Decreased 4-1BB expression on CD4+CD25^high regulatory T cells in peripheral blood of patients with multiple sclerosis

Abstract

As a tumour necrosis factor receptor superfamily member, 4-1BB (CD137) is preferentially expressed in CD4+CD25+ regulatory T cells (Tregs) and has been suggested to play an important role in regulating the generation or function of Tregs. Recent studies of human Tregs have shown that blood CD4+CD25^high T cells were much closer to Tregs in terms of their functionality. Furthermore, CD4+CD25^high Tregs have been found to have a decreased effector function in patients with multiple sclerosis (MS). In this study, we examined the expression of 4-1BB and soluble 4-1BB (s4-1BB) protein levels in the peripheral blood of MS patients. Compared with healthy controls, MS patients had decreased 4-1BB expression in their CD4+C25^high Tregs and increased plasma s4-1BB protein levels. Moreover, the plasma s4-1BB levels of MS patients were shown to be inversely correlated with the 4-1BB surface expression of CD4+CD25^high Tregs. The down-regulated 4-1BB expression on CD4+CD25^high Tregs of MS patients may be involved in the impaired immunoactivity of these Tregs. The elevated s4-1BB levels may, at least in part, function as a self-regulatory attempt to inhibit antigen-driven proliferation of Tregs or their immunosuppressive activity.

Introduction

Multiple sclerosis (MS) is an inflammatory demyelinating disease of the central nervous system (CNS) causing a variable degree of axonal damage. Although the aetiology of MS remains unclear, it is generally considered that a T cell-mediated autoimmune response against CNS myelin underlines the pathogenesis of the disease [1,2]. Recent studies have provided strong evidence that besides CD4+ helper type 1 T cells, a variety of other immune cells such as B cells, CD8+ T cells, natural killer (NK) T cells and CD4+CD25+ regulatory T cells (Tregs) seems to be involved in disease pathogenesis by inducing or controlling the immune response in MS [3,4].

CD4+CD25+ Tregs, which normally reside in the thymus as natural Tregs, are believed to play a critical role in maintenance of immunological tolerance. Under normal physiological conditions, Tregs are generated by interaction with immature antigen-presenting cells, such as bone-marrow derived dendritic cells, in the periphery. They are normally anergic, but they can be activated by exposure to antigens or to high concentrations of interleukin-2 (IL-2). In addition, several other members of CD28 or tumor necrosis factor receptor superfamily (tumour necrosis factor receptor superfamily, TNFRSF) have been suggested to have certain roles in regulating Treg generation or function, including glucocorticoid-induced tumor necrosis factor receptor (GITR), cytotoxic T lymphocyte antigen-4 (CTLA-4), inducible costimulator, CD40 [5] and 4-1BB (CD137) [6–8].

In various studies involving human Tregs, it has recently been found that a population of CD4+CD25^high T cells consistently showed regulatory activity much closer to that described for murine cells in vitro, rather than the cells expressing medium to low levels of CD25 [9–11]. So far, various studies have reported that these well-defined Tregs were dysfunctional in MS. Although the frequency of CD4+CD25^high Tregs in the blood and cerebrospinal fluid (CSF) of MS patients was similar to that of healthy controls (HC) [12], the inhibitory effect on proliferation of effector T cells or myelin oligodendrocyte protein-reactive T cells by Tregs from MS patients was reduced compared with healthy individuals [12,13]. These findings suggest that the capacity of the cells to suppress the activity of encephalitogenic T cells could be impaired in MS. Interestingly, a recent study demonstrated a higher frequency of CD4+CD25^high Tregs in the peripheral blood of MS patients at relapse compared with HC [14]. Nevertheless, further studies are needed to elucidate the distribution of Tregs in MS and the possible role of 4-1BB in skewed Treg homeostasis during the inflammatory process. In the present study, we have analysed peripheral blood mononuclear cells (PBMC) for CD4+CD25^high Tregs as well as their 4-1BB expression in patients with this disease to address these issues.

