Cleavage of A1 by ADAM10 during Secondary Necrosis Generates a Monocytic ''Find-Me'' Signal

This information is current as Karin E. Blume, Szabolcs Soeroes, Hildegard Keppeler, of September 25, 2021. Stefan Stevanovic, Dorothee Kretschmer, Maren Rautenberg, Sebastian Wesselborg and Kirsten Lauber J Immunol 2012; 188:135-145; Prepublished online 23 November 2011;

doi: 10.4049/jimmunol.1004073 Downloaded from http://www.jimmunol.org/content/188/1/135

Supplementary http://www.jimmunol.org/content/suppl/2011/11/23/jimmunol.100407 Material 3.DC1 http://www.jimmunol.org/ References This article cites 46 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/188/1/135.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 25, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2011 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Cleavage of by ADAM10 during Secondary Necrosis Generates a Monocytic “Find-Me” Signal

Karin E. Blume,* Szabolcs Soeroes,* Hildegard Keppeler,* Stefan Stevanovic,† Dorothee Kretschmer,‡ Maren Rautenberg,‡ Sebastian Wesselborg,*,x and Kirsten Lauber*,{

Annexin A1 is an intracellular calcium/phospholipid-binding that is involved in membrane organization and the regulation of the immune system. It has been attributed an anti-inflammatory role at various control levels, and recently we could show that annexin A1 externalization during secondary necrosis provides an important fail-safe mechanism counteracting inflammatory responses when the timely clearance of apoptotic cells has failed. As such, annexin A1 promotes the engulfment of dying cells and dampens the postphagocytic production of proinflammatory cytokines. In our current follow-up study, we report that exposure of annexin A1 during secondary necrosis coincided with proteolytic processing within its unique N-terminal domain by ADAM10. Downloaded from Most importantly, we demonstrate that the released peptide and culture supernatants of secondary necrotic, annexin A1- externalizing cells induced chemoattraction of monocytes, which was clearly reduced in annexin A1- or ADAM10-knockdown cells. Thus, altogether our findings indicate that annexin A1 externalization and its proteolytic processing into a chemotactic peptide represent final events during apoptosis, which after the transition to secondary necrosis contribute to the recruitment of monocytes and the prevention of inflammation. The Journal of Immunology, 2012, 188: 135–145. http://www.jimmunol.org/

