Beneficial effect of co-polymer 1 on cytokine production by CD4 T cells in multiple sclerosis

Abstract

Multiple sclerosis (MS) has been associated with an imbalance in the T helper type 1 (Th1) and Th2 subsets. We investigated, at the single-cell level, the synthesis of pro-inflammatory cytokines by CD4 and CD8 T cells from MS patients. We report the relationship between priming of CD4 and CD8 T cells for interleukin-2 (IL-2), interferon-γ (IFN-γ) and tumour necrosis factor-α (TNF-α) and disease evolution in MS patients, clinically subdivided into relapsing–remitting MS (RRMS) in remission, RRMS in relapse, or chronic progressive MS (CPMS). Moreover, we report the in vivo influence of co-polymer 1 (COP) treatment on the pattern of cytokine producers in RRMS patients. We show that the frequency of CD4 T cells primed for TNF-α synthesis increased in all stages of MS, including RRMS remitting, and was normalized to control values in COP-treated patients (43·2 ± 11·8% in treated patients versus 47 ± 7·3% in RRMS remitting versus 40·3 ± 8% in controls). In addition, a significant decrease in the frequency of CD4 T cells primed for IL-2 was found in COP-treated patients as compared to the other groups of patients, reaching values below that of controls (59·1 ± 9·9% in treated patients versus 70 ± 11·6% in RRMS remitting versus 67·1 ± 7·4% in controls). Unexpectedly, COP-treated patients also showed a significantly decreased priming for IFN-γ at the CD4 T-cell level (9·1 ± 3·4% in treated patients versus 18·8 ± 0·6.4% in RRMS remitting versus 15·4 ± 4·7% in controls), but not at the CD8 T-cell level. This bystander suppression on the inflammatory cells should be considered in the monitoring of MS patients submitted to COP treatment, in order to evaluate better its clinical efficacy.

Introduction

Multiple sclerosis (MS), a frequent cause of neurological disability in young adults, is an inflammatory disease of the central nervous system (CNS) that affects primarily the CNS myelin and the oligodendrocytes, and may lead to axonal degeneration.^1,2 The cause and the pathogenesis of MS are still unknown, although it has been postulated that several immunopathological mechanisms are involved.³ In particular, clinical and biological findings suggest that MS is mediated by macrophages, T cells and T-cell-derived cytokines, which may promote or limit the disease.^4,5 Indeed, inflammatory cytokines, such as interleukin-1 (IL-1), IL-6 and tumour necrosis factor-α (TNF-α), are detected in the CNS of animals with experimental autoimmune encephalomyelitis (EAE) and in patients with MS.^5,6 Myelin-reactive CD4 cells of the T helper type 1 (Th1) subtype are capable of adoptively transferring EAE to näve animals,^7–9 whereas neuroantigen-specific Th2 cells are not,¹⁰ and the relapsing form of EAE is associated with the absence of up-regulation of Th2 cytokines in the CNS.¹¹ Moreover, EAE can be suppressed by a recently identified CD4 Th subset (Th3), producing mainly transforming growth factor-β (TGF-β), and down-regulating Th1 cells.¹² The pro-inflammatory cytokine TNF-α, produced by both Th1 and Th2 cell subsets, is also involved in the pathogenesis of MS and EAE. Indeed, TNF-α is up-regulated in mononuclear cells of MS patients,¹³ it is detected in MS plaques¹⁴ and increased TNF-α production has been associated with the clinical activity of MS.^15–17 In addition, TNF-α has been shown to express a cytotoxic activity on oligodendrocytes in vitro.¹⁸ However, a protective role of TNF-α in vivo has also been reported. For example, treatment of chronic progressive MS patients with anti-TNF-α monoclonal antibody increased the clinical activity^19,20 and a potent anti-inflammatory activity of TNF-α was found in EAE.²¹ IL-2 and interferon-γ (IFN-γ) are also associated with MS, although their precise role remains to be elucidated. IL-2 and soluble IL-2 receptors have been found in the serum and CSF of MS patients²² and a significant positive correlation was reported between the proportion of IL-2-secreting peripheral blood mononuclear cells (PBMC) and disease activity.²³ The relation between IFN-γ production and disease activity remains unclear, since some studies did not find increased IFN-γ production by patients' PBMC before exacerbation of MS,²⁴ while others reported a small correlation between IFN-γ secretion by blood T cells and disease activity.²⁵ However, a protective role of IFN-γ is suggested by the finding that IFN-γ knockout mice are susceptible to the induction of EAE and IFN-γ treatment confers resistance to EAE on normal mice.^26,27

