TLR9 Sensing of Self-DNA Controls Cell-Mediated Immunity to Listeria Infection via Rapid Conversion of Conventional CD4+ T Cells to Treg

Abstract

CD4⁺ T lymphocytes are crucial for controlling a range of innate and adaptive immune effectors. For CD8⁺ cytotoxic T lymphocyte (CTL) responses, CD4⁺ T cells can function as helpers (T_H) to amplify magnitude and functionality or as regulatory cells (T_reg) capable of profound inhibition. It is unclear what determines differentiation to these phenotypes and whether pathogens provoke alternate programs. We find that, depending on the size of initial dose, Listeria infection drives CD4⁺ T cells to act as T_H or induces rapid polyclonal conversion to immunosuppressive T_reg. Conversion to T_reg depends on the TLR9 and IL-12 pathways elicited by CD8α⁺ dendritic cell (DC) sensing of danger-associated neutrophil self-DNA. These findings resolve long-standing questions regarding the conditional requirement for T_H amongst pathogens and reveal a remarkable degree of plasticity in the function of CD4⁺ T cells, which can be quickly converted to T_regin vivo by infection-mediated immune modulation.

Figure 1.. Induction of T Cell Responses by LD and HD Listeria Infection

(A–D) Groups of C57BL/6 mice systemically infected with LD or HD Lm DactA-Ova. (A) Kinetics of total CD8⁺ T cells, Ova_257–264/H-2K^b-tetramer-binding (H-2K^b-Ova⁺) CD8⁺ T cells, total CD4⁺ T cells, and CD25⁺FoxP3⁺CD4⁺ T cells (T_reg) in the spleens of infected mice (n = 4–8 per group). (B) Frequency of CD25⁺FoxP3⁺CD4⁺ T_reg within the total CD4⁺ T cell fraction (n = 7 per group). (C) Change in FoxP3⁺:FoxP3⁻ CD4⁺ T cell ratio after HD Lm ΔactA-Ova infection relative to the naive cohort (D FoxP3⁺:FoxP3⁻) (n = 4 per group). (D) Absolute number of Ova-specific CD8⁺ T cells and Ova-specific IFN-γ⁺ effectors at day 7 post-infection ± CD4⁺ T cell depletion (n = 6 per group). (E and F) C57BL/6 and Foxp3^DTR-EGFP mice infected with LD or HD Lm ΔactA-Ova with DT delivery to all animals before and throughout infection. (E) Absolute number of Ova-specific CD8⁺ T cells and Ova-specific IFN-g⁺ effectors at day 7 ± CD4⁺ T cell depletion (n = 6 per group). (F) H-2K^b-Ova relative affinity of splenic CD8⁺ T cells calculated as the geometric mean fluorescence intensity (geoMFI) of Ova_257–264/H-2K^b-tetramer to anti- TCRβ FACS stain ratio compared between strain and Lm dosage (n = 10 per group). (G) Survival of C57BL/6 versus Foxp3^DTR-EGFP mice infected with WT Lm Ova, with DT injections spanning the immune response (n = 10 per group).

Figure 2.. FoxP3⁺ T_reg Suppress Primary CD8⁺ T Cell Responses with a Different Kinetic Compared to T_H

(A–C) C57BL/6 mice infected with LD or HD Lm ΔactA-Ova. (A and B) (A) Frequency with CD44^hi inset (red box) and (B) absolute number of Ova-specific CD8⁺ T cells 7 days post-infection among total CD44^lo-hiCD8⁺ T cells in the spleens of mice ± CD4⁺ T cell depletion either prior to infection or at 3, 4, or 5 days post-infection (n = 5–6 per group). (C) The absolute number of splenic Ova-specific CD8⁺ T cells at day 7 in mice that had received CD40L-blocking antibody 1 day before or 3 days after infection (n = 6 per group). (D) C57BL/6 and Foxp3^DTR-EGFP mice i.v. infected with LD or HD Lm ΔactA-Ova with DT delivery to all animals before and throughout infection. Absolute number of splenic Ova-specific CD8⁺ T cells at day 7 in animals that had received CD40L-blocking antibody 1 day before infection (n = 6 per group).