nnexin A1 (anx A1, also known as annexin I or lipocortin properties (6, 7). Based on its ability to bind phospholipids, in I) is a glucocorticoid-inducible member of the annexin particular phosphatidylserine (PS), anx A1 can couple membranes A superfamily, which comprises 13 members that have and has been implicated in different cellular processes, such as specific biological functions in membrane reorganization and the differentiation and proliferation, membrane trafficking, regulation of intracellular calcium levels (1, 2). All have and , and organization of the (8, 9). Ex- the common ability to bind calcium and phospholipids coordi- tracellularly, bound to the surfaces of monocytes and neutrophils, nated by four or more highly conserved domains, the so-called anx A1 has been described to inhibit their extravasation into annexin boxes (3, 4). The different functions of the annexins are inflamed tissues (10, 11). Anx A1 is also known to be involved in by guest on September 25, 2021 determined by their unique N-terminal domains. Anx A1 exhibits apoptosis. Thus, it has been reported to be externalized by to-date a 46-aa-long N terminus that in the absence of calcium is inserted unknown mechanisms and to bind to PS-rich plaques on the sur- into the protein core and upon calcium binding is extruded by faces of dying cells, where it functions as a bridging protein fa- a conformational change. This domain is known to undergo cilitating their phagocytic uptake (12, 13). Our previous studies posttranslational modifications, including phosphorylation, which have shown that externalization of anx A1 is a rather late apoptotic fine-tune calcium-dependent binding of anx A1 to membranes (5). event occurring after the transition from apoptosis to secondary Additionally, peptides derived from the anx A1 N terminus can be necrosis. In this context, externalized anx A1 potently inhibits the generated by different and exhibit anti-inflammatory release of proinflammatory cytokines by macrophages that have engulfed secondary necrotic cells (14). Secondary, postapoptotic necrosis occurs if the phagocytic re- *Department of Internal Medicine I, Eberhard Karls University, 72076 Tuebingen, moval of apoptotic cells fails, for instance when the phagocytic Germany; †Department of Immunology, Interfacultary Institute for Cell Biology, Eberhard Karls University, 72076 Tuebingen, Germany; ‡Cellular and Molecular process is directly impaired or when massive apoptosis overwhelms Microbiology Division, Interfaculty Institute of Microbiology and Infection Medi- the available scavenging capacity. Such situations have been de- cine, Eberhard Karls University, 72076 Tuebingen, Germany; xInstitute of Molecular { scribed in the context of acute inflammation associated with Medicine, Heinrich Heine University, 40255 Duesseldorf, Germany; and Division of Molecular Oncology, Department of Radiation Oncology, Ludwig Maximilians Uni- massive neutrophil accumulation, ischemia, drug-induced hepati- versity Munich, 81377 Munich, Germany tis, or bacterial infections (15). During the transition from apo- Received for publication December 16, 2010. Accepted for publication October 26, ptosis to secondary necrosis, the integrity of the plasma membrane 2011. becomes compromised leading to the leakage of potentially cy- This work was supported by grants from the Deutsche Forschungsgemeinschaft (We totoxic and antigenic intracellular contents into the surrounding 1801/2-3, SFB 685, and SFB 914). tissue. This in turn can promote and propagate the onset of in- Address correspondence and reprint requests to Dr. Kirsten Lauber, Molecular On- flammatory and autoimmune responses. Consequently, several auto- cology, Department of Radiation Oncology, Ludwig Maximilians University, March- ioninistrasse 15, D-81377 Munich, Germany. E-mail address: kirsten.lauber@med. immune diseases, such as systemic lupus erythematosus, and uni-muenchen.de chronic inflammatory conditions, such as rheumatoid arthritis, The online version of this article contains supplemental material. have been linked to a deficient and/or delayed clearance of apo- Abbreviations used in this article: anx 1, annexin 1; PARP, poly(ADP-ribose)poly- ptotic cells (15, 16). merase; PI, propidium iodide; PS, phosphatidylserine; qRT-PCR, quantitative real- In the current study, we have investigated the phenomenon of anx time polymerase chain reaction; siRNA, small interfering RNA. A1 externalization during secondary necrosis in greater depth. We Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00 observed that anx A1 externalization coincided with proteolytic www.jimmunol.org/cgi/doi/10.4049/jimmunol.1004073 136 AN ANX A1 PEPTIDE AS ATTRACTION SIGNAL OF 2˚ NECROTIC CELLS cleavage of anx A1 within its unique N-terminal domain. Phar- Production of culture supernatants of viable and secondary macological inhibition and RNA interference experiments identi- necrotic cells fied ADAM10 as the responsible in this scenario. Mapping Cells (2 3 106 per milliliter) were UV-irradiated with 10 mJ/cm2 and of the cleavage site by Edman degradation revealed a small peptide incubated in serum-free medium for the indicated times. Supernatants were comprising the N-terminal 7 aa that was generated and presumably collected by centrifugation at 4000 3 g for 10 min and stored at 270˚C. released from the secondary necrotic cell surface. Most interest- Flow cytometry ingly, this peptide and cell culture supernatants from secondary necrotic anx A1-externalizing cells induced monocyte migration. Flow cytometry was performed with a FACSCalibur (BD Biosciences) and In contrast, culture supernatants of secondary necrotic anx A1- CellQuest analysis software. Externalization of PS was measured by staining with -FITC/propidium iodide (PI) (Roche) according to silenced and ADAM10-silenced cells displayed a profound re- the manufacturer’s instructions. Cells positive for annexin A5–FITC but duction in their monocyte-attracting potential. Thus, our results negative for PI staining were considered apoptotic, whereas cells double- show that externalized anx A1 exerts multiple functions on diverse positive for annexin A5–FITC and PI staining were considered secondary levels during secondary necrosis not only by promoting the necrotic (14). DNA fragmentation as a marker of apoptosis and secondary necrosis was determined by the percentage of sub-G1 nuclei, and anx A1 phagocytic removal of dying cells and preventing proinflammatory externalization was assessed by indirect FACS immunostaining as de- cytokine production but also by recruiting monocytes to the sites of scribed previously (14, 18). For surface staining of ADAM10 and secondary necrosis. ADAM17 on native cells, cells were suspended in ice-cold staining buffer (RPMI 1640 medium plus 2% heat-inactivated human serum) supple- mented with FITC- or PE-conjugated mAb and incubated for 30 min on Materials and Methods ice. After two washing steps in staining buffer, cells were resuspended Cells and cell lines in staining buffer and analyzed by flow cytometry. For staining of per- Downloaded from meabilized cells (Supplemental Fig. 3), the Cytofix/Cytoperm kit (BD Jurkat, THP-1, MOLT-4, HuT78, HL60 (all from American Type Culture Biosciences) was used according to the manufacturer’s recommendations. Collection), and U937 cells (a gift from Ralph Hass) were cultured in RPMI 1640 medium supplemented with 10% heat-inactivated FCS, 100 U/ml SDS-PAGE, silver staining, and Western blot analysis penicillin, 0.1 mg/ml streptomycin, and 10 mM HEPES (all from Invi- trogen Life Technologies, Karlsruhe, Germany). Cells were grown at 37˚C SDS-PAGE and Western blot analyses were performed as described previ- in a 5% CO2 atmosphere and maintained in log phase. Human PBMCs ously (19) with mAbs against anx A1 (from BD Biosciences), poly(ADP- were prepared by Ficoll density gradient centrifugation. Subsequently, ribose)polymerase (PARP; Axxora), and vinculin (Sigma-Aldrich). Semi- http://www.jimmunol.org/ CD14+ monocytes were isolated by MACS cell separation according to the quantitative Western blot analyses were carried out by using IRDye-coupled manufacturer’s instructions (Miltenyi). PHA-stimulated human T lym- secondary Abs and the Licor Odyssey imaging platform. Silver staining of phoblasts (PHA blasts) were generated as described previously (14). SDS-PA gels was performed as described (20). Abs, small interfering RNA oligonucleotides, and other Electroporation with siRNA oligonucleotides reagents Electroporation with siRNA oligonucleotides was performed twice (at day Staurosporine was purchased from Roche Molecular Biochemicals 0 and day 3) with the Pulser II plus Capacity Extender II (Bio-Rad) and 0.4-cm gap cuvettes as described previously (14). In brief, 5 3 106 (Penzberg, Germany). The following mAbs were used for flow cytometry m and immunoblot analyses: anti-anx A1 (from BD Biosciences, Heidelberg, Jurkat cells were electroporated with 2 M siRNA oligonucleotides in 500 m m by guest on September 25, 2021 Germany), anti-vinculin (Sigma-Aldrich, Taufkirchen, Germany), anti–poly l OptiMEM medium (Invitrogen) by a single pulse (800 F, 200 V, time constant 20–30 ms). The cells were cultured for 3 d before electroporation (ADP-ribose)polymerase (Axxora, Lo¨rrach, Germany), anti-ADAM10 and anti-ADAM17 conjugated to FITC or PE (R&D Systems, Heidelberg, was repeated. All following experiments were conducted at day 6. Germany). In Fig. 4C, a polyclonal anti-anx A1 Ab (Invitrogen) was used. Quantitative real-time PCR analysis Chemically modified small interfering RNA (siRNA) oligonucleotides (stealth RNAi) were purchased from Invitrogen. The following sequences The detection of ADAM10 and ADAM17 mRNA levels was performed by were used: anx A1 nt893, 59-GGA UUA UGG UUU CCC GUU CUG quantitative real-time polymerase chain reaction (qRT-PCR) analysis with AAA U-39; anx A1 scramble, 59-GGA UAG GUU CUC CUG CUG UAU an ABI Prism 7000 Sequence Detection System (Applied Biosystems, UAA U-39; ADAM10 nt715, 59-CGA GAA GCU GUG AUU GCC CAG Foster City, CA) and qPCR Maxima Mastermix (Fermentas). The following AUA U-39; ADAM10 nt2064, 59-GAU CAU GCU AAU GGC UGG AUU probes and primers (Sigma-Aldrich) were used: ADAM10 forward, 59-CGG UAU U-39; ADAM10 scramble, 59-CGA UCG AAG UGC GUU GAC AAC ACG AGA AGC TGT GAT T-39; ADAM10 reverse, 59-TTA CGG CAA GUA U-39; ADAM17 nt90, 59-GAG ACU CGA GAA GCU UGA ATT CCG GAG AAG TCT GT-39; ADAM17 forward, 59-CGA GTA CAG UUC UUU G-39; ADAM17 nt969, 59-UGA UAU AGC UGA GGA AGC GAT GTA ATT GAA CGA TTT-39; ADAM17 reverse, 59-CCA ACG ATG AUC UAA A-39; ADAM17 scramble, 59-GAG AGC UGA AGG UUC TTG TCT GCT AAA AAC T-39; ALAS-1 forward, 59-TCC ACT GCA CUU AUC AUU G-39. Different protease inhibitors were used to charac- GCA GTA CAC TAC CA-39; ALAS-1 reverse, 59-ACG GAA GCT GTG terize the protease responsible for anx A1 cleavage: Elastase inhibitor TGC CAT CT-39; ALAS-1 probe, VIC-59-AAA GAA ACC CCT CCG (Bachem, Weil am Rhein, Germany), Pefabloc SC and protease-inhibitor GCC AGT GAG AA-39-TAMRA. Total RNA from 2 3 106 cells was mixture (Roche), aprotinin and o-phenanthroline (Sigma-Aldrich), GM extracted with the Nucleo-Spin RNA II kit (Macherey & Nagel, Dueren, 6001 and TAPI-2 (Merck Chemicals, Darmstadt, Germany), GI 254023X Germany). One microgram RNA was reverse transcribed with 200 U and GW 280264X (both from Andreas Ludwig, RWTH Aachen, Ger- RevertAid reverse transcriptase in the presence 50 mM random hexamers, many), and z-VAD-fmk (Bachem). Recombinant human MCP-1, SDF-1a, 400 mM desoxyribonucleoside triphosphates, and 1.6 U/ml RiboLock (all ADAM10, and ADAM17 ectodomains were obtained from R&D Systems. from Fermentas). Forty to eighty nanograms of the resulting cDNA was The N-terminal anx A1 peptide (aa 2–7, AMVSEF) and the corresponding applied to the following qRT-PCR analyses (20 ml final volume) with 200 reverse sequence (aa 7–2, FESVMA) were synthesized in an automated nM primers and 100 nM probe or 300 nM primers in the case of SYBR peptide synthesizer (433A; Applied Biosystems, Darmstadt, Germany) Green-based detection. For the study of siRNA knockdown effects, relative following the Fmoc/tBu strategy. Synthesis products were analyzed by quantification was performed using the standard curve method. The results HPLC (Young Lin Acme 9000; Young Lin, Anyang, Korea) and electrospray were normalized on the reference gene ALAS-1, and the cell population ionization-time of flight mass spectrometry (Q-ToF; Waters, Eschborn, transfected with the scramble control oligonucleotide was used as cali- Germany). brator. All experiments were performed in duplicates and are presented as mean values. Induction of apoptosis and secondary necrosis Transmigration assay Cells (1 3 106 or 1 3 105 per well) were cultured in 24-well or 96-well plates, respectively. Apoptosis was induced by addition of 2.5 mM staur- Transmigration assays were carried out with slight modifications as de- osporine, or cells were UV-irradiated with 10 mJ/cm2 in the UV Stra- scribed previously (21). Briefly, 1 3 105 calcein-labeled THP-1 cells per talinker 2400 (Stratagene) as described previously (17) and incubated for well were used in a final volume of 75 ml. Three hundred microliters the indicated times. Protease inhibitors were added 5 h after stimulation of cell culture supernatants or chemokines dissolved in serum-free RPMI with staurosporine or UV irradiation. 1640 medium was placed into the lower chamber of 5-mmpore The Journal of Immunology 137

