SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

Reestablishing immune tolerance and long-term suppression of disease represent major therapeutic goals in rheumatoid arthritis (RA). Dendritic cells (DCs) likely play a central role in such regulation via the expansion and/or induction of Treg cells. The present study was undertaken to explore the contribution of DCs to the development of Treg cells in a human autoimmune disease setting.

Methods

DC subsets were characterized by flow cytometry in the peripheral blood and synovial fluid of patients with RA. Proliferation of and cytokine release by naive CD4+CD25− T cells were measured in cocultures of these cells with DCs from patients with RA and healthy controls. The suppressive capacity of DC-polarized T cells was explored in vitro by a standard suppression assay.

Results

Only very low numbers of both plasmacytoid DCs (CD303+) and myeloid DCs (CD1c+) were present in the peripheral blood of patients with active RA. In contrast, patients with therapy-induced remission of RA exhibited higher numbers of circulating plasmacytoid DCs. Mature plasmacytoid DCs from RA patients with low disease activity, but not those from healthy controls, expressed high levels of indoleamine 2,3-dioxygenase and promoted the differentiation of allogeneic naive CD4+CD25− T cells into interleukin-10–secreting Treg cells, or Tr1 cells, that showed poor proliferation in vitro. Importantly, these plasmacytoid DC–primed Treg cells potently suppressed the proliferation of autologous naive CD4+ T cells, in a dose-dependent manner.

Conclusion

These results demonstrate, for the first time, that human plasmacytoid DCs may be educated within the rheumatoid microenvironment to acquire a tolerogenic phenotype. Modulation of the immune response by plasmacytoid DCs might provide novel immune-based therapies in autoimmunity and transplantation.

Rheumatoid arthritis (RA) is a chronic inflammatory autoimmune disease characterized by synovial inflammation, which is orchestrated by both innate and adaptive immune responses. Major hallmarks of autoimmunity in this disease include the appearance of serum autoantibodies and the persistent activation of self-reactive CD4+ cells (1). Although therapeutic strategies targeting cytokines and B or T cells have had a major clinical impact, disease in many patients remains refractory to current biologic interventions, and for patients who show a response to therapy, true remission associated with reestablishment of immune tolerance is rare. This has increased the interest in exploring strategies to reestablish immune tolerance and provide long-term disease suppression (2).

Dendritic cells (DCs) are professional antigen-presenting cells (APCs) that can induce either immunity or tolerance. Myeloid DCs and plasmacytoid DCs represent the 2 major subsets of DCs, with human plasmacytoid DCs defined as CD45+CD123+CD303+CD11c− cells and myeloid DCs defined as CD45+CD1a+ CD11c+CD1c+ cells. Although both subsets exhibit a functional plasticity in directing T cell responses (3), current evidence supports a predominant role of plasmacytoid DCs in the maintenance of tolerance through the expansion/induction of Treg cells (4–7).

To date, 2 major subsets of Treg cells have been described: thymus-derived Foxp3+ Treg cells and adaptive interleukin-10 (IL-10)–producing Treg cells (Tr1 cells) generated in the peripheral blood (8). The importance of Treg cells in the maintenance of tolerance is highlighted by the development of spontaneous multiorgan autoimmunity following Treg cell deletion in rodents (9). In humans, reduced numbers of Treg cells and impaired function of Treg cells have been reported in patients with autoimmune diseases such as rheumatoid arthritis (RA), systemic lupus erythematosus, and multiple sclerosis (10–12). The molecular pathways involved in prompting plasmacytoid DCs to promote Treg cell development are not fully understood. Expression of the indoleamine 2,3-dioxygenase (IDO) enzyme by plasmacytoid DCs is thought to play a significant role in plasmacytoid DC–mediated Treg cell induction (13–16). Whether human DCs exploit similar mechanisms of Treg cell induction that might facilitate reestablishment of tolerance in a disease setting remains to be seen.

Cytokine-directed therapy may have an impact on Treg cell function. Thus, Treg cells in patients with active RA are functionally defective. Anti–tumor necrosis factor α (anti-TNFα) therapy reverses this defect (10, 17, 18), through conversion of naive T cells into Foxp3+ Treg cells (18). Whether therapy may also have an impact on DCs, in terms of changing their effects on Treg cells, has not been explored. Herein, we present findings demonstrating that plasmacytoid DCs isolated from patients with RA whose disease is in remission induce an additional, distinct population of Treg cells, Tr1 regulatory cells, in a manner that is dependent on the presence of IDO. Unlike the previously described anti-TNFα–induced Treg cells (18), this population of genuine Tr1 cells does not express Foxp3 but, instead, secretes high levels of IL-10. Most notably, these cells suppress the proliferation of autologous naive CD4+ cells, in a dose-dependent manner.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Patient population.

