MerTK-dependent efferocytosis by monocytic-MDSCs mediates resolution of ischemia/reperfusion injury after lung transplant

Abstract

Lung transplantation (LTx) outcomes are impeded by ischemia/reperfusion injury (IRI) and subsequent chronic lung allograft dysfunction (CLAD). We examined the undefined role of receptor Mer tyrosine kinase (MerTK) on monocytic myeloid-derived suppressor cells (M-MDSCs) in efferocytosis to facilitate resolution of lung IRI. Single-cell RNA sequencing of lung tissue and bronchoalveolar lavage (BAL) from patients after LTx were analyzed. Murine lung hilar ligation and allogeneic orthotopic LTx models of IRI were used with BALB/c (WT), Cebpb-/- (MDSC-deficient), Mertk-/-, or MerTK-cleavage-resistant mice. A significant downregulation in MerTK-related efferocytosis genes in M-MDSC populations of patients with CLAD was observed compared with healthy individuals. In the murine IRI model, a significant increase in M-MDSCs, MerTK expression, and efferocytosis and attenuation of lung dysfunction was observed in WT mice during injury resolution that was absent in Cebpb-/- and Mertk-/- mice. Adoptive transfer of M-MDSCs in Cebpb-/- mice significantly attenuated lung dysfunction and inflammation. Additionally, in a murine orthotopic LTx model, increases in M-MDSCs were associated with resolution of lung IRI in the transplant recipients. In vitro studies demonstrated the ability of M-MDSCs to efferocytose apoptotic neutrophils in a MerTK-dependent manner. Our results suggest that MerTK-dependent efferocytosis by M-MDSCs can substantially contribute to the resolution of post-LTx IRI.

Keywords: Immunology; Innate immunity; Monocytes; Surgery; Transplantation.

Figure 1. Single-cell RNA-sequencing analysis of myeloid cells reveals downregulation of efferocytosis-related genes in M-MDSCs of patients with CLAD.

(A) Uniform manifold approximation and projection (UMAP) visualization of 18 myeloid cell clusters in lung tissue of patients with CLAD and donor controls. (B) Venn diagram outlining identification of DEGs for MerTK-dependent efferocytosis. Downregulated genes in CLAD (blue) and DT (yellow) are described. (C) Volcano plot of DEGs of myeloid cell cluster 4. Genes identified by blue dots are ERGs with differential expression of P < 0.05. Orange dots are other DEGs with P < 0.05. Gray dots are genes that are not significant (P > 0.05). (D) Quantification of sol-MER levels in BAL of patients showed significant increase on day 1 after LTx compared with day 0. Box plots show the interquartile range, median (line), and minimum and maximum (whiskers). Data analyzed by Mann-Whitney test; *P < 0.0001; n = 8–10/group.

Figure 2. Increase in M-MDSCs is associated with the endogenous resolution of lung IRI.

(A) Representative schematic depicting the hilar ligation IRI model where left lung undergoes 1 hour of ischemia followed by 6 or 24 hours of reperfusion in BALB/c (WT) and Cebpb^–/– mice. (B–D) Significant lung dysfunction was observed in WT and Cebpb^–/– mice following 6 hours compared with sham controls that was mitigated after 24 hours. However, lung dysfunction was significantly higher in Cebpb^–/– mice compared with WT mice after 24 hours. *P < 0.0001 vs. WT sham; ^#P < 0.0001 vs. Cebpb^–/– sham; ^δP < 0.0001 vs. WT IRI (6 hours); ^ΦP < 0.05 vs. WT IRI (24 hours); n = 6–10/group. (E and F) The percentage of M-MDSCs was significantly upregulated in WT mice following IRI (24 hours) compared with IRI (6 hours) or sham mice, as analyzed by flow cytometry. A markedly significant mitigation of M-MDSCs was observed in Cebpb^–/– mice compared with WT mice. *P = 0.0005 vs. WT sham; ^#P < 0.009 vs. WT sham and Cebpb^–/– IRI (6 hours); ^δP < 0.0001 vs. WT IRI (6 hours) and Cebpb^–/– IRI (24 hours); n = 6–10/group. Data analyzed by 1-way ANOVA and Tukey’s multiple comparisons test.