Multiscreen-MIC chemotaxis chambers (Millipore, Billerica, MA). The and neutrophils as model systems, we now wanted to analyze filter was adjusted, the stained cell suspension was added on top, and the whether anx A1 externalization can also be observed in other cell assay was incubated for 120 min at 37˚C. Subsequently, the transmigrated types. To this end, different cell lines were stimulated with cells were collected by centrifugation and lysed in 100 ml lysis buffer (25 mM Tris-phosphate, pH 7.8, 2 mM EDTA, 10% glycerol, 1% Triton X- staurosporine to undergo apoptosis followed by secondary ne- 100). When migration of primary human monocytes was analyzed, mi- crosis, and anx A1 externalization was measured by surface im- gration assay was performed in medium containing 5% autologous serum munostaining and subsequent FACS analysis. Notably, exposure supplemented with the indicated stimuli for 4 h. Subsequently, cells at- of anx A1 was only observed in Jurkat cells, MOLT-4 cells, and tached to the upper side of the filter were removed with a cotton swab, and cells attached to the lower side of the filter were lysed. Green fluorescence primary human PHA lymphoblasts, but not in HuT78, HL60, was analyzed, and transmigration was calculated as percentage of total U937, or THP-1 cells, although staurosporine treatment potently cells deployed (mean values 6 SD from quadruplicates). induced apoptosis and secondary necrosis in all cell types used Chemotaxis assay in IBIDI m-slide chemotaxis chambers (Fig. 1A,1B). Western blot analysis was performed to ensure that anx A1 actually was expressed in all the cell types tested. Sur- Directional chemotaxis of primary human monocytes was analyzed by live prisingly, protein extracts of secondary necrotic cells that had m cell tracking in IBIDI -slide chemotaxis chambers according to the displayed anx A1 exposure (Fig. 1A) revealed an anti-anx A1 manufacturer’s instructions (IBIDI, Munich, Germany). Briefly, mono- cytes were seeded into the observation area of the chamber, and adherence reactive band of 36 kDa in addition to the expected full-length was allowed for 15 min. Nonadherent cells were carefully washed away, protein of 38 kDa (Fig. 1C). This band presumably constituted and the reservoirs of the chamber were filled with medium supplemented a proteolytic cleavage product of anx A1. Of note, although re- with 5% autologous serum. The chemotactic stimulus was added to the vealing a certain cell-type specificity, anx A1 externalization and upper reservoir, the slide was mounted on the heated stage of an Axi- oObserver inverted microscope (Zeiss, Goettingen, Germany), and after cleavage into the 36-kDa fragment were not only restricted to Downloaded from 30 min of equilibration, time-lapse video microscopy was performed for hematogenous cells but were also observed in non-hematogenous 4hat35 magnification. Subsequently, migration of 30 randomly picked cells, including MDA-MB 231 breast cancer and Colo 800 mel- cells was manually tracked with the ImageJ manual tracking plug-in, and anoma cells (data not shown). data were analyzed with the IBIDI chemotaxis and migration tool (IBIDI).

Proteolytic digestion of recombinant anx A1 http://www.jimmunol.org/ Recombinant forms of human anx A1 (aa 1–346, aa 47–346, and aa 1–46) were prokaryotically expressed as His6-tagged fusion and purified as described previously (14). Membrane proteins were prepared from Jurkat cells 18 h after UV irradiation (10 mJ/cm2). Cells were collected by centrifugation, washed in ice-cold PBS supplemented with 10 mM EDTA, and resuspended in 1 ml buffer A containing 20 mM HEPES, pH 7.5, 1.5 mM MgCl2, 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, 0.1 mM PMSF, 10 mg/ml leupeptin, and 10 mg/ml aprotinin. The cells were ho- mogenized in a Dounce homogenizer, and the homogenates were cleared 3 at 1000 g for 10 min at 4˚C to remove debris. The supernatant was by guest on September 25, 2021 centrifuged at 50,000 3 g for 2 h, and the resulting membrane pellet was resuspended in 25 mM Tris-HCl pH 9, 2.5 mM ZnCl2. Subsequently, the membrane protein extract of 3 3 107 cells was mixed with 1 mg re- combinant human anx A1, incubated for the indicated times at 37˚C, and subjected to SDS-PAGE with subsequent immunoblot analysis. Digestion of 1 mg recombinant anx A1 with 100 ng recombinant ADAM10 or ADAM17 ectodomain was performed accordingly. Edman degradation Forty micrograms anx A1 (aa 1–346) was mixed with 4 mg ADAM10 ectodomain (R&D Systems) in 25 mM Na-phosphate pH 9, 2.5 mM ZnCl2, 0.005% Brij35, and 5 mM CaCl2 and incubated for 4 h at 37˚C. Subse- quently, the reaction products were applied to N-terminal Edman degra- dation. Fifteen microliters of sample solution were applied to a trifluoroacetic acid-treated glass filter disk coated with 0.75 mg BioBrene Plus (Applied Biosystems). Filters were not precycled. Sequencing was carried out in a protein sequencer (494A Procise; Applied Biosystems) following the manufacturer’s protocols. PTH-amino acids were separated online by HPLC using a 2.1 3 250 mm C18 column, detected according to their UV absorbance at 269 nm, and quantitated with reference to a 10 pmol PTH- amino acid standard (all materials from Applied Biosystems). The cleav- age site was identified as the newly appearing N-terminal sequence in FIGURE 1. Anx A1 externalization during secondary necrosis is re- comparison with the N-terminal sequence of anx A1 that was incubated stricted to certain cell types and coincides with anx A1 cleavage. A, Anx with heat-inactivated (95˚C, 20 min) ADAM10 ectodomain prior to Edman A1 externalization is restricted to certain cell types. Cells were left un- degradation. treated or stimulated with 2.5 mM staurosporine (“Stauro”) for 12 h to undergo apoptosis followed by secondary necrosis. Subsequently, cells Results were stained with anti-anx A1/anti-mouse IgG–Cy2 and analyzed by flow