Patients with RA (n = 35) fulfilling the revised classification criteria for RA from the American College of Rheumatology (formerly, the American Rheumatism Association) (19) were included in this study (Table 1). Disease activity was assessed by the Disease Activity Score in 28 joints (DAS28), and disease remission was defined as a DAS below 2.6 (20). A total of 20 patients with active RA and 15 with inactive RA (those achieving remission) were evaluated after they had been treated with either anti-TNFα therapy (infliximab at 3 mg/kg intravenously at weeks 0, 2, 4, and 8 and every 8 weeks thereafter) or methotrexate (10–15 mg/week orally). Ten healthy individuals were used as controls. All subjects gave their written informed consent prior to study enrollment.

Table 1. Characteristics of the patients with rheumatoid arthritis (RA)*
 Active RAInactive RA
  • *

    Except where indicated otherwise, values are the mean ± SD. Patients with missing data are excluded. DAS28 = Disease Activity Score in 28 joints; anti-TNFα = anti–tumor necrosis factor α; MTX = methotrexate.

  • Defined as those achieving disease remission from 6 months up to 2.5 years after the start of therapy for RA.

Sex, no. female/no. male11/44/8
Age, years66.2 ± 8.1852.5 ± 20.58
RA duration, years12.6 ± 11.749.41 ± 8.54
DAS285.06 ± 1.021.91 ± 0.48
Therapy, no. of patients  
 Anti-TNFα158
 MTX04

Media and reagents.

The following fluorescence-conjugated antibodies were used: CD45–phycoerythrin-cyanine-5 (PC-5) (clone J33), CD69–phycoerythrin (PE) (clone TP1.55.3), and CD4–PC-5 (clone 13B8.2) (all from Beckman-Coulter, Miami, FL); CD303–fluorescein isothiocyanate (FITC) (blood dendritic cell antigen 2 [BDCA-2], clone AC144), CD123-PE (IL-3R, clone AC145), and CD1c-PE (BDCA-1, clone AD5-8E7) (all from Miltenyi Biotec, Bergisch Gladbach, Germany); CD11c-FITC (clone 3.9), programmed death ligand 1 (PDL-1)–PE (clone MIH1), inducible costimulator ligand (ICOS-L)–PE (clone MIH12), CD80-PE (B7-1, clone 2D10), CD86-PE (B7-2, clone IT2.2), HLA–DR–PE (clone LN3), and Foxp3-PE (clone 3G3) (all from eBioscience, San Diego, CA); CD40-FITC (clone 5C3), CD25-FITC (clone M-A251), and CD62L-FITC (clone Dreg-56) (all from BD Biosciences PharMingen, San Diego, CA); and CCR7-PE (R&D Systems, Abingdon, UK). For 5,6-carboxyfluorescein succinimidyl ester (CFSE) staining of CD4+ T cells, the CellTrace CFSE cell proliferation kit (Invitrogen, Eugene, OR) was used.

Cell cultures were performed in RPMI 1640 supplemented with 10% fetal bovine serum, penicillin (100 units/ml), streptomycin (100 μg/ml), and 2 mML-glutamine (all from Gibco, Carlsbad, CA). Recombinant human IL-3 (R&D Systems) and recombinant human granulocyte–macrophage colony-stimulating factor (PeproTech, London, UK) were used in DC cultures. For activation of DCs, anti-human CD40L (PeproTech) was used. For stimulation of T cells, we used 5 μg/ml plate-bound anti-human purified CD3 (OKT-3 clone; eBioscience) and 1 μg/ml soluble anti-human purified CD28 (clone CD28.3; eBioscience). In some experiments, soluble anti-human CD3 (50 ng/ml) was used. The IDO inhibitor 1-methyl-DL-tryptophan (1-MT; Sigma-Aldrich, Munich, Germany) was dissolved in DMSO and added to in vitro cultures at a concentration of 10 μM.

Cell isolation.

Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Histopaque (Sigma-Aldrich) density-gradient centrifugation of heparinized venous blood aspirates. Synovial fluid (SF) samples were obtained from patients with active RA, and mononuclear cells were isolated from the SF using treatment with hyaluronidase solution, comprising 1 mg hyaluronidase (Sigma-Aldrich) per ml of hyaluronidase buffer (1M phosphate buffer, pH 7, 5M NaCl, 7.5% bovine serum albumin [BSA]), for 10 minutes at room temperature, followed by Ficoll-Histopaque density-gradient centrifugation.

Flow cytometry analysis.

PBMCs were magnetically separated using anti-CD3 microbeads (Miltenyi Biotec), following the manufacturer's instructions. The phenotypic characterization of the DC subsets was performed on the CD3− cell fraction, whereas the CD3+ cell population was extracellularly stained for CD4, CD25, and CD62L, followed by intracellular Foxp3 staining, according to the manufacturer's protocol. Matched IgG isotypes were used as negative controls. All stainings were conducted in phosphate buffered saline (PBS)/5% fetal calf serum for 20 minutes at 4°C. Flow cytometry was performed with an Epics Elite model flow cytometer (Beckman-Coulter), and data were analyzed with WinMDI software.

