SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

Dendritic cells (DC) are professional antigen-presenting cells that are capable of both activating immune responses and inducing tolerance. Several studies have revealed efficiency of therapeutic vaccination with tolerogenic DC (tolDC) in inhibition of experimental autoimmunity. The purpose of this study was to compare four different protocols for generation of tolDC – the antidiabetic drug troglitazone (TGZ DC), NF-κB inhibitor BAY 11-7082 (BAY DC), prostaglandin D2 metabolite 15d-PGJ2 (PGJ DC) and a combination of dexamethasone and 1α,25-dihydroxyvitamin D3 (DexVD3 DC) regarding phenotype, cytokine production and T cell stimulatory capacity. TGZ DC and BAY DC had a phenotype comparable to immature DC, while DexVD3 DC were more macrophage like. Analysis of cytokine production using cell culture supernatants from all DC populations revealed that DexVD3 DC were efficient producers of IL-10 and produced less pro-inflammatory cytokines. T cells primed with DexVD3 DC showed reduced proliferation, and further analyses of these T cells revealed that functionally effective type 1 regulatory T cells (Tr1) but not FoxP3+ Treg were induced. Furthermore, DexVD3 DC promoted the induction of regulatory B cells (Breg). Together, these results indicate that DexVD3 DC have the best potential to be used in a tolerogenic antigen-presenting cell-based immunotherapy setting.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

Autoimmune diseases have a complex aetiology and affect a large proportion of the human population [1]. Most of the diseases are chronic, lead to serious disabilities and suffering of the patient together with absence from work. Despite the use of B cell depleting antibodies and cytokine antagonists, a drug-free remission is rarely observed. Therefore, there is still a need for individualized and more effective lasting treatment. Therapeutic vaccination of patients with autologous monocyte-derived tolerogenic dendritic cells could be one of the clinical modalities [2].

Dendritic cells (DC) are professional antigen-presenting cells (APC) that bridge the innate and adaptive immune system [3]. In addition to the ability to recognize and eliminate ‘foreign’, DC are capable to circulate through tissues and lymphoid organs, taking up and presenting peptides from apoptotic cells and other self-antigens to T cells [4]. Through this mechanism, they play a critical role in the continuous induction of peripheral tolerance, thereby preventing hyperreactivity and autoimmunity [5, 6].

Tolerogenic DC (tolDC) are characterized by low production of pro-inflammatory and high production of anti-inflammatory cytokines, reduced expression of costimulatory molecules, high levels of inhibitory molecules and the ability to induce regulatory T cells (Treg) [7]. Several Treg subsets have been described, including naturally occurring Treg (CD4+CD25+FoxP3+) [8] and IL-10-producing T regulatory type 1 (Tr1) cells [9, 10]. In contrast to murine data, the role of FoxP3 as a specific marker for human Treg is controversial, and expression of FoxP3 does not necessarily correlate with suppressive activity of CD4+ T cells [11, 12]. Therefore, tolDC studies that refer only to FoxP3+ T cells might lack substantial data. Besides induction of Treg, a recent clinical trial indicated that tolDC might promote regulatory B cells (Breg) [13]. Although the data on human Breg are limited, they could be considered as one of the targets of tolDC therapy because they have been shown to promote Tr1 cells and suppress Th17-mediated inflammation in mice [14].

So far, several protocols for the in vitro generation of tolDC from monocytes have been described. These include genetic modifications or the use of pharmacological compounds, including IL-10 [15], NF-κB inhibitors [16], vitamin D receptor agonists [17] and PPAR-γ agonists [18]. To date, there are two published studies comparing some of these protocols but coming to different conclusions as to the optimal compound required for the generation of tolDC [19, 20]. Despite the thorough methodology and GMP conditions, both studies used different settings, different compounds and different read-out systems complicating the collation between them. In addition to that, none of them studied the treatment of DC with the combination of dexamethasone and 1alpha,25-dihydroxyvitamin D3 (DexVD3 DC). These drugs were recently shown to induce tolDC with good migratory capacity and to be effective in inhibition of collagen-induced arthritis in a mouse model [21]. Furthermore, the combination was approved for use in a clinical trial (AUTODECRA NCT01352858) at the Newcastle University.

