The mannose 6‐phosphate/insulin‐like growth factor 2 receptor mediates plasminogen‐induced efferocytosis

Abstract The plasminogen system is harnessed in a wide variety of physiological processes, such as fibrinolysis, cell migration, or efferocytosis; and accordingly, it is essential upon inflammation, tissue remodeling, wound healing, and for homeostatic maintenance in general. Previously, we identified a plasminogen receptor in the mannose 6‐phosphate/insulin‐like growth factor 2 receptor (M6P/IGF2R, CD222). Here, we demonstrate by means of genetic knockdown, knockout, and rescue approaches combined with functional studies that M6P/IGF2R is up‐regulated on the surface of macrophages, recognizes plasminogen exposed on the surface of apoptotic cells, and mediates plasminogen‐induced efferocytosis. The level of uptake of plasminogen‐coated apoptotic cells inversely correlates with the TNF‐α production by phagocytes indicating tissue clearance without inflammation by this mechanism. Our results reveal an up‐to‐now undetermined function of M6P/IGF2R in clearance of apoptotic cells, which is crucial for tissue homeostasis.


INTRODUCTION
The plasminogen activation system, one of the major blood plasma proteolytic pathways, is best known for its role in dissolution of fibrin clots when they are no longer needed. However, the active serine protease plasmin, a central player of the system, participates in many other physiological as well as pathological processes. Activation of plasminogen (Plg), a zymogen of plasmin, must therefore be tightly balanced.
We have previously demonstrated that the mannose 6-phosphate/ insulin-like growth factor 2 receptor (M6P/IGF2R, CD222) binds and internalizes Plg. 8,9 M6P/IGF2R is an ubiquitously expressed type I transmembrane glycoprotein that is largely present in the Golgi apparatus and endosomes, and to a minor extent, also on the cell surface. 10 M6P/IGF2R has an essential role in the selective delivery of both newly synthesized enzymes and extracellular ligands bearing the recognition marker mannose 6-phosphate (M6P) to lysosomes. 11 In an M6Pindependent manner, it internalizes insulin-like growth factor 2 (IGF2), a potent mitogen. 12 We have pinpointed that M6P/IGF2R via its amino-terminal region binds 8 and internalizes Plg. 9 The crystal structure of this region, resolved by Olson and colleagues, has provided further insights into the shape of this Plg-binding site. 13 Here, we show that M6P/IGF2R expression increases during

Cells
The human monocytic cell line THP-1 and the human T cell line Jurkat E6.1, both from ATCC were cultured in RPMI-1640 medium (Invitrogen) supplemented with 100 U/ml penicillin, 100 g/ml streptomycin, 2 mmol/l L-glutamine and 10% heat-inactivated FCS (Sigma-Aldrich). When needed, apoptosis was induced by adding 200 ng/ml staurosporine (SSP). All the cells were grown in a humidified atmosphere at 37 • C and 5% CO 2 . All THP-1 cells, i.e., control-and Hill, NJ; 50 ng/ml) as previously described. 18,19 Following a 7-day differentiation, macrophages were rested in macrophage serum free medium (Invitrogen), supplemented with 2 mmol/l L-glutamine and 2% FCS for 2 days prior to efferocytosis assays.

Flow cytometry and cell surface binding assay
Cells were harvested, washed with PBS containing 1% BSA and 0.02% sodium azide, blocked with 2% beriglobin for 15 min on ice and afterward incubated for 30 min on ice with a specific mAb, either fluorescently labeled or unlabeled. For the latter, a second step staining was done with FITC-conjugated F(ab')2 anti-mouse IgG+IgM antibodies (An der Grub, Kaumberg, Austria). Prior to analysis, the cells were washed again and briefly incubated with DAPI. Apoptotic cells were discriminated by staining with DAPI and the Annexin V-Pacific Blue conjugate (BioLegend). Flow cytometry was performed with an LSR II flow cytometer (Becton Dickinson). Data acquisition was executed with the FACS DIVA software. Data analysis was accomplished with the FlowJo software (Treestar Inc., Ashland, OR). 20
ZFN-based gene knockout was performed as previously described. 15 Briefly, Cys2His2-based zinc fingers were created to specifically target the genomic sequence of M6P/IGF2R. The ZFN cassettes were cloned into pMLM290/pMLM292 and pMSM800/ pMLM802 (both from Addgene) and these constructs were elec-