Anx A1 externalization during secondary necrosis is restricted cytometry. Mouse IgG1 served as isotype control. Mean values of dupli- to certain cell types and coincides with anx A1 cleavage cates are shown. B, Induction of DNA fragmentation in different cell types. Cells were stimulated as in A. After 24 h, cell death was measured by flow We have previously shown that anx A1 is externalized during cytometry and calculated as percentage of subdiploid nuclei (mean values secondary necrosis and provides a dual anti-inflammatory fail- of duplicates). C, A 36-kDa cleavage fragment of anx A1 can be observed safe mechanism by promoting dying cell removal and preventing in anx A1-externalizing cells. Cells were stimulated as in A, and protein proinflammatory cytokine production (14). Because this study extracts of 1 3 106 cells per lane were subjected to immunoblot analysis mainly used Jurkat cells, primary PHA lymphoblasts, monocytes, with anti-anx A1 Ab. Vinculin served as a loading control. 138 AN ANX A1 PEPTIDE AS ATTRACTION SIGNAL OF 2˚ NECROTIC CELLS

Anx A1 cleavage occurs on the outer cell surface during externalization and cleavage in UV-irradiated Jurkat cells. Again, secondary but not primary necrosis we observed that the cleavage of anx A1 into p36 occurred in We first analyzed the time course of anx A1 cleavage during parallel to the appearance of anx A1 on the surfaces of secondary staurosporine-mediated apoptosis and secondary necrosis. Notably, necrotic cells. Semiquantitative Western blot analyses revealed the appearance of the p36 fragment was observed as late as 6 h after that anx A1 p36 accounted for ∼20–30% of total anx A1 by 18 h apoptosis induction (Fig. 2A, left panel), a time point at which after UV irradiation or staurosporine treatment, respectively (Fig. a substantial proportion of cells had already lost plasma mem- 2C,2D). brane integrity and thus were secondary necrotic (Fig. 2B). In To address the issue that externalization and cleavage of anx A1 contrast, cleavage of PARP, a prototypical caspase-substrate, was are causally linked events, we studied the subcellular localization detected already 3 h after apoptosis-induction suggesting that of anx A1 p38 and p36. To this end, secondary necrotic cells were different mechanisms account for the proteolytic processing of treated with the calcium chelator EGTA to detach externalized, PS- these two proteins. Notably, we did not observe cleavage of anx bound anx A1 from the cell surface, and the collected supernatants A1 in heat-treated primary necrotic cells (Fig. 2A, right panel). and cell pellets were applied to SDS-PAGE with subsequent im- From these findings, it can be concluded that the proteolytic munoblot analysis. Notably, the p36 fragment was only found in the processing of anx A1 constitutes a late event in the time course of supernatant but not in the residual cell pellet of EGTA-treated cells apoptosis occurring after the transition to secondary necrosis. indicating that it was exclusively located on the outer cell surface Because we have recently shown that externalization of anx A1 (Fig. 2E). It should be noted that annexin A2, an annexin that we displays similar kinetics, it was conceivable that externalization have reported not to be externalized during secondary necrosis and cleavage of anx A1 are causally linked events (14). Extending (14), was not detected in the EGTA supernatant implying that Downloaded from our studies to a different death stimulus, we analyzed anx A1 EGTA treatment did not release intracellular annexins. Washing

FIGURE 2. Anx A1 cleavage occurs on the outer cell surface during secondary but not primary necrosis. http://www.jimmunol.org/ A, Anx A1 cleavage occurs during secondary necrosis. Jurkat cells were heat treated (60˚C) or stimulated with 2.5 mM staurosporine for the indicated times to un- dergo primary necrosis or apoptosis followed by sec- ondary necrosis. Subsequently, protein extracts of 1 3 106 cells per lane were separated by SDS-PAGE and subjected to immunoblot analysis using anti-anx A1, anti-PARP, and anti-vinculin Abs. B, Time-course analysis of PS exposure and loss of plasma membrane integrity. Jurkat cells were stimulated as in A. Subse- by guest on September 25, 2021 quently, exposure of PS and plasma membrane per- meability were verified by annexin A5 (anx A5)–FITC/ PI staining. Representative dot plots of triplicates are shown. C, Time-course analysis of anx A1 external- ization and loss of plasma membrane integrity after UV irradiation. Jurkat cells were UV-irradiated with 10 mJ/cm2 and incubated for the indicated times. Subse- quently, externalization of anx A1 and plasma mem- brane permeability were verified by anti-anx A1/PI staining. Representative dot plots of triplicates are shown. D, Semiquantitative Western blot analysis of anx A1 processing. Jurkat cells were UV-irradiated or treated with staurosporine as in C or A, respectively, and proteolytic processing of anx A1 into p36 was monitored by immunoblot analysis. Integrated pixel intensities were quantified, and the amount of anx A1 p36 is presented as percentage of total anx A1. E, The 36-kDa cleavage fragment of anx A1 is exclusively found on the outside of secondary necrotic cells. Jurkat cells were stimulated with 2.5 mM staurosporine and incubated for 18 h. Then, cells were washed twice with

