Neutrophils promote clearance of nuclear debris following acid-induced lung injury

Neutrophil recruitment to sites of sterile injury is commonly observed, but is of unknown significance. Data from a number of groups suggest that sterile inflammatory responses are deleterious and perpetuate pathology.^1-3  Other studies show that the integrated response to sterile injury removes inciting materials and facilitates repair.⁴  Although sterile lesions exhibit commonalities with infection,⁵  resolution of sterile injury remains poorly understood.

We previously demonstrated that neutrophil accumulation in the lung after acid aspiration is necessary for repair.⁶  Mechanisms by which neutrophils facilitate lung repair are lacking, though in models of sterile liver injury, neutrophils^7,8  are necessary for debris clearance, which precedes repair. In these systems, neutrophils have been shown to take up small DNA particles that presumably allowed for improved revascularization.⁷  In models of mechanical lung injury, neutrophils and neutrophil-derived matrix metalloproteinase-9 were shown to be critical for epithelial and matrix repair.⁹  Similarly, in the central nervous system, clearance of myelin debris is essential for repair and remyelination.^10-12 

We show extensive debris accumulation in sterile lung injury in neutrophil-depleted mice and hypothesize that neutrophil clearance of DNA-containing debris is critical for repair. Identification of DNase-containing infiltrating neutrophils suggests that DNA clearance is specifically targeted. These neutrophils are able to degrade and internalize DNA in vitro, through an MyD88-dependent pathway. Interestingly, intratracheal DNase I supplementation enhanced lesion repair in neutropenic mice, suggesting a mechanism in lung repair.

Mice were housed in specific pathogen-free conditions at the Children’s Hospital of Philadelphia (CHOP) animal facility. Protocols were approved by the CHOP Institutional Animal Care and Use Committee. GCSF^−/−-C57BL/6J (002398) and wild-type (WT)-C57BL/6J mice (The Jackson Laboratory) were used. Sedated mice lungs were injured with 0.1 N HCl as described.^6,13 

Lungs were fixed under constant pressure with 10% formaldehyde, dehydrated with ethanol washes, embedded in paraffin, and sectioned at 5 μm. Hematoxylin-and-eosin staining was used as previously described.⁶ 

Following deparaffinization, samples were treated with antigen-unmasking solution (Vector Laboratories). Immunohistochemistry antibodies included: podoplanin clone 8.1.1 (University of Iowa), prosurfactant protein C (EMD Millipore), and Ki67 (Abcam). The TSA biotin staining kit (Perkin-Elmer) was used to prevent cross-reactivity.

Bronchoalveolar lavage (BAL) and cell counts were analyzed as previously described.¹⁴  BAL and serum testing were performed with enzyme-linked immunosorbent assay (ELISA) kits: mouse granulocyte colony-stimulating factor (G-CSF; R&D Systems) and immunoglobulin M (IgM; eBioscience). Protein quantitation was performed by BCA protein assay (ThermoScientific).

Single-cell suspensions were generated by chopping lung, enzymatic digestion in 480 U/mL collagenase I (Life Technologies), 50 U/mL dispase (Collaborative Research), and 0.33 U/mL DNase I (Roche) and passage through 100-μM and 40-μM cell strainers (BD Falcon) following 30-minute incubation. Red blood cells were lysed with ACK solution and cells were resuspended in fluorescence-activated cell sorting buffer. Fixation and permeabilization were conducted per protocol (FOXP3 kit; eBioscience). Antibodies for fluorescence-activated cell sorting (Accuri-C6) included: Ki67-allophycocyanin (BioLegend) and prosurfactant protein C (EMD Millipore) conjugated to DyLight488 (Abcam). Data were analyzed on CFlow-Plus (BD Biosciences).

DNA content was analyzed by Nanodrop and by Bioanalyzer.