In the current study, we addressed the mechanisms responsible for the tolerogenic activity of DexVD3 DC and compared this protocol for tolDC generation with other pharmacological compounds suitable for clinical trials – the antidiabetic drug troglitazone, a PPAR-γ agonist, (TGZ DC), NF-κB inhibitor BAY 11-7082 (BAY DC) and prostaglandin D2 metabolite 15d-PGJ2 (PGJ DC). DexVD3 DC induced the highest amounts of functional Tr1 cells as well as Breg making them superior candidates for the use in a tolDC-based immunotherapy setting.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

Generation of tolDC

Dendritic cells were generated from monocytes isolated from buffy coat preparations from healthy blood donors (Blood Bank, Haukeland University Hospital, Bergen, Norway) as previously described [22]. Briefly, peripheral blood mononuclear cells were separated by a density gradient centrifugation and the monocytes were then isolated by plastic adherence in X-VIVO20 medium (Cambrex Bioscience, Verviers, Belgium). The monocytes were cultured in RPMI medium (Cambrex Bioscience, Verviers, Belgium) supplemented with 10% FCS (PAA, Pasching, Austria), 2 mm glutamine and antibiotics (100 U/ml penicillin and 100 μg/ml streptomycin (Sigma-Aldrich, St. Louis, MO)) at 1 × 106 cells/ml in the presence of IL-4 (20 ng/ml; ImmunoTools, Friesoythe; Germany) and GM-CSF (100 ng/ml; ImmunoTools, Friesoythe; Germany) for 6 days. Cytokines were replenished every 2–3 days. TGZ DC, BAY DC and PGJ DC were generated by addition of 10 μm troglitazone, 10 μm 15d-PGJ2 or 1.5 μm BAY 11-7082 at days 3 and 6. DexVD3 DC were generated by addition of 1 μm dexamethasone (Dex) at day 3 and Dex plus 0.1 nm 1α,25-dihydroxyvitamin D3 at day 6. Because DMSO was used as a solvent for all compounds, the equivalent amount of DMSO was added to a control group (DMSO DC) on days 3 and 6. On day 6, a part of all populations were stimulated with LPS (100 ng/ml) for 24 h. Cells were harvested on day 7.

All compounds were manufactured by Enzo Life Sciences, Laussen, Switzerland, except for DMSO and dexamethasone (Sigma-Aldrich, St. Louis, MO).

Flow cytometry

Immunostaining was performed as described previously [22]. Briefly, after 5-min incubation with Fc receptor (FcR) block (Miltenyi, Germany), cells were stained with a titrated amount of antibodies for 10 min at room temperature before being washed and immediately analysed on a FACSCanto I (DC phenotyping) or LSRFortessa (proliferation and coculture experiments) cytometer (BD Biosciences, Heidelberg, Germany). For the intracellular staining, the cells were fixed using a Cytofix/Cytoperm fixation solution (BD Biosciences, Heidelberg, Germany) according to recommendations of the manufacturer. All subsequent analyses were performed with FlowJo software (Tree Star, Ashland, OR). One per cent false-positive events were accepted in the negative controls. The antibodies used were CD1-PE (NA1/34-HLK), CD14-FITC (UCHM1), HLA-DR-APC (HL-39), CD38-Alexa Fluor 647 (AT13/5), CD8-PE (LT8), CD86-FITC (BU63), CD83-PE (HB15e), CD80-APC (MEM-233), CD40-FITC (LOB7/6), CD274-FITC (MIH6), CD273-Alexa Fluor 647 (MIH14), all from AbD Serotec (Düsseldorf, Germany); CD19 PE-Cy7 (SJ25C1), IL-10-APC (JES3-19F1), CD25-Alexa Fluor 700 (M-A251) and FoxP3-PerCP-Cy5.5 (236a/E7) from BD Biosciences (Heidelberg, Germany); CCR7-PE, (150503), DC-SIGN-PE (120507) from R&D Systems (Minneapolis, MN); CD4-APC (MEM-241) from ImmunoTools (Friesoythe, Germany).