Figure 3. Pro-inflammatory cytokine expression was significantly increased in Cebpb^–/– mice after IRI.

(A–G) Pro-inflammatory cytokine and chemokine levels in BAL fluid were significantly increased in WT and Cebpb^–/– mice after 6 hours of IRI compared with sham controls and were mitigated in WT mice following 24 hours but not in Cebpb^–/– mice; *P < 0.0001 vs. WT sham; ^#P < 0.0001 vs. Cebpb^–/– sham; ^δP < 0.0001 vs. WT IRI (6 hours); ^ΦP < 0.003 vs. WT IRI (24 hours) n = 4–5/group. (H) Antiinflammatory IL-10 expression was significantly increased in WT mice following 24 hours compared with all other groups. **P < 0.0001 vs. other groups; n = 4–5/group. Data analyzed by 1-way ANOVA and Tukey’s multiple comparisons test.

Figure 4. Cebpb^–/– mice experienced sustained neutrophil infiltration and activation following IRI.

(A–C) PMN infiltration in lung tissue and activation (MPO levels in BAL) were significantly increased in WT and Cebpb^–/– mice after 6 hours compared with sham controls, which were attenuated in WT mice following 24 hours but not in Cebpb^–/– mice. *P < 0.0001 vs. WT sham; ^#P < 0.02 vs. Cebpb^–/– sham; ^δP < 0.0001 vs. WT IRI (6 hours); ^ΦP < 0.01 vs. WT IRI (24 hours); n = 5/group. PMNs were quantified per high-power field (HPF). Scale bars represent 100 μm. (D) sol-MER levels in BAL were significantly increased in both WT and Cebpb^–/– mice after 6 hours compared with respective sham controls. These levels were mitigated in WT mice and significantly decreased in Cebpb^–/– mice after IRI (24 hours) compared with IRI (6 hours). Box plots show the interquartile range, median (line), and minimum and maximum (whiskers). *P < 0.0001 vs. WT sham; ^#P < 0.0001 vs. Cebpb^–/– sham; ^δP < 0.0001 vs. WT IRI (6 hours); ^ΦP < 0.0004 vs. WT IRI (6 hours and 24 hours) and Cebpb^–/– IRI (6 hours); n = 8–10/group. Data analyzed by 1-way ANOVA and Tukey’s multiple comparisons test. (E and F) Efferocytosis of neutrophils by M-MDSCs was significantly increased after 24 hours compared with 6 hours in WT mice following IRI. *P = 0.0016 vs. IRI (6 hours); n = 5–8/group. Data analyzed by Mann-Whitney test.

Figure 5. Administration of M-MDSCs attenuates pulmonary dysfunction after IRI.

(A) Schematic depicting adoptive transfer of WT M-MDSCs prior to IRI (6 hours) in WT mice. (B–D) Treatment with M-MDSCs significantly improved lung dysfunction compared with untreated mice. *P < 0.003 vs. IRI; n = 6–11/group. (E–G) Pro-inflammatory cytokines in BAL fluid were significantly reduced, and antiinflammatory IL-10 expression was significantly increased in M-MDSC–treated mice compared with untreated mice. *P < 0.05 vs. IRI; n = 9/group. (H) Representative images of histological staining for PMNs after IRI with and without M-MDSC treatment. Scale bars represent 100 μm. (I and J) Neutrophil infiltration in lung tissue and MPO levels in BAL were significantly attenuated in mice treated with M-MDSCs. *P < 0.008 vs. IRI; n = 5/group. (K and L) Adoptively transferred M-MDSCs (CellTrace Violet) demonstrated a significant increase in efferocytosis of apoptotic neutrophils compared with endogenous M-MDSCs (Ly6C⁺iNOS⁺) in lung tissue. *P = 0.004 vs. IRI; n = 5–9/group. Data analyzed by Mann-Whitney test.

