (Tr1), CD4 transforming growth factor β (TGFβ)-secreting Th3 cells, CD8 T suppressor cells, and some natural killer T (NKT) cells (Fig. 1).4
The central role of natural Treg in maintenance of self-tolerance was confirmed by evidence that their deficiency results in fatal autoimmune diseases and chronic inflammation with lymphoproliferation (4, 6). In humans, mutation of FoxP3 causes Treg deficiency and the immune dysregulation, polyendocrinopathy, enteropathy, X-linked (IPEX) syndrome. In mice, a naturally occurring FoxP3 mutation also causes systemic autoimmunity, which can be reproduced experimentally by deletion of Treg cells.
Natural Treg cells generated in the thymus recognize virtually all autoantigens and preferentially migrate to peripheral lymphoid organs where exposure to peptide antigens leads to activation of antigen-specific suppressive functions to prevent conversion of immature DCs expressing autoantigenic peptides to mature DCs capable of activating naïve autoreactive T cells. In contrast to naïve T cells, which require high levels of both class I and II MHC-antigen complexes and costimulatory CD80/CD86 molecules for activation, iTreg can be fully activated by semimature DCs (smDCs) expressing low levels of both MHC-antigen complexes and costimulatory CD80/CD86.4 The state of maturation of the DCs is of particular interest, since smDCs in mice induced optimal antigen-specific expansion of CD4+CD25+FOXP3+ Treg cells in vitro.10 Presentation of peptide antigen with submaximal costimulation appears to be essential for activating Treg function in animal models of autoimmunity.11
Type 2 AIH is ideally suited to explore the role of iTreg in pathogenesis and their potential therapeutic use. In contrast to type 1 AIH, in which the hepatic autoantigens are poorly defined,3 the autoantigenic epitopes for B, CD4, and CD8 T cells in type 2 AIH are located on cytochrome P450IID6 (CYP2D6).2 The immunodominant autoantigenic B cell epitope is CYP2D6193-212, but additional minor epitopes have also been defined. Epitopes CYP2D6193-212, CYP2D6217-260, and CYP2D6305-348 are recognized by B, CD4, and CD8 T cells. In addition, type 2 AIH is strongly associated with two class II HLA-DR alleles: HLA-DRB1*0701 (DR7) and HLA-DRB1*0301 (DR3), which allows selection of patients with and without these alleles for studies.2
At the time of diagnosis, both the quantity and function of CD4+CD25+FoxP3+ iTreg cells in peripheral blood are deficient in patients with type 2 AIH.12, 13 However, successful therapy with corticosteroids and/or azathioprine partially restored the circulating numbers and functions of iTreg,12, 13 indicating that reduction of inflammatory disease activity and deleterious effector T cell functions facilitated iTreg function. In children with type 2 AIH, the quantities of iTreg were significantly inversely correlated with disease severity as well as with titers of anti–soluble liver antigen (SLA) and anti-LKM1 autoantibodies.13 While the inverse correlation with autoantibody titers has been interpreted as evidence of a pathogenetic role for autoantibodies, a plausible alternative explanation is that the paucity of functional iTreg permitted unregulated CD4 Th cytokine stimulation of antibody secretion. iTreg isolated from peripheral blood mononuclear cells (PBMCs) of afflicted children were unable to inhibit secretion of interferon (IFN)γ by CD4 or CD8 T cells.12, 13 Evidence that polyclonal expansion of iTreg from PBMCs could partially overcome these deficiencies underscored the importance of iTreg in the pathogenesis of type 2 AIH and their potential therapeutic use.14
The study of Longhi et al.15 in this issue of Hepatology extends these findings in type 2 AIH by demonstrating the ability to produce high-affinity CYP2D6-antigen-specific Tregs from the peripheral blood of patients that suppress CYP2D6-specific T cell responses of both activated CD4 T cells and CD8 CTL more potently than polyclonal Treg.14 CYP2D6-antigen-specific iTreg cells exhibited the differentiated CD4+CD25+highCD45RO+highCD62L+highCD127low phenotype of functional iTreg cells. The iTreg cells with the highest TCR affinity for CYP2D6 peptide antigen-HLA class II complexes were also the most potent suppressors of target cell proliferation, cytokine secretion, and cytotoxic function. The investigators also showed convincingly that the CYP2D6-specific suppressor functions of these iTreg could be substantially increased by coculture with smDCs pulsed with CYP2D6 peptide antigens to enhance antigen processing, presentation, and bilateral cytokine production by the iTreg and smDC. By expressing high levels of MHC class I and II molecules with low levels of costimulatory CD80/CD86 the human smDCs reproduced the optimal conditions for antigen-specific induction of iTreg observed in mice.10, 11 Coculture of CYP2D6-pulsed smDCs and iTreg significantly decreased IFNγ-secreting cells, and pretreatment of iTreg with anti-IFNγ antibodies resulted in increased iTreg suppressor activity.
