|Home | About | Journals | Submit | Contact Us | Français|
DC in the CNS have emerged as the major rate-limiting factor for immune invasion and subsequent neuroinflammation during EAE. The mechanism of how this is regulated by brain-localized DC remains unknown. Here, we describe the ability of brain-localized DC expressing B7-H1 molecules to recruit CD8+ T cells to the site of inflammation. Using intracerebral microinjections of B7-homologue 1-deficient DC, we demonstrate a substantial brain infiltration of CD8+ T cells displaying a regulatory phenotype (CD122+) and function, resulting in a decrease of EAE peak clinical values. The recruitment of regulatory-type CD8+ T cells into the CNS and the role of brain DC expressing B7-homologue 1 molecules in this process open up the possibility of DC-targeted therapeutic manipulation of neuroinflammatory diseases.
DC emerge as a crucial immune cell population during development and maintenance of CNS-specific autoimmunity and inflammation. The presence of CNS DC during neuroinflammation is sufficient for T-cell re-activation within the CNS . The frequency, distribution, and phenotype of DC represent limiting factors in the induction as well as the effector phase of autoimmune demyelinating disease . Also, the development and function of Treg cells is closely linked to DC, forming an intimate bidirectional influence [2-4]. Thus, DC were considered relevant players in the immunopathogenesis of T-cell-mediated diseases (reviewed in ). Studies of CD8+ Treg cells in the context of neuroinflammation have received less attention, although many attempts to understand the cellular mechanisms of CD8+ Treg-mediated suppression during EAE have recently been made [6-9].
The potential of DC and Treg as therapeutic targets in autoimmune disorders including multiple sclerosis has therefore been recognized. Accordingly, the generation of regulatory CD8+ T-cell subsets through T-cell interaction with immature DC  or plasmacytoid DC  has proven a tolerogenic role of DC in the generation of suppressor cells in humans. To delineate the factors that might be influencing DC function during their interaction with T cells, we here used DC isolated from mice deficient for the inhibitory molecule B7 homologue 1 (B7-H1), also known as programmed death ligand 1 . The co-stimulation of B7-H1 expressed on human and murine mature DC  plays a key role in regulating T-cell activation and tolerance .
The increasing number of myelin oligodendrocyte glycoprotein (MOG)-loaded DC in the brain accelerates the onset of EAE, resulting in exacerbated clinical severity and longer extent of the disease . Whether B7-H1-regulated interactions on brain-localized DC would influence the phenotypic profile of DC-recruited T cells into the CNS is currently not known. Using intracerebral (i.c.) microinjections of DC deficient for B7-H1, we show an unexpected beneficial effect of these cells resulting in a decreased EAE peak clinical score and a mild delay in the onset of this disease. This is accompanied by increased neuroantigen-specific CD8+ Treg recruitment into the CNS. These neuroantigen-specific CD8+ and CD122-expressing Treg appear to be suppressors of MOG TCR-specific CD4+ T-cell proliferation in vitro. Our data describe the importance of B7-H1 regulation on DC for the CD8+ Treg recruitment into the CNS during neuroinflammation.
We have previously described that the i.c. injection of WT DC loaded with MOG peptide significantly accelerated the onset of EAE in comparison with i.c. PBS-injected animals . We have also demonstrated that antigen specificity was a critical contributing factor in this process, as MOG-irrelevant OVA antigenpulsed DC (DCOVA) microinjected into the brain induced the preferential recruitment of OVA-specific T cells into the CNS and did not lead to the accumulation of other antigen-specific T cells . Analyzing the effects of i.c. injection of single versus two DC populations pulsed with two different antigens on antigen-specific T-cell recruitment and function in the CNS, we showed that antigen-specific T cells responded to i.c. injected DC and entered the brain in detectable numbers only when DC were pulsed with their specific antigen . Finally, only a modest acceleration of EAE clinical scores and cellular infiltration was observed in response to i.c. DCOVA injections . These experiments provide evidence that the augmented accumulation of antigen-specific CD4+ and CD8+ T cells in the brain in response to i.c. WT DC contributes to the amplification of neuroinflammation at early and late stages of EAE, suggesting a critical importance of DC frequency and function for the outcome of the disease. However, the molecules involved in DC–T-cell interaction and the exact mechanism of DC function in the CNS remain unknown. Understanding the mechanism of DC function during CNS autoimmunity should help in the development of a cell-based therapy in which the tolerogenic properties of DC could be induced as a means of suppressing pathogenic T cells. One approach is to study the impact of DC on CNS-mediated immune reactions in the absence (deficiency) or over-expression of tolerogenic molecules on DC.