Materials and methods

Subjects

Twenty patients with clinically definite MS (8 men and 12 women; mean age 47·8 ± 12·5 years) were included in this study. All the patients had stable relapsing-remitting MS (RRMS). None of the patients were treated with immunosuppressants, including corticosteroids, for at least 6 months before the time of sample collection. The Expanded Disability Status Scale (EDSS) score was used to reflect the disease severity at the time of blood sampling [15]. In addition, five of these RRMS patients receiving treatment with interferon-b1a (IFN-b1a, Rebif) were chosen for a serial study. Nineteen patients with stroke (6 men and 13 women; mean age 45·4 ± 8·5 years) were enrolled as other neurological disease (OND) controls. Twenty healthy volunteers were included as normal controls (9 men and 11 women; mean age 41·5 ± 9·1 years). All participating subjects gave their informed consent before the start of the study.

Sample collection

Heparinized blood specimens were collected between 09.00 and 11.00. For the serial study, blood specimens of the patients receiving treatment with IFN-β 1a were collected twice after 2 and 4 weeks of the therapy. PBMC were isolated by density gradient centrifugation on Lymphoprep (Nycomed, Oslo, Norway) using a standard protocol. After centrifugation, plasma was stored at −70°C in small aliquots and thawed just before further use.

Magnetic separation of CD4+CD25+ Treg

To examine the 4-1BB mRNA expression in CD4+CD25+ Tregs, freshly isolated PBMC were resuspended in PBS with 2% FCS, and incubated with magnetic beads coated with the monoclonal antibodies according to the supplier's directions (Dynal, Oslo, Norway). Positive cells were washed with buffer and the pellets were snap frozen in liquid nitrogen and stored at −70°C until further use.

Flow cytometry

Peripheral blood mononuclear cells of the patients and control subjects were characterized for the expression of 4-1BB and GITR by three-colour direct immunofluorescence and flow cytometry using a FACScan (Becton Dickinson, CA, USA). Cells were washed with PBS (pH 7·2) and resuspended in RPMI1640 medium. The following monoclonal antibodies were added to 2 × 10⁵ cells according to the recommendation of the manufacturer: Peridinin chlorophyll protein-labelled anti-IgG1, anti-CD4, phycoerythrin-labelled anti-IgG1, anti-4-1BB, anti-GITR, and fluorescence isothiocyanate-labelled anti-IgG1, anti-CD25 (Becton & Dickinson, CA, USA). After a 15-min incubation at 4°C, the cells were washed twice with staining buffer and analysed on the FACScan using the CellQuest software.

RNA preparation and cDNA synthesis

Total RNA was isolated with RNeasy® Mini Kit (Qiagen, Germany) following a standard protocol. Reverse transcription was performed with TaqMan RT reagents (Applied Biosystems, Foster City, CA, USA) using random hexamers as primers. The reaction was performed at 25°C for 10 min, 48°C for 30 min and 95°C for 5 min on a 9600 GeneAmp PCR system (Applied Biosystems, Norwalk, CT, USA).

Real-time PCR

4-1BB mRNA was quantified using cDNA-specific primers as described elsewhere [16]. For 4-1BB, the forward primer was 5′-TCA TTG CAG GAT CCT TGT AGT AAC-3′ and the reverse primer was 5′-GGC AGG TCC ACG GTC AAA G-3′. PCR reactions contained ∼25 ng of cDNA, 100 nM of forward and reverse primers, and a TaqMan universal PCR master mix (Applied Biosystems, Foster City, CA, USA). 18S was selected as the endogenous control as described elsewhere [17] and quantified using pre-developed assays from Applied Biosystems (CA, USA). Real-time PCR was performed on an ABI PRISM 7700 Sequence Detector. The PCR conditions for 4-1BB and 18S was one hold at 50°C for 2 min, one hold at 95°C for 10 min, and 40 cycles at 95°C for 15 s, 58°C 1 min, 72°C 45 s.

Quantification of soluble 4-1BB levels in plasma

Plasma soluble 4-1BB (s4-1BB) levels were measured by an enzyme-linked immunosorbent assay (ELISA) with a detection limit of 1 ng/ml. Briefly, microtiter plates (Corning Costar, NY, USA) were coated overnight with 100 µl/well of anti-human 4-1BB antibody (R&D Systems, UK) at 1µg/ml; non-specific binding was blocked by 300 µl of PBS containing 1% BSA, 5% sucrose and 0·05% NaN₃. One hundred µl of plasma or standards of recombinant human 4-1BB: Fc protein (Alexis Biochemicals, CA, USA) with a two-fold dilution series starting at 20 ng/ml were added in duplicate. One hundred µl of biotinylated anti-4-1BB antibody (Neomarker, CA, USA) at 0·5 µg/ml was added to each well, followed by the stepwise addition of 100 µl streptavidin horse-radish peroxidase (R&D Systems, UK) and 100 µl of substrate solution (R&D Systems, UK) to each well. 50 µl of stop solution (1 M H₂SO₄) was used per well to stop the reaction. The evaluation was performed at 450 nm in a microplate reader (Labsystems, Finland). All assays were performed simultaneously in a blinded fashion.

