Deletion of c-FLIP from CD11bhi Macrophages Prevents Development of Bleomycin-induced Lung Fibrosis

Abstract

Idiopathic pulmonary fibrosis is a progressive lung disease with complex pathophysiology and fatal prognosis. Macrophages (MΦ) contribute to the development of lung fibrosis; however, the underlying mechanisms and specific MΦ subsets involved remain unclear. During lung injury, two subsets of lung MΦ coexist: Siglec-F^hi resident alveolar MΦ and a mixed population of CD11b^hi MΦ that primarily mature from immigrating monocytes. Using a novel inducible transgenic system driven by a fragment of the human CD68 promoter, we targeted deletion of the antiapoptotic protein cellular FADD-like IL-1β-converting enzyme-inhibitory protein (c-FLIP) to CD11b^hi MΦ. Upon loss of c-FLIP, CD11b^hi MΦ became susceptible to cell death. Using this system, we were able to show that eliminating CD11b^hi MΦ present 7-14 days after bleomycin injury was sufficient to protect mice from fibrosis. RNA-seq analysis of lung MΦ present during this time showed that CD11b^hi MΦ, but not Siglec-F^hi MΦ, expressed high levels of profibrotic chemokines and growth factors. Human MΦ from patients with idiopathic pulmonary fibrosis expressed many of the same profibrotic chemokines identified in murine CD11b^hi MΦ. Elimination of monocyte-derived MΦ may help in the treatment of fibrosis. We identify c-FLIP and the associated extrinsic cell death program as a potential pathway through which these profibrotic MΦ may be pharmacologically targeted.

Keywords: IPF; RNA-seq; c-FLIP; fibrosis; macrophage.

Figure 1.

CD11b^hi macrophages (MΦ) drive an increase in total lung MΦ numbers after bleomycin. (A) Flow cytometry gating strategy used on lung digest and BAL to isolate Siglec-F^hi and CD11b^hi MΦ. Representative gating is shown on lung digest 14 days after bleomycin. Live cells are selected using size (forward scatter [FSC]) and granularity (side scatter [SSC]), doublets are excluded, and leukocytes are selected using CD45. Ly6G⁺ neutrophils are excluded, and CD64 and Mertk coexpression is used to select MΦ. (B) Siglec-F and CD11b are used to divide MΦ into two subpopulations. Representative dot plots from lung digest 4, 7, and 14 days after bleomycin are shown. (C and D) Total numbers of Siglec-F^hi and CD11b^hi MΦ were assessed in naive mice and mice 7 and 14 days after bleomycin. (C) MΦ in lung digest. (D) MΦ in BAL. Data from n = 3–5 mice/group shown as mean (±SEM). **P < 0.01, ***P < 0.001. IT = intratracheal.

Figure 1.

CD11b^hi macrophages (MΦ) drive an increase in total lung MΦ numbers after bleomycin. (A) Flow cytometry gating strategy used on lung digest and BAL to isolate Siglec-F^hi and CD11b^hi MΦ. Representative gating is shown on lung digest 14 days after bleomycin. Live cells are selected using size (forward scatter [FSC]) and granularity (side scatter [SSC]), doublets are excluded, and leukocytes are selected using CD45. Ly6G⁺ neutrophils are excluded, and CD64 and Mertk coexpression is used to select MΦ. (B) Siglec-F and CD11b are used to divide MΦ into two subpopulations. Representative dot plots from lung digest 4, 7, and 14 days after bleomycin are shown. (C and D) Total numbers of Siglec-F^hi and CD11b^hi MΦ were assessed in naive mice and mice 7 and 14 days after bleomycin. (C) MΦ in lung digest. (D) MΦ in BAL. Data from n = 3–5 mice/group shown as mean (±SEM). **P < 0.01, ***P < 0.001. IT = intratracheal.

Figure 2.

