Plasmacytoid dendritic cells (pDCs) have both stimulatory and regulatory effects on T cells. pDCs are a major CNS-infiltrating dendritic cell population during experimental autoimmune encephalomyelitis but, unlike myeloid dendritic cells, have a minor role in T cell activation and epitope spreading. We show that depletion of pDCs during either the acute or relapse phases of experimental autoimmune encephalomyelitis resulted in exacerbation of disease severity. pDC depletion significantly enhanced CNS but not peripheral CD4+ T cell activation, as well as IL-17 and IFN-γ production. Moreover, CNS pDCs suppressed CNS myeloid dendritic cell-driven production of IL-17, IFN-γ, and IL-10 in an IDO-independent manner. The data demonstrate that pDCs play a critical regulatory role in negatively regulating pathogenic CNS CD4+ T cell responses, highlighting a new role for pDCs in inflammatory autoimmune disease.
Experimental autoimmune encephalomyelitis (EAE)3 is a widely used model for multiple sclerosis (MS) (1) that is initiated and driven by myelin-specific CD4+ T cells producing IFN-γ and TNF (by Th1) and IL-17 (by Th17) in the CNS (2, 3). Disease progression in relapsing EAE (R-EAE) is characterized by “epitope spreading,” where new T cell responses to myelin epitopes distinct from the priming epitope develop (4). Our recent evidence indicates that peripherally derived myeloid dendritic cells (mDCs) prime naive CD4+ T cells in the CNS, inducing a Th17 dominant phenotype (5, 6). In contrast, CNS plasmacytoid dendritic cells (pDCs) are inefficient at inducing the proliferation of and cytokine production by naive and activated myelin-specific CD4+ T cells, although the pDCs can ingest myelin proteins (6).
Enhanced numbers of activated pDCs have been described in MS, and activated pDCs are associated with Sjögren’s syndrome, lupus, and psoriasis (7, 8, 9, 10, 11). It was therefore important to understand the role of pDCs in MS/EAE pathogenesis. Using a mAb (anti-mPCDA-1) to deplete pDCs from the CNS following R-EAE induction, we show that pDC depletion caused the rapid exacerbation of EAE severity in both the primary acute and relapse phases. Mechanistically, pDC depletion did not affect the frequency of myelin-specific CD4+ T cells in peripheral lymphoid organs but markedly enhanced CNS CD4+ T cell activation as well as IL-17 and IFN-γ production. Moreover, CNS pDCs suppressed CNS mDC-driven production of IL-17, IFN-γ, and IL-10 in an IDO-independent manner.
Materials and Methods
Female SJL/J mice were purchased from Harlan Sprague Dawley (SJL/JCrHsd) or Taconic Farms (SJL/JCrNtac). Mice were housed and cared for according to Northwestern University (Chicago, IL) Institutional Animal Care and Use Committee-approved protocols.
Induction of EAE
As previously reported using 50 μg of proteolipid protein peptide 139–151 (PLP139–151) (6).
Peptides and antibodies
PLP139–151 (HSLGKWLGHPDKF) was synthesized to >95% purity by Genemed Synthesis. Conjugated Abs were purchased from BD Pharmingen or eBioscience.
Depletion of pDCs
Mice received i.p. injections of 250 μg of anti-mouse plasmacytoid dendritic cell antigen-1 (anti-mPDCA-1) (clone JF05-1C2, rat IgG2b; Miltenyi Biotec) or purified rat IgG2b (eBioscience) every other day for four treatments.
Isolation of cells from secondary lymphoid tissues and CNS
As previously reported (6).
Flow cytometric analysis and gating
For analysis of cytokines, CNS cells were cultured for 4 h in R10 medium (6) plus GolgiStop (BD Biosciences) according to the manufacturer’s recommendations. Cells were stained with five- or six-color Ab mixtures. Amine-reactive, fixable, live/dead viability dye was used according to the manufacturer’s instructions (Molecular Probes/Invitrogen) and dead cells were excluded. Data were acquired on an BD LSR II cytometer (BD Biosciences) and analyzed using FACSDiva (BD Biosciences) or Flow Jo (Tree Star) software.
CD4+ T cell activation assay
Cell populations were flow sorted as described in Ref. 6 to >98% purity. Sorted populations were >95% pure. APCs (2 × 104) were cocultured with 105 CD4+ T cells from the CNS or spleen for 96 h with R10 medium, 5 μg/ml antiCD3, and 200 μM 1-methyl-d-tryptophan (1-MT; Sigma-Aldrich) prepared according to Ref. 12 . Carrier solution alone was added as a control. Cells were assessed for viability by flow cytometry as described above, and cytokines in the supernatants were assessed by cytokine bead array for levels of IL-10, IL-17, and IFN-γ according to manufacturers instructions (Upstate Biotechnology).
