Functional defect in regulatory T cells in myasthenia gravis

Abstract

Forkhead box P3 (FOXP3)⁺ is a transcription factor necessary for the function of regulatory T cells (T_reg cells). T_reg cells maintain immune homeostasis and self-tolerance, and play an important role in the prevention of autoimmune disease. Here, we discuss the role of T_reg cells in the pathogenesis of myasthenia gravis (MG) and review evidence indicating that a significant defect in T_reg cell in vitro suppressive function exists in MG patients, without an alteration in circulating frequency. This functional defect is associated with a reduced expression of key functional molecules such as FOXP3 on isolated T_reg cells and appears to be more pronounced in immunosuppression-naive MG patients. In vitro administration of granulocyte-macrophage colony-stimulating factor (GM-CSF) enhanced the suppressive function of T_reg cells and up-regulated FOXP3 expression. These findings indicate a clinically relevant T_reg cell–intrinsic defect in immune regulation in MG that may reveal a novel therapeutic target.

Introduction

Acquired myasthenia gravis (MG) is an autoimmune disorder of the neuromuscular junction in which patients experience fluctuating skeletal muscle weakness that often affects selected muscle groups preferentially. The target of the autoimmune attack in most cases is the skeletal muscle acetylcholine receptor (AChR), but in others, non-AChR components of the neuromuscular junction, such as the muscle-specific receptor tyrosine kinase, are targeted (reviewed in Ref. 1). The precise origin of the autoimmune response in MG is unknown, but abnormalities of the thymus gland (hyperplasia and neoplasia) almost certainly play a role in patients with anti-AChR antibodies^(2,3) and genetic predisposition is also likely to influence which patients develop the disorder.^(4–6) High-affinity, anti-AChR antibodies bind to the neuromuscular junction, activate complement and accelerate AChR destruction, while causing failure of neuromuscular transmission and the resulting myasthenic symptoms (reviewed in Ref. 1). However, auto reactive AChR-specific CD4⁺ T cells, which can be detected in most MG patients,^(7–10) likely play an important role in MG, through cognate interactions with B cells leading to and the synthesis of anti-AChR antibodies.⁽¹¹⁾

Regulatory T cells and immune tolerance

Immune tolerance to self antigens is initially achieved during thymic development by the clonal deletion of potentially autoreactive T cells. However, some of these pathogenic cells, including some with reactivity to skeletal muscle AChR, survive clonal deletion in normal individuals, and are kept in check by peripheral tolerance mechanisms, most notably by a specialized subset of CD4⁺ T cells called regulatory T cells (T_reg cells).⁽¹²⁾ T_reg cells are a subpopulation of T cells that act to suppress activation of other immune cells and thereby maintain immune system homeostasis and self-tolerance.⁽¹³⁾ Current theories regarding the pathogenesis of autoimmune disease hypothesize that a functional deficiency in T_reg cells could result in a failure to suppress autoreactive T cell responses.^(14,15)

The thymus gland is the primary source of T_reg cells which constitute approximately 5–10% of the peripheral CD4⁺ T cell population, and play a crucial role in the maintenance of immune homeostasis.⁽¹³⁾ T_reg cells actively mediate self-tolerance and thus control autoimmunity by suppressing the activation, proliferation and cytokine production of effector autoreactive T cells that arise de novo or escape thymic deletion.^(13,16) The transcription factor forkhead box protein P3 (FOXP3) is expressed in CD4⁺ CD25⁺ T_reg cells and is the master regulator for the development and function of these cells. ^(17,18) FOXP3 gene transfer has been shown to convert naive CD4⁺ CD25⁻ T cells into a functional regulatory lymphocyte population, demonstrating the pivotal role of FOXP3 in T_reg cell biology.⁽¹⁸⁾

A deficiency or dysfunction of T_reg cells has been shown to contribute to the pathogenesis and development of many autoimmune diseases.^(19–22) Functional defects have been demonstrated in T_reg cells isolated from patients with diverse autoimmune diseases, suggesting that in autoimmunity, T_reg cell suppressive capacity may be diminished. It has also recently been reported that the source of the apparent T_reg cell dysfunction may not necessarily be primarily intrinsic to T_reg cells, and that effector T cells may acquire resistance to T_reg cell suppression in certain autoimmune diseases. ^{(19, 23–25)}