RNA isolation and real-time PCR
Total RNA was extracted from the cells with TRIzol reagent (Invitrogen) supplemented with -mercaptoethanol for RNAse inhibition.
cDNA was synthesized from 500 ng total RNA using SuperScript III reverse transcriptase (Invitrogen). Quantitative PCR was carried out in duplicates using the TaqMan R Gene Expression Assay System (Invitrogen) in a CFX96 Touch Real Time PCR Detection System (Bio-Rad, Hercules, CA). To measure M6P/IGF2R expression, probe set Hs00974500_m1 was used, together with the probe set Hs03044281_g1 for the endogenous gene YWHAZ and analyzed by the 2 -ΔΔCT method. 22 Results are reported relative to the values for one of the monocyte samples, which were set to 1.

Efferocytosis assay
As phagocytic cells, we used primary monocyte-derived macrophages, and 7Pg to Plg, and control mAb AFP-01, all at 5 g/ml). For the mAbs experiments, the cells were co-treated with 2% beriglobin to block Fc receptors. Afterward, the phagocytes fed with the CFSE-labeled apoptotic cells were thoroughly washed from the non-uptaken apoptotic Jurkat T cells and then harvested by trypsinization. By this process, any bound apoptotic cell should be cleaved off from the phagocytes' surface. Flow cytometry was used to quantify the percentage of cells that phagocytosed apoptotic cells labeled with either CFSE or eFluor 670. The cell-free culture supernatant was then harvested and stored at −80 • C. TNF-was measured from the supernatants by the Luminex technology as done previously. 23

Confocal microscopy
Confocal microscopy was performed using the Leica TCS SP8 STED-3X system equipped with an inverted microscope Leica DMi8, using an HC PL APO CS2 100×/1.40 OIL objective. A diode laser with 405 nm excitation was used for UV excited dyes, all other fluorophores were excited using a white pulsed laser and the emission detection wavelengths were selected using a Leica SP detector. CFSE (green) was excited at 495 nm and emission was collected at 505-560 nm.
AF647 and eFluor 670 were excited at 647 nm and emission was collected at 660-710 nm. PE was excited at 560 nm and emission was collected at 570-620 nm. DAPI emission was collected from 415 to 470 nm. Z-stacks were taken in slices of 300 nm. The primary human macrophages were harvested after an above-described 7-day differentiation and cultured for 2 days directly on microscope slides in macrophage serum free medium, supplemented with 2 mmol/l L-glutamine and 2% FCS. In one setting, after the efferocytosis assay with the CFSE-or eFluor 670-labeled apoptotic Jurkat T cells the macrophages or fibroblasts were fixed with 4% formaldehyde, permeabilized with 0.1% saponin, blocked with 2% beriglobin, and stained

Statistical analysis
All experiments were performed at least 3 times in at least triplicates.
The data were expressed as mean values with SD. Statistical significance was evaluated by using a Student's t-test; a value of * p < 0.05, F I G U R E 1 M6P/IGF2R expression increases during monocyte differentiation to macrophages. (A) Cell-surface expression of M6P/IGF2R on isolated PBMC and MACS-enriched monocytes from healthy donors was evaluated with mAb MEM-238-AF647 and flow cytometry. In parallel, MOPC-21-AF647 was used as an isotype control mAb, displayed by the cut-off gates. The same analysis was performed with macrophages differentiated from MACS-sorted monocytes during a 7-day culture with recombinant human M-CSF (50 ng/ml) followed by 2 days resting in macrophage serum free medium. (B) Primary human monocytes and monocyte-derived macrophages from (A) were lysed and RNA was extracted. cDNA was synthesized from total RNA and gene expression was measured by real-time PCR as described in the Material and Methods section with TaqMan primer sets for human M6P/IGF2R and YWHAZ as endogenous control. The M6P/IGF2R mean expression values relative to that of monocytes ± SD from 3 donors is shown F I G U R E 2 Plg marks apoptotic cells. Jurkat T cells were stained on ice with Plg-AF647, Annexin V-Pacific blue and DAPI, and analyzed by flow cytometry to discriminate early (Annexin V + ) and late (Annexin V + / DAPI + ) apoptotic cells (AC) from viable (Annexin V − / DAPI − ) cells. Optionally, we co-incubated the cells with Plg-AF647 and TA (5 mmol/l) ** p < 0.005, or *** p < 0.0005 was considered to be significant or highly significant, respectively. Analysis and graphing were performed using GraphPad Prism 5 (GraphPad Software, San Diego, CA).

