Published January 4, 2017, doi:10.4049/jimmunol.1601495 The Journal of Immunology

Inactivation of Rab11a GTPase in Macrophages Facilitates of Apoptotic Neutrophils

Chunling Jiang,*,† Zheng Liu,*,‡ Rong Hu,x Lulong Bo,*,‡ Richard D. Minshall,*,{ Asrar B. Malik,{ and Guochang Hu*,{

The timely and efficient clearance of apoptotic neutrophils by macrophages (efferocytosis) is required for the resolution of inflam- mation and tissue repair, but the regulatory mechanisms remain unclear. In this study, we investigated the role of the small GTPase Ras-related in brain ()11a in regulating efferocytosis, and on this basis the resolution of inflammatory lung injury. We observed that apoptotic neutrophil feeding induced a rapid loss of Rab11a activity in bone marrow–derived macrophages and found that depletion of Rab11a in macrophages by small interfering RNA dramatically increased the phagocytosis of apoptotic neutrophils compared with control cells. Additionally, overexpression of wild-type Rab11a inhibited macrophage efferocytosis, whereas overexpression of dominant-negative Rab11a (Rab11a S25N) increased the clearance of apoptotic neutro- phils. Rab11a knockdown also increased the surface level of CD36 in macrophages, but it reduced cell surface expression of a disintegrin and metalloproteinase (ADAM) 17. Depletion of ADAM17 rescued the decreased surface CD36 expression found in macrophages overexpressing wild-type Rab11a. Also, blockade of CD36 abolished the augmented efferocytosis seen in Rab11a- depleted macrophages. In mice challenged with endotoxin, intratracheal instillation of Rab11a-depleted macrophages reduced neutrophil count in bronchoalveolar lavage fluid, increased the number of macrophages containing apoptotic neutrophils, and prevented inflammatory lung injury. Thus, Rab11a inactivation in macrophages as a result of apoptotic cell binding initiates phagocytosis of apoptotic neutrophils via the modulation of ADAM17-mediated CD36 cell surface expression. Our results raise the possibility that inhibition of Rab11a activity in macrophages is a promising strategy for activating the resolution of inflammatory lung injury. The Journal of Immunology, 2017, 198: 000–000.

cute inflammation is the survival response of a host flammatory PMNs and tissue debris to restore tissue homeostasis against invading pathogenic microorganisms, and it is (2). After tissue injury, recruited PMNs undergo apoptosis and are A characterized by rapid emigration of polymorphonuclear subsequently phagocytized by macrophages via a process termed neutrophils (PMNs) from blood vessels to sites of infection. efferocytosis. Efferocytosis by macrophages is critical for tissue

by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. The recruited PMNs are required for the neutralization and removal remodeling, modulation of immune responses, and resolution of of microbes but can also cause significant damage to host tissues inflammation (3). Clearance of apoptotic PMNs initiates a phe- when the acute inflammatory response is uncontrolled and ex- notype switch from M1 to M2 macrophages, which secrete anti- cessive (1). Thus, efficient resolution of inflammation depends on inflammatory cytokines, such as TGF-b and IL-10, that in turn not only cessation of PMN infiltration but also the recruitment and inhibits the release of inflammatory cytokines such as TNF-a and differentiation of macrophages, which remove the spent proin- IL-6 (4, 5). Efferocytosis by a set of macrophages also triggers the production of specialized proresolving mediators (e.g., resolvins and lipoxins) that signal the restoration of vascular integrity and *Department of Anesthesiology, University of Illinois College of Medicine, Chicago, regeneration of injured tissues (6). † https://www.jimmunol.org IL 60612; Department of Anesthesiology, West China Hospital, Sichuan University, Efferocytosis is a coordinately orchestrated process involving the Chengdu, Sichuan 610041, China; ‡Department of Anesthesiology and Intensive Care, Changhai Hospital, Shanghai 200433, China; xUndergraduate Program, recruitment of macrophages to inflammatory sites, recognition of Department of Biology, Washington University in St. Louis, St. Louis, MO { the marked apoptotic cells, engulfment, and ultimately removal of 63130; and Department of Pharmacology, University of Illinois College of apoptotic cells (7). An essential recognition or “find me” signal is Medicine, Chicago, IL 60612 the plasma membrane inner leaflet phospholipid phosphati- ORCID: 0000-0001-6787-1837 (L.B.). dylserine (PS) localized on the outer surface of apoptotic cells, Received for publication August 29, 2016. Accepted for publication December 6, Downloaded from 2016. and additional signals include the release of chemokines and lysophosphatidylserine by apoptotic cells, which induces macro- This work was supported by National Institutes of Health/National Heart, Lung, and Blood Institute Grant HL104092 (to G.H.) and by Natural Science Foundation of phage recruitment (5, 7). In response, the macrophage increases China Grant 81470259 (to G.H.). the expression of bridge molecules and cell surface receptors that Address correspondence and reprint requests to Dr. Guochang Hu, Department of act as tethers for the “catch me” step (3, 7). This then proceeds to Anesthesiology, University of Illinois, College of Medicine, 1740 West Taylor Street, the “eat me” step when the macrophage engulfs apoptotic cells Suite 3200W, MC 515, Chicago, IL 60612-7239. E-mail address: [email protected] through generation of large phagosomes (also known as effer- The online version of this article contains supplemental material. osomes) to degrade the ingested cells (3, 7). Recognition and Abbreviations used in this article: ADAM, a disintegrin and metalloproteinase; BAL, bronchoalveolar lavage; BMDM, bone marrow–derived macrophage; i.t., intratra- binding of apoptotic cells by receptors on the macrophage surface cheal(ly); MPO, myeloperoxidase; PMN, polymorphonuclear neutrophil; PS, phos- is thus a key event for macrophage recruitment and efferocytosis phatidylserine; Rab, Ras-related protein in brain; SE, succinimidyl ester; siRNA, (8, 9). Some efferocytosis receptors that directly recognize PS small interfering RNA; WT, wild-type. include brain-specific angiogenesis inhibitor-1, stabilin 1 and 2, Copyright Ó 2017 by The American Association of Immunologists, Inc. 0022-1767/17/$30.00 and members of the T cell Ig mucin domain (3, 10).

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1601495 2 Rab11a REGULATES EFFEROCYTOSIS