Mixed leukocyte reaction (MLR).

The myeloid DC and plasmacytoid DC populations were isolated from PBMCs obtained from patients with inactive RA, using the BDCA-1 (CD1c) and BDCA-4 (CD304) magnetic microbead DC isolation kits (Miltenyi Biotec), respectively, according to the manufacturer's instructions. The purity of the isolated populations ranged from 75% to 95%. To maintain cell viability, myeloid DCs and plasmacytoid DCs were cultured in flat-bottomed 96-well plates with 2 ng/ml GM-CSF and 1 ng/ml IL-3, respectively. After 1–2 days of culture, DCs were activated for 24 hours in the presence of 1 μg/ml anti-CD40L. CD4+CD25− T cells (responders) were isolated from heparinized cord blood, using CD4 and CD25 magnetic microbeads (Miltenyi Biotec), and were subsequently labeled with 2 μM CFSE in labeling buffer (PBS, 0.1% BSA) for 10 minutes at 37°C.

MLRs were set up by coculturing 104 myeloid DCs or plasmacytoid DCs with 2 × 104 allogeneic CFSE-labeled CD25− T cells in round-bottomed 96-well plates for 5–7 days, in the presence or absence of soluble anti-human CD3. The proliferation of T cells was determined by flow cytometry based on the dilution of CFSE. Culture supernatants were collected to assess the presence of cytokines. In other experiments, MLRs were performed in the presence of the specific antagonist of the IDO enzyme, 1-MT (1 mM). Where indicated, the MLR was performed in the presence of either anti–ICOS-L (10 μg/ml) or anti-TNFα and anti–IL-10 (10 μg/ml).

Cytokine detection.

DC-primed CD4+ T cells were collected and washed after 5 days of primary culture and were restimulated (106 cells/ml) with anti-human CD3/anti-human CD28 for 24 hours. The levels of IL-10 and interferon-γ (IFNγ) in the culture supernatants were measured using the respective human cytokine enzyme-linked immunosorbent assay kits (eBioscience). The detection limits were 15.65 pg/ml for IFNγ and 16.25 pg/ml for IL-10.

Analysis of IDO expression by real-time polymerase chain reaction (PCR).

Plasmacytoid DCs (2 × 105) isolated from either patients with inactive RA or healthy controls were resuspended in TRIzol reagent (Sigma-Aldrich), and total RNA was extracted using the RNeasy Mini Kit (Qiagen, Chatsworth, CA). RNA was then reverse-transcribed into complementary DNA (cDNA) using ThermoScript Reverse Transcriptase (Invitrogen).

Expression of the IDO enzyme by plasmacytoid DCs was determined by real-time PCR using gene-specific primers for the human gene IDO (forward 5′-CATGCTGCTCAGTTCCTCCAG-3′, reverse 5′-CAGAGCTTTCACACAGGCGTC-3′). The PCR was run with 5 μl undiluted cDNA template in 20-μl reactions, performed in duplicate, on an Applied Biosystems Prism 7000 SDS (Foster City, CA) using 1× iTaq SYBR Green Master Mix (Bio-Rad, Richmond, CA) and IDO primers at a concentration of 0.4 μM or 0.25 μM. Amplification of GAPDH with gene-specific primers (forward 5′-CATGTTCCAATATGATTCCACC-3′, reverse 5′-GATGGGATTTCCATTGATGAC-3′) was used as the reference. The temperature profile consisted of an initial step at 95°C for 10 minutes (for Taq activation), followed by 40 cycles of 95°C for 15 seconds and 56°C for 1 minute. A standard curve was determined by serially diluting a control cDNA sample. The relative quantification method with reference to the standard curve was then applied to evaluate the expression of the IDO gene after normalization to the expression values for the GAPDH reference gene.

Suppression assay.

DC-primed T cells (from a single donor) were collected and cultured with 2 × 104 naive autologous CD4+CD25− T cells (from the same donor) in the presence of 104 irradiated (40-Gy) allogeneic CD3-depleted monocytes (from a different donor) as APCs, in round-bottomed 96-well plates. Titration of DC-primed T cells was performed, resulting in naive T cell:DC-primed T cell ratios of l:1, 1:0.5, and 1:0.1. After 4 days, 1 μCi 3H-thymidine (Amersham Biosciences, Roosendaal, The Netherlands) was added for the last 18 hours of culture. Cells were harvested, and the incorporated radioactivity was measured using a Beckman-Coulter beta counter.

Statistical analysis.

The nonparametric Mann-Whitney U test was used for statistical comparisons between groups. Results are expressed as the mean ± SEM. P values (2-tailed) less than 0.05 were considered statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Increased frequencies of Foxp3+ Treg cells in patients with inactive RA.