Cellular FADD-like IL-1β–converting enzyme–inhibitory protein (cFLIP)^Δ/Δ mice delete c-FLIP in CD11b^hi MΦ and sensitize CD11b^hi MΦ to caspase-8–dependent cell death. (A–D) Caspase activity was measured using Caspase-Glo luminescent assays. MΦ were collected by lavage 7 days after bleomycin and treated with doxycycline (20 ng/ml) in vitro for 36 hours to induce deletion of c-FLIP. Doxycycline was removed and cells were treated with α-Fas activating antibody (5 μg/ml) for a further 8 hours (caspase-8) or 12 hours (caspase-3). Some wells were treated with the selective caspase-8 inhibitor, Z-IETD-FMK, at 10 μM for 30 minutes before α-Fas. (A and B) Caspase-8 activity of CD11b^hi MΦ from cFLIP^Δ/Δ (A) or cFLIP^fl/fl mice (B) after doxycycline treatment and/or stimulation with activating α-Fas antibody. (C and D) Caspase-3 activity of CD11b^hi MΦ from cFLIP^Δ/Δ mice (C) or cFLIP^fl/fl mice (D) after doxycycline treatment and stimulation with activating α-Fas antibody. (E and F) Control Tomato-reporter mice and cFLIP^Δ/Δ-Tomato mice were treated with bleomycin and then started on doxycycline at Day 7. At Day 10, lungs were collected, inflated, and frozen with optimal cutting temperature compound for histology. Double-stranded DNA breaks were labeled with terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL; green), and tissue sections were costained with CD68 (teal) and DAPI (blue), along with endogenous Tomato reporter (red). Tomato⁺ and Tomato⁻ MΦ were counted and assessed for TUNEL positivity. Representative images are shown, with white arrowheads pointing to Tomato⁻ viable MΦ, the red arrowhead pointing to a Tomato⁺ viable MΦ, and the green arrowhead pointing to a TUNEL⁺ Tomato⁺ apoptotic MΦ. Data from n = 4–7 mice from three separate experiments shown as mean (±SEM). *P < 0.05. NS = not significant.

Figure 3.

CD11b^hi MΦ, but not Siglec-F^hi MΦ, numbers are decreased in cFLIP^Δ/Δ mice after bleomycin. Flow cytometry was used to assess total MΦ numbers 0, 4, 7, and 14 days after bleomycin in lavage and lung digests. cFLIP^Δ/Δ and cFLIP^fl/fl mice were started on doxycycline 1 day after bleomycin. (A–C) Total MΦ, Siglec-F^hi MΦ, and CD11b^hi MΦ numbers in BAL. (D–F) Total MΦ, SiglecF⁺ MΦ, and CD11b^hi MΦ numbers in lung digests. (G) Bleomycin-treated mice were given IT α-CD45 to label alveolar cells (IT CD45). BAL was performed 3 minutes later to sample alveolar cells and remove unbound antibody. Lung digestion was performed immediately after lavage. BAL and lung digest was processed for flow cytometry; Siglec-F^hi MΦ and CD11b^hi MΦ were separated as described in Figure 1A. IT CD45 staining is shown in BAL and lung digest. BAL staining acts as a positive control; lung digest from mice not given IT CD45 help set the negative gate. (H) CD11b^hi MΦ can be separated into high-and low-SSC populations using SSC versus CD11c. (I) Low-SSC CD11b^hi MΦ correlate with IT⁻ tissue MΦ, whereas high-SSC CD11b^hi MΦ correlate with IT⁺ alveolar MΦ. (J) Representative dot plots showing high- and low-SSC subpopulations in cFLIP^fl/fl and cFLIP^Δ/Δ mice 14 days after bleomycin. (K) Total numbers of high- and low-SSC CD11b^hi MΦ 14 days after bleomycin corresponding to alveolar and tissue CD11b^hi MΦ, respectively. Data from n = 3–5 mice/group shown as mean (±SEM). *P < 0.05, **P < 0.01.

Figure 3.