In accord with their prior observations that patients immunosuppressed with prednisolone and/or azathioprine mediated greater Treg suppressive activity in vitro, the present study showed that the efficiency of suppression of the smDC-Treg system was also superior in patients on immunosuppressive therapy. Also, in patients studied before and again during treatment with prednisolone/azathioprine, the suppressor function of CYP2D6 pulsed smDC and iTreg increased on therapy, suggesting that prednisolone/azathioprine may have enhanced the numbers and/or functions of circulating iTreg or smDC precursors. An effect of immunosuppression on smDC was provided by a study of patients with myasthenia gravis showing that prednisolone augmented iTreg function by down-regulating DC expression of costimulatory molecules and inhibiting DC maturation.16 Thus, successful treatment with corticosteroids/azathioprine might enhance the functions of both components of the smDC-Treg suppressor system.
The smDC/Treg suppressor system was also effective in suppressing the cytotoxic functions of CD8 CTL against CYP2D6 antigens. Of particular interest was the finding that smDC/Treg suppressed not only CD8 CTL cytotoxicity against target cells expressing the specific CYP2D6 antigenic epitopes used to pulse the smDC but also suppressed CD8 CTL cytotoxicity against targets expressing epitopes from other CYP2D6 regions. Such Treg “bystander” suppression represents an important attribute of the smDC/Treg suppressor system in type 2 AIH, in which not all CYP2D6 antigenic epitopes for the CD8 TCR repertoire are known and epitope-determinant spreading is expected.17
iTreg generated by coculture with CYP2D6-pulsed smDC expressed substantial quantities of CXCR3, a chemokine receptor associated with trafficking to sites of inflammation, including the liver.18 The fact that the iTreg specifically recognize CYP2D6 antigenic peptides could also aid in their trafficking to the liver, since antigen-specific T cells preferentially migrate to sites of antigen expression.19
Therapeutic adoptive transfer of autologous iTregs to suppress autoimmune effector cells in vivo represents the holy grail of studies of the ex vivo induction of antigen-specific iTreg.4, 6 However, both theoretical and practical obstacles must be overcome to make this goal a reality. Studies of dose-response relationships and duration of action, which will likely favor the use of antigen-specific iTregs rather than polyclonal Tregs, must be performed in experimental animal models to optimize conditions for survival and function of the adoptively transferred iTreg. Whether iTreg should be targeted to lymphoid compartments to abrogate generation of new autoimmune effector cells or directed to inflamed organs to inhibit activated effector cells causing immunopathology are important, unanswered questions. In particular, the fate and function of adoptively transferred iTreg in the liver must be determined because Kupffer cell expression of programmed death receptor-ligand-1 (PD-L1) mediates apoptosis of activated T cells and inactivates T cell functions.20 In long-established autoimmune diseases, expanded iTreg populations may not be as functional as desired because they may have been derived from remnant, “defective” Treg populations. In that case, a period of intense therapy targeting pathogenetic mechanisms of individual diseases might be required to restore iTreg precursors with sufficient functional capacities for ex vivo expansion. In some diseases, including type 1 and 2 AIH, non–antigen-specific mechanisms of chronic inflammation (Fig. 1) involving activated macrophages, neutrophils, T cells expressing T cell receptors comprised of γδ chains (Tγδ), NKT cells, and NK cells may be insensitive to iTreg control.21, 22 Despite these caveats and concerns, the remarkable progress in the generation and characterization of CYP2D6-antigen-specific iTreg cells bodes well for their ultimate introduction into therapeutic trials.