Immunotherapy with Costimulatory Dendritic Cells To Control ...

6 downloads 262 Views 2MB Size Report
and Langerhans cells are extensively ablated develop severe au- toimmune ... anergy, and regulation, both through induction of functional reg- ulatory T cells .... duced numbers and costimulatory function of conventional DCs. Although .... recipient. Recipients were between 5 and 8 wk old, and all showed signs of auto-.
The Journal of Immunology

Immunotherapy with Costimulatory Dendritic Cells To Control Autoimmune Inflammation Brendan J. O’Sullivan,*,1 Saparna Pai,*,1 Shayna Street,* Xiayou An,* Kelli P. A. MacDonald,† Michele Wong,‡ Geoffrey Strutton,‡ Steve Gerondakis,x Raymond J. Steptoe,* Barbara Fazekas de St. Groth,{,‖ Geoffrey R. Hill,† and Ranjeny Thomas* Costimulation-deficient dendritic cells (DCs) prevent autoimmune disease in mouse models. However, autoimmune-prone mice and humans fail to control expansion of peripheral autoreactive effector memory T cells (TEMs), which resist immunoregulation by costimulation-deficient DCs. In contrast, activation of DC costimulation may be coupled with regulatory capacity. To test whether costimulatory DCs control TEMs and attenuate established autoimmune disease, we used RelB-deficient mice, which have multiorgan inflammation, expanded peripheral autoreactive TEMs, and dysfunctional Foxp3+ regulatory T cells (Tregs) cells and conventional DCs. TEMs were regulated by Foxp3+ Tregs when costimulated by CD3/CD28-coated beads or wild-type DCs but not DCs deficient in RelB or CD80/CD86. After transfer, RelB and CD80/CD86-sufficient DCs restored tolerance and achieved a longterm cure of autoimmune disease through costimulation of TEM and Foxp3+ Treg IFN-g production, as well as induction of IDO by host APCs. IDO was required for regulation of TEMs and suppression of organ inflammation. Our data challenge the paradigm that costimulation-deficient DCs are required to regulate established autoimmune disease to avoid TEM activation and demonstrate cooperative cross-talk between costimulatory DCs, IFN-g, and IDO-dependent immune regulation. IFN-g and IDO activity may be good surrogate biomarkers measured against clinical efficacy in trials of autoimmune disease immunoregulation. The Journal of Immunology, 2011, 187: 4018–4030.

D

endritic cells (DCs) comprise a heterogeneous group of APCs with many specialized functions, including uptake, processing, and presentation of Ag and costimulation of naive T cells for immune priming. Moreover, mice in which DCs and Langerhans cells are extensively ablated develop severe autoimmune disease associated with myeloid proliferation and T and B lymphocyte activation, demonstrating the essential constitutive role that DCs play in peripheral self-tolerance (1). The mechanisms by which DCs maintain peripheral tolerance are still unclear, but they include central or peripheral clonal deletion, ignorance, anergy, and regulation, both through induction of functional regulatory T cells (Tregs) and through immunosuppressive mecha*University of Queensland Diamantina Institute, Princess Alexandra Hospital, Brisbane, Queensland 4102, Australia; †Queensland Institute of Medical Research, Herston, Queensland 4029, Australia; ‡Queensland Health Pathology Services, Princess Alexandra Hospital, Brisbane, Queensland 4102, Australia; xBurnet Institute, Prahran, Victoria 3004, Australia; {Centenary Institute of Cancer Medicine and Cell Biology, Newtown, New South Wales 2042, Australia; and ‖Discipline of Dermatology, University of Sydney, Sydney, New South Wales 2006, Australia 1

B.J.O. and S.P. contributed equally to this work.

Received for publication June 15, 2011. Accepted for publication August 10, 2011. The work was supported by National Health and Medical Research Council Grants 351439, 569755, and 569938. B.J.O. was supported by a Queensland Smart State Research Fellowship; S.S. and K.P.A.M. were supported by National Health and Medical Research Council career development awards; G.R.H. and S.G. were supported by National Health and Medical Research Council Research Fellowships; and R.T. was supported by Arthritis Queensland and an Australian Research Council Future Fellowship. Address correspondence and reprint requests to Prof. Ranjeny Thomas, University of Queensland Diamantina Institute, Princess Alexandra Hospital, Building 1, Research Wing, Fourth Floor, Brisbane, Queensland 4102, Australia. E-mail address: ranjeny. [email protected] The online version of this article contains supplemental material. Abbreviations used in this article: CIA, collagen-induced arthritis; DC, dendritic cell; 1-MT, 1-methyl-dl-tryptophan; TEM, effector memory T cell; Treg, regulatory T cell. Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1101727

nisms (2). The elucidation of tolerance mechanisms has important implications for development of Ag-specific immunotherapeutic strategies for autoimmune and allergic disease and allografting. Considerable evidence has demonstrated that resting or immature DCs promote peripheral tolerance constitutively, through the presentation of endogenous or apoptotic-derived self-Ag to naive CD4+ T cells, the cross-presentation of tissue-derived self-Ag to CD8+ T cells, and deletion or regulation of potentially selfreactive peripheral T cells (3–7). These principles have been used to demonstrate the usefulness of resting or costimulationdeficient DCs in tolerizing immunotherapeutic strategies. For example, transfer of hematopoietic stem cell precursors of resting DCs encoding self-Ag regulates Ag-specific CD4+ T cells in autoimmune-prone mice (8), and transfer of Ag-loaded DCs generated from mice deficient in the RelB subunit of NF-kB or CD40 regulates naive or effector CD4+ T cells in nonautoimmuneprone C57BL/6 mice through induction of Ag-specific Tregs (9, 10). In contrast, activated or mature DCs costimulate T cells through cell surface CD80/CD86 and secreted proinflammatory cytokines. DC activation may also induce regulatory molecules, including programmed cell death 1 ligands 1 and 2, IL-10, and the enzyme IDO (11–13). Through catabolism of tryptophan, functional IDO promotes tolerance through Foxp3+ Treg differentiation and function, as well as inhibition of T cell proliferation (14, 15). The expression of IDO by costimulatory DCs is further reinforced under conditions of immune activation, after reverse signaling by T cells through GITR ligand or CD80/CD86 or stimulation by IFN-g or TLR ligands. Recruitment of IDO+ macrophages, B cells, and stromal cells also occurs through this mechanism (16, 17). Despite the evidence for use of costimulation-deficient DCs for Ag-specific prevention of autoimmune disease, autoimmune-prone mice and humans with central tolerance defects fail to control

The Journal of Immunology expansion of autoreactive effector memory T cells (TEMs) in the periphery, and these autoreactive T cells resist Ag-specific regulation using or targeting costimulation-deficient DCs. For example, in the NOD model of type 1 diabetes, the effectiveness of tolerizing immunotherapies, such as anti-CD40L, rapamycin, or DCs treated with NF-kB inhibitors, is greater when administered to young mice than to mice with advanced insulitis and expanded TEMs (18–23). Identification of immunotherapeutic DC strategies for autoreactive TEMs is critically important for clinical translation to control existing autoimmune disease, because it is very challenging to trial at-risk patients, and disease-specific TEMs and even subclinical organ inflammation can be identified in patients prior to symptoms (24, 25). Given this background, we hypothesized that, in contrast to naive T cells, immunotherapy using DCs with costimulatory capacity would be required for effective control of TEMs and attenuation of established autoimmune disease. To test this, we made use of RelB-deficient mice, which have a central tolerance defect, expansion of autoreactive TEMs in the periphery, and multiorgan inflammatory disease, coupled with reduced numbers and costimulatory function of conventional DCs. Although microbial inflammatory or T cell-derived signals normally induce the nuclear translocation and transcriptional activity of RelB in DCs, inflammatory induction of MHC class II and costimulatory molecules is impaired in RelB2/2 DCs (10, 26). T cell stimulatory function of RelB2/2 DCs, including priming and cross-priming of exogenous Ag, and control of infections are deficient (10, 27, 28). Mice lacking RelB develop thymic medullary atrophy (29, 30) and, consequently, have impaired thymic negative selection and T cell autoreactivity (31). Consistent with their central tolerance and DC defects, RelB2/2 mice develop spontaneous Th2 T cell-dependent multisystem autoimmune, myeloproliferative, and allergic disease from birth, characterized by failure to thrive; splenomegaly (lymph nodes are absent); inflammatory disease of skin, eyes, lungs, liver, and pancreas; high levels of serum IgE; and death by ∼3 mo of age (29, 30). T celldeficient RelB2/2 mice do not develop inflammatory disease, indicating a defect in the control of effector T cell or Treg functions in these mice either directly or as result of the lack of interaction with mature DCs with high RelB activity (32–34). We show in this article that TEMs in this model are efficiently regulated by Foxp3+ Tregs when costimulated by RelB-sufficient DCs but not DCs deficient in RelB or CD80/CD86. After transfer, a small number of RelB and CD80/CD86-sufficient DCs could restore tolerance and achieve a long-term cure of autoimmune disease in RelB2/2 mice through costimulation of IFN-g production by TEMs and Foxp3+ Tregs, leading to RelB-independent induction of IDO expression and function by host APCs. In turn, IDO was required for regulation of TEM function and suppression of organ inflammation. Our data demonstrated the cooperative relationship between costimulatory DCs and IDO-dependent immune regulation through IFN-g–dependent cross-talk.

4019 RelB2/2 mice were bred with B6.SJLPtprca (H-2b CD45.1) mice to generate RelB+/2 (CD45.1, CD45.2) progeny. Five million splenic CD11c+ DCs from these mice were injected into RelB2/2 recipients to track migration of RelB+/2 DCs.