ELISPOT assays were performed as previously described (13) with 106 cells plus 10 μM PLP139–151.
Immunohistochemistry was performed on 6-μm-thick frozen cerebellar and lumbar spinal cord sections from PBS-perfused mice as previously described (13). Staining was analyzed using a Leica DM5000B fluorescent microscope and Advanced SPOT software.
Differences between groups were determined using an unpaired Student’s t test and the Mann Whitney U test.
Results and Discussion
Anti-mPDCA-1 mAb efficiently and specifically depletes CNS-infiltrating pDCs during EAE
pDCs are a minor subset of dendritic cells (DCs) in the secondary lymphoid tissues of most mouse strains (14), comprising 23% of DCs and 1.15% of cells in the lymph nodes (LNs) of SJL/J mice (data not shown). Strikingly, CNS infiltrates during EAE contain 37.7% pDCs, 5.4% of the total CNS mononuclear cell population (6). To investigate the role of pDCs during EAE, anti-mPDCA-1 mAb (15) was used to deplete pDCs during EAE onset. One day after pDC depletion, pDCs were depleted from LNs (not shown) (15) and CNS (92.4 ± 0.9% of CNS pDCs) (Fig. 1, A and B). During the relapse phase of EAE, 9 days following pDC depletion the number of CNS pDCs returned to control levels in pDC-depleted animals (Fig. 1,A), and, importantly, anti-mPDCA-1 treatment did not affect the numbers of CNS mDCs or macrophages (Fig. 1,B). Following pDC depletion, a few mPDCA-1+ cells remained in the meningeal area of the CNS; however, no mPDCA-1+ pDCs were detected in the parenchyma of the spinal cord or cerebellum (Fig. 1 C). Inefficient clearing of pDCs from the blood vessel-rich areas of the CNS correlated with low level mPDCA-1 expression on blood CD11c+B220+ pDCs (data not shown).
pDCs regulate the severity of EAE
The depletion of pDCs at the onset of R-EAE resulted in a significant exacerbation of peak clinical disease (Fig. 2,A). Clinical scores in pDC-depleted mice returned to control levels 2–3 days after the last anti-mPDCA-1 mAb treatment, consistent with a report showing that pDC numbers recover 3–5 days following anti-mPDCA-1 depletion (15). That pDC depletion caused an immediate increase in clinical severity with a rapid return to control levels upon pDC reconstitution suggests that pDCs have a direct, acute regulatory effect on CNS autoimmune disease. pDCs were then depleted during the primary relapse of EAE, leading to enhanced EAE and the abrogation of secondary remission (Fig. 2 B).
The depletion of pDCs could have affected the priming of pathogenic T cells in the periphery (16), explaining the clinical outcome of pDC depletion (Fig. 2, A and B). However, the frequency of IL-17-, IFN-γ-, and IL-2-producing CD4+ T cells specific for the immunizing peptide in the LNs 1 day following the final mPCDA-1 mAb injection was unchanged compared with controls (Fig. 2 C). In fact, PLP139–151-specific Th17 cells were significantly reduced in the spleens of pDC-depleted mice, which may be reflective of reduced EAE severity (3) and not enhanced severity as observed in the clinical experiments.
pDCs modulate the activation and frequency of Th1 and Th17 cells in the CNS
Clinical EAE generally correlates with the number and activation status of CNS-infiltrating effector CD4+ T cells and inversely with regulatory T cell (Treg) numbers (17). We found that pDC depletion did not significantly affect the numbers of CNS CD4+ T cells and Foxp3+ CD4+ Tregs (p = 0.2; Fig. 3 A). We previously showed that CNS pDCs isolated during EAE poorly activate both naive and activated myelin-specific T cells in comparison to mDCs (6). Because there was little change in the T cell numbers, it is unlikely that the primary function of pDC during EAE is to prime and expand CD4+ T cells in the CNS.
Strikingly, however, in contrast to peripheral responses the CNS CD4+ T cells were highly activated and produced more IL-17 and IFN-γ in the absence of pDCs. Following pDC depletion, CNS CD4+ T cells were significantly more activated than controls as assessed by down-regulation of CD45RB and up-regulation of CD25 (Fig. 3,B). Endogenous production of inflammatory cytokines by CNS CD4+ T cells was determined by incubating CNS isolates, which contain pathogenic T cells and DCs presenting endogenous myelin peptides (6), with GolgiStop for 4 h and then analyzing the accumulated cells expressing IFN-γ and IL-17. The frequency of CNS Th17 cells was increased by an average of 1.6 ± 0.38-fold and that of Th1 cells by 1.6 ± 0.27-fold, and more cytokine per cell (enhanced mean fluorescence intensity) were produced in pDC-depleted mice (Fig. 3 C). Thus, CNS pDCs promote the accumulation of CD4+ T cells and Tregs in the target organ but strongly modulate the activation status of CD4+ T cells and, importantly, the frequency of CNS Th1 (IFN-γ) and Th17 (IL-17) cells.
