SEARCH

SEARCH BY CITATION

Keywords:

  • Autoreactive CD8+ T cells;
  • β1-integrin;
  • Carbon monoxide;
  • Dendritic cells;
  • Diabetes

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

Carbon monoxide (CO) treatment improves pathogenic outcome of autoimmune diseases by promoting tolerance. However, the mechanism behind this protective tolerance is not yet defined. Here, we show in a transgenic mouse model for autoimmune diabetes that ex vivo gaseous CO (gCO)-treated DCs loaded with pancreatic β-cell peptides protect mice from disease. This protection is peptide-restricted, independent of IL-10 secretion by DCs and of CD4+ T cells. Although no differences were observed in autoreactive CD8+ T-cell function from gCO-treated versus untreated DC-immunized groups, gCO-treated DCs strongly inhibited accumulation of autoreactive CD8+ T cells in the pancreas. Interestingly, induction of β1-integrin was curtailed when CD8+ T cells were primed with gCO-treated DCs, and the capacity of these CD8+ T cells to lyse isolated islet was dramatically impaired. Thus, immunotherapy using CO-treated DCs appears to be an original strategy to control autoimmune disease.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

Autoimmune diseases such as type 1 diabetes (T1D) are mainly chronic and impact the quality of life. Current therapy for T1D is based on daily injection of insulin; although insulin delivery methods are improving, this treatment adjusts glycemia imperfectly, resulting in hyperglycemic events ultimately leading to cardiovascular complications. Recent immunotherapeutic trials have focused on systemic immunoregulation protocols with success exemplified by anti-CD3 Ab treatment [1, 2]. However, by indiscriminately modifying the pathogenic autoimmune process together with beneficial immune responses, serious side effects may occur. An attractive alternative is to use Ag-specific approaches to selectively silence autoreactive T cells. As pivotal players in T-cell activation, DCs have been extensively studied to induce peptide-specific tolerance in numerous pathologies (reviewed in [3]). DCs also specifically regulate the T-cell expression of integrins, which are major molecules implicated in migration and function of effector T cells in nonlymphoid tissues (reviewed in [4]).

A recent approach to induce tolerance has emerged using heme oxygenase-1 (HO-1) inducers or carbon monoxide (CO). Endogenous CO is only produced from the oxidative degradation of hemoproteins (e.g. hemoglobin, myoglobin, cytochrome) by the heme oxygenase enzymes, in particular HO-1 [5]. Up-regulation of the HO-1 pathway has a significant protective effect against induced [6] and spontaneous [7, 8] autoimmune diseases, and allergy [9]. Several reports have also shown that overexpression of HO-1 and CO treatment can be beneficial to graft survival [10-14]. The precise underlying mechanisms for HO-1-based protection are not yet completely understood, but appear to involve heme degradation products, such as CO on DCs (reviewed in [15]). In an attempt to elucidate this mechanism, we have shown that immature DCs express HO-1 and that HO-1 overexpression inhibits the LPS-induced immunogenicity of human, rat, and mice DCs [16, 17]. Moreover, we have recently demonstrated that not only HO-1 but also CO modify polyI:C (TLR-3 ligand) and LPS (TLR-4 ligand)-induced DC maturation leading to decreased phenotypic maturation and proinflammatory cytokine production with conserved IL-10 production [17]. We have also developed a transgenic mouse model where DCs trigger autoimmune diabetes. In this model, when DCs treated with the CO-releasing molecule (CORM)-2 were used to immunize transgenic mice, diabetes development was inhibited [17] showing that CORM-treatment of DCs dramatically reduces their immunogenic capabilities.

Previous studies have shown that systemic HO-1 induction or CO treatment of diabetic prone NOD mice lead to protection against diabetes onset [7, 8]. However, the mechanisms involved in this protection are still unknown. In the present study, we have explored the in vivo consequences of gaseous CO (gCO)-induced DC modifications in a double transgenic mouse model of T1D. We found that gCO-treated DCs inhibits diabetes incidence, decreases β1-integrin expression on autoreactive CD8+ T cells and inhibits their accumulation into the pancreas.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

DCs treated with gCO reduced diabetes induction in a peptide-specific manner

As CORMs might have CO nonrelated effects [18], we used gaseous CO (referred to as gCO) in all our experiments as a physiologically relevant form of CO. Two hours of in vitro treatment of DCs by gCO (350 ppm) led to the inhibition of IL-12 secretion but conservation of IL-10 secretion (Fig. 1A) as observed with CORM-2 [17]. This treatment was nontoxic to the cells (data not shown). Interestingly, as with CORM-2 [17], we also observed a slight decrease in co-stimulatory molecules expression on CO-treated DCs although not statistically significant (Supporting Information Fig. 1).

image

Figure 1. gCO treatment modifies cytokine production and inhibits DC pathogenicity in a peptide-dependent manner. (A) DCs (BM-derived DCs, bmDCs) were sorted for CD11c and incubated at 5% CO2, matured with 0.5 μg/mL LPS and treated or not with gCO (350 ppm for 2 h). Twenty-four hours post culture, supernatant was collected for IL-12 and IL-10 detection by ELISA. (A) Data are shown as mean + SEM of 30 samples pooled from eight independent experiments. (B) Mice were transferred with 1 × 106 anti-HA512–520 TCR transgenic CD8+ T cells. The following day, mice were immunized with 3 × 104 HA512–520-loaded mature DCs treated (DC-LPS/HA/gCO) or not (“DC-LPS/HA” group) with gCO (350 ppm for 2 h) and monitored for diabetes development(C) CO-treated DCs were either loaded with H2Kd-irrelevant NP147–155 or relevant HA512–520 peptides and i.v. co-injected with HA512–520 or NP147–155 loaded DCs into recipients. Mice were monitored for diabetes development by glycosuria. (B and C) Data shown are from the indicated number of mice pooled from at least three independent experiments. ***p < 0.001 with paired t-test.

