Eomesodermin-expressing T-helper cells are essential for chronic neuroinflammation

Abstract

Development of acute experimental autoimmune encephalomyelitis (EAE) depends on Th17 cells expressing the nuclear factor NR4A2. However, in mice lacking NR4A2 in T cells, a late-onset disease is still inducible, despite a great reduction in acute inflammation. We here reveal that development of this late onset disease depends on cytotoxic T-cell-like CD4(+) T cells expressing the T-box transcription factor Eomesodermin (Eomes). T-cell-specific deletion of the Eomes gene remarkably ameliorates the late-onset EAE. Strikingly, similar Eomes(+) CD4(+) T cells are increased in the peripheral blood and cerebrospinal fluid from patients in a progressive state of multiple sclerosis. Collective data indicate an involvement of granzyme B and protease-activated receptor-1 in the neuroinflammation mediated by Eomes(+) CD4(+) T cells.

Figure 1. Mice lacking NR4A2 in CD4⁺ T cells are protected from early/acute EAE but develop late/chronic EAE signs.

(a) Clinical EAE score. Disease scoring for control NR4A2^fl/fl (Control, black squares) and Cre-CD4/NR4A2^fl/fl (NR4A2 cKO, grey circles). The mice with C57BL/6 background were immunized with MOG_35–55 peptide emulsified in complete Freund’s adjuvant. Error bars represent s.e.m. (b) Histopathology of EAE. The spinal cords were fixed in formal-saline, paraffin embedded and sectioned before microphotography. Panels show adjacent sections from a representative mouse stained with haematoxylin and eosin (top row) or Luxol fast blue (bottom row) at × 4 (left) or × 10 (right) magnification; top panels are from a control mouse and bottom panels are from a NR4A2 cKO mouse; left panels are from day 18 post EAE induction, right panels are from day 32 post EAE induction from 1 of 2 independent experiments. Scale bars, 200 μm. (c) IL-17 and IFN-γ production by CNS CD4⁺ T cells as measured by intracellular cytokine staining following PMA/ionomycin stimulation in the presence of GolgiPlug for 5 h. The cells were isolated by enzymatic digestion of pooled CNS tissues during either early (Day 17) or late phase (Day 28) of EAE. Data are representative of at least three independent experiments, each using pools of three to five mice for each genotype and time point.

Figure 2. Late/chronic EAE is associated with Eomes expression by pathogenic CNS-infiltrating CD4⁺ T cells.

(a) Differential expression of immune-associated genes. CD4⁺ T cells were isolated and sorted from control and NR4A2 cKO CNS from mice with early or late EAE, and analysed by expression microarray. Gene expression was confirmed by quantitative reverse transcriptase–PCR (qRT–PCR) and relative expressions of selected genes are summarized by heat map. (b) Differential expression of eomes in the CNS CD4⁺ T cells determined by qRT-PCR. (c,d) Intracellular flow cytometry for Eomes expression of freshly isolated CNS T cells. Unstimulated cells were stained intracellularly. (e–g) Flow cytometric analyses for co-expression of IL-17, IFN-γ, CD27 and CD11a, with Eomes in CD4⁺ CNS-infiltrating T cells (Day 27). The cells were stained after stimulation with PMA/ionomycin in the presence of GolgiPlug for 5 h. FACS staining plots are representative of at least three independent EAE experiments.

Figure 3. Role of Eomes expressed by CD4⁺ T cells in chronic CNS inflammation.

(a) siRNA treatment for late/chronic EAE. eomes-specific (unfilled symbols) or control scrambled (solid symbols) siRNA was stabilized in an atellocollagen substrate and injected i.v. into control (red) or NR4A2 cKO mice (blue) on day 15 post EAE induction, to generate systemic gene knockdown. Plots show EAE scores for mice. Error bars represent s.e.m. for groups of five mice. (b) Induction of EAE in NR4A2/Eomes DKO mice compared with NR4A2 cKO mice. Plot shows clinical EAE scores for control (red symbols), NR4A2 cKO (blue symbols) and Eomes/NR4A2 dKO (magenta symbols) mice immunized with MOG_35–55. Error bars represent s.e.m. for groups of five mice. (c) Exacerbation of EAE after transfer of CNS T cells. CD45⁺TCR⁺CD4⁺CD11b⁻F4/80⁻ FACS-sorted CNS T cells were isolated from control, NR4A2 cKO or NR4A2/Eomes DKO mice with late-stage EAE on day 28 post EAE induction. T cells were restimulated polyclonally with anti-CD3 monoclonal antibody for 48 h before adoptive transfer into NR4A2 cKO recipient mice that had been immunized with MOG_35–55 7 days before. Plots show clinical scores (left) and cumulative disease scores (right), and are representative of two independent experiments. Error bars represent s.e.m. for clinical scores for groups of four or five mice and light lines show the 95% confidence interval for cumulative disease scores.