Also, other efferocytosis receptors can indirectly recognize PS Isolation and culture of mouse bone marrow–derived via interaction with specific bridging molecules in the serum or macrophages produced by macrophages themselves; these receptors include Murine bone marrow–derived macrophages (BMDMs) were isolated as integrins avb3/avb5, CD36, and Tyro3/Axl/Mer, each of which described (19). In brief, mice were sacrificed by rapid cervical dislo- recognizes the bridging molecules milk fat globule–epidermal cation. Bone marrow was flushed from femurs and tibias with 2–5 ml of growth factor 8 (lactaherin), thrombospondin-1, and protein S/ PBS (without Ca2+ and Mg2+). Collected bone marrow cell suspensions 3 growth arrest–specific 6, respectively (3, 10, 11). were centrifuged for 10 min at 500 g at room temperature. Cell pellets were then resuspended in macrophage complete medium Ras-related protein in brain (Rab) , the largest subfamily (DMEM/F12 with 10% FBS, 20% L929 cells–conditioned medium, of the of small G , serve as master regulators 10 mM L-glutamine, 100 IU/ml penicillin, and 100 mg/ml streptomy- of intracellular membrane trafficking, including cargo selection, cin). Cells (4 3 105) were then added to each sterile plastic petri dish vesicle budding, moving, tethering, docking, and targeting (12, 13). in 10 ml of macrophage complete medium and incubated at 37˚C and 5% CO2. On day 3, another 5 ml of macrophage complete medium was Based on distinct subfamily-specific sequence motifs, Rab pro- added to each dish. The adherent cells after 7 d in culture were .95% teins are grouped into ∼70 subfamilies. In particular, the Rab11 pure macrophages, as assessed by the expression of cell-surface subfamily members Rab11a, Rab11b, and Rab25 are shown to markers HLA-DR, CD11b, and CD206, and these BMDMs were used be central regulators of endocytic membrane recycling in both for all relevant experiments. polarized and nonpolarized cells (14). Rab11a is among the most Isolation of PMNs and induction of apoptosis prominent recycling components and plays important roles in broader intracellular domains, encompassing the trans- PMNs were isolated from mouse peripheral blood by density-gradient Golgi network, post-Golgi secretory vesicles, and the recycling Ficoll-Histopaque as described previously (20). Apoptosis of PMNs was induced by the exposure of cells to UV irradiation (254 nm, UVS-26, 6-W endosome (15–17). Interestingly, Rab11a has also been reported to 2 bulb, 0.02 J/s/cm ) for 15 min and then incubation for 3 h in a CO2 (5% [v/v]) be involved in macrophage phagocytosis, and Rab11a activation incubator at 37˚C. Approximately 91% of PMNs in this study were apoptotic induced FcgR-mediated internalization of IgG-opsonized particles asverifiedbyannexinV+ cells using flow cytometric analysis. (18). Thus, we addressed the possibility that Rab11a is essential Phagocytosis of IgG-opsonized latex beads for regulating phagocytosis of apoptotic PMNs by macrophages, thereby mediating resolution of inflammation. FITC-conjugated polystyrene latex beads covalently were opsonized We found that Rab11a inactivation enhanced macrophage with mouse IgG1 (Sigma-Aldrich) to study the FcgR-mediated phagocytosis. In brief, latex beads (diameter, 1.0 mm) were washed efferocytosis and consequently the resolution of lung inflam- three times and then incubated in HBSS containing 3 mg/ml mouse mation and injury via inhibiting translocation of a disintegrin and IgG1 at 4˚C on a rotator for 8 h to allow sufficient binding. After in- metalloproteinase (ADAM) 17 from the cytoplasm to the cell surface, cubation, the beads were resuspended and washed three times with cold 6 which prevented cleavage of surface CD36 on macrophages. Our HBSS to remove any unbound IgG1. Macrophages (10 per well, seeded in six-well plates) were incubated with IgG1-opsonized beads at a ratio of results showed that inhibition of Rab11a activation in macrophages 10 beads per cell at 37˚C for 20 min, washed with ice-cold HBSS, and may be a promising strategy for the resolution of inflammatory lung fixed in 4% paraformaldehyde. Phagocytosis was visualized by a DS-Fi2 injury through the effective clearance of dying PMNs. fluorescence microscope (Nikon), and images were captured using a Nikon DS-U3 camera. The phagocytic index was calculated as the Materials and Methods number of latex beads internalized by 100 macrophages counted in 10 by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. Reagents random fields (21). LPS (Escherichia coli O55:B5, L2880) was from Sigma-Aldrich. Anti- Macrophage efferocytosis assay in vitro + + Rab11a (no. 2413), Na /K ATPase (no. 3010), and GAPDH (no. 2118) 3 7 Abs were from Cell Signaling Technology. Anti-HRP–conjugated Abs Apoptotic PMNs (2–3 10 /ml) were labeled by incubation of cells with were obtained from Santa Cruz Biotechnology. Anti-CD36 (ab23680, CellTracker Red at 37˚C for 15 min, whereas mouse BMDMs or J774A.1 ab133625) Abs, anti-ADAM17 (ab57484, ab39163) Abs, and a myelo- macrophages were seeded in a 24-well plate and incubated with Cell- Tracker Green for 15 min. Macrophages were then overlaid with apoptotic peroxidase (MPO) assay kit were ordered from Abcam. Cytokine ELISA PMNs (1:10 ratio) for 2 h. After vigorous washing with PBS, the macro- kits were purchased from BioLegend. Rab11a and ADAM17 small inter- phages engulfing labeled apoptotic PMNs were analyzed using fluores- fering RNAs (siRNAs) were from Dharmacon. CellTracker Green and Red, pHrodo–succinimidyl ester (SE), Alexa Fluor 594– or 488–conjugated Abs, cence microscopy. Phagocytic activity was expressed as the phagocytic https://www.jimmunol.org Lipofectamine 2000, Lipofectamine RNAiMAX transfection reagent, a index, calculated as the percentage of macrophages containing at least one Mem-PER Plus membrane protein extraction kit, BSA, FBS, and heat- ingested PMN. inactivated FBS were from Invitrogen. The Rab11 activation assay kit was Assay of efferocytosis by alveolar macrophages in vivo from NewEast Biosciences. Radioimmunoprecipitation assay buffer was from Pierce Biotechnology. Bicinchoninic acid kits and sample buffer were Bronchoalveolar lavage (BAL) was performed by intratracheal (i.t.) in- from Bio-Rad Laboratories. DMEM/F12 and nonenzymatic cell dissociation jection of 1 ml of PBS followed by gentle aspiration. The lavage was solution were from CellGro. Wild-type (WT) Rab11a and dominant- repeated three times. The pooled BAL fluid was centrifuged, and cell pellets

Downloaded from negative mutant Rab11a S25N plasmids were a gift from Dr. Wei Guo were suspended in PBS. Cell suspensions were cytospun onto slides with a (University of Pennsylvania, Philadelphia, PA). cytocentrifuge (Shandon). Slides were stained with Diff-Quik dye (Dade Behring) and examined at magnifications of 320 and 340 by light mi- Mice croscopy. A minimum of 300 macrophages were scored for each sample. C57BL/6J male mice (25–30 g, 8–12 wk of age) obtained from The Jackson Phagocytosis was expressed as the percentage of macrophages ingesting Laboratory (Bar Harbor, ME) were inbred in microisolator cages under apoptotic cells and apoptotic bodies (22). specific pathogen-free conditions and fed with autoclaved food. All animal The phagocytosis of apoptotic PMNs by alveolar macrophages in vivo procedures were approved by the Institutional Animal Care and Use Com- was carried out by i.t. instillation of exogenous mouse apoptotic PMNs mittee prior to execution. All studies were performed under anesthesia using labeled with the pH-sensitive fluorescent dye pHrodo Red–SE (20 ng/ml) at either 1–3% isoflurane (inhalation) or ketamine (90 mg/kg, i.p.). room temperature for 30 min. pHrodo Red–SE-labeled apoptotic PMNs (1.0 3 107 in 100 ml) were i.t. instilled 2 d following CellTracker Green– 6 Cell culture labeled BMDM (2.0 3 10 ) transplantation. After 3 h of PMN instillation, BAL was performed. Cells in BAL fluid were washed, resuspended, and J774A.1 murine macrophages (American Type Culture Collection) were analyzed by flow cytometry (23). pHrodo-SE was chosen to allow the cultured in DMEM supplemented with 10% FBS at 37˚C in CO2 (5% [v/v]) detection of only ingested apoptotic PMNs, due to increased light emission incubator. The cells were washed three times in PBS to remove media and in the acidic environment of the phagosomes of phagocytes. Attached nonadherent cells prior to transfection. The cells between passages 4 and apoptotic PMNs were excluded because pHrodo-SE was nonfluorescent at 10 were used throughout this study. neutral pH (24). The Journal of Immunology 3