RESULTS
By staining PBMCs, we found very low surface expression of M6P/IGF2R. Only about 20% of PBMC displayed markedly higher M6P/IGF2R surface expression (Fig. 1A). We identified this M6P/IGF2R + subpopulation as CD14 + monocytes (Fig. 1A). The M6P/IGF2R surface expression increased when we differentiated the primary monocytes toward macrophages using M-CSF (Fig. 1A) and this increase was also recapitulated at the level of mRNA (Fig. 1B).
Based on this, we concluded that the overexpression of M6P/IGF2R on differentiated macrophages was regulated at the level of transcription.
We showed earlier that M6P/IGF2R binds and internalizes Plg and thereby regulates the proteolytic activity of this powerful enzyme. 8,9 Because Plg efficiently coats apoptotic cells, 5-7 we asked whether another function of M6P/IGF2R might be the Plg-mediated efferocytosis of apoptotic cells by macrophages. In our first experiment, we analyzed if Plg bound specifically to apoptotic cells also in our hands. By means of flow cytometric analysis allowing a discrimination of apoptotic from viable cells via the combined staining with Annexin V and DAPI, we observed a strong and specific binding of Alexa Fluor (AF)-488 conjugated Plg to apoptotic but not to viable Jurkat T cells (Fig. 2).
We observed similar results with Annexin V and propidium iodide costaining (data not shown). The binding of Plg to apoptotic cells was completely blocked in the presence of tranexamic acid (TA), a lysine analogue that blocks Plg binding to Plg receptors, suggesting that lysine-binding sites within kringle domains were implicated in the binding of Plg to apoptotic cells (Fig. 2).
Based on these observations, we examined the role of M6P/IGF2R in the uptake of Plg-coated apoptotic cells. We co-cultured M-CSF-differentiated human macrophages with CFSE-labeled apoptotic Jurkat T cells and evaluated efferocytosis by flow cytometry (Fig. 3).
Since the late apoptotic cells displayed more binding of Plg than the early apoptotic cells (Fig. 2), we induced apoptosis of Jurkat cells by  (Fig. 3B). We found the same pattern with the anti-Plg mAbs: 4Pg inhibited efferocytosis whereas 7Pg, recognizing a different epitope on Plg, did not (Fig. 3B). The mAb MEM-240 recognizes an epitope within the extracellular repeat domains 6 to 9 of M6P/IGF2R 14 and mAb 4Pg an epitope within the catalytic part of Plg. 24 We were able previously to coprecipitate the Plg-M6P/IGF2R complex from human serum with these two mAbs, 16 suggesting that they do not interfere with the Plg-M6P/IGF2R binding but are able, maybe due to steric hindrance, to inhibit the efferocytosis process.
We performed the efferocytosis experiments also with early apoptotic cells that we generated by treatment of Jurkat T cells with SSP for 9 h. Their uptake was 50% compared to late apoptotic cells (Fig. 3B,   right). However, the positive effect of Plg pretreatment on efferocytosis was again substantial (increase from 25 to 40% macrophages that engulfed Plg-labeled apoptotic cell bodies). Our data are in line with a previous report showing that Plg binding to apoptotic cells exhibits slightly delayed kinetics compared to the phosphatidylserine exposure on the apoptotic cell surface. 6 We analyzed the engulfment of apoptotic cells by macrophages also by confocal microscopy. We fixed and stained macrophages with fluorescently labeled nonblocking anti-M6P/IGF2R mAb MEM-238 and analyzed the staining in relation to CFSE-positive apoptotic cell bodies.