CD11b^hi MΦ, but not Siglec-F^hi MΦ, numbers are decreased in cFLIP^Δ/Δ mice after bleomycin. Flow cytometry was used to assess total MΦ numbers 0, 4, 7, and 14 days after bleomycin in lavage and lung digests. cFLIP^Δ/Δ and cFLIP^fl/fl mice were started on doxycycline 1 day after bleomycin. (A–C) Total MΦ, Siglec-F^hi MΦ, and CD11b^hi MΦ numbers in BAL. (D–F) Total MΦ, SiglecF⁺ MΦ, and CD11b^hi MΦ numbers in lung digests. (G) Bleomycin-treated mice were given IT α-CD45 to label alveolar cells (IT CD45). BAL was performed 3 minutes later to sample alveolar cells and remove unbound antibody. Lung digestion was performed immediately after lavage. BAL and lung digest was processed for flow cytometry; Siglec-F^hi MΦ and CD11b^hi MΦ were separated as described in Figure 1A. IT CD45 staining is shown in BAL and lung digest. BAL staining acts as a positive control; lung digest from mice not given IT CD45 help set the negative gate. (H) CD11b^hi MΦ can be separated into high-and low-SSC populations using SSC versus CD11c. (I) Low-SSC CD11b^hi MΦ correlate with IT⁻ tissue MΦ, whereas high-SSC CD11b^hi MΦ correlate with IT⁺ alveolar MΦ. (J) Representative dot plots showing high- and low-SSC subpopulations in cFLIP^fl/fl and cFLIP^Δ/Δ mice 14 days after bleomycin. (K) Total numbers of high- and low-SSC CD11b^hi MΦ 14 days after bleomycin corresponding to alveolar and tissue CD11b^hi MΦ, respectively. Data from n = 3–5 mice/group shown as mean (±SEM). *P < 0.05, **P < 0.01.

Figure 4.

Deletion of c-FLIP from CD11b^hi MΦ protects against the development of fibrosis. cFLIP^fl/fl and cFLIP^Δ/Δ mice were treated with bleomycin and started on doxycycline 1 day later; fibrosis was assessed at Day 14. Doxycycline-only controls comprised of both cFLIP^fl/fl and cFLIP^Δ/Δ mice were given doxycycline for a matched 13 days before death. (A) Static compliance measured by flexiVent. (B) Hydroxyproline levels in the right lung. (C and D) Trichrome stain on paraffin-embedded lung sections from mice started on doxycycline 1 day after bleomycin and isolated at Day 14. Collagen stains blue–green. Zoom image of pleural surface highlights collagen deposition. (E–G) cFLIP^fl/fl and cFLIP^Δ/Δ mice were treated with bleomycin and started on doxycycline 7 days later. Fibrosis was assessed at Day 14. (E) MΦ numbers from lung digest of the left lung. (F) Static compliance measured by flexiVent. (G) Hydroxyproline content in the right lung. Data from n = 6–11 mice/group shown as mean (±SEM). *P < 0.05. Bleo = bleomycin; Dox = doxycycline.

Figure 5.

Global analysis of CD11b^hi MΦ and Siglec-F^hi MΦ transcriptomes shows that the two MΦ populations are distinct. Resident Siglec-F^hi alveolar MΦ or resident interstitial MΦ (IMs) from naive lungs (naive Day 0), and Siglec-F^hi and CD11b^hi MΦ from bleomycin-treated lungs after 8 days (Day 8 bleo) were sorted for RNA-seq analysis. (A) Individual replicates of Siglec-F^hi MΦ of IMs from naive lungs and Siglec-F^hi MΦ and CD11b^hi MΦ from bleomycin-treated lungs compared using pairwise Pearson correlation. (B) Heat map showing gene expression in transcripts per million (TPM) on a log scale of MΦ-lineage–associated genes and non–MΦ-lineage–associated genes.

Figure 6.

CD11b^hi MΦ express a profibrotic signature, including high expression of multiple profibrotic chemokines, that is validated in human MΦ from patients with idiopathic pulmonary fibrosis (IPF). (A) Heat maps showing expression of genes encoding chemokines. Unsupervised clustering was used to highlight associations between expression of these genes and MΦ subtypes. Raw expression is displayed on a log scale as average TPM per group (green–purple color scale) and scaled expression shows relative expression of each gene as the change from row mean scaled to a standard deviation of 1 (red–blue color scale). (B) Chemokines with significantly different expression in BAL MΦ from patients with IPF compared with normal volunteers were identified from publically available data in the Gene Expression Omnibus (accession no. GSE49072). Unsupervised clustering was used to highlight associations between expression of these genes and disease state. Scaled expression shows relative expression of each gene as the change from row mean scaled to a standard deviation of 1. Chemokines upregulated in both human IPF MΦ and murine CD11b^hi MΦ compared with control human MΦ or murine SiglecF^hi MΦ after bleomycin are highlighted in red. *Gene with no direct murine/human homolog.