Tolerogenic Dendritic Cells Derived from Donors with Natural Rubber Latex Allergy Modulate Allergen-Specific T-Cell Responses and IgE Production

Natural rubber latex (NRL; Hevea brasiliensis) allergy is an IgE-mediated reaction to latex proteins. When latex glove exposure is the main sensitizing agent, Hev b 5 is one of the major allergens. Dendritic cells (DC), the main antigen presenting cells, modulated with pharmacological agents can restore tolerance in several experimental models, including allergy. In the current study, we aimed to generate DC with tolerogenic properties from NRL-allergic patients and evaluate their ability to modulate allergen-specific T and B cell responses. Here we show that dexamethasone-treated DC (dxDC) differentiated into a subset of DC, characterized by low expression of MHC class II, CD40, CD80, CD86 and CD83 molecules. Compared with LPS-matured DC, dxDC secreted lower IL-12 and higher IL-10 after CD40L activation, and induced lower alloantigenic T cell proliferation. We also show that dxDC pulsed with the dominant Hev b 5 T-cell epitope peptide, Hev b 546–65 , inhibited both proliferation of Hev b 5-specific T-cell lines and the production of Hev b 5-specific IgE. Additionally, dxDC induced a subpopulation of IL-10-producing regulatory T cells that suppressed proliferation of Hev b 5-primed T cells. In conclusion, dxDC generated from NRL-allergic patients can modulate allergen-specific T-cell responses and IgE production, supporting their potential use in allergen-specific immunotherapy.


Introduction
Natural rubber latex (NRL; Hevea brasiliensis) allergy is an IgEmediated reaction to latex proteins, first described in 1979 and increasingly recognized in subsequent years [1,2,3]. Several NRL allergenic proteins are described, Hev b 1-13. Of these, Hev b 2, Hev b 5, Hev b 6 and Hev b 13 are the major allergens for health care workers, typically sensitized by exposure to NRL gloves [4]. Hev b 5 is a strong antigen and one of the most important latex allergens, with a high prevalence of immunoglobulin (Ig)-E reactivity in health care workers [5,6].
Pharmacotherapy for allergic diseases, such as antihistamines or corticosteroids, ameliorates symptoms but does not stop progression; the only therapy that modifies progression of allergic diseases is allergen specific immunotherapy (SIT). However, application of this potentially curative treatment is restricted, largely due to the risk of serious adverse events, especially in asthmatics and for potent allergens such as peanut, seafood and latex [7].
Mechanisms described for regulation of the immune response to allergens include those driven by naturally occurring CD4+CD25+ Treg and inducible populations of allergen-specific, interleukin (IL)-10-secreting Type 1 regulatory T cells (Tr1) [8,9]. Dendritic cells (DC) play a pivotal role in the immunoregulatory mechanisms underlying successful SIT, dampening allergic immune responses by absence of co-stimulation [10] or increasing IL-10 production to expand allergen-specific Tr1 cells [11,12]. These features can be observed at an immature stage of DC differentiation [13] and in a specialized DC subset, called tolerogenic DC (t-DC) [14].
At present, allergen avoidance and symptomatic treatment for adverse reactions are the only therapeutic strategies available for NRL allergy [15]. Therefore, t-DC could provide a novel treatment strategy for inducing allergen-specific desensitization in NRL-allergic patients. Several agents have been used to generate t-DC, including IL-10 [13], transforming growth factor-b1 (TGF-b1), the active form of vitamin D3 [16], neuropeptides [17], corticosteroids and cyclosporin, amongst others [18,19].
Glucocorticoids (GC) are potent immunosuppressive and antiinflammatory agents used to treat autoimmune diseases and prevent graft rejection [20]. Studies have indicated that GC affects the phenotype of monocyte-derived DC, modulating their differentiation, maturation and function [21,22].
In the current study, we aimed to generate t-DC from NRLallergic patients using dexamethasone (dx), and determine whether dx-treated DC (dxDC) can modulate NLR-specific T-cell responses and prime T cells to become allergen-specific IL-10producing cells. We demonstrate that dxDC have reduced alloantigen and allergen-specific T-cell stimulatory capacity and can inhibit Th2 differentiation of naïve CD4+ T cells. In addition, dxDC induced the expansion of an IL-10-producing regulatory Tcell subset with suppressive activity. These findings indicate that dxDC may be useful for antigen-specific tolerization of allergic immune responses in NRL-allergic patients.

