Autoreactive Tbet-positive CD4 T cells develop independent of classic Th1 cytokine signaling during experimental autoimmune encephalomyelitis

Abstract

Many autoimmune chronic inflammatory diseases, including multiple sclerosis, are associated with the presence of Th1 and Th17 effector CD4 T cells. Paradoxically, the principal Th1 cytokine IFN-γ does not appear necessary for disease, but the key Th1-associated transcription factor Tbet has been reported to be essential for disease development. This conundrum propelled us to investigate the regulation of this transcription factor during autoimmunity. Following the onset of experimental autoimmune encephalomyelitis, we observed a preferential upregulation of Tbet by CD4 T cells within the CNS, but not the secondary lymphoid organs. These Tbet-positive CD4 T cells were capable of producing the cytokine IFN-γ, and a proportion of these cells produced both IFN-γ and IL-17A. Interestingly, these Tbet-positive cells were present in high frequencies during disease in IFN-γ-deficient mice. Moreover, we found that CD4 T cells from IFN-γ-deficient/IFN-γ reporter mice upregulated the Thy1.1 reporter, indicating the presence of Th1 or Th1-like, Tbet-positive CD4 T cells even in the absence of the cardinal Th1 cytokine IFN-γ. These IFN-γ-deficient Th1-like cells not only maintain multiple Th1 properties but also exhibit increased expression of genes associated with the Th17 phenotype. We further examined the requirement of other Th1-associated molecules in controlling Tbet expression during experimental autoimmune encephalomyelitis and noted that STAT1, IL-12, and IFN-γ were dispensable for the induction of Tbet in vivo. Hence, this study highlights the complex regulation of Tbet and the potential unrecognized role for Th1 cells during autoimmunity.

Figure 1. Tbet is essential and highly expressed in infiltrating CD4 T cells in CNS during EAE

EAE was induced in WT C57BL/6 mice by MOG_35–55 peptide immunization. (a) Disease severity in WT mice was monitored daily by the criteria described in Materials and Methods. Data represent two independent experiments with 5 mice in each group (means ± SEM). (b) Lymphocytes were purified from brain (BR), spinal cord (SC), spleen (SPL), and inguinal lymph nodes (LN) of WT mice 20 days after EAE induction. Samples were analyzed for Tbet expression ex vivo by intracellular staining with anti-Tbet (black histogram) and mouse IgG1κ isotype control (grey histogram) mAbs. Plots were gated on CD4 positive cells. (c) Single cell suspensions from the various tissues were restimulated with PMA/ionomycin for 4 hrs, and Tbet expression was evaluated by FACS in IFNγ single positive (blue line), IL-17A single positive (green line), and IFNγ/IL-17A double positive cells (red line) and shown by overlapping histograms. Tbet isotype control is shown in grey line. (d) Cells collected from BR and SC were restimulated with PMA/ionomycin for 4 hours and the production of IFNγ, IL-17A, TNFα, IL-2, and GM-CSF was analyzed by flow cytometry among CD4+ Tbet+ population. Data is representative of two experiments with a total of 8–9 mice.

Figure 2. IFNγ signaling is not required for Tbet expression in pathogenic CD4 T cells in CNS

EAE was induced in WT, IFNγ-deficient, and IFNγR-deficient mice by MOG_35–55 peptide immunization. (a) Disease severity in WT and IFNγ-deficient mice after EAE induction. Data is cumulative of two experiments with 5–9 mice total (means ± SEM). (b, c) Tbet expression was examined directly ex vivo in WT and IFNγ-deficient lymphocytes. (b) Representative histograms gated on CD4 positive cells are shown and percentage of Tbet-positive CD4 T cells, as well as the MFI of the Tbet staining is noted. Tbet isotype control staining is shown by the grey histograms. (c) Graphs indicate the percentages of Tbet-positive CD4 T cells in CNS from the WT and IFNγ-deficient mice. (d) Disease severity in WT and IFNγR-deficient mice after EAE induction. Results are pooled; 8–10 mice in two independent experiments (means ± SEM). (e, f) Tbet expression was evaluated in WT and IFNγR-deficient lymphocytes at 19–21 days after EAE induction. (e) Representative histograms display the Tbet staining in gated CD4 T cells. The percentage of Tbet-positive cells is noted, as well as the MFI of the Tbet staining. Isotype control staining is represented by the grey histograms. (f) Graphs denote the percentage of Tbet-positive cells in WT and IFNγR-deficient mice. Data is combined from 2 independent experiments.