TBS supplemented with 5 mM CaCl2 or 10 mM EGTA, respectively. Washing supernatants were pooled, and proteins were collected by TCA precipitation. Protein extracts of 1 3 106 washed cells per lane (“Cells”) together with TCA precipitated proteins from super- natants (“Sup”) were subjected to SDS-PAGE and subsequent immunoblot analysis with anti-anx A1. Annexin A2 (Anx A2), an annexin that is not exter- nalized during secondary necrosis, served as a control to show that EGTA treatment only detaches annexins that are bound to the outer cell surface. The Journal of Immunology 139

with CaCl2 was performed as a control to exclude the possibility and proteinase 3 (6, 7). However, to date no study has addressed that mechanical cell disruption and subsequent release of intra- the cleavage of anx A1 during secondary necrosis. To clarify cellular anx A1 accounted for its detection in the EGTA super- whether elastase or proteinase 3 are involved in anx A1 processing natant. In the presence of CaCl2, virtually no anx A1 was found in during secondary necrosis, we used different protease inhibitors: the supernatant. a specific elastase inhibitor and the serine protease inhibitors The exclusive extracellular localization of anx A1 p36, the ki- aprotinin and Pefabloc targeting both elastase and proteinase 3. netics of anx A1 cleavage, and our finding that only protein extracts However, neither of these protease inhibitors did significantly of anx A1-externalizing cells reveal the p36 fragment allow the interfere with anx A1 cleavage during secondary necrosis as conclusion that anx A1 is cleaved after it has been translocated to detected by immunoblot analysis (Fig. 3A). Hence, apparently an the outer cell surface during secondary necrosis. Consequently, we unrelated protease must be involved. wanted to identify the respective protease. Intending to narrow down the number of putative protease candidates, we incubated recombinant anx A1 with cell-free Anx A1 cleavage during secondary necrosis occurs supernatants or the membrane fraction of secondary necrotic downstream of ADAM10 cells and subjected the products to SDS-PAGE and immunoblot To date, several reports have described the proteolytic cleavage of analysis. Of note, only the coincubation with the membrane anx A1 by different proteases, including human leukocyte elastase fraction resulted in the generation of anx A1 p36 suggesting that Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 3. Anx A1 cleavage is downstream of ADAM10. A, Anx A1 cleavage cannot be blocked by inhibition of elastase or proteinase 3. Jurkat cells were stimulated with 2.5 mM staurosporine in the absence or presence of elastase inhibitor (Ela-I), aprotinin, or Pefabloc for 18 h. Subsequently, cells were lysed, and anx A1 processing was monitored by immunoblot analysis. B, Anx A1 cleavage is mediated by a membrane-resident protease. One microgram of purified recombinant human anx A1 was incubated with the culture supernatant of 4 3 106 secondary necrotic Jurkat cells per milliliter or the membrane fraction of 3 3 107 secondary necrotic Jurkat cells at 37˚C for the indicated times, and anx A1 cleavage was examined by immunoblot analysis. C, Anx A1 cleavage is blocked by o-phenanthroline. Jurkat cells were stimulated as in A in the absence or presence of the metalloproteinase inhibitor o-phenanthroline. Anx A1 processing was detected by immunoblot analysis. Vinculin was used as a loading control. D, Addition of the broad-range matrix metalloproteinase inhibitor GM 6001 blocks anx A1 cleavage. Jurkat cells were stimulated as in A in the presence of 0–100 mM GM 6001. Afterwards, anx A1 cleavage was monitored by immunoblot analysis. PARP was used as a loading and apoptosis/secondary necrosis control. E, Proteolytic processing of anx A1 can be blocked by the ADAM10 inhibitor GI 254023X. Jurkat cells were stimulated as in A in the absence or presence of 10 mM of the ADAM10 inhibitor GI 254023X (GI) or the ADAM10/17 inhibitor GW 280264X (GW). Subsequently, anti-anx A1 immunoblot analysis was performed with protein extracts as in D. F, Analysis of ADAM10 and ADAM17 knockdown efficiency by qRT-PCR. Knockdown of ADAM10 and ADAM17 expression was carried out by electroporation of Jurkat cells with two different ADAM10- and ADAM17-specific oligonucleotides and a scramble control oligonucleotide as described in Materials and Methods. Total RNA was prepared, reversely transcribed, and the resulting cDNA was used for qRT-PCR as described in Materials and Methods. Relative ADAM10/17 mRNA levels were normalized on the endogenous control ALAS-1, and the ADAM10/17 mRNA level in Jurkat cells that were treated with the scramble control siRNA was set as 100% calibrator. G, Anx A1 cleavage is strongly inhibited in ADAM10 silenced cells. siRNA- mediatedknockdownofADAM10orADAM17expressionwasperformedasinF. Subsequently, cells were stimulated to undergo secondary necrosis as in A,and anx A1 cleavage was monitored by immunoblot analysis. The amount of anx A1 p36 compared with total anx A1 was calculated from integrated pixel intensities, and the inhibition of cleavage is presented as percent of the scramble control. PARP served as a loading and apoptosis/secondary necrosis control. 140 AN ANX A1 PEPTIDE AS ATTRACTION SIGNAL OF 2˚ NECROTIC CELLS a membrane-resident but not a soluble protease mediated anx shown (26). We therefore sought to analyze whether this chemo- A1 cleavage (Fig. 3B). This observation was in accordance with tactic activity is preserved within the first seven amino acids and previous reports showing that membrane-bound (metallo-)protei- applied the synthetic peptide to a chemotaxis assay with THP-1 nases, including members of the ADAM family, are activated monocytes. Comparable with the classical chemokine MCP-1, anx during apoptosis and perform ectodomain shedding of various A1 aa 2–7 stimulated monocyte chemotaxis in a dose-dependent transmembrane proteins (22, 23). To investigate whether anx A1 manner (Fig. 5A). The reverse sequence anx A1 (aa 7–2) served cleavage is indeed mediated by a membrane-localized metallopro- as a negative control that displayed profoundly less chemotactic teinase, we next used the broad-range metalloproteinase inhibitor potential. We also analyzed the migratory response of primary o-phenanthroline and the more selective matrix metalloproteinase human monocytes toward anx A1 (aa 2–7) and observed that anx inhibitor GM 6001 targeting various matrix metalloproteinases A1 (aa 2–7) potently induced monocyte migration—although to and ADAMs. Both inhibitors reduced secondary necrosis-asso- a lesser extent than anx A1 (aa 2–26) or the classical chemokine ciated anx A1 processing in a dose-dependent manner (Fig. 3C, SDF-1a (Fig. 5B). Thus, our results reveal that anx A1 aa 2–7 is 3D). Remarkably, o-phenanthroline had previously been described endowed with chemotactic activity rendering it likely to be a to inhibit anx A1 truncation in the context of calcium-induced monocytic attraction signal of secondary necrotic cells. neutrophil activation (24). Finally, by using the ADAM10 inhib- We tested our hypothesis that the N-terminal anx A1 peptide itor GI 254023X and the ADAM10/17 inhibitor GW 280264X might be involved in phagocyte recruitment by initially charac- (25), we observed that ADAM10 was crucially involved in anx A1 terizing the biochemical properties of the chemotactic activity in cleavage, as both inhibitors profoundly mitigated anx A1 pro- culture supernatants of secondary necrotic cells (Supplemental Fig. cessing, and GI 254023X almost completely abolished it (Fig. 1). Indeed, secondary necrotic cells released attraction signals for Downloaded from 3E). THP-1 and primary human monocytes, respectively (Supplemental To corroborate further the results obtained with the ADAM Fig. 1A,1B,Fig.5B). Inhibiting cell death by addition of the poly- inhibitors, we next silenced the expression of ADAM10 and caspase inhibitor z-VAD-fmk profoundly reduced this release ADAM17 by using different siRNA oligonucleotides. qRT-PCR (Supplemental Fig. 1A,1B). Checkerboard analysis and direc- analysis revealed a knockdown efficiency of ∼60% for ADAM10 tional chemotaxis assays in IBIDI m-slide chemotaxis chambers