Ethics Statements
The study was approved by the local medical ethics committee (Hospital Clínico Universidad de Chile Mm Nu 304). All patients provided written informed consent. After the samples had been collected, each patient was allocated a trial number, demographic data collected, and the database anonymised.

Study Subjects
Peripheral venous blood from 8 NRL-allergic patients (1-8) was collected. All patients had positive skin prick tests ($3 mm above saline negative control) to latex extract (Stallergenes, Paris, France) and increased levels of IgE specific to an extract of NRL (Immulite 2000 analyzer; Siemens Healthcare, Germany). All patients were no atopic or sensitized to other allergens unrelated to latex. Three of eight tested patients (NRL-1, NRL-5, NRL-7) were positive for specific IgE against recombinant Hev b 5 protein, determined by ELISA. The demographic and clinical characteristics of the patients are reported in Table 1.

Preparation of T and B Cells
PBMC were obtained from peripheral blood by density gradient separation with Ficoll-Hypaque (Axis-Shield, Oslo, Norway). CD45RA+ CD4+ T cells and CD27-B cells were isolated from PBMC using antibody-coated magnetic Microbeads (MACS, MiltenyiBiotec Inc, CA, USA), according to the manufacturer's protocol. Separation was assessed by flow cytometry (purity 90% and 97%, respectively).

Generation of DC
DC were differentiated from monocytes using established methods [23]. Briefly PBMCs were obtained from buffy coats. Cells (3610 7 /well) were incubated in serum-free AIM-V therapeutic medium (Gibco BLR, Paisley, UK) at 37uC, 5% CO 2 for 2 h in a six-well plate (Falcon Becton Dickinson, Hershey, PA, USA). Non-adherent cells were removed, and the remaining cells were incubated for 7 days in the presence of 500 U/ml recombinant human IL-4 (rhIL-4) (US Biological, Swampscott, MA, USA) and 800 U/ml of GM-CSF (Shering Plough, Brinny Co., Ireland). The cultures were maintained for 7 days, replacing the medium every 2 days.

FITC-dextran Endocytosis Assay
Dextran uptake activity was assessed by incubating 0.5610 6 DC with FITC-conjugated dextran (Molecular Probes, Eugene, OR) (0.2 mg/mL) for 2 h at 37uC in the dark. Cells were washed carefully with PBS and FITC-dextran uptake was quantified by flow cytometry. Background of dextran incorporation was assessed by incubating DC on ice.

Measurement of DC and T-cell Cytokine Production
DC (2610 4 ) were incubated with an irradiated CD40Lexpressing 3T3 fibroblast cell line (cell ratio 10:1) at 37uC and 5% CO 2 overnight. IL-10 and IL-12 producing cells were enumerated using an ELISPOT Ready-SET-Go!H according to the manufacturer's instructions (eBioscience). Spots were counted using A.EL.VIS ELISPOT Analysis Software (Hannover, Germany). T-cell production of IL-4 and IFN-c was also evaluated by ELISPOT Ready-SET-Go!H (eBioscience). Tumoral necrosis factor (TNF)-a production was measured by intracellular cytokine staining and samples were analyzed by flow cytometry (FACScanto; Becton Dickinson).

Proliferation Assays
Allogeneic PBMC or Hev b 5-specific T-cell lines, generated using established methods [26], were labeled with CFSE (5 mM per 1610 7 cells) (Renovar, USA) for 15 min at 37uC. Cells were washed extensively and 2610 5 cells/well were cultured with Hev b 5 46-65 peptide-pulsed DC in round-bottomed 96-well plates in serum-free AIM-V medium (Gibco BLR) for 5 days. Type II human collagen (CII) 259-263 peptide (GIAGFKGEQGPKGET) (GenScript) was used as a control. CD4+ T-cell proliferation was determined by CFSE dilution analysis by flow cytometry (FACScanto; Becton Dickinson). Apoptosis of T cells was measured using an Annexin V Apoptosis Detection Kit APC (eBioscience).