Therefore, we examined the impact of costimulatory molecule-deficient DC on immune responses of CD4+ versus CD8+ T cells in the context of neuroinflammation by using MOG-induced EAE. For this, we used the model in which MOG antigen was delivered by B7-H1-deficient or WT DC injected directly into the CNS. Equal volume of sterile PBS injected into the brain was used as a control. Five days post injection the animals were immunized with MOG-EAE. In this experimental setup, animals were injected i.c. with PBS, WT DC or B7-H1-/- DC and compared (Fig. 1A, upper panel). Interestingly, the deficiency of B7-H1 expression on i.c. injected DC resulted in a delay in EAE onset compared with that of the group injected with WT DC. (Fig. 1A, lower panel), but was still earlier than the EAE onset of i.c. PBS control mice. Additionally, the peak clinical score of EAE decreased in mice i.c. injected with B7-H1-/- DC. These results indicate that there might be a dual role for the inhibitory function of B7-H1 on DC in the induction and effector phases of neuroinflammatory disease.
To dissect the mechanism of B7-H1-/- DC-mediated amelioration of EAE, we analyzed T-cell infiltration into the CNS and periphery at different stages of immune response using flow cytometry. Similar to previously published data , we observed a significant accumulation of both CD4+ and CD8+ T cells in the CNS of mice injected with WT DC compared with PBS-injected control group during EAE onset (day 7 post EAE) (Fig. 1B and C). The accumulation of CD4+ and CD8+ T cells was also significantly amplified in mice i.c. injected with B7-H1-/- DC compared with both PBS and WT DC groups (Fig. 1B and C). Notably, an amplified infiltration of CD8+ T cells in the CNS was observed in the group i.c. injected with B7-H1-/- DC (Fig. 1B, lower right panel, and Fig. 1C), which increased substantially within the CNS during the resolution of EAE (data not shown).
A rapid proliferation of CD4+ Treg within the CNS during the natural resolution of autoimmune inflammation has already been described . To address the question of whether an increase in the number of CD8+ T cells in the CNS was triggered by modification of the CNS milieu by B7-H1-/- DC injection or by an increase in their peripheral proliferation, we also combined in vivo labeling with BrdU and ex vivo staining for determination of cell phenotype in the periphery (cervical LN and spleen) and CNS 7 days post EAE immunization (Fig. 2A). The level of BrdU incorporation in the cervical LN and spleen in the group injected with WT DC was elevated in comparison with the PBS and B7-H1-/-DC-treated groups, indicating increased T-cell proliferation outside the CNS before EAE onset (Fig. 2B, left and middle panels). These data are in accordance with the observed EAE course. We observed a CNS-restricted increase in the level of proliferation among CD8− T cells, which incorporated BrdU in all experimental groups (Fig. 2B, right panel). However, a higher percentage of CNS-infiltrating BrdU+CD8+ T cells was observed in the B7-H1-/- DC-treated group compared with both PBS and WT DC-treated animals (Fig. 2B, right panel). These results suggest an early infiltration of expanded CD8+ T cells in response to i.c. DC injection in the target organ (CNS) during neuroinflammation. This infiltration seems to be facilitated by the absence of B7-H1 expression on DC, loaded with MOG-peptide and directly injected into the CNS.
During MOG-induced EAE, brain-derived CD8+ T cells were shown to be the likely cell type that is able to regulate the disease in animal models [18, 19]. To analyze the phenotype of CNS-recruited CD8+ T cells, we used different markers described for Treg cells such as CD25, FoxP3, and CTLA-4 (reviewed in ). To further distinguish between effector and regulatory CD8+ T cells, we used CD122 as a marker recently identified to be expressed on natural CD8+ Treg cells [20-22]. A beneficial role of CD8+CD122+ natural Treg was recently demonstrated during the recovery phase of EAE . In accordance with previously published data , CD8+CD122+ T cells isolated from the brain after i.c. injections were negative for CTLA-4 and FoxP3 expression but expressed moderate levels of CD25, which could be due to the activation status of these cells in the brain during EAE (data not shown). Consistent with previous observations, the percentage of natural CD8+ Treg based on CD122 expression constituted around 10% of all CD8+ splenocytes obtained from 6–7 wk old mice  (data not shown). Generally, a significantly higher proportion of CD8+CD122+ natural Treg could be found among proliferating BrdU+-incorporated versus BrdU− lymphocytes on day 7 post EAE (Fig. 2C, left and middle panels) in the cervical LN. Consistent with delayed EAE course, the frequency of proliferating BrdU+CD8+CD122+ Treg was increased in the group of i.c. PBS-injected group (22±2%) and this was even amplified in B7-H1-/- DC-injected animals (39±3%) compared with WT DC-treated group (13±3.5%) (Fig. 2C, left panel). This data suggests the significance of emergence and recruitment of a regulatory CD8+ T-cell population during EAE confinement.
