Interferon Beta–Induced Restoration of Regulatory T-Cell Function in Multiple Sclerosis Is Prompted by an Increase in Newly Generated Naive Regulatory T Cells FREE

Reduced prevalence or suppressive capacity of CD4⁺CD25⁺FOXP3⁺ regulatory T (T_reg) cells have been described in various autoimmune diseases in human beings.^{1 - 11} We and others have shown that T_reg cells derived from patients with relapsing-remitting multiple sclerosis (RRMS) demonstrate functional impairment as their potential to inhibit myelin-specific and antigen nonspecific T-cell proliferation is diminished compared with that in healthy individuals.^{8 - 11} Moreover, T_reg cells delay the onset and progression of experimental autoimmune encephalomyelitis and promote recovery from active disease.^{1 ,12} Hence, T_reg-cell dysfunction in multiple sclerosis (MS) may be causally related to facilitated activation and migration of autoreactive effector T (T_eff) cells into the central nervous system. This implies that pharmacologic modulation of T_reg-cell prevalence or function has the potential to beneficially influence the severity and long-term course of MS. In accord with this assumption, recent observations of experimental autoimmune encephalomyelitis indicate that enhancement of T_reg-cell effector functions by either adoptive transfer of preactivated T_reg cells or by direct in vivo stimulation of T_reg cells provides protection from disease.^{12 - 14}

Interferon beta is an immunomodulatory drug with established therapeutic efficacy in MS. It decreases relapse rates and slows disease progression as measured clinically and by using magnetic resonance imaging.^{15 - 16} The mechanisms behind these beneficial effects are not fully understood. Numerous studies have demonstrated anti-inflammatory properties of interferon beta treatment such as downregulation of activated T cells, apoptotic elimination of T lymphocytes, and amelioration of blood-brain barrier dysfunction.^{15 - 16} There is emerging evidence that interferon beta may also affect the T_reg-cell compartment.^{17 - 18} We recently showed that the prevalence of CD31-coexpressing naive helper T cells (T_H) and naive T_reg cells that have just entered the circulation as recent thymic emigrants (RTE-T_reg cells) (CD4⁺CD45RA⁺CD45RO⁻CD31⁺ RTE-T_H and CD4⁺CD45RA⁺CD45RO⁻CD31⁺FOXP3⁺ RTE-T_reg cells) are decreased in patients with RRMS.¹⁹ We further demonstrated that the prevalence of RTE-T_reg cells is critical for suppressive function of circulating T_reg cells, which suggests that a shift in the homeostatic composition of T_reg- cell subsets related to a reduced thymic-dependent de novo generation of RTE-T_reg cells with compensatory expansion of memory T_reg cells contributes to the MS-related functional T_reg cell impairment.

We assessed the effects of interferon beta treatment on the T_reg-cell compartment in 20 patients with RRMS undergoing immunomodulatory therapy. We determined prevalence, phenotype, and suppressive function of T_reg cells before treatment and after 3 and 6 months of therapy.

We demonstrate that exposure to interferon beta increases T_reg cell inhibitory capacity in most patients with RRMS. Whereas restoration of T_reg-cell function coincided with only slightly enhanced prevalence of T_reg cells in the peripheral T-cell compartment, we observed a marked increase in the RTE-T_reg cell subsets and a parallel decrease in CD4⁺CD45RA⁻CD45RO⁺FOXP3⁺ memory T_reg cells. Thus, the increase of T_reg cell inhibitory capacity mediated by interferon beta can be explained by its effect on the homeostatic balance within the T_reg-cell compartment.

Blood specimens were obtained from 20 patients with RRMS (mean age, 36.8 years; age range, 21-53 years) and 18 healthy control individuals (31.1 years; 19-54 years). All patients had definite RRMS according to the criteria of McDonald et al²⁰ or Poser et al²¹ and had experienced, on average, 2 relapses (range, 1-4). Mean duration of disease was 3 years (range, 0-8 years). The mean Expanded Disability Status Scale was 1.9 (range, 1-3). Eighteen patients received interferon beta-1a drugs (11 patients received Rebif [Serono, Inc, Randolph, Massachusetts], and 7 patients received Avonex [Biogen, Inc, Cambridge, Massachusetts]). Two patients received interferon beta-1b (Betaferon; Bayer Schering Pharma AG, Berlin, Germany). Rebif was given at a dose of 22 μg 3 times per week initially; 4 patients continued to receive Rebif, 44 μg/wk, 3 times per week after 4 weeks. Blood samples for analysis of T_reg-cell function were obtained at 3 and 6 months after treatment. During the observation period, clinical status in all patients remained stable.