CNS pDCs actively suppress IL-17, IFN-γ, and IL-10 production by CNS CD4+ T cells in an IDO-independent manner
We next sought to determine the mechanism by which pDCs regulate CNS CD4+ T cell activation and cytokine production. pDCs have been implicated in inducing T cell anergy through IFN-α and IL-10 production (18) or TGF-β (19). Using quantitative PCR, we have previously shown that CNS pDCs expressed low levels of TGF-β transcripts (6) and similar levels of IFN-α4 mRNA (not shown) compared with other CNS APCs. In addition, IL-10 levels were lower in CNS pDC-CNS T cell cocultures (Fig. 4 A) and CNS pDCs stimulated with CD40L (not shown) compared with CNS mDCs. Based on the low production by CNS pDCs, it is unlikely that TGF-β, IL-10, or IFN-α4 is a dominant pathway for pDC suppression of CD4+ T cells in the CNS.
pDCs are known to produce the T cell inhibitor IDO in response to IFNs (12, 20). To determine whether CNS pDCs produce IDO that suppresses CD4+ T cells during EAE, CNS pDCs and mDCs isolated during acute R-EAE were cocultured with CD4+ T cells from the CNS and spleen of the same animals with or without 1-MT. pDCs were poor stimulators of CNS CD4+ T cell survival and expansion ex vivo, and the addition of 1-MT had little effect on CNS T cell viability in the presence of pDCs or mDCs (not shown). In agreement with our previously published work (6), CNS mDCs supported the highest levels of IL-17, IFN-γ, and IL-10 production by both splenic and CNS-derived T cells (Fig. 4). The IDO inhibitor 1-MT enhanced mDC-induced CD4+ T cell IL-17 and IFN-γ secretion and decreased IL-10 production. However, 1-MT had no affect on cytokine production in CD4+ T cells cultured with CNS pDCs. Most profoundly, when CNS pDCs were cocultured with CNS mDC and CNS T cells, production of IL-17, IFN-γ,and IL-10 was significantly suppressed (Fig. 4 A). These results indicate that CNS pDCs regulate CD4+ T cell cytokines in an active manner that dominates that of mDC-driving Th17 cells in the CNS (6) and that regulation is via an IDO-independent pathway. IFN-β modulates IFN-γ and IL-17 production by human PBMCs (21); thus, pDC production of IFN-β (22) may play a primary role in pDC modulation of Th1/Th17 activation during R-EAE concordant with the observation that IFN-β treatment is therapeutic in both EAE and MS (23, 24). We are currently investigating the role of IFN-β production by pDCs during EAE.
In summary, we demonstrate an acute, dominant regulatory role for pDCs in CNS autoimmune disease. pDC depletion results in exacerbated EAE but, once pDCs return to normal levels, relapse severity returns to control levels. Normal relapses suggest that the priming of naive T cells in the CNS is unaffected. pDCs suppress mDC-dependent induction/expansion of CNS Th17 and Th1 cells (Fig. 4 A). pDC suppression of T cell cytokine production is IDO independent but is not due to the killing of T cells, because in CNS mDC/pDC cocultures CNS CD4+ T cells have enhanced viability (not shown). These data support a dominant regulatory role for pDCs during EAE in that pDCs recruited to the CNS limit pathology by regulating T cell activation and cytokine production. Treatments that support and expand regulatory pDCs may therefore be attractive therapies for T cell-mediated autoimmune diseases.
We thank Mat Degutes (Northwestern University, Chicago, IL) for technical assistance, James Marvin (NWU) for cell sorting, Dr. Xunrong Luo (Northwestern University), and Drs. Qizhi Tang and Jeff Bluestone (University of California, San Francisco, CA) for helpful comments and discussion.
Drs. Fischer and Dzionek are currently employees of Miltenyi Biotec GmbH, which produced the mPDCA-1 antibody used in the study to deplete plasmacytoid dendritic cells.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported in part by U.S. Public Health Service, National Institutes of Health Research Grant NS-030871, National Multiple Sclerosis Society (NMSS) Research Grant RG 3793-A-7, NMSS Postdoctoral Fellowship Grant FG 1563 A-1 (to S.L.B.), and a grant from the Myelin Repair Foundation.
Abbreviations used in this paper: EAE, experimental autoimmune encephalomyelitis; DC, dendritic cell; LN, lymph node; mDC, myeloid DC; mPDCA, mouse plasmacytoid dendritic cell antigen-1; MS, multiple sclerosis; 1-MT, 1-methyl-d-tryptophan; pDC, plasmacytoid DC; PLP139–151, proteolipid protein peptide 139–151; R-EAE, relapsing EAE; Treg, regulatory T cell.