Download figure to PowerPoint

As the mechanism of CO-induced protection is unknown, we wished to dissect the in vivo consequences of CO-induced DCs modifications using a simplified mouse model for auto-immune diabetes where islet destruction is induced by DC-primed CD8+ T cells. Mice expressing influenza virus HA in pancreatic β cells (InsHA mice [19]) received naïve anti-HA512–520 CD8+ T cells from TCR transgenic CL4 Thy1.1+ mice [20]. In this experimental setting, we have previously shown that the adoptive transfer of naïve HA512–520 specific CD8+ T cells does not induce diabetes unless co-transferred with LPS-matured HA512–520 peptide-loaded DCs [17]. We next assessed whether gCO-treated DCs were able to inhibit diabetes initiated by CD8+ T cells (Fig. 1B). gCO-treated DCs loaded with relevant (HA512–520) peptide did not trigger diabetes, as observed previously with CORM-treated DCs [17]. These results indicate not only that this pharmacological agent can modify the DC phenotype but also that the physiological form of CO affects DC biology and function. We also asked whether inhibition of CD8+ T-cell pathogenicity by gCO-treated DC was peptide specific. DCs treated with gCO and loaded with an irrelevant peptide were unable to inhibit diabetes in mice co-transferred with DCs loaded with the relevant peptide (HA512–520) on the same day (Fig. 1C). Since only gCO-treated DCs with the relevant peptide recognized by the diabetogenic CD8+ T cells abrogated diabetes whereas DCs loaded with the irrelevant peptide did not, these results indicate that protection induced by gCO-treated DCs was firstly peptide-restricted and secondly not the result of nonspecific suppressive mechanisms by gCO-treated DCs.

The inhibitory effect of gCO-treated DCs on CD8+ T cells is IL-10 and CD4+ T-cell independent

As IL-10 secretion, known for its anti-inflammatory properties, was preserved when DCs were treated with CO [17], we questioned the contribution of this cytokine secretion in CO-mediated resistance to diabetes. We showed that IL-10 KO DCs were as effective as WT-DCs to induce CO-mediated CD8+ T-cell modification and to abrogate diabetes development (Fig. 2A) demonstrating that IL-10 secretion by DCs is dispensable for CO-mediated inhibition of CD8+ T-cell pathogenicity.

image

Figure 2. The inhibitory effect of gCO-treated DCs on CD8+ T cells is IL-10 and CD4+ T-cell independent. (A–C) Ins-HA mice were transferred with 1 × 106 anti-HA512–520 TCR transgenic CD8+ T cells. The following day, mice were immunized with 3 × 104 mature DCs treated or not with gCO (350 ppm for 2 h) as indicated. Mice were monitored for diabetes development by glycosuria. (A) IL-10 KO (IL-10KO) or IL-10 WT (IL-10wt) DCs were used to immunize recipient mice. (B) Ins-HA Rag KO (RAG2−/−) mice were used as recipients. (C) Mice depleted of Treg cells by three injections of anti-CD25 mAb (PC61 clone) were used as recipients before assessing their resistance to diabetes. Rat IgG2a was used as isotype control (IC) and administered with the same dose and schedule as PC61 mAb injection. (A–C) Data shown are from the indicated numbers of mice pooled from at least three independent experiments.

Download figure to PowerPoint

Some authors have described the inhibition of CD8+ T-cell function by means of CD4+ T cells, notably Treg cells in the context of diabetes (reviewed in [2]). To test this hypothesis, Treg cells were depleted in recipient mice using anti-CD25 Ab leading to complete CD4+CD25+ depletion (data not shown). The results show that these cells were unnecessary to mediate the inhibitory effect of gCO (Fig. 2B). To confirm this we used InsHA mice on a RAG2 KO background as recipients for CO-treated or untreated DC immunization. In this situation, where endogenous B and T cells were absent, HA-loaded matured DCs were still able to trigger diabetes with slightly delayed kinetics (Fig. 2C). Thus, in the absence of any CD4+ T cells (including CD4+ Treg cells), gCO-mediated resistance to diabetes was still observed by contrast to untreated DC-immunized mice.

gCO-treated DCs show normal in vitro and in vivo CD8+ T-cell priming

The resistance to diabetes induction observed previously (Fig. 1B) could be due to lack of islet-specific T-cell expansion, clonal deletion, or anergy as a consequence of CD8+ T-cell priming impairment. These hypotheses are strengthened by the disrupted cytokine secretion and by the slight decrease in co-stimulatory molecules expression observed in gCO-treated DCs (Fig. 1A and Supporting Information Fig. 1).

To test these possibilities, we primed CD8+ T cells with gCO-treated or untreated DCs in vitro. Surprisingly, no difference was observed in CD8+ T-cell proliferation (Supporting Information Fig. 2A) or IL-2 (Supporting Information Fig. 2B) secretion. Similar results were obtained at low peptide concentrations (data not shown). Furthermore, CD25 (Supporting Information Fig. 2C) or CD69 (Supporting Information Fig. 2D) early activation markers were not affected by gCO DC treatment. Similar results were obtained when CO-releasing molecule (CORM-2, 100 μM) was used as a CO provider in the same settings (data not shown).

In order to have an in vivo picture of CD8+ T-cell priming by gCO-treated DC, we transferred HA-specific CFSE-loaded CD8+ T cells genetically tagged with Thy1.1+ marker in InsHA mice before DC immunization. At the time of diabetes onset in untreated DC-immunized mice, mesenteric, pancreatic LNs, and spleen cells were harvested. The percentage of positive cells for the activation marker CD44 (Fig. 3A and B) were comparable in gCO-treated or untreated DC groups, suggesting that priming of autoreactive CD8+ T cells was identical. The percentage of donor cells divided in recipient mice was similar in gCO-treated DC and untreated-DC immunized groups (Fig. 3C). Similarly, the percentages and numbers of donor CD8+ T cells in different LN territories or spleen were comparable in both groups at diabetes onset (Fig. 3D). We also observed similar expansion kinetics of anti-HA CD8+ T cells in HA expressing mice at day 4, 8, and 11, again demonstrating comparable levels of priming (Fig. 3E). Finally, we addressed the question of the long-term expansion of gCO-treated DC-primed CD8+ T cells in BALB/c mice (Fig. 3F). CD8+ T-cell expansion kinetics at day 8, 15, and 30 was indistinguishable when untreated DC or gCO-treated DC immunized mice were compared demonstrating that CO treatment, despite the observed protection, did not altered CD8+ T-cell priming by DCs.