Cell purification CD11c+ DCs were isolated from disaggregated, collagenase-treated spleens using CD11c-conjugated immunomagnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany). The negative fraction was immunomagnetically depleted of non-T cells and CD8+ cells. Enriched CD4+ T cells were fractionated into CD252 and CD25+ populations. CD4+ CD252 T cells were further purified using CD4-coated beads. Splenic CD11c+ DCs and CD4+CD252 and CD4+CD25+ T cells were .90% pure, by flow cytometry. CD44hi and CD44lo T cells were purified from enriched CD4+ T cells by staining with CD44-PE and then selecting with anti-PE beads.

Suppression assay To assess suppressor function of CD4+CD25+ T cells, 25 3 103 purified splenic DCs were incubated with 50 3 103 CD4+CD252 T cells and 50 3 103 CD4+CD25+ T cells in the presence of 1 mg/ml anti-CD3 (clone 2KC11; BD) for 3 d in 96-well round-bottom plates, followed by [3H]thymidine (1 mCi/well; Perkin Elmer, Wellesley, MA) overnight. Results are expressed as mean cpm 6 SEM of triplicate wells.

In vitro T cell cytokine production Purified T cells were plated at 50 3 103 cells/well and stimulated with immobilized anti-CD3 (BD Pharmingen, San Diego, CA) for 72 h. Supernatants were assayed for cytokine levels using mouse Th1/Th2 and inflammation kits (BD). For intracellular staining, T cells were stimulated for 4 h with PMA/ionomycin in the presence of brefeldin and then stained for surface markers and intracellular cytokines.

Flow cytometric analysis Organs were disaggregated and filtered, and RBCs were lysed. Cells were stained for 30 min on ice in blocking buffer. Ab clones included CD4 (GK1.1), Foxp3 (FJK-16s), CD44 (IM78.1), CD62L (MEL14), CD69 (H1.2F3), CD103 (2E7), GITR (DTA-1), CD11c (N418), CD45.1 (A20), CD40 (3/23), CD80 (16-10A1), and CD86 (GL-1).

In vivo Ab and BrdU administration RelB+/2 DC-treated or untreated RelB2/2 mice were injected i.p. with 2 mg BrdU (Roche, Basel, Switzerland) every 12 h for 3 d. Splenocytes were harvested 3 d later, stained for surface CD4 and CD25, and then fixed, permeabilized, and stained with either FITC-labeled anti-BrdU (Serotec, Raleigh, NC) or isotype-control mAb.

DC transfer RelB+/2 splenic CD11c+ DCs were purified by MACS or sorted. Either 5 3 106 MACS-purified DCs (.90% CD11chi) or 2 3 106 sorted CD11chi (.98% CD11chi) DCs were injected i.v. into each RelB2/2 recipient. Recipients were between 5 and 8 wk old, and all showed signs of autoimmune disease before treatment. Mice were monitored every 5 d for weight gain, gait, agility, alopecia, progression of lacrimal disease/ keratoconjunctivitis, dermatitis, joint swelling, and feeding. Some mice received BrdU before sacrifice. To test the role of CD80 and CD86, 1 3 106 MACS-purified DCs from CD80/CD862/2 mice or from B6.SJLPtprca mice were injected i.v. into each RelB2/2 recipient. Transferred B6. SJLPtprca DCs were detected in similar number to transferred B6. SJLPtprca RelB+/2 DCs 2 wk later.

Materials and Methods

Inhibition of IDO with 1-methyl-dl-tryptophan

Mice

1-methyl-dl-tryptophan (1-MT) was obtained from Sigma-Aldrich and prepared as a 20-mmol/l stock in 0.1 N NaOH, adjusted to pH 7.4, and stored at 220˚C protected from light. RelB2/2 mice treated or not with DCs were given 2 mg/ml 1-MT solution, supplemented with aspartame, in foil-wrapped standard autoclaved drinking water bottles. Mice drank an average of 5 ml/d, and water was replaced as needed for 3 wk.

Experiments were approved by the University of Queensland Animal Ethics Committee. All mice were maintained in specific pathogen-free conditions. Mice homozygous for an insertional mutation in the RelB gene (designated RelB2/2), on a C57BL/6 background, were generated as described (29) and used for experiments at 4–9 wk of age. RelB+/2 mice display neither pathology nor immune dysfunction and were used as the source of RelBcompetent DCs. CD80/CD862/2 mice (35) were obtained from Jackson Laboratories. RelB2/2 radiation chimeras, generated by transferring bone marrow from RelB2/2 (or RelB+/2 bone marrow for control chimeras) into lethally irradiated C57BL/6 hosts (Animal Resource Centre, Perth, Australia), were used after $8 wk of reconstitution. For transfer studies,

Histology Formalin-fixed liver, spleen, and pancreas from RelB+/2 DC-treated and control RelB2/2 mice, as well as RelB+/2 mice, were sectioned and stained with H&E. BrdU incorporation in tissue sections was revealed with HRP-conjugated anti-mouse Ig (Silenus, Victoria, Australia) and

4020

IMMUNOTHERAPY WITH COSTIMULATORY DENDRITIC CELLS

diaminobenzidine (Biocare Medical, Concord, CA); hematoxylin identified cell nuclei. Images were acquired on a Nikon 50i microscope with NIS-Elements software (Nikon, Melville, NY) using a Nikon Plan Fluor 43/0.13 ‘/2 WD 17.2 or 403/0.75 Ph2 DLL ‘/0.17 WD 0.66 objective at room temperature.

Immunofluorescence microscopy Tissues were harvested and immersed in OCT freezing medium, and 10-mm sections were prepared using a microtome. Sections were fixed in ice-cold acetone, blocked with 5% BSA and 0.1% Tween-20, incubated with mAb in blocking buffer, washed, and mounted with fluorescent mounting medium. Images were acquired on an LSM-510-meta confocal microscope with Zen 2008 software (Carl Zeiss Micro Imaging, Gottingen, Germany) using a Zeiss Plan-Apochromatic 253/0.1 1-mm Korr Dic M27 objective at room temperature. Ab clones included RelB (C-20), CD11c-allophycocyanin (N418), CD4-PE (RM4-5), Foxp3-Alexa Flour 488 (FJK-16s), and CD45.1-PE (A20). DAPI identified cell nuclei.

Statistical analysis Unpaired, two-tailed Student t tests, } = 0.05, assessed whether the means of two normally distributed groups differed significantly. The Mann– Whitney U test (unpaired) was used when means were not normally distributed or for sample sizes , 10. One-way ANOVA analysis, with the Bonferroni multiple comparison posttest, was used to compare multiple means. All error bars represent SEM.

Results Reduced numbers and costimulatory capacity of splenic conventional DCs in RelB2/2 mice We first characterized the effects of RelB deficiency on splenic DC numbers, cell surface phenotype, and response to microbial and T cell-derived signals. As previously described, the number of MHC class II+CD11c+ conventional DCs in RelB2/2 mice was reduced compared with RelB+/2 mice (Fig. 1A, 1B), as the result of the specific reduction in CD4+CD82 DCs (Fig. 1A, 1C) (33, 36, 37). Although DCs freshly isolated from spleens of RelB2/2 mice expressed higher cell surface CD40 and MHC class II and equivalent levels of CD80 and CD86 relative to RelB+/2 DCs (Fig. 1A, 1D), after culture in medium or in the presence of anti-CD40 or LPS, expression of CD40, CD80, and CD86 increased significantly only if DCs expressed RelB, with the exception of a nonsignificant increase in CD86 by RelB2/2 DCs stimulated with LPS. There was no difference in viability between cultured cells from each strain. Similarly, in contrast to RelB+/2 DCs, RelB2/2 DCs secreted lower levels of IL-1b, IL-6, and IL-12p70 in response to anti-CD40 or LPS and no significant difference in IL-10 secretion (Fig. 1E), further demonstrating their reduced costimulatory capacity. Because these data are consistent with previous analyses of DCs from RelB2/2 radiation chimeras, which have no inflammatory disease, this reduction in specific cytokine secretion by DCs is RelB intrinsic and not due to DC “exhaustion” in the face of chronic autoimmune inflammation (34). In response to LPS, RelB2/2 and RelB+/2 DCs produced similar levels of TNF. Thus, consistent with their prior activation upon exposure to chronic autoimmune inflammatory stimuli, including TNF, splenic RelB2/2 DCs resembled tolerant semi-mature DCs, which are resistant to the LPS- or anti-CD40–mediated upregulation of costimulatory capacity observed for RelB+/2 DCs (38, 39). This tolerant phenotype was evident whether splenic RelB2/2 DCs were CD8+, CD4+, or double negative (data not shown). These data indicated that RelB2/2 mice, which develop severe multisystem autoimmune disease, also have a reduction in numbers of splenic DCs and their capacity for upregulation of costimulatory molecules and cytokines upon activation. Therefore, they represent an ideal model in which to test the relationship between DC costimulatory function and autoimmune disease.