Rab11a and ADAM17 knockdown in macrophages subjected to PAGE (10–12%) and the separated proteins were transferred onto nitrocellulose membranes (Invitrogen). The membranes were blocked BMDMs grown in six-well plates were transfected at 50–70% confluence with 5% nonfat milk, probed with primary Abs overnight at 4˚C, and then with 50 nM Rab11a siRNA, ADAM17 siRNA, or nontargeting control incubated with HRP-conjugated secondary Abs (1:3000–1:5000) at room siRNA (Dharmacon) according to the protocol provided by the manufac- temperature for 1 h. The protein bands were detected with an Odyssey Fc turer. Successful knockdown of Rab11a or ADAM17 in macrophages was imager (LI-COR Biosciences). Relative band densities of the various confirmed by Western blot analysis. All experiments were performed 48 h proteins were measured from scanned films using ImageJ Software (Na- posttransfection. tional Institutes of Health). Immunoprecipitation analysis was conducted as described previously Rab11a cDNA transfection (17, 19). Cell lysates were precleared with 1 mg of normal rabbit IgG and BMDMs were transfected with pcDNA6 plasmid vector containing 20 ml of protein A+G agarose beads for 2 h at 4˚C and then incubated cDNAs of GFP-tagged WT Rab11a or dominant-negative mutant Rab11a overnight at 4˚C with primary Ab, followed by addition of 25 mlof S25N. Macrophages were incubated with a mixture of 4 mgofcDNAand protein A/G Plus–agarose and further incubated at 4˚C for 2 h. The 10 ml of Lipofectamine 2000 (Invitrogen) in each well of a six-well samples were electrophoresed on SDS-PAGE gels (10–12%) and sub- plate. At 48 h posttransfection, successful expression of exogenous sequently transferred to nitrocellulose membranes. The protein bands are Rab11a was confirmed by Western blot analysis of cell lysates. detected with digital imaging equipment and analyzed with ImageJ software. Flow cytometry Depletion of alveolar macrophages in mice For analysis of PMN apoptosis, PMN suspensions were centrifuged for 5 min at 400 3 g and washed twice with cold PBS (without Ca2+/Mg2+) Depletion of alveolar macrophages was performed as described previously containing 1% BSA. The pellets were resuspended in ice-cold PBS (19, 25). The clodronate liposome (Encapsula NanoSciences) was i.t. de- (without Ca2+/Mg2+) with 1% BSA at a final cell concentration of 2 3 livered to the anesthetized mice with ketamine (90 mg/kg, i.p.). Lavage- 107/ml. PMNs were incubated with FITC-conjugated annexin V (BD able alveolar macrophage count was reduced by 95% at 2 d following Biosciences) and propidium iodide for 45 min on ice, washed, and an- clodronate liposome nebulization. alyzed by flow cytometry (Becton Dickinson LSR II). Early apoptotic PMNs were defined as those annexin V+/propidium iodide2, and necrotic Murine model of LPS-induced acute lung injury and PMNs as propidium iodide+ nonpermeabilized cells. For analyzing sur- pulmonary macrophage transplantation face expression of CD36 and ADAM17, cell suspensions were centri- BMDMs were isolated and cultured as described above. BMDMs were fuged for 5 min at 200 3 g and washed twice with cold PBS (without Ca2 transfected with a control siRNA or Rab11a siRNA. Alveolar macro- +/Mg2+) with 1% BSA. Nonspecific Ab-binding sites were blocked by phage–depleted mice were challenged with LPS (3.5 mg/kg, i.t.). At d 3 incubation in PBS containing 10% goat serum and 0.3 M glycine for 30 after LPS challenge, PBS or BMDMs transfected with a scrambled min at room temperature. Cells were then incubated with anti-CD36, siRNA or Rab11a siRNA (2 3 106 cells, 200 ml of total volume each) anti-ADAM17 Ab, or isotype control Ab for another 30 min at room were given to alveolar macrophage–depleted mice via i.t. instillation as temperature. After washing, cells were stained with Alexa Fluor 594– described previously (26). Mice were sacrificed and lung injury was conjugated goat anti-mouse or anti-rabbit IgG secondary Ab fluorescein evaluated by analysis of BAL fluid, wet/dry lung weight ratio mea- for 30 min at room temperature, washed, and subjected to flow cyto- surement, and biochemical/immunological analysis of lung tissue at metric analysis. different time points. Rab11a-GTP pull-down assay Lung tissue MPO activity Rab11a pull-down assay was performed using Rab11 activation assay kit MPO activity in the lung as a marker of PMN sequestration was measured (NewEast Biosciences) according to the manufacturer’s instructions (17). by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. according to the manufacturer’s protocol (27). Briefly, lung tissue (∼50 mg) Cells were washed once with ice-cold PBS and lysed with cell lysis buffer. was homogenized in 1 ml of cold sample buffer and centrifuged at 16,000 Lysates were centrifuged at 13,000 3 g for 10 min, and the supernatant 3 g for 30 min at 4˚C. MPO activity was measured in 100 ml of super- was incubated with 1.0 mg of anti-active Rab11 Ab coupled with protein natants in duplicate by using development reagent at 450 nm and expressed A/G agarose. The reaction mixture was gently rocked for 1 h at 4˚C. The as change in absorbance per milligram of protein. beads were then collected by centrifugation at 5000 3 g, incubated at 4˚C for 1 min, and washed twice with lysis buffer. Beads were finally resus- Quantification of PMN infiltration in the lung pended in Laemmli sample buffer (30 ml) and boiled for immunoblotting with anti-Rab11a Ab. The total level of Rab11a was measured by Western At the end of the experiments, BAL was performed to enumerate the blot analysis of cell lysates. absolute number of PMNs in the intra-alveolar inflammatory exudates as described previously (27). In brief, lungs were lavaged by an i.t. injection Immunofluorescence https://www.jimmunol.org of 1 ml of PBS followed by gentle aspiration. The procedure was re- peated three times. After the collected BAL fluid was centrifuged, cell Immunofluorescence was performed as described previously with slight pellets were suspended in PBS. Cell suspension (300 ml) was cytocen- modifications (19). Briefly, cells plated on coverslips were fixed with 4% trifuged onto a glass slide at 300 rpm for 5 min with a cytocentrifuge paraformaldehyde for 10 min, washed three times with PBS for 10 min, (Shandon). Slides were stained with Diff-Quik dye (Dade Behring) and permeabilized with 0.2% Triton X-100 in PBS for 5 min, incubated with 3 3 anti-Rab11a, anti-CD36, or anti-ADAM17 (1:100) Ab at 4˚C overnight, examined at magnifications of 20 and 40 by light microscopy. The percentage of PMNs was calculated after counting 300 cells in randomly followed by incubation with Alexa Fluor 594– or Alexa Fluor 488–con- selected fields. jugated secondary Abs (1:1000) for 1 h. Cell nuclei were stained with Downloaded from DAPI (5 mg/ml) for 10 min. The cells were rinsed three times and finally Lung vascular injury and edema formation mounted on glass slides using ProLong antifade mounting medium (Mo- lecular Probes). Confocal images were acquired with a laser-scanning BAL protein concentration was measured using a protein assay kit (Bio-Rad confocal microscope (Zeiss LSM 510 META) using a mercury lamp and Laboratories) to evaluate permeability of the alveolar–capillary barriers UV filter set to detect DAPI (band pass 385–470 nm emission), 488 nm (27). At the end of experiments, lungs were weighed, dried, and reweighed. excitation laser line to Alexa Fluor 488 (band pass 505–550 nm emission), Wet-to-dry lung weight ratio was used as an index of lung water content and 568 nm excitation laser line to Alexa Fluor 594 (excitation/emission and edema. ∼590/617 nm). Optical sections had a thickness of 1 mm (pinhole set to 1 Airy unit). Lung histology and lung injury scoring Western blotting and immunoprecipitation At the end of animal experiments, the lungs were fixed by instillation of 4% paraformaldehyde (Sigma-Aldrich) in PBS and held for 2 h under 25 cm Cells were lysed by radioimmunoprecipitation assay buffer supplemented H2O pressure. The fixed lungs were washed with PBS and dehydrated in with 1 mM PMSF, 1 mM Na4VO3, protease, and phosphatase inhibitor 70% ethanol before paraffin embedding. Lung tissue sections (4.0 mm) mixture. A membrane protein extraction kit was used to extract membrane were stained with H&E. proteins from cells. The cell lysates were sonicated for 10 s and centri- The severity of lung injury was evaluated independently by two blinded fuged at 10,000 3 g for 10 min at 4˚C. Protein concentration was mea- investigators according to the histological semiquantitative scoring system sured using a protein assay kit (Bio-Rad Laboratories). The samples were as described previously (28). Each investigator scored all lung fields per 4 Rab11a REGULATES EFFEROCYTOSIS

slide at 320 magnification. Within each field, points were assigned on a beads between groups were evaluated by a two-way ANOVA. Data are scale from 1 to 5 for the following criteria: 1, normal; 2, focal (,50% lung expressed as mean 6 SEM. Differences were considered significant when section) interstitial congestion and inflammatory cell infiltration; 3, diffuse p , 0.05. (.50% lung section) interstitial congestion and inflammatory cell infil- tration; 4, focal (,50% lung section) consolidation and inflammatory cell infiltration; and 5, diffuse (.50% lung section) consolidation and in- Results flammatory cell infiltration. Injury scores were averaged for the two in- Depletion of Rab11a enhances macrophage phagocytosis of vestigators. The mean score was used for comparison between groups. apoptotic PMNs Measurements of cytokine levels Rab11 was shown to be expressed in multiple intracellular The levels of TNF-a, IL-6, TGF-b1, and IL-10 in BAL were measured compartments, including phagosomes of macrophages. In with commercial ELISA kits (BioLegend) according to the manufac- murine peritoneal macrophages and the RAW cell line, Rab11 is turer’s instructions. Each value represents the means of triplicate required for FcgR-mediated phagocytosis (18), raising the determinations. possibility that Rab11a activation facilitates efferocytosis and Statistical analysis is involved in subsequent resolution of inflammation. There- One-way ANOVA and a Student–Newman–Keuls test for post hoc com- fore, we first genetically manipulated Rab11a expression in parisons were used to test differences between control and experi- macrophages and examined its role in the phagocytosis of ap- mental groups. The differences in phagocytosis of IgG-opsonized latex optotic PMNs. As shown in Fig. 1A (left), Rab11a expression by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. https://www.jimmunol.org Downloaded from

FIGURE 1. Depletion of Rab11a enhances macrophage phagocytosis of apoptotic PMNs. J774A.1 macrophages (A) and BMDMs (B and C) were transfected with scrambled (Sc) or Rab11a siRNA (si). At 48 h after siRNA transfection, the transfected macrophages were labeled with CellTracker Green and incubated with CellTracker Red–labeled apoptotic PMNs (A and B) at a 1:10 ratio for 2 h. The cells were then mounted on a slide and analyzed by fluorescence microscopy. Scale bars, 10 mm. (A) Rab11a knockdown increased the phagocytosis of apoptotic PMNs in J774A.1 macrophages. Left, Rab11a protein expression in macrophages at 48 h after siRNA transfection; middle, representative fluorescent images showing macrophages engulfing apoptotic PMNs; right, phagocytic index based on the fluorescent images. (B) Rab11a knockdown increased the phagocytosis of apoptotic PMNs in BMDMs. Left, Rab11a protein expression in BMDMs at 48 h after siRNA transfection; middle, representative fluorescent images of macrophage engulfing apoptotic PMNs; right, phagocytic index based on the fluorescent images. (C) Rab11a knockdown decreased the phagocytosis of IgG-opsonized fluorescent latex beads in BMDMs. The transfected BMDMs were labeled with CellTracker Red and incubated with FITC-labeled, IgG-opsonized latex beads (green) for 20 min. The cells were then mounted on a slide and analyzed by fluorescence microscopy. Left, representative fluorescent images showing macrophages engulfing IgG-opsonized latex beads; right, phagocytic index based on the fluorescent images. Scale bar, 10 mm. Data were obtained from three inde- pendent cultures, each performed in triplicate. Each bar represents the mean 6 SEM (n = 3). Statistical significance was calculated by a Student t test (A and B) or by a two-way ANOVA followed by a Student t test (C). *p , 0.05 versus scrambled siRNA groups. The Journal of Immunology 5

