Phagocyte Chemoattraction Is Induced through the Mcp-1–Ccr2 Axis during Efferocytosis

Apoptotic cells generated during development and for tissue homeostasis are swiftly and continuously removed by phagocytes via a process called efferocytosis. Efficient efferocytosis can be achieved via transcriptional modulation in phagocytes that have engulfed apoptotic cells. However, such modulation and its effect on efferocytosis are not completely understood. Here, we report that phagocytes are recruited to apoptotic cells being cleared through the Mcp-1–Ccr2 axis, which facilitates clearance of apoptotic cells. We identified Mcp-1 as a modulated transcript using a microarray and found that Mcp-1 secretion was augmented in phagocytes engulfing apoptotic cells. This augmented Mcp-1 secretion was impaired by blocking phagolysosomal degradation of apoptotic cells. Conditioned medium from wild type (WT) phagocytes promoted cell migration, but that from Mcp-1−/− phagocytes did not. In addition, blockade of Ccr2, the receptor for Mcp-1, abrogated cell migration to conditioned medium from phagocytes incubated with apoptotic cells. The intrinsic efferocytosis activity of Mcp-1−/− and Ccr2−/− phagocytes was unaltered, but clearance of apoptotic cells was less efficient in the peritoneum of Mcp-1−/− and Ccr2−/− mice than in that of WT mice because fewer Ccr2-positive phagocytes were recruited. Taken together, our findings demonstrate a mechanism by which not only apoptotic cells but also phagocytes induce chemoattraction to recruit phagocytes to sites where apoptotic cells are cleared for efficient efferocytosis.


Introduction
Apoptotic cells generated during development and for tissue homeostasis are swiftly and continuously removed through a process called efferocytosis [1,2]. Efficient efferocytosis is dependent on both phagocytes and apoptotic cells. Apoptotic cells do not passively wait to be cleared, but actively recruit phagocytes by secreting chemoattractants called 'find-me' signals such as nucleotides and modulate gene programs in the neighboring cells within a tissue by releasing metabolites as 'goodbye' signals [3][4][5]. In addition, apoptotic cells expose 'eat me' signals and simultaneously blunt 'do-not-eat-me' signals to be specifically engulfed [6,7]. The best-known eat-me signal is phosphatidylserine (PS), which is generally located in the inner leaflet of the plasma membrane in normal cells but translocates to the outer leaflet of the plasma membrane in apoptotic cells [8]. By contrast, CD47, a 'do-not-eat-me' signal, binds to SIRPα on phagocytes and disables phagocytosis [9][10][11]. However, CD47 clustering is disrupted in apoptotic cells and the interaction of CD47 with SIRPα is weakened [12]. Thus, these regulated eat-me and do-not-eat-me signals enable phagocytes to specifically engulf apoptotic cells. Phagocytes also express corresponding receptors for these signals and thus distinguish cells to be engulfed from those not to be engulfed [11]. For example, Tim-4 and Mertk, which are well-characterized engulfment receptors for apoptotic cells, directly recognize apoptotic cells by binding to PS on these cells and indirectly recognize apoptotic cells via bridging molecules such as Gas 6 [13][14][15][16]. Thus, phagocytes can sense cells to be cleared through interactions between PS and its receptors.
During efferocytosis, the transcriptional and translational programs of phagocytes are modulated for efficient clearance of apoptotic cells after they are engulfed and/or recognized. Transcription of genes involved in various processes is regulated in phagocytes, which increases the competence of phagocytes to recognize and engulf apoptotic cells promptly, and recruits neutrophils to promote digestion of ingested apoptotic cells [17][18][19][20]. Additionally, the levels of some proteins are translationally or post-translationally modulated in phagocytes for efficient efferocytosis after engulfment of apoptotic cells. For example, Ucp-2, Drp-1, and Orai1, which are involved in proton gradient dissipation in mitochondria and calcium flux, are upregulated in phagocytes engulfing apoptotic cells, causing phagocytes to engulf multiple apoptotic cells continuously [21][22][23].
The genes involved in efferocytosis and its signaling pathways have been identified over the past decades. Consequently, considerable progress has been made, especially with regard to molecules involved in recognition and ingestion of apoptotic cells. However, less is known about genes that are regulated in phagocytes during efferocytosis. Therefore, in this study, we performed a microarray to identify transcriptionally regulated genes and to evaluate their effects on efferocytosis. Mcp-1 was transcriptionally modulated during efferocytosis. It was upregulated in phagocytes incubated with apoptotic cells and this was dependent on degradation of apoptotic cells in phagocytes. Upregulation of Mcp-1 in phagocytes augmented secretion of Mcp-1, which promoted phagocyte chemoattraction. Depletion of Mcp-1 or Ccr2, the receptor for Mcp-1, in phagocytes did not alter their intrinsic efferocytosis activity; however, clearance of apoptotic cells was less efficient in Mcp-1 −/− and Ccr2 −/− mice due to a defect in phagocyte chemoattraction. Taken together, our findings imply that more phagocytes are recruited to apoptotic cells being engulfed through the Mcp-1-Ccr2 axis and that this facilitates clearance of apoptotic cells.