Autoreactive RelB2/2 TEMs are not suppressed by Foxp3+ Tregs unless costimulated by RelB+/2 DCs We next analyzed the function of TEMs and Tregs as a result of interaction with RelB2/2 DCs. RelB2/2 mice were relatively lymphopenic (Fig. 2A), and the percentage of CD4+Foxp3+ Tregs was increased 2-fold in the spleen compared with RelB+/2 mice (data not shown), similar to other lymphopenic autoimmune mouse models (e.g., where TCR signaling is attenuated) (40). Hence, the numbers of splenic CD4+Foxp3+ T cells in RelB2/2 mice were comparable to RelB+/2 mice (Fig. 2A). Splenic CD4+Foxp32 T cells from RelB2/2 mice were enriched in activated TEMs relative to RelB+/2 mice, with increased proportions of CD4+CD44hi, CD69+, and CD62Llo cells (Fig. 2B). Splenic CD4+Foxp3+ Tregs isolated from RelB2/2 mice were also enriched in CD44hi and CD62Llo cells and expressed higher levels of activation markers associated with Treg function, including CD69, CD103, and GITR, than did RelB+/2 Tregs (Fig. 2E). When stimulated with anti-CD3 mAb, the RelB2/2 CD4+CD44hi TEMs were functional and secreted higher levels of IL-4, IL-5, and IL-10 and reduced levels of IFN-g and IL-2 than did RelB+/2 CD4+CD44hi cells (Fig. 2D). The proportion of CD4+Foxp3+ Tregs was similar in spleen, pancreas, liver, and lung of RelB2/2 mice (Fig. 2E). The increased numbers of activated CD4+CD44hi cells preferentially secreting Th2-type cytokines in RelB2/2 mice are consistent with peripheral expansion of autoreactive TEMs, in the absence of effective regulation by Foxp3+ Tregs, and with the autoimmune inflammatory dermatitis and eosinophilic infiltrates, as described in lung and skin of these mice (41). A deficiency in Treg function is supported by the pathologic similarities (scaling eyelids and tail, splenomegaly, and lymphocytic infiltrates in the skin, liver, and lungs) between Foxp3-deficient mice and RelB2/2 mice. Therefore, we tested the capacity of RelB2/2 splenic CD4+ CD252 T cells, which are enriched in TEM, to be suppressed by autologous Tregs in vitro in the presence of splenic DCs and antiCD3. RelB2/2 DCs failed to stimulate the proliferation of RelB2/2 CD4+CD252 T cells in the presence of anti-CD3, and RelB2/2 CD4+CD252 T cells could not be suppressed by CD4+CD25+ Tregs when stimulated by RelB2/2 DCs (Fig. 3A). When stimulated by RelB+/2 DCs, RelB2/2 CD4+CD252 T cell proliferation increased; concomitantly, RelB2/2 Tregs suppressed the proliferation of RelB2/2 CD4+CD252 T cells, suggesting that RelB2/2 DCs fail to deliver the necessary signals to support suppression of TEM proliferation by Tregs (Fig. 3A). Once stimulated by RelB+/2 DCs, RelB2/2 Tregs suppressed RelB2/2 CD4+CD252 T cells even more effectively than did RelB+/2 Tregs, consistent with their elevated expression of Treg activation markers. Unlike RelB2/2 CD4+CD252 T cells, RelB+/2 CD4+CD252 T cells were signaled to proliferate, albeit weakly, by RelB2/2 DCs. In this case, RelB2/2 Tregs again suppressed CD4+CD252 T cell proliferation (Fig. 3A). Indeed, proliferation of RelB2/2 CD4+CD252 T cells was significantly reduced, compared with RelB+/2 CD4+ CD252 T cells, either in response to RelB+/2 DCs (p , 0.05) or RelB2/2 DCs (p , 0.001) and anti-CD3 (Fig. 3A). The difference in proliferative response of RelB2/2 and RelB+/2 CD4+CD252 T cells is consistent with the activated TEM phenotype of RelB2/2 CD4+CD44hi T cells, which have been shown to produce low levels of IL-2 and to upregulate CD25 when stimulated, leading to a net consumption of IL-2 (42). Thus, responsiveness of effector T cells to suppression when weakly costimulated by RelB2/2 DCs depends on the enrichment in autoreactive TEMs, because RelB+/2 CD4+CD252 T cells remained susceptible to suppression by RelB2/2 Tregs. The data indicated that activated RelB2/2 peripheral TEMs escape suppression by Foxp3+ Tregs if DC signaling

The Journal of Immunology

4021

FIGURE 1. Reduced numbers and costimulatory capacity of splenic conventional DCs in RelB2/2 mice. Collagenase-digested and disaggregated splenocytes from RelB+/2 and RelB2/2 mice were stained for MHCII+CD11c+ DCs (T, B, and NK cells excluded) and CD4 and CD8. A, Proportion of CD4+ and CD8+ splenic DCs. B, Number of splenic MHCII+CD11c+ DCs. C, Number of splenic DC subsets. Results are representative of four individual mice. D, Splenocytes were cultured with or without anti-CD40 or LPS for 24 h in medium, and MHCII+CD11c+ DCs were stained for CD40, CD80, and CD86 and analyzed by flow cytometry. Mean fluorescence intensity (MFI) before and after culture is shown. E, Cytokine production from CD11c beadpurified RelB+/2 or RelB2/2 DCs cultured for 72 h with or without anti-CD40 or LPS. Data show mean 6 SEM of two separate experiments analyzing 5–10 individual mice. *p , 0.05, **p # 0.01, ***p # 0.001 (B, C, t test; D, E, one-way ANOVA followed by Bonferroni post hoc test).

4022

IMMUNOTHERAPY WITH COSTIMULATORY DENDRITIC CELLS

FIGURE 2. Expansion of activated Th2 TEMs in periphery of RelB2/2 mice. Collagenase-digested and disaggregated splenocytes from RelB+/2 and RelB2/2 mice were stained for CD4-PE and Foxp3FITC. A, CD4+ and Foxp3+CD4+ cells were enumerated by flow cytometry, and groups were compared. Each symbol represents an individual mouse. *p # 0.05 (Mann– Whitney U test, tested over two experiments). Splenocytes were stained with T cell activation markers as shown, and CD4+Foxp32 (B) or CD4+Foxp3+ (C) T cells were gated; the proportion of CD4+ cells expressing Foxp3 was 14.2 6 0.9% (mean 6 SEM, n = 4) in RelB+/2 mice and 32.4 6 1.2% (mean 6 SEM, n = 7) in RelB2/2 mice. D, CD4+CD44hi effector T cells purified from RelB+/2 or RelB2/2 spleens were cultured for 3 d with immobilized anti-CD3. Cytokine production was measured in supernatants by cytokine bead array. Data are mean 6 SEM of duplicate samples from individual mice and are representative of three separate experiments. *p # 0.05, **p # 0.01, ***p # 0.001; t test. E, Cells isolated from collagenase-digested tissues from RelB2/2 mice were stained for CD4, CD44, and Foxp3 and analyzed by flow cytometry. The percentage of Foxp3+ T cells in the CD44+CD4+ population is indicated. Representative of seven mice in two separate experiments.

is inadequate because of a RelB deficit. In functional assays, the capacity of Tregs to suppress CD4+CD252 T cells was not restored by addition of anti–IL-6 or anti–IL-1 to the cultures (data not shown), indicating that the defect did not result from T cell resistance due to excessive proinflammatory cytokine secretion; rather, expression of RelB by DCs is required for appropriate signaling of RelB2/2 TEMs (43, 44). Therefore, we tested whether costimulation of T cells by anti-CD3/anti-CD28 mAb-coated beads could provide the necessary signal for suppression of RelB2/2 TEMs when stimulated by RelB2/2 DCs. The additional bead-derived costimulatory signals promoted the proliferative response and associated IL-2 production by RelB2/2 CD4+CD252 T cells and made them more susceptible to suppression by RelB2/2 Tregs (Fig. 3B).

Anti-CD3/CD28 mAb-coated beads enhanced production of both Th1 and Th2 cytokines by TEMs but did not shift the cytokine profile from Th2 to Th1 (Fig. 3B). Thus, the failure of RelB2/2 Tregs to suppress CD4+CD252 T cells when stimulated by RelB2/2 DCs is not due to a failure to suppress the predominant Th2 cytokines secreted by RelB2/2 TEMs. Rather, when adequately costimulated, RelB2/2 TEMs secreted even higher levels of Th2 cytokines, as well as Th1 cytokines, and all cytokines became concomitantly susceptible to suppression. Addition of exogenous IL-2 to cultures of RelB2/2 CD4+CD252 TEMs or Tregs stimulated by RelB2/2 DCs enhanced proliferation (p , 0.05, Fig. 3B), although to a lesser extent than did CD3/CD28 mAb-coated beads or RelB+/2 DCs (p , 0.05; Fig. 3A, 3B), consistent with the conclusion that net IL-2 consumption underlies the T cell

The Journal of Immunology

4023

FIGURE 3. Activated RelB2/2 TEMs are not suppressed by Foxp3+ Tregs unless costimulated by RelB+/2 DCs. A, RelB2/2 or RelB+/2 CD4+CD252 T cells were stimulated with CD11c+ splenic DCs purified from RelB2/2 or RelB+/2 mice and anti-CD3, with or without RelB2/2 or RelB+/2 CD4+ CD25+ T cells. Mean (6 SEM) proliferative response is shown. B, RelB2/2 CD4+CD252 T cells were stimulated with RelB2/2 splenic DCs and anti-CD3 in the absence or presence of 10 U/ml IL-2 or CD3/CD28 mAb-coated beads, with or without RelB2/2 CD4+CD25+ T cells. Mean (6 SEM) proliferative response and cytokine production are shown. Results are representative of three separate experiments. *p # 0.05, ** p # 0.01, ***p # 0.001; one-way ANOVA, followed by the Bonferroni post hoc test.

proliferation defect, at least in part, in the presence of costimulation-poor RelB2/2 DCs (42). Transfer of RelB+/2 costimulatory DCs arrests autoimmune inflammatory disease in RelB2/2 mice To demonstrate the principle that increasing costimulation restores the capacity of Tregs to suppress TEMs mediating Th2 inflammation in vivo, we adoptively transferred RelB+/2 DCs into RelB2/2 recipients to promote suppression of inflammation by the Foxp3+ Tregs already present in the periphery in high numbers. To track the DCs, we crossed RelB2/2 mice with CD45.1 congenic mice to generate RelB+/2 DC donors expressing CD45.1 and CD45.2. CD11c+ splenic DCs purified from these mice were transferred i.v. into CD45.2 RelB2/2 recipients between 5 and 8 wk of age, when all had developed autoimmune disease. Three days later, the mice were sacrificed, and the distribution and phenotype of donor DCs were analyzed. Donor CD45.1+CD11c+