Cytokine determination

Cell-free supernatants were stored in aliquots at −20 °C until analysis. Sandwich ELISA was performed on culture supernatants from the DC populations to determine IL-12p70 and IL-10 (both from BioLegend, San Diego, CA) secretion according to the recommendations of the manufacturer. Supernatants from DC cultures as well as DC-NAC cocultures were analysed with a 25-plex Luminex assay cytokine and chemokine kit (Invitrogen, Carlsbad, CA) and run on a Luminex 100 System (Luminex Corporation, Austin, TX) according to the manufacturer's instructions.

DC stimulatory Capacity

To analyse the T cell stimulatory capacity of the various tolDC populations, an autologous mixed lymphocyte reaction (MLR) was performed using tuberculin-purified protein derivative (PPD, Statens Serum Institutt, Copenhagen, Denmark) as a recall antigen. To this end, autologous PBMC depleted for monocytes (non-adherent cells, NAC) were thawed and allowed to rest overnight before being labelled with CellTrace Violet Cell Proliferation Kit (Invitrogen, Carlsbad, CA) according to the manufacturer's recommendations. Two hundred thousand CellTrace Violet-labelled lymphocytes were then cocultured with forty thousand autologous tolDC previously incubated with PPD (1 μg/ml) for 24 h. After 5 days, the cells were harvested, stained for CD4 and CD8 and analysed on a LSRFortessa flow cytometer.

Suppression experiments

To analyse the suppressive capacity of T cells primed with the various tolDC populations and the induction of regulatory T and B cells, autologous NAC were thawed and allowed to rest overnight before priming with tolDC (1:6 ratio) for 5 days. Then, the non-adherent lymphocytes were harvested, washed and rested for another 5 days. After the rest, the cells were harvested, washed, counted and divided into 2 parts – one was phenotyped on a LSRFortessa flow cytometer and another was used for suppression experiments.

Dendritic cell-primed autologous NAC were labelled with CellTrace Violet Cell Proliferation Kit (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. tolDC and autologous NAC were thawed and allowed to rest overnight. Then, the naïve NAC (responder cells) were labelled with CFDA-SE (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions to prevent convergence with DC-primed NAC. Responder cells were incubated with DC-primed cells and in the presence of DMSO DC or tolDC (ratio 2:1:0.2).

In some experiments, the DC-primed autologous NAC were depleted of CD25+ cells using a CD25 MicroBeads II kit (Miltenyi, Germany) according to manufacturer's instructions prior to coculture with naïve NAC and DC to test the effect of CD25+FoxP3+ cells on the proliferation.

All coculture experiments and resting phases were carried out in X-VIVO20 medium supplemented with IL-2 (50 U/ml; ImmunoTools, Friesoythe; Germany).

Statistical analysis

Mann–Whitney U-test was used for statistical analyses. Significance was set at P < 0.05. To calculate differences in surface marker expression levels, DexVD3 DC were compared to the other DC populations (unstimulated vs. unstimulated and LPS stimulated vs. LPS stimulated). All statistical calculations were performed with Prism 5 (GraphPad Software, Inc., USA).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

DexVD3 DC have a macrophage-like phenotype and express high amounts of inhibitory molecules