FIGURE 2. Rab11a inactivation promotes clearance of apoptotic PMNs. (A) Time-dependent Rab11a inactivation in BMDMs in response to apoptotic PMN (AC) feeding. BMDMs were cocultured with apoptotic PMNs at a 1:10 ratio for ∼15 min. Lysates from macrophages were analyzed by an Rab11a activation assay kit. Left, pull-down experiment showing the content of Rab11-GTP in BMDMs stimulated with ACs; right, quantification of three in- dependent experiments is provided. (B) Endogenous and exogenous expression of Rab11a protein in BMDMs analyzed by Western blots. BMDMs were transfected with vector, dominant-negative (S25N), or WT Rab11a cDNA. (C) Representative fluorescent imaging of phagocytosis of ACs. The transfected by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. BMDMs were labeled with CellTracker Green. After incubation of macrophages with CellTracker Red–labeled apoptotic PMNs for 2 h, macrophage efferocytosis was visualized using fluorescence microscopy. Scale bar, 10 mm. (D) Phagocytic index. Data were obtained from three independent cultures, each performed in triplicate, and expressed as mean 6 SEM. The p value was calculated by one-way ANOVA followed by a Student t test. *p , 0.05 versus control (A) or vector group (D), †p , 0.05 versus 5-min group (A) or S25N group (D), ‡p , 0.05 versus 10 min group (A).

was reduced by .90% at 48 h posttransfection with a specific At 15 min, GTP-Rab11a was almost completely lost (Fig. 2A). To siRNA. Depletion of Rab11a in fact enhanced phagocytosis of elucidate the role of Rab11a activity in regulating phagocytosis of apoptotic PMNs by J774A.1 macrophages compared with apoptotic PMNs, BMDMs were transfected with GFP-tagged WT https://www.jimmunol.org control cells (Fig. 1A). Consistently, in BMDMs, silencing of Rab11a or dominant-negative Rab11a-S25N plasmid (Fig. 2B). Rab11a also induced engulfment of apoptotic PMNs (Fig. 1B). Macrophages expressing WT Rab11a showed a decreased en- To validate the role of Rab11a in the FcgR-mediated phagocytosis, gulfment of apoptotic PMNs as compared with the vector control FITC-conjugated polystyrene latex beads were opsonized with group. In contrast, overexpression of Rab11a-S25N dramatically mouse IgG1. We observed that Rab11a knockdown inhibited en- augmented phagocytosis of apoptotic PMNs (Fig. 2C, 2D). Thus, gulfment of IgG-opsonized beads compared with control siRNA. Rab11a inactivation promotes macrophage phagocytosis of apo-

Downloaded from Approximately 80% of control cells contained four or more beads ptotic PMNs. whereas .40% of Rab11a-depleted macrophages had no beads, Rab11a negatively regulates macrophage efferocytosis through and the overall distribution was shifted toward lower numbers downregulation of cell surface expression of CD36 (Fig. 1C). Thus, these results demonstrate that Rab11a is a neg- ative regulator of macrophage efferocytosis. Efferocytosis involves recognition, binding, ingestion, and di- gestion of apoptotic cells by phagocytes. A variety of redundant Rab11a inactivation is required for clearance of apoptotic receptors expressed on the surface of macrophages serve a key PMNs role in mediating the recognition and uptake of apoptotic cells Based on our finding that Rab11a inhibited the phagocytosis of (29). We found a time-dependent increase in the surface ex- apoptotic PMNs by macrophages, we next asked whether GTP/ pression of a class B scavenger receptor CD36 in BMDMs in GDP-dependent activity of Rab11a is required for the phago- response to apoptotic PMN exposure, which coincided with cytic clearance of apoptotic PMNs. Rab11a activity during effer- decreased release of soluble CD36 without significant changes ocytosis was determined by the relative amount of GTP-bound in its cytoplasmic fraction (Supplemental Fig. 1). The surface Rab11a (active). As shown in Fig. 2A, inclusion of apoptotic expression of CD36 was also found to increase in Rab11a- PMNs suppressed Rab11a activation in a time-dependent manner. depleted BMDMs by siRNA compared with that in control 6 Rab11a REGULATES EFFEROCYTOSIS

FIGURE 3. Rab11a regulates macrophage cell surface expression of CD36. (A) Rab11a knockdown increased surface expression of CD36. BMDMs were transfected with a scrambled (Sc) or Rab11a siRNA (si). At 48 h posttransfection, cells were stained with an anti-Rab11a Ab (green) and an anti- by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. CD36 Ab (red) targeting an extracellular epitope, and cells were left intact (not permeabilized). Left, representative confocal images showing surface expression of CD36 by confocal microscopy; right, quantitative fluorescence intensity of membrane CD36 expression. n =4.*p , 0.05 versus Sc si control group. Scale bar, 10 mm. (B) Effects of Rab11a activation on CD36 surface expression. BMDMs were transfected with vector, dominant- negative (S25N), or WT Rab11a cDNA. At 48 h posttransfection, cells were stained with an anti-CD36 Ab. Left, endogenous and exogenous expression of Rab11a protein in BMDMs analyzed by Western blots; middle, surface expression of CD36 was assessed by flow cytometry; right, quantitative data showing changes in mean fluorescence intensity (MFI) of CD36. n =4.*p , 0.05 versus vector group, †p , 0.05 versus S25N group. (C)CD36 blockade abolished Rab11a depletion–mediated increase in phagocytosis of apoptotic PMNs. At 48 h after transfection of BMDMs with Sc siRNA or Rab11a siRNA, cells were treated with CD36-blocking Ab (20 mg/ml) or isotype Ab for 2 h. Macrophages were then labeled with CellTracker Green and coincubated with CellTracker Red–labeled apoptotic PMNs for 2 h. Phagocytosis of apoptotic PMNs was visualized using fluorescence microscopy. https://www.jimmunol.org Left, Rab11a protein expression in macrophages at 48 h after siRNA transfection; middle, representative fluorescent images showing macrophages engulfing apoptotic PMNs; right, phagocytic index based on the fluorescent images. Scale bar, 10 mm. n = 4. Statistical significance was calculated by a Student t test (A) or by a one-way ANOVA followed by a Student t test (B and C). *p , 0.05 versus Sc siRNA plus isotype groups, †p , 0.05 versus corresponding isotype group.

siRNA-treated cells (Fig. 3A). Flow cytometry analysis Rab11a facilitates surface expression of ADAM17 in revealed that overexpression of dominant-negative Rab11a- macrophages Downloaded from S25N stimulated CD36 surface expression whereas macro- Ectodomain shedding of the extracellular portion of transmem- phages expressing WT Rab11a exhibited a decreased surface brane proteins at or near the cell surface is a recognized mech- level of CD36 protein as compared with the vector control anism for regulating surface expression and function of receptors group (Fig. 3B). To investigate the role of CD36 in the en- (31). ADAM17 is a sheddase responsible for ectodomain hanced apoptotic PMN uptake by Rab11a-depleted macro- shedding of a variety of cell surface proteins, including CD36. phages, the inhibitory effect by anti-CD36–blocking Ab was The cleavage of surface receptors CD36 by ADAM17 leads to evaluated (Fig. 3C). An anti-CD36–blocking Ab (no. ab23680; inhibition of efferocytosis (32). Because Rab11a is crucial in 20 mg/ml for 2 h) (30) not only abolished the augmented regulating intracellular and membrane trafficking (14–17), we efferocytosis in Rab11a-depleted BMDMs but also signifi- tested the concept that Rab11a mediates the transport of cantly dampened efferocytosis in scrambled siRNA-treated ADAM17 from cytosol to cell surface. We thus studied whether macrophages (Fig. 3C). Taken together, these data identify surfaceexpressionofADAM17inBMDMswasmodifiedin CD36 as the major receptor responsible for the enhanced response to apoptotic PMN feeding. ADAM17 protein expres- efferocytosis seen in Rab11a-depleted or -inactivated sion exhibited a time-dependent decrease on the cell surface and macrophages. a concomitant increase in cytoplasmic fraction in response to The Journal of Immunology 7

FIGURE 4. Rab11a regulates macrophage cell surface ADAM17 expression. (A) Rab11a colocalized with ADAM17 in BMDMs. BMDMs were immunostained with anti-Rab11a (green) and anti-ADAM17 (red) Abs. Representative confocal images show colocalization (yellow) of Rab11a and ADAM17. Scale bar, 10 mm. (B) Rab11a associated with ADAM17 in BMDMs. BMDMs were incubated with apoptotic PMNs (ACs) for 15 min, immunoprecipitated (IP) with Rab11a, and immunoblotted (IB) with an anti-ADAM17 Ab. (C) Effects of Rab11a knockdown on cell surface expression of ADAM17. BMDMs were transfected with a scrambled (Sc) siRNA or Rab11a siRNA (si). Left, Rab11a protein expression in macrophages at 48 h after

by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. siRNA transfection; middle, surface expression of ADAM17 was analyzed by flow cytometry; right, quantitative data showing changes in mean fluo- rescence intensity (MFI) of ADAM17. n = 4. Statistical significance was calculated by a one-way ANOVA followed by Student t test. *p , 0.05 versus isotype groups, †p , 0.05 versus Sc siRNA group. (D) Effects of Rab11a depletion on subcellular localization of ADAM17 in BMDMs. Top (left), im- munoblot showing the levels of ADAM17 expression on cell surface; top (right), immunoblot showing the levels of ADAM17 expression in cytosolic fraction; bottom, immunoblot showing the total levels of ADAM17 expression. (E) Protein quantification by densitometry. Bar graph shows the relative abundance of ADAM17 protein (normalized to that of loading controls) from three independent experiments. Statistical significance was calculated bya Student t test. *p , 0.05 versus corresponding Sc siRNA groups.