Figure 3. IFNγ signaling inhibits IL-17A production in CNS infiltrating CD4 T cells during EAE

(a, b) Lymphocytes isolated from WT and IFNγ-deficient mice with EAE were restimulated with PMA/ionomycin, and subsequently stained for IL-17A production. (a) Representative plots are gated on CD4 positive T cells and the values indicate the percent of cells staining IL-17A-positive. (b) Pooled data of the frequency of IL-17A-positive CD4 T cells from multiple experiments are shown (n = 12–14, *** p < 0.005). (c, d) IL-17A production was analyzed in WT and IFNγR-deficient CD4 T cells following restimulation with PMA/ionomycin. (c) Representative plots show gated CD4 positive T cells and the percent IL-17A-positive cells is noted. (d) Collective data for the percent of IL-17A-positive CD4 T cells from multiple experiments are shown (n = 10, *** p < 0.005).

Figure 4. Functional Th1 cells are present during EAE irrespective of IFNγ signaling

EAE was induced in WT and IFNγR-deficient mice and the CD4 T cells were analyzed for the ability to produce IFNγ after restimulation with PMA/ionomycin. (a) Representative plots show IFNγ staining of gated CD4 T cells. The frequency of IFNγ-producing CD4 T cells is noted. (b) Cumulative data of the percentage of IFNγ-positive CD4 T cells present in WT and IFNγR-deficient mice with EAE (n = 8–10).

Figure 5. The emergence of Th1 or “Th1-like” cells in during EAE

(a–c) EAE was induced in WT IFNγ BAC-In (IFNγ BI) and IFNγ-deficient IFNγ BAC-In (IFNγ−/− × IFNγ BI) transgenic mice and IFNγ reporter molecule (Thy1.1) expression was assessed 20–21 days later. (a) Representative contour plots are gated on CD4 T cells and the values indicate the percentage of Thy1.1-positive cells present in the gate. (b) Combined data from two separate experiments are shown (n = 7). (c) Cells from the BR and SC of WT and IFNγ-deficient IFNγ BI mice were stimulated for 4 hr with PMA/ionomycin and subsequently stained for CD4 and Thy1.1 in conjunction with IFNγ (left panel), Tbet (middle panel), or IL-17A (right panel). (d-f) Thy1.1-positive CD4 T cells were FACS-sorted from the CNS of WT and IFNγ-deficient IFNγ BI mice and real-time PCR was performed for genes related to the (d) Th1, (e) Th17, and (f) Th2/Treg lineages. Relative gene expression of IFNγ deficient compared to WT cells is shown as the ΔΔCt value, in which ΔΔCt = (ΔCt of the IFNγ-deficient cells – ΔCt of the WT cells). Data were combined from three independent experiments except the AHR data is from two experiments.

Figure 6. The development of Tbet-positive effector CD4 T cells during EAE is independent of STAT1, IL-12, and IFNγ signaling

(a, b) WT and STAT1-deficient mice in either 129S6 or B6 background were actively induced with EAE. (a) The frequencies of Tbet-positive CD4 T cells in B6 cohorts were evaluated. (b) Combined data are shown for both 129S6 and B6 genetic backgrounds. (c, d) EAE was induced in WT and IL-12p35 deficient mice. CD4 T cells were examined by FACS for Tbet expression 20 days after disease induction. (e, f) WT and IL-12p35-deficient mice were immunized with MOG_35–55 peptide and treated α-IFNγ mAb (100μg) every three days from day 0–9. Tbet expression was analyzed on day 13–21 after immunization. All histograms show gated CD4 T cells and isotype control staining is represented by the grey histograms. Experiments were performed 2–3 times, 5–15 mice total.