The phenotype of the different tolDC populations was analysed by flow cytometry (Fig. 1). TGZ DC and BAY DC had a similar phenotype comparable to immature DC, characterized by high expression of CD1a and low/absent expression of CD14, while DexVD3 DC had a more macrophage-like phenotype with high expression of CD14 and reduced CD1a on the cell surface. PGJ DC also retained CD14 expression together with CD1a. Upon stimulation with LPS, all DC populations had increased surface expression of MHC class II molecules and CD40. DexVD3 DC had very low level of CD86 expression, which did not increase after LPS stimulation (P < 0.001 for both unstimulated and stimulated populations). CD274 (PD-L1) was expressed by all DC groups at a similar level before stimulation and was upregulated after the stimulation. However, DexVD3 DC had the lowest expression level of CD274 (P ≤ 0.01 for all groups except for stimulated PGJ DC). DexVD3 DC showed the highest level of expression of CD273 (PD-L2, P ≤ 0.01 for both unstimulated and stimulated populations), except for mature DMSO DC and CD80 (P < 0.01 for stimulated populations). Surface molecules involved in migration (CCR7 and CD38) were expressed at similar levels by all groups and upregulated after LPS stimulation (data not shown).

image

Figure 1. DexVD3 DC have a macrophage-like phenotype and express high levels of inhibitory molecules. The percentage positive cells or median fluorescence intensity (MFI) is shown. MFI was used when all DC populations showed more than 90 per cent positivity. The distribution is shown by the max/min as well as the 25, 50 and 75 quartiles and represents 5 independent experiments. DMSO−, immature control DC; DMSO+, LPS-stimulated control DC; BAY−, immature DC generated in the presence of NF-κB inhibitor BAY 11-7082; BAY+, LPS-stimulated DC generated in the presence of BAY 11-7082; TGZ−, immature DC generated in the presence of PPAR-γ agonist troglitazone; TGZ+, LPS-stimulated DC generated in the presence of troglitazone, DexVD3−, immature DC generated in the presence of dexamethasone and 1alpha, 25-dihydroxyvitamin D3; DexVD3+, LPS-stimulated DC generated in the presence of dexamethasone and 1alpha,25-dihydroxyvitamin D3; PGJ−, DC generated in the presence of prostaglandin D2 metabolite 15d-PGJ2; PGJ+, LPS-stimulated DC generated in the presence of 15d-PGJ2.

Download figure to PowerPoint

DexVD3 DC are effective IL-10 producers and favour anti-inflammatory environment

First, we analysed the production of IL-12p70 and IL-10 using sandwich ELISA. None of the DC populations produced IL-12p70 without LPS stimulation, but only DexVD3 DC and PGJ DC did not produce IL-12p70 after LPS stimulation (Fig. 2). Interestingly, the anti-inflammatory cytokine IL-10 was present in supernatants of all DC populations without LPS stimulation and increased after the addition of it. DexVD3 DC produced significantly higher amounts of IL-10 both in unstimulated and stimulated state compared to other tolDC populations as well as immature and mature DMSO DC (P < 0.01 for both unstimulated and stimulated populations).

image

Figure 2. DexVD3 DC are effective IL-10 producers. DexVD3 DC secrete significantly higher quantities of anti-inflammatory cytokines (P < 0.01 for both unstimulated and stimulated state) and do not produce IL-12p70 upon LPS stimulation. The amounts of IL-10 and IL-12p70 in the supernatants were analysed using sandwich ELISA. The concentrations are given in pg/ml, and distribution is shown by the max/min as well as the 25, 50 and 75 quartiles, n = 5. See Fig. 1 for definitions.

Download figure to PowerPoint

Next, the supernatants from LPS-stimulated DC populations as well as immature DMSO DC were tested using the 25-plex Luminex assay (Figure S1). DexVD3 DC, unlike other tolDC groups and similarly to immature DMSO DC, produced less pro-inflammatory cytokines IL-5, IL-6 and TNF-α. Similar to ELISA results, supernatants from DexVD3 DC had the highest amounts of IL-10. Interestingly, we detected relatively high levels of IL-8 in DexVD3 DC supernatants that might be favourable for migration of these cells.