apoptotic PMN feeding (Supplemental Fig. 2), suggesting that To evaluate whether Rab11a inhibits surface expression of CD36 https://www.jimmunol.org ADAM17 is unable to move to the during effer- by modulation of ADAM17, macrophages were simultaneously ocytosis. We next addressed whether ADAM17 localized in the transfected with ADAM17 siRNA with or without WT Rab11a Rab11-positive compartments. We observed a high degree of plasmid (Fig. 5A). Flow cytometry analysis showed that trans- colocalization between Rab11a and ADAM17 in BMDMs by con- fection with WT Rab11a plasmid resulted in decreased surface focal immunofluorescence (Fig. 4A). Coimmunoprecipitation data CD36, whereas ADAM17 siRNA transfection completely rescued confirmed this observation and showed further that the association of this effect and augmented surface expression of CD36 (Fig. 5B,

Downloaded from Rab11a and ADAM17 was increased in response to apoptotic PMNs 5C). Taken together, these observations show the important role of (Fig. 4B). Cell surface expression of ADAM17 was analyzed using Rab11a in regulating the cell surface expression of CD36 through flow cytometry to measure the amount of immunoreactive ADAM17 ADAM17 signaling. present in the plasma membrane. We found that cell surface ADAM17 expression was markedly reduced in Rab11a knockdown Reduced expression of Rab11a in macrophages promotes BMDMs compared with scrambled siRNA-treated cells (Fig. 4C). clearance of apoptotic PMNs and accelerates resolution of To ascertain whether Rab11a controlled intracellular ADAM17 inflammation trafficking, we examined the effects of Rab11a depletion by a spe- Clearance of apoptotic PMNs favors resolution of inflammation cific siRNA on the cell surface and cytoplasmic expression of by downregulating the inflammatory phenotype of activated ADAM17 in BMDMs. As measured by immunoblotting, Rab11a macrophages (3, 29). Based on our in vitro findings that Rab11a knockdown decreased cell surface ADAM17 by ∼65% but increased inactivation in macrophages was involved in phagocytosis of cytoplasmic ADAM17 2-fold (Fig. 4D, 4E), demonstrating the cy- apoptotic PMNs (efferocytosis), we addressed whether Rab11a tosolic accumulation of ADAM17 in the absence of Rab11a. regulates resolution of lung inflammation through modulation Overall, these results indicate that Rab11a promotes ADAM17 of macrophage efferocytosis using an established in vivo model trafficking to the cell surface from the cytosol. (19, 26). By depleting alveolar macrophages in mice and then 8 Rab11a REGULATES EFFEROCYTOSIS

FIGURE 5. Rab11a deletion increase surface levels of CD36 through inhibiting ADAM17 trafficking. BMDMs were transfected with scrambled (Sc) or ADAM17 siRNA (si) and plasmids encoding vector or GFP-tagged WT Rab11a. (A) Protein expression of endogenous Rab11a and ADAM17 and exogenous expression of Rab11a protein in BMDMs were analyzed by Western blots. (B) Surface expression of CD36 was assessed by flow cytometry. (C) Quantitative data showing changes in mean fluorescence intensity (MFI) of CD36. n = 4. Statistical significance was cal- culated by a one-way ANOVA followed by a Student t test. *p , 0.05 versus control groups (vector plus Sc), †p , 0.05 versus corresponding Sc group (Rab11a WT).

transplanting BMDMs that were genetically modified to have and infiltration of inflammatory cells at day 5 following LPS

by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. differential Rab11a expression, we specifically determined the challenge, in comparison with untreated mice (Fig. 6F). These effect of Rab11a expression on the ability of macrophages to histopathological alterations in lung tissues were dramatically resolve LPS-induced lung inflammatory injury. As shown in ameliorated following administration of control siRNA-treated Fig. 6A, mice were first depleted of alveolar macrophages and BMDMs and further improved by administration of Rab11a- then challenged with LPS, followedbyi.t.deliveryofBMDMs treated BMDMs, as seen in the significantly reduced lung in- transfected with a scrambled siRNA or specific Rab11a siRNA jury score (Fig. 6G). Consistently, at day 2 after injection of (Supplemental Fig. 3A). Following i.t. instillation, the spatial BMDMs (day 5 following LPS challenge), the number of distribution of macrophages in alveolar macrophage–depleted macrophages containing apoptotic bodies or cells in BAL fluid lungs receiving BMDMs treated with a scrambled siRNA or was much higher in Rab11a siRNA–treated lungs than in https://www.jimmunol.org Rab11 siRNA is uniform and comparable (Supplemental Fig. control siRNA-treated lungs (Fig. 7A, 7B). To further demon- 3B). As expected, LPS caused a significant increase in PMN strate the role of Rab11a in the clearance of apoptotic PMNs counts in BAL fluid (Fig. 6B), lung MPO levels (Fig. 6C), in vivo, apoptotic mouse pHrodo Red (SE)–labeled PMNs were protein level in BAL fluid (Fig. 6D), and edema formation i.t. injected into mice at day 2 following delivery of BMDMs, (Fig. 6E) in control mice (without alveolar macrophage de- andBALfluidwasharvested3hlaterfordeterminationof pletion) and alveolar macrophage–depleted mice at days 1, 3, macrophage phagocytosis of PMNs. As shown in Fig. 7C and

Downloaded from and 5. Mice depleted of alveolar macrophages showed the 7D, the number of macrophages containing apoptotic PMNs delayed resolution of neutrophil infiltration (Fig. 6B, 6C), ex- in BAL was greater in mice receiving Rab11a siRNA-treated uded protein (Fig. 6D), and lung edema (Fig. 6E) at days 7 and BMDMs than in those receiving control siRNA-treated 10 following LPS challenge, whereas these delayed responses BMDMs. To elucidate the potential mechanisms responsible were reversed by administration of control siRNA-treated for the accelerated resolution of lung inflammation by Rab11a BMDMs (Fig. 6B–E). Importantly, alveolar macrophage– depletion, we examined the levels of key inflammatory medi- depleted mice receiving Rab11 siRNA-transfected BMDMs ators in BAL fluid from mice receiving control and Rab11a showed further decreased PMN counts in BAL fluid (Fig. 6B), siRNA-treated BMDMs. We found that the levels of proin- lung MPO (Fig. 6C), protein level in BAL fluid (Fig. 6D), and flammatory cytokines TNF-a (Fig. 8A) and IL-6 (Fig. 8B) in edema formation (Fig. 6E) compared with mice receiving the BAL fluid were elevated and reached peak levels at day 1 scrambled siRNA-treated BMDMs at days 5, 7, and 10 fol- (TNF-a) or 3 (IL-6) following LPS challenge and then gradu- lowing LPS challenge. H&E staining indicated that lung tissues ally decreased thereafter in control and alveolar macrophage– from control mice (without alveolar macrophage depletion) and depleted mice. At days 5, 7, and 10 following LPS challenge, alveolar macrophage–depleted mice were presented with severe alveolar macrophage–depleted mice receiving scrambled histological changes, including alveolar congestion, exudates, siRNA-treated BMDMs showed lower levels of TNF-a (Fig. The Journal of Immunology 9 by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd.

FIGURE 6. Role of macrophage Rab11a in resolution of lung inflammation and injury. (A) Experimental protocols of induction and time course of https://www.jimmunol.org resolution of lung inflammation after LPS challenge in WT mice. Following depletion of alveolar macrophages with clodronate liposome (CLOD), mice were i.t. instilled with LPS. BMDMs isolated from donor mice were cultured and transfected with a scramble siRNA (Sc) or Rab11a siRNA (si). After 48 h, the efficiency of transfection was evaluated by Western blot analysis. BMDMs treated with a Sc siRNA or Rab11a siRNA were i.t. injected into alveolar macrophage–depleted mice. Resolution of lung inflammatory injury was evaluated at days 1, 3, 5, 7, and 10 after LPS challenge as described in Materials and Methods. n = 6 animals per group per time point. (B) Neutrophil (PMN) counts in the BAL fluid. (C) PMN sequestration in lungs as assessed by MPO activity. (D) Pulmonary vascular protein permeability as determined by protein concentration of BAL fluid. (E) Pulmonary edema formation measured by F 3 G

Downloaded from wet-to-dry (W/D) lung weight ratio. ( ) Histological analysis of lung tissue by H&E staining (original magnification, 40). ( ) Histopathological mean lung injury scores from low-power (320) sections. Measurements were performed in triplicate for data analysis. Statistical significance was calculated by a two- (B–E) or one-way (G) ANOVA followed by a Tukey multiple comparison tests. *p , 0.05 versus corresponding CLOD plus LPS groups, †p , 0.05 versus corresponding Sc siRNA groups, ‡p , 0.05 versus corresponding LPS groups.