We detected zones of colocalization of engulfed apoptotic bodies and M6P/IGF2R (Fig. 4A). The 3D cell analysis also showed intracellular colocalization of M6P/IGF2R with CFSE-positive engulfed apoptotic cells ( Supplementary Fig. 1). Next, we analyzed living macrophages To test whether M6P/IGF2R was specifically responsible for the clearance of Plg-coated apoptotic cells, we employed the human monocytic THP-1 cell line. THP-1 cells display a relatively high surface expression of M6P/IGF2R 9 (Fig. 5) and are capable of phagocytosis when differentiated. 25 We modified THP-1 cells by means of genetic knockdown, knockout, and rescue approaches: First, we used control and M6P/IGF2R-silenced THP-1 cells that had been previously generated by RNA interference (RNAi) 9 ; second, via sequence-specific ZFNs, we generated a genetic M6P/IGF2R knockout in THP-1 cells that was, third, rescued by retroviral transduction of the recombinant human M6P/IGF2R construct (Fig. 5A-C), as described previously for Jurkat T cells. 15 We differentiated the various THP-1 cells to phagocytes by using PMA and subjected the cells to the efferocytosis assay with apoptotic CFSE-labeled Jurkat T cells. These cells displayed, in contrast to the primary macrophages, a 5 times lower capacity to engulf apoptotic cells-50% vs. 10% (Fig. 6A). M6P/IGF2R silencing at positions 6588 (shM6PR-1) and 4525 (shM6PR-2) led to a decrease in efferocytosis, as only ∼6% of M6P/IGF2R-silenced THP-1 cells contained CFSE-labeled apoptotic bodies (Fig. 6A-C). Moreover, the efferocytosis capacity of the M6P/IGF2R-silenced cells did not increase when the apoptotic cells were pretreated with Plg. In contrast, the control cells displayed Engulfment of apoptotic cells is known to reprogram macrophages from a proinflammatory to a pro-resolution state, characterized by suppression of pro-inflammatory cytokine production (e.g., TNF-, IL-12, and IL-1 ) and potentiation of secretion of anti-inflammatory cytokines IL-10, TGF-, and other pro-resolving mediators. [26][27][28] To ascertain the role of M6P/IGF2R in the aforementioned macrophage reprogramming, we measured TNF-levels from supernatants of the THP-1 phagocytes that had been fed (or not) with apoptotic Jurkat cells (Fig. 7B). Under control conditions, TNF-was potently produced by PMA-differentiated THP-1 phagocytes, in agreement with published data. 29 However, the cocultivation of THP-1 KO-CTR cells with apoptotic cells led to a robust, nearly 90% decrease in TNF-production. Upon M6P/IGF2R-knockout, the suppression of TNF-production was less pronounced, resulting in approximately 68% decrease, while re-expression of M6P/IGF2R in the M6P/IGF2R knockout cells was associated with more powerful suppression of TNF-(84% decrease; Fig. 7B, middle). The Plg-pretreatment of apoptotic cells led to even more pronounced suppression of TNF-production with a similar pattern observed with non-pretreated apoptotic cells (Fig. 7B, right). Finally, we verified our data also with fibroblasts, because Hall and colleagues showed that fibroblasts, although non-professional phagocytes, were also able to engulf apoptotic cells. 28,30 Furthermore, we reported previously that re-expression of human M6P-IGF2R in mouse fibroblasts derived from M6P-IGF2R knockout mice resulted in an increased Plg internalization. 9 Indeed, these fibroblasts expressing human M6P/IGF2R executed efferocytosis. Although the level of efferocytosis of Plg-pretreated apoptotic cells in these cells was low (6%), this experiment is a further proof that M6P-IGF2R is responsible for Plg-mediated efferocytosis, because the human M6P-IGF2R-negative mouse fibroblasts scored negligible (Fig. 8A). We analyzed the engulfment of apoptotic cells by fibroblasts also by confocal microscopy.
We fixed and stained the fibroblasts with the directly labeled anti-M6P/IGF2 mAb MEM-238 and analyzed the uptake of CFSE-positive apoptotic cells. Also in this case, the results showed the intracellular colocalization of M6P/IGF2R with the CFSE-positive fragments of apoptotic cells (Fig. 8B).