Statistics

The data are presented as mean ± standard deviation of the mean value (surface expression of 4-1BB and GITR) or median with range (s4-1BB,4-1BB mRNA). Data with a normal distribution were analysed with one-way analysis of variance (anova) with Student-Newman-Keul's post-hoc test or Pearson's correlation test. Data with a non-normal distribution were analysed with Kruskal-Wallis anova or Spearman's correlation test. A P value of less than 0·05 was considered to be statistically significant.

Results

Cell surface expression of 4-1BB and GITR

Multiple sclersis patients showed decreased expression of 4-1BB on CD4+CD25^high T cells (Fig. 1) compared with HC (P < 0·05, Table 1), but no significant difference of 4-1BB expression was found on CD4+CD25− T cells between MS and either of the control groups. On CD4+CD25^high T cells or CD4⁺CD25− T cells (Fig. 1), no difference was found between MS patients and the control groups with regard to GITR expression (Table 1). Additionally, there was nosignificant difference in the frequency of CD4+CD25^high T cells between the three groups. In five MS patients and five healthy individuals, we confirmed that all of the CD4⁺CD25^high T cells expressed surface markers of HLA-DR, CD45RO, CD62L and CTLA-4 (data not shown). Moreover, five MS patients who were treated with IFN-b1a demonstrated a trend towards a continuous increase in 4-1BB expression on CD4+CD25^high T cells after 2 and 4 weeks of treatment (Fig. 2A), while the CD4+CD25^high T cells frequency or their GITR expression presented slight or irregular changes (Fig. 2B-E).

Fig. 1.

Region 1 (R1) is selected to set CD4+CD25^high T cells gate for 4-1BB and GITR analysis. Control staining with isotype control antibodies was used as control to define the gate.

Table 1.

CD4+CD25^high and CD4+ CD25− T cells spontaneously expressing 4-1BB and GITR in the peripheral blood of patients with multiple sclerosis (MS), other neurological diseases (OND) and healthy controls (HC).

CD4+CD25high		4-1BB (%)		GITR (%)
	(%)	CD4+CD25high	CD4+CD25−	CD4+CD25high	CD4+CD25−
MS (n = 20)	1·16 ± 0·85	1·56 ± 1·47*	0·52 ± 0·47	5·73 ± 1·3	0·79 ± 0·34
OND (n = 19)	0·85 ± 0·48	1·82 ± 1·63	0·53 ± 0·49	5·23 ± 0·83	0·71 ± 0·33
HC (n = 20)	0·91 ± 0·72	3·05 ± 2·04	0·7 ± 0·66	5·09 ± 0·83	0·58 ± 0·26
P value (anova)	0·3617	0·0196	0·5133	0·1229	0·0949

^*P < 0·05 for post-hoc comparison with healthy controls.

Fig. 2.

Comparison of 4-1BB mRNA levels of blood CD4+CD25+ Treg between patients with MS, other neurological diseases (OND) and healthy controls (HC). Horizontal lines indicate median values.

Quantification of 4-1BB mRNA expression

In isolated CD4+CD25+ Tregs (Fig. 3), there was a lower 4-1BB mRNA expression in MS patients than that of HC (P < 0·05), but no significant difference was found between MS and OND patients.

Fig. 3.

Serial study of (A–E) CD4+CD25^high T cells as well as their GITR or 4-1BB surface expression in peripheral blood of five MS patients before treatment and after 2 (14 days) and 4 weeks (28 days) of treatment with IFN-b1a. 14 d = 14 days; 28 d = 28 days.

ELISA

The plasma s4-1BB levels were calculated using a standard curve. There was an increase of plasma s4-1BB levels in MS patients as compared with those in HC (P < 0·05; Fig. 4), but no differences were found between OND and HC groups. In addition, five MS patients who were treated with IFN-b1a showed a continuous decrease in plasma s4-1BB levels after 2 and 4 weeks of treatment (Fig. 5).