and ADAM17 using different specific siRNAs (Fig. 3F). Most confirmed that cell migration was due to chemotaxis (directed http://www.jimmunol.org/ importantly, secondary necrotic ADAM10-silenced cells displayed migration) and not chemokinesis (Supplemental Fig. 1C, Fig. 5C). a convincingly reduced amount of proteolytically processed anx In contrast to the classical chemokine SDF-1a, the chemotactic A1 p36, whereas ADAM17 knockdown had only a moderate yet activity in supernatants of secondary necrotic cells was heat- detectable inhibitory effect on anx A1 cleavage (Fig. 3G). Hence, stabile and could not be inactivated by incubation at 90˚C for ADAM10 appeared to be of crucial importance in this scenario. 40 min (Supplemental Fig. 1D). Ultrafiltration revealed that sub- , 7 stances of an apparent molecular mass 3 kDa accounted for Anx A1 is directly cleaved by ADAM10 after F ∼50% of the chemotactic activity (Supplemental Fig. 1E). Nota- The question that needed to be addressed at this point was whether bly, Jurkat and MOLT-4 cells, two cell lines externalizing and anx A1 was directly cleaved by ADAM10 or by a protease activated processing anx A1 during secondary necrosis (Fig. 1), released by guest on September 25, 2021 downstream of ADAM10. To this end, purified recombinant anx A1 profoundly more chemotactic activity than non–anx A1-exter- was incubated with purified ADAM10, and the reaction products nalizing THP-1 cells. Furthermore, this could be strongly re- were applied to SDS-PAGE and subsequent immunoblot analysis. duced by addition of a commercial protease inhibitor mixture As shown in Fig. 4B, ADAM10 specifically cleaved anx A1 into (Supplemental Fig. 1F). Taken together, the described biochemi- p36, and the cleavage was fully abrogated in the presence of the cal characteristics support the notion of peptide-like attraction ADAM inhibitor TAPI-2. Notably, no such cleavage was observed signals mediating phagocyte recruitment by secondary necrotic by coincubation of anx A1 with ADAM17 (Fig. 4E). To map ap- cells. proximately the cleavage site, we also digested the purified N- To address specifically the contribution of anx A1-derived pep- terminal domain as well as an N-terminally truncated mutant of tides in this scenario, we measured monocyte chemotaxis with anx A1 with ADAM10 (Fig. 4B–D). Proteolytic processing was supernatants of secondary necrotic anx A1 knockdown cells. In- only observed in case of the anx A1 N terminus suggesting that the triguingly, we observed a substantial reduction in monocyte at- cleavage site must be localized within the first 46 aa of anx A1. traction (Fig. 5D,5E). Of note, this was not due to the fact that the Subsequent Edman degradation of full-length anx A1 cleaved by process of apoptosis or secondary necrosis was impaired by the ADAM10 revealed the cleavage site to be located after F7 (Fig. 4F). lack of anx A1 (data not shown and Ref. 14). Instead, anx A1 Thus far, the current study and our previously published ob- apparently was crucially required for the generation of secondary servations (14) have shown that certain cell types translocate anx necrotic cell-derived “find-me” signals supporting our hypothesis A1 to the outer cell surface during secondary necrosis, where it is that the N-terminal peptide liberated by ADAM10-mediated proteolytically processed by ADAM10 at F7. This cleavage yields cleavage of anx A1 might play a role. This was further con- a 36-kDa large core fragment, which stays attached to PS-rich firmed by our observation that supernatants of secondary necrotic plaques on the cell surface, and an N-terminal peptide of 7 aa, cells generated in the presence of GM 6001 or GI 254023X dis- which is presumably released. Whereas the cell-bound core played a strongly reduced chemotactic potential (Fig. 6A,6B). fragment has been implicated in the phagocytic removal of the Finally, silencing ADAM10 expression by RNA interference con- dying cell corpses and in the regulation of the postphagocytic vincingly revealed that a reduction in ADAM10 expression in cytokine response (14), the function of the released N-terminal secondary necrotic cells was paralleled by a robustly decreased peptide remained elusive. extent of phagocyte recruitment (Fig. 6C,6D). In conclusion, our study shows for the first time that anx A1 is The anx A1 N-terminal peptide acts as a “find-me” signal of externalized and subsequently cleaved by ADAM10 during sec- secondary necrotic cells ondary necrosis, thus liberating an N-terminal peptide of 7 aa, For larger peptides derived from the anx A1 N terminus, including which crucially contributes to chemotactic phagocyte recruitment aa 2–26, induction of leukocyte chemotaxis has previously been by secondary necrotic cells. The Journal of Immunology 141 Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 4. Anx A1 is directly cleaved by ADAM10 after F7. A, Domain structure of different anx A1 constructs. Arabic numbers depict the amino acid position, and annexin repeats are numbered I–IV. B, Recombinant human anx A1 (aa 1–346) is processed by recombinant human ADAM10. One mi- crogram of recombinant human anx A1 (aa 1–346) was incubated with 100 ng recombinant human ADAM10 ectodomain in the presence or absence of 100 mM of the matrix metalloproteinase inhibitor TAPI-2 at 37˚C for the indicated times. Subsequently, anx A1 cleavage was detected by SDS-PAGE and immunoblot analysis with an anti-anx A1 Ab. C, Recombinant human anx A1 core domain (aa 47–346) is not cleaved by ADAM10. Incubation of anx A1 (aa 47–346) with ADAM10 ectodomain was performed as in B. For immunoblot analysis, a polyclonal anti-anx A1 Ab was used. D, The cleavage site of ADAM10 is located within the unique N-terminal domain of anx A1 (aa 1–46). One microgram of the recombinant human anx A1 N-terminal domain (aa 1–46) was incubated with ADAM10 as in B. Cleavage fragments were separated by SDS-PAGE and visualized by subsequent silver staining. E, Recombinant human anx A1 (aa 1–346) is not processed by recombinant human ADAM17. One microgram of recombinant human anx A1 (aa 1–346) was incubated with 100 ng of recombinant human ADAM17 ectodomain as in B. F, Identification of the ADAM10 cleavage site within the anx A1 N-terminal domain. Recombinant human anx A1 (aa 1–346) was incubated with native or heat-inactivated recombinant human ADAM10 as in A. Subsequently, the reaction mixture was subjected to N-terminal Edman degradation. The N-terminal sequence newly generated by incubation with active ADAM10 was L8 KQA.