DexVD3 DC have low T cell stimulatory capacity

To understand the functional effect of phenotype and cytokine environment of the different tolDC populations, we tested their ability to stimulate autologous T cells using PPD as a recall antigen. The PPD-primed DC were cocultured with autologous NAC, and the level of proliferation was measured by the dilution of CellTrace Violet. Mature control DMSO DC induced up to 41.8% T cell proliferation (median 33.6%, Fig. 3A), while the proliferation of T cells cocultured with BAY DC, TGZ DC and PGJ DC as well as immature DMSO DC was around 20–25% (BAY DC: median 25.5%; TGZ DC: median 23%; PGJ DC: median 23.4%; immature DMSO DC: median 23.9%). Consistent with their phenotype and cytokine profile, DexVD3 DC induced significantly less proliferation of T cells compared to immature or mature DMSO DC with only up to 18% (median 16.2%; P = 0.0286).

image

Figure 3. DexVD3 DC have a low T cell stimulatory capacity (A). PPD-primed DC were cocultured with CellTrace Violet-labelled autologous NAC (1:5 ratio), and the level of proliferation was measured by the dilution of CellTrace Violet resulting in reduction in MFI after 5 days. Data are presented as median ± SD and represent 3 independent experiments. Dashed line is set at the median level of proliferation induced by immature DC (23.9%). (B) After the coculture with various tolDC populations for 5 days, NAC were harvested, washed, counted, plated at 1 × 106 cells/well and rested for 5 days in the presence of IL-2 (50 U/ml). Subsequently, the cells were harvested and counted using a Casy cell counter. NAC cocultured with all DC groups, except DexVD3 DC, proliferated intensively during the rest phase. The dashed line indicates the amount of NAC before rest, that is, 1 × 106. Data are presented as median ± SD, n = 5. The tolDC population that was used to stimulate NAC before the rest phase is shown on the x-axis. (C) DexVD3 DC-primed NAC suppress proliferation of T cells induced by mature DMSO DC. Naïve NAC (responder cells labelled with CFSE) were cocultured with different tolDC-primed NAC (effector cells labelled with CellTrace Violet) in the presence of PPD-loaded mature DMSO DC (ratio 2:1:0.2) for 5 days. CellTrace Violet positive cells were gated out, and the level of proliferation of CFSE-labelled responder cells was evaluated as reduction in MFI. DexVD3 DC-primed NAC significantly inhibited proliferation of responder cells when compared to proliferation in presence of mature DMSO DC-primed NAC. Data are presented as median ± SD, n = 3. The gating strategy is presented in Figure S2. The DC population that was used to stimulate the effector cells is shown on the x-axis. See Fig. 1 for definitions.

Download figure to PowerPoint

Furthermore, the effect of DexVD3 DC on T cell proliferation was similar even after the removal of DC from culture and resting of the remaining cells in the presence of IL-2. The cell number was evaluated after a 5-day resting period using a Casy cell counter. Only NAC that were primed by DexVD3 DC did not proliferate during the rest phase, while all other tolDC induced proliferation similar to mature DMSO DC (Fig. 3B).

DexVD3 DC induce Tr1 cells that are functionally effective

Because the aim of the therapeutic vaccination with tolDC is to overcome the ongoing autoimmune reaction and prevent the induction of new autoreactive cells, we examined the ability of NAC primed with different tolDC populations to suppress T cell stimulation in the presence of mature DMSO DC loaded with antigen (PPD) (Fig. 3C). Co-incubation of NAC primed with DexVD3 DC resulted in a significant inhibition of proliferation of responder cells (P = 0.02), defined as persistence of high MFI of CFSE-labelled cells (median 11900), when compared to proliferation of responder cells with effector NAC primed with mature DMSO DC (median 5486). Other tolDC treatments were similar to immature DMSO DC (BAY DC: median 8452; TGZ DC: median 9787; PGJ DC: median 10200; immature DMSO DC: median 8843).