We first sought to confirm whether Treg cells had been restored in the peripheral blood of patients with RA who had been successfully treated and achieved remission of their RA, as reported previously (17, 18). As shown in Figure 1A, the percentages of Foxp3+CD4+ T cells in patients with inactive RA were increased 2–3-fold compared with those in patients with active RA. Of interest, although the CD4+Foxp3+ T cells from all 3 groups were also CD25high (results not shown), they exhibited a distinct pattern of expression for the CD62L molecule. Thus, in patients with inactive RA, almost 34% of the CD4+Foxp3+ cells did not express CD62L, whereas in healthy controls and patients with active RA, the majority of the CD4+Foxp3+ cells (>85%) were CD62L+ (Figure 1B). Taken together, these findings support the notion of the de novo generation of Treg cells, rather than the expansion of these cells from preexisting Treg cells, in the peripheral blood of patients with therapy-induced remission of RA.

thumbnail image

Figure 1. Increase in the levels of Foxp3+CD4+ T cells in the peripheral blood of patients with rheumatoid arthritis (RA) responding to therapy. T cells were isolated from peripheral blood mononuclear cells obtained from healthy controls, patients with active RA, and patients with inactive RA, and assessed by fluorescence-activated cell sorting (FACS) analysis. A, Intracellular Foxp3 expression was determined by staining CD3+ T cells with fluorescence-conjugated antibodies against CD4 and Foxp3. Values are the percentages of gated CD4+Foxp3+ T cells. B, The gated CD4+Foxp3+ T cell population was also assessed by FACS analysis for the expression of the CD62L molecule. Representative results are shown for 10 healthy controls, 20 patients with active RA, and 15 patients with inactive RA.

Download figure to PowerPoint

Restoration of the levels of plasmacytoid and myeloid DCs in the peripheral blood of patients with RA achieving remission.

Since DCs have been linked to the induction of tolerance through the generation of Treg cells (21), we reasoned that DCs may be involved in the generation and/or expansion of Treg cells in patients with RA whose disease is in therapy-induced remission. To this end, CD3-depleted peripheral blood leukocytes were stained for the CD303 antigen (BDCA-2) and the CD1c antigen (BDCA-1), which are expressed on the surface of plasmacytoid DCs and the surface of myeloid DCs, respectively. The results were then analyzed by flow cytometry. As shown in Figure 2A, expression of myeloid DCs (CD45+CD1c+CD11c+CD303−) was severely curtailed in the peripheral blood of patients with active RA, in contrast to that in patients with inactive RA, in whom the myeloid DC levels were similar to those in healthy controls. The decrease in myeloid DC numbers in patients with active RA was statistically significant compared with that in either healthy controls or patients with inactive RA (Figure 2B). Similar to the findings regarding myeloid DCs, CD303+ cells were virtually absent (<0.1%) in the peripheral blood of patients with active RA, whereas the levels of CD303+ cells in patients who had achieved remission were restored (∼0.5%) and were comparable with the levels in healthy controls (∼0.7%) (Figure 2C).

thumbnail image

Figure 2. Restoration of the levels of myeloid and plasmacytoid dendritic cells (DCs) in the peripheral blood of patients with rheumatoid arthritis (RA) in remission. CD3-depleted peripheral blood mononuclear cells from healthy controls (n = 10), patients with active RA (n = 20), and patients with inactive RA (n = 15) were stained with myeloid DC (CD1c)– and plasmacytoid DC (CD303)–specific antibodies, and expression of these molecules in the peripheral blood was assessed by flow cytometry. A and B, The proportion of CD45+CD1c+CD11c+CD303− myeloid DCs (A) and the relative frequencies of CD1c+ myeloid DCs per 3 × 105 events (B) were determined. C and D, The fraction of CD45+CD303+CD1c−CD11c− plasmacytoid DCs (C) and the frequencies of CD303+ plasmacytoid DCs per 3 × 105 events (D) were determined. Each symbol in B and D represents the results from an independent experiment, and the horizontal lines indicate the mean. P values (bars) were calculated by Mann-Whitney U test. E, The CD45+CD303+ fraction of cells from patients with inactive RA (left) was also assessed to determine the percentage staining positive (upper right) or negative (lower right) for CD123 (interleukin-3 receptor α). Results in E are representative of 1 of 5 independent experiments.

Download figure to PowerPoint

We also determined the relative frequencies of CD45+CD303+CD1c−CD11c−CD3− plasmacytoid DCs in the peripheral blood of patients with active RA and patients with inactive RA and found that the numbers of plasmacytoid DCs (CD303+ cells) in patients with active RA (mean 250 per 3 × 105 cells) were significantly decreased as compared with those in either patients with inactive RA (mean 1,700 per 3 × 105 cells; P < 0.005) or healthy individuals (mean 2,100 per 3 × 105 cells; P < 0.05) (Figure 2D).