Fig. 4.

Comparison of plasma s4-1BB levels between patients with MS, other neurological diseases (OND) and healthy controls (HC). Horizontal lines indicate median values.

Fig. 5.

Serial study of plasma s4-1BB levels in five MS patients before treatment and after 2 (14 days) and 4 weeks (28 days) of treatment with IFN-b1a. 14 d = 14 days; 28 d = 28 days.

Correlation of 4-1BB surface expression with s4-1BB protein release

Neither 4-1BB surface expression on CD4+CD25^{−T cells}, nor the mRNA expression in CD4+CD25+ Tregs correlated with plasma s4-1BB levels in MS patients, except for a significantly inverse correlation for 4-1BB surface expression on the CD4+CD25^high T cells (r = −0·6028, P = 0·0104).

Correlation of CD4+CD25^high T cell frequency or s4-1BB levels and EDSS

There was no correlation between the s4-1BB levels or CD4+CD25^high T cell frequency and EDSS of MS patients.

Discussion

4-1BB is a TNFRSF member expressed by a variety of different cells, including activated T lymphocytes [18,19], NK cells [20], activated eosinophils [21], monocytes [22] and CD4+CD25+ Tregs [23,24]. Like CD28, 4-1BB functions as a potent costimulator to promote the activation, proliferation and survival of CD4+ and CD8+ T cells [25]. In an animal model of MS-experimental autoimmune encephalomyelitis (EAE), the use of an agonistic anti-4-1BB monoclonal antibody was reported to remarkably reduce the incidence and severity of EAE by increasing activation-induced cell death in CD4+ T cells [26]. Interestingly, 4-1BB is produced in both membrane-bound and soluble forms. The s4-1BB protein has been shown to be increased in the blood and CSF of MS patients compared with HC [27], although its exact mechanism remains elusive. These findings indicate that the 4-1BB may be involved in the inflammatory process of MS.

In our study, MS patients showed decreased 4-1BB expression on CD4+CD25^high Tregs, instead of CD4+CD25− T cells. This indicates that the dysregulation of this molecule in CD4+ T cells may predominantly occur in CD4+CD25+ Tregs, rather than CD4+CD25− responder T cells. Moreover, the decreased 4-1BB mRNA expression in CD4+CD25+ Tregs of MS patients further substantiates our findings, as one previous study showed that isolated CD4+CD25+ T cells through magnetic cell separation contained about 50% of CD4+CD25^high T cells and demonstrated a suppressive activity compared with that of CD4+CD25+ Tregs purified by fluorescence-activated cell sorting [28]. In line with our results, a recent study reported that MS patients had decreased surface expression of 4-1BB ligand (4-1BBL) on blood plasmacytoid dendritic cells (pDC), which is considered to potentially influence the ability of the pDC to interact with the immune system, and particularly to generate regulatory cells [29]. Taking into consideration all these findings, this suggests that there is an aberrant 4-1BB co-stimulation existing in Tregs of MS patients.

4-1BB was preferentially expressed in CD4+CD25+ Treg as compared with non-regulatory CD4+ T cells. Various studies have shown that 4-1BB co-stimulation induced the proliferation of CD4+CD25+ Tregs in vivo and in vitro[6,8,30]. In our serial study, although there was non-significant dynamic change of CD4+CD25^high Treg frequency in the MS patients, 4-1BB expression on the cells tended to increase after 2 and 4 weeks of treatment with IFN-b1a. This is in parallel with a very recent MS study of a remarkably enhanced immunosuppressive function of CD4+CD25+ Tregs after 3 and 6 months of IFNβ-1a therapy [31], indicating a positive role that 4-1BB could play in regulating immunoactivity of CD4+CD25^high Tregs. However, further serial studies are needed with non-treated MS patients to confirm this finding. Nevertheless, our results strongly suggest that down-regulated 4-1BB expression on CD4+CD25^high Tregs may be involved in impairing their immunosuppressive capability.