The T_reg cells from 2 healthy control subjects and 2 patients with MS were tested in parallel for their capacity to inhibit both syngeneic and allogeneic responder T cells ex vivo. The protocol was approved by the University of Heidelberg Ethics Committee, and all study participants gave written informed consent.

Monoclonal antibodies (mAbs) were obtained from BD Pharmingen (Heidelberg, Germany; CD4-APC.Cy7, CD45RO-PE.Cy7, CD45RA-ECD, CD31-FITC), Miltenyi Biotech Inc (Auburn, California; CD25-PE), and eBioscience, Inc (San Diego, California; FOXP3-APC). FOXP3 staining was performed according to the manufacturer's protocol. For 6-color flow cytometric analysis, cells were first stained with surface mAb specific for CD4, CD25, CD45RO, CD45RA, and CD31, followed by FOXP3 intracellular staining. Flow cytometry was performed immediately with a cytometer (FACSCanto; BD Biosciences, San Jose, California) and analyzed with commercially available software (FACSDiva; BD Biosciences). Quantification of T_reg-cells and T_reg-cell subsets was performed as previously described.¹⁹

In short, stained peripheral blood mononuclear cells were gated on CD4⁺ cells and analyzed for expression of FOXP3 and CD25. Inasmuch as only CD4⁺ T cells expressing the FOXP3⁺ gene product scurf in exhibit suppressive function,²² CD4⁺CD25⁺FOXP3⁺ cells were defined as T_reg cells and CD4⁺FOXP3^- cells as T_H. T_reg and T_H cells were further analyzed for their CD45RA/CD45RO surface expression to identify CD45RA⁻CD45RO⁺ memory and CD45RA⁺CD45RO−naive subsets. RTE-T_reg and RTE-T_H were identified by coexpression of the CD31 molecule on naive T_reg and T_H cells (Figure 1).

Peripheral blood mononuclear cells were isolated from 50 mL of peripheral blood using density gradient centrifugation with Ficoll-Hypaque (Biochrom AG, Berlin, Germany). CD4⁺ T cells were isolated from peripheral blood mononuclear cells using a negative CD4⁺ T-cell isolation kit (Dynal Biotech GmbH, Hamburg, Germany). CD25^high T_reg and CD25^low/CD4⁺CD25^int T_eff were isolated from the pure untouched CD4⁺ T cells using CD25 magnetic beads (Dynal Biotech GmbH). Immune magnetic separation constantly yielded highly pure T_reg cells with more than 90% CD4⁺CD25^high cells, as previously described.⁹ Purity of random samples was further tested using intracellular FOXP3 staining, which revealed more than 93% FOXP3⁺ cells in preparations of total T_reg cells.¹⁹ Furthermore, flow cytometry analysis revealed constant purities of both T_reg- and T_eff-cell preparations obtained from serial blood specimens during the observation period.

Cell proliferation assays were performed as described previously.^{9 ,19} In brief, 10⁵ freshly isolated T_eff cells were incubated in 96-well plates (Nunc GmbH & Co, Wiesbaden, Germany) in 200-μL culture medium alone or in 4:1 coculture with 2.5 × 10⁴ total T_reg cells. For T-cell receptor stimulation, soluble anti-CD3 (1 μg/mL) and anti-CD28 mAbs (1 μg/mL) were added to the culture medium. After 4 days at 37°C in 5% carbon dioxide, 1 μCi of tritium-labeled thymidine per well was added for an additional 16 hours. Proliferation was measured using a scintillation counter. Inhibition rate (percent) of T_reg cells in coculture experiments was defined as follows:

[1-3(H) Thymidine Uptake (in Counts per Minute [cpm]) in T_reg/T_eff Coculture / cpm of T_eff Alone] × 100.

Each experiment was performed in triplicate. The CD3 and CD28 mAbs used for cell culture experiments were purified from hybridoma supernatants using protein A affinity purification as described previously.²³ Cell culture media and mAbs were endotoxin-free (<10 pg/mL) as assessed by the Limulus assay (Sigma-Aldrich, Taufkirchen, Germany).