Discussion leased and can stimulate inflammatory and autoimmune re- The development of chronic inflammation and autoimmunity is actions. known to be promoted by defects or delays in the phagocytic Our previous studies have demonstrated that anx A1 translocates clearance of apoptotic cells and the concomitant transition to to the surface of secondary necrotic cells. There, it promotes the secondary necrosis (16). Insufficient apoptotic cell removal might phagocytic uptake of dying cells and efficiently dampens the se- be due to direct impairments in the phagocytic process or massive cretion of proinflammatory cytokines by macrophages that have apoptosis overwhelming the phagocytic capacity in the context of ingested secondary necrotic cells. These findings suggest that anx acute inflammation, bacterial infections, ischemia, drug-induced A1 exposure represents an important anti-inflammatory fail-safe hepatitis, or tumor chemotherapy/radiotherapy (15). During sec- mechanism after the transition from apoptosis to secondary ne- ondary necrosis, the plasma membrane becomes permeable, and crosis. In our current follow-up study, we have investigated the intracellular contents, including cytotoxic metabolic intermediates, phenomenon of anx A1 externalization during secondary necrosis hydrolytic enzymes, danger signals, and autoantigens, are re- in greater detail. In this study, we show that after its translocation to 142 AN ANX A1 PEPTIDE AS ATTRACTION SIGNAL OF 2˚ NECROTIC CELLS Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021

FIGURE 5. The anx A1 N terminus acts as a “find-me” signal of secondary necrotic cells. A, THP-1 monocytes migrate toward anx A1 (aa 2–7). Anx A1 (aa 2–7) or MCP-1 were added at different concentrations to the lower chamber of a double-chamber plate. Transmigration assay with THP-1 cells was performed as described under Materials and Methods. The reverse sequence of anx A1 (aa 7–2) served as a control. Error bars represent SD of quad- ruplicates. B, Primary human monocytes migrate toward anx A1 (aa 2–7) and to supernatants of secondary necrotic cells. Left panel, Anx A1 (aa 2–7) or anx A1 (aa 2–26) were added at different concentrations to the lower chamber of a double-chamber plate. Transmigration assay with primary human monocytes was performed as described under Materials and Methods. Right panel, Cell-free supernatants of secondary necrotic Jurkat cells were used as chemotactic stimulus. Assay medium and recombinant SDF-1a (200 ng/ml) served as controls. Error bars represent SD of quadruplicates. C, Chemotactic response of primary human monocytes toward anx A1 (aa 2–7) and supernatants of secondary necrotic cells. Monocyte chemotaxis toward anx A1 (aa 2–7, 10 mM) and supernatants of secondary necrotic Jurkat cells was analyzed by live cell tracking in IBIDI m-slide chemotaxis chambers as described in Materials and Methods. DMSO and anx A1 (aa 2–26, 10 mM) served as negative and positive control, respectively. Trajectory paths of 30 randomly picked single cells are shown. Black paths depict cells with net migration upward, and red paths depict cells with net migration downward. The filled blue circle represents the center of mass after 4 h of migration (center of mass coordinates are given). D, Release of “find-me” signals during secondary necrosis is reduced in anx A1 knockdown cells. Knockdown of anx A1 expression in Jurkat and MOLT-4 cells was carried out as in Fig. 3G. On day 5 after the first electroporation, apoptosis and subsequent secondary necrosis were induced by UV irradiation (10 mJ/cm2). Cell-free culture supernatants were collected and subjected to a transmigration assay with THP-1 cells as in A. Mean values + SD are shown. E, Immunoblot analysis of anx A1-silenced or scramble control cell extracts confirmed the efficient downregulation of anx A1 expression but not of the control protein vinculin (10 mg protein per lane). The integrated intensity of the anx A1 bands was normalized on the integrated intensity of the vinculin bands, and scramble transfected cells were set as 1.00 calibrator. the outer leaflet of the plasma membrane, anx A1 was proteolyt- phenomenon but rather restricted to certain cell types. Further- ically processed at the cell surface. We could identify ADAM10 as more, we observed that protein extracts from cells, which exhibited the responsible protease, which cleaved anx A1 within its unique anx A1 externalization in the course of secondary necrosis, con- N-terminal domain after F7. Thereby, a small peptide was gener- tained a proteolytically processed, 36-kDa form of anx A1 in ated and released from the secondary necrotic cell surface, which addition to the full-length protein of 38 kDa. These findings to- in turn induced chemoattraction of monocytes. Hence, anx A1 gether with our kinetic analyses of anx A1 cleavage and our ob- apparently constitutes a multifunctional protein that acts at dif- servation that the cleaved 36-kDa form was exclusively found on ferent stages and levels of cell death and dying cell clearance. the dying cell surface support the conclusion that anx A1 is The results presented here show that externalization of anx A1 truncated after its translocation to the cell surface, although other during secondary necrosis is not a general, cell type-independent reports have shown that anx A1 can also be intracellularly pro- The Journal of Immunology 143 Downloaded from