Regulatory T cells and immune regulation in myasthenia gravis

While it is likely that T_reg cells exert their effects on T helper cells, thereby regulating B cell function and antibody secretion, there is also evidence that T_reg cells may have a direct effect on B cells.⁽²⁶⁾ Specifically, in MG, it may be hypothesized that dysfunctional T_reg cells permit the production of anti-AChR autoantibodies, a hypothesis that remains to be specifically tested. Investigators examining the relative frequencies and function of T_reg cells in the peripheral circulation and thymi of MG patients have reported conflicting results including reductions in T_reg cell numbers,^(27–29) impaired regulatory function,⁽³⁰⁾ or no defect^.(31,32) Aside from measuring T_reg cell numbers, and ability to suppress non-specific T cell proliferation, more detailed analyses of T_reg cell– intrinsic (T_reg phenotype, cytokine production, and suppressive function) versus extrinsic factors (T helper/effector cells, dendritic cells), and ability to suppress specific AChR-induced T proliferation have not been performed. The reported demonstration of a functional impairment of thymic T_reg cells in MG patients is an interesting finding and suggests an origin for T_reg cell dysfunction in the thymus, possibly owing to thymic pathology (hyperplasia or neoplasia). ⁽³³⁾

Importantly, in all of the studies referenced above, investigators utilized a single-step enrichment protocol in which T_reg cells were identified and isolated by high surface expression of CD25. Unfortunately, the level of expression of CD25 that defines T_reg cells has been inconsistently reported in the literature, and CD25 is also expressed by recently activated T cells, resulting in the possible inclusion of effector T cells in isolated T_reg cell populations.⁽³⁴⁾ More recent studies have shown that T_reg cells express low or absent levels of the IL-7 receptor α-chain (CD127) and that this expression is inversely correlated with FOXP3 expression and with suppressive function,^(35,36) and that it may provide a more precise method for the isolation of a pure T_reg cell population.

Impaired suppressive function of isolated MG T_reg cells

We have investigated the existence of a functional defect in CD4⁺ CD25^high CD127^low/− T_reg cells in MG patients by isolating this population of cells using flow-based sorting and co-culturing them with autologous CD4⁺ CD25⁻ T responder cells (T_resp) at different T_reg:T_resp ratios, in the presence of irradiated autologous antigen presenting cells (APCs). We performed T cell proliferation studies (CFSE dilution) stimulating the cell cultures with anti-CD3 Abs and analyzed cytokine profiles from cell culture supernatants. We found that the suppressive ability of isolated T_reg cells from MG patients (MG T_reg cells) was consistently reduced compared to age-matched controls analyzed concurrently.⁽³⁷⁾ As expected, T_resp cells from MG patients produced higher levels of pro-inflammatory cytokines (IL-6 and IFN-γ) and lower levels of IL-10, compared with controls. Furthermore, when isolated CD4⁺ CD25^high CD127^low/− cells (T_reg cells) were added to the cultures, IL-6 and IL-17 levels actually increased, without altering IL-10 levels in MG patients.⁽³⁷⁾ This result suggests not only a functional inability of MG T_reg cells to suppress the production of pro-inflammatory cytokines, but also the possible production of IL-6 and IL-17 by CD4⁺ CD25^high CD127^low/− cells in MG. Along these lines, it has been recently reported that peripheral T_reg cells may differentiate into IL-17–producing cells upon T cell receptor stimulation in the presence of IL-6.⁽³⁸⁾ Thus, enhanced T_reg cell plasticity in MG may underlie the observed disturbance in T_reg cell suppressive function in MG, through functional conversion of cells with a T_reg cell phenotype into pathogenic Th17 cells. This phenomenon may be associated with the level of FOXP3 expression within a T cell (see below).

Dysfunctional MG T_reg cells express attenuated FOXP3 levels

The level of expression of FOXP3 plays a critical role in the development and function of T_reg cells. A number of factors, some positive, some negative, interact to collectively drive FOXP3 gene expression and then maintain its expression in T_reg cells. Reduced FOXP3 expression in circulating CD4⁺ CD25⁺ T cells^(33,39) and CD4⁺ CD25⁺ thymocytes⁽³³⁾ has been observed in MG patients. We isolated CD4⁺ CD25^high CD127^low/− T_reg cells from peripheral blood mononuclear cells (PBMCs) and found no alteration in the reservoir of T_reg cells in the peripheral blood of MG patients compared to controls. We confirmed that these cells were largely (>90%) FOXP3-expressing cells, by flow cytometry–based phenotypic analysis, and were highly suppressive in functional suppressor assays.⁽³⁷⁾ However, we observed a significant down-regulation of the relative expression of FOXP3 at the protein and mRNA levels in isolated CD4⁺ CD25^high CD127^low/− cells in MG patients, compared with healthy controls. ⁽³⁷⁾