image

Figure 3. CO treatment of mDCs preserves their in vivo CD8+ T-cell priming capacities. (A–E) InsHA or (F) BALB/c mice were transferred with 1 × 106 CFSE-loaded Thy1.1+ anti-HA512–520 TCR transgenic CD8+ T cells. The following day, mice were immunized with 3 × 104 HA512–520-loaded mature DCs treated (“LPS/HA/CO” group) or not (“LPS/HA” group) with gCO (350 ppm for 2 h). Animals were sacrificed 10 days postimmunization except in (E) and (F). Lymphocytes from spleen, pancreatic draining LNs (PLNs), and nonpancreatic LNs (nonPLNs) were isolated and stained with anti-CD8 and Thy1.1 antibodies. CD44 on donor cells (CD8+ Thy1.1+) was assessed as percentage of donor cells among CD8+ T cells in different organs. (A) A representative flow cytometric analysis is presented, together with (B) a quantitative analysis of three independent experiments. (C, D) Proliferation of donor cells was evaluated at day 10 as (C) the percentage or (D) number of donor cells by flow cytometry, gating on CD8+ Thy1.1+ cells in different organs. (B–D) Each symbol represents an individual mouse, bar represents the mean. (E, F) Kinetics of CD8+ Thy1.1+ cell expansion in spleen was analyzed in (E) InsHA or (F) BALB/c mice. (E, F) Data are shown as mean ± SEM of at least five mice per group. Data shown are pooled from three (B, C) and two (E, F) independent experiments.

Download figure to PowerPoint

Even if autoreactive CD8+ T-cell activation was identical in gCO-treated and untreated DCs immunized groups, their function could however be deeply affected. However, comparable levels of IFN-γ-expressing cells were noted between the two groups, in terms of percentage in spleen or pancreatic LNs (Fig. 4A). Additionally, a nonsignificant decrease (p > 0.05) in the absolute numbers of IFN-γ+ donors cells was observed (Fig. 4B). An in vivo CTL assay was performed; no differences in terms of cytolytic function whether CD8+ T cells were activated with gCO-treated or untreated DCs was observed (Fig. 4C and D).

image

Figure 4. Peripheral function of autoreactive CD8+ T cells primed by CO-treated DCs is preserved. All mice were treated as in Fig. 3. On day 10, the proportion of IFN-γ+ donor cells was determined in splenocytes and PLNs by intracellular cytokine staining assay. (A) A representative density plot is shown and (B) quantification of IFN-γ-secreting donor cells in spleen and in PLNs pooled from three independent experiments is reported. Positive control was obtained using PMA/ionomycin as unspecific T-cell activator. Negative controls were stimulated cells in the same conditions as above using an irrelevant peptide (H2Kd NP147–155 peptide). Each symbol represents an individual mouse, bar represents the mean. (C, D) On day 10, syngeneic CFSE-labeled splenocytes were either loaded with HA512–520 (CFSE high) or irrelevant NP147–155 (CFSE low) peptide and intravenously administered to mice. The percentage of specific lysis was determined by flow cytometry 12 h after transfer. (C) A representative flow cytometric analysis of lysis is shown and (D) quantification of CTL lysis activity is reported. (D) One representative experiment out of three is presented.

Download figure to PowerPoint

Altogether these results show that in vivo, modification of autoreactive CD8+ T-cell responses by gCO-treated DCs is not mediated by clonal deletion, anergy, or functional inactivation.

Autoreactive CD8+ T cells primed by gCO-treated DCs show impaired pancreatic accumulation

As diabetes incidence was significantly different in gCO-treated versus untreated DCs immunized groups, we asked whether the accumulation of autoreactive CD8+ T cells in the pancreas was impaired. The histological analysis of the pancreas 10 days following gCO-treated or untreated DC immunization showed that the severity of insulitis correlated with the development of disease (Fig. 5A). To have a more detailed picture of the local process, we isolated pancreas infiltrating lymphocytes and analyzed this fraction by flow cytometry. Interestingly, numbers of donor cells was dramatically reduced in gCO-treated DCs immunized mice compared to untreated DC immunized mice (Fig. 5B). These data show that β-islet-specific CD8+ T cells are fully functional but impaired in their accumulation in the pancreas when primed with gCO-treated DCs.

image

Figure 5. Autoreactive CD8+ T cells show impaired pancreatic accumulation when stimulated with CO-treated DCs. (A and B). Mice were treated as in Fig. 3. The extent of insulitis was determined at 14 days either by (A) histological or (B) flow cytometric analysis in two different experiments for each technique. (A) The extent of insulitis was determined as the percentage of normal, peri-infiltrated, infiltrated <50%, or infiltrated >50% islets (n = 6 mice/group) after hematoxylin–eosin coloration. Total numbers of analyzed islets for DC-LPS and DC-LPS/gCO immunized group are 127 and 128, respectively. (B) The extent of insulitis was determined as the number of donor cells (CD8+Thy1.1+) after isolation of infiltrating lymphocytes in the pancreas. Each symbol represents an individual mouse and bar represents the mean. **p < 0.01 with paired t-test.

Download figure to PowerPoint

gCO-treated DC-primed CD8+ T cells exhibit lower β1-integrin and impaired islet lytic capabilities

Next we explored the molecular mechanisms supporting the accumulation defect in the pancreas of autoreactive CD8+ T cells following gCO-treated DCs priming. However, because the enzymatic procedure used to isolate intrapancreatic infiltrating lymphocytes reduced their numbers, we could not perform direct in vivo phenotyping or assess the function of gCO-treated DC-primed CD8+ T cells versus untreated DC. Moreover, we hypothesized that CD8+ T cells in the periphery were not relevant for analysis because these cells have not migrated into the pancreas. To overcome this technical issue, we co-cultured ex vivo anti-HA CD8+ T cells with gCO-treated or untreated DCs. This method allowed us to screen a wide variety of surface molecules linked to T lymphocyte migration (Supporting Information Fig. 3). Most of these markers were induced following CD8+ T cell–DC encounter. However, none of these markers on gCO-treated DC-primed CD8+ T cells differed significantly compared with those of untreated DC-primed CD8+ T cells, except for β1-integrin (CD29) expression (Fig. 6A). This result was confirmed in six independent experiments (Fig. 6B). We also confirmed that the transfer of these gCO-treated DCs primed CD8+ T cells to InsHA mice, did not trigger diabetes by contrast to untreated DC-primed CD8+ T cells (Supporting Information Table 1). Taken together, these data show that homing receptor β1-integrin induction is impaired following activation by gCO-treated DCs, most probably accounting for the reduced accumulation of autoreactive CD8+ T cells in the pancreas.