IgE Production
Autologous naïve B cells (1610 5 ), naïve T cells (2.5610 5 ), Hev b 5 46-65 peptide-pulsed DC (2.5610 4 ) and CD40L-expressing fibroblasts (2.5610 3 ) were co-cultured in round-bottomed 96-well plates in the presence of rhIL-4 (1000 IU/ml) (eBioscience). After 10 days, supernatants were harvested and assessed for total and Hev b 5-specific IgE levels by Serum samples were tested for specific IgE using our standard ELISA protocol. In brief, ELISA plates (Falcon Becton Dickinson) were coated with rhev b 5 (2.5 mg/ml) [27] in 0.1 M bicarbonate buffer (pH 9.6). After blocked, diluted plasma (1/10) were added. IgE were quantified with biotinylated anti-human IgE mAb (BD Pharmingen, USA) diluted 1/1000. Development was gone with substrate solution (ATBS/H 2 O 2 ). Plates were read at 460 nm using an ELISA plate reader. Background values obtained for sera and mAb on wells uncoated with Ag were subtracted from values obtained on wells coated with Ag. Values were considered positive when they differed from control supernatant values .2 times the SD.

Statistical Analysis
Results are presented as mean 6 SD. The Kruskal-Wallis test with Dunn's Multiple Comparison post-test was used to compare the mean values of cell surface marker expression, cytokine and IgE production between different cell culture conditions. Prolifer-

Characterization of Tolerogenic dxDC
Analysis of the phenotype and function of dxDC showed a lower level of HLA II (p = 0.0058), CD80 (p = 0.0117), CD86 (p = 0.0058), CD40 (p = 0.0055) and CD83 (p = 0.0058) expression compared with mDC, while the expression of the chemokine receptor CCR7 was similar to mDC ( Figure 1A). In contrast, dxDC expressed equivalent levels of surface markers to untreated iDC that were used as an immature control. To further characterize dxDC, their ability to take up FITC-dextran was examined. dxDC displayed similar uptake to iDC, and higher than mDC, which showed characteristically reduced activity ( Figure 1B). The suppressive effect of dx on DC was not affected by LPS, indicating that dx can induce durable immaturity of DC in terms of HLA II, CD86 and CD83 expression and production of the pivotal pro-inflammatory cytokine, TNF-a ( Figure 2).
As cytokines secreted by DC are important in determining the differentiation fate of T cells, we assessed the cytokine profile of our DC populations. dxDC secreted high levels of IL-10 and low levels of IL-12 upon activation via CD40L. iDC and mDC produced significantly greater levels of IL-12 and lower levels of IL-10 compared with dxDC ( Figure 3). These results indicate that dxDC have an anti-inflammatory profile characterized by high production of IL-10.

Modulation of T-cell Response by Tolerogenic dxDC
Stimulation of allogeneic CD4+ T cells by dxDC at different stimulator:effector ratios induced a lower proliferative CD4+ percentage compared with allogeneic CD4+ T cells stimulated with mDC or iDC (p,0.05) ( Figure 4A). To discard that the inhibitory effect of dxDC was mediated by a mechanism of T cell death, we examined the level of apoptosis and necrosis in CD4+ T cells by Annexin V staining and propidium iodide dye. No significant differences were observed in the fraction of apoptotic cells to T cells co-cultured with DC subjected to different treatments ( Figure 4B).
To determine the modulation of dxDC in NRL-specific T-cell response, oligoclonal Hev b 5-specific T-cell lines were generated from 3 patients with Hev b 5-specific IgE reactivity (patients 1, 5 and 7). These T-cell lines were stimulated with Hev b 5 46-65pulsed dxDC or with Hev b 5 46-65 -pulsed mDC. We observed that while peptide-pulsed mDC induced an intense proliferative response, stimulation with peptide-pulsed dxDC induced hyporesponsiveness of specific T-cell lines to their cognate antigen (p = 0.013) ( Figure 5). T-cell lines stimulated with mDC pulsed with the irrelevant peptide CII 259-263 did not show proliferation at all ( Figure S1).