One milliliter of BAL was albumin and IgG depleted (Sigma-Aldrich) per protocol. Following trichloroacetic acid precipitation, samples were run on 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis gel, Coomassie stained, excised in 3 segments, destained with 50% methanol/1.25% acetic acid, reduced with 5 mM dithiothreitol, alkylated with 20 mM iodoacetamide, washed with 20 mM ammonium bicarbonate, and dehydrated with acetonitrile. Gel was incubated overnight at 37°C with trypsin. Peptides were extracted with 0.3% trifluoroacetic acid followed by 50% acetonitrile. Tryptic digests were analyzed by liquid chromatography with tandem mass spectrometry (MS/MS) on a QExactive HF mass spectrometer (ThermoFisher). Peptides were separated by reversed-phase high-performance liquid chromatography and eluted at 300 nL/min. The mass spectrometer repetitively scanned m/z from 300 to 1400 (R = 240 000) followed by data-dependent MS/MS scans on the 20 most abundant ions.

Sequences were aligned with SEQUEST and were analyzed in Scaffold. Differential expressions were analyzed using Ingenuity Pathway Analysis (Qiagen). Proteins >2 standard deviations from identity were considered differentially expressed. Peptides were analyzed by sample and quantity. Significance was defined as peptides present in at least 2 samples.

Means were compared using the Student t test or 2-way analysis of variance. P values <.05 were considered significant.

Absence of G-CSF is associated with severe neutropenia.^15,16  BAL content of acid-induced lung injury in G-CSF^−/− mice had significantly reduced neutrophils,⁶  and tissue specimens 24 hours after injury had an 81.1% reduction in neutrophils (P = .036) compared with WT.⁶  BAL-resident macrophage numbers were not significantly altered in G-CSF^−/− mice compared with WT mice at baseline and after lung injury (supplemental Figure 1, available on the Blood Web site). Acid-injured lung tissue from G-CSF^−/− mice also had significant alveolar debris compared with WT mice (Figure 1A).

A proteomic analysis of albumin-depleted BAL samples from G-CSF^−/− and WT mice was reexamined for repair-associated factors.⁶  As seen in Figure 1B, BAL of neutropenic mice had increased histones and nucleus-associated proteins (supplemental Table 1). Nucleic acids present in the BAL were increased by 48.5% in G-CSF^−/− mice compared with WT (P < .001; Figure 1C). Size-distribution analysis showed a threefold increase (P = .0178) of larger DNA fragments in G-CSF^−/− BAL compared with WT (Figure 1D-E). Intra-alveolar Sytox staining was used to quantify extracellular DNA. As seen in Figure 1F, uninjured G-CSF^−/− and WT mice did not have increased levels of extracellular DNA. Twenty-four hours postinjury, however, there was a twofold increase (P < .01) in extracellular DNA debris in the G-CSF^−/− mice relative to WT.

To determine whether neutrophils degrade extracellular DNA-containing debris, RNA sequencing was performed in bone marrow (BM)–derived and BAL-isolated neutrophils 24 hours after injury. There was a differential increase in DNA-degrading enzyme gene expression (trex1, dnase1I3, wrn, and isg20) in BAL-derived neutrophils, suggesting that neutrophils acquire DNA-degradation mechanisms following BM release (Figure 2A). Differential increased expression of TREX1 (P = .0002), DNASE1L3 (P = .0045), and WRN (P < .0001) was confirmed by western blot and ELISA on BAL neutrophils compared with BM neutrophils (Figure 2B).

Sytox-labeled salmon sperm nuclei were incubated with BAL-derived neutrophils for 1 hour to determine whether neutrophils directly degrade extracellular DNA. As seen in Figure 2C, no small DNA fragments were detected in samples without neutrophils, but 10-kb fragments were seen in samples incubated with neutrophils. These data suggest that neutrophils degrade extracellular DNA.

BAL-derived neutrophils were incubated with or without Sytox-labeled salmon sperm nuclei and stained with Ly6G antibody. Sytox-labeled DNA was identified within neutrophils (Figure 2D) showing the capability to phagocytose extracellular DNA. Analysis by flow cytometry of BAL-derived neutrophils coincubated with Sytox-labeled salmon sperm showed significant increase of neutrophils containing Sytox-labeled DNA compared with control (P = .005; Figure 2E) This was verified by incubating BAL neutrophils with Sytox-labeled DNA using live imaging. Over a 4-hour period, neutrophils were observed binding and phagocytosing Sytox-labeled DNA, offering direct visualization of this process (supplemental Video 1; Figure 2F).