image

Figure 6. gCO-treated DC-primed CD8+ T cells exhibit β1-integrin downregulation and impaired islet lytic capabilities. A total of 1 × 106 anti-HA512–520 TCR transgenic CD8+ T cells were co-cultured or not (w/o DC) with 2 × 105 HA512–520-loaded mature DCs treated (DC-LPS/HA/gCO) or not (“DC-LPS/HA” group) with gCO (350 ppm for 2 h). CD8+ T cells were analyzed for β1-integrin (CD29) expression 48 h after co-culture. Dashed line histogram: CD8+ T cells without DCs; solid black line histogram: LPS-treated DCs; gray line histogram, LPS and gCO-treated DCs. (A) One representative experiment is shown and (B) a quantitative analysis of data pooled from six independent experiments is presented. (C) Freshly isolated islets from InsHA mice were co-cultured with 0.065 × 106 DC-activated CD8+ CL4 Thy1.1+ T cells for 16 h. The percentage of surviving islets was determined by dithizone incorporation. Data shown are pooled from four independent experiments, each using 28 islets per condition. *p < 0.05, **p < 0.001, ***p < 0.001 with one-way ANOVA test.

Download figure to PowerPoint

Since a defect in integrin induction was observed, we next addressed the question of functional capacity of gCO-treated DC-primed CD8+ T cells to adhere and lysate islets. Indeed, β1-integrin has been described to be essential for pathogenic T-cell infiltration in the pancreas [21]. Moreover, β1-integrin is directly involved in target cell recognition and also for lysing abilities of these cytotoxic T cells [22]. Isolated islets from InsHA mice were co-incubated with untreated or gCO-treated DCs primed CD8+ T cells. Interestingly, the lytic activity of gCO-treated DC-primed CD8+ T cells was inhibited compared with that of untreated DC-primed CD8+ T cells (Fig. 6C). This reduced cytotoxic ability was also observed when untreated DC-primed CD8+ T cells were previously incubated with the β1-integrin ligand laminin or with an anti-β1-integrin Ab (Fig. 6C). Interestingly, the loss of lytic activity of gCO-treated DCs primed CD8+ T cells was only observed when islets were used as target but not when splenocytes were used as target in a classical CTL assay (Supporting Information Fig. 4). This last result is in agreement with that of the in vivo CTL assay (Fig. 4C) and suggests that the diminished β1-integrin expression on gCO-treated DC-primed CD8+ T cells may account for the inability to lysate islets.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

We [16, 17] and others [23] have shown that induction of HO-1 expression by chemical inducers or gene transfer inhibited LPS-induced phenotypic maturation and proinflammatory cytokine expression of DCs while preserving IL-10 expression. In a recent paper, we have demonstrated that CO is the only HO-1 end product capable of recapitulating these effects and that CORM-treated DCs led to a decrease of the IL-12/IL-10 ratio and loss of in vivo DC immunogenicity toward CD8+ T cells in mice [17]. Results concerning the effect of HO-1 overexpression on DCs have been questioned since HO-1 chemical inducers (CoPP) were used [24]. However, we show here the same cytokine secretion profile and effect on immunogenicity when DC was in vitro treated with gCO, the physiological form of CO, strongly arguing that CO is the molecule responsible for the previously described HO-1 effects on DCs. Moreover, we demonstrated that gCO-treated DCs significantly reduces diabetes incidence in the InsHA model. Neither T and B cells or IL-10 played a role in the protection against diabetes provided by CO-treated DCs since Rag-2 and IL-10 deficient mice were also protected. We can not exclude that endogenous DCs could play a role in this model but if they do it should include incorporation of OVA peptides as well as factors from CO-treated DCs. Surprisingly, gCO-treated DC did not modify peripheral CD8+ T-cell functions (proliferation, IFN-γ secretion, and CTL capacities) but rather inhibited their ability to accumulate in the pancreas. Unaffected CD8+ T-cell IFN-γ secretion or proliferation by gCO-treated DCs is in accordance with previous observations showing that IL-12 is dispensable to CD8+ T-cell priming in the context of high concentration of Ag/strong TCR affinity and TLR ligands [25, 26].

In this study, we show that CD8+ T cells primed by gCO-treated DCs were impaired in their accumulation in the pancreas. Interestingly, inhibition of alloreactive CD8+ T-cell infiltration without affecting IFN-γ secretion capacities has already been observed when HO-1-expressing donor aortas were transplanted into recipients [27]. These results, in the same line as ours, suggest that the HO-1/CO-mediated inhibition of CD8+ T-cell accumulation in tissue may not be organ restricted. Moreover, inhibition of CD8+ T-cell infiltration has also been observed in pancreas after systemic induction of HO-1 by pharmaceutical inducer [7] without receiving much attention.

Tolerization of autoreactive CD8+ T cells by gCO-treated DCs may affect homing of these cells. Homing is a multistep process that allows CD8+ T cells to cross into tissues through the endothelium of microcapillary vessels. Each of these steps (tethering, slow rolling, activation-dependent arrest, and diapedesis) is controlled by different interactions of surface molecules and multiple signaling pathways triggered by such interactions. Some aspects of diabetogenic T-cell homing to the pancreas, such as the role of homing receptor/adhesion molecules (as CD62L/MAdCAM-1, CD44/hyaluronate, α4β7/MadCAM, α4β7, α4β1/VCAM-1, and CD103/E-cadherin) [28-30] and chemokine receptors (CXCR3, CCR5, CCR7) in response to ligand gradients [31-33], have been addressed previously. We have screened many adhesion molecules and chemokine receptors and found that β1-integrin (CD29) was significantly down-regulated when CD8+ T cells were primed with gCO-treated DCs. Interestingly, this homing receptor imprinting has been shown to be determinant for autoreactive migration to the pancreas in the OVA/OT-I model [21] and for target cell recognition and CTL activation [22]. Moreover, as described in Figure 6C, β1-integrin blockade at the autoreactive CD8+ T-cell surface by specific Ab is sufficient to dramatically reduce islet destruction. This result suggests that adhesion-modulating therapies could be useful in T1D. Interestingly, this strategy has already proved to be efficient in human autoimmunes disease such as Crohn's disease [34] and MS [35] with anti-α4-integrin Ab (Natalizumab®).