We went on to evaluate the expression of FOXP3-associated proteins including those responsible for key features of Treg function such as cytotoxic T lymphocyte antigen-4 (CTLA-4). We have shown that a significantly lower percentage of isolated T_reg cells from MG patients express CTLA-4,⁽³⁷⁾ consistent with the impairment in suppressive function when these cells are used in suppressive assays. MHC II expression on human T_reg cells has been reported to identify a distinct population of T_reg cells with high FOXP3 expression and with the capability to mediate strong contact-dependent suppression.⁽⁴⁰⁾ HLA-DR expressing T_reg cells have also been hypothesized to be involved in homeostatic maintenance of T_reg cells in vivo via presentation of self- antigens,⁽⁴⁰⁾ so that a deficiency in this subset of T_reg cells may impact on tolerance to self-antigens like the AChR. Accordingly, we have found that a reduced number of CD4⁺ CD25^high CD127^low/− cells express HLA-DR in MG patients.⁽³⁷⁾

It should be pointed out that the demonstrated functional defect in T_reg cells that has been reported in MG to date appears to be a global one (non-specific) since there is a failure to normally suppress T cell proliferation in response to non-specific T cell receptor engagement (anti-CD3 antibodies). One may hypothesize that T_reg cell suppression of AChR-specific T cells is perturbed in MG, and that restoration of function in this population of T_reg cells may be a focused strategy for treatment. The first step to address this possibility would be to investigate the question of whether T_reg cells from MG patients have a reduced ability to suppress AChR-reactive T cells. While the number of circulating AChR-specific T cells is obviously low, previous investigators have identified in many patients with MG myasthenogenic AChR peptides that induce T cell proliferation.⁽⁴¹⁾ Based on this work, we synthesized two peptides representing sequences of the human AChR-α subunit (p195–212 and p257–259), in which at least one of the two peptides have been determined to induce proliferation of peripheral blood lymphocytes in approximately 70% of patients with MG.⁽⁴¹⁾ These peptides were used in suppression assay performed as described above, using AChR peptide rather than anti-CD3 stimulation. To assess T_reg cell suppression of AChR-stimulated T cell proliferation, previously activated CD4⁺ CD25^high CD127^low/− cells were added to the culture. T cell proliferation of 4.2–15% in response to one of the two AChR peptides was observed, none of which were significantly suppressed by the addition of T_reg cells.⁽³⁷⁾ Little or no proliferation in response to the AChR-α peptide was seen in the age-matched healthy control population. These studies indicate that the observed defect in suppressive function of T_reg cells in MG extends to AChR-specific T cell responses.

Defects in in vitro T_reg cell suppression in autoimmune disease may potentially be explained by T_reg cell–intrinsic defects, resistance of T_resp cells, or the pro-inflammatory properties of antigen presenting cells (APCs).^(23–25) Our data showing reduced cellular expression of FOXP3 in MG T_reg cells argues for a T_reg cell–intrinsic defect. To further test this, we have co-cultured MG T_reg cells from individuals with T_resp from healthy controls, and have also performed mirror image experiments, determining the degree of in vitro suppression of both of polyclonal and AChR-stimulated T cell proliferation. Both polyclonal and AChR-activated T_resp cells from MG patients could be effectively suppressed using T_reg cells isolated from healthy controls, while polyclonal-activated T_resp cells from controls were not effectively suppressed using T_reg cells isolated from MG patients, which strongly suggests ₍₃₇₎ a primary intrinsic defect in isolated T_reg cells in MG.⁽³⁷⁾