DCs were found in the spleen, liver, and lung but not the thymus. CD80 and CD86 were expressed by a higher proportion of RelB+/2 donor DCs than RelB2/2 recipient DCs (Fig. 4A). CD40 was expressed by a lower proportion of RelB+/2 donor DCs than recipient RelB2/2 DCs (Fig. 4A), consistent with strain differences ex vivo (Fig. 1D). By immunofluorescence staining, Foxp3+ Tregs were found in close proximity to nuclear RelB+CD11c+ DCs in the splenic peri-arteriolar lymphoid sheath in RelB+/2 mice (Fig. 4B). No RelB staining was observed in spleens from RelB2/2 mice (data not shown). In RelB2/2 mice treated with RelB+/2 DCs, Foxp3+ Tregs contacted CD11c+ DCs. Although infrequent, CD45.1+ donor DCs also contacted Foxp3+ Tregs (Fig. 4C). After transfer, donor DCs expressed RelB in the cytoplasm and the nucleus (Fig. 4D). In suppression assays using CD4+CD252 T cells and Tregs isolated from spleens of untreated (Fig. 5A, left) or RelB+/2 DC-treated mice (Fig. 5A, right), Tregs from untreated RelB2/2 mice were unable to suppress proliferation of CD4+

4024

IMMUNOTHERAPY WITH COSTIMULATORY DENDRITIC CELLS

FIGURE 4. Transferred RelB+/2 DCs contact Foxp3+ Tregs in spleens of RelB2/2 mice. A, A total of 5 3 106 CD11c+ splenic DCs purified from congenic RelB+/2 mice (CD45.1+, CD45.2+, RelB 2/+) was transferred i.v. into RelB2/2 (CD45.2+) mice. Five days later, cells were isolated from collagenase-digested spleens and liver and stained for CD45.1 (donor cells), CD11c, CD40, CD80, and CD86, with contour plots representing the average distribution. The percentages of CD40+, CD80+, and CD86+ DCs (minus isotype control staining) are indicated. Data are representative of two transfers. B, Immunofluorescence staining of frozen spleen sections obtained from RelB+/2 mice (Foxp3-Alexa Fluor 488, RelB-Alexa Fluor 555, CD11c-allophycocyanin) showing close proximity of nuclear RelB+ DCs (red) to Foxp3+ T cells (green). C and D, Spleens from mice, treated as in A, were harvested, frozen in OCT, sectioned, and stained for Foxp3-Alexa Fluor 488, CD45.1-PE, and CD11c-allophycocyanin. Donor DCs expressing RelB (red) are in close proximity to Foxp3+ Tregs (green) (C) and express RelB; nuclei are stained with DAPI (cyan) (D). Original magnification 3250. Scale bar, 25 mm. Results are representative of three separate experiments.

CD252 T cells, whereas CD4+CD252 T cells purified from RelB+/2 DC-treated RelB2/2 mice were suppressed by Tregs from these mice (Fig. 5A). These data were confirmed in vivo by analysis of BrdU incorporation by CD4+CD252 and CD4+CD25+ T cells in RelB2/2 mice that were treated or not with DCs. BrdU incorporation by CD4+ T cells in untreated RelB2/2 mice was increased 2–4-fold relative to RelB+/2 mice. This was significantly reduced after transfer of RelB+/2 DCs, consistent with suppression of TEM proliferation (Fig. 5B). Thus, donor DCs expressing RelB, and increased CD80 and CD86 in the proinflammatory environment, traffic to inflamed organs, contact Tregs, and promote suppression of autoreactive TEMs. To determine the consequences for autoimmune pathology, the physical condition of RelB2/2 mice was monitored for 20 d after the i.v. transfer of RelB+/2 DCs and compared with the physical condition of control RelB2/2 and RelB+/2 mice that had not received DCs. RelB2/2 mice that received RelB+/2 DCs gained weight compared with untreated RelB2/2 mice, eventually reaching a weight similar to RelB+/2 mice (Fig. 5C). The hair of treated mice regenerated and gait, tail and spine posture, agility, and eye luster improved rapidly and remained at a healthy level (Supplemental Video 1). Conversely, untreated RelB2/2 mice or mice receiving RelB2/2 chimera DCs gained little weight and suffered obvious disability after 10–15 d to the extent that they were sacrificed after 20 d for humane reasons. Organ pathology in treated mice correlated with their physical condition (Fig. 5D). When stained with H&E, tissue sections from untreated RelB2/2 pancreas showed extensive inflammation, with

infiltration of lymphoid and myeloid cells into the exocrine pancreas and islets, disruption of acinar cell architecture, atrophy of pancreatic and islet cells, and numerous proliferating cells in the inflammatory infiltrate that incorporated BrdU. In contrast, stained sections of pancreas from DC-treated RelB2/2 mice showed minimal inflammation with normal acinar cell architecture and pancreatic and islet cell morphology, with only rare parenchymal cells incorporating BrdU. These findings suggested that the pancreas had undergone regeneration in the DC-treated mice. In stained liver sections, heavy peri-portal infiltration and BrdU incorporation by inflammatory cells observed in RelB2/2 mice were no longer observed after transfer of RelB+/2 DCs. In the spleen, extramedullary hematopoiesis with increased numbers of erythroblasts and megakaryocytes and inflammatory infiltration of the red pulp were present. In RelB2/2 mice receiving RelB+/2 DCs, infiltration and number of erythroblasts and megakaryocytes were reduced, resulting in larger areas of red pulp (Fig. 5D). Importantly, a single DC treatment afforded long-term immunoregulation, with mice surviving .1 y. Analysis of organs of 12-mo-old DC-treated RelB2/2 mice revealed scattered benign infiltrates containing mostly lymphocytes (liver, Fig. 5E). Given the capacity of anti-CD3/CD28 mAb-coated beads to restore suppression in vitro, as well as the increased levels of expression of CD80 and CD86 by RelB+/2 DCs transferred to RelB2/2 mice, we tested the capacity of DCs purified from spleens of CD80 and CD86 doubly deficient mice to suppress autoimmune inflammatory disease. Mice treated with DCs lacking CD80 and CD86 showed no improvement in appearance or weight

The Journal of Immunology

4025

FIGURE 5. Transfer of RelB+/2 DCs arrests autoimmune inflammatory disease. A, Suppression assays, as described for Fig. 2, were carried out using cells isolated from the spleens of RelB2/2 mice or RelB2/2 mice, which were treated for 7 d with RelB+/2 CD11c+ DCs. **p , 0.01; t test. B, RelB2/2 mice were injected i.v. with either saline or with RelB+/2 CD11c+ DCs, along with 2 mg BrdU twice daily for 3 d. Harvested splenocytes were stained with anti-CD4, anti-CD25, and anti-BrdU Ab. Values represent the mean fold increase (6 SEM) in BrdU incorporation by RelB2/2 splenic CD4+CD25+ and CD4+CD252 T cells relative to RelB+/2 splenic CD4+CD25+ and CD4+CD252 T cells, analyzed by flow cytometry, of two or more experiments containing at least two mice examined individually. *p , 0.05, **p , 0.01; t test. C, Weight gain over 20 d following treatment with 5 3 106 CD11c+ splenic DCs purified from RelB+/2, RelB2/2 radiation chimeras (RelB2/2 chi), or CD80/862/2 mice. *p , 0.05, **p # 0.01, one-way ANOVA, followed by the Bonferroni post hoc test. D, H&E and anti-BrdU stain of organs collected from RelB+/2, RelB2/2, and RelB2/2 mice previously treated with RelB+/2 DCs. Arrows indicate inflammatory infiltrate; arrowhead (spleen) depicts megakaryocyte. Original magnification 340. Scale bar, 100 mm. E, H&E stain of liver isolated from RelB2/2 mice treated with RelB+/2 DCs for 12 mo. Original magnification 3400. Scale bar, 1000 mm (upper panel, liver) and 20 mm (lower panel, infiltrate). Results are representative of three separate experiments.

(Fig. 5C). Thus, costimulation by CD80 and CD86 is sufficient for DC-mediated regulation of autoimmune inflammatory disease in RelB2/2 mice. Costimulatory RelB+ DCs control autoreactive effector memory CD4+ T cell expansion and function We next determined whether the reduction in organ inflammation after RelB+/2 DC transfer was indeed associated with reduction in the number or function of pathogenic TEMs. Foxp32CD4+ T cells isolated from RelB+/2 DC-treated, but not CD80/CD86-deficient DC-treated, mice expressed lower levels of CD44, CD69, and GITR and higher levels of CD62L compared with untreated RelB2/2 mice, consistent with the reduced expansion of TEMs after RelB+/2 DC transfer (Figs. 5B, 6A, 6B). In contrast, expression of activation markers by CD4+Foxp3+ Tregs did not change after DC treatment (Fig. 6A), and there was no change in the proportion of CD4+ T cells expressing Foxp3 (untreated mice: 21.6 6 6.2%, n = 8; DC-treated mice: 22.7 6 6.7%, n = 5). Following DC transfer to RelB2/2 mice, the percentage of TUNEL+CD44hi TEMs was 4.7 after 3 d and 3.9 after 5 d, whereas 3.5% of CD44hi TEMs were TUNEL+ in untreated RelB2/2 mice. Thus, RelB+/2 DCs with greater costimulatory capacity did not kill TEMs by activation-induced cell death (45). The myeloproliferative disease of RelB2/2 mice is associated with the secretion of growth factors and chemokines by Th2 TEMs (46, 47). In RelB2/2 mice, CD11b+ myeloid cells comprise Gr12 inflammatory macrophages, eosinophils, and basophils and Gr1+ monocytes, myeloid progenitors, and neutrophils (Fig. 6C). Associated with the

reduction in Th2 effector function by TEMs, CD11b+ myeloid cells were reduced in RelB2/2 mice treated with RelB+/2, but not CD80/CD86-deficient, DCs (Fig. 6C). Consistent with increased costimulation by RelB+/2 DCs, 7 d after RelB+/2 DC transfer, cultures of splenic CD4+CD252 T cells stimulated by autologous DCs and anti-CD3 produced higher levels of IFN-g, IL-6, and TNF and lower levels of IL-5 when cells were derived from treated RelB2/2 mice than from untreated mice (Fig. 7A). Similarly, CD4+CD44hi splenic T cells stimulated with anti-CD3 produced lower levels of IL-5 when purified from RelB+/2 DC-treated RelB2/2 mice compared with untreated RelB2/2 mice (Fig. 7B). In the serum of RelB2/2 mice treated with RelB+/2 DCs, levels of IL-2 were increased and levels of IL-4 and IL-5 were decreased relative to RelB+/2 or untreated RelB2/2 mice (Fig. 7C). These results indicated that immunotherapy with costimulatory DCs regulates TEMs and inflammatory consequences of autoimmune disease driven by a central tolerance defect and costimulation-deficient semi-mature conventional DCs. The increase in serum IL-2 after RelB+/2 DC treatment is consistent with TEM costimulation by transferred RelB+/2 DCs, similar to the increased IL-2 secretion by RelB2/2 TEMs when costimulated by anti-CD3/28–coated beads (Fig. 3B). IDO is required for suppression of inflammation in RelB2/2 mice IDO is induced in response to a variety of signals, including IFN-g and back-signaling of CD80/CD86 by CTLA4 (48). Given the low levels of IFN-g in RelB2/2 mice and the requirement for CD80/