Previous studies have demonstrated that HO-1 induction or CO treatment reduce the expression of the β1-integrin ligand VCAM-1 in endothelial cells [36-38]. Here, for the first time we demonstrate that CO treatment is also able to reduce β1-integrin expression on CD8+ T cells following DC priming. By itself, β1-integrin downregulation might explain this inhibition of islet destruction, however we do not know whether this defect of homing receptor induction is sufficient to explain the inhibition of diabetes occurrence in the gCO-treated DC group.

Studies on immune disorders reveal that by enhancing systemic HO-1 expression, or one of its end products, namely CO, it is possible to prevent allergy [9, 23, 39] or autoimmune diseases such as T1D [8] or MS [6]. However, neither the cells nor the mechanisms involved in this protection are known. Processing and presentation of self-Ags by steady-state DCs in draining LNs is now thought to be a major component of the maintenance of peripheral immune tolerance [40, 41]. The introduction of gCO-treated DCs which home not only to spleen but also to pancreatic LNs [42] may contribute to the tolerization of autoreactive T cells in these organs, leading to a decrease of their pathogenic potential and ultimately to a lower diabetes incidence. Thus, gCO-induced tolerance could be an interesting approach to Ag-specific immunotherapy of auto-immune diseases. Several Ag-specific strategies have already been devised to promote deletion or anergy of pathogenic autoreactive T cells [43-47], or to convert them into innocuous or even regulatory cells [48-52]. Our results showing that tolerization is induced by gCO-treated mature DCs in a T1D mouse model suggest that isolation, treatment of mature DCs with gCO, and re-infusion to individuals may represent a new way to tolerize autoreactive CD8+ T cells. Another interesting approach would be to induce HO-1 expression in DCs that are capable of cross-presentation, including the recently described dermal DCs [53]. In any event, immunomodulation by gCO emerges as a promising therapeutic alternative for organ-specific autoimmune diseases.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

Mice

The InsHA transgenic mice [19] express HA in pancreatic islets and the CL4 Thy1.1+ transgenic mice [54] express a TCR-specific for the H2Kd restricted (IYSTVASSL) epitope of HA and the Thy1.1 marker. InsHA, RAG2−/−, and CL4 Thy1.1+ mice were obtained through CDTA (Orléans, France). All animal breeding and experiments were performed under specific pathogen-free conditions in accordance with the European Union Guidelines.

Cell preparation, culture, and treatments

Treatment of DCs

BM-derived DCs were prepared as described [17]. CD11c+ cells were incubated 2 h at 37°C in poly(2-hydroxyethyl metacrylate) (Sigma) precoated 24-well Nunc plates in a 60 L anesthetic chamber filled with air with 5% CO2 with or without 350 ppm of CO (controlled by CO-detector, Kidde 900–0146). We obtained similar results with 300 to 450 ppm. During this treatment, DCs were matured with 0.5 μg/mL of LPS (Escherichia coli O111:B4; Sigma-Aldrich) and pulsed with HA512–520(IYSTVASSL) or NP147–155(TYQRTRALV) peptide (5 μM, Sigma-Aldrich) and washed thoroughly. For all experiments presented, 24 h viability was above 90% and IL-12 secretion inhibition higher than 50%.

Co-culture of DCs/CD8+ T cells

CO-treated mature DCs were co-cultured with CD8+ T cells isolated by magnetic selection (Miltenyibiotec >95% pure) from CL4 Thy1.1+ splenocytes at DC:CD8+ ratio 1:5 during 48 h.

CD25+ T-cell depletion

Recipient mice were Treg-depleted by three injections (at days −8, −5, and −2) of 100 μg of anti-CD25 mAb (PC61 clone/BioXcell) before assessing their resistance to diabetes. Rat IgG2a (BioXcell) was used as isotype control and administered with the same dose and schedule as PC61 mAb injection. CD25+ cell depletions have been monitored by cytometry.

In vitro and in vivo proliferation assays

Proliferation of CD8+ cells after DC co-culture was quantified as described in [17]. In vivo proliferation was assessed by loading CD8+ Thy1.1+CL4+ with 5 mM Vybrant® CFSE (Molecular Probes™, Invitrogen) for 5 min at room temperature. For adoptive transfer studies, 1.5 × 106 CD8+ Thy1.1+ CFSE-labeled cells were injected intravenously.

In vivo and in vitro killing assay

Spleen cells from BALB/c mice were pulsed with 5 μM of relevant H2-Kd (IYSTVASSL, HA512–520) or irrelevant H2-Kd (TYQRTRALV, NP147–155) peptide for 1 h at 37°C and incubated 5 min at RT in PBS containing 5 and 0.5 μM CFSE, respectively. For in vivo cytotoxic test, 3 × 106 of HA512–520- and NP147–155-pulsed splenocytes were co-injected i.v. into recipient mice. Mice were killed 16 h later, and spleen cells were analyzed by flow cytometry. Cytolytic activity was determined by calculating the percentage of specific lysis using the following formula: 100 – (((% HA512–520-pulsed splenocytes/NP147–155-pulsed splenocytes) with activated CL4 CD8+ T cells/(% HA512–520-pulsed splenocytes/% NP147–155-pulsed splenocytes) with nonactivated CL4 CD8 T cells) × 100). For in vitro assay, 5 × 104 of HA512–520- and NP147–155-pulsed splenocytes or isolated islets from InsHA mice were co-cultured, respectively, with 2 × 105 or 6.5 × 104 of DC-activated CD8+ CL4 Thy1.1+ T cells for 16 h.

