Ablation of Transcription Factor IRF4 Promotes Transplant Acceptance by Driving Allogenic CD4+ T Cell Dysfunction

Abstract

CD4⁺ T cells orchestrate immune responses and destruction of allogeneic organ transplants, but how this process is regulated on a transcriptional level remains unclear. Here, we demonstrated that interferon regulatory factor 4 (IRF4) was a key transcriptional determinant controlling T cell responses during transplantation. IRF4 deletion in mice resulted in progressive establishment of CD4⁺ T cell dysfunction and long-term allograft survival. Mechanistically, IRF4 repressed PD-1, Helios, and other molecules associated with T cell dysfunction. In the absence of IRF4, chromatin accessibility and binding of Helios at PD-1 cis-regulatory elements were increased, resulting in enhanced PD-1 expression and CD4⁺ T cell dysfunction. The dysfunctional state of Irf4-deficient T cells was initially reversible by PD-1 ligand blockade, but it progressively developed into an irreversible state. Hence, IRF4 controls a core regulatory circuit of CD4⁺ T cell dysfunction, and targeting IRF4 represents a potential therapeutic strategy for achieving transplant acceptance.

Keywords: T cell dysfunction; interferon regulatory factor 4; programmed cell death protein 1; transcriptional regulation; transplant acceptance.

Figure 1. IRF4 is overexpressed in graft-infiltrating T cells, and Irf4-deficient T cells do not mediate heart allograft rejection

(A and B) Flow cytometry analysis of IRF4 expression in CD4⁺ (left) and CD8⁺ (right) T cells from the spleens and allografts of WT B6 recipient mice at 7 days after Balb/c heart transplantation (HTx). CD4⁺ and CD8⁺ T cells from the spleens of naïve B6 mice were used as controls. Representative histograms (A) and IRF4 MFI values (B) were shown. Data in (B) are presented as mean ± SD (n = 4). **P<0.01; unpaired Student’s t-test. Data are representative of two independent experiments. (C and D) WT B6 and Irf4^fl/flCd4-Cre mice were transplanted with Balb/c hearts. (C) The percentage of allograft survival after transplantation (n=6). **P<0.01; Mann-Whitney test. (D) H&E stained sections of heart allografts harvested from WT B6 recipients at day 7 post-transplant, or from Irf4^fl/flCd4-Cre recipients at day 7 or day 100 post-transplant. Three mice per group were analyzed, with ten graft sections per mouse. Two representative images (top and bottom) per group are shown. See also Figure S1.

Figure 2. The dysfunction of Irf4-deficent T cells in transplantation is correlated with impaired cytokine production and graft infiltration

(A) Balb/c heart allograft survival in Irf4^fl/flCd4-Cre mice that were adoptively transferred with 2 or 20 million (M) indicated T cells. (B) Balb/c heart allograft survival in Irf4^fl/flCd4-Cre mice that were treated with rat IgG or an anti-CD25 (αCD25) mAb on indicated days. (C-H) Rag1^−/− mice were co-injected with 2 × 10⁷ (20M) CD45.1⁺ WT and 20M CD45.2⁺ (CD45.1⁻) Irf4^−/− T cells on day −1, and received Balb/c heart allografts on day 0. Splenocytes and graft-infiltrating cells were isolated on day 9 for flow cytometry analysis. (C) Schematic of the experimental design. (D) Representative contour plots showing % transferred CD45.1⁺ WT and CD45.1⁻ Irf4^−/− T cells (CD4⁺ or CD4⁻, gated on CD3⁺ cells) in spleens on day −1 (left) and day 9 (right). (E) The left plot displays % CD4⁺ (CD3⁺CD8⁻) and CD8⁺ (CD3⁺CD8⁺) T cells among CD45⁺ graft-infiltrating cells on day 9, and the right plots display % CD45.1⁺ WT and CD45.1⁻ Irf4^−/− cell populations among those CD4⁺ and CD8⁺ infiltrating T cells. (F) Expression of indicated markers on CD4⁺ (top) and CD8⁺ (bottom) populations of transferred CD45.1⁺ WT and CD45.1⁻ Irf4^−/− T cells in spleens on day 9. (G and H) Plots (left) and the bar graphs (right) depict the percentage expression of indicated molecules in co-transferred CD45.1⁺ WT and CD45.1⁻ Irf4^−/− CD4⁺ (G) or CD8⁺ (H) T cells in spleens on day 9. Data are mean ± SD. **P<0.01; unpaired student’s t-test. Data are representative of three independent experiments with three to four mice in each group (C-H). See also Figure S2.