Imbalanced homeostatic composition of circulating T_reg cells in MG

While the existing data suggests that the relative numbers of circulating T_reg cells is unchanged in MG patients, the observed T_reg cell dysfunction may be linked to an imbalanced homeostatic composition of circulating T_reg cell subsets. There are two major types of CD4⁺ FOXP3⁺ T_reg cells.⁽⁴²⁾ Natural T_reg cells (nT_reg cells) constitute a stable subset derived from the thymus and are thought to control reactivity to self-antigens. Induced or adaptive T_reg cells (iT_reg cells) are less stable, being derived from CD4⁺ CD25⁻ T cells in the periphery upon antigen exposure. There is currently no good method for isolating viable nT_reg cells for functional studies, since the most reliable method relies on the presence of an epigenetic signature. However, cells recently exiting the thymus, including nT_reg cells, may be identified by expression of CD45RA and CD31 (recent thymic emigrants). In addition, important biological differences between naive (CD45RA⁺) and memory (CD45RO⁺) CD4⁺ CD25^high T_reg cells have been described.^(43,44) Differences in T_reg cell subsets in patients with autoimmune disease (SLE, T1D) compared to controls have also been reported when CD45RA is used to differentiate T_reg cell subsets,⁽⁴⁵⁾ suggesting that alterations in the relative composition of T_reg cell subsets may underlie autoimmunity. Therefore, we have analyzed the frequency of circulating recent thymic emigrant (CD31⁺) and naive (CD45RA⁺) CD4⁺ CD25^high CD127^low/− T_reg cells in MG patients. We have found that the prevalence of recent thymic emigrant and naive T_reg cells is significantly decreased in MG patients compared to healthy controls. ⁽³⁷⁾

This finding raises the question of whether T_reg cells may exhibit multiple differences from healthy controls that are likely to play roles in the complex pathogenesis of MG, including reduced thymic output of naive T_reg cells, impaired naive T_reg cell suppression, and distinct impairments in specific subsets of memory T_reg cells. The reduced percentage of circulating naive and RTE T_reg cells is particularly intriguing given the high incidence of thymic pathology (hyperplasia and neoplasia) in MG. Further studies correlating T_reg cell subset phenotype and function with thymic pathology will help to better characterize the relationship of circulating T_reg cell alterations and thymic T_reg cell development.

Are CD4⁺ CD25^high CD127^low/− T_reg cells affected by immunotherapy?

Most MG patients in our study were treated with prednisone or other immunosuppressive drugs, and it is conceivable that these drugs may have an effect on T_reg cell function. Subset analyses of T_reg cell suppressive function and FOXP3 expression in MG patients treated with immunosuppressants (IS) compared to those that were immunsuppression-naive at the time of blood collection, showed that T_reg cells from MG patients treated with IS more effectively suppressed T cell proliferation and expressed higher levels of FOXP3 (Fig. 1). In addition, a correlation was also found between manual muscle testing (MMT) scores and T_reg cell suppressive function.⁽³⁷⁾ These findings agree with those of other investigators,^(46,47) and suggest that the observed T_reg cell defects are clinically relevant and that the beneficial effects of IS therapy are at least partly mediated through enhancement of T_reg cell function.

Figure 1.

FOXP3 expression and suppressive function of T_reg cells in treated versus untreated MG patients. Intensity of FOXP3 expression (MFI = mean fluorescent intensity) in CD4⁺ CD25^high CD127^low/− T_reg cells. (A) Representative data for immunosuppressive naive (MG - IS naive) and immunosuppressed (MG + IS) subjects. (B) MFI of FOXP3 expression in isolated CD4⁺ CD25^high CD127^low/− cells. (C) Percent suppression of T_resp cell proliferation by T_reg cells. Data in (B) and (C) represent n = 5 for MG-IS naive subjects and n = 12 for MG-IS subjects. Results are expressed as mean ± SEM.

GM-CSF enhances T_reg cell function in MG

The expansion of T_reg cells for cellular therapy is a common experimental strategy applied to a number of autoimmune conditions.^(48,49) Unfortunately, expansion of dysfunctional MG T_reg cells is unlikely to provide a therapeutic benefit, and the identification of agents that enhance the function of T_reg cells would be a more appropriate strategy. GM-CSF is an important hematopoietic growth factor and immune modulator that has a profound effect on various circulating immune cells.⁽⁵⁰⁾ In vivo administration of GM-CSF has been shown to prevent or attenuate autoimmunity in a number of mouse models of autoimmune disease by expanding dendritic cells and inducing an expansion of T_reg cells.^(51–55) Furthermore, we have used the murine experimental model of MG to demonstrate that the therapeutic administration of GM-CSF produces clinical improvement, an expansion of the numbers of splenic FOXP3⁺ cells, a suppression of AChR-induced T cell proliferation, diminished production of anti-AChR antibodies, and a reduction of the deposition of IgG and complement at the muscle endplate, resulting in preservation of the normal postsynaptic endplate morphology and functional AChRs.^(53,54) More importantly, adoptively transferred GM-CSF-induced Tregs potently suppressed ongoing MG in recipient mice, confirming their critical role in GM-CSF’s effects.⁽⁵⁵⁾

Based on these preclinical studies, we have investigated the effects of GM-CSF treatment in a case of refractory myasthenic crisis.⁽⁵⁶⁾ Clinical improvement in this case was associated with expansion in the circulating numbers of FOXP3⁺ cells, increase in FOXP3 expression levels in CD4⁺ CD25^high CD127^low/− T_reg cells, early improvement in T_reg cell suppressive capacity for AChR-α–induced T cell proliferation, and subsequent enhancement in T_reg cell suppression of polyclonal T cell proliferation.⁽⁵⁶⁾ Although the use of multiple immunosuppressive drugs may have contributed to these findings, the observed effects on T_reg cells are consistent with preclinical studies.