We used the nonparametric 2-tailed Mann-Whitney test to analyze for differences in cell counts and ex vivo T_reg cell–mediated suppressions at various times. P < .05 was considered significant. Statistical analyses were performed using commercially available software (SPSS version 14; SPSS, Inc, Chicago, Illinois).

As described by Haas et al,¹⁹ mean (SD) RTE-T_H cells were significantly reduced in patients with MS compared with healthy controls (11.5% [4.2%]; median, 12.2%; vs 25.2% [6.1%]; median, 26.3%; P < .01; Figure 2A). Similarly, total naive T_H cells were decreased (28.6% [12.3%]; median, 26.5%; vs 45.3 [12.0%]; median, 43.6%; P = .03), whereas memory cells among CD4⁺ T cells were increased (54.0% [14.2%]; median, 54.5%; vs 42.0% [4.8%]; median, 41.0%; P < .01).

The mean (SD) peripheral CD4⁺ T-cell population in patients with MS contained slightly lower numbers of FOXP3⁺ T_reg cells compared with healthy controls (4.6% [1.5%]; median, 4.9%; vs 5.6% [1.1%]; median, 5.2%; P = .08; Figure 2B). Patient-derived T_reg cells harbored reduced prevalence of naive T_reg cells (24.2% [6.4%]; median, 26.0%; vs 30.7% [9.5%]; median, 29.8%; P = .09) and RTE-T_reg cells (3.5% [1.2%]; median, 3.5%; vs 8.2% [4.4%]; median, 8.4%; P < .01) but moderately higher prevalence of memory T_reg cells (62.6% [6.7%]; median, 60.0%; vs 56.3% [11.0%]; median, 53.7%; P = .12), as previously reported.¹⁹

Treatment with interferon beta was associated with decreased prevalence of leukocytes and lymphocytes (data not shown), whereas the prevalence of CD4⁺ T cells was unchanged (Figure 2A). Mean (SD) naive T_H and RTE-T_H cells had markedly increased at 3 and 6 months after treatment (naive T_H cells, 3 months: 37.7% [9.3%]; median, 39.1%; P = .06; and 6 months: 49.2% [14.6%]; median, 45.5%; P < .01; RTE-T_H cells, 3 months: 20.8% [8.6%]; median, 22.5%; P = .01; and 6 months: 25.3% [10.9%]; median, 23.7%; P = .04). Conversely, memory T_H cells had decreased (3 months: 50.9% [8.8%]; median, 50.5%; P = .03; and 6 months: 38.6% [13.1%]; median, 40.4%; P = .02).

The prevalence of FOXP3⁺ T_reg in CD4⁺ T cells was slightly elevated at 3 and 6 months after therapy (5.3% [1.6%]; median, 6.0%; P = .15; and 5.6% [1.8%]; median, 5.9%; P = .11), though the changes were not statistically significant (Figure 2B). Naive T_reg and RTE-T_reg cells had increased at 3 and 6 months (3 months: 28.3% [8.5%]; median, 28.7%; P = .08; and 6.8% [3.1%]; median, 7.4%; P = .03; and 6 months: 34.7% [11.2%]; median, 35.0%; P = .11; and 9.0% [3.3%]; median, 8.5%; P < .01). Coincidental with the increases in RTE-T_reg cells, memory-T_reg cells had decreased (3 months: 54.2% [12.4%]; median, 52.6%; P = .09; and 6 months: 48.7% [10.9%]; median, 46.9%; P = .05).

Similar to recent reports,^{8 - 9 ,11 ,18} mean (SD) patient-derived T_reg cells exhibited markedly reduced in vitro inhibitory activity compared with healthy control-derived T_reg cells (32.3% [13.0%]; median, 27.4%; vs 64.8% [15.3%]; median, 66.8%; P = .01; Figure 3A). Low suppression rates (<30%) were noted in 16 of 20 patients.