Figure 3.. Inflammation-Driven Polyclonal Conversion of Conventional CD4⁺ T Cells to T_reg

(A) TCR engagement and surface CD69 upregulation by conventional CD4⁺ T cells (FoxP3⁻), T_reg (CD25⁺FoxP3⁺), and CD8⁺ T cell populations in naive (blue) and at 24 h following LD (red) and HD (green) Lm ΔactA-Ova infection of Nr4a1^GFP mice (n = 4 per group). (B and C) Conventional CD4⁺EGFP⁻ T cells were isolated from CD45.1⁺ Foxp3^EGFP mice and transferred into C57BL/6 mice followed by a 24-h rest period. The percentage of splenic CD45.1⁺FoxP3-EGFP⁻ T cells that expressed EGFP under control of the Foxp3 promoter (denoted as FoxP3-EGFP) after assessment of (B) LD and HD Lm ΔactA-Ova day 1 (n = 3–4 per group) and 7 (n = 3 per group), as well as (C) titrated WT Lm Ova day 1 (n = 3–4 per group) infections.

Figure 4.. Converted T_reg Mediate Suppression of CD8⁺ T Cell Responses to Listeria In Vitro and In Vivo

(A) In vitro CD3/CD28-driven CD45.2⁺CD8⁺ T_resp proliferation represented as percent suppression and replication index at day 3 after co-culture of naive CellTrace Violet (CTV)-labeled CD45.2⁺CD8⁺ T_resp with graded populations of converted CD45.1⁺CD4⁺FoxP3-EGFP⁺ and non-converted CD45.1⁺CD4⁺FoxP3- EGFP⁻ T cells re-isolated from C57BL/6 recipient mice that underwent 24 h HD Lm ΔactA-Ova infection (n = 4 per group). (B) Absolute number of splenic Ova-specific CD8⁺ T cells at day 7 post-infection of C57BL/6 and Foxp3^DTR-EGFP mice infected with HD Lm ΔactA-Ova. FoxP3⁺ T_reg depletion was achieved via limiting DT injection to a pre-infection phase (days −3 to −1) or continuing DT administration during the immune response (days −3 to 5) (n = 4 per group). (C and D) Naive CD4⁺EGFP⁻ T cells were isolated from CD45.1⁺ Foxp3^EGFP mice and transferred into C57BL/6 and Foxp3^DTR-EGFP mice at day −1 that received a full DT regimen relative to a day 0 HD Lm ΔactA-Ova infection. The frequency and absolute number of splenic (C) Ova-specific CD8⁺ T cells and (D) Ova-specific IFN-g⁺ effectors together with change in IFN-γ production per cell (Δ 3 10³ IFN-γ geoMFI) at day 7 post-infection (n = 4 per group).

Figure 5.. Role of TLR9 and IL-12 (p70) in T_reg Conversion

(A and B) Groups of C57BL/6 and TLR-deficient mice infected with LD or HD Lm ΔactA-Ova. (A) Absolute number of Ova-specific CD8⁺ T cells at day 7 post-infection with HD Lm ΔactA-Ova compared among splenic isolates of Tlr2^−/−, Tlr4^−/−, Tlr7^−/−, and Tlr9^M7Btlr mice (n = 6–10 per group). (B) Absolute number of Ova-specific CD8⁺ T cells at day 7 in the spleens of Tlr9^M7Btlr mice ± CD4⁺ T cell depletion (n = 6–8 per group). (C and D) Groups of C57BL/6 and Tlr9^M7Btlr mice systemically infected with HD Lm ΔactA-Ova. (C) Fold change in serum cytokines and chemokines from 0–6 h post-infection. (D) Selected serum cytokine/chemokine concentration at 6 h post-infection displaying significance in (C) (n = 6 per group). (E) Foxp3^EGFP mice infected with HD Lm ΔactA-Ova for 24 h. Ingenuity Pathway Analysis (IPA) upstream regulator prediction of differential gene expression profiles comparing total FoxP3-EGFP⁺ T_reg sorted from naive versus infected mice (top 50 significant hits; p = gene network overlap) (n = 4 per group). (F) Kinetic display of IFN-a2/4, IFN-b, IFN-g, and IL-12 (p70) serum concentration in C57BL/6 and Tlr9^M7Btlr mice after exposure to HD Lm DactA-Ova (n = 6 per group).