4026

IMMUNOTHERAPY WITH COSTIMULATORY DENDRITIC CELLS

FIGURE 6. DC costimulation by transferred DCs is required for control of peripheral inflammation. Splenocytes from untreated, RelB+/2 DC-treated, or CD802/2CD862/2 DC-treated RelB2/2 mice were stained with T cell activation markers as shown. CD4+ Foxp32 or CD4+Foxp3+ T cells were gated (A), and the proportion of CD44hi cells (left panel) and CD69+ cells (right panel) among CD4+ T cells was determined (B). *p # 0.05, **p , 0.01; one-way ANOVA, followed by the Bonferroni post hoc test. C, CD11b+Gr1+ and CD11b+Gr12 granulocytes from RelB+/2 or RelB 2/2 mice were sorted, and cytospins were prepared and stained with Giemsa (left panel). Right panel, Percentage of CD11b+ cells isolated from spleens from untreated, RelB+/2 DC-treated, or CD802/2CD862/2 DC-treated RelB2/2 mice, as determined by flow cytometry. Original magnification 3600. Data points represent individual mice analyzed over two separate experiments. **p # 0.01, ***p # 0.001, one-way ANOVA, followed by the Bonferroni post hoc test.

CD86 expression for restoration of tolerance and suppression of T cell proliferation in RelB2/2 recipients by RelB+/2 DCs, we hypothesized that RelB+/2 DC transfer promotes IDO activity. To test this, RelB2/2 mice were treated with DCs or left untreated for 3 wk. Groups of mice were administered 1-MT or vehicle in the

drinking water. Although the noncanonical NF-kB pathway (inhibitor of NFk-B kinasea/NFk-B–inducing kinase) signaling was shown to be required for CD40L and GITR-Ig–induced IDO expression (49, 50), basal and LPS/IFN-g–induced IDO expression by splenic CD11c+ DCs and MHC class II+CD11c2 B cells from FIGURE 7. Transferred RelB+/2 DCs reduce peripheral TEM-driven inflammation. A, Cytokine production from cultures of CD4+CD252 T cells and DCs purified from spleens of RelB+/2 DCtreated or untreated RelB2/2 mice, and anti-CD3 mAb. B, CD4+CD44hi T cells purified from DCtreated or untreated RelB2/2 spleens were cultured for 72 h with immobilized anti-CD3, and cytokine levels were measured in supernatants by cytokine bead array. *p # 0.05, **p # 0.01; oneway ANOVA, followed by the Bonferroni post hoc test. Data represent the mean of duplicate samples from individual mice and are representative of three separate experiments. C, Cytokine levels were measured at sacrifice in the serum of untreated, RelB+/2, and RelB+/2 DC-treated RelB2/2 mice. Data represent measurements from individual mice analyzed in duplicate.

The Journal of Immunology RelB+/2 and RelB2/2 mice was equivalent (Fig. 8A, 8B). These data indicated that IDO is not intrinsically deficient in RelB2/2 mice and is inducible in response to LPS/IFN-g. Consistent with our hypothesis, basal IDO expression was increased in RelB2/2 mice treated with RelB+/2 DCs relative to untreated controls (Fig. 8B). Although expressed by APCs in untreated RelB2/2 mice, IDO did not effectively suppress autoimmune disease until further induced by RelB+/2 DCs. IDO function was required for disease suppression and functional immune restoration, because RelB2/2 mice treated with RelB+/2 DCs had less liver damage (aspartate aminotransferase) and higher proportions of IFN-g+ effector T cells and Tregs and lower proportions of IL-4+ and IL-5+ T cells. These proportions, as well as liver damage, were reversed by 1MT in RelB+/2 DC-treated mice (Fig. 8C, 8D). Taken together, our data indicated that DCs restore tolerance and achieve a longterm cure of autoimmune disease in RelB2/2 mice through CD80/ CD86 costimulation of T cell IFN-g production by effector T cells and Tregs, as well as induction of IDO expression and function by host APCs. This is required for regulation of effector T cell function and suppression of organ inflammation.

Discussion Functional Foxp3+ Tregs are required to maintain peripheral tolerance, to prevent autoimmune disease, and to control inflammation through suppression of innate immunity and effector T cells (51, 52). Although functional proliferating peripheral Foxp3+ Tregs infiltrate the inflamed organs and spleen of RelB2/2 mice (in equivalent numbers to RelB+/2 mice), they are incapable of controlling proliferation and cytokine production by autoreactive TEMs when signaled by semi-mature DCs deficient in RelB or CD80/CD862/2 DCs. In vitro, we showed that autoreactive CD44hi TEMs escape suppression by RelB2/2 Tregs as a result of insufficient costimulation by RelB2/2 DCs and a net consumption of IL-2, consistent with a previous study (42). It is noteworthy that patients with autoimmune disease show similar defects in TCR signaling and replicative capacity (53, 54). Remarkably, tolerance is restored in RelB2/2 mice as a result of suppression of Th2-type autoreactive TEMs in vivo by Tregs in an IDO-dependent manner after transfer of costimulatory DCs, even when severe peripheral organ inflammation is present. Consistent with this CD80/86dependent costimulation, IL-2 increased in the serum of treated mice. Costimulation through CD28 is a well-described mechanism for increased IL-2 production by T cells (55). Moreover tolerance is long-lived, resulting in survival for $12 mo. Thus, DC-encoded costimulation restores tolerance through a process of selfperpetuating peripheral regulation. IDO is an important mediator of tolerance, which is expressed by multiple APCs, including plasmacytoid DCs, macrophages, and B cells (16, 48, 56–58). In DCs, IDO is induced by IFN-g and other T cell-derived signals, such as CD40L, GITR, and CTLA4 (15, 48, 50, 59). After CD40 stimulation of human DCs or GITR ligand stimulation of murine DCs, induced IDO was blocked by inhibition of alternate NF-kB pathway inhibitor of NFk-B kinasea or NFk-B–inducing kinase signaling, which would be assumed to involve RelB (49, 50). As shown in this article, DC IDO expression and induction by LPS/IFN-g may also occur in a RelBindependent fashion. In support of a signaling pathway independent of the alternate NF-kB pathway, IDO is induced through STAT1 and IRF1 signaling downstream of the IFN-g receptor in DCs (60, 61). In the current study, IDO (although expressed) did not control autoimmune disease until induced by RelB+/2 DCs. Subsequently, IDO function was required for disease suppression and functional immune restoration, including induction of IFN-g+ T cells, and reduction of Th2 cells. Of in-

4027 terest, expression of IDO by host APCs and IFN-g production by T cells were increased relative to untreated RelB2/2 mice as long as 3 wk after DC transfer, when donor DCs were no longer detected (data not shown), and mice remained tolerant 12 mo later. Given that RelB2/2 TEMs were resistant to suppression, and few T cells expressed IFN-g prior to DC treatment, we concluded that RelB+/2 DCs induced functional IFN-g+ TEMs and Tregs as a result of CD80/CD86 costimulation, and this IFN-g had a prolonged feed-forward effect on IDO expressed by DCs and other APCs in the host. This mechanism of tolerance differs from the conversion or expansion of Foxp3+ Tregs that is described in vivo after transfer or targeting of costimulation-deficient DCs (62, 63) or that occurs in vitro after stimulation of Foxp32 T cells with mature CD80/CD86+ DCs (64, 65). In contrast, the proportion of Foxp3+ Tregs did not increase; indeed, their proliferation was arrested after RelB+/2 DC transfer to RelB2/2 mice. Our conclusions that IFN-g and IDO are critical components of effective therapeutic immunoregulation of autoimmune disease are strongly supported by the literature in strains of mice prone to spontaneous Th1/Th17-mediated autoimmune disease. Although TEM and Treg IFN-g was induced in the current studies, regulatory IFN-g could also derive from NK or NKT cells to similar regulatory effect. For example, Tan and colleagues (66, 67) showed that CFA (a DC activator) suppressed disease in prediabetic NOD mice through immunoregulation dependent upon IFN-g production by NK cells. Moreover, disease onset is accelerated in NOD mice when T cell costimulation is deficient (68). Progression to diabetes can be prevented in prediabetic or recently diabetic NOD mice by treatment with anti-CD3 mAb, which enhances Treg proliferation and function, but not by the use of immature or NF-kB–deficient DCs (21, 69–72). Recently, autoantigenic peptide-MHC–coated nanoparticles, despite paradoxical expansion of disease-relevant CD8+ TEMs, were shown to suppress established diabetes in an IDO- and IFN-g–dependent manner (73). In humans with type 1 diabetes, we showed that, similar to RelB2/2 DCs, peripheral blood monocytes and DCs have a tolerant phenotype, with reduced p65 and RelB activity in response to LPS (74). Pancreatic function in patients with recent-onset diabetes is also improved after treatment with anti-CD3 Abs, leading to IFN-g secretion by TEMs and Tregs (74–79). In the collagen-induced arthritis (CIA) model of rheumatoid arthritis, several tolerizing DC strategies have been reported (80, 81), including DCs generated in the presence of dexamethasone and vitamin D, followed by LPS stimulation, which suppressed established disease. Notably, these DCs induced higher T cell expression of IFN-g compared with untreated mice with CIA (82). Furthermore, CD80/ CD86-dependent suppression of established CIA was observed after injection of exosomes derived from IDO+ DCs (83). Therefore, the evidence in mice and humans is consistent with the idea that autoreactive TEMs can be regulated by either enhancing T cell signaling or by promoting IFN-g–dependent cross-talk to IDOproducing costimulatory APCs. It will be of interest to determine whether DCs expressing CD80/CD86 suppress, rather than flare, autoimmune disease and the role played by IFN-g and IDO in a human proof-of-concept study. The current studies demonstrated that activated (RelB+CD80/ CD86+) DCs are required to achieve long-term tolerance in an autoimmune model where DC numbers and costimulatory function are reduced but Th2 cytokine-producing and IL-2–consuming TEMs are abundant. Of interest, in the MRL/lpr model of lupus, Treg expansion declines over time as a result of deficient IL-2 production by TEMs. In vitro, deficient Treg proliferation was observed in the presence of DCs from MRL/lpr, but not MRL/+, mice (84). In that model, transfer of healthy Tregs attenuated