Co-polymer 1 (COP, glatiramer acetate, Copaxone) is a standardized mixture of synthetic polypeptides (poly Y, E, A, K), which is effective in suppressing EAE and has beneficial effects on the clinical course and magnetic resonance imaging (MRI)-defined brain lesions of patients with MS.²⁸ On the basis of in vitro and in vivo studies in EAE, it has been proposed that COP acts by two basic mechanisms: first, by competition with the encephalitogenic proteins proteolipid protein (PLP),²⁹ myelin oligodendrocyte glycoprotein (MOG)³⁰ and myelin basic protein (MBP)³¹ at the major histocompatibility complex (MHC) and T-cell antigen receptor (TCR) level; and, second, by induction of Th2 regulatory T cells in both EAE mice^32–34 and MS patients.³⁵ To evaluate whether a specific commitment to type-1/pro-inflammatory cytokine synthesis is found at the T-cell level in MS patients, and to determine the influence of COP on the cytokine profile, we have performed a single cell analysis of IL-2, IFN-γ and TNF-α synthesis by peripheral blood CD4 and CD8 T cells from MS patients. We report for the first time the relationship between priming of CD4 and CD8 T cells for synthesis of these pro-inflammatory cytokines and disease evolution in patients clinically subdivided into relapsing–remitting MS (RRMS) in remission, RRMS in relapse, or chronic progressive MS (CPMS). Moreover, the in vivo influence of COP treatment on the pattern of cytokine producers was evaluated in a subgroup of RRMS patients.

Materials and methods

Patients

This study included 61 patients diagnosed as having definite MS according to Poser criteria.³⁶ The clinical characteristics of patients are summarized in Table 1. These patients were divided into relapsing–remitting multiple sclerosis in remission (RRMS-remitting), or in relapse (RRMS-relapsing) and secondary chronic progressive multiple sclerosis (CPMS). Some patients receiving COP (Teva Pharmaceutical Industries, Kfar Sava, Israel, 20 mg subcutaneously daily) for at least 1 year were also analysed. To be eligible, MS patients should have received no steroids or other drugs at least during the last 3 months, and should have no infection at the time of blood drawing. The disability score according to Kurtzke's expanded disability status scale (EDSS)³⁷ ranged from 0 to 7·5. Healthy donors without allergy or inflammatory or autoimmune disease at the time of blood drawing were analysed for comparison.

Table 1.

Clinical characteristics of the healthy donors and MS patients studied

Clinical category	n	Age (years)	EDSS	Sex ratio
Control	13	23–56; 36	0	1·5
RRMS remitting	19	24–54; 35	[0–3]; 1·4	3·5
RRMS relapsing	12	22–43; 34	[2–3·5]; 2·8	4·3
CPMS	13	27–53; 44	[3–7·5]; 4·7	4·5
RRMS – COP	7	25–53; 35	[0–5]; 2	4

EDSS, Kurtzke's expanded disability status scale; RRMS, relapsing–remitting multiple sclerosis; CPMS, chronic progressive multiple sclerosis; COP, co-polymer.

Lymphocyte isolation and stimulation

PBMC were isolated from ethylenediaminetetraacetic acid anti-coagulated blood by Ficoll–Hypaque (Pharmacia, Uppsala, Sweden) gradient centrifugation immediately after venepuncture. Cells were then washed three times in Hanks' balanced salt solution and resuspended in RPMI-1640 (Gibco, Geneva, Switzerland) medium supplemented with 10% heat-inactivated fetal calf serum (FCS), 10⁴ IU penicillin and 10 µg/ml streptomycin, 20 mm HEPES, and 2 mm l-glutamine (Gibco, Geneva, Switzerland).

