Fas determines differential fates of resident and recruited macrophages during resolution of acute lung injury

Abstract

Rationale: During acute lung injury (ALI) the macrophage pool expands markedly as inflammatory monocytes migrate from the circulation to the airspaces. As inflammation resolves, macrophage numbers return to preinjury levels and normal tissue structure and function are restored.

Objectives: To determine the fate of resident and recruited macrophages during the resolution of ALI in mice and to elucidate the mechanisms responsible for macrophage removal.

Methods: ALI was induced in mice using influenza A (H1N1; PR8) infection and LPS instillation. Dye labeling techniques, bone marrow transplantation, and surface immunophenotyping were used to distinguish resident and recruited macrophages during inflammation and to study the role of Fas in determining macrophage fate during resolving ALI.

Measurements and main results: During acute and resolving lung injury from influenza A and LPS, a high proportion of the original resident alveolar macrophages persisted. In contrast, recruited macrophages exhibited robust accumulation in early inflammation, followed by a progressive decline in their number. This decline was mediated by apoptosis with local phagocytic clearance. Recruited macrophages expressed high levels of the death receptor Fas and were rapidly depleted from the airspaces by Fas-activating antibodies. In contrast, macrophage depletion was inhibited in mice treated with Fas-blocking antibodies and in chimeras with Fas-deficient bone marrow. Caspase-8 inhibition prevented macrophage apoptosis and delayed the resolution of ALI.

Conclusions: These findings indicate that Fas-induced apoptosis of recruited macrophages is essential for complete resolution of ALI.

Figure 1.

Resident alveolar macrophage turnover at baseline. (A–D) Bronchoalveolar lavage (BAL) was performed on lung-shielded Green fluorescent protein (GFP) chimeras after bone marrow transplantation (n = 8 per time point). (A) Total alveolar macrophage counts in BAL. The dashed line represents the mean for all samples. (B) Alveolar macrophage GFP expression. Macrophages were identified by their forward-scatter and side-scatter properties (R1) and the macrophage marker, F4/80 (R2). GFP was used to discriminate alveolar macrophages (AM) of pulmonary origin (R3) from macrophages of bone marrow origin (R4). Representative flow cytometry plots from BAL performed 30 days after transplantation are shown. (C and D) The percentage and total number of GFP-negative alveolar macrophages (i.e., recipient origin) are shown (n = 6–8 per time point). The dashed line represents the mean for all samples. (E) Naive mice were treated with intratracheal PKH26-PCL dye. BAL was performed at 1, 15, or 30 days later. PKH fluorescence intensity was assessed on F4/80⁺ macrophages. (F) Peritoneal lavage from abdomen-shielded GFP chimeras (n = 4 per time point). F4/80⁺ peritoneal macrophages (PM) were assessed for GFP expression.

Figure 2.

Macrophage kinetics in the lungs during LPS-induced injury. (A and B) Lung-shielded bone marrow chimeras were treated with high-dose (200 μg) intratracheal LPS (n = 12 per time point). (A) Leukocyte counts from bronchoalveolar lavage (BAL). (B) Resident (Green fluorescent protein negative [GFP⁻]) and recruited (GFP⁺) macrophages discriminated on the basis of GFP expression. (C and D) BAL from mice treated with PKH26-PCL followed by LPS (200 μg) (n = 6 per group). (C) PKH26 fluorescent intensity on F4/80⁺ alveolar macrophages (AM). Representative histograms are shown. (D) F4/80⁺ PKH^high and F4/80⁺ PKH^low cell populations sorted from BAL. Wright-Giemsa–stained cytospin samples from LPS Day 6 are shown (×60). (E) PKH labeling on F4/80⁺ macrophages lavaged from GFP bone marrow chimeras treated with PKH26-PCL followed by LPS. The corresponding histogram displays PKH fluorescence for resident (F4/80⁺ GFP⁻) and recruited (F4/80⁺ GFP⁺) AM. (F) Resident and recruited AM counts in BAL from wild-type mice treated with PKH and LPS (200 μg). Resident macrophages were defined as F4/80⁺ PKH^high cells. Recruited macrophages were defined as F4/80⁺ PKH^low cells (n = 6 per group). * P < 0.05 versus Day 0.

Figure 3.

Lung injury after low-dose LPS (20 μg) in lung-shielded bone marrow chimeras (A and B) and wild-type mice (C and D). (A) Leukocyte counts from BAL. (B) Resident (F4/80⁺ Green fluorescent protein [GFP]⁻) and recruited (F4/80⁺ GFP⁺) macrophages discriminated on the basis of GFP expression (n = 8 per group). (C) Histologic lung injury scores from tissue sections stained with hematoxylin and eosin. (D) Albumin concentrations in BAL fluid (n = 6 per group). * P < 0.05 versus Day 0.

Figure 4.

Differential expression of CD11b and CD11c distinguishes resident and recruited alveolar macrophages. (A) Dot plots representing CD11b and CD11c staining on macrophage subpopulations lavaged from lung-shielded Green fluorescent protein (GFP) chimeras treated with LPS (20 μg). Macrophages were identified with F4/80 mAb (R1). Resident alveolar macrophages were defined as GFP⁻ cells (R2), whereas recruited macrophages were defined as GFP⁺ (R3). Dot plots representing CD11b and CD11c staining on macrophage subpopulations shown for each time point are representative of three independent bone marrow transplant experiments. (B and C) Resident (F4/80⁺ CD11c^high CD11b^low) and recruited (F4/80⁺ CD11c^low CD11b^high) macrophages from wild-type mice treated with 20 μg LPS (n = 8 mice per time point). * P < 0.05 versus Day 0. FSC = Forward Scatter.

