Inflammatory monocytes orchestrate innate antifungal immunity in the lung

Abstract

Aspergillus fumigatus is an environmental fungus that causes invasive aspergillosis (IA) in immunocompromised patients. Although -CC-chemokine receptor-2 (CCR2) and Ly6C-expressing inflammatory monocytes (CCR2⁺Mo) and their derivatives initiate adaptive pulmonary immune responses, their role in coordinating innate immune responses in the lung remain poorly defined. Using conditional and antibody-mediated cell ablation strategies, we found that CCR2⁺Mo and monocyte-derived dendritic cells (Mo-DCs) are essential for innate defense against inhaled conidia. By harnessing fluorescent Aspergillus reporter (FLARE) conidia that report fungal cell association and viability in vivo, we identify two mechanisms by which CCR2⁺Mo and Mo-DCs exert innate antifungal activity. First, CCR2⁺Mo and Mo-DCs condition the lung inflammatory milieu to augment neutrophil conidiacidal activity. Second, conidial uptake by CCR2⁺Mo temporally coincided with their differentiation into Mo-DCs, a process that resulted in direct conidial killing. Our findings illustrate both indirect and direct functions for CCR2⁺Mo and their derivatives in innate antifungal immunity in the lung.

Figure 1. CCR2⁺ cells protect against Invasive Aspergillosis.

A–B) CCR2 depleter (solid gray line) and control B6 non-transgenic littermates (solid black line) were treated with 250 ng of DT i.p. on day −1, +1, and +3. Neutrophil depleted mice (dashed black line) were B6 mice injected with 1A8 (anti-Ly6G antibodies) daily. (A) All animals were infected with 8×10⁷ live A.fumigatus conidia. The graph shows Kaplan-Meier survival of individual groups pooled from two independent experiments with 4–5 mice per group per experiment. Statistical analysis was performed with log-rank test and Bonferroni correction for multiple comparisons: WT vs. CCR2 depleter P = 0.0002, WT vs anti-Ly6G treated P = 0.0003. (B) Kaplan-Meier survival of DT-treated B6 (solid black line, inoculum 6×10⁷ conidia) and CCR2 depleter mice (6×10⁷ conidia, dashed black line; 4×10⁷ conidia, solid grey line). Statistical analysis was performed as described in (A). WT vs. CCR2 depleter 6×10⁷ p = <0.0001, WT vs CCR2 depleter 4×10⁷ p = 0.001. Data shown is for five mice per group. (C) Representative photomicrographs of formalin-fixed GMS-stained lung sections collected at the indicated times p.i. from DT-treated CCR2 depleter (top row) and B6 mice (bottom row). Naïve animals were sacrificed at day +6 and received 3 doses of DT. Sections shown are for one mouse per group and are representative of 3–5 mice that were examined per group per time point in two independent experiments.

Figure 2. CCR2⁺ NK cells and innate lymphocytes are dispensable for innate defense against IA.

(A) Representative plots of CD45⁺ lung cells obtained from control B6, DT-treated CCR2 depleter mice, and RAG^−/−γC^−/− mice one day p.i. with 8×10⁷ A.fumigatus conidia and analyzed for NK1.1 expression. B–D) The bar graphs show the total number of lung (B) NK1.1⁺ cells, (C) CD11b⁺Ly6G⁺Ly6C⁺ neutrophils, or (D) CD11b⁺Ly6G⁻Ly6C⁺ monocytes (CCR2⁺Mo) in DT-treated CCR2 depleter (gray bars), control mice (white bars), or RAG^−/−γC^−/− (black bars) at day +1 and +2 p.i. (E–F) The scatter plots show the mean ± SEM of lung CFUs recovered from control (white circles), DT-treated CCR2 depleter mice (gray circles) or RAG^−/−γC^−/− (black circles) at day +1 and +2 p.i. (B–F) Data shown is for mean ± SEM for 4–5 mice per group from one of two representative experiments. Mann-Whitney test used for statistical analyses, * p<0.05, **p<0.01. G) The photomicrograph shows GMS-stained lung tissue from a representative RAG^−/−γC^−/− mouse on day +3 p.i.

Figure 3. CCR2⁺ cells are dispensable for the production of neutrophil chemokines and neutrophil recruitment.

(A–E) Control and CCR2 depleter mice were treated with DT and infected with 6×10⁷ conidia on day 0 and euthanized at the indicated times for ELISA of lung homogenates and FACS analysis of lung single cell suspensions. (A–B) The scatter plots show mean ± SEM lung (A) CXCL1 and (B) CXCL2 levels at 48 h p.i. in CCR2 depleter (white circles) and control B6 mice (black circles). (C–E) Representative FACS plots (day+1 p.i.) from CCR2 depleter (C, top row) and control B6 mice (C, bottom row) gated on lung CD45⁺CD11b⁺ cells and analyzed for Ly6C and Ly6G. Monocytes (Mo) are identified as Ly6C⁺Ly6G⁻ cells while neutrophils (Ne) are identified as Ly6G⁺Ly6C⁺cells. (D) The graph shows mean number (±SEM) of monocytes recovered from the lung of DT-treated B6 mice (black circles) or CCR2 depleter mice (white triangles) at the indicated time points p.i. Pooled data shown from three independent experiments (3–5 mice per group and per expt.). (E) The scatter plots show mean ± SEM of number of neutrophils recovered from the lung of CCR2 depleter mice (white circles) or control littermates (black circles) at various times after infection. Each symbol represents one mouse. Data is cumulative for two or three independent experiments with 3–5 mice per group per time point. (F–G) The bar graphs show the mean number (±SEM) of lung monocytes (F) and neutrophils (G) recovered from anti-Ly6G-treated and control mice as described in Figure 1. Statistical analyses were performed using Mann Whitney tests, n.s (not significant), * p<0.05.