The s4-1BB is released by activated lymphocytes and is generated by proteolytic cleavage from the cell surface. The s4-1BB level is inversely correlated with lymphocytes proliferation, and positively reflects the degree of activation-induced cell death caused by mitogen stimulation [32]. The functional significance of elevated s4-1BB levels in inflammatory diseases such as MS and rheumatoid arthritis (RA) remains unclear [27,33,34], but it has been suggested to act as a negative feedback control of the ongoing inflammation [32]. Hence, we speculate that the increased s4-1BB release in MS patients may be related to the immunoregulation of CD4+CD25+ Tregs due to the preferential expression of 4-1BB in these cells, rather than the CD4+CD25− responder T cells. Indeed, our study demonstrated an inverse correlation between the plasma s4-1BB levels and 4-1BB surface expression on CD4+CD25^high Tregs. Furthermore, we found that plasma s4-1BB levels of MS patients tended to decrease following treatment, in agreement with another study that showed that elevated serum s4-1BB levels of untreated RA patients rapidly decreased after receiving immunosuppressive treatment [34]. These results indicate that increased s4-1BB levels, probably due to more release from activated lymphocytes including CD4+CD25^high Treg, may at least partially function as a self-regulatory attempt to inhibit antigen-driven proliferation of Tregs or their immunosuppressive function in the process of MS.

As a transmembrane protein, GITR is expressed at low levels in resting responder T cells [35–37], but at high levels in activated T cells [35–41] and CD4+CD25+ Tregs [24,35]. Generally, it is considered that GITR triggering could have at least five distinct effects: (i) costimulation and activation of effector T cells; (ii) inhibition of Treg activity by down-regulating molecules important for suppression; (iii) decrease of the specific sensitivity of effector T cells to Treg-mediated suppression; (iv) partial deletion of Tregs; and (v) potential promotion of Tregs proliferation and increase of suppression [42]. In our study, the MS patients, however, showed a slight increase of GITR surface expression on CD4+CD25^high Treg compared with HC. In addition, there were only irregular changes of GITR expression on CD4+CD25^high Tregs after 2 and 4 weeks of IFN-β1b therapy, in contrast to a recent study that observed a trend towards increasing proportions of CD4+CD25+GITR+ T cells after 6 months of IFN-β1b treatment as compared with those before treatment [31]. These findings suggest that GITR may have a complex or fine regulatory effect on the immunoactivity of Tregs in different stages of MS. However, this issue needs to be further ascertained.

So far, the impaired function of peripheral blood CD4+CD25^high Tregs has been reported in patients with MS, type I diabetes, psoriasis, autoimmune myasthenia gravis, RA and active systemic lupus erythematosus [43–48], implicating a common mechanism of Treg defect involved in the autoimmune diseases. However, the cellular defect seems to play a different or complex role in the pathogenesis of autoimmune diseases, as in a large population-based study, no excess of common autoimmune diseases were identified in MS patients or their families, including multicase pedigrees [49]. We speculate that the Treg dysfunction may exert different effects on the development of each autoimmune disease with its particular anatomical predilection, because human CD4+CD25^high Tregs are heterogeneous and each cell subpopulation demonstrates distinct mechanisms of suppression. Moreover, the ability of the Tregs to suppress CD4+CD25−T responder cells depends on the situation of inflammatory environments, where the CD4+CD25− T responder cells are activated with different signal strengths [50]. On the other hand, difference in genetics may also influence the effector outcome of Tregs in these autoimmune diseases, as most of the human autoimmune diseases are complex genetic disorders comprised of multiple common allelic variants that can, in combination with environmental factors, lead to development of a pathologic response. In genome-wide studies with enough power to detect small effects, the human leukocyte antigen (HLA) has consistently been the only clear-cut locus linked to MS [51,52], although one non-HLA gene, IL-7 receptor α chain (IL7R), has been recently reported to be associated with MS by several groups [52–54]. Most non-HLA genes linked to the pathogenesis of common autoimmune diseases, however, showed no consistent association with MS [55–58]. Nevertheless, more evidence is needed to answer this question.

In conclusion, this study shows lower 4-1BB expression in CD4+CD25^high Tregs of MS patients than those of HC. Our results strongly suggest that the down-regulation of 4-1BB expression on CD4+CD25^high Treg may be involved in impaired immunoactivity of Tregs, while the elevated s4-1BB levels may, at least in part, function as a self-regulatory attempt to inhibit antigen-driven proliferation of Tregs or their immunosuppressive activity. Further studies of this costimulatory molecule in various subtypes of MS and after treatment of the disease are needed to gain better insight into the significance and potential therapeutic implications of our findings.