To further confirm that CD303 is specifically expressed on plasmacytoid DCs, we performed costaining for CD123 (IL-3 receptor α), which is commonly used as a marker for plasmacytoid DCs. As shown in Figure 2E, the entire population of CD303+ cells (>96%) also stained positive for CD123.

The decrease in DCs in the peripheral blood of patients with active RA prompted us to explore whether some of these cells may be accumulating at the site of inflammation within the joint. An increased accumulation of myeloid DCs (∼15%) in the SF of patients with active RA was observed. In contrast, we failed to detect CD303+ plasmacytoid DCs in the SF of the same RA patients (results not shown).

Induction of IL-10–secreting T cells from naive CD4+ T cells by mature plasmacytoid DCs in patients with RA responding to therapy.

We next sought to determine whether plasmacytoid DCs isolated from RA patients showing a response to therapy have the ability to induce T cells with a regulatory phenotype. To this end, myeloid DCs and plasmacytoid DCs were isolated from the peripheral blood of patients with inactive RA and cultured in the presence of GM-CSF and IL-3, respectively, and then matured with CD40L for 24 hours. The maturation status of the DC subsets was confirmed on the basis of the expression of HLA–DR, CD80, CD86, CD40, and the chemokine receptor CCR7 (results not shown).

To detect the ability of CD40L-matured myeloid and plasmacytoid DCs to activate and expand naive T cells, we cultured CFSE-labeled naive CD4+CD25− T cells, which were isolated from cord blood, in the presence of CD40L-matured plasmacytoid DCs and myeloid DCs, and monitored cell proliferation based on dilution of the CFSE staining after 6 days. Naive T cells cocultured with mature plasmacytoid DCs were hyporesponsive, as demonstrated by their delayed proliferation, in contrast to that seen in cocultures with mature myeloid DCs, in which a robust expansion of naive T cells occurred (Figure 3A). Specifically, more than 90% of the myeloid DC–primed T cells progressed to the third division, whereas only 40% of the T cells cultured with plasmacytoid DCs entered the third division (gate M3 in Figure 3A). This was further confirmed on the basis of the increased expression of the early activation marker CD69 on CD4+ T cells. Thus, the majority (>90%) of the CFSE−CD4+ T cells cultured with myeloid DCs were CD69− after 6 days of culture, whereas almost 35% of the plasmacytoid DC–primed T cells were in the early activation stage (CFSE+CD69+) (Figure 3B).

thumbnail image

Figure 3. Induction of interleukin-10 (IL-10)–secreting T cells by mature plasmacytoid dendritic cells (pDCs) isolated from patients with rheumatoid arthritis (RA) in remission. Myeloid DCs (mDCs) and pDCs isolated from the peripheral blood mononuclear cells of patients with inactive RA were cultured with 2 ng/ml granulocyte–macrophage colony-stimulating factor and 1 ng/ml IL-3, respectively, and matured upon culture with 1 μg/ml anti-CD40L for 24 hours. Allogeneic naive 5,6-carboxyfluorescein succinimidyl ester (CFSE)–labeled CD4+CD25− T cells were isolated from cord blood and were cultured with either mature mDCs or mature pDCs in a DC-to–T cell ratio of 1:2. As the control, CFSE-labeled naive T cells were used. The proliferation was optimized using soluble human purified anti-CD3 (50 ng/ml). A, Histograms show the dilution of CFSE staining in CD4+CD25− T cells after 6 days of coculture with DC subsets. Gates depict individual cell cycles. B, The surface expression of the CD69 molecule was assessed on T cells cultured with mDCs or pDCs. The values indicate quadrant percentages. C, DC-primed T cells were collected after 6 days of coculture and restimulated with 5 μg/ml of plate-bound anti-CD3 (aCD3) and 1 μg/ml of soluble anti-CD28 monoclonal antibodies. Culture supernatants were collected 24 hours later and assessed by cytokine enzyme-linked immunosorbent assay for the presence of interferon-γ (IFNγ) and IL-10. The detection limit of the assay was 15.65 pg/ml for IFNγ and 16.25 pg/ml for IL-10. Bars show the mean and SEM representative results from 1 of at least 3 independent experiments.

Download figure to PowerPoint

We next assessed the ability of DC-primed T cells to secrete cytokines upon in vitro restimulation with anti-human CD3/anti-human CD28. T cells cultured in the presence of myeloid DCs produced significant amounts of IFNγ (250–500 pg/ml), whereas T cells cultured with plasmacytoid DCs produced lower amounts of IFNγ (80–100 pg/ml) (Figure 3C). Most notably, plasmacytoid DC–primed T cells secreted large amounts of IL-10 (800–1,200 pg/ml), in contrast to myeloid DC–primed T cells, in which IL-10 was not detectable (Figure 3C). Collectively, these results show that T cells cultured in the presence of plasmacytoid DCs from patients with RA in remission proliferate poorly and secrete significant amounts of IL-10 in response to polyclonal stimulation, indicating that they have acquired features of Treg cells.