Figure 4. Diminished neutrophil conidiacidal activity in CCR2 depleter mice.

CCR2 depleter and control mice were treated with 10/gm DT on day −1 and day 0 and infected with 3×10⁷ FLARE conidia. (A) Representative FACS plots of lung neutrophils isolated from CCR2 depleter mice and control mice and analyzed for dsRed and AF633 fluorescence. Plots show the frequencies of neutrophils that contain live (red gate) or killed conidia (blue gate) at 36 h p.i. (B) The scatter plots pooled from 2 experiments show the average frequency (± SEM) of lung neutrophil conidial uptake (R1+R2) and (C) lung neutrophil conidial viability (R1/(R1+R2) in CCR2 depleter and control mice. ^*p<0.05 by Mann-Whitney test. (D) Representative FACS plots of bone marrow neutrophils isolated from control or CCR2 depleter mice and cultured in vitro with FLARE conidia. Neutrophils were identified as CD45⁺CD11b⁺Ly6G⁺ cells and analyzed for dsRed and AF633 fluorescence as shown. (E and F) The scatter plots pooled from 2 experiments show the average frequency (± SEM) of bone marrow in vitro neutrophil conidial uptake (R1+R2)(E) and in vitro conidial viability (R1/(R1+R2) in bone marrow neutrophils isolated from CCR2 depleter and control mice (F). ^**p<0.01 by Mann-Whitney test.

Figure 5. Inflammatory responses of CCR2⁺Mo and Mo-DC during respiratory fungal infection.

Lung CCR2⁺Mo (GFP⁺CD45⁺CD11b⁺CD11c⁻Nk1.1⁻) and Mo-DC (GFP⁺CD45⁺CD11b⁺CD11c⁺NK1.1⁻) were FACS sorted 48 h p.i. from CCR2 reporter mice (purity >97% for all sorts) for transcriptome analysis by RNA-seq (A) or for quantitative RT-PCR (B). Control CCR2⁺Mo were also isolated from the lung of uninfected CCR2 reporter mice (naïve sample) to >97% purity. (A) Gene expression data shown in A is for one experiment and representative of 3 independent biological replicates and three idependent sequencing reactions using SOLiD sequencing platform. Differences in gene expression are shown as fragments per kilobase (FPKM) as calculated using Cufflinks and R software. (B) The graphs show expression of specific transcripts in the indicated cell populations by qRT-PCR using Taq-Man probes normalized to GAPDH. Data shown is mean ±SEM pooled from two separte experiments. (C) The graph shows pulmonary Nos2 induction in DT-treated CCR2 depleter and control mice at the indicated time points p.i. Data shown is mean ±SEM pooled from two separte experiments with 3 mice per group per time point. (D–E) The scatterplots show mean ± SEM lung (D) IL-12p70 and (E) TNF levels at 48 h p.i. in CCR2 depleter (grey circles) and control B6 mice (black circles) as in Figure 3A.

Figure 6. CCR2⁺Mo differentiate into Mo-DC and efficiently kill A.fumigatus.

CCR2 reporter mice were infected with FLARE conidia and lung cell suspensions were enumerated and examined by (A) imaging cytometry and (B–H) flow cytometry. (A) Imaging cytometry of lung GFP⁺ (CCR2⁺) cells from FLARE-infected CCR2 reporter mice 36 h p.i. The micrograph depicts dsRed⁺AF633⁺ and dsRed⁻AF633⁺ monocytes that contain live and killed conidia, respectively. BF, bright-field. (B–E) The graphs show the total number (mean ± SEM) of lung CCR2⁺Mo (white circles) and Mo-DCs (black circles) at the indicated times p.i. . CCR2⁺Mo (white circles), and Mo-DC (black circles) were identified as shown in Figure S1. (B) Data shows total recrutiment of each subset over time. (C) The graph shows the total number of CCR2⁺Mo (white circles) or Mo-DCs (black circles) that contain engulfed conidia. (D–E) The graph shows the total number of CCR2⁺Mo (white circles) or Mo-DCs (black circles) that contain (D) live or (E) killed conidia. (F–H) Comparison of CCR2⁺Mo, Mo-DC and neutrophil conidiacidal activity. The scatter plots show the frequency of fungus-engaged (F) CCR2⁺Mo, (G) Mo-DC, and (H) neutrophils that contain live (red circles) or killed (blue circles) FLARE conidia at the indicated times p.i. Results are pooled from two experiments. (I) The graph shows lung CFUs from DT-treated B6 controls (white circles), DT-treated CCR2 depleter (black squares), and anti-Ly6G-treated B6 mice (grey triangles) at day +1 p.i.. Each symbol represents one mouse. Results are for one experiment representative of two individual experimenst for all data shown in this figure. Statistical analysis was performed using a Mann Whitney test.