We have also performed studies examining the in vitro effects of GM-CSF in T cell proliferation/suppression assays from MG patients and healthy controls. Total CD4⁺ and/or CD4⁺ CD25^high CD127^low/− T_reg cells depleted CD4⁺ T cells were cultured with autologus APCs in the presence of anti-CD3 and or anti-CD3 plus GM-CSF. We have found that the addition of GM-CSF to lymphocyte cultures ameliorates the suppressive defect in MG T_reg cells (Fig. 2A and 2B), and markedly enhances expression of FOXP3 in MG T_reg cells at the mRNA and protein level (Fig. 2C and 2D). This suggests that GM-CSF may be effectively used as an enhancer of T_reg cell suppressive function, as has been reported for a number of other agents (i.e., rapamycin, vitamin D analogues).^(57,58) Additional studies are needed to confirm these findings, define the requirements for GM- CSF’s in vitro effects, and also determine its actions on non-T_reg cells.

Figure 2.

In vitro GM-CSF treatment to lymphocytes enhances FOXP3 expression and suppressive function of T_reg cells in MG. CFSE-labeled total CD4⁺ T cells and T_reg cell–depleted CD4⁺ T cells were stimulated with human anti-CD3 alone or anti-CD3 plus GM-CSF in the presence of autologus irradiated APCs. Cultures consisting of total CD4⁺ cells (black bars) and CD4⁺ cells after removal of CD4⁺ CD25^high CD127^low/− cells (grey bars), were compared to evaluate suppressive function of T_reg cells. After 5 days of culture, percentage of T cell proliferation was analyzed based on CFSE dilution. The bar diagram represents percentage T_resp cell proliferation in response to anti-CD3 alone (A) and anit-CD3 plus GM-CSF (B) for four MG patients and four healthy controls.(C) FOXP3 mRNA expression was determined by multiplex PCR using isolated CD4⁺ CD25^high CD127^low/−T_reg cells obtained from a control subject, MG patient and an MG patient’s PBMCs treated with GM-CSF. Results are expressed as relative FOXP3 mRNA expression. Results are expressed as mean ± SEM. (D) Representative flow cytometry plots illustrate mean florescent intensity of FOXP3 expression within isolated CD4⁺ CD25^high CD127^low/− T_reg cells.

Conclusions

Defective immune regulation is present in patients with MG, characterized by defects in suppressive function and expression of FOXP3 in CD4⁺ CD25^high CD127^low/− cells. T_reg cell–mediated suppression of both polyclonal and autoantigen (AChR)-specific T cell responses is impaired and may correlate with disease severity, being notably less severe in the subgroup of patients treated with immunosuppressive drugs. This defect primarily resides within the isolated population of CD4⁺ CD25^high CD127^low/− cells which display reduced or unstable FOXP3 expression. While it has been generally thought that FOXP3 expression serves as an on-off switch endowing T lymphocytes with suppressive ability, emerging evidence, including our own work showing an association between attenuated FOXP3 expression in T_reg cells and human MG, suggests a paradigm in which alterations in FOXP3 expression in lymphocytes may lead to loss of immune tolerance in a dose-dependent manner (Fig. 3). Furthermore, GM-CSF has a profound in vitro beneficial effect on T_reg cell function as well as FOXP3 expression, and further investigations of the mechanisms responsible for these effects are warranted.

Figure 3.

Hypothetical model linking T_reg cell suppressive function to FOXP3 expression level. In this model, high expression of FOXP3 in T cells endows them with enhanced suppressive capacity and a characteristic phenotype. Absence of FOXP3 expression is associated with no suppressive capacity, but reduced FOXP3 expression may result in an intermediate phenotype without suppressive function, and possibly a tendency for enhanced functional plasticity.

In sum, investigation of the nature and subset localization of the T_reg cell defect in MG may reveal critical factors that regulate immune responses in MG and other autoimmune diseases, leading to the identification of novel therapeutic targets.