To get insights into mechanism of action of DexVD3 DC, the phenotype of the proliferating cells after the first coculture with the various tolDC populations was analysed. Surprisingly, the percentage of CD4+CD25+FoxP3+ Treg cells was lower in NAC primed with DexVD3 DC (Fig. 4A) and the remaining cells contained high amounts of CD4+IL-10+CD25FoxP3 Tr1 cells (Fig. 4B). The 25-plex Luminex assay revealed that supernatants from NAC primed with DexVD3 DC produced less GM-CSF, TNF-α, IFN-γ, IL-2R, IL-5, IL-12 and MIP-1a (P < 0.05 for IL-12 DexVD3 DC vs. any group). Interestingly, the level of IL-2 was higher compared to the other groups (Figure S4).

image

Figure 4. DexVD3 DC induce Tr1 cells and promote Breg. After the 5-day rest period, the NAC, previously primed with the different tolDC populations, were phenotyped on a LSRFortessa flow cytometer. (A) CD4+ cells were gated, and the proportion of CD25+FoxP3+ cells amongst them was evaluated. Data are presented as median ± SD, n = 3. (B) The percentage of CD4+IL-10+ CD25FoxP3 Tr1 cells from the same experiments is shown. Data are presented as median ± SD, n = 3. (C) NAC primed with DexVD3 DC contained highest amount of CD19+IL-10+ double-positive Breg. Data are presented as median ± SD, n = 3. The gating strategy is presented in Figure S3. The DC population that was used to stimulate NAC is shown on the x-axis. See Fig. 1 for definitions.

Download figure to PowerPoint

To address the effect of CD4+CD25+FoxP3+ T cells, we depleted the cultures of CD25+ cells. Unexpectedly, this resulted in a clear reduction in proliferation of CFSE-labelled responder cells in most of the groups while having little effect on the suppression of proliferation by NAC primed with immature DMSO DC or DexVD3 DC (Fig. 5A, B). Analysis of supernatants from these cocultures showed that depletion of CD25+ cells leads to decreased levels of most of the cytokines and chemokines analysed (Figure S5).

image

Figure 5. Depletion of CD25+ effector cells results in reduced proliferation of responder cells in the presence of mature DMSO DC. Naïve responder cells were labelled with CFSE, and tolDC-primed NAC were labelled with CellTrace Violet after the depletion of CD25+ cells. (A) The gating was performed on the CellTrace Violet negative cells, and the percentage of unproliferated CFSE-labelled cells is shown. (B) The reduction in MFI on CFSE-labelled responder cells is shown. Data are presented as mean ± SD, n = 3. Overlaid dotted bars show the results after the depletion of CD25+ cells from the effector population. The DC population that was used to stimulate the effector cells is shown on the x-axis. See Fig. 1 for definitions.

Download figure to PowerPoint

Dendritic cells are also known as regulators of B cells, and they could interact directly and induce B cell differentiation [23]. We hypothesized that the effect of proliferation suppression could be not only due to Treg cells because we used monocyte-depleted PBMC but also due to the presence of B cells with regulatory activity. The Breg population was defined as CD19+IL-10+ and was detectable in all coculture experiments. NAC primed with DexVD3 DC contained the highest amount of Breg (up to 2.4%, Fig. 4C).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

Different tolDC populations share similar features such as low expression of costimulatory molecules and high production of immunosuppressive cytokines but differ in the expression of other surface markers and functional properties. Here, we compared four different protocols for the generation of tolDC with drugs suitable for use in clinical trials that were not included in previously published comparison studies [19, 20].

While the phenotype of TGZ DC and BAY DC was similar to immature DMSO DC, the use of dexamethasone and 1alpha,25-dihydroxyvitamin D3 resulted in generation of tolDC that were more macrophage like with high CD14 expression and little CD1a expression. Due to retention of CD14 expression by DexVD3 DC and PGJ DC, it might be more appropriate to call them tolerogenic APC instead of tolDC because they do not resemble DC [24].

In our experiments, LPS was used for the stimulation of tolDC as it was shown to improve antigen presentation and migration of tolDC without alteration of their tolerogenic capacity [25]. Indeed, all populations upregulated MHC class II and costimulatory/inhibitory molecules after the addition of LPS. In addition, DexVD3 DC upregulated CD80 (B7-1) molecules, which are known for their high affinity to CTLA-4 on T cells [26].