After 3 and 6 months of interferon beta treatment, mean (SD) inhibition rates had increased significantly to 65.5% (15.4%; median, 63.3%; P < .01) and 58.3% (12.9%; median, 59.7%; P < .01). A clear increase in T_reg-cell inhibition rates (>10%) was present in 15 patients, whereas 5 individuals exhibited only small increases in T_reg-mediated suppression (0%-10%). At 6 months after therapy, low suppression rates (<30%) were present in only 1 patient. In contrast, the functional impairment was unchanged when T_reg cells were serially derived from patients without immunomodulatory treatment, as shown previously.⁹

The treatment-associated increase in T_reg-cell suppressive activities was linked to increased RTE-T_reg cell prevalence in 14 of 15 patients, whereas in 4 of 5 individuals without clear improvement in T_reg cell suppressive activities, prevalence of RTE-T_reg cells had not changed at follow-up (Figure 3B).

The increase of T_reg cell–mediated suppression and the increased RTE-T_reg-cell prevalence in patients with MS receiving interferon beta therapy occurred independently of the interferon beta preparation used for treatment and of clinical parameters including Expanded Disability Status Scale score, disease duration, and number of relapses.

We reciprocally mixed patient T_reg cells before and 4 weeks after initiation of interferon beta treatment with anti-CD3/CD28-stimulated syngeneic T_eff and in parallel with healthy control T_eff, and vice versa, in 2 independent experiments. Patient T_reg cells exhibited weak antiproliferative effects (mean [SD]) on both autologous (29.5% [9.8%]) and allogeneic (21.0% [10.5%]) T_eff, whereas healthy control T_reg cells mediated a stronger reduction of proliferation of both T_eff-cell subsets (autologous T_eff, 54.8% [6.5%]; allogeneic T_eff, 49.9% [13.1%]). After 4 weeks of treatment, T_reg cells from the same patients exhibited markedly improved suppression of both healthy control T_eff (53.0% [10.6%]) and patient T_eff cells (55.4% [13.5%]). Conversely, healthy control T_reg cell–mediated inhibition remained stable (autologous T_eff, 48.6% [7.7%]; and allogeneic T_eff cells, 49.0% [9.2%]; data not shown). This observation confirms that direct targeting of T_reg cells contributes to enhanced T_reg-cell suppressive capacity prompted by interferon beta therapy.

Mean (SD) proliferative response of 1 × 10⁵ T_eff cells from 20 treatment-naive patients was 45.2 (35.5) × 10³ cpm at day 5 after polyclonal stimulation in vitro. During therapy, T_eff-cell immune responses remained stable (3 months: 38.4 [27.2] × 10³ cpm; and 6 months: 39.1 [16.0] × 10³ cpm; data not shown). Likewise, T_reg cells remained anergic in response to polyclonal stimuli during the observation period. Proliferative responses of 2.5 × 10⁴ T_reg cells did not change either, and ranged from 0.1 to 2.2 × 10³ cpm at baseline and from 0.1 to 1.9 × 10³ cpm after 6 months of treatment.

The functional T_reg-cell defect detectable in MS may imbalance the interaction between responder T cells and T_reg cells and compromise peripheral immune regulation.^{8 - 9 ,11} Herein, we show that T_reg-cell dysfunction is counteracted by pharmacologic modulation because interferon beta treatment markedly augments the T_reg-cell suppressive capacity independent of the interferon beta preparation used for treatment (data not shown). In 16 of 20 patients, T_reg cells produced low suppression of T_eff cell-proliferation at baseline and became a more potent T_eff cell-suppressor after treatment. These observations are in accord with a previous study that reported an augmentive effect of interferon beta on T_reg-cell function in 10 patients with RRMS after 3 and 6 months of treatment.¹⁷ In this study, T_reg cells exposed to interferon beta in vivo for 4 weeks showed an improved suppressive effector function ex vivo toward allogeneic IFN-β–naive T_eff cells compared with baseline suppression, which suggests a direct effect of interferon beta on T_reg cells.