Figure 3. IRF4 represses a set of molecules associated with CD4⁺ T cell dysfunction

(A) Flow cytometry analysis of cell surface molecules expressed on naïve WT CD4⁺ T cells (gray shades), or on activated WT (black lines) or Irf4^−/− (red lines) CD4⁺ T cells one day after stimulation with B6 APCs (mitomycin C-treated, T-cell–depleted B6 splenocytes) and soluble anti-CD3 mAb. (B-E) WT and Irf4^−/− CD4⁺ T cells were activated for 2 days. RNA was analyzed by microarray and quantitative real-time PCR, and Helios expression was analyzed by flow cytometry. (B) Heat map showing the normalized expression scores (relative to row mean) of selected genes from WT or Irf4^−/− CD4⁺ T cells. Two RNA samples of each group were obtained from two independent culture experiments of pooled T cells from n = 3 mice per sample. (C) GO categories enrichment analysis of 438 upregulated genes in Irf4^−/− CD4⁺ T cells in accordance with biological process. The horizontal axis shows −log10 of the P-value. (D) Relative changes of mRNA expression of the indicated genes in Irf4^−/− CD4⁺ T cells compared to WT CD4⁺ T cells determined by quantitative real-time PCR. Data are mean ± SD. (E) Flow cytometry analysis of Helios expression in WT and Irf4^−/− CD4⁺ T cells. Data are representative of three experiments with triplicate samples (A, D, E).

Figure 4. Upregulation of PD-1 in activated Irf4^−/− CD4⁺ T cells through increased chromatin accessibility and Helios binding at PD-1 cis-regulatory elements

(A) Histograms show PD-1 expression (shades and lines) and MFI (numbers) on freshly isolated naïve WT CD4⁺ T cells (gray), or on activated WT (black) or Irf4^−/− (red) CD4⁺ T cells at indicated days after stimulation with B6 APCs and soluble anti-CD3 mAb. Line graph (right) displays change in PD-1 MFI with time after activation. (B) PD-1 expression on co-cultured CD45.1⁺ WT and CD45.1⁻ Irf4^−/− CD4⁺ T cells 3 days after activation. (C) PD-1 expression on activated Irf4^−/− CD4⁺ T cells transduced with a retroviral vector expressing GFP alone (Ctrl) or with retrovirus expressing Irf4-GFP (IRF4). Numbers in contour plots (left) and the bar graph (right) indicate PD-1 MFI of gated GFP⁺ cells. (D) ChIP analysis of H3Ac, H4Ac, H3K4me3, and H3K9me3 at the PD-1 cis-regulatory elements (−3.7, CR-C, CR-B, and +17.1) in WT and Irf4^−/− CD4⁺ T cells 2 days after activation. (E) PD-1 and Helios expressions on WT and Irf4^−/− CD4⁺ T cells at 2 days after activation. (F) ChIP analysis of the enrichment of Helios at the PD-1 cis-regulatory elements in WT and Irf4^−/− CD4⁺ T cells at 2 days after activation. (G) PD-1 expression on activated WT CD4⁺ T cells transduced with a retroviral vector expressing GFP alone (Ctrl) or with retrovirus expressing Ikzf2-GFP (Helios). Numbers in contour plots (left) and the bar graph (right) indicate PD-1 MFI of gated GFP⁺ cells. (H) Helios (left three panels) and PD-1 (right two panels) expression by GFP⁺ Irf4^−/− CD4⁺ T cells that were transduced with a retroviral vector co-expressing GFP and shRNA sequences for Helios (sh-Helios) or containing GFP alone (sh-Ctrl). Numbers in plots and the bar graphs indicate Helios and PD-1 MFI of gated GFP⁺ cells. *P<0.05 and **P<0.01 (unpaired student’s t-test). Data are representative of three independent experiments. See also Figure S3.

Figure 5. Responsiveness to checkpoint blockade defines the dysfunctional states of Irf4-deficient T cells after transplantation

(A) Balb/c heart graft survival in Irf4^fl/flCd4-Cre mice that were adoptively transferred with 2 × 10⁷ WT or Irf4^−/− TEa cells on day −1. (B) CD45.1⁺ congenic mice were transferred with 5 × 10⁶ (5M) CellTrace Violet (CTV)-labeled CD45.2⁺ WT or Irf4^−/− TEa cells on day −1, received Balb/c heart transplants (HTx) or left un-transplanted (no Tx) on day 0, followed by analysis of splenocytes on day 6. Contour plots show co-expression of CD45.2 with CTV or PD-1, gated on CD4⁺ cells; or show co-expression of Foxp3 with intracellular CTLA-4, gated on CD4⁺CD45.2⁺ TCRα2⁺ TEa cells. Histograms display CTV, PD-1, and intracellular CTLA-4 expressions, gated on TEa cells. Data are representative of three experiments. (C) Balb/c heart graft survival in Irf4^fl/flCd4-Cre mice that were treated with rat IgG, anti-PD-L1, anti-CTLA-4, or anti-PD-L1 plus anti-CTLA-4 mAbs on days 0, 3, and 5 post-transplant. (D and E) Dilution of CTV (D) indicates proliferation of CD45.1⁺ CD4⁺ T cells in the absence or presence of Treg cells from indicated groups. Quantification of the results (E) is presented as division index. (F and G) Irf4^fl/flCd4-Cre mice were transplanted with Balb/c hearts and treated with anti-PD-L1 plus anti-CTLA-4 mAbs, together with 400 μg anti-CD4 or anti-CD8 depleting mAb on days 0, 3, and 5. Counter plots (F) show efficacy of T cell depletion on day 4. The graph (G) shows heart graft survival. (H) Balb/c heart graft survival in Irf4^fl/flCd4-Cre mice that were treated with anti-PD-L1 plus anti-CTLA-4 mAbs starting from day 0 (on days 0, 3, and 5), day 7 (on days 7, 10, and 12), or day 30 (on days 30, 33, and 35) post-transplant. (I and J) Histogram (I) shows % GFP⁺ cells in CD4⁺CD69⁺ Irf4^−/− cells following 3-day stimulation with Balb/c splenic DCs and 1-day IRF4-GFP viral transduction. The graph (J) shows Balb/c heart graft survival in Irf4^fl/flCd4-Cre mice that were transferred on day 1 with 1 × 10⁶ GFP⁺ Irf4^−/− CD4 T cells (transduced with IRF4-GFP or GFP-Ctrl). P<0.05 and **P<0.01; Mann-Whitney test (A, C, G, H and J). See also Figure S4.