Monoclonal antibodies

Mouse monoclonal antibodies (mAbs) specific for human surface antigens used in this study included anti-CD3 [immunoglobulin G (IgG) 1k, clone SK7] and anti-CD8 (IgG1k, clone SK1) (Becton Dickinson, Basel, Switzerland) mAbs. All surface mAbs were conjugated to fluorescein isothiocyanate (FITC). Control mAbs included mouse FITC- or phycoerythrin (PE) -conjugated IgG1k or IgG2a mAbs (Becton Dickinson). For intracellular detection of cytokines, the following PE-conjugated mAbs were used: anti-IL-2 (clone MQ1-17H12) and anti-TNF-α (clone Mab11) (Pharmingen, La Jolla, CA), anti-IFN-γ (clone 25-723) (Becton Dickinson). All mAbs were used at 1/100 dilution, except the anti-IFN-γ which was diluted at 1/50.

Quantification of IL-2-, IFN-γ- and TNF-α-producing T cells by intracellular detection of cytokines

PBMC from healthy donors or MS patients were stimulated for 16 hr with PI [20 ng/ml phorbol 12-myristate 13-acetate (Sigma Chemical Co., Geneva, Switzerland) and 1 µg/ml ionomycin (Sigma)]. Brefeldin A (Sigma) was added at 10 µg/ml during the last 14 hr to disrupt intracellular protein transport and cause cytokine accumulation within the Golgi apparatus. Enumeration at the single cell level of cytokine-producing peripheral T cells was performed as previously described.³⁸ Briefly, 10⁵ stimulated PBMC were washed in phosphate-buffered saline with 1% bovine serum albumin and 0·1% sodium azide (PBS-BSA-NaN₃) and incubated in the same buffer with FITC-conjugated anti-CD3 or anti-CD8 mAbs, for 30 min. After washing, cells were fixed in PBS-BSA-NaN₃ containing 1% paraformaldéhyde (PFA) for 15 min at 4°. Fixed cells were permeabilized at room temperature with 0·05% saponin (SAP) in PBS-BSA-NaN₃ for 15 min. Intracellular cytokine staining was performed with PE-conjugated anti-cytokine mAbs in SAP-PBS-BSA-NaN₃ for 30 min at 4°. Doubled stained cells were washed in PBS-BSA-NaN₃, fixed with 1% PFA in PBS-BSA-NaN₃ and immediately applied to the flow cytometer (FACscalibur, Becton Dickinson). Because of the previously reported down-modulation of CD4 molecule following phorbol ester stimulation^38–40 and according to previous studies, the percentage of CD4 T cells producing a given cytokine was deduced from the difference between the percentage of cytokine-producing CD3⁺ T cells and the percentage of cytokine-producing CD8⁺ T cells. For each sample, 10000 doubled-stained cells were acquired following size forward scatter (FSC) and granularity side scatter (SSC) criteria, and were analyzed with Winmdi software (version 2.6 and 2.7). Cytokine production was evaluated on both living and apoptotic T cells. Cytokine-positive cells were identified according to background staining (less than 1% of the analysed subset) obtained with isotype-matched control mAbs.

Statistical analyses

Statistical analyses included the Mann–Whitney test. A P-value < 0·05 was considered significant.

Results

CD4 and CD8 T-cell subset distribution following PI stimulation

PBMC from healthy donors or MS patients were stimulated for 16 hr with PI and the proportions of CD4⁺ and CD8⁺ T cells within total PBMC were determined at the end of the culture by fluorescence-activated cell sorter (FACS) analysis following staining with subset-specific antibodies. The efficacy of PI stimulation was tested for each sample by following the expression of the early activation marker CD69 within the CD3⁺ subset. Thus, the percentage of CD69⁺ CD3⁺ T cells rose from 2% to more than 95% in all groups of donors (data not shown). When all MS patients were considered, no significant difference in the proportions of CD4 and CD8 T cells following PI stimulation between patients and control subjects was found. However, when patients were subdivided into RRMS remitting/relapsing and CPMS, the proportions of CD4 and CD8 T cells were significantly different in CPMS versus healthy donors (respectively 62·5 ± 4·5% versus 49·9 ± 7·4%, P < 0·05 for CD4 T cells, and 13·8 ± 3·4% versus 22·1 ± 4·3% P < 0·001 for CD8 T cells). In contrast, when RRMS patients were compared to controls, no significant difference was observed at both CD4 (55 ± 4·8% in remitting and 55 ± 5·8% in relapsing versus 49·9 ± 7·4% in controls) and CD8 T cell level (22·9 ± 4·4% in remitting and 24·9 ± 4·6% in relapsing versus 22·1 ± 4·3% in controls). A similar observation was made for RRMS patients treated with COP (CD4: 56 ± 5·9%; CD8: 24·4 ± 3·8%).