Flow cytometry

Lymphocytes from different organs were stained with anti-CD8 (clone 53–6.7), anti-CD44 (clone IM7), anti-CD25 (clone PC61), anti-CD69 (clone H1.2F3), anti-CD90.1/Thy1.1 (clone OX-7), anti-CD29/β1-integrin (clone Ha2/5), anti-CD49d/α4 (R1–2), anti-LPAM-1/α4ß7 (clone DATK32), anti-CD103/αEß7 (clone M290), anti-CD62E (clone 10E9.6), anti-CD62L (clone Mel14), anti-CD183/CXCR3 (clone CXCR3–173), anti-CD195/CCR5 (clone C34–3448), anti-CD197/CCR7 (clone 4B12), and anti-IFN-γ (clone XMG1.2). All antibodies are from BD Biosciences. Staining was assessed using a FACSaria flow cytometer and Diva 6.1 software (Becton Dickinson). Isotype antibodies (Immunotech) were used as a negative control.

Cytokine measurement

Mouse ELISA kits were used for IL-12p70, IL-10, IL-2, and IFN-γ (BD Pharmingen). IFN-γ intracellular staining was performed as described in [19].

Diabetes

At least 8-week-old InsHA mice were injected i.v. with 1 × 106 CD8+ T cells (purity >95%) isolated from CL4 Thy1.1+ mice (Miltenyi Biotech). We obtained the same results with 0.5 to 1.5 × 106 CD8+ T cells. The following day, mice were injected i.v. with 3 × 104 matured HA512–520 loaded CD11c+ DCs from BALB/c mice. We obtained the same results with 1.5 × 104 to 3 × 104 DCs. For challenge experiments, 6 × 104 HA512–520 loaded LPS-matured DCs were injected 11 days after the first immunization. Mice were considered diabetic when urine glycosuria was above 5.5 mmol/L in two successive measurements.

Evaluation of insulitis

By histology

Pancreata were snap-frozen and cryosections (8 μm thick) were acetone-fixed. Sections were stained with H&E (Thermo Electron Corp.), and the degree of insulitis was evaluated microscopically.

By flow cytometry

Pancreata were digested with Collagenase P (1 mg/mL) in the presence of DNAse (1 mg/mL). Lymphocytes were isolated at the interface of a 40% and 60% Percoll gradient. Cells were washed and stained with anti-CD8 and Thy1.1 antibodies.

Statistical analysis

For diabetes incidence, significance was calculated using the log-rank test. For all other parameters, significance was calculated by t-paired test (Fig. 1A) or Mann–Whitney nonparametric t-test (other figures) using Prism software, indicated as follows in the figures: *p < 0.05, **p < 0.01, and ***p < 0.001.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