Mice
C57BL/6 were purchased from Taconic bioscience. Tim-4 −/− mice (RBRC04895) were obtained from Riken BioResource Center (Tsukuba, Japan). Mertk −/− and Ccr2 RFP mice were purchased from Jackson Laboratory, USA. Mcp-1 −/− mice were a generous gift from SungHoon Back at University of Ulsan. All mice were maintained and housed in equipped animal facility with temperature at 20-25 • C and humidity at 30-70%, under the same dark/light cycle (12:12). All experiments using mice were approved by the animal care and ethics committees of GIST in accordance with the national institutes of health guide for the care and use of laboratory animals.

Quantitative PCR
Total RNA from BMDMs incubated with or without apoptotic thymocytes was extracted using RNeasy Plus Mini Kit (74136, Qiagen, Dusseldorf, Germany). After that, cDNA was synthesized with SuperScript ® III First-Strand Synthesis System (18080-051, Invitrogen, Waltham, MA, USA). The levels of the indicated transcripts were measured using StepOnePlus real-time PCR system (Applied Biosystem, Waltham, MA, USA).

Preparation of Apoptotic Cells
Apoptotic Jurkat cells were prepared by irradiating Jurkat cells in DPBS with 100 mJ/cm 2 ultraviolet-C (UVC). The equal volume of complete RPMI medium was immediately added and the cells were further incubated at 37 • C for 2 h. Apoptotic thymocytes were prepared as previously described. Briefly, thymocytes derived from 4-to 6-week-old C57BL/6 mice were stained with 50 µM TAMRA-SE or 0.5 µM CellTracker for 30 min and washed with RPMI medium containing 10% serum. Then, apoptosis in thymocytes was induced using 50 µM dexamethasone at 37 • C for 4 h in a 5% CO 2 incubator.

ELISA
Peritoneal macrophages derived from the indicated mice or BMDMs were incubated with apoptotic Jurkat cells, PS beads, or PS liposomes in the absence or presence of Mfge8 D89E (10 µg/mL), cytochalasin D, or bafilomycin A1 for 8 h. Conditioned medium from the cells was collected, and Mcp-1 or Mcp-3 was measured using ELISA kit (900-K126 and 900-K123, PeproTech, Cranbury, NJ, USA) and ABTS ELISA buffer kit (900-K00, Pepro-Tech, Cranbury, NJ, USA) according to the manufacturer's protocol. Briefly, a 96-well plate was coated with an Mcp-1-or Mcp-3-capturing antibody for 12 h and blocked by blocking solution containing 1% BSA in PBS. One-hundred microliters of conditioned medium was added to the plate. After that, a detection antibody-avidin-HRP-and ABTS substrate were sequentially added and incubated for 2 h, 30 min, and 30 min, respectively. Fluorescence was measured using a microplate reader (VersaMax, Molecular Devices, San Jose, CA, USA).