dxDC Modulate T Cell Differentiation Hindering Subsequent B-cell Help for IgE Secretion
In order to evaluate the T-helper cytokine profile driven by dxDC, naïve CD4+ T cells from 3 patients (1, 5 and 7) were primed with autologous Hev b 5 46-65 pulsed dxDC (T Hev b 5-dxDC ) or mDC (T Hev b 5-mDC ), and IL-4 and IFN-c secreting T cells were measured (Table 2). dxDC stimulated greater IL-4 and IFN-c production by CD4+ T cells than mDC, but with a more pronounced effect on IL-4 production. As IL-4 favours the differentiation of naïve B cells into IgE antibody-secreting cells, we assessed the ability of T Hev b 5-dxDC to induce total and allergen-specific IgE production. For this purpose, we co-cultured Hev b 5 46-65 peptide-pulsed dxDC with autologous nä ive CD4+T cells, naïve B cells and CD40Ltransfected fibroblasts in the presence of IL-4 for 10 days. Interestingly, total IgE ( Figure 6A) and Hev b 5-specific IgE ( Figure 6B) production decreased in the presence of dxDC compared to stimulation with mDC.
dxDC Induce Regulatory T Cells IL-10 secreting DC have been shown to induce Tr1 cells in vitro by a direct effect of this cytokine on undifferentiated CD4+ T cells [28,29,30]. According to these observations, we explored whether dxDC could induce the secretion of IL-10 by naïve CD4+ T cells. For this purpose, T Hev b 5-dxDC were re-stimulated with PMA/ ionomycin and IL-10 secretion was evaluated by ELISPOT. As expected, T Hev b 5-dxDC secreted significantly more IL-10 than T Hev b 5-mDC (P,0.002) ( Figure 7C). To confirm that dxDC can induce a true Tr1 population, the suppressive ability of T Hev b 5-dxDC was tested. Consistent with the high levels of IL-10 secretion,