This project was funded by the Fondation Centaure, Fondation Progreffe, and ROTRF grant to IA, JDRF grant to PB (5–2010-640) and from the IMBIO network funded by the Région Pays de la Loire through the core facility Research and Development for Clinical Transfer.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information
  • 1
    Herold, K. C., Gitelman, S. E., Masharani, U., Hagopian, W., Bisikirska, B., Donaldson, D., Rother, K. et al., A single course of anti-CD3 monoclonal antibody hOKT3gamma1(Ala-Ala) results in improvement in C-peptide responses and clinical parameters for at least 2 years after onset of type 1 diabetes. Diabetes 2005. 54: 17631769.
  • 2
    Luo, X., Herold, K. C. and Miller, S. D., Immunotherapy of type 1 diabetes: where are we and where should we be going? Immunity 2010. 32: 488499.
  • 3
    Manicassamy, S. and Pulendran, B., Dendritic cell control of tolerogenic responses. Immunol. Rev. 241: 206227.
  • 4
    Denucci, C. C., Mitchell, J. S. and Shimizu, Y., Integrin function in T-cell homing to lymphoid and nonlymphoid sites: getting there and staying there. Crit. Rev. Immunol. 2009. 29: 87109.
  • 5
    Owens, E. O., Endogenous carbon monoxide production in disease. Clin. Biochem. 2010. 43: 11831188.
  • 6
    Chora, A. A., Fontoura, P., Cunha, A., Pais, T. F., Cardoso, S., Ho, P. P., Lee, L. Y. et al., Heme oxygenase-1 and carbon monoxide suppress autoimmune neuroinflammation. J. Clin. Invest. 2007. 117: 438447.
  • 7
    Li, M., Peterson, S., Husney, D., Inaba, M., Guo, K., Kappas, A., Ikehara, S. et al., Long-lasting expression of HO-1 delays progression of type I diabetes in NOD mice. Cell Cycle 2007. 6: 567571.
  • 8
    Hu, C. M., Lin, H. H., Chiang, M. T., Chang, P. F. and Chau, L. Y., Systemic expression of heme oxygenase-1 ameliorates type 1 diabetes in NOD mice. Diabetes 2007. 56: 12401247.
  • 9
    Xia, Z. W., Zhong, W. W., Xu, L. Q., Sun, J. L., Shen, Q. X., Wang, J. G., Shao, J. et al., Heme oxygenase-1-mediated CD4+CD25high regulatory T cells suppress allergic airway inflammation. J. Immunol. 2006. 177: 59365945.
  • 10
    Chauveau, C., Bouchet, D., Roussel, J., Mathieu, P., Braudeau, C., Renaudin, K., Tesson, L. et al., Gene transfer of heme oxygenase-1 and carbon monoxide delivery inhibit chronic rejection. Am. J. Transplant. 2002. 2: 581592.
  • 11
    Braudeau, C., Bouchet, D., Tesson, L., Iyer, S., Remy, S., Buelow, R., Anegon, I. et al., Induction of long-term cardiac allograft survival by heme oxygenase-1 gene transfer. Gene Ther. 2004. 11: 701710.
  • 12
    Soares, M. P., Lin, Y., Anrather, J., Csizmadia, E., Takigami, K., Sato, K., Grey, S. T. et al., Expression of heme oxygenase-1 can determine cardiac xenograft survival. Nat. Med. 1998. 4: 10731077.
  • 13
    Bach, F., Hancock, W. and Ferran, C., Protective genes expressed in endothelial cells: a regulatory response to injury. Immunol. Today 1997. 18: 483486.
  • 14
    Hancock, W. W., Buelow, R., Sayegh, M. H. and Turka, L. A., Antibody-induced transplant arteriosclerosis is prevented by graft expression of anti-oxidant and anti-apoptotic genes. Nat. Med. 1998. 4: 13921396.
  • 15
    Blancou, P. and Anegon, I., Editorial: heme oxygenase-1 and dendritic cells: what else? J. Leukoc. Biol. 2010. 87: 185187.
  • 16
    Chauveau, C., Remy, S., Royer, P., Hill, M., Tanguy-Royer, S., Hubert, F. X., Tesson, L. et al., Heme oxygenase-1 expression inhibits dendritic cell maturation and pro-inflammatory function but conserves IL-10 expression. Blood 2005. 106: 16941702.
  • 17
    Remy, S., Blancou, P., Tesson, L., Tardif, V., Brion, R., Royer, P. J., Motterlini, R. et al., Carbon monoxide inhibits TLR-induced dendritic cell immunogenicity. J. Immunol. 2009. 182: 18771884.
  • 18
    Dong, D. L., Chen, C., Huang, W., Chen, Y., Zhang, X. L., Li, Z., Li, Y. et al., Tricarbonyldichlororuthenium (II) dimer (CORM2) activates non-selective cation current in human endothelial cells independently of carbon monoxide releasing. Eur. J. Pharmacol. 2008. 590: 99104.
  • 19
    Lo, D., Freedman, J., Hesse, S., Palmiter, R. D., Brinster, R. L. and Sherman, L. A., Peripheral tolerance to an islet cell-specific hemagglutinin transgene affects both CD4+ and CD8+ T cells. Eur. J. Immunol. 1992. 22: 10131022.
  • 20
    Bercovici, N., Heurtier, A., Vizler, C., Pardigon, N., Cambouris, C., Desreumaux, P. and Liblau, R., Systemic administration of agonist peptide blocks the progression of spontaneous CD8-mediated autoimmune diabetes in transgenic mice without bystander damage. J. Immunol. 2000. 165: 202210.
  • 21
    Hanninen, A., Nurmela, R., Maksimow, M., Heino, J., Jalkanen, S. and Kurts, C., Islet beta-cell-specific T cells can use different homing mechanisms to infiltrate and destroy pancreatic islets. Am. J. Pathol. 2007. 170: 240250.
  • 22
    Doucey, M. A., Legler, D. F., Faroudi, M., Boucheron, N., Baumgaertner, P., Naeher, D., Cebecauer, M. et al., The beta1 and beta3 integrins promote T cell receptor-mediated cytotoxic T lymphocyte activation. J. Biol. Chem. 2003. 278: 2698326991.
  • 23
    Listopad, J., Asadullah, K., Sievers, C., Ritter, T., Meisel, C., Sabat, R. and Docke, W. D., Heme oxygenase-1 inhibits T cell-dependent skin inflammation and differentiation and function of antigen-presenting cells. Exp. Dermatol. 2007. 16: 661670.
  • 24
    Mashreghi, M. F., Klemz, R., Knosalla, I. S., Gerstmayer, B., Janssen, U., Buelow, R., Jozkowicz, A. et al., Inhibition of dendritic cell maturation and function is independent of heme oxygenase 1 but requires the activation of STAT3. J. Immunol. 2008. 180: 79197930.
  • 25
    Nembrini, C., Abel, B., Kopf, M. and Marsland, B. J., Strong TCR signaling, TLR ligands, and cytokine redundancies ensure robust development of type 1 effector T cells. J. Immunol. 2006. 176: 71807188.
  • 26
    Curtsinger, J. M., Lins, D. C. and Mescher, M. F., Signal 3 determines tolerance versus full activation of naive CD8 T cells: dissociating proliferation and development of effector function. J. Exp. Med. 2003. 197: 11411151.
  • 27
    Clarke, H. M., Shrivastava, S., Motterlini, R., Sawle, P., Chen, D. and Dorling, A., Donor HO-1 expression inhibits intimal hyperplasia in unmanipulated graft recipients: a potential role for CD8+ T-cell modulation by carbon monoxide. Transplantation 2009. 88: 653661.
  • 28
    Savinov, A. Y., Wong, F. S., Stonebraker, A. C. and Chervonsky, A. V., Presentation of antigen by endothelial cells and chemoattraction are required for homing of insulin-specific CD8+ T cells. J. Exp. Med. 2003. 197: 643656.
  • 29
    Savinov, A. Y., Rozanov, D. V., Golubkov, V. S., Wong, F. S. and Strongin, A. Y., Inhibition of membrane type-1 matrix metalloproteinase by cancer drugs interferes with the homing of diabetogenic T cells into the pancreas. J. Biol. Chem. 2005. 280: 2775527758.