4028

IMMUNOTHERAPY WITH COSTIMULATORY DENDRITIC CELLS

FIGURE 8. IDO activity is required for DC control of organ inflammation. A, Splenocytes isolated from RelB+/2 mice or 8-wk-old RelB2/2 mice were cultured for 24 h with LPS and IFN-g and stained for MHC class II and CD11c and intracellular IDO. IDO expression in freshly isolated and LPS/IFN-g– stimulated DCs (blue) and MHC class II+ APCs (red) is shown. B, CD11c+ DCs freshly isolated from RelB+/2, RelB2/2, or DC-treated RelB2/2 mice or cultured for 24 h with LPS and IFN-g were stained for intracellular IDO. The difference in mean fluorescence intensity (MFI) of IDO and isotype-stained samples is shown. C, RelB2/2 mice treated with or without DCs and 1-MT for 3 wk were sacrificed, and serum aspartate transaminase was measured (upper left panel). CD4+ splenic T cells from those mice were cultured for 4 h with PMA/ionomycin and stained for Foxp3 and intracellular IFN-g, IL-4, and IL-5. Proportion of CD4+Foxp32 cells staining positive for IFN-g (upper right panel), IL-4 (lower left panel), and IL-5 (lower right panel). D, Proportion of CD4+Foxp3+ cells expressing IFN-g. Data points represent individual mice analyzed over two separate experiments. *p # 0.05, **p # 0.01, *** p # 0.001; one-way ANOVA, followed by the Bonferroni post hoc test.

The Journal of Immunology disease, presumably also associated with a feed-forward effect on host DCs. Unlike the MRL/lpr model, even RelB2/2 Tregs from sick mice are functional ex vivo, and Treg numbers are maintained in vivo. The RelB2/2 model identifies a feedback loop among costimulatory DCs, IFN-g, and IDO that effectively restores regulation of TEMs by endogenous Foxp3+ Tregs. It challenges the current paradigm that DCs deficient in costimulation are required to regulate established autoimmune disease to avoid TEM activation and disease flares. Furthermore, clinical trials of immunotherapies designed to restore tolerance, including anti-CD3, and various DC preparations are either in progress or planned in type 1 diabetes and rheumatoid arthritis (85, 86). The current studies suggested that host DC responsiveness to activation with TLR ligands and IFN-g, effector and regulatory cell IFN-g production, and IDO expression and activity would be important baseline disease biomarkers to consider in preclinical studies to determine likely responsiveness to immunoregulatory strategies. Furthermore, in addition to the analysis of Treg numbers and function, IFN-g production and IDO activity would be valuable biomarkers of outcome measured against clinical efficacy in tolerance trials.

Disclosures The authors have no financial conflicts of interest.

References 1. Ohnmacht, C., A. Pullner, S. B. King, I. Drexler, S. Meier, T. Brocker, and D. Voehringer. 2009. Constitutive ablation of dendritic cells breaks self-tolerance of CD4 T cells and results in spontaneous fatal autoimmunity. J. Exp. Med. 206: 549–559. 2. Thomas, R. 2010. The balancing act of autoimmunity: central and peripheral tolerance versus infection control. Int. Rev. Immunol. 29: 211–233. 3. Kurts, C., W. R. Heath, H. Kosaka, J. F. Miller, and F. R. Carbone. 1998. The peripheral deletion of autoreactive CD8+ T cells induced by cross-presentation of self-antigens involves signaling through CD95 (Fas, Apo-1). J. Exp. Med. 188: 415–420. 4. Belz, G. T., G. M. Behrens, C. M. Smith, J. F. Miller, C. Jones, K. Lejon, C. G. Fathman, S. N. Mueller, K. Shortman, F. R. Carbone, and W. R. Heath. 2002. The CD8alpha(+) dendritic cell is responsible for inducing peripheral selftolerance to tissue-associated antigens. J. Exp. Med. 196: 1099–1104. 5. Scheinecker, C., R. McHugh, E. M. Shevach, and R. N. Germain. 2002. Constitutive presentation of a natural tissue autoantigen exclusively by dendritic cells in the draining lymph node. J. Exp. Med. 196: 1079–1090. 6. Turley, S., L. Poirot, M. Hattori, C. Benoist, and D. Mathis. 2003. Physiological beta cell death triggers priming of self-reactive T cells by dendritic cells in a type-1 diabetes model. J. Exp. Med. 198: 1527–1537. 7. del Rio, M. L., J. I. Rodriguez-Barbosa, E. Kremmer, and R. Fo¨rster. 2007. CD103- and CD103+ bronchial lymph node dendritic cells are specialized in presenting and cross-presenting innocuous antigen to CD4+ and CD8+ T cells. J. Immunol. 178: 6861–6866. 8. Steptoe, R. J., J. M. Ritchie, and L. C. Harrison. 2003. Transfer of hematopoietic stem cells encoding autoantigen prevents autoimmune diabetes. J. Clin. Invest. 111: 1357–1363. 9. Martin, E., C. Capini, E. Duggan, V. P. Lutzky, P. Stumbles, A. R. Pettit, B. O’Sullivan, and R. Thomas. 2007. Antigen-specific suppression of established arthritis in mice by dendritic cells deficient in NF-kappaB. Arthritis Rheum. 56: 2255–2266. 10. Martin, E., B. O’Sullivan, P. Low, and R. Thomas. 2003. Antigen-specific suppression of a primed immune response by dendritic cells mediated by regulatory T cells secreting interleukin-10. Immunity 18: 155–167. 11. Cobbold, S. P., E. Adams, K. F. Nolan, F. S. Regateiro, and H. Waldmann. 2010. Connecting the mechanisms of T-cell regulation: dendritic cells as the missing link. Immunol. Rev. 236: 203–218. 12. Ruffner, M. A., S. H. Kim, N. R. Bianco, L. M. Francisco, A. H. Sharpe, and P. D. Robbins. 2009. B7-1/2, but not PD-L1/2 molecules, are required on IL-10treated tolerogenic DC and DC-derived exosomes for in vivo function. Eur. J. Immunol. 39: 3084–3090. 13. Levings, M. K., S. Gregori, E. Tresoldi, S. Cazzaniga, C. Bonini, and M. G. Roncarolo. 2005. Differentiation of Tr1 cells by immature dendritic cells requires IL-10 but not CD25+CD4+ Tr cells. Blood 105: 1162–1169. 14. Baban, B., P. R. Chandler, M. D. Sharma, J. Pihkala, P. A. Koni, D. H. Munn, and A. L. Mellor. 2009. IDO activates regulatory T cells and blocks their conversion into Th17-like T cells. J. Immunol. 183: 2475–2483. 15. Ju¨rgens, B., U. Hainz, D. Fuchs, T. Felzmann, and A. Heitger. 2009. Interferongamma-triggered indoleamine 2,3-dioxygenase competence in human monocytederived dendritic cells induces regulatory activity in allogeneic T cells. Blood 114: 3235–3243.