Discussion
iDC resident in peripheral tissues in a steady-state can induce Tcell tolerance [31]. Similarly, DC prepared ex vivo and exposed to antigens in the absence of full-maturation stimuli down-regulate immunity and induce tolerance. Similarly, the use of immunosuppressants such as corticosteroids can induce human DC to acquire tolerogenic properties [18,19]. In this study, we demonstrate the feasibility of generating t-DC from NRL-allergy patients using dexamethasone. We demonstrate that dexamethasone induces an immature phenotype on DC, with low expression levels of HLA II and co-stimulatory molecules such as CD40, CD80, CD86 and CD83 and high particle uptake ability. However, the expression of the homing receptor CCR7 was not affected by dexamethasone, showing similar levels to mDC. This would enable dxDC to migrate to lymph nodes, as would be required if they were to be used in immunotherapy. Moreover, if tolerance restitution is the goal of immunotherapy, as in NRL allergy, it is essential that DC maintain an immature state in the presence of inflammatory factors. Therefore, we evaluated the resistance of the dxDC phenotype to LPS exposure. Our results demonstrate that dxDC, unlike iDC, were unaffected by LPS stimulation in terms of HLA II, CD86 and CD83 expression and production of the pivotal pro-inflammatory cytokine, TNF-a.
Low antigen presenting and co-stimulatory molecule expression levels, strongly suggest that one of the most likely mechanisms of T-cell tolerization by dxDC is through anergy induction, due to the lack of sufficient first and second activation signals [32].
Consequently, we showed that dxDC have weak allogeneic Tcell stimulatory activity. This effect was not due to deletion of alloreactive T cells, because an increase in apoptosis was ruled out. Since the aim of immunotherapy is to induce antigen-specific tolerance without impairing other immune functions, we evaluated the ability of dxDC to present the immunodominant T-cell epitope for the NRL allergen Hev b 5 [26]. Our results demonstrate that dxDC have a reduced ability to present antigen to Hev b 5 specific oligoclonal T-cell lines compared with mDC. This is consistent with the results obtained from our allogeneic experiments, and with previous data from Steinbrink et al. showing that clonal anergy was induced in an influenza hemagglutininspecific CD4+ T-cell clone via stimulation with IL-10 treated DC [29].
As noted earlier, dxDC display low levels of activation signal 1 (HLA II) and 2 (co-stimulatory molecules), however, a third signal, established by secreted cytokines, is also relevant in guiding T-cell polarization [33,34,35]. IL-10-secreting DC have been associated with tolerance induction against common antigens [36,37] and with poor T-cell stimulatory function [38,39]. Furthermore, IL-10 expression has been observed in DC located in lung tissue and the intestine, suggesting an important role in maintaining local T-cell tolerance to common environmental antigens [36]. Our study shows that dxDC produce large amounts of IL-10 and low levels of IL-12 upon CD40L stimulation; these findings are in agreement with previous work using dexamethasone as a tolerogenic agent [40,41].
As demonstrated previously, IL-10-secreting DC induce regulatory T cells upon activation [42,43]. In line with this evidence, our study shows that naïve CD4+ T cells that were primed by dxDC (T Hev b 5-dxDC ) differentiated into an IL-10+ T-cell population, compatible with a Tr1 cell phenotype [44]. These findings agree with those described by Bosma et al. for myeloid DC isolated from peripheral blood and tolerized with dexamethasone and LPS [45]. When we assessed the regulatory properties of T Hev  b5-dxDC cells on Hev b 5-specific-CD4+ memory T cells, we observed inhibition of T-cell proliferation, where the suppressor effect was independent of T Hev b 5-dxDC cell number. This observation is compatible with an immunomodulatory mechanism of T-cell response, mediated by cytokines such as IL-10 [46], however, we do not exclude regulatory activity by other molecules. The determination of the full range of cytokines and/or molecules involved in the suppressor activity of T Hev b 5-dx-DC needs further investigation.
In this study, we analyzed the influence of dxDC on T cell help for total and allergen-specific IgE production by autologous B cells from NRL-allergic patients. We observed that the presence of dxDC inhibits the secretion of total and Hev b 5-specific IgE by B cells. The reduction of IgE production could be explained by the induction of a Tr1 population. Kanjarawi et al. reported that CD4+CD25+ regulatory T cells were able to modulate total and b-lactoglobulin-specific serum IgE production in a food allergy murine model [47]. IL-10 appears to be an important cytokine in the success of specific immunotherapy in allergy since it has been described to be increased in blood and affected tissues of treated patients [48,49,50]. Akdis et al. reported that Tr1 cells consistently represent the dominant T cell subset specific for common environmental allergens in healthy individuals [8]. They suggested that Tr1 cells might control allergen-specific Th2 responses in allergic and healthy individuals. Pacciani et al. demonstrated that IL-10-producing DC can induce specific T cells with suppressive activity on allergen-specific Th2 cells from house dust mite-allergic patients [30]. In addition, experiments in murine models showed that DC transduced with IL-10 negatively regulate allergic inflammation of airways by inducing IL-10-producing T cells [51]. T cells transduced to express IL-10 have also been shown to inhibit airway hyper-responsiveness induced by Th2 cells [52]. Taken together, these studies suggest that the DC-induced deviation in T cell cytokine production towards IL-10 could be an important factor in preventing allergic responses.
In summary, our results demonstrate that Hev b 5-pulsed dxDC-primed nä ive T cells become Tr1 cells, which modulate allergen-specific T-cell responses. These findings indicate that dxDC may be a useful immunotherapy tool for antigen-specific down-regulation of allergic immune responses in NRL-allergic patients.