Inflammatory Flt3l is essential to mobilize dendritic cells and for T cell responses during Plasmodium infection

Abstract

Innate sensing mechanisms trigger a variety of humoral and cellular events that are essential to adaptive immune responses. Here we describe an innate sensing pathway triggered by Plasmodium infection that regulates dendritic cell homeostasis and adaptive immunity through Flt3 ligand (Flt3l) release. Plasmodium-induced Flt3l release in mice requires Toll-like receptor (TLR) activation and type I interferon (IFN) production. We found that type I IFN supports the upregulation of xanthine dehydrogenase, which metabolizes the xanthine accumulating in infected erythrocytes to uric acid. Uric acid crystals trigger mast cells to release soluble Flt3l from a pre-synthesized membrane-associated precursor. During infection, Flt3l preferentially stimulates expansion of the CD8-α(+) dendritic cell subset or its BDCA3(+) human dendritic cell equivalent and has a substantial impact on the magnitude of T cell activation, mostly in the CD8(+) compartment. Our findings highlight a new mechanism that regulates dendritic cell homeostasis and T cell responses to infection.

Figure 1. Systemic Flt3L increase, Flt3-dependent BM activation and DC expansion during P. chabaudi infection

a : Curve shows serum Flt3L (pg.ml⁻¹) in wild type mice on d0-d6 after P. chabaudi infection. b: Changes in spleen leukocyte (black) and DC numbers (white:CD8α⁻, grey CD8α⁺) during the first 8 days of P. chabaudi infection in CD57BL/6 mice. See supplementary figure for DC gating. Absolute numbers are normalized to naïve mice. (3–7 mice analyzed independently for each time point). c, d : Bar graphs show absolute numbers of CD8α⁺ (c), CD8α⁻ (d) DCs in the spleens of wild type (WT) or Flt3^−/− mice normalized to WT controls before and 7 days after infection with P. chabaudi. Each dot represents an individual mouse. Mean, SEMs, and t-tests are shown. e, f : P. chabaudi-induced alterations in early, multipotent progenitors in WT (e, f) and Flt3^−/− mice (f). e : Bar graph shows the percentage of Lin⁻(CD3, CD4, CD8, CD19, B220, CD11b, CD11c, NK1.1, Ter119) CD117⁺/Sca1⁺ HSC-containing fraction in the BM during the first 5 days of P. chabaudi infection (mean+/− SEM, t-tests are shown) of a representative experiment with 3 mice analyzed independently at each time point. f : Bar graph indicate the percentage (mean+/− SEM of 4 mice analyzed independently for each group) of Lin⁻CD117⁺Sca1⁺ CD150⁺ and Lin⁻CD117⁺Sca1⁺ CD150⁻ in the BM of WT or Flt3^−/− mice at d6 after P. chabaudi infection. g, h : P. chabaudi-induced alterations in monocyte/DC committed progenitors in WT (g, h), Flt3^−/− and Flt3L^−/− mice (h). Bar graphs show the percentage of Lin⁻CD117⁺Sca1⁻CD115⁺ (MDP and CDP) in the BM (mean+/− SEM, t-test are shown) of a representative experiment with 3 mice analyzed independently at each time point.

Figure 2. Role of Flt3L responsive CD8α⁺/CD103⁺ langerin positive DCs in T cell activation during P. chabaudi infection

Control (DT-treated WT) or langerin-DTR transgenic mice were infected with P. chabaudi. Depletion of CD8a+ spleen DCs during P. chabaudi infection using Langerin-DTR mice was measured at d7 (a). DT had no effect on T cell activation (Supplementary Fig.3e). T cell activation was assessed at d7 in the spleen by measuring the percentage of CD62L^low/CD44^high cells inside TCRβ⁺CD8⁺ (b, upper panel) or TCRβ⁺/CD4⁺ (d, lower panel). Quantitation is shown in (c,d,f,g). Relative (c, f) and absolute (d, g) accumulation of activated TCRβ⁺CD4⁺(c, d) and TCRβ⁺CD8⁺(f, g) cells in the spleen are shown. Each dot represents an individual mouse, mean, SEM and some t-tests are shown.

Figure 3. MCs release inflammatory Flt3L release during Plasmodium infection

a : Production of soluble Flt3L by radioresistant cells during P. chabaudi infection. Serum Flt3L levels in BM chimeras (Flt3L^+/+ or Flt3L^−/− donor BM and recipient irradiated Flt3L^+/+ or Flt3L^−/− host), naïve (left) or at d4 after P. chabaudi infection (right). b, c : MC express high levels of surface Flt3L that is lost during P. chabaudi infection. b : Histogram shows surface levels of membrane Flt3L in FcεRI⁺CD117⁺ peritoneal MCs from naïve mice before (black) or after P. chabaudi infection (red, d2) or Flt3L^−/− mice (green). c : Quantification of membrane Flt3L fluorescence in individual mice before (black) or after P. chabaudi infection (red, d2 and d3). Absolute values are normalized to 1 in Flt3L^−/− MCs used as a negative control. Each dot is an individual mouse. d : Deficient Flt3L serum increased levels in Kit^W-sh MC-deficient mice infected with P. chabaudi (d2,d4,d6). Each dot represents an individual mouse. e–g: Deficient CD8a⁺ DC expansion in Kit^W-sh MC-deficient mice infected with P. chabaudi. e : Lin ⁻ (CD3,CD19,Ter119,Nk1.1, Gr1) CD11c^highMHCII^high DC compartment in spleen of WT or Kit^W-sh mice, before or after P. chabaudi infection (d8) with or without Flt3L injection. f, g : Absolute numbers per spleen of CD8α⁺ (f) or CD8α⁻ (g) DCs, values are normalized to WT naïve. Each dot is an individual mouse. h–j: Deficient CD8⁺ T cell activation in Kit^W-sh MC-deficient mice infected with P. chabaudi. T cell activation was assessed at d8 in spleen of WT or Kit^W-sh mice during P. chabaudi infection (d8, with or without Flt3L injection) by measuring the percentage of CD62L^low/CD44^high cells inside TCRβ⁺CD8⁺ (h, upper panel) or TCRβ⁺/CD4⁺ (h, lower panel). Absolute (i, j) accumulation of activated TCRβ⁺CD4⁺(i) and TCRβ⁺CD8⁺(j) cells in the spleen are shown. Each dot represents an individual mouse, mean, SEM and some t-tests are shown.

Figure 4. Role of uric acid sensing by MCs leads in Flt3L release, DC expansion and T cell responses during Plasmodium infection

a–c. Uric acid triggers MCs to release Flt3L. PBS or preformed uric acid crystals in PBS (200ug or 2000ug) were injected i.p. into WT mice. Mice were sacrificed 2 hours post injection and CD117⁺FcγRI⁺ peritoneal MCs were analyzed for surface Flt3L levels (a). Histogram shows Flt3L levels with Flt3L^−/− MCs as a negative control. (b), as is (a) but bar graph summarizes results for 3 mice for each condition. (c) Serum Flt3L levels from the same mice is shown in (b). d. Hyperuricemia during the early phases of Plasmodium blood stage infection. Uric acid levels in the serum of mice before or after infection with P. chabaudi (d1, d2, d4). e. Uric acid inhibition (UAI) decreases Flt3L release during P. chabaudi infection. WT mice infected with P.chabaudi were treated with allopurinol and uricase (d0 and d1, here termed as UAI) or a control vehicle injection. Serum Flt3L was measured at d2 post infection. f. Uric acid inhibition (UAI) decreases partially MC activation and Flt3L release during P. chabaudi infection. Histogram plots show examples of peritoneal MCs phenotypes at d1 post- P.chabaudi infection (surface Flt3L, left, surcface Lamp1, right). Mice were treated with allopurinol and uricase (d0 and d1, here termed as UAI) or control (vehicle). Quantification of Lamp1 surface upregulation (left bar graph panel) or Flt3L surface staining (right bar graph panel) in various experiments. Each dot represents an individual mouse, mean, SEM and some t-tests are shown. g. Deficient CD8a⁺ DC expansion in Uric Acid Inhibited (UAI treatment) mice infected with P. chabaudi. Lin (CD3,CD19,Ter119,Nk1.1, Gr1) CD11c^highMHCII^high DC compartment in spleen of WT or UAI-treated mice (allopurinol injection, daily), before or after P. chabaudi infection (d8). Absolute numbers per spleen of CD8α⁺ (left bar graph panel) or CD8α⁻ (right bar graph panel) DCs, values are normalized to WT naïve. Each dot is an individual mouse, mean, SEM and some t-tests are shown. h. Deficient T cell activation in UAI in Uric Acid Inhibited (UAI treatment) mice infected with P. chabaudi. T cell activation was assessed at d8 in the spleen by measuring the percentage of CD62L^low/CD44^high cells inside TCRβ⁺CD8⁺ (upper flow plot panel) or TCRβ⁺/CD4⁺ (lower flow plot panel). Absolute accumulation of activated TCRβ⁺CD8⁺(left bar graph panel) and TCRβ⁺CD4⁺(right bar graph panel) cells in the spleen are shown. Each dot represents an individual mouse, mean, SEM and some t-tests are shown.

Figure 5. Role of TLR and type I IFN-dependent signaling in uric acid up-regulation, Flt3L release, DC expansion and T cell activation during Plasmodium blood stage infection

a : Impaired serum Flt3L response in Myd88^−/−/TRIF^−/−, Myd88^−/−, TLR9^−/− but not TRIF^−/−. WT or TLR mutant mice mice were infected with P. chabaudi and serum Flt3L measured at d2 after infection. b : Impaired serum IFN-α response in Myd88^−/−/TRIF^−/−. c : Decreased serum Flt3L response in type I IFN receptor deficient mice. WT or IFNα/βRI^−/− mice were infected with P. chabaudi and serum Flt3L measured at d2 and d4 after infection. d : Type I IFN signaling is required in radiosensitive cells for an efficient Flt3L response. Irradiated WT or IFNα/βRI^−/− recipients were injected with WT or IFNα/βRI^−/− BM and infected at least 8 weeks after with P. chabaudi. Flt3L serum levels were measured after 4d. e : Type I IFN receptor mediates increased Xanthine dehydrogenase production and uric acid levels upregulation. XdH mRNA was measured by qPCR on total lung mRNAs in WT or IFNα/βRI^−/− mice, before or after P. chabaudi infection (d1 and d2, left panel). XdH activity (middle panel) and Uric acid levels (right panel) were measured in total lung extracts at d1.5 after infection with P. chabaudi in WT or IFNα/βRI^−/− mice. f : Decreased MC activation and Flt3L release in type I IFN receptor deficient mice. Histogram plots show examples of peritoneal MCs phenotypes at d1 post- P.chabaudi infection (surface Flt3L, left, surface Lamp1, right) in WT or type I IFN receptor deficient mice. Quantification of Lamp1 surface upregulation (left bar graph) or Flt3L surface staining (right bar graph) in various experiments. g : Decreased spleen CD8α⁺ DC expansion in type I IFN receptor deficient mice. Lin⁻ (CD3,CD19,Ter119,Nk1.1) CD11c^highMHCII^high spleen DCs from WT or IFNα/βRI^−/− mice, naïve or after P. chabaudi infection (d8). Absolute numbers of CD8α⁺ DCs (left bar graph panel), and CD8α⁻ DCs (right bar graph panel) in the various groups as indicated, values are normalized to WT naïve for comparison. h : Deficient T cell activation in type I IFN receptor deficient mice. T cell activation was assessed at d8 in the spleen by measuring the percentage of CD62L^low/CD44^high cells inside TCRβ⁺CD8⁺ (upper panel) or TCRβ⁺/CD4⁺ (lower panel). Absolute accumulation of activated TCRβ⁺CD4⁺(left bar graph panel) and TCRβ⁺CD8⁺(right bar graph panel) cells in the spleen are shown. Each dot represents an individual mouse, mean, SEM and some t-tests are shown.

Figure 6. Systemic Flt3L increase, BDCA3⁺ DCs expansion and CD8⁺ T cell responses in patients infected by P. falciparum

a. Plasma Flt3L concentration in healthy children and children suffering from mild or severe malaria. Numbers indicate the numbers of Flt3L positive patients for each group. Each dot indicates a patient. Mean and Kruskal Wallis significance are shown. b–d. Circulating BDCA1⁺ (b), BDCA2⁺(c), BDCA3⁺(d) DCs for a subgroup of patients (experimental values taken from plotted against plasma Flt3L concentration (measured in this study). Each dot indicates a patient. Median and interquartile range is shown. e,f. Circulating (e) or spleen (f) BDCA1⁺, BDCA2⁺, BDCA3⁺ DCs in the blood of humanized mice treated or not 8 days with recFlt3L (10µg/day) after gating on human CD45⁺, lin-, CD14⁻,CD16⁻ (Supplementary Fig.15a). Results are quantified as percent of humanCD45+ postive cells (e,f). A representative experiment is shown. g,h. Activation status of circulating CD4⁺ (g) and CD8⁺ (h) T cells in patients classified upon their Flt3L levels. Median and interquartile range is shown.