Efferocytosis Assay
Efferocytosis assay was performed as previously described [24]. Basically, peritoneal macrophages derived from the indicated mice were incubated with TAMRA-labeled (C1171, Life Technologies, USA) apoptotic thymocytes for 15 min. A ratio of 1:10 (phagocytes to apoptotic cells) was used. To validate the effects of cytochalasin D, bafilomycin A1, or Mcp-1 on efferocytosis, the phagocytes were incubated with TAMRA-stained apoptotic thymocytes in the presence or absence of cytochalasin D for 8 h, bafilomycin A1 for 8 h, or Mcp-1 for 15 min. After that, the phagocytes were extensively washed with ice-cold PBS, trypsinized, and analyzed by flow cytometry (BD FACS Canto II, Franklin Lakes, NJ, USA). For in vivo efferocytosis assay, 1 × 10 7 of TAMRA-labeled apoptotic thymocytes in 300 µL PBS were intraperitoneally injected into 8-10-week-old mice. Two hours after injection, peritoneal exudates were collected and stained with an FITC-conjugated anti-F4/80 antibody and analyzed by flow cytometry. TAMRA positive and F4/80 negative cells were considered as unengulfed apoptotic cells which were counted using Sphero AccuCount particles (ACBP-50-10, Spherotech, Lake Forest, IL, USA).

Migration Assay
Transwell migration assays were performed by applying 100 µL of THP-1 cells at 2 × 10 6 /mL to the upper chamber as the conditioned medium in the lower chamber (600 µL) of 8 µm pore size transwells (35224, SPL, Pocheon, Korea) at 37 • C for 6 h. The number of migrated THP-1 cells was manually determined using hemocytometer and the percentage of migrated cells was calculated as the percentage of input cells. To disrupt the gradient of Mcp-1, recombinant Mcp-1 (50 ng/mL) was added to the upper chamber with THP-1 cells. 0.025 units/mL apyrase (M0398S, NEB, Ipswich, MA, USA) was used for 5 min at room temperature to remove nucleotides in conditioned medium, and an anti-CCR2 antibody (5 µg) was pre-incubated with THP-1 cells for 30 min.

Statistical Analysis
All data are shown as mean ± standard deviation. Each experiment was repeated at least three times independently. Data were analyzed using the GraphPad Prism 7 software (Prism 7, GraphPad Software, La Jolla, CA, USA). It was considered that differences were statistically significant when the p-values were less than 0.05.

Phagocytes Release Mcp-1 during Engulfment of Apoptotic Cells
To identify genes that are modulated during efferocytosis, we performed a microarray in which RNA from bone marrow-derived macrophages (BMDMs) incubated with apoptotic thymocytes was compared with that from BMDMs not incubated with apoptotic thymocytes. In total, 283 genes were upregulated, and 80 genes were downregulated more than 1.5-fold in BMDMs incubated with apoptotic cells compared with control BMDMs (Figure 1a). Subsequent analysis revealed that expression of genes linked to cell differentiation, cell migration, cell proliferation, and immune responses was modulated in phagocytes incubated with apoptotic cells (Figure 1b). We decided to study the functional relevance of genes belonging to the chemokine family, which is related to cell migration, for efferocytosis because their function in this process is unknown. Three genes-Mcp-1, Mcp-3, and Cxcl2were selected because they were upregulated to a relatively large extent in the microarray analysis and their transcripts were undetectable in thymocytes (Supplementary table). Quantitative PCR analysis confirmed the transcriptional changes of Mcp-1 and Mcp-3, but not of Cxcl2, detected in the microarray (Figure 1c). These two genes were also upregulated in peritoneal macrophages incubated with apoptotic cells. Moreover, they were upregulated more in peritoneal macrophages than in BMDMs ( Figure 1d). Notably, expression of Mcp-1 and Mcp-3 was only marginally detected in apoptotic thymocytes, suggesting that these two genes are upregulated specifically in phagocytes. We next tested whether the increased levels of these two transcripts lead to increases of