To determine how extracellular DNA is degraded, MyD88^−/− neutrophils were incubated with Sytox-labeled nuclei. MyD88^−/− neutrophils do not engulf nuclei (supplemental Figure 2). Furthermore, MyD88^−/− neutrophils failed to degrade extracellular DNA, indicating an MyD88-dependent process (Figure 2G). Expression profiling of BAL neutrophils compared with BM neutrophils during sterile injury (Figure 2H) shows upregulation of Toll-like receptor (TLR) and TLR-associated genes that can sense nucleic acids including: Lrrfip1, Ticam1, and TLR13. Thus, the TLR pathways represent a mechanism through which MyD88 may be necessary for reparative neutrophil function.

G-CSF^−/− mice were treated with 50 μg of intranasal DNaseI¹⁷  24 and 48 hours after acid instillation. As seen in Figure 2I, DNase normalized the repair deficit seen in G-CSF^−/− mice compared with controls, indicating the importance of DNA degradation in tissue repair.

Inflammation has long been recognized as a double-edged sword, in which tissue damage consequent to neutrophil recruitment is balanced against host defense.^18-20  We suggest that beneficial aspects of inflammation extend beyond pathogen clearance, and as others have previously hypothesized, debris disposal is a critical function of recruited neutrophils.²¹  We recently identified neutrophils as vital for alveolar epithelial regeneration.⁶  These findings parallel data that neutrophils promote regeneration in various organs²²  including liver⁷  and skin.²³  Similar to our data in lung tissue, neutrophils were also important for DNA clearance to promote revascularization in the model of liver injury.⁷  Our data along with previous studies showing importance of neutrophil-derived matrix metalloproteinase-9 in matrix processing after mechanical lung injury⁹  implicate the critical role neutrophils have in lung repair and/or regeneration.

We show an important new mechanism of this regenerative pathway by the observation that neutropenia is associated with increased nuclear debris in BAL fluid and in situ. Interestingly, RNA sequencing revealed that alveolar, but not BM, neutrophils express DNA-degradation genes. The kinetics of this neutrophil-acquired phenotype and the location in which it occurs warrant further investigation. Furthermore, alveolar neutrophils incubated with salmon sperm nuclei degrade DNA in an MyD88-dependent manner. Importantly, administration of DNase I proved that DNA-debris degradation improves alveolar repair.

These data elucidate new neutrophil functions during inflammation. Although our data suggest that G-CSF does not mediate macrophage recruitment, macrophages are important in phagocytosis and thus are important for further study of debris clearance and tissue repair. Although the pathologic effects of neutrophils have been described, our data showing neutrophil-dependent DNA-debris clearance and increased alveolar repair highlight benefits of neutrophil responses during inflammation. Of note, DNase I is currently in trials for acute respiratory distress syndrome. The mechanism through which it may improve pathology might in part be clearance of DNA debris. Thus, additional studies are warranted on this unique aspect of neutrophil function in disease pathology.

This work was supported by the following National Institutes of Health grants: 1R01AI099479 from the National Institute of Allergy and Infectious Diseases; 5R01HL10583402, 5T32HL07586, and T32HL715041 from the National Heart, Lung, and Blood Institute; KL2TR001879 from the National Center for Advancing Translational Sciences; and 1F32HL131079-01 from the National Heart, Lung, and Blood Institute.

Contribution: J.H.O., A.J.P., and G.S.W. designed the experiments and wrote the paper; A.J.P. performed acid instillation, collected BAL, and fixed lungs; J.H.O., A.J.P., and N.D. collected BAL, and performed cell counts and ELISAs; A.J.P. and G.S.W. did confocal microscopy and image editing; K.G. and M.P. assisted with neutrophil engulfment assays; K.R. assisted with experimental design and neutrophil enrichment/isolation; L.A.S., G.S.W., A.J.P., and S.H.S. generated the proteomic data; and A.J.P. and P.W. performed and analyzed the flow cytometry experiments and data.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: G. Scott Worthen, Abramson Research Center, Children’s Hospital of Philadelphia, 416H, 3615 Civic Center Blvd, Philadelphia, PA 19104; e-mail: worthen@email.chop.edu; and Joseph H. Oved, Abramson Research Center, Children's Hospital of Philadelphia 303E, 3615 Civic Center Blvd, Philadelphia, PA 19104; e-mail: ovedj@email.chop.edu.