Figure 5.

Recruited macrophages undergo apoptosis during resolving inflammation. (A and B) Annexin V and propidium iodide staining on resident and recruited alveolar macrophages lavaged from CD45.1 bone marrow chimeras. Resident macrophages were F4/80⁺ CD45.1⁻ and recruited macrophages were F4/80⁺ CD45.1⁺. n = 12 for all groups except Day 9 where n = 8. * P < 0.05 versus Day 0. (C) Activated caspase-3 staining on resident (F4/80⁺ CD11c^high CD11b^low) and recruited (F4/80⁺ CD11c^low CD11b^high) macrophages. n = 10 per group; * P < 0.05 versus Day 0. (D) Macrophage cell death in paraffin-embedded tissue sections using terminal deoxynucleotidyltransferase-mediated dUDP nick-end labeling (TUNEL; red). Macrophages were identified using Mac-3 Ab (green) and were distinguished from autofluorescent erythrocytes based on cell morphology and the presence of nuclear staining with DAPI (blue). Sections incubated with DNase-1 provided a positive control for TUNEL. Negative controls for TUNEL were processed without the enzyme terminal deoxynucleotidyl transferase. Apoptotic macrophages displayed intense TUNEL staining that colocalized with DAPI nuclear staining (large arrow). Macrophages containing ingested apoptotic cells were also present (short arrowhead). (E) Cytospin samples from LPS-treated chimeras were used to identify macrophages that had ingested apoptotic cells. Representative images are shown (×60). Arrows indicate ingested apoptotic cells.

Figure 6.

Recruited mononuclear phagocytes express high levels of Fas. (A) Fas expression on recruited (F4/80⁺ CD11c^low CD11b^high) and resident (F4/80⁺ CD11c^high CD11b^low) macrophages from LPS-treated mice. (B) Fas mean fluorescent intensity (MFI) for resident and recruited macrophages. Δ MFI = MFI Fas – MFI Isotype. n = 8 mice per group. ** P < 0.001 versus resident Δ MFI.

Figure 7.

Administration of a Fas-activating antibody depletes recruited mononuclear phagocytes from LPS-injured lungs. LPS-injured mice were treated with an agonistic Fas antibody (or isotype control) during resolving inflammation. (A) BAL macrophage counts 24 hours after Ab administration. (B) CD11c and CD11b expression on F4/80⁺ cells. (C) Recruited (F4/80⁺ CD11c^low/CD11b^high) and resident (F4/80⁺ CD11c^high/CD11b^low) macrophages were quantified for Fas antibody–treated mice (white bars) and mice treated with the isotype control antibody (black bars). n = 12 mice per group. * P < 0.05 for Fas versus isotype.

Figure 8.

Fas is required for contraction of the inflammatory macrophage pool. (A) Wild-type mice were treated with LPS (20 μg) followed a Fas-blocking antibody (administered on LPS Days 4 and 7) or isotype control. Macrophage counts from lavage. n = 10 mice per group; * P < 0.05 versus isotype control. (B–D) Bone marrow chimeras were created by adoptively transferring bone marrow from Fas knockout (Fas -> Green Fluorescent Protein [GFP]) or wild-type mice (WT -> GFP) into GFP-expressing recipients that were irradiated with lead shields protecting the lungs. (B) Total macrophages in lavage on Day 4 (n = 4 per group) and Day 10 (n = 8 for wild-type chimeras, n = 11 for Fas chimeras). (C) Recruited (F4/80⁺ GFP⁻) macrophages in lavage on Day 10. (D) Histologic lung injury scores for LPS-treated chimeras. * P < 0.05, ** P < 0.01 for Fas knockout versus wild-type chimeras.

Figure 9.

Caspase-8 inhibition prevents contraction of the macrophage pool. Wild-type mice were treated with LPS followed 4 days later by a selective inhibitor of caspase-8 (Z-IETD-FMK, 0.1 μM), phosphate-buffered saline, or control (Z-FA-FMK 0.1 μM) daily for 6 days. Gray bars represent treatment-naive mice. (A) Macrophage counts in bronchoalveolar lavage (BAL) (n = 10 per group). (B) Lung injury scores from tissue sections (n = 5 per group). (C) Albumin concentrations in BAL (n = 5 per group). * P < 0.05 versus control; # P = 0.09 versus control.

Figure 10.

Lung injury after influenza A infection (H1N1; PR8). Lung injury was assessed using (A) bronchoalveolar lavage (BAL) cell counts, (B) BAL albumin concentrations, and (C) lung histology. (D) Resident (F4/80⁺ CD11c^high CD11b^low) and recruited (F4/80⁺ CD11c^low CD11b^high) macrophage levels in BAL fluid. (E) Macrophage cell death quantified in tissue sections using terminal deoxynucleotidyltransferase-mediated dUDP nick-end labeling (TUNEL). n = 5 mice per group * P < 0.05 versus Day 0.