Suppression of naive CD4+ T cell proliferation by mature plasmacytoid DC–induced IL-10–producing Treg cells in patients with RA in remission.

To determine whether IL-10–producing plasmacytoid DC–primed T cells are Treg cells, plasmacytoid DC–primed T cells were cultured with CFSE-labeled autologous naive CD4+CD25− T cells in the presence of irradiated allogeneic CD3-depleted monocytes as APCs; myeloid DC–primed T cells were used as the control. As shown in Figure 4A, plasmacytoid DC–primed T cells inhibited the proliferation of the autologous naive T cells, in a dose-dependent manner. In contrast, the addition of myeloid DC–primed T cells into the culture resulted in an increased proliferation of the autologous naive T cells, when compared with that in cultures with naive T cells alone (Figure 4B). These results indicate that mature plasmacytoid DCs in patients with RA in remission have an intrinsic ability to induce the generation of Treg cells that display a potent suppressive activity on autologous T cells in vitro.

thumbnail image

Figure 4. Suppression of autologous T cell proliferation by plasmacytoid dendritic cell (pDC)–primed T cells in vitro. Naive CD4+CD25− T cells from the peripheral blood of a single donor were cocultured with either pDCs (A) or myeloid DCs (mDCs) (B) from patients with rheumatoid arthritis in remission, and then matured with CD40L. After 7 days of culture, CD4+ T cells were separated from DCs and used for functional assays, in which 2 × 104 freshly purified autologous CD4+CD25− T cells from the same donor were cultured with 104 irradiated antigen-presenting cells from a second donor, and the DC-primed T cells from the first donor were added at graded doses. T cell proliferation was assessed in a 4-day mixed leukocyte reaction assay, with3H-thymidine added during the last 18 hours of culture. Bars show the mean ± SEM representative results from 1 of 4 independent experiments.

Download figure to PowerPoint

Increased expression of IDO mRNA, but not ICOS-L and PDL-1 molecules, by mature plasmacytoid DCs in patients with inactive RA.

Several lines of evidence indicate an important role of negative costimulatory molecules, such as ICOS-L and PDL-1 (22, 23), as well as the expression of the enzyme IDO (24) by DCs in the induction of Treg cells. As shown in Figure 5A, plasmacytoid DCs from patients with inactive RA and healthy controls expressed comparable levels of either ICOS-L or PDL-1 upon maturation, suggesting that the expression of these molecules most likely does not account for the polarization of T cells by plasmacytoid DCs toward Treg cells in these patients. In contrast, plasmacytoid DCs from patients with RA in remission displayed a 4–8-fold increase in IDO mRNA levels as compared with that in healthy individuals, independent of the maturation status of the plasmacytoid DCs (Figure 5B).

thumbnail image

Figure 5. Induction of interleukin-10 (IL-10)–producing Treg cells by plasmacytoid dendritic cells (pDCs) in an indoleamine 2,3-dioxygenase (IDO)–dependent manner. A, Surface expression of inducible costimulator ligand (ICOS-L) and programmed death ligand 1 (PDL-1) on mature pDCs from patients with inactive rheumatoid arthritis (RA) (solid line) and healthy controls (shaded line). Results obtained with isotype-matched control antibody are also shown (filled histogram). B, IDO mRNA expression on pDCs from the same patients with inactive RA and healthy controls as in A. IDO-specific mRNA levels were quantified by quantitative reverse transcription–polymerase chain reaction and normalized to GADPH mRNA levels. Bars show the mean and SD results with or without maturation by CD40L. C, Expression levels of myeloid DCs (mDCs) (left) and pDCs (right) from patients with inactive RA in cocultures with allogeneic 5,6-carboxyfluorescein succinimidyl ester (CFSE)–labeled CD4+CD25− naive T cells, in the presence (open histogram) or absence (filled histogram) of the IDO-specific antagonist 1-methyl-DL-tryptophan (1-MT). CFSE dilution of the depicted CD4+ gated cells is shown after 5 days of culture. D, Surface expression of CD69 on CFSE+CD4+CD25− T cells cultured with pDCs, with (right) or without (left) 1-MT. The values indicate quadrant percentages. E, Secretion of IL-10 in DC-primed T cells in the presence or absence of 1-MT. IL-10 levels were measured by enzyme-linked immunosorbent assay in the supernatants collected 24 hours after restimulation. Bars show the mean and SEM representative results of 1 of at least 3 independent experiments.

Download figure to PowerPoint

Induction of IL-10–producing Treg cells by plasmacytoid DCs in an IDO-dependent manner.

To examine whether the expression of IDO by mature plasmacytoid DCs in patients with RA in remission is required for the differentiation of naive T cells into IL-10–producing Treg cells, we set up cocultures of naive T cells and DCs in the presence or absence of the IDO-specific inhibitor 1-MT. As shown in Figure 5C, mature myeloid DCs induced a robust proliferation of naive CD4+CD25− T cells, irrespective of the addition of 1-MT in vitro. In cocultures with plasmacytoid DCs, naive T cells proliferated poorly when cultured with plasmacytoid DCs alone (Figures 3A and 5C), whereas this phenotype was altered in the presence of the IDO inhibitor, since a strong proliferation of naive T cells occurred after inhibition of IDO (Figure 5C).