4029 16. Johnson, B. A., III, D. J. Kahler, B. Baban, P. R. Chandler, B. Kang, M. Shimoda, P. A. Koni, J. Pihkala, B. Vilagos, M. Busslinger, et al. 2010. Blymphoid cells with attributes of dendritic cells regulate T cells via indoleamine 2,3-dioxygenase. Proc. Natl. Acad. Sci. USA 107: 10644–10648. 17. Mao, R., J. Zhang, D. Jiang, D. Cai, J. M. Levy, A. Cuconati, T. M. Block, J.-T. Guo, and H. Guo. 2011. Indoleamine 2,3-dioxygenase mediates the antiviral effect of gamma interferon against hepatitis B virus in human hepatocyte-derived cells. J. Virol. 85: 1048–1057. 18. Shoda, L. K., D. L. Young, S. Ramanujan, C. C. Whiting, M. A. Atkinson, J. A. Bluestone, G. S. Eisenbarth, D. Mathis, A. A. Rossini, S. E. Campbell, et al. 2005. A comprehensive review of interventions in the NOD mouse and implications for translation. Immunity 23: 115–126. 19. Balasa, B., T. Krahl, G. Patstone, J. Lee, R. Tisch, H. O. McDevitt, and N. Sarvetnick. 1997. CD40 ligand-CD40 interactions are necessary for the initiation of insulitis and diabetes in nonobese diabetic mice. J. Immunol. 159: 4620–4627. 20. Feili-Hariri, M., D. H. Falkner, A. Gambotto, G. D. Papworth, S. C. Watkins, P. D. Robbins, and P. A. Morel. 2003. Dendritic cells transduced to express interleukin-4 prevent diabetes in nonobese diabetic mice with advanced insulitis. Hum. Gene Ther. 14: 13–23. 21. Ma, L., S. Qian, X. Liang, L. Wang, J. E. Woodward, N. Giannoukakis, P. D. Robbins, S. Bertera, M. Trucco, J. J. Fung, and L. Lu. 2003. Prevention of diabetes in NOD mice by administration of dendritic cells deficient in nuclear transcription factor-kappaB activity. Diabetes 52: 1976–1985. 22. Nitta, Y., F. Tashiro, M. Tokui, A. Shimada, I. Takei, K. Tabayashi, and J. Miyazaki. 1998. Systemic delivery of interleukin 10 by intramuscular injection of expression plasmid DNA prevents autoimmune diabetes in nonobese diabetic mice. Hum. Gene Ther. 9: 1701–1707. 23. Bertin-Maghit, S., D. Pang, B. O’Sullivan, S. Best, E. Duggan, S. Paul, H. Thomas, T. W. H. Kay, L. C. Harrison, R. Steptoe, and R. Thomas. 2011. Interleukin-1b produced in response to islet autoantigen presentation differentiates T-helper 17 cells at the expense of regulatory T-cells: Implications for the timing of tolerizing immunotherapy. Diabetes 60: 248–257. 24. Standifer, N. E., E. A. Burwell, V. H. Gersuk, C. J. Greenbaum, and G. T. Nepom. 2009. Changes in autoreactive T cell avidity during type 1 diabetes development. Clin. Immunol. 132: 312–320. 25. Kraan, M. C., H. Versendaal, M. Jonker, B. Bresnihan, W. J. Post, B. A. t Hart, F. C. Breedveld, and P. P. Tak. 1998. Asymptomatic synovitis precedes clinically manifest arthritis. Arthritis Rheum. 41: 1481–1488. 26. O’Sullivan, B. J., and R. Thomas. 2002. CD40 ligation conditions dendritic cell antigen-presenting function through sustained activation of NF-kappaB. J. Immunol. 168: 5491–5498. 27. Moore, F., S. Buonocore, E. Aksoy, N. Ouled-Haddou, S. Goriely, E. Lazarova, F. Paulart, C. Heirman, E. Vaeremans, K. Thielemans, et al. 2007. An alternative pathway of NF-kappaB activation results in maturation and T cell priming activity of dendritic cells overexpressing a mutated IkappaBalpha. J. Immunol. 178: 1301–1311. 28. Weih, F., G. Warr, H. Yang, and R. Bravo. 1997. Multifocal defects in immune responses in RelB-deficient mice. J. Immunol. 158: 5211–5218. 29. Burkly, L., C. Hession, L. Ogata, C. Reilly, L. A. Marconi, D. Olson, R. Tizard, R. Cate, and D. Lo. 1995. Expression of relB is required for the development of thymic medulla and dendritic cells. Nature 373: 531–536. 30. Weih, F., D. Carrasco, S. K. Durham, D. S. Barton, C. A. Rizzo, R. P. Ryseck, S. A. Lira, and R. Bravo. 1995. Multiorgan inflammation and hematopoietic abnormalities in mice with a targeted disruption of RelB, a member of the NFkappa B/Rel family. Cell 80: 331–340. 31. Laufer, T. M., J. DeKoning, J. S. Markowitz, D. Lo, and L. H. Glimcher. 1996. Unopposed positive selection and autoreactivity in mice expressing class II MHC only on thymic cortex. Nature 383: 81–85. 32. Weih, F., S. K. Durham, D. S. Barton, W. C. Sha, D. Baltimore, and R. Bravo. 1996. Both multiorgan inflammation and myeloid hyperplasia in RelB-deficient mice are T cell dependent. J. Immunol. 157: 3974–3979. 33. Castiglioni, P., C. Lu, D. Lo, M. Croft, P. Langlade-Demoyen, M. Zanetti, and M. Gerloni. 2003. CD4 T cell priming in dendritic cell-deficient mice. Int. Immunol. 15: 127–136. 34. MacDonald, K. P., R. D. Kuns, V. Rowe, E. S. Morris, T. Banovic, H. Bofinger, B. O’Sullivan, K. A. Markey, A. L. Don, R. Thomas, and G. R. Hill. 2007. Effector and regulatory T-cell function is differentially regulated by RelB within antigen-presenting cells during GVHD. Blood 109: 5049–5057. 35. Schweitzer, A. N., and A. H. Sharpe. 1998. Studies using antigen-presenting cells lacking expression of both B7-1 (CD80) and B7-2 (CD86) show distinct requirements for B7 molecules during priming versus restimulation of Th2 but not Th1 cytokine production. J. Immunol. 161: 2762–2771. 36. Wu, L., A. D’Amico, K. D. Winkel, M. Suter, D. Lo, and K. Shortman. 1998. RelB is essential for the development of myeloid-related CD8alpha2 dendritic cells but not of lymphoid-related CD8alpha+ dendritic cells. Immunity 9: 839– 847. 37. Kobayashi, T., P. T. Walsh, M. C. Walsh, K. M. Speirs, E. Chiffoleau, C. G. King, W. W. Hancock, J. H. Caamano, C. A. Hunter, P. Scott, et al. 2003. TRAF6 is a critical factor for dendritic cell maturation and development. Immunity 19: 353–363. 38. Menges, M., S. Ro¨ssner, C. Voigtla¨nder, H. Schindler, N. A. Kukutsch, C. Bogdan, K. Erb, G. Schuler, and M. B. Lutz. 2002. Repetitive injections of dendritic cells matured with tumor necrosis factor alpha induce antigen-specific protection of mice from autoimmunity. J. Exp. Med. 195: 15–21. 39. Geisel, J., F. Kahl, M. Mu¨ller, H. Wagner, C. J. Kirschning, I. B. Autenrieth, and J. S. Frick. 2007. IL-6 and maturation govern TLR2 and TLR4 induced TLR

4030

40.

41.

42. 43. 44.

45.

46. 47.

48.

49.

50.

51.

52. 53.

54.

55.

56.

57.

58.

59.

60. 61.

62.

63.

IMMUNOTHERAPY WITH COSTIMULATORY DENDRITIC CELLS

agonist tolerance and cross-tolerance in dendritic cells. J. Immunol. 179: 5811– 5818. Sakaguchi, N., T. Takahashi, H. Hata, T. Nomura, T. Tagami, S. Yamazaki, T. Sakihama, T. Matsutani, I. Negishi, S. Nakatsuru, and S. Sakaguchi. 2003. Altered thymic T-cell selection due to a mutation of the ZAP-70 gene causes autoimmune arthritis in mice. Nature 426: 454–460. Barton, D., H. HogenEsch, and F. Weih. 2000. Mice lacking the transcription factor RelB develop T cell-dependent skin lesions similar to human atopic dermatitis. Eur. J. Immunol. 30: 2323–2332. Ishimaru, N., H. Kishimoto, Y. Hayashi, and J. Sprent. 2006. Regulation of naive T cell function by the NF-kappaB2 pathway. Nat. Immunol. 7: 763–772. Pasare, C., and R. Medzhitov. 2003. Toll pathway-dependent blockade of CD4+ CD25+ T cell-mediated suppression by dendritic cells. Science 299: 1033–1036. O’Sullivan, B. J., H. E. Thomas, S. Pai, P. Santamaria, Y. Iwakura, R. J. Steptoe, T. W. Kay, and R. Thomas. 2006. IL-1 beta breaks tolerance through expansion of CD25+ effector T cells. J. Immunol. 176: 7278–7287. Lenardo, M., K. M. Chan, F. Hornung, H. McFarland, R. Siegel, J. Wang, and L. Zheng. 1999. Mature T lymphocyte apoptosis—immune regulation in a dynamic and unpredictable antigenic environment. Annu. Rev. Immunol. 17: 221– 253. Xia, Y., M. E. Pauza, L. Feng, and D. Lo. 1997. RelB regulation of chemokine expression modulates local inflammation. Am. J. Pathol. 151: 375–387. Lampinen, M., M. Carlson, L. D. Ha˚kansson, and P. Venge. 2004. Cytokineregulated accumulation of eosinophils in inflammatory disease. Allergy 59: 793– 805. Munn, D. H., M. D. Sharma, and A. L. Mellor. 2004. Ligation of B7-1/B7-2 by human CD4+ T cells triggers indoleamine 2,3-dioxygenase activity in dendritic cells. J. Immunol. 172: 4100–4110. Tas, S. W., M. J. Vervoordeldonk, N. Hajji, J. H. Schuitemaker, K. F. van der Sluijs, M. J. May, S. Ghosh, M. L. Kapsenberg, P. P. Tak, and E. C. de Jong. 2007. Noncanonical NF-kappaB signaling in dendritic cells is required for indoleamine 2,3-dioxygenase (IDO) induction and immune regulation. Blood 110: 1540–1549. Grohmann, U., C. Volpi, F. Fallarino, S. Bozza, R. Bianchi, C. Vacca, C. Orabona, M. L. Belladonna, E. Ayroldi, G. Nocentini, et al. 2007. Reverse signaling through GITR ligand enables dexamethasone to activate IDO in allergy. Nat. Med. 13: 579–586. Powrie, F., R. Correa-Oliveira, S. Mauze, and R. L. Coffman. 1994. Regulatory interactions between CD45RBhigh and CD45RBlow CD4+ T cells are important for the balance between protective and pathogenic cell-mediated immunity. J. Exp. Med. 179: 589–600. O’Garra, A., and P. Vieira. 2004. Regulatory T cells and mechanisms of immune system control. Nat. Med. 10: 801–805. Emery, P., G. S. Panayi, and A. M. Nouri. 1984. Interleukin-2 reverses deficient cell-mediated immune responses in rheumatoid arthritis. Clin. Exp. Immunol. 57: 123–129. Koetz, K., E. Bryl, K. Spickschen, W. M. O’Fallon, J. J. Goronzy, and C. M. Weyand. 2000. T cell homeostasis in patients with rheumatoid arthritis. Proc. Natl. Acad. Sci. USA 97: 9203–9208. Umlauf, S. W., B. Beverly, O. Lantz, and R. H. Schwartz. 1995. Regulation of interleukin 2 gene expression by CD28 costimulation in mouse T-cell clones: both nuclear and cytoplasmic RNAs are regulated with complex kinetics. Mol. Cell. Biol. 15: 3197–3205. Fallarino, F., C. Vacca, C. Orabona, M. L. Belladonna, R. Bianchi, B. Marshall, D. B. Keskin, A. L. Mellor, M. C. Fioretti, U. Grohmann, and P. Puccetti. 2002. Functional expression of indoleamine 2,3-dioxygenase by murine CD8 alpha(+) dendritic cells. Int. Immunol. 14: 65–68. Munn, D. H., E. Shafizadeh, J. T. Attwood, I. Bondarev, A. Pashine, and A. L. Mellor. 1999. Inhibition of T cell proliferation by macrophage tryptophan catabolism. J. Exp. Med. 189: 1363–1372. Munn, D. H., M. D. Sharma, J. R. Lee, K. G. Jhaver, T. S. Johnson, D. B. Keskin, B. Marshall, P. Chandler, S. J. Antonia, R. Burgess, et al. 2002. Potential regulatory function of human dendritic cells expressing indoleamine 2,3-dioxygenase. Science 297: 1867–1870. Puccetti, P., and U. Grohmann. 2007. IDO and regulatory T cells: a role for reverse signalling and non-canonical NF-kappaB activation. Nat. Rev. Immunol. 7: 817–823. Platanias, L. C. 2005. Mechanisms of type-I- and type-II-interferon-mediated signalling. Nat. Rev. Immunol. 5: 375–386. Jeong, Y.-I., S. W. Kim, I. D. Jung, J. S. Lee, J. H. Chang, C.-M. Lee, S. H. Chun, M.-S. Yoon, G. T. Kim, S. W. Ryu, et al. 2009. Curcumin suppresses the induction of indoleamine 2,3-dioxygenase by blocking the Janus-activated kinaseprotein kinase Cdelta-STAT1 signaling pathway in interferon-g-stimulated murine dendritic cells. J. Biol. Chem. 284: 3700–3708. Chorny, A., E. Gonzalez-Rey, A. Fernandez-Martin, D. Pozo, D. Ganea, and M. Delgado. 2005. Vasoactive intestinal peptide induces regulatory dendritic cells with therapeutic effects on autoimmune disorders. Proc. Natl. Acad. Sci. USA 102: 13562–13567. Capini, C., M. Jaturanpinyo, H. I. Chang, S. Mutalik, A. McNally, S. Street, R. Steptoe, B. O’Sullivan, N. Davies, and R. Thomas. 2009. Antigen-specific