Single-cell analysis of cytokine synthesis by peripheral blood CD3 T cells from MS patients

To measure accurately Th1 populations and to determine their respective proportions at different stages of MS, a method of single-cell analysis by flow cytometry was used. PI-stimulated PBMC from control donors or MS patients were surface stained with anti-CD3 mAbs and intracellularly stained with mAbs against TNF-α, IL-2, or IFN-γ. Figure 1 shows representative stainings from a control donor and a CPMS patient.

Figure 1.

Single-cell analysis of CD3 T cells producing cytokines. PBMC from a control donor and a CPMS patient were stimulated for 16 hr with PMA and ionomycin. Brefeldin A was added to the cultured medium for the last 12 hr. After a first staining with anti-CD3 mAbs, the cells were permeabilized and co-stained with anti-cytokine mAbs. Analysis of cytokine-producing cells was performed on total (living+apoptotic) T cells, as shown by the gate R1 on the size/granularity dot plots. Numbers in quadrants of the dot plots represent the percentage of PBMC. Numbers in brackets correspond to the percentage of cytokine-positive cells among CD3 T cells.

With regard to size and granularity parameters of stimulated lymphocytes, no difference was observed between the control and MS patients. Analysis of the frequency of cytokine-producing CD3 T cells among PBMC revealed that the percentage of CD3⁺ TNF-α⁺ cells was increased in the CPMS patient compared to the control (43·1% versus 31·1%) and, similarly, an increase in the percentage of CD3⁺ IL-2⁺ T cells was observed in the CPMS patient (56·2% versus 34·7%). In contrast, the proportion of CD3⁺ IFN-γ⁺ T cells was comparable in CPMS and control donor (15·7% versus 14·5%).

This experimental approach was applied on blood samples from 13 healthy donors and 61 MS patients classified according to the clinical stage of the disease. The pattern of cytokine synthesis following polyclonal stimulation is independent of the gender both in controls and MS patients (data not shown). It is noteworthy that MS is associated with a significantly increased frequency within the PBMC of CD3 T cells primed for TNF-α synthesis, whether patients were RRMS remitting (P < 0·05), RRMS relapsing (P < 0·001), or CPMS (P < 0·05) Table 2. A similar significant increase was observed in the proportions of TNF-α⁺ CD3⁺ T cells among CD3 T cells for RRMS relapsing and CPMS patients compared to controls (Table 2). With regard to IL-2 production, an increased percentage of CD3⁺ IL-2⁺ T cells within PBMC was observed in all groups of MS patients, and it was significantly different from controls in RRMS remitting (P < 0·01) and in CPMS (P < 0·01) patients. The proportion of CD3⁺ IL-2⁺ T cells within CD3⁺ T cells was significantly increased in CPMS patients only (P < 0·01 versus controls). Interestingly, the frequency of PI-induced IFN-γ-producing cells within PBMC or CD3⁺ T cells was comparable among MS patients, whatever their clinical stage, and similar to that of controls (Table 2).

Table 2.

Single cell analysis of TNF-α, IL-2 and IFN-γ production by CD3 T cells following 16 hr PI stimulation

	Healthy controls (n = 13)	RRMS remitting (n = 18)	RRMS relapsing (n = 12)	CPMS (n = 13)	RRMS COP (n = 7)‡
% of TNF-α+ CD3+	42·3 ± 8·1	46·5 ± 6·1	52·7 ± 8·8	51·6 ± 6·9	44·4 ± 9
		ns*	P < 0·01*	P < 0·01*	ns*; ns†
% of IL-2+ CD3+	58·8 ± 7·1	65·1 ± 10·8	61·6 ± 9·9	68·7 ± 8·4	53·1 ± 6·7
		ns*	ns*	P < 0·01*	ns*; P < 0·02†
% of IFN-γ+ CD3+	23·8 ± 5·1	25·6 ± 5·5	24·8 ± 8·2	20 ± 4·3	17 ± 4·1
		ns*	ns*	ns*	ns*; P < 0·01†

Results are expressed as mean±SD.

ns, not statistically significant.