As a control, we observed an altered ratio between encephalitogenic and CD4+CD25+ FoxP3-expressing Treg cells towards encephalitogenic phenotype in i.c. WT DC-treated group in comparison with i.c. PBS control animals . Interestingly, encephalitogenic and CD4+CD25+ FoxP3-expressing Treg were similarly distributed between WT DC and B7-H1-/- DC-injected groups during the EAE (data not shown).
The expression of chemokine receptor 6 (CCR6) on Treg enables their migration to the inflamed sites . Therefore, we analyzed the expression of CCR6 on proliferating versus non-proliferating CD8+CD122+ Treg (Fig. 2C, right panel). Remarkably, only proliferating BrdU+CD8+CD122+ T cells revealed CCR6 expression, whereas non-proliferating BrdU−CD8+CD122+ were negative for CCR6, suggesting a propensity of peripheral T cells to get recruited to the inflamed target organ. Therefore, it seems to be very likely that the frequency and nature (e.g. the expression of B7-H1 molecules) of i.c. DC might be critical for the selective recruitment of CD8+ T cells with a regulatory phenotype into the target organ during CNS neuroinflammation.
Indeed, on day 14 post EAE, a large part of CD8+ T cells detected in the CNS of all experimental groups of animals were proliferating BrdU+ T cells (data not shown and Fig. 3A, experimental scheme). In particular, a significant accumulation of BrdU+CD8+ T cells in the CNS could be detected in B7-H1-/- DC-injected group compared with WT DC and PBS-injected groups (Fig. 3B). A higher amount of CD8+BrdU+ proliferating cells could also be found among CD122+ cells than their CD122− counterparts (Fig. 3C). Further, a strong accumulation of proliferating CD8+CD122+ Treg cells was found in B7-H1-/- DC-injected group compared with PBS and WT DC groups. These results suggest a targeted recruitment and expansion of CD8+ T cells with regulatory phenotype into the CNS, where these cells continue to proliferate during the course of EAE.
We next sought to test the capacity of sorted CNS- versus periphery-derived CD8+ T cells to suppress the proliferation of CD4+ 2D2 T cells in vitro. Although the majority of CNS CD8+ T cells expressed CD122 marker in all experimental groups (data not shown), only CD8+ T cells sorted from the CNS but not from the periphery of i.c. B7-H1-/- DC-injected animals displayed suppressive capabilities in an in vitro proliferation assay (CNS: Fig. 3D and E; periphery: data not shown). This suggests that i.c. B7-H1-/- DC injection results in a selective CD8+ Treg infiltration into the CNS, and that the majority of these infiltrated CD8+ T cells are responsive to MOG peptide, which enables their ability to suppress MOG-specific immune responses leading to EAE delay. The observation of increased frequency of proliferating CD4+ and CD8+ T cells in the brain in the absence of B7-H1 signal is more obvious when looking at the CD8+ T cells. Subsequently, either CD4+ T cells might have some additional proliferation blockers that are missing on the CD8+ T cells, or CD8+ T cells survive or get recruited more into the CNS.
In summary, we describe the recruitment of CD8+ Treg into the CNS during neuroinflammation in response to a single i.c. injection of DC deficient for the tolerogenic B7-H1 molecule. Our data in mice support a theory that DC can be instrumental in maintaining the balance between the generation of encephalitogenic and regulatory T cells. Interestingly, a positive upregulation of B7-H1 on DC during interferon-beta treatment of multiple sclerosis patients resulting in enhanced inhibitory properties of DC and contributing to immunoregulatory mechanisms during the disease has been described in humans . Upregulation of parenchymal B7-H1 exerts inhibitory function in the CNS and confines the expansion as well as persistence of encephalitogenic T cells in the CNS (e.g. [27-30]). The level of B7-H1 on immunogenic DC injected directly into the CNS, however, seems to be decisive for the specific recruitment of suppressive, FoxP3− CD8+ T cells into the target organ of the inflammation. This knowledge may be used in efforts to develop or to optimize cell-based therapies in humans.
Four- to six-wk-old female C57BL/6 mice were purchased from Harlan Winkelmann (Borchen, Germany). B7-H1-/- mice were generated by Chen (Baltimore, USA) . 2D2 Tg mice were a gift from Kuchroo (Harvard Medical School, USA) . Mice were kept under specific pathogen-free conditions in our animal facility according to German guidelines for animal care. All experiments were conducted according to animal experimental ethics committee guidelines and were approved by the local authorities (Regierung von Unterfranken; 55.2-2531.01-48/08).