64.

65.

66.

67. 68.

69.

70. 71.

72.

73.

74.

75.

76. 77.

78.

79.

80.

81.

82.

83.

84.

85.

86.

suppression of inflammatory arthritis using liposomes. J. Immunol. 182: 3556– 3565. Yamazaki, S., T. Iyoda, K. Tarbell, K. Olson, K. Velinzon, K. Inaba, and R. M. Steinman. 2003. Direct expansion of functional CD25+ CD4+ regulatory T cells by antigen-processing dendritic cells. J. Exp. Med. 198: 235–247. Banerjee, D. K., M. V. Dhodapkar, E. Matayeva, R. M. Steinman, and K. M. Dhodapkar. 2006. Expansion of FOXP3high regulatory T cells by human dendritic cells (DCs) in vitro and after injection of cytokine-matured DCs in myeloma patients. Blood 108: 2655–2661. Lee, I.-F., H. Qin, J. Trudeau, J. Dutz, and R. Tan. 2004. Regulation of autoimmune diabetes by complete Freund’s adjuvant is mediated by NK cells. J. Immunol. 172: 937–942. Lee, I. F., H. Qin, J. J. Priatel, and R. Tan. 2008. Critical role for IFN-g in natural killer cell-mediated protection from diabetes. Eur. J. Immunol. 38: 82–89. Lenschow, D. J., K. C. Herold, L. Rhee, B. Patel, A. Koons, H. Y. Qin, E. Fuchs, B. Singh, C. B. Thompson, and J. A. Bluestone. 1996. CD28/B7 regulation of Th1 and Th2 subsets in the development of autoimmune diabetes. Immunity 5: 285–293. Amrani, A., J. Verdaguer, P. Serra, S. Tafuro, R. Tan, and P. Santamaria. 2000. Progression of autoimmune diabetes driven by avidity maturation of a T-cell population. Nature 406: 739–742. Chatenoud, L., J. Primo, and J. F. Bach. 1997. CD3 antibody-induced dominant self tolerance in overtly diabetic NOD mice. J. Immunol. 158: 2947–2954. Nishio, J., M. Feuerer, J. Wong, D. Mathis, and C. Benoist. 2010. Anti-CD3 therapy permits regulatory T cells to surmount T cell receptor-specified peripheral niche constraints. J. Exp. Med. 207: 1879–1889. Bertin-Maghit, S., D. Pang, B. O’Sullivan, S. Best, E. Duggan, S. Paul, H. Thomas, T. W. Kay, L. C. Harrison, R. Steptoe, and R. Thomas. 2011. Interleukin-1b produced in response to islet autoantigen presentation differentiates T-helper 17 cells at the expense of regulatory T cells: Implications for the timing of tolerizing immunotherapy. Diabetes 60: 248–257. Tsai, S., A. Shameli, J. Yamanouchi, X. Clemente-Casares, J. Wang, P. Serra, Y. Yang, Z. Medarova, A. Moore, and P. Santamaria. 2010. Reversal of autoimmunity by boosting memory-like autoregulatory T cells. Immunity 32: 568–580. Mollah, Z. U., S. Pai, C. Moore, B. J. O’Sullivan, M. J. Harrison, J. Peng, K. Phillips, J. B. Prins, J. Cardinal, and R. Thomas. 2008. Abnormal NF-kappa B function characterizes human type 1 diabetes dendritic cells and monocytes. J. Immunol. 180: 3166–3175. Vang, T., M. Congia, M. D. Macis, L. Musumeci, V. Orru´, P. Zavattari, K. Nika, L. Tautz, K. Taske´n, F. Cucca, et al. 2005. Autoimmune-associated lymphoid tyrosine phosphatase is a gain-of-function variant. Nat. Genet. 37: 1317–1319. Jansen, A., M. van Hagen, and H. A. Drexhage. 1995. Defective maturation and function of antigen-presenting cells in type 1 diabetes. Lancet 345: 491–492. Eibl, N., M. Spatz, G. F. Fischer, W. R. Mayr, A. Samstag, H. M. Wolf, G. Schernthaner, and M. M. Eibl. 2002. Impaired primary immune response in type-1 diabetes: results from a controlled vaccination study. Clin. Immunol. 103: 249–259. Herold, K. C., S. E. Gitelman, U. Masharani, W. Hagopian, B. Bisikirska, D. Donaldson, K. Rother, B. Diamond, D. M. Harlan, and J. A. Bluestone. 2005. 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 54: 1763–1769. Putnam, A. L., T. M. Brusko, M. R. Lee, W. Liu, G. L. Szot, T. Ghosh, M. A. Atkinson, and J. A. Bluestone. 2009. Expansion of human regulatory Tcells from patients with type 1 diabetes. Diabetes 58: 652–662. Morita, Y., J. Yang, R. Gupta, K. Shimizu, E. A. Shelden, J. Endres, J. J. Mule´, K. T. McDonagh, and D. A. Fox. 2001. Dendritic cells genetically engineered to express IL-4 inhibit murine collagen-induced arthritis. J. Clin. Invest. 107: 1275– 1284. van Duivenvoorde, L. M., P. Louis-Plence, F. Apparailly, E. I. van der Voort, T. W. Huizinga, C. Jorgensen, and R. E. Toes. 2004. Antigen-specific immunomodulation of collagen-induced arthritis with tumor necrosis factor-stimulated dendritic cells. Arthritis Rheum. 50: 3354–3364. Stoop, J. N., R. A. Harry, A. von Delwig, J. D. Isaacs, J. H. Robinson, and C. M. Hilkens. 2010. Therapeutic effect of tolerogenic dendritic cells in established collagen-induced arthritis is associated with a reduction in Th17 responses. Arthritis Rheum. 62: 3656–3665. Bianco, N. R., S. H. Kim, M. A. Ruffner, and P. D. Robbins. 2009. Therapeutic effect of exosomes from indoleamine 2,3-dioxygenase-positive dendritic cells in collagen-induced arthritis and delayed-type hypersensitivity disease models. Arthritis Rheum. 60: 380–389. Yang, C.-H., L. Tian, G.-S. Ling, N. J. Trendell-Smith, L. Ma, C.-K. Lo, D. I. Stott, F. Y. Liew, and F.-P. Huang. 2008. Immunological mechanisms and clinical implications of regulatory T cell deficiency in a systemic autoimmune disorder: roles of IL-2 versus IL-15. Eur. J. Immunol. 38: 1664–1676. Hilkens, C. M., J. D. Isaacs, and A. W. Thomson. 2010. Development of dendritic cell-based immunotherapy for autoimmunity. Int. Rev. Immunol. 29: 156– 183. Chatenoud, L. 2010. Immune therapy for type 1 diabetes mellitus-what is unique about anti-CD3 antibodies? Nat. Rev. Endocrinol. 6: 149–157.