RRMS, relapsing–remitting multiple sclerosis; CPMS, chronic progressive multiple sclerosis; COP, copolymer.

^*Statistically significant in comparison with healthy controls.

^†Statistically significant in comparison with RRMS remitting patients.

^‡n = 6 for IFN-γ analyses.

Differential alteration in the priming for cytokine synthesis of CD4 and CD8 T-cell subsets from MS patients

The respective contribution of CD4 and CD8 T-cell subsets to TNF-α, IL-2 and IFN-γ synthesis by stimulated PBMC was assessed (Fig. 2). Compared to controls, a significant increase in the frequency of CD4 T cells producing TNF-α within PBMC was found in all MS patients, particularly in CPMS (P < 0·001 versus controls). In contrast, the proportion of CD8 T cells synthesizing TNF-α was unchanged in MS patients, whatever their clinical stage (Fig. 2). Therefore the global increase in the proportion of T cells primed for TNF-α synthesis in MS patients is directly correlated to the increased capacity of the CD4 T-cell subset to synthesize this cytokine. Similarly, an increased priming for IL-2 synthesis was found in CD4 T cells from MS patients, which was particularly pronounced in CPMS (P < 0·001 versus controls). The concomitant increased proportion of patients' CD4 T cells synthesizing TNF-α and IL-2 is probably the consequence of the co-production of both cytokines by these subsets as recently demonstrated.⁴¹ At the CD8 T-cell level, the proportion of IL-2 producers within PBMC was unchanged in MS patients compared to controls (Fig. 2). Finally, as previously found at the CD3 T-cell level (Table 2), IFN-γ expression by CD4 T cells was comparable in MS patients and controls (Fig. 2).

Figure 2.

Comparative analysis of cytokine production by CD4⁺ and CD8⁺ T cells. PBMC from 13 controls (C), 19 RRMS remitting (I), 12 RRMS relapsing (II), 13 CPMS (III), and seven RRMS-COP were stimulated as described in Fig. 1. Each dot represents one donor. Horizontal bars indicate the mean value in each group. Statistical analyses included the Mann–Whitney test which compared the different groups of patients (*) to the controls and RRMS remitting patients to RRMS-COP patients (#). The P-values are indicated in each quadrant; NS, not significant.

In contrast, the frequency of IFN-γ-producing CD8 T cells within PBMC was decreased in CPMS patients compared to the other groups (P < 0·05). This reduction is the consequence of the reduced number of total CD8 T cells in PI-stimulated PBMC from the CPMS group (cf. the first paragraph).

MS is associated with an increased capacity of CD4 T-cell subset to produce TNF-α and IL-2

Figure 3 shows the frequency of CD4 or CD8 T cells producing TNF-α, IL-2, or IFN-γ from all groups of donors. The relative frequency of TNF-α-producing CD4 T cells within the CD4 subset was increased in all groups of patients compared with healthy subjects (mean percentage 40·3 ± 8% in controls, 47 ± 7·3% in RRMS remitting (P < 0·01), 55·6 ± 9·3% in RRMS relapsing (P < 0·001), 52·1 ± 7·1% in CPMS patients (P < 0·001). The frequency of IL-2-producing CD4⁺ T cells within the subset was similar in RRMS patients compared to controls (70 ± 11·6 in RRMS remitting, 69 ± 12·4 in RRMS relapsing versus 67·1 ± 7·4 in controls), whereas it was slightly but significantly increased in CPMS patients (75·2 ± 8·8%; P < 0·05). With regard to IFN-γ synthesis, the frequency of CD4⁺ T cells synthesizing IFN-γ was similar in MS patients compared to controls (18·8 ± 6·4% in RRMS remitting, 18·4 ± 6·7% in RRMS relapsing, 15·8 ± 3·9% in CPMS and 15·4 ± 4·7% in controls). It is noteworthy that, whatever MS clinical stage, the CD8 T-cell subset was not significantly altered in its capacity to produce the three type 1 cytokines analysed in this study (between 45% and 50% for TNF-α; between 43 and 50% for IL-2; and between 38 and 42% for IFN-γ).