The reversal in plasmacytoid DC–induced T cell proliferation in the presence of the IDO inhibitor was further confirmed by measuring the expression of CD69 by proliferating T cells. In the absence of 1-MT inhibition, only 40% of the CD4+ T cells divided (CFSE−CD4+) and down-regulated the CD69 molecule (Figure 5D). In contrast, with the addition of the IDO inhibitor, more than 90% of the cells entered cell division, with the majority of cells (∼86%) being CFSE−CD4+CD69−; only 6% of the cells were activated (CFSE+CD4+CD69+) but not dividing (Figure 5D).

Furthermore, T cells primed with myeloid DCs with or without the addition of 1-MT did not produce significant amounts of IL-10 (<50 pg/ml) (Figure 5E). In contrast, plasmacytoid DC–primed T cells secreted large amounts of IL-10 upon restimulation (1,400–1,750 pg/ml), and the addition of the IDO inhibitor during plasmacytoid DC–T cell cocultures effectively abolished the secretion of IL-10 by T cells (180–250 pg/ml) (Figure 5E). In contrast, blocking of the ICOS pathway or blocking of the secretion of cytokines by DCs using neutralizing antibodies did not affect the ability of plasmacytoid DCs to polarize naive T cells toward Treg cells (detailed results available from the corresponding author upon request). Taken together, these findings suggest that plasmacytoid DCs from RA patients whose disease has responded to therapy have the capacity to polarize naive T cells toward Treg cells that secrete high levels of IL-10 in an IDO-dependent manner.

Increased levels of IL-10–producing T cells in the peripheral blood of patients with inactive RA.

To determine whether disease remission is associated with an increase in IL-10–producing T cells in the peripheral blood, we examined the secretion of IL-10 upon polyclonal stimulation of CD4+CD25− T cells isolated from patients with inactive RA, patients with active RA, and healthy controls. Although both T cells from patients with active RA and those from healthy individuals showed increased levels of IL-10 upon stimulation, the levels of IL-10 secreted from stimulated T cells from patients with RA in remission were 3–4-fold higher (details available from the corresponding author upon request). Collectively, these results confirm that patients with RA in remission display increased numbers of IL-10–producing T cells, a proportion of which might have been induced upon interaction with the IDO-expressing plasmacytoid DCs.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

In this study, we have shown that the remission of RA is associated with the restoration of plasmacytoid DC levels in the peripheral blood, and we have provided evidence of an important role for this DC subset in the maintenance of tolerance through the induction of IL-10–secreting Tr1 regulatory cells. Amelioration of disease activity in RA patients resulted in the reestablishment of the levels of both plasmacytoid DCs and myeloid DCs in the circulation. Because plasmacytoid DCs migrate to the sites of inflammation, we sought to determine whether the lack of plasmacytoid DCs in the peripheral blood of patients with active RA may be due to the recruitment of plasmacytoid DCs at the synovium. Although we were unable to detect CD303+ plasmacytoid DCs in the SF of patients with inactive RA, plasmacytoid DCs may still be present at the synovial membrane (25).

There is ample evidence for the functional plasticity of plasmacytoid DCs in directing T cell responses, which mainly depends on their maturation status (26, 27). Thus, virus-infected plasmacytoid DCs elicit a potent Th1 response (28), whereas IL-3–treated plasmacytoid DCs prime Th2 cells in an OX40 ligand–dependent mechanism (29). In addition, plasmacytoid DCs induce Treg cell–mediated differentiation of naive T cells, which depends on the activation stimulus used (4, 5, 13). Our findings are consistent with those in previous studies in that our results also support the notion that amelioration of an autoimmune disease in humans may be associated with plasmacytoid DC–induced Treg cell development. Based on these observations, we speculate that in patients with active RA, plasmacytoid DCs are exposed to a cytokine milieu and adapt to a tolerogenic phenotype. During disease remission, tolerogenic plasmacytoid DCs are restored in the peripheral blood, where they are able to polarize naive T cells toward IL-10–secreting Treg cells. Although it is not clear whether remission of RA is directly mediated through plasmacytoid DC–primed Treg cells, our results demonstrate that IL-10–secreting Treg cells can potently suppress the activation of naive T cells in vitro, implying a possible role in the homeostatic control of inflammation in vivo.