Figure 4. Eomes⁺CD4⁺ T cells mediate cytotoxicity.

Pooled CNS T cells were isolated from NR4A2 cKO mice with late-stage EAE. Purified CD45⁺TcRβ⁺CD4⁺ cells and CD45⁺TcRβ⁺CD4⁻ cells were prepared by FACS sorting. These T cells were restimulated with or without anti-CD3 antibodies in the presence of CNS-infiltrating myeloid cells and fluorescently conjugated anti-CD107a antibodies. After 4 h stimulation, cells were then stained intracellularly for expression of Eomes and granzyme B. Plots show granzyme B expression (a) and CD107a surface trapping (b) for unstimulated (left) and restimulated (right) CNS CD4⁺ T cells (top) and CNS CD4⁻ T cells (bottom). Data are representative of three independent experiments.

Figure 5. Increased proportions of Eomes⁺CD4⁺ T cells in patients with SPMS.

(a) Eomes expression by peripheral blood T cells from HCs (plotted black), patients with RRMS (plotted blue) and patients with SPMS (plotted red) as measured by flow cytometry. (b) Proportions of Eomes⁺ cells (%) among CD3⁺CD4⁺ PBMCs. Horizontal lines shows mean % for each patient group with s.e.m. error bars. Sample group information is detailed in Supplementary Table 1. ****P<0.0001 and ***P<0.001 as measured by a two-tailed T-test.

Figure 6. Accumulation of Eomes⁺CD4⁺ T cells in the CSF.

a) Eomes expression by CD3⁺ T cells in peripheral blood or CSF samples from patients with SPMS. (b) Proportions (%) of Eomes⁺ cells among CD3⁺CD4⁺ from PBMCs and CSF. Results from each patient are connected with lines **P<0.01 with a Wilcoxon matched-pairs signed rank test.

Figure 7. Eomes⁺CD4⁺ T cells from SPMS patients have characteristics of cytotoxic cells.

(a) PBMCs from SPMS patients were cultured with or without anti-CD3 and anti-CD28 antibodies in the presence of fluorescently conjugated anti-CD107a antibodies and brefaldin A. After 5 h stimulation, cells were then stained intracellularly for expression of Eomes. Plots show CD107a surface trapping for unstimulated (left) and restimulated (right) for CD3⁺CD4⁺ T cells and granzyme B, and perforin 1 staining shown in b. (c) PBMCs from SPMS patients were stimulated with PMA/ionomycin in the presence of GolgiPlug. After 5 h stimulation, cells were stained intracellularly with antibodies against granzyme B, IFN-γ, IL-17 and IL-10. Plots show CD3⁺CD4⁺ PBMCs from a representative SPMS patient from four SPMS patients tested.

Figure 8. Effective treatment for late/chronic neuroinflammation.

(a) Effects of granzyme B-specific siRNA for late/chronic EAE. NR4A2 cKO mice were immunized with MOG_35–55 and five of the ten mice were treated with granzyme B-specific siRNA (anti-gmzb siRNA) stabilized in atellocollagen on day 17 post EAE induction (NR4A2 cKO+anti-grzb siRNA). As control, WT mice were immunized at the same day (control) (n=5 for each group). In the left panel, clinical EAE scores are shown with error bars representing s.e.m. *P<0.05 for treated versus untreated NR4A2 cKO mice as tested by a Fisher’s least significant difference (LSD) test. In the right panel, solid lines represent cumulative disease scores, with dashed lines showing 95% confidence intervals. ***P<0.001 tested by linear regression analysis. Data are representative of two independent experiments. (b) Effects of PAR antagonist FR171113 for late/chronic EAE. Groups of wild-type (WT) mice (control) or NR4A2 cKO mice were immunized with MOG_35–55. One group of each genotype was treated daily from day 17 post EAE induction by intraperitoneal injection of PAR antagonist FR171113 or carboxymethyl cellulose (CMC) as vehicle. Left panel shows clinical EAE scores with error bars representing s.e.m. *P<0.05 for treated NR4A2 cKO versus untreated NR4A2 cKO mice as tested by a Fisher’s LSD test. Right panel shows cumulative disease, with dashed lines showing 95% confidence intervals. ***P<0.001 tested by linear regression analysis. n⩾5 and data are representative of three independent experiments.