DISCUSSION
Apoptosis, a genetically programmed process of cell death, is an essential physiological process involved in tissue remodeling, hematopoiesis, and inflammation. When dysregulated, it crucially contributes to the development of many pathologies including tumor progression, 31 neurodegenerative disorders, 32 or serious inflammatory conditions. 33 Apoptosis involves many intracellular molecular pathways and culminates in the removal of apoptotic bodies by tissue resident macrophages or other surrounding cells in a process called efferocytosis. 33 Rapid efferocytosis guarantees that no inflammation is triggered by apoptotic cells and no damage of the tissue occurs, in contrast to necrotic cell death. 31 [53][54][55][56][57][58]. In contrast to them, M6P/IGF2R down-regulates plasmin generation on the cell surface. 8,9 Studies in the mouse have demonstrated that the Plg/plasmin system influences macrophages by transcriptional modulation of several genes to increase their efferocytosis activity 7 and by promoting their reprogramming from the proinflammatory (M1) to the resolution (M2) type. 57 However, these studies have not identified the Plg receptor on the phagocytes. Here, we show by means of loss-of-and gain-offunction genetic approaches that M6P/IGF2R facilitates not only the Plg-induced efferocytosis but also contributes to the efferocytosismediated macrophage switch by restraining production of the proinfammatory cytokine TNF-. The genetic knock-out of M6P/IGF2R in THP-1-derived macrophages abrogates the Plg-induced efferocytosis and increases the TNF-production indicating that M6P/IGF2R contributes significantly to the Plg-mediated clearance of apoptotic cells and macrophage reprogramming. We have previously shown that M6P/IGF2R internalizes Plg leading to its degradation, and thus serves as a negative regulator of Plg activation. 9 Thus, the simultaneous apoptotic clearance and the control of the Plg/plasmin system by M6P/IGF2R might ensure that apoptotic cells do not progress to secondary necrosis associated with generation of proteolytically processed neoantigens, which underlies the pathogenesis of chronic inflammatory autoimmune diseases. 27,59 When we incubated the apoptotic cells with Plg, the phagocytic capacity of the macrophages was enhanced, whereas TA reduced efferocytosis below the basal level. These results indicate that Plgmediated efferocytosis occurs also in the basic human Plg-free setup of the assay. This might be either Plg independent, or due to the bovine Plg present in serum that could have pre-coated apoptotic bodies, since the bovine Plg molecule can also bind to human M6P/IGF2R. 60 Further, the genetic knockout of M6P/IGF2R in phagocytes did not eliminate their efferocytic capacity completely suggesting that other receptor(s) were implicated in recognizing exposed phosphatidylserine and/or other "eat-me" signals on apoptotic cells, or alternatively, other Plg receptors expressed on macrophages might be involved.
In the majority of cell types, the protein transport routes of ubiquitously expressed M6P/IGF2R are mostly restricted to the Trans-Golgi-network and endosomal compartments. As few as 5-10% of the M6P/IGF2R molecules are typically displayed on the cell surface, 61 where they bind and internalize various extracellular ligands, such as IGF2, 62-64 heparanase, 65,66 leukemia inhibitory factor, 67 proliferin, 68,69 TGF-, 70 or Plg. 8,71 In contrast to Plg, we did not observe that other ligands of M6P/IGF2R, such as IGF2 or M6P, directly affected efferocytosis (data not shown). However, it is possible that through the modulation of surface expression of M6P/IGF2R by specific ligands also Plg-dependent efferocytosis is modulated.
Surface expression of M6P/IGF2R can change fast upon ligand binding. 72 On T lymphocytes the surface expression of M6P/IGF2R is very low, although it can be upregulated upon activation. 15,61 Similarly, low surface expression of M6P/IGF2R was detected on B lymphocytes. 73

DISCLOSURES
The authors declare no competing financial interests.