  • 30
    Fousteri, G., Dave, A., Juntti, T. and von Herrath, M., CD103 is dispensable for anti-viral immunity and autoimmunity in a mouse model of virally-induced autoimmune diabetes. J. Autoimmun. 2009. 32: 7077.
  • 31
    Forster, R., Schubel, A., Breitfeld, D., Kremmer, E., Renner-Muller, I., Wolf, E. and Lipp, M., CCR7 coordinates the primary immune response by establishing functional microenvironments in secondary lymphoid organs. Cell 1999. 99: 2333.
  • 32
    Carvalho-Pinto, C., Garcia, M. I., Gomez, L., Ballesteros, A., Zaballos, A., Flores, J. M., Mellado, M. et al., Leukocyte attraction through the CCR5 receptor controls progress from insulitis to diabetes in non-obese diabetic mice. Eur. J. Immunol. 2004. 34: 548557.
  • 33
    Frigerio, S., Junt, T., Lu, B., Gerard, C., Zumsteg, U., Hollander, G. A. and Piali, L., Beta cells are responsible for CXCR3-mediated T-cell infiltration in insulitis. Nat. Med. 2002. 8: 14141420.
  • 34
    Ghosh, S., Goldin, E., Gordon, F. H., Malchow, H. A., Rask-Madsen, J., Rutgeerts, P., Vyhnalek, P. et al., Natalizumab for active Crohn's disease. N. Engl. J. Med. 2003. 348: 2432.
  • 35
    Miller, D. H., Khan, O. A., Sheremata, W. A., Blumhardt, L. D., Rice, G. P., Libonati, M. A., Willmer-Hulme, A. J. et al., A controlled trial of Natalizumab for relapsing multiple sclerosis. N. Engl. J. Med. 2003. 348: 1523.
  • 36
    Soares, M. P., Seldon, M. P., Gregoire, I. P., Vassilevskaia, T., Berberat, P. O., Yu, J., Tsui, T. Y. et al., Heme oxygenase-1 modulates the expression of adhesion molecules associated with endothelial cell activation. J. Immunol. 2004. 172: 35533563.
  • 37
    Seldon, M. P., Silva, G., Pejanovic, N., Larsen, R., Gregoire, I. P., Filipe, J., Anrather, J. et al., Heme oxygenase-1 inhibits the expression of adhesion molecules associated with endothelial cell activation via inhibition of NF-kappaB RelA phosphorylation at serine 276. J. Immunol. 2007. 179: 78407851.
  • 38
    Bergstraesser, C., Hoeger, S., Song, H., Ermantraut, L., Hottenrot, M., Czymai, T., Schmidt, M. et al., Inhibition of VCAM-1 expression in endothelial cells by CORM-3: the role of the ubiquitin-proteasome system, p38, and mitochondrial respiration. Free Radic. Biol. Med. 2012. 52: 794802.
  • 39
    Kirino, M., Kirino, Y., Takeno, M., Nagashima, Y., Takahashi, K., Kobayashi, M., Murakami, S. et al., Heme oxygenase 1 attenuates the development of atopic dermatitis-like lesions in mice: implications for human disease. J. Allergy Clin. Immunol. 2008. 122: 290297, 297 e291–e298.
  • 40
    Morelli, A. E. and Thomson, A. W., Tolerogenic dendritic cells and the quest for transplant tolerance. Nat. Rev. Immunol. 2007. 7: 610621.
  • 41
    Steinman, R. M., Hawiger, D. and Nussenzweig, M. C., Tolerogenic dendritic cells. Annu. Rev. Immunol. 2003. 21: 685711.
  • 42
    Creusot, R. J., Yaghoubi, S. S., Chang, P., Chia, J., Contag, C. H., Gambhir, S. S. and Fathman, C. G., Lymphoid-tissue-specific homing of bone-marrow-derived dendritic cells. Blood 2009. 113: 66386647.
  • 43
    Mori, F., Bianchi, L., Pucci, N., Azzari, C., De Martino, M. and Novembre, E., CD4+CD25+Foxp3+ T regulatory cells are not involved in oral desensitization. Int. J. Immunopathol. Pharmacol. 2010. 23: 359361.
  • 44
    Falb, D., Briner, T. J., Sunshine, G. H., Bourque, C. R., Luqman, M., Gefter, M. L. and Kamradt, T., Peripheral tolerance in T cell receptor-transgenic mice: evidence for T cell anergy. Eur. J. Immunol. 1996. 26: 130135.
  • 45
    Kearney, E. R., Pape, K. A., Loh, D. Y. and Jenkins, M. K., Visualization of peptide-specific T cell immunity and peripheral tolerance induction in vivo. Immunity 1994. 1: 327339.
  • 46
    Brocke, S., Gijbels, K., Allegretta, M., Ferber, I., Piercy, C., Blankenstein, T., Martin, R. et al., Treatment of experimental encephalomyelitis with a peptide analogue of myelin basic protein. Nature 1996. 379: 343346.
  • 47
    Raz, I., Elias, D., Avron, A., Tamir, M., Metzger, M. and Cohen, I. R., Beta-cell function in new-onset type 1 diabetes and immunomodulation with a heat-shock protein peptide (DiaPep277): a randomised, double-blind, phase II trial. Lancet 2001. 358: 17491753.
  • 48
    Petersen, J. S., Bregenholt, S., Apostolopolous, V., Homann, D., Wolfe, T., Hughes, A., De Jongh, K. et al., Coupling of oral human or porcine insulin to the B subunit of cholera toxin (CTB) overcomes critical antigenic differences for prevention of type I diabetes. Clin. Exp. Immunol. 2003. 134: 3845.
  • 49
    Goubier, A., Dubois, B., Gheit, H., Joubert, G., Villard-Truc, F., Asselin-Paturel, C., Trinchieri, G. et al., Plasmacytoid dendritic cells mediate oral tolerance. Immunity 2008. 29: 464475.
  • 50
    Burkhart, C., Liu, G. Y., Anderton, S. M., Metzler, B. and Wraith, D. C., Peptide-induced T cell regulation of experimental autoimmune encephalomyelitis: a role for IL-10. Int. Immunol. 1999. 11: 16251634.
  • 51
    Apostolou, I. and Von Boehmer, H., The TCR-HA, INS-HA transgenic model of autoimmune diabetes: limitations and expectations. J. Autoimmun. 2004. 22: 111114.
  • 52
    Piaggio, E., Mars, L. T., Cassan, C., Cabarrocas, J., Hofstatter, M., Desbois, S., Bergereau, E. et al., Multimerized T cell epitopes protect from experimental autoimmune diabetes by inducing dominant tolerance. Proc. Natl. Acad. Sci. USA 2007. 104: 93939398.
  • 53
    Bedoui, S., Whitney, P. G., Waithman, J., Eidsmo, L., Wakim, L., Caminschi, I., Allan, R. S. et al., Cross-presentation of viral and self antigens by skin-derived CD103+ dendritic cells. Nat. Immunol. 2009. 10: 488495.
  • 54
    Morgan, D. J., Liblau, R., Scott, B., Fleck, S., McDevitt, H. O., Sarvetnick, N., Lo, D. et al., CD8(+) T cell-mediated spontaneous diabetes in neonatal mice. J. Immunol. 1996. 157: 978983.
Abbreviations
CO

carbon monoxide

CORM

CO-releasing molecule

gCO

gazeous carbon monoxide

HO

heme oxygenase

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results
  5. Discussion
  6. Materials and methods
  7. Acknowledgements
  8. Conflict of interest
  9. References
  10. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors.

Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset.

Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
eji2478-sup-0001-S1.pdf772K

Table S1. Transfer of CD8+ T cells primed in vitro with gCO-treated DCs does not induce diabetes

Figure S1. gCO treatment does not modify mature DC phenotype.

Figure S2. gCO-treated DCs induce normal in vitro CD8+ T-cell priming and function.

Figure S3. Screening of homing and chemokine receptors following DC-CD8 co-culture.

Figure S4. Autoreactive CD8+ T cells show normal in vitro lytic activity when stimulated with CO-treated DCs.

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.