Figure 3.

Comparative analyses of cytokines produced by the CD4 and CD8 T-cell subsets from controls and MS patients. PBMC were stimulated and analysed as described in Fig. 1. Histograms represent the percentage of the mean value±standard deviation for the CD4 and CD8 T cells among the different groups of donors (I, controls; II, RRMS remitting; III, RRMS relapsing; IV, CPMS; V, RRMS-COP). Statistical significance was assessed by the Mann–Whitney test. *P represents the P-value compared with the group of controls, whereas #P represents the P-value compared with the RRMS remitting group.

Beneficial effect of COP on cytokine synthesis

We have evaluated the in vivo influence of COP treatment given to RRMS patients for at least 1 year. As shown in Table 1, the disability score of these patients ranged from 0 to 5 with a mean of 2 and it was close to that of RRMS remitting patients (mean 1·4). Determination of the frequency of peripheral T cells producing TNF-α, IL-2 and IFN-γ following PI stimulation indicated a lack of alteration in cytokine synthesis in those patients. Indeed, Table 2 shows that the frequency of TNF-α- and IL-2-producing CD3 T cells within PBMC was similar in COP-treated patients and in controls (36 ± 6·6% versus 30·3 ± 5·9% for TNF-α, 43 ± 4·6% versus 42·6 ± 6·6 for IL-2) while it was significantly increased in RRMS remitting patients. Importantly, analysis of the capacity of CD4 T cells to produce cytokines (Fig. 3) revealed that the frequency of CD4 T cells primed for TNF-α synthesis, which was significantly increased in all stages of MS, including RRMS remitting, was normalized to control values in COP-treated patients (43·2 ± 11·8% in treated patients versus 47 ± 7·3% in RRMS remitting versus 40·3 ± 8% in controls). In addition, a significant decrease in the frequency of CD4 T cells primed for IL-2 was found in COP-treated patients compared to the other groups of patients, reaching values below that of controls (59·1 ± 9·9% in treated patients versus 70 ± 11·6% in RRMS remitting versus 67·1 ± 7·4% in controls). Unexpectedly, COP-treated patients showed a significantly decreased priming for IFN-γ at the CD4 T-cell level (9·1 ± 3·4% in treated patients versus 18·8 ± 6·4% in RRMS remitting versus 15·4 ± 4·7% in controls (Fig. 3). No significant modification was observed at the CD8 T-cell level.

Discussion

This single-cell analysis of the priming of peripheral T cells for type 1/pro-inflammatory cytokine synthesis revealed that MS is associated with an increased frequency of T cells primed for TNF-α and IL-2 synthesis, which is related to the stage of the disease, while no modification was observed for IFN-γ synthesis. These alterations concerned exclusively the CD4 T-cell subset while the frequency of cytokine-producing CD8 T cells was similar to that of control donors. In addition, we show that COP-1 treatment has beneficial effects on this ex vivo altered cytokine profile.

The increased T cell priming for TNF-α that we report is consistent with previous studies showing higher levels of TNF-α mRNA and increased mitogen-induced TNF-α production by whole blood from CPMS patients⁴² or MS patients prior to clinical manifestations.^6,15,16 An important technical aspect of our study is that we analysed the cytokine profile of freshly isolated PBMC by intracellular double-fluorescence flow cytometry which allowed the precise identification of the cytokine pattern of each T-cell subset, and we show for the first time that increased priming for TNF-α synthesis concerns exclusively the CD4 T-cell subset of MS patients. The influence of TNF-α on MS pathology remains controversial. Treatment of mice with anti-TNF-α or soluble TNF-α receptor I abrogates autoimmune demyelination or prevents chronic relapsing EAE,^43,44 while TNF-α augments EAE.⁴⁵ In contrast, TNF-α shows anti-inflammatory and protective effects in several animal models, in particular in autoimmune-mediated demyelination.⁴⁶ It was reported that the lack of TNF-α in mice predisposes them to MOG-induced demyelination²¹ and induces EAE.⁴⁷ In humans, anti-TNF-α therapy induces an increased MRI signal, and recombinant soluble TNF-α receptor p55 immunoglobulin fusion protein was reported to induce increased exacerbations in MS patients.^20,48 These observations suggest that TNF-α has a suppressive role on autoimmunity, and several mechanisms can be proposed. For example, local expression of TNF-α may be able to tolerize auto-reactive CD4 T cells, shifting the cytokine pattern toward a Th2 profile,⁴⁹ or TNF-α may induce cell death of autoimmune mature T cells,⁵⁰ or chronic exposure to TNF-α may attenuate the TCR transduction pathway.⁵¹ In MS patients, TNF-α may be overexpressed at the initial stage of a microbial infection, thus priming an antigen-specific Th1 response, and consequently to molecular mimicry mechanisms, it could induce or exacerbate the autoimmune disease.^52,53 However, at the end of a relapse or in CPMS, TNF-α may slow down the autoimmune response.