8A) and IL-6 (Fig. 8B), which further decreased in alveolar CD36 blockade abolishes enhanced resolution of lung macrophage–depleted mice receiving Rab11a-silenced BMDMs inflammation following transplantation of Rab11a-depleted (Fig. 8A, 8B). In marked contrast, alveolar macrophage–depleted macrophages mice receiving Rab11a-silenced BMDMs had a remarkably in- We sought to examine the contribution of CD36-mediated effer- creased level of anti-inflammatory cytokines TGF-b1 (Fig. 8C) and ocytosis in the enhanced resolution of lung inflammatory injury by IL-10 (Fig. 8D) compared with those receiving scrambled siRNA- Rab11a-depleted macrophages. CD36-blocking or isotype control transfected BMDMs at days 5, 7, and 10 following LPS challenge. Ab was i.v. injected at the same time as i.t. instillation of BMDMs These data clearly indicate that macrophage Rab11a regulates into the lung. CD36 blockade completely abrogated enhanced resolution of LPS-induced lung inflammation and injury by mod- resolution of acute inflammatory lung injury by Rab11a depletion, ulating efferocytosis. as evidenced by increased PMN counts (Fig. 9A) and protein 10 Rab11a REGULATES EFFEROCYTOSIS

FIGURE 7. Rab11 depletion increases phagocytosis of apoptotic bodies or PMNs by alveolar macrophages. Experimental protocols of induction and time course of resolution of lung inflammation after LPS challenge are shown as in Fig. 6A. Alveolar macrophage–depleted mice were i.t. administered with vehicle (PBS, control) or BMDMs transfected with scrambled (Sc) or Rab11a siRNA (si). (A) Representative photomicrographs of cytospin prep- arations of BAL cells 2 d after injection of BMDMs. Arrows indicate apoptotic bodies. Original magnification, 340. (B) Quantification of BAL fluid macrophages (Mf) containing apoptotic bodies. n = 6 per group. *p , 0.05 versus control group, †p , 0.05 versus Sc si group. (C) Effects of Rab11a depletion on phagocytosis of apoptotic PMNs in vivo following LPS challenge. pHrodo Red (SE)–labeled apoptotic PMNs (1.0 3 107) were i.t. instilled 2 d following CellTracker Green–labeled BMDM transplantation (2.0 3 106). At 3 h after instillation of apoptotic PMNs, BAL was performed and cells were washed and then analyzed by flow cytometry (pH to 10 with 0.003 M sodium carbonate). Representative flow cytometric dot plots demonstrating changes in the proportion of macrophages engulfing pHrodo-stained apoptotic PMNs are shown. (D) Phagocytic index was calculated by average percentage of macrophage-containing apoptotic PMNs. n = 6 per group. Measurements were performed in triplicate for data analysis. Statistical significance was cal- culated by a one-way ANOVA followed by a Student t test. *p , 0.05 versus control group, †p , 0.05 versus Sc si group. by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. (Fig. 9B) in BAL fluid, edema formation (Fig. 9C), and lung in- for promoting phagocytosis of apoptotic PMNs. This is evident by jury score (Fig. 9D) compared with isotype control Ab. Thus, the findings that upon exposure of macrophages to apoptotic PMNs, CD36-mediated phagocytosis of apoptotic cells may serve a key Rab11a activity was rapidly reduced, and that depletion or inac- role in accelerating resolution of lung inflammatory injury by tivation of Rab11a in macrophages strongly increased phagocytosis Rab11a depletion following LPS challenge (Fig. 10). of apoptotic PMNs. In a key experiment, we found that i.t. in- stilling of macrophages in which Rab11a was silenced induced Discussion the uptake of apoptotic PMNs. In contrast to the role of reduced In this study, we identified a novel role of macrophage-expressed Rab11a in promoting macrophage efferocytosis, we found that https://www.jimmunol.org Rab11a GTPase in negatively regulating efferocytosis, thereby Rab11a knockdown in macrophages inhibited phagocytosis of IgG- resolving acute lung inflammatory injury induced by sepsis. Our opsonized beads, consistent with previous findings (18). Macro- results showed that interaction of macrophages with apoptotic cells phage efferocytosis is regulated by signaling mechanisms different induced reduction of Rab11a activity in macrophages, which was from phagocytosis of cells opsonized with IgG (33–35). A study central for upregulating the surface expression of CD36 through showed that LPS and TNF-a inhibited the phagocytosis of apo- blocking Rab11a-mediated ADAM17 trafficking to the macro- ptotic PMNs by human macrophages while having no effect on the

Downloaded from phage cell membrane. We demonstrated that this pivotal step was phagocytosis of IgG-opsonized erythrocytes (33). Moreover, the important for the induction of phagocytosis of apoptotic PMNs. To efferocytic engulfment of apoptotic cells by macrophages sup- elucidate the in vivo functional relevance of these observations, we pressed proinflammatory signaling and activated proresolving showed that i.t. instillation of macrophages, in which Rab11a was signaling, whereas macrophage engulfment of IgG-opsonized genetically depleted, enhanced the clearance of apoptotic PMNs pathogens by FcgR activated the release of proinflammatory cy- and promoted resolution of lung inflammation and injury following tokines (36, 37). Although the phagocytosis of bacteria and apo- LPS challenge. These findings clearly indicate that Rab11a acts ptotic cells relies on the same cellular machinery, the signaling as a negative regulator of phagocytosis of apoptotic PMNs and mechanisms are quite different (38). Our results combined with efferocytosis-dependent inflammation resolution. Therefore, tar- other findings (18) suggest a specialized role of Rab11a in me- geting Rab11a activity in macrophages is a potential novel ther- diating phagocytosis of apoptotic PMNs as opposed to pathogens apeutic strategy to reduce inflammation and injury associated with bound to IgG. sepsis. The scavenger receptor CD36, a transmembrane glycoprotein A crucial finding was that Rab11a activity itself in macrophages member of the class B scavenger receptor family, is expressed on functioned as a brake on efferocytosis signaling; thus, Rab11 in- the surface of multiple cell types, including macrophages (39). activation during efferocytosis is an important mechanism critical CD36 is important for apoptotic cell recognition and uptake (40) The Journal of Immunology 11 by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd.

FIGURE 8. Macrophage Rab11a regulates the release of cytokines. Experimental protocols of induction and time course of resolution of lung inflammation after LPS challenge are shown as in Fig. 6A. (A–D)Levelsof TNF-a (A), IL-6 (B), TGF-b1(C), and IL-10 (D) in BAL fluid measured by ELISA. n = 6 per group. Measurements were performed in triplicate for data analysis. Statistical significance was calculated by a two-way ANOVA followed by a Tukey multiple comparison tests. *p , 0.05 versus corre- † https://www.jimmunol.org sponding CLOD plus LPS groups, p , 0.05 versus corresponding Sc siRNA groups, ‡p , 0.05 versus corresponding LPS groups. CLOD, clodronate liposome; Sc, scrambled; si, siRNA.

and serves as a critical regulator of phagocytosis in sepsis (41). We observed that Rab11a activation downregulated CD36 surface

Downloaded from expression, which in turn decreased efferocytosis and delayed FIGURE 9. CD36 blockade abolishes Rab11a depletion-induced reso- resolution of lung inflammation and injury. Confocal imaging and lution of lung inflammation. Experimental protocols of induction and time flow cytometry data showed higher surface expression of CD36 in course of resolution of lung inflammation after LPS challenge are shown as Rab11a-depleted or Rab11a-inactivated macrophages as compared in Fig. 6A. CD36-blocking or isotype control (Iso) Ab was i.v. injected with controls. Importantly, CD36 blockade abrogated the en- whereas BMDMs were i.t. admonished into the lung. At day 2 after in- A hanced efferocytosis response and accelerated resolution of lung jection of BMDMs, BAL and lung tissue were collected and tested. ( ) inflammation and injury seen following Rab11a depletion. Thus, PMNs in BAL fluid were enumerated to evaluate lung airspace inflam- mation. (B) Pulmonary vascular protein permeability as determined by relative CD36 expression on the macrophage cell surface has a protein concentration of BAL fluid. (C) Pulmonary edema formation requisite role in the mechanism of Rab11a-mediated phagocytosis measured by wet-to-dry (W/D) lung weight ratio. (D) Histopathological of apoptotic cells. mean lung injury scores from low-power (320) sections. n = 6 animals per An important question is how Rab11a regulates CD36 surface group. Measurements were performed in triplicate for data analysis. Sta- expression in macrophages. CD36 localizes in Rab11-containing tistical significance was calculated by a one-way ANOVA followed by a vesicle (42), suggesting that Rab11 may facilitate the trafficking Student t test. *p , 0.05 versus corresponding Iso group, †p , 0.05 versus of CD36 to the cell membrane through recycling . Sc plus Iso group. Iso, isotype; Sc, scrambled; si, siRNA. However, this does not appear to be the case in the present study 12 Rab11a REGULATES EFFEROCYTOSIS by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd.