Cytokine profiling revealed that DexVD3 DC were the only populations that did not produce IL12p70 upon LPS stimulation, while IL-10 was present in all DC culture supernatants. As expected, addition of LPS increased IL-10 production by all tolDC and control populations [27]. However, DexVD3 DC produced by far the highest amounts of IL-10 and in addition to that less pro-inflammatory cytokines IL-5, IL-6 and TNF-α, similarly to immature DMSO DC and unlike other tolDC groups. Induction of functionally active Tregs is an obligatory outcome of tolDC therapy. In murine studies, the transcription factor FoxP3 can be used as a reliable marker to define Treg within CD4+CD25+ T cells [28, 29]. In humans, however, more recent data showed that FoxP3 can be transiently expressed and is present in a non-suppressive population of cells [30]. In addition, its expression is inducible, and it is present in activated effector and memory T cells [31]. Our results support these findings because FoxP3 was expressed on a large proportion of CD4+ T cells without subsequent suppression of effector cell proliferation. Furthermore, this observation was confirmed in a series of depletion experiments where CD25+ cells were removed from tolDC-primed NAC before the suppression experiments. In our experiments, the proliferation of responder cells cultured with mature DMSO DC, BAY DC, TGZ DC and PGJ DC dropped down to proliferation rates of cells cultured with immature DMSO DC and DexVD3 DC in the absence of CD25+ cells, and the production of pro-inflammatory cytokines and chemokines was reduced.

Interestingly, we detected a larger proportion of Tr1 cells amongst the T cells primed with DexVD3 DC, which did not express FoxP3 and secreted considerable amounts of IL-10. This is in line with results from a collagen-induced arthritis mouse model where injection of DexVD3 DC resulted in an increase in the number of CD4+IL-10+ Tr1 cells and reduction in Th17 effector cells [21]. It was shown previously that VD3 acts differentially on different DC lineages. Dermal DC treated with VD3 induced Tr1 cells, while Langerhans cells induced classical Treg [32]. Our study might shed light on the ongoing debate on so-called inflammatory DC that have a monocyte-derived origin [33]. Our data suggest that inflammatory DC might simply resemble dermal DC because they induce the same type of Treg.

Despite the discovery of B cells with suppressive activity 20 years ago [34], B lymphocytes have been mostly considered as cells promoting activation of autoreactive T cells and producing antibodies. Only during the last decade, Breg have been more intensively studied. Their phenotype is still debated and has been defined as a combination of various surface markers (e.g. CD19, CD21 and CD23) together with the expression of IL-10 [35-37]. They represent up to 0.6% of blood B cells and are present in the blood of patients with autoimmune diseases, such as rheumatoid arthritis, systemic lupus erythematosus (SLE), primary Sjögren's syndrome and multiple sclerosis [38] in increased numbers. However, in contrast to Breg isolated from blood of healthy subjects, Breg isolated from blood of patients with SLE produce less IL-10 and are impaired in their suppressive capacity [39]. In our experiments, an increase in Breg frequencies during coculture of NAC with DexVD3 DC was observed. A similar effect was detected in recipients of tolDC therapy in a recently conducted phase I clinical trial for type I diabetes [13]. The patients enrolled in the study showed an increased frequency of Breg in peripheral blood, and follow-up in vitro experiments revealed that these tolDC stimulated the proliferation of novel Breg in NOD and C57BL/6 mice through CD40 and IL-12p40 secretion [40]. Because Breg in patients with SLE were found to be refractory to CD40 stimulation [39], we believe that tolDC could promote the development of new Breg through secretion of IL-10 or yet undiscovered mechanisms; however, further studies are required to confirm the suppressive capacity of such Breg promoted by tolDC.

In conclusion, we show here that the combination of dexamethasone and 1alpha,25-dihydroxyvitamin D3 is superior to other tolDC inducing compounds used in this study concerning induction of regulatory T and B cells and has therefore the best potential to be used in immunotherapy of autoimmune diseases.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

We thank Nicolas Delaleu for helpful discussions, Harald Wiker for supplying PPD and Marianne Enger from the Core facility for flow cytometry, Gade institute, University of Bergen, for help with organization of flow cytometry experiments.