Phagolysosomal Acidification Is Necessary for Mcp-1 Secretion
Next, we investigated the mechanism by which secretion of Mcp-1 from phagocytes increases during efferocytosis. We first investigated whether a factor in the conditioned medium of apoptotic cells (apoptotic supernatants) stimulates secretion of Mcp-1. Mcp-1 secretion was not elevated by apoptotic supernatants but was robustly increased by apoptotic cells (Figures 2a and S1), suggesting that apoptotic cells are crucial for release of Mcp-1 by phagocytes. Thus, we next investigated whether binding of apoptotic cells to phagocytes is important for Mcp-1 secretion. To this end, binding of apoptotic cells to phagocytes was blocked by Mfge8 D89E , which binds to PS on apoptotic cells but not to integrins on phagocytes [25]. Treatment of apoptotic cells with Mfge8 D89E abolished not only efferocytosis, but also the elevation of Mcp-1 secretion by peritoneal macrophages (Figures 2b and S2). In addition, peritoneal macrophages derived from Tim-4 −/− and Mertk −/− mice secreted substantially less Mcp-1 than wild type (WT) controls when they were incubated with apoptotic cells (Figure 2c). These data imply that PS recognition is necessary for Mcp-1 secretion during efferocytosis. We next investigated whether PS recognition is sufficient for Mcp-1 secretion. To address this, we allowed phagocytes to bind to apoptotic cells, but not to internalize them, using cytochalasin D, an inhibitor of actin polymerization. Cytochalasin D reduced Mcp-1 secretion by peritoneal macrophages incubated with apoptotic cells in a dose-dependent manner, which was paralleled by a similar decrease in the percentage of phagocytes engulfing apoptotic cells (Figure 2d,e). This suggests that binding of apoptotic cells to phagocytes is insufficient to induce Mcp-1 secretion and that internalization of apoptotic cells or a subsequent step(s) is necessary for elevation of Mcp-1 secretion. The level of Mcp-1 secretion was commensurate with the percentage of phagocytes engulfing apoptotic cells, and PS recognition was necessary but not sufficient for elevation of Mcp-1 secretion. Therefore, degradation of apoptotic cells in phagocytes is likely crucial for elevation of Mcp-1 secretion. To investigate this, we blocked phagolysosomal degradation of apoptotic cells using bafilomycin A1, which inhibits acidification of phagolysosomes and thus degradation of phagolysosomal cargos. Bafilomycin A1 diminished Mcp-1 secretion by peritoneal macrophages incubated with apoptotic cells without affecting the efficiency of efferocytosis (Figure 2f,g), implying that degradation of apoptotic cells in phagocytes is required for elevation of Mcp-1 secretion during efferocytosis. This notion was supported by experiments using indigestible or simplified surrogates mimicking apoptotic cells, namely, PS beads and PS liposomes, respectively. Incubation with neither PS beads nor PS liposomes elevated Mcp-1 secretion by peritoneal macrophages (Figure 2h,i). In summary, these data suggest that phagolysosomal degradation of apoptotic cells is necessary for elevation of Mcp-1 secretion by phagocytes. In addition, they indicate that engulfment receptors for apoptotic cells are not directly linked to elevation of Mcp-1 secretion during efferocytosis and that Tim-4 −/− and Mertk −/− peritoneal macrophages secrete less Mcp-1 because they contain fewer engulfed apoptotic cells than control phagocytes.