For i.c. immunization, mice were anesthetized by i.p. injection of a ketaneze–Rompun mixture. DC (2.5 × 105) loaded for 4 h with MOG35–55 peptide (EVGWYRSPFSRVVHLYRNGK; synthesized and HPLC purified by R. Volkmer, Charite, Berlin, Germany) (10 μg/mL) and matured with LPS (10 μg/mL) in 20 μL of PBS or in sterile PBS alone, were injected into the right frontal lobe with an insulin syringe attached to a penetrating depth controller as previously described [31-33].
DC from WT or B7-H1-/- mice were prepared as described previously [34, 35]. The single-cell suspension of bone marrow cells was cultured in RPMI 1640 (BioWhittaker, Verviers, Belgium), supplemented with 10% FCS, 2 mM L-Glutamine (PAA Laboratories, Pasching, Germany), 50 μM 2-Mercaptoethanol, antibiotics (penicillin 100 U/mL/streptomycin 10 μg/mL) (Biochrom, Berlin, Germany) and 20 ng/mL mGM-CSF (Peprotech, Hamburg, Germany), and the medium was changed at days 3 and 6. Bone marrow-derived DC were harvested on day 8 and characterized as more than 85% pure CD11c+ cells.
EAE was induced by injecting the mice subcutaneously (into the flanks) with 200 μL of an emulsion containing 100 μg of MOG35–55 peptide supplemented with 250 μg of Mycobacterium tuberculosis H37Ra (Difco) in incomplete Freund’s adjuvant oil. Pertussis toxin (List Biological Laboratories, Campbell, CA, USA) (400 ng/mouse) was injected i.p. on days 0 and 2. Where indicated, 2.5 × 105 DC pre-loaded with MOG35–55 were i.c. injected 5 days before immunization. Clinical signs of EAE were recorded on daily basis and graded on a standard scale as previously described .
Mice were perfused through the left cardiac ventricle with cold PBS. Spleen, lymph nodes, and CNS mononuclear cells were isolated as described previously . The isolated cells were washed with cold HBSS and resuspended either in staining buffer (PBS containing 1% BSA and 0.1% NaN3) for a direct cell-surface staining, or in culture medium (RPMI supplemented with 10% FBS) for cell culture experiments. Brain lymphocytes were isolated with a Percoll gradient (50%/30%) and centrifugation as described previously .
DC isolated from WT DC animals were seeded at a density of 5 × 103 cells/well and loaded with MOG35–55 peptide (10 μg/mL). Purified and MACS-sorted (Miltenyi Biotec, Bergish Gladbach, Germany) 5 × 104 2D2 Tg TCR CD4+ T cells were CFSE-stained (2.5 μM, Molecular Probes, Eugene, OR, USA) as described previously  and cultured either alone or in combination with aforementioned DC with and without specific antigen, or in a combination with different ratios of CNS- or periphery-sorted CD8+ T cells in 96-well culture plates for 3 days in vitro. Cell proliferation was assessed based on CFSE dilution and analyzed by flow cytometry.
Single-cell suspensions of splenocytes or lymphocytes from cervical lymph nodes (periphery), or brains combined with spinal cord tissue (CNS) were stained with surface antibodies (anti-CD4-PerCp, anti-CD8-PE, anti-CD122-FITC, anti-CD11b-PerCp, anti-CD11c-APC, and anti-PIR-A/B, BD Pharmingen) in the presence of FcγRII/FcγRIII-specific antibody (clone 2.4G2) to block unspecific binding. Anti-CCR6-PE mAb was purchased from R&D Systems (Germany). FACS data were collected on a four-color FACSCalibur cytometer (BD Biosciences), and data were analyzed using FlowJo software (TreeStar) version 7.2.1.
Mice were treated i.p. with 1 mg BrdU (BD Pharmingen) in PBS 18 h before cell harvest. Following harvest, BrdU incorporation was detected using BrdU-APC Flow Kit (BD Pharmingen) as described in the manufacturer’s instructions.
Statistical evaluations of results were determined using two-tailed Student’s t test.
We thank T. Moritz and B. Reuter for excellent technical assistance. The authors would also like to thank Melissa G. Harris for careful review of this paper. This work was supported by grants from the BMBF 01gz0707 (to H.W.) and SFB581 (TP A8 to H.W.).
Conflict of interest: The authors declare no financial or commercial conflict of interest.