Funding

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

This work was supported by the Broegelmann Foundation and the Bergen Research Foundation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Funding
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
sji12039-sup-0001-FigureS1.pdfapplication/PDF22KFigure S1 DexVD3 DC have an anti-inflammatory cytokine profile. The cell-free supernatants were collected when harvesting the DC populations and stored in aliquots at −20 °C until analysis. The amount of cytokines/chemokines secreted were analyzed using a 25-plex bead assay and the concentrations are given in pg/ml. Cytokines and chemokines that showed results below the detection limit (IL-1β, IP-10, MCP-1, MIP-1a, MIP-1b) are not shown. GM-CSF and IL-4 were part of the added cytokines and therefore not considered in the analysis. Data are shown as mean ± SD and represent three independent experiments. See Fig. 1 for definitions.
sji12039-sup-0002-FigureS2.pdfapplication/PDF132KFigure S2 Gating strategy for the analysis of the suppressive capacity of T cells primed with various tolDC populations. Naïve responder cells were labelled with CFSE and tolDC-primed NAC were labelled with CellTrace Violet. This allowed exclusive gating on CellTrace Violet negative cells and further analysis of proliferation of CFSE-labelled responder cells. The top left dot plot depicts cells in a live gate based on forward scatter and side scatter, the top right histogram depicts gating on CellTrace Violet negative cells. The left bottom histogram shows proliferation of CFSE-labelled CellTrace Violet negative responder cells after co-culture with immature DMSO DC-primed NAC. The right bottom histogram is an overlay of proliferation histograms of CFSE-labelled CellTrace Violet negative responder cells after co-culture with different tolDC-primed NAC. MFI values for the different populations are shown in the legend.
sji12039-sup-0003-FigureS3.pdfapplication/PDF1939KFigure S3 Gating strategy for the analysis of NAC phenotype. The top row of dot plots shows cells in a live gate based on forward scatter and side scatter, the second row depicts gating on CD4+ cells based on unstained negative controls. 3rd row depicts the proportion of the CD25FoxP3 cells amongst the CD4+ NAC. 4th row shows the proportion of CD4+IL-10+CD25FoxP3- Tr1 cells. The bottom row shows gating on CD19+IL-10+ double positive Breg cells based on unstained negative controls. The DC population that was used to stimulate NAC is shown on the top. See Fig. 1 for definitions.
sji12039-sup-0004-FigureS4.pdfapplication/PDF25KFigure S4 Cytokine and chemokine profiling of culture supernatants from NAC primed with various DC populations. The cell-free culture supernatants were collected after co-culture of DC and NAC and stored in aliquots at −20 °C until analysis. The amount of cytokines/chemokines secreted was analyzed using a 25-plex bead assay and the concentrations are given in pg/ml. Cytokines and chemokines that showed results below the detection limit (IL-4, IP-10, IL-1β, MCP-1, MIP-1b) are not shown. Data are shown as mean ± SD and represent three independent experiments. Only differences in IL-12 production are statistically significant (P < 0.05). The tolDC population that was used to stimulate NAC is shown on the x-axis. See Fig. 1 for definitions.
sji12039-sup-0005-FigureS5.pdfapplication/PDF34KFigure S5 Cytokine and chemokine profiling of culture supernatants from suppression experiments. The supernatants were collected before the FACS analysis of proliferation of the responder cells and stored in aliquots at −20 °C until analysis. The amount of cytokines/chemokines secreted were analyzed using a 25-plex bead assay and the concentrations are given in pg/ml. Cytokines and chemokines that showed results below the detection limit (IL-1β, IL-4, IL-17, MIP-1a, MIP-1b) are not shown. Data are shown as mean ± SD and represent three independent experiments. The DC population that was used to stimulate the effector cells is shown on the x-axis. See Fig. 1 for definitions. CD25 defines the depletion of CD25+ from the effector population.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.