FIGURE 10. Model of Rab11a in regulating phagocytosis of apoptotic PMNs.

because depletion of Rab11a increased CD36 surface expression demonstrate the crucial role of Rab11a activity in mediating https://www.jimmunol.org in macrophages. Surface expression of CD36 is highly regulated ADAM17 trafficking in macrophages. Finally, ADAM17 knock- by its cleavage and shedding (32). The transmembrane protease down completely reversed the decreased surface expression of ADAM17 proteolytically cleaves efferocytosis receptors, includ- CD36 caused by overexpression of WT Rab11a protein. Collec- ing CD36, tyrosine kinase Mer, and lectin-like oxidized low- tively, our results show the requisite role of Rab11a in down- density lipoprotein receptor 1 on the cell surface (43, 44). regulating the surface expression of CD36 in macrophages ADAM17 membrane-proximal cleavage of its substrates releases through facilitating the trafficking of ADAM17 to the macrophage

Downloaded from almost the entire extracellular domain of CD36 (32). Our study plasma membrane. supports the novel role of Rab11a in mediating ADAM17 trans- To address the in vivo relevance of our findings, we used an port to the macrophage plasma membrane, thereby regulating established murine model of LPS-induced lung inflammation and phagocytosis of apoptotic PMNs through cleavage of CD36 and injury (19). We demonstrated that Rab11a in macrophages func- reducing CD36 cell surface expression. We showed that in re- tioned as a negative regulator of efferocytosis-dependent resolu- sponse to apoptotic cell feeding, Rab11a activity was rapidly re- tion of lung inflammation and injury. Efferocytosis is a central duced, meanwhile, ADAM17 expression on the cell surface mechanism for restoration of tissue homeostasis and is critical for decreased in association with an increased cytoplasmic ADAM17 recovery from inflammation and injury (45). We observed that fraction. Furthermore, Rab11a was highly colocalized with airway instillation of Rab11a-silenced macrophages enhanced the ADAM17 in macrophages under basal conditions. Although the anti-inflammatory response and accelerated resolution of lung association between Rab11a and ADAM17 increased in response inflammation and injury. This proresolution response was associ- to apoptotic PMN exposure, ADAM17 trafficking to the cell ated with increased numbers of macrophages containing either membrane decreased due to the inhibition of Rab11a activity. In apoptotic bodies or PMNs in BAL fluid. Also, inhibition of Rab11-depleted macrophages, ADAM17 accumulated in the cy- efferocytosis by anti-CD36–blocking Ab reversed the protective tosol and was unable to reach the cell surface. These results effect of instilled Rab11a-depleted macrophages. Although our The Journal of Immunology 13

results support the role of CD36 downregulation in reducing 19. Yang, Z., D. Sun, Z. Yan, A. B. Reynolds, J. W. Christman, R. D. Minshall, A. B. Malik, Y. Zhang, and G. Hu. 2014. Differential role for p120-catenin in clearance of apoptotic PMNs and enhancing inflammation in re- regulation of TLR4 signaling in macrophages. J. Immunol. 193: 1931–1941. sponse to LPS (39), they describe, to our knowledge for the first 20. Hu, G., R. D. Ye, M. C. Dinauer, A. B. Malik, and R. D. Minshall. 2008. time, the underlying Rab11a- and ADAM17-regulated mechanism Neutrophil caveolin-1 expression contributes to mechanism of lung inflamma- tion and injury. Am. J. Physiol. Lung Cell. Mol. Physiol. 294: L178–L186. of fine-tuning the cell surface expression of CD36 and 21. Kang, J., K. H. Park, J. J. Kim, E. K. Jo, M. K. Han, and U. H. Kim. 2012. The efferocytosis-mediated resolution of inflammation. role of CD38 in Fcg receptor (FcgR)-mediated phagocytosis in murine macro- In summary, the present studies demonstrate the essential role of phages. J. Biol. Chem. 287: 14502–14514. 22. Moon, C., Y. J. Lee, H. J. Park, Y. H. Chong, and J. L. Kang. 2010. N-acetylcysteine Rab11a activation in macrophages as a “choke point” for regu- inhibits RhoA and promotes apoptotic cell clearance during intense lung inflam- lating phagocytosis of PMN and thereby resolving inflammatory mation. Am. J. Respir. Crit. Care Med. 181: 374–387. 23. Miksa, M., H. Komura, R. Wu, K. G. Shah, and P. Wang. 2009. A novel method lung injury induced by endotoxin. Apoptotic PMN feeding in- to determine the engulfment of apoptotic cells by macrophages using pHrodo duced Rab11a inactivation in macrophages, which in turn in- succinimidyl ester. J. Immunol. Methods 342: 71–77. creased surface levels of CD36 through blocking ADAM17 24. Esmann, L., C. Idel, A. Sarkar, L. Hellberg, M. Behnen, S. Mo¨ller, G. van Zandbergen, M. Klinger, J. Ko¨hl, U. Bussmeyer, et al. 2010. Phagocytosis of trafficking to the cell membrane and subsequent ADAM17- apoptotic cells by neutrophil granulocytes: diminished proinflammatory neu- mediated cleavage of CD36. Increased CD36 surface expression trophil functions in the presence of apoptotic cells. J. Immunol. 184: 391–400. in macrophages as result of Rab11a inactivation thus promotes 25. Wu, J., Z. Yan, D. E. Schwartz, J. Yu, A. B. Malik, and G. Hu. 2013. Activation of NLRP3 inflammasome in alveolar macrophages contributes to mechanical effective clearance of apoptotic PMNs and resolution of inflam- stretch-induced lung inflammation and injury. J. Immunol. 190: 3590–3599. mation and tissue injury (Fig. 10). Our findings provide new in- 26. Suzuki, T., P. Arumugam, T. Sakagami, N. Lachmann, C. Chalk, A. Sallese, sights into mechanisms of efferocytosis regulated by Rab11a, S. Abe, C. Trapnell, B. Carey, T. Moritz, et al. 2014. Pulmonary macrophage transplantation therapy. Nature 514: 450–454. showing that enhanced macrophage efferocytosis due to Rab11a 27. Wang, Y. L., A. B. Malik, Y. Sun, S. Hu, A. B. Reynolds, R. D. Minshall, and inactivation accelerates resolution of lung inflammation. The re- G. Hu. 2011. Innate immune function of the adherens junction protein p120- catenin in endothelial response to endotoxin. J. Immunol. 186: 3180–3187. sults point to a potential therapeutic approach whereby modulat- 28. D’Alessio, F. R., K. Tsushima, N. R. Aggarwal, J. R. Mock, Y. Eto, ing Rab11a activity in macrophages enhances phagocytosis of B. T. Garibaldi, D. C. Files, C. R. Avalos, J. V. Rodriguez, A. T. Waickman, et al. dying PMNs and resolves acute lung injury. 2012. Resolution of experimental lung injury by monocyte-derived inducible nitric oxide synthase. J. Immunol. 189: 2234–2245. 29. McCubbrey, A. L., and J. L. Curtis. 2013. Efferocytosis and lung disease. Chest Acknowledgments 143: 1750–1757. 30. Valente, A. J., T. Yoshida, R. A. Clark, P. Delafontaine, U. Siebenlist, and We thank Maricela Castellon (Department of Anesthesiology, University of B. Chandrasekar. 2013. Advanced oxidation protein products induce car- Illinois College of Medicine) for technical assistance. diomyocyte death via Nox2/Rac1/superoxide-dependent TRAF3IP2/JNK sig- naling. Free Radic. Biol. Med. 60: 125–135. 31. Liu, C., P. Xu, S. Lamouille, J. Xu, and R. Derynck. 2009. TACE-mediated Disclosures ectodomain shedding of the type I TGF-b receptor downregulates TGF-b sig- The authors have no financial conflicts of interest. naling. Mol. Cell 35: 26–36. 32. Driscoll, W. S., T. Vaisar, J. Tang, C. L. Wilson, and E. W. Raines. 2013. Macrophage ADAM17 deficiency augments CD36-dependent apoptotic cell uptake and the linked anti-inflammatory phenotype. Circ. Res. 113: 52–61. References 33. Michlewska, S., I. Dransfield, I. L. Megson, and A. G. Rossi. 2009. Macrophage 1. Rock, K. L., J. J. Lai, and H. Kono. 2011. Innate and adaptive immune responses phagocytosis of apoptotic neutrophils is critically regulated by the opposing to cell death. Immunol. Rev. 243: 191–205. actions of pro-inflammatory and anti-inflammatory agents: key role for TNF-a.