Both naturally occurring Treg cells and Tr1 cells have a crucial role in the induction and maintenance of tolerance to self and foreign antigens (30). Our results indicate that T cells primed by mature plasmacytoid DCs from patients with RA in remission have the cardinal features of Tr1 regulatory cells (31), since they secrete high levels of IL-10 and low levels of IFNγ, and they proliferate poorly following polyclonal T cell receptor–mediated stimulation. Most notably, they suppress the activation and proliferation of autologous naive CD4+ T cells, in a dose-dependent manner. Previous studies have convincingly shown that treatment of RA patients with infliximab led to the induction of Foxp3+CD25highCD62L− Treg cells that were distinct from natural Treg cells and Tr1 cells (18). Although we confirmed the presence of Foxp3+CD4+CD62L− T cells in patients with inactive RA (Figure 1A), we were unable to detect Foxp3 expression by plasmacytoid DC–primed IL-10–producing Tr1 cells (results not shown). These results suggest that restoration of tolerance during an autoimmune disease may be mediated by distinct types of Treg cells with specialized functions. Support for this hypothesis has been provided in mouse models, in which distinct types of Treg cells could act in concert to suppress autoimmune phenomena (32, 33).

Secretion of high levels of IL-10 by plasmacytoid DC–primed T cells could have a beneficial role during induction and/or maintenance of remission in these patients. Therefore, the plasmacytoid DC–mediated secretion of IL-10 by Tr1 cells in patients with RA in remission may exert pleiotropic effects toward maintenance of immune tolerance. Secretion of IL-10 by plasmacytoid DC–primed Treg cells might also increase the regulatory activity of other Treg cell subsets (34) or restore the suppressive capacity of Foxp3+ Treg cells that were previously defective. In addition, IL-10 can indirectly inhibit immune responses by abolishing the antigen-presenting ability of APCs, either through down-regulation of the expression of the major histocompatibility complex and costimulatory molecules or through inhibition of cytokine secretion by these cells (35). Consistent with our findings, an increase in serum IL-10 levels has been demonstrated in patients who have exhibited a prolonged clinical response upon anti-TNFα treatment (36, 37).

Our results clearly demonstrate that IDO expression by mature plasmacytoid DCs isolated from patients with RA in remission was required for the induction of IL-10–producing Treg cells. Blockage of IDO activity by the inhibitor 1-MT completely reversed the T cell polarization. This is in line with mouse and human studies demonstrating the crucial role of IDO in Treg cell induction (13–15, 38–42). This reversal was IDO-specific, since blockage of the ICOS pathway or neutralization of cytokines such as TNFα and IL-10 did not affect the ability of plasmacytoid DCs to polarize naive T cells toward IL-10–producing Treg cells. Recent study findings have suggested that the ICOS pathway is essential for the induction of anergic T cells with a regulatory phenotype in immature myeloid DCs (43), and therefore it might not be required for the induction of Tr1 cells.

Although the exact molecular mechanism that links Treg cell induction with IDO-expressing plasmacytoid DCs is still unresolved, it has been proposed that interactions between Treg cells and plasmacytoid DCs might initiate a so-called reverse-signaling mechanism that leads to the activation of NF-κB signaling and up-regulation of IDO (24, 44). Subsequently, the combined effects of IDO expression and tryptophan depletion may induce Treg cells from naive T cells (45). What drives the up-regulation of IDO on plasmacytoid DCs in patients with RA who have experienced disease remission remains to be determined. IDO is strongly up-regulated on plasmacytoid DCs that have been exposed to IFNγ, a cytokine present during the course of RA (46). We postulate that plasmacytoid DCs exposed to the cytokine milieu in RA increase the expression of IDO and, upon contact with naive T cells, mediate the conversion of these cells toward Treg cells.

In summary, we have shown that plasmacytoid DCs isolated ex vivo from patients with RA in remission display increased IDO expression and prime naive CD4+ T cells to differentiate in vitro into IL-10–producing Tr1 regulatory cells that potently suppress the proliferation of autologous naive CD4+ T cells. Amelioration of the disease was correlated with the reappearance of circulating plasmacytoid DCs in RA patients, raising the possibility that IDO-expressing plasmacytoid DC–induced Treg cells might be involved, at least in part, in the restoration of immune homeostasis. It is tempting to speculate that plasmacytoid DCs exposed to an autoimmune environment have been conditioned to express a tolerogenic phenotype that induces the de novo generation of Treg cells with the ability to control ongoing immune responses. A better understanding of the pathways involved in the plasmacytoid DC–Treg cell crosstalk will facilitate the exploration of the potential use of cell-based therapies for the induction of tolerance in transplantation and autoimmunity.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Verginis 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.

Study conception and design. Verginis.

Acquisition of data. Kavousanaki, Makrigiannakis, Verginis.

Analysis and interpretation of data. Kavousanaki, Makrigiannakis, Boumpas, Verginis.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We thank H. Kritikos, P. Sidiropoulos, and E. Houstoulaki for patient care and the collection of samples, G. Bertsias and I. Tassiulas for critical review of the manuscript, E. Koutala and C. Choulaki for technical assistance, and all of the patients and healthy volunteers for participating in the study.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES