Flt3 ligand expands bona fide innate lymphoid cell precursors in vivo

Abstract

A common helper-like innate lymphoid precursor (CHILP) restricted to the innate lymphoid cells (ILC) lineage has been recently characterized. While specific requirements of transcription factors for CHILPs development has been partially described, their ability to sense cytokines and react to peripheral inflammation remains unaddressed. Here, we found that systemic increase in Flt3L levels correlated with the expansion of Lineage (Lin)^negα4β7⁺ precursors in the adult murine bone marrow. Expanded Lin^negα4β7⁺ precursors were bona fide CHILPs as seen by their ability to differentiate into all helper ILCs subsets but cNK in vivo. Interestingly, Flt3L-expanded CHILPs transferred into lymphopenic mice preferentially reconstituted the small intestine. While we did not observe changes in serum Flt3L during DSS-induced colitis in mice or plasma from inflammatory bowel disease (IBD) patients, elevated Flt3L levels were detected in acute malaria patients. Interestingly, while CHILP numbers were stable during the course of DSS-induced colitis, they expanded following increased serum Flt3L levels in malaria-infected mice, hence suggesting a role of the Flt3L-ILC axis in malaria. Collectively, our results indicate that Flt3L expands CHILPs in the bone marrow, which might be associated with specific inflammatory conditions.

Figure 1

Flt3L-producing tumors expand CHILPs in the BM. Mice were injected with 2 × 10⁶ B16-Flt3L or B16/PBS as a control. Two weeks after tumor injection, Flt3L serum levels were analyzed by ELISA and CHILPs and ILC2P were analyzed by FACS on BM single-cell suspension. (a) Flt3L serum levels in B16/PBS and B16-Flt3L injected mice (n = 3/group, 2 experiments). Representative dot plots (b) and total number (c) of Lin⁻α4β7⁺ cells in the BM of wild-type mice injected with B16-Flt3L or B16/PBS (n = 6/group, 3 experiments). Representative dot plots (d) and absolute numbers (e) of CHILPs and ILC2P in the BM of Id2 ^Gfp/+ mice injected with B16-Flt3L or B16/PBS (n = 6/group, 3 experiments). (f) Correlation plot between Flt3L serum levels determined by ELISA and BM CHILPs absolute numbers determined by FACS in mice treated with B16-Flt3L or B16 (n = 6, 2 experiments). Representative dot plots (g) and absolute numbers (h) of ILCP in the bone marrow of wild type mice injected with B16-Flt3L or B16 cells (n = 5 to 7/group, 3 experiments). *P < 0.05; **P < 0.01; ns, not significant; Student’s t-test (c,e,h), Linear regression with Pearson’s correlation analysis (f). Error bars represent SEM in all panels.

Figure 2

Flt3L-producing tumors do not expand mature ILCs in periphery. Mice were injected with 2 × 10⁶ B16-Flt3L or B16/PBS as a control. Two weeks after tumor injection, ILCs were analyzed by FACS in the lungs, small intestine and colonic lamina propria. (a) Representative dot plots showing NK1.1⁺T-bet⁺Eomes⁻ ILC1, NK1.1⁺T-bet⁺Eomes⁺ cNK, GATA-3⁺ ILC2, RORγt⁺ ILC3 and NKp46⁺ and CCR6⁺ ILC3 subsets gated on CD45⁺ Lin (CD3, CD19)^neg CD90⁺ cells in the colonic lamina propria and (b) quantification of absolute numbers (n = 5 to 7/group, 3 experiments). Representative dot plots (c) and absolute numbers (d) of ILC2 and ILC3 in the small intestine lamina propria and lungs of B16-Flt3L and B16/PBS injected mice (n = 5 to 8/group, 3 experiments). (e) Representative dot plots of IL22⁺ RORγt⁺ ILC3 and IL-5⁺ GATA-3⁺ ILC2 in the small intestine of B16-Flt3L and B16 injected mice (n = 3 to 4, 2 experiments). *P < 0.05; ns, not significant; Student’s t-test (b,d,e). Error bars represent SEM in all panels.

Figure 3

Flt3L expands bona fide CHILPs. FACS-sorted CHILPs were isolated from PBS- (CHILP^control) or B16-Flt3L-injected (CHILP^B16-Flt3L) Id2 ^Gfp/+ mice. CLPs were isolated from PBS-injected mice. The indicated highly purified cell populations were adoptively transferred into congenic alymphoid recipients, which were analyzed 5 to 10 weeks later. (a) Schematics of the experimental design. (b) Representative dot plots showing the analysis of ILCs (CD45.1⁺ CD90⁺ CD3⁻) and T cells (CD45.1⁺ CD90⁺ CD3⁺) from the small intestine lamina propria (SILP). Displayed plots are gated on CD45.1⁺ donor-derived cells. (c) Pie plots showing the frequencies of ILCs, T cells and the rest out of all the donor-derived cells (CD45.1⁺) from different tissues (n = 4 for CHILP^B16-Flt3L group; n = 2 for CHILP^Control group; n = 3 for CLP group, 2 experiments). (d) Engraftment index of CHILP^Control and CHILP^B16-Flt3L in different organs (n = 2 to 4, 2 experiments). (e) Representative dot plots of RORγt⁺ ILC3 and GATA-3⁺ ILC2 gated on total intestinal donor-derived ILCs (n = 4–6/group, 3 experiments). Representative dot plot of RORγt⁺ ILC3 and GATA-3⁺ ILC2 gated on total intestinal ILCs from a non-transferred wild-type C57BL/6 mouse is shown as a control. (f) Representative dot plots showing NK1.1⁺T-bet⁺ cells gated on donor derived CD90⁺ cells in the liver (non-transferred C57BL/6 wild type is shown as a control) and representative histograms showing Eomes⁺ cNK. ns, not significant; **P < 0.01; one-way ANOVA with Bonferroni’s multiple comparisons test (d). Average and standard deviation are displayed in all panels.

Figure 4

Unchanged serum Flt3L levels and CHILP numbers during intestinal inflammation. (a–d) Wild type mice were treated for 7 days with 2.5% DSS in drinking water followed by 7 days of regular water. Mice were sacrificed and analyzed at the indicated time points. (a) Body weight monitored during the course of the treatment. Body weight of mice sacrificed at the different time points are indicated (n = 3/time point). (b) Quantification of CHILPs numbers in the BM of mice treated with DSS at the indicated time point (n = 3 to 11/time point, 3 experiments) (c) mRNA transcript expression for Flt3l in colon was measured by quantitative PCR at the indicated time point (n = 3/time point) (d) Serum levels of Flt3L determined by ELISA at the indicated time points (n = 2 to 6/time point, 2 experiments). (e) FLT3L plasma levels in healthy controls, newly diagnosed IBD patients and chronic Crohn’s disease (CD) and Ulcerative colitis (UC) patients. Each dot represents an individual patient. *P < 0.05; **P < 0.01; ns, not significant; two-way ANOVA with Bonferroni’s multiple comparisons test (b,d) and one-way ANOVA with Bonferroni’s multiple comparisons test (e). Error bars represent SD in all panels.

Figure 5

CHILPs expand in the BM during malaria. (a) Plasma FLT3L in healthy controls and patients suffering from acute P. falciparum malaria. Each dot represents an individual healthy control (n = 28) or Plasmodium falciparum infected patient (acute malaria; n = 39). Dashed line represents the average Flt3L levels in healthy donors. (b) Correlation plot between plasma FLT3L levels and percentage of parasitemia in patients suffering from acute malaria. Each dot represents an individual patient. Coefficient of determination was calculated by Pearson’s correlation analysis. (c,d) Wild type mice were infected with P. berghei ANKA and were analyzed at day 5, and 7 after infection. (c) Schematics of the experimental design. (d) Quantification of CHILPs number in the BM (black dots) and Flt3L serum levels (red dots) at the indicated time points (n = 3 at day 0 and 5; n = 2 at day 7). *P < 0.05; **P < 0.01; ***P < 0.001; ns, not significant; one-way ANOVA with Bonferroni’s multiple comparisons test (d). Error bars represent SD for CHILPs numbers and SEM for Flt3L serum levels.