Figure 6.. TLR9/NF-κB-Mediated Induction of IL-12 in CD8α⁺ DCs

(A and B) Tlr2, Tlr4, Tlr7, and Tlr9 mRNA normalized to Hprt1 mRNA in (A) whole spleen and (B) sorted DC and T cell populations in C57BL/6 mice infected with LD or HD Lm ΔactA-Ova (n = 3 per group). (C and D) p65-GFP knockin mice challenged for 24 h with HD Lm ΔactA-Ova. (C) Frequency of p65-GFP and TLR9 expression in B cells, CD8a⁻ DC/Mac, CD8a⁺ DCs, plasmacytoid DCs (pDCs), CD103⁺ DCs, red pulp macrophages (RP Macs), and marginal zone Macs (MZ Macs) (n = 3 per group). (D) Representative ImageStream images displaying TLR9^hi APC subcellular localization of p65-GFP (green) and DAPI (red) and quantitative p65-GFP nuclear translocation (n = 3 per group). (E) C57BL/6 and Tlr9^M7Btlr mice exposed to HD Lm ΔactA-Ova for 24 h. Frequency and absolute number of IL-12 (p40)⁺ and IFN-β⁺ APC subsets after PMA/ionomycin stimulation (n = 4 per group). (F) C57BL/6 and Batf3^−/− mice infected with HD Lm ΔactA-Ova for 24 h. Change in FoxP3⁺:FoxP3⁻ CD4⁺ T cell ratio at day 1 post-infection relative to the naive state (D FoxP3⁺:FoxP3⁻) (n = 7 per group).

Figure 7.. Release of Self-DNA Drives IL-12 (p70) Production, FoxP3 Induction, and T_reg Conversion from Conventional CD4⁺ T Cells

(A) C57BL/6 mice administered 50 μg CpG-A, CpG-B, and CpG control ODNs at days −1 and 0. Frequency of splenic CD25⁺FoxP3⁺CD4⁺ T_reg and change in FoxP3⁺:FoxP3⁻ CD4⁺ T cell ratio relative to CpG controls (D FoxP3⁺:FoxP3⁻) at day 1 (n = 4 per group). (B) C57BL/6 and Foxp3^DTR-EGFP mice administered CpG ODNs as in (A) and infected with LD Lm ΔactA-Ova at day 0 and co-delivered a complete DT regimen. Absolute number of splenic Ova-specific CD8⁺ T cells at day 7 post-infection (n = 6 per group). (C) Splenic Ova-specific CD8⁺ T cell response observed in CpG-treated or -infected C57BL/6 as in (B), with blockade of IL-12 (p70) (n = 4 per group). (D) Quantification of mtDNA (12 s and Trlev), nDNA (Neb and Becn1), and Lm DNA (Ova, mpl, and plcB) in cell- and RNA-free serum samples after HD Lm ΔactA- Ova infection of C57BL/6 mice (n = 6 per group). (E) Serum ALT concentration within LD versus HD Lm ΔactA-Ova-infected C57BL/6 mice (n = 6 per group). (F) Effect of neutrophil depletion before HD Lm ΔactA-Ova infection of C57BL/6 mice on mtDNA (12 s and Trlev) in cell- and RNA-free serum samples isolated at 4 h post-infection (n = 6 per group). (G) Absolute number of splenic Ova-specific CD8⁺ T cells at day 7 post-infection in C57BL/6 and Foxp3^DTR-EGFP mice administered a complete DT regimen, systemic HD Lm ΔactA-Ova, and selective DNase I delivery (n = 4 per group). (H) Survival of untreated versus DNase I-treated C57BL/6 mice infected with WT Lm Ova (n = 10–16 per group). (I and J) In vitro co-expression of FoxP3 and T-bet in conventional CD4⁺EGFP⁻ T cells isolated from Foxp3^EGFP or Il12rb2^−/−Foxp3^EGFP mice cultured in the presence or absence of CD3/CD28-mediated activation, rmTGF-b1, and rmIL-12 (p70) (n = 4 per group).