by guest on October 1, 2021. Copyright 2017 Pageant Media Ltd. 2. Ortega-Go´mez, A., M. Perretti, and O. Soehnlein. 2013. Resolution of inflam- FASEB J. 23: 844–854. mation: an integrated view. EMBO Mol. Med. 5: 661–674. 34. Rossi, A. G., J. C. McCutcheon, N. Roy, E. R. Chilvers, C. Haslett, and 3. Korns, D., S. C. Frasch, R. Fernandez-Boyanapalli, P. M. Henson, and I. Dransfield. 1998. Regulation of macrophage phagocytosis of apoptotic cells by D. L. Bratton. 2011. Modulation of macrophage efferocytosis in inflammation. cAMP. J. Immunol. 160: 3562–3568. Front. Immunol. 2: 57. 35. Michlewska, S., A. McColl, A. G. Rossi, I. L. Megson, and I. Dransfield. 2007. 4. Voll, R. E., M. Herrmann, E. A. Roth, C. Stach, J. R. Kalden, and I. Girkontaite. Clearance of dying cells and autoimmunity. Autoimmunity 40: 267–273. 1997. Immunosuppressive effects of apoptotic cells. Nature 390: 350–351. 36. Yamamoto, K., and R. B. Johnston, Jr. 1984. Dissociation of phagocytosis from 5. Fadok, V. A., D. L. Bratton, D. M. Rose, A. Pearson, R. A. Ezekewitz, and stimulation of the oxidative metabolic burst in macrophages. J. Exp. Med. 159: P. M. Henson. 2000. A receptor for phosphatidylserine-specific clearance of 405–416. apoptotic cells. Nature 405: 85–90. 37. Fadok, V. A., D. L. Bratton, A. Konowal, P. W. Freed, J. Y. Westcott, and 6. Basil, M. C., and B. D. Levy. 2016. Specialized pro-resolving mediators: endog- P. M. Henson. 1998. Macrophages that have ingested apoptotic cells in vitro enous regulators of infection and inflammation. Nat. Rev. Immunol. 16: 51–67. inhibit proinflammatory cytokine production through autocrine/paracrine https://www.jimmunol.org 7. Headland, S. E., and L. V. Norling. 2015. The resolution of inflammation: mechanisms involving TGF-beta, PGE2, and PAF. J. Clin. Invest. 101: 890–898. principles and challenges. Semin. Immunol. 27: 149–160. 38. Gruenberg, J., and F. R. Maxfield. 1995. Membrane transport in the endocytic 8. Ravichandran, K. S., and U. Lorenz. 2007. Engulfment of apoptotic cells: signals pathway. Curr. Opin. Cell Biol. 7: 552–563. for a good meal. Nat. Rev. Immunol. 7: 964–974. 39. Zamora, C., E. Canto´, J. C. Nieto, M. Angels Ortiz, C. Juarez, and S. Vidal. 9. Serhan, C. N., and J. Savill. 2005. Resolution of inflammation: the beginning 2012. Functional consequences of CD36 downregulation by TLR signals. Cy- programs the end. Nat. Immunol. 6: 1191–1197. tokine 60: 257–265. 10. Poon, I. K., C. D. Lucas, A. G. Rossi, and K. S. Ravichandran. 2014. Apoptotic cell 40. Greenberg, M. E., M. Sun, R. Zhang, M. Febbraio, R. Silverstein, and clearance: basic biology and therapeutic potential. Nat. Rev. Immunol. 14: 166–180. S. L. Hazen. 2006. Oxidized phosphatidylserine-CD36 interactions play an es- 11. Hochreiter-Hufford, A., and K. S. Ravichandran. 2013. Clearing the dead: ap- sential role in macrophage-dependent phagocytosis of apoptotic cells. J. Exp. Downloaded from optotic cell sensing, recognition, engulfment, and digestion. Cold Spring Harb. Med. 203: 2613–2625. Perspect. Biol. 5: a008748. 41. Leelahavanichkul, A., A. V. Bocharov, R. Kurlander, I. N. Baranova, 12. Schwartz, S. L., C. Cao, O. Pylypenko, A. Rak, and A. Wandinger-Ness. 2007. T. G. Vishnyakova, A. C. Souza, X. Hu, K. Doi, B. Vaisman, M. Amar, et al. Rab GTPases at a glance. J. Cell Sci. 120: 3905–3910. 2012. Class B scavenger receptor types I and II and CD36 targeting improves 13. Stenmark, H. 2009. Rab GTPases as coordinators of vesicle traffic. Nat. Rev. sepsis survival and acute outcomes in mice. J. Immunol. 188: 2749–2758. Mol. Cell Biol. 10: 513–525. 42. Kessler, A., E. Tomas, D. Immler, H. E. Meyer, A. Zorzano, and J. Eckel. 2000. 14. Bhuin, T., and J. K. Roy. 2015. Rab11 in disease progression. Int. J. Mol. Cell. Rab11 is associated with GLUT4-containing vesicles and redistributes in re- Med. 4: 1–8. sponse to insulin. Diabetologia 43: 1518–1527. 15. Welz, T., J. Wellbourne-Wood, and E. Kerkhoff. 2014. Orchestration of cell 43. Thorp, E., T. Vaisar, M. Subramanian, L. Mautner, C. Blobel, and I. Tabas. 2011. surface proteins by Rab11. Trends Cell Biol. 24: 407–415. Shedding of the Mer tyrosine kinase receptor is mediated by ADAM17 protein 16. Grant, B. D., and J. G. Donaldson. 2009. Pathways and mechanisms of endocytic through a pathway involving reactive oxygen species, protein kinase Cd, and p38 recycling. Nat. Rev. Mol. Cell Biol. 10: 597–608. mitogen-activated protein kinase (MAPK). J. Biol. Chem. 286: 33335–33344. 17. Yan, Z., Z. G. Wang, N. Segev, S. Hu, R. D. Minshall, R. O. Dull, M. Zhang, 44. Zhao, X. Q., M. W. Zhang, F. Wang, Y. X. Zhao, J. J. Li, X. P. Wang, P. L. Bu, A. B. Malik, and G. Hu. 2016. Rab11a mediates vascular endothelial-cadherin J. M. Yang, X. L. Liu, M. X. Zhang, et al. 2011. CRP enhances soluble LOX-1 recycling and controls endothelial barrier function. Arterioscler. Thromb. Vasc. release from macrophages by activating TNF-a converting . J. Lipid Res. Biol. 36: 339–349. 52: 923–933. 18. Cox, D., D. J. Lee, B. M. Dale, J. Calafat, and S. Greenberg. 2000. A Rab11- 45.Patel,B.V.,M.R.Wilson,andM.Takata. 2012. Resolution of acute lung containing rapidly recycling compartment in macrophages that promotes injury and inflammation: a translational mouse model. Eur. Respir. J. 39: phagocytosis. Proc. Natl. Acad. Sci. USA 97: 680–685. 1162–1170. Supplemental Figure 1

A membrane cytosolic ACs (min) 0 30 60 120 ACs (min) 0 30 60 120 CD36 CD36 Na+/K+ ATPase GAPDH

ACs (min) 0 30 60 120 CD36 (total lysate) CD36 (Supernatant) GAPDH

Membrane

B 1 0 0 Total lysate Supernatant *†

8 0 *†

6 0 †# *† * 4 0 CD36 *† *†# 2 0

(Relative intensity) (Relative 0 ACs (min) 0 30 60 120

Supplemental Figure 1. Effects of apoptotic PMN challenge on total and surface expression of CD36 in macrophages and soluble CD36 in the medium. BMDMs were coincubated with apoptotic PMNs at 1:10 ratio for indicated time and lysates separated into cytosolic and membrane fractions followed by Western blotting with CD36 antibody to determine CD36 localization. A. Representative Western blots for CD36 and corresponding loading controls. B. Protein quantification by densitometry. Bar graph showing the relative abundance of CD36 protein (normalized to corresponding loading controls). n = 3. Measurements were performed in triplicate for data analysis. Statistical significance was calculated by one-way ANOVA followed by Student’s t-test. *P < 0.05 vs. corresponding control (untreated) groups; † P＜0.05 compared with corresponding 30 min groups, # P＜0.05 vs. corresponding 60 min groups. ACs = apoptotic PMNs. Supplemental Figure 2

A membrane cytosolic ACs (min) 0 30 60 120 ACs (min) 0 30 60 120 ADAM17 ADAM17 Na+/K+ ATPase GAPDH

total lysate ACs (min) 0 30 60 120 ADAM17 GAPDH

Membrane B Cytosolic

1 0 0 Total lysate

† 8 0 *

* †

6 0 *†

4 0 * † ADAM17

2 0 (Relative intensity) (Relative

0 ACs(min) 0 30 60 120

Supplemental Figure 2. Effects of apoptotic PMN challenge on total, surface and cytosolic expression of ADAM17 in macrophages. BMDMs were coincubated with apoptotic PMNs at 1:10 ratio for indicated time and lysates separated into cytosolic and membrane fractions followed by Western blotting with ADAM17 antibody to determine ADAM17 localization. A. Representative Western blots for ADAM17 and corresponding loading controls. B. Protein quantification by densitometry. Bar graph showing the relative abundance of ADAM17 protein (normalized to corresponding loading controls). n = 3. Measurements were performed in triplicate for data analysis. Statistical significance was calculated by one-way ANOVA followed by Student’s t- test. *P < 0.05 vs. corresponding control (untreated) groups; † P＜0.05 compared with corresponding 30 min groups. ACs = apoptotic PMNs. Supplemental Figure 3

A siRNA Sc Rab11a Rab 11a

GAPDH

B Lung tissue F4/80

CLOD CLOD CLOD Control Sc si BMDM Rab11a si BMDM

Supplemental Figure 3. Effects of intratracheal instillation of BMDMs treated with scrambled siRNA or Rab11a siRNA on reconstitution of macrophages in the lung. BMDMs were transfected with scrambled (Sc) siRNA or Rab11a siRNA. A. effective knockdown of p120 expression was confirmed by Western blotting at 48 h posttransfection. B. Photomicrographs showing reconstitution of macrophages in the lung. BMDMs transfected with a scrambled siRNA or p120 siRNA (2 × 106 cells, 200 μl total volume each) were intratracheally instilled into alveolar macrophage-depleted mouse lungs. After 2 d, mice were sacrificed and lungs frozen sections (8 μm) were incubated with F4/80 antibody overnight and then with Alexa Fluor 488–conjugated secondary antibody. Photomicrographs were obtained with an Nikon Eclipse E400 microscope with a 20× objective. Scale bar = 20 μm.