IL-2 and soluble IL-2 receptor are found in the serum and CSF of patients suffering from MS,^22,54,55 increased production of IL-2 by T cells following OKT3 stimulation has been reported during MS exacerbations⁵⁶ and a positive correlation between IL-2-secreting PBMC and serial gadolinium MRI disease activity has been found in MS patients.²³ Our observation of increased priming of MS patients' T cells for IL-2 synthesis in CPMS and some RRMS relapsing patients is thus consistent with these observations, but we reveal that increased IL-2 production is related to the CD4 T-cell subset, which is the major source of IL-2.³⁹

The beneficial effect of COP treatment has been associated with a shift from Th1 to Th2/Th3 cytokine profile, COP-treated patients showing increased levels of IL-10, TGF-β and IL-4 in peripheral blood cells,⁵⁷ and COP-reactive T-cell lines from COP-treated patients being predominantly of the Th2 type.⁵⁸ Our data extend these observations and show that COP-treatment suppresses the type 1/pro-inflammatory pattern of patients' T cells by reducing the frequency of CD4 T cells primed for TNF-α and IL-2 synthesis, reaching values of healthy donors.

The role of IFN-γ in MS disease remains unclear. A disease-promoting role is suggested by the finding that transgenic expression of IFN-γ in the CNS is associated with demyelination,⁵⁹ IFN-γ-producing, myelin-reactive CD4 T cells can adoptively transfer EAE⁷ and, in vitro, IFN-γ elicits a number of effects on oligodendrocytes, including cell death.⁶⁰ In humans, IFN-γ was found to promote MS and to increase the exacerbation rate.⁶¹ In contrast a protective role of IFN-γ is suggested by the findings that IFN-γ administration in Lewis rats suppresses EAE,⁶² IFN-γ or IFN-γ receptor knock-out mice can develop EAE^26,63 and anti-IFN-γ antibodies exacerbate EAE disease.⁶⁴ Our flow cytometry analysis did not show any alteration in the frequency of CD4 and CD8 T cells primed for IFN-γ in MS patients, whatever their clinical stage. These data are compatible with the reported lack of relationship between IFN-γ production and MRI-defined brain lesions in MS patients²⁴ and the unchanged frequency of IFN-γ-producing blood T cells in MS patients,⁶⁵ but they are different from those reported by Becher et al.²⁵ who found a weak increased proportion of IFN-γ–T-cell producers in CPMS patients. Immunoregulatory therapy such as IFN-β was reported to down-regulate IFN-γ expression in both CD4 and CD8 T cells.^25,65 Our study shows that COP treatment also down-regulates IFN-γ expression, particularly at the CD4 T-cell level. Regarding the possible mechanism of action of COP in vivo, it has been suggested that the therapeutic effect of COP in MS is related to the induction of Th2 cells,^33,35,57,58 and our study shows that it is associated with the suppression of the Th1-biased cytokine profile. This is in agreement with the recent demonstration that COP suppresses IFN-γ production in EAE.⁶⁶ Thus COP may induce COP-reactive Th2 cells which will compete inside the CNS with MBP by both TCR antagonism and MHC class II blocking^31,33 and may exert bystander suppression on other inflammatory cells.^34,67 In conclusion, the beneficial effect of COP on MS disease may be associated with abrogation of the priming of T cells for IFN-γ and IL-2 synthesis. The modest effect of COP on TNF-α synthesis might have some relevance with recent data showing the absence of deleterious effect of this cytokine on MS.