Figure 6. Checkpoint blockade reverses the initial dysfunction of Irf4-deficient CD4⁺ T cells by restoring their ability to undergo proliferation and secrete IFN-γ

(A) As shown in the schematic, CD45.1⁺ congenic mice were transferred with 5 × 10⁶ (5M) CD45.2⁺ WT or Irf4^−/− TEa cells on day −1, received Balb/c heart transplants (HTx) or left un-transplanted (no Tx) on day 0, treated with rat IgG (Irf4^−/− TEa (IgG)) or anti-PD-L1 plus anti-CTLA-4 mAbs (Irf4^−/− TEa (P+C)) on days 0, 3, and 5, followed by flow cytometry analysis of TEa cells in spleens on day 6. (B) % transferred TEa cells among CD4⁺ splenocytes (top row) and expression of indicated molecules by TEa splenocytes (other rows) on day 6 post-transplant. (C) Bar graphs show % TEa cells among CD4⁺ splenocytes and numbers of TEa splenocytes (top row), and % Ki67⁺, CD98 MFI, GLUT1 MFI, CD71 MFI, % IFN-γ⁺, and % Foxp3⁺ of TEa cells (other rows). *P<0.05 (unpaired student’s t-test). Data are mean ± SD (C) and are representative of three experiments (B and C). See also Figure S5.

Figure 7. Trametinib inhibits IRF4 expression in T cells, prevents EAE development, and prolongs allograft survival

(A) IRF4 expression (left) and MFI (right) in freshly isolated naïve B6 CD4⁺ T cells, or in CD4⁺ T cells that were activated for 2 days in the presence of DMSO vehicle or varying concentrations of trametinib. (B) Dilution of CTV (left) indicates proliferation of CD4⁺ T cells that were activated for 3 days in the presence of DMSO or 100 nM trametinib. Quantification of the results (right) is presented as division index. (C) Contour plots (left) and bar graphs (right) display frequencies of IFN-γ, IL-17, and Foxp3 expressing cells in CD4⁺ T cells that were cultured under Th1, Th17, and inducible Treg (iTreg) polarizing conditions for 3 days in the presence of DMSO or 100 nM trametinib. (D and E) 10-week-old B6 female mice subjected to MOG35-55–induced EAE were treated with corn oil or 3 mg/kg Trametinib every other day from day 0 to day 12 post immunization. (D) Clinical scores of mice in each group. (E) Contour plots (6 panels on the top) display the frequency of CD4⁺TCRβ⁺ T cells among CD45⁺ cells in the brain tissues at 18–20 days post induction of EAE, and expression of GM-CSF, IL-17A, IFN-γ by those CD4⁺ T cells. Bar graphs (4 panels on the bottom) indicate the number of CD4⁺ T cells in the brain tissues, and frequencies of IFN-γ⁺, IL-17A⁺, GM-CSF⁺ cells among them. (F) Percentage Balb/c heart allograft survival in B6 recipients that were treated with Trametinib or corn oil every other day from day 0 to day 12 post-transplant. (G) CD45.1⁺ mice were transferred with 5 × 10⁶ CD45.2⁺ WT TEa cells on day −1, received Balb/c heart transplants on day 0 and treated with corn oil or 3 mg/kg Trametinib on days 0, 2, 4, and 6, followed by analysis of splenocytes on day 7. Dot plots show co-expression of CD45.2 with PD-1 (left) or Helios (right), gated on CD4⁺ cells. Bar graphs display PD-1 MFI and % Helios⁺ cells of transferred CD45.2⁺ TEa cells. Data are mean ± SD (A-C, E, and G) and are representative of two to three independent experiments. **P<0.01; unpaired student’s t-test (A-C, E, and G); Mann-Whitney test (D and F). See also Figures S6 and S7.