Mcp-1 Promotes Phagocyte Chemoattraction
Mcp-1 is a member of the CC chemokine family that mediates chemotaxis and induces monocyte chemoattraction [26]. Thus, we hypothesized that, in addition to find-me signals, Mcp-1 secreted by phagocytes engulfing apoptotic cells also functions as a chemoattractant to recruit more phagocytes to sites where apoptotic cells are being cleared. To test this, we first investigated whether Mcp-1 secreted by engulfing phagocytes induces phagocyte migration using a transwell migration assay. Conditioned medium from phagocytes incubated with apoptotic cells enhanced migration of THP-1 monocytes, whereas conditioned medium from Mcp-1 −/− peritoneal macrophages incubated with apoptotic cells did not (Figure 3a). Addition of purified Mcp-1 to the upper chamber of the transwell abrogated migration of THP-1 cells to the lower chamber (Figure 3b), suggesting that Mcp-1 in conditioned medium induces phagocyte migration. Noticeably, the enhanced migration of phagocytes was not caused by nucleotides such as ATP released from apoptotic cells that function as chemoattractants because apyrase, a ATP-diphosphatase, had no effect on cell migration induced by conditioned medium, and the level of ATP was much lower in conditioned medium than in apoptotic supernatants, which is likely due to degradation of ATP by CD39, an ectonucleotidase that catalyzes hydrolysis of triphosphonucleosides to the monophosphonucleoside derivative, expressed in peritoneal macrophages (Figure 3c-e) [27,28]. Ccr2 is the receptor for Mcp-1 and is highly expressed in monocytes [29]. Thus, we next investigated whether the enhanced migration of THP-1 cells to conditioned medium is dependent on Ccr2. To this end, we disabled Ccr2 using an anti-Ccr2 antibody and performed a transwell migration assay. Blockade of Ccr2 with an anti-Ccr2 antibody appreciably diminished migration of THP-1 cells to conditioned medium ( Figure 3f). These data suggest that phagocyte migration is induced through the Mcp-1-Ccr2 axis during clearance of apoptotic cells.

Clearance of Apoptotic Cells Is Impaired in Mcp-1 −/− and Ccr2 −/− Mice
Next, to investigate the relevance of Mcp-1 to efferocytosis, we first evaluated its effects on efferocytosis. Efferocytosis by Mcp-1 −/− peritoneal macrophages was similar to that by WT peritoneal macrophages, as indicated by the percentage and MFI (mean fluorescence intensity, representing the relative number of apoptotic cells per phagocyte) of phagocytes that engulfed apoptotic cells (Figure 4a). Moreover, addition of purified Mcp-1 did not affect the efficiency of efferocytosis by peritoneal macrophages (Figure 4b). These data imply that Mcp-1 is unrelated to the intrinsic efferocytosis capability.
A crucial feature of efferocytosis is recruitment of phagocytes to sites where apoptotic cells are generated to facilitate their clearance, which is achieved by find-me signals released from apoptotic cells. Mcp-1 was released from phagocytes during efferocytosis and enhanced phagocyte chemoattraction, and therefore may play a similar role as find-me signals to facilitate recruitment of phagocytes and thus clearance of apoptotic cells. Thus, we next investigated whether the inability of Mcp-1 −/− phagocytes to induce cell migration is linked to a defect of apoptotic cell clearance in vivo, although the intrinsic efferocytosis ability of Mcp-1 −/− peritoneal macrophages remains unchanged. To this end, apoptotic cells were intraperitoneally injected and unengulfed apoptotic cells were measured. There were more unengulfed apoptotic cells in the Mcp-1 −/− peritoneum than in the WT peritoneum (Figures 4c and S3a). Interestingly, fewer Ccr2-positive cells were recruited to the Mcp-1 −/− peritoneum than to the WT peritoneum (Figure 4d), which correlated with the number of unengulfed apoptotic cells, implying that recruitment of fewer phagocytes to the Mcp-1 −/− peritoneum leads to an increase in unengulfed apoptotic cells. Furthermore, we validated the defect of apoptotic cell clearance in Mcp-1 −/− mice using Ccr2 −/− mice. Ccr2 deletion is indicated by RFP expression in cells expressing Ccr2. Using this feature, we evaluated recruitment of Ccr2-expressing cells to the peritoneum after intraperitoneal injection of apoptotic cells. Efferocytosis by Ccr2 −/− peritoneal macrophages was comparable with that by Ccr2 +/− peritoneal macrophages (Figure 4e). However, there were more unengulfed apoptotic cells in the Ccr2 −/− peritoneum than in the Ccr2 +/− peritoneum (Figures 4f and S3b). In addition, similar to Mcp-1 −/− mice, fewer RFP-positive cells were recruited to the Ccr2 −/− peritoneum than to the Ccr2 +/− peritoneum (Figure 4g), suggesting that phagocytes are recruited through the Mcp-1-Ccr2 axis during efferocytosis. Collectively, transcriptional programs are modulated in phagocytes engulfing apoptotic cells, and this facilitates prompt and continuous clearance of apoptotic cells. Our observations demonstrate a mechanism by which this is achieved. Specifically, phagocyte chemoattraction is induced through the Mcp-1-Ccr2 axis during efferocytosis, resulting in recruitment of more phagocytes to apoptotic cells being phagocytosed and efficient clearance of these cells.