Figure 6. M-MDSC–mediated resolution of lung IRI is mediated via MerTK.

(A) Schematic depicting adoptive transfer of M-MDSCs prior to IRI (24 hours) in Cebpb^–/– mice. (B–D) Treatment with M-MDSCs from WT mice, but not from Mertk^–/– mice, significantly mitigated lung dysfunction compared with untreated mice. *P < 0.006 vs. other groups; n = 7–10/group. (E–G) Expression of pro-inflammatory cytokines was significantly decreased in mice treated with WT M-MDSCs but not with Mertk^–/– M-MDSCs. *P < 0.05 vs. IRI; ^#P < 0.05 vs. IRI+WT M-MDSCs; n = 6–8/group. (H and I) PMN infiltration was significantly attenuated by treatment with WT-derived M-MDSCs but not Mertk^–/–-derived M-MDSCs. *P < 0.0001 vs. IRI; ^#P = 0.0005 vs. IRI+WT M-MDSCs; n = 5–7/group. Scale bars represent 100 μm. (J) MPO levels in BAL were significantly mitigated by WT M-MDSCs but not by Mertk^–/– M-MDSCs. *P < 0.05 vs. IRI alone; ^#P < 0.05 vs. IRI + WT M-MDSCs; n = 5–7/group. Data analyzed by 1-way ANOVA and Tukey’s multiple comparisons test.

Figure 7. MerTK mediates M-MDSC–dependent efferocytosis in vitro.

(A) Schematic depicting MerTK-mediated efferocytosis of apoptotic PMNs by M-MDSCs. (B) Representative immunofluorescence images demonstrating colocalization (shown by arrows) of M-MDSCs’ (red) engulfment of apoptotic PMNs (green). DAPI is shown in blue. Original magnification, 40×, representative images are shown. (C and D) Representative flow cytometry plots showing engulfment of apoptotic PMNs by WT or Mertk^–/–-derived M-MDSCs. Quantification in lung tissue showed a significant increase in efferocytosis of apoptotic PMNs by WT-derived M-MDSCs but was absent in Mertk^–/–-derived M-MDSCs. *P < 0.0001 vs. other groups; n = 15/group. Data analyzed by 1-way ANOVA and Tukey’s multiple comparisons test.

Figure 8. Resolution of IRI in an allogeneic orthotopic brain-dead model of IRI after LTx is associated with increase in M-MDSCs.

(A) Schematic depicting brain-dead model of LTx in C57BL/6 donors and WT (BALB/c) recipients. (B and C) Representative flow cytometry plots and quantification of M-MDSCs in lung tissue from LTx recipients. The percentage of M-MDSCs was significantly upregulated after 72 hours of reperfusion to 24 hours. *P = 0.03 vs. 24 hours; n = 5–6/group. (D) Albumin levels in BAL were significantly mitigated following 72 hours of reperfusion compared with 24 hours of reperfusion. *P = 0.0079; n = 5/group. (E–G) Expression of pro-inflammatory cytokines in BAL was significantly mitigated and IL-10 expression was significantly increased after 72 hours compared with 24 hours of reperfusion. *P < 0.04; n = 5/group. (H and I) PMN infiltration in lung tissue was significantly abrogated following 72 hours of reperfusion compared with 24 hours. *P = 0.01; n = 4–5/group. Scale bars represent 100 μm. (J) MPO expression in BAL was significantly mitigated following 72 hours of reperfusion. *P = 0.0079; n = 5/group. (K) sol-MER levels in BAL were significantly decreased in LTx recipient mice following 72 hours of reperfusion compared with 24 hours. *P = 0.016; n = 5/group. Data analyzed by Mann-Whitney test.