We have recently shown that T_reg-cell dysfunction in patients with RRMS is linked to an imbalanced homeostatic composition of circulating T_reg cells, because prevalence of newly generated naive T_reg cells is significantly decreased along with a coincidental increase in T_reg cells exhibiting a memory-phenotype.¹⁹ Moreover, previous findings consistently demonstrate contraction of the entire naive T-cell compartment in MS.^{24 - 25} The altered T_reg-cell homeostasis effects T_reg-function as prevalences of RTE-T_reg cells correlate with T_reg-cell inhibitory properties and depletion of RTE-T_reg-cell prompts decreased suppression rates.¹⁹ In an attempt to establish whether interferon beta mediates changes in the composition of peripheral T_reg cells, we demonstrated that restoration of T_reg-cell suppressive function induced by interferon beta therapy is associated with a shift in T_reg-cell homeostasis, resulting in an increase in RTE-T_reg-cell prevalence to levels found in healthy controls and a coincidental decrease in memory cells. This effect was consistently detectable during treatment with all 3 interferon beta preparations (data not shown). The treatment-associated increase in RTE-T_reg cells correlated with upregulation of T_reg-cell inhibitory function in 14 of 15 patients, whereas RTE-T_reg-cell prevalence remained unaltered in 4 of 5 patients with stable low levels of T_reg cell suppressive activity, which suggests a causal link between T_reg-cell homeostasis and function. Therapy with interferon beta also promoted an increase in naive T_H and RTE-T_H cells, corroborating earlier observations with other immunomodulatory agents used for treatment of MS. Upregulation of naive T cells together with downregulation of memory T cells was reported during therapy with linomide.²⁶ Moreover, in a recent longitudinal study, glatiramer acetate prompted a significant increase in the percentage of CD4⁺CD45RA⁺ T cells.²⁷

In this study, interferon beta therapy was accompanied by a mild decrease in total leukocytes and lymphocytes, as previously reported.²⁸ The relative percentages of CD4⁺ T cells remained unaltered, and we observed a mild though nonsignificant increase in the percentage of FOXP3⁺ T_reg cells (approximately 1%). In contrast, FOXP3⁺ T_reg cells were moderately increased (<2%) in 15 patients with RRMS treated with interferon beta compared with 40 patients who received no treatment.¹⁸ However, the study by Venken et al¹⁸ used a nonprospective cross-sectional design to evaluate FOXP3⁺ T_reg cells in patients with heterogeneous treatment durations and did not include follow-up assessments. In line with our data, CD25^high T_reg-cell prevalence did not differ in patients with untreated MS compared with patients undergoing therapy with disease-modifying drugs including interferon beta.²⁹

In conclusion, IFN-β markedly upregulates the T_reg-cell inhibitory capacity and reverses T_reg cell dysfunction associated with RRMS. This effect seems to be conferred by drug-induced normalization of peripheral T_reg-cell homeostasis. The extent to which this effect contributes to stabilization of clinical disease activity remains to be determined in long-term follow-up studies. These findings, however, clearly demonstrate that T_reg cells are responsive to pharmacologic modulation, and efforts to selectively enhance T_reg-cell expansion or T_reg-cell function are a highly promising new treatment option in MS.

Correspondence: Brigitte Wildemann, MD, Division of Molecular Neuroimmunology, Department of Neurology, University of Heidelberg, Im Neuenheimer Feld 350, D-69120 Heidelberg, Germany (brigitte_wildemann@med.uni-heidelberg.de).

Accepted for Publication: February 22, 2008.

Author Contributions: Dr Korporal had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Korporal and Haas contributed equally to this study. Study concept and design: Korporal, Haas, Suri-Payer, and Wildemann. Acquisition of data: Korporal, Haas, Balint, Fritzsching, Moeller, and Fritz. Analysis and interpretation of data: Haas, Schwarz, and Wildemann. Drafting of the manuscript: Korporal, Haas, and Wildemann. Critical revision of the manuscript for important intellectual content: Korporal, Haas, Balint, Fritzsching, Schwarz, Moeller, Fritz, Suri-Payer, and Wildemann. Statistical analysis: Haas and Schwarz. Obtained funding: Suri-Payer and Wildemann. Administrative, technical, and material support: Korporal, Haas, Balint, Fritzsching, Moeller, Fritz, and Wildemann. Study supervision: Korporal, Haas, and Wildemann.

Financial Disclosure: None reported.

Funding/Support: This study was supported by grants 1.319.110/01/11 and 1.01.1/04/003 from the Gemeinnützige Hertie-Stiftung; grants SFB 405, 5H, and SFB 571, B7 from Deutsche Forschungsgemeinschaft; and a Young Investigator Award from the Faculty of Medicine, University of Heidelberg (B.F.)

Role of the Sponsor: None of the funders had a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or preparation, review, or approval of the manuscript.

Additional Contributions: We thank the patients with RRMS and the healthy controls for their participation in this study.