Discussion
Apoptotic cells are not passive cargos engulfed by phagocytes, but actively play roles for efficient efferocytosis. They secrete various molecules, including nucleotides, S1P, and lactoferrin, which induce chemotaxis [3,[30][31][32]. These molecules recruit specific professional phagocytes such as monocytes and macrophages but repel undesired phagocytes such as neutrophils and help phagocytes find apoptotic cells promptly, which contributes to efficient clearance of apoptotic cells. In this study, we showed that secretion of Mcp-1 by phagocytes is induced during efferocytosis, and that this leads to recruitment of more phagocytes to sites where apoptotic cells are engulfed and thus enhances clearance of these cells. Thus, Mcp-1 functions similarly to find-me signals. Mcp-1 is released by phagocytes instead of apoptotic cells and can therefore be called a 'help-me' signal. Clearance of apoptotic cells was less efficient in Ccr2 −/− mice than in Mcp-1 −/− mice. This phenomenon may be due to the redundancy of chemokines released by phagocytes engulfing apoptotic cells. Secretion of Mcp-3, as well as Mcp-1, was elevated in phagocytes engulfing apoptotic cells (Figure 1f). In addition, phagocytes engulfing apoptotic cells may release other chemokines besides Mcp-1 and Mcp-3. Indeed, Ccr2 is a promiscuous receptor for a variety of chemokines including CC and CXC chemokines [33]. Thus, Ccr2 may affect phagocytic recruitment much more than a chemokine that binds to it, resulting in the more severe defect of apoptotic cell clearance in Ccr2 −/− mice. It remains to be elucidated whether phagocytes engulfing apoptotic cells secrete other chemokines that bind to Ccr2 as a receptor.
A recent study showed that apoptotic cell-derived metabolites metabolized in phagocytes upregulate Dbl by stabilizing Dbl mRNA, leading to engulfment of multiple apoptotic cells by phagocytes [34]. In this study, we showed that indigestible or simplified surrogates of apoptotic cells, namely, PS beads and PS liposomes, respectively, failed to elevate Mcp-1 secretion. In addition, phagolysosomal degradation of apoptotic cells was necessary for elevation of Mcp-1 secretion by phagocytes. These findings indicate that metabolites derived from apoptotic cells likely regulate the level of Mcp-1 and that upregulation of Mcp-1 during efferocytosis can be due not only to transcriptional regulation but also alteration of Mcp-1 stability.
In summary, our observations suggest that during efferocytosis, phagocytes release Mcp-1 to recruit more phagocytes to sites where apoptotic cells are being engulfed and thus facilitate clearance of these cells. Due to the correlation between defects in efferocytosis and autoimmune-related diseases, phagocyte chemoattraction through the Mcp-1-Ccr2 axis may have implications to develop therapeutics for these diseases.