FIGURE 6. ADAM10 contributes to the release of “find-me” signals from secondary necrotic cells. A, GM 6001 inhibits the release of monocytic chemoattractants during secondary necrosis. Jurkat, MOLT-4, and THP-1 cells were UV-irradiated and incubated for 12 or 18 h in the presence or absence http://www.jimmunol.org/ of the broad-range matrix metalloproteinase inhibitor GM 6001 (100 mM). Cell-free culture supernatants were collected and analyzed for their chemotactic potential as in Fig. 5A. Error bars represent SD of quadruplicates. B, The release of secondary necrotic cell-derived attraction signals is strongly decreased in the presence of the ADAM10-specific inhibitor GI 254023X. Jurkat cells were UV-irradiated and incubated for 18 h in the absence or presence of 10 mM of the ADAM10 inhibitor GI 254023X (GI) or the ADAM10/17 inhibitor GW 280264X (GW). Supernatants were collected and applied to a transmigration assay with THP-1 cells as in Fig. 5A. Error bars represent SD of quadruplicates. C, Evaluation of ADAM10 knockdown efficiency by FACS analysis. Knockdown of ADAM10 expression in Jurkat cells was carried out as in Fig. 3G. On day 5 after the first electroporation, cells were fixed, permeabilized, and stained with anti-ADAM10–PE Ab to detect ADAM10 expression level by FACS analysis or IgG-2b-PE isotype control. Left panel, Representative histograms are shown. Right panel, Median PE fluorescence of the histograms in the left panel. D, The release of monocytic attraction signals is strongly decreased in ADAM10-silenced secondary necrotic cells. ADAM10-silenced cells were induced to undergo secondary necrosis by UV irradiation. Cell culture supernatants were harvested and analyzed for their chemotactic potential in a transmigration assay with THP-1 cells as in Fig. 5A. Mean values + SD by guest on September 25, 2021 of quadruplicates are given. cessed by caspase-3 during apoptosis (27). Our pharmacological which is not accessible for the ectoprotease ADAM10. How- inhibition and RNA interference studies identified ADAM10 as ever, after its translocation to the cell surface and the calcium- the responsible protease that cleaved anx A1 during secondary dependent binding to PS-rich plaques with concomitant exposition necrosis. ADAM10 is well known to execute the processing of of its N-terminal domain (1, 35), the latter becomes available for various transmembrane and membrane-associated proteins, in- ADAM10-mediated shedding. Our finding that only anx A1- cluding MICA, CD46, CD95L, and E- (22, 28–30). No- externalizing cell types displayed anx A1 cleavage during sec- tably, during neutrophil apoptosis, ADAM17 has been implicated ondary necrosis, although all other non–anx A1-externalizing cells in the generation of a soluble IL-6 receptor fragment, which expressed ADAM10 on their surfaces (Supplemental Fig. 3), governs monocyte/macrophage recruitment for the resolution of strongly supports this notion. Therefore, it can be concluded that inflammation (23). However, in this report, IL-6 receptor cleavage externalization of anx A1 is the crucial and pace-limiting step in was observed early during apoptosis, as long as the plasma this context. membrane integrity was still intact. Nevertheless, with the iden- Sequencing of the cleavage site by Edman degradation has tification of the anx A1-derived N-terminal peptide, our study adds revealed that ADAM10 cleaved anx A1 after F7 within its N- a second “find-me” signal that is generated in an ADAM- terminal domain. This is, to our knowledge, the first time that dependent fashion to the scenario of phagocytic dying cell cleavage at this position is reported, although other cleavage sites clearance. within the anx A1 N-terminal domain have been described. For The mechanisms that orchestrate ADAM10 activation during instance, during neutrophil activation, human leukocyte elastase apoptosis and secondary necrosis are far from being understood. and proteinase 3 have been shown to process anx A1 at aa 26 and Enhanced activity due to an upregulation of surface expression can between aa 29 and 33, respectively (6, 7, 36). The corresponding be excluded in this context, as we did not detect any significant peptide (aa 2–26) has been implicated in the induction of mono- increase in ADAM10 surface staining during apoptosis or sec- cyte chemotaxis, inhibition of neutrophil extravasation, and other ondary necrosis (Supplemental Fig. 2). However, there is accu- anti-inflammatory processes, such as reduction of neutrophil- mulating evidence that high intracellular calcium concentrations dependent mouse skin edema, inhibition of neutrophil accumula- and proteolytic maturation are involved (31–33). Because both tion in zymosan-induced peritonitis, protection from experimen- events might also occur in viable cells or during the early phases tally induced renal ischemia–reperfusion injury, and amelioration of apoptosis (34), the question arises why anx A1 processing is of acute carrageenan-induced inflammation (37–39). Intriguingly, such a late event in the course of cell death. A feasible explanation it was also reported that N-terminal anx A1 peptides promote the would be spatial availability. Anx A1 is an intracellular protein, macrophage-mediated phagocytosis of apoptotic cells (40). The 144 AN ANX A1 PEPTIDE AS ATTRACTION SIGNAL OF 2˚ NECROTIC CELLS peptide described in the current study (aa 2–7) is much shorter in 13. Fan, X., S. Krahling, D. Smith, P. Williamson, and R. A. Schlegel. 2004. Macrophage surface expression of annexins I and II in the phagocytosis of length. Nevertheless, it convincingly induced monocyte chemo- apoptotic lymphocytes. Mol. Biol. Cell 15: 2863–2872. taxis. Most importantly, supernatants of secondary necrotic anx 14. Blume, K. E., S. Soeroes, M. Waibel, H. Keppeler, S. Wesselborg, M. Herrmann, A1 or ADAM10 knockdown cells displayed a profound reduction K. Schulze-Osthoff, and K. Lauber. 2009. Cell surface externalization of annexin A1 as a failsafe mechanism preventing inflammatory responses during secondary in their chemotactic potential implying that this peptide essentially necrosis. J. Immunol. 183: 8138–8147. contributes to monocyte recruitment by secondary necrotic cells. 15. Silva, M. T., A. do Vale, and N. M. dos Santos. 2008. Secondary necrosis in The receptors that mediate the effects of anx A1 (aa 2–26) have multicellular animals: an outcome of apoptosis with pathogenic implications. Apoptosis 13: 463–482. been identified as members of the formyl peptide receptor family 16. Mun˜oz, L. E., K. Lauber, M. Schiller, A. A. Manfredi, and M. Herrmann. 2010. (41–43). Receptor binding and activation has been shown for all The role of defective clearance of apoptotic cells in systemic autoimmunity. Nat. three members of the formyl peptide receptor family (26). Nota- Rev. Rheumatol. 6: 280–289. 17. Lauber, K., E. Bohn, S. M. Kro¨ber, Y. J. Xiao, S. G. Blumenthal, bly, THP-1 monocytes that were used in the current study express R. K. Lindemann, P. Marini, C. Wiedig, A. Zobywalski, S. Baksh, et al. 2003. FPR1 and FPR2 but not FPR3 at detectable mRNA levels (Sup- Apoptotic cells induce migration of phagocytes via caspase-3-mediated release of a lipid attraction signal. Cell 113: 717–730. plemental Fig. 4). However, in the case of FPR1, the crucial motif 18. Lauber, K., H. A. Appel, S. F. Schlosser, M. Gregor, K. Schulze-Osthoff, and for receptor binding, activation and desensitization was mapped to S. Wesselborg. 2001. The adapter protein apoptotic protease-activating factor-1 anx A1 aa 9–12 (44). This renders FPR1 unlikely to be involved (Apaf-1) is proteolytically processed during apoptosis. J. Biol. Chem. 276: 29772–29781. in monocyte migration stimulated by anx A1 aa 2–7, and the re- 19. Berg, C. P., G. M. Stein, H. Keppeler, M. Gregor, S. Wesselborg, and K. Lauber. sponsible receptor remains to be identified in future studies. 2008. Apoptosis-associated antigens recognized by autoantibodies in patients Overall, anx A1 represents a protein with various functions with the autoimmune liver disease primary biliary cirrhosis. Apoptosis 13: 63– 75. at different levels of cell death, dying cell clearance, and anti- 20. Rabilloud, T., G. Carpentier, and P. Tarroux. 1988. Improvement and simplifi- Downloaded from inflammation (45). The current study extends this spectrum to cation of low-background silver staining of proteins by using sodium dithionite. a novel function in the context of dying cell removal: monocyte Electrophoresis 9: 288–291. 21. Peter, C., M. Waibel, C. G. Radu, L. V. Yang, O. N. Witte, K. Schulze-Osthoff, recruitment by an N-terminal, anx A1-derived peptide that is S. Wesselborg, and K. Lauber. 2008. Migration to apoptotic “find-me” signals is generated by ADAM10 during secondary necrosis. The multi- mediated via the phagocyte receptor G2A. J. Biol. Chem. 283: 5296–5305. functional properties of anx A1 represent attractive targets for 22. Hakulinen, J., and J. Keski-Oja. 2006. ADAM10-mediated release of comple- ment membrane cofactor protein during apoptosis of epithelial cells. J. Biol. novel therapeutic strategies in the context of chronic inflammation Chem. 281: 21369–21376. http://www.jimmunol.org/ and autoimmunity. In this line, local administration of anx A1 or 23. Chalaris, A., B. Rabe, K. Paliga, H. Lange, T. Laskay, C. A. Fielding, S. A. Jones, S. Rose-John, and J. Scheller. 2007. Apoptosis is a natural stimulus peptides derived from its N terminus has been reported to alleviate of IL6R shedding and contributes to the proinflammatory trans-signaling func- various types of inflammatory reactions (46), and it is tempting to tion of neutrophils. Blood 110: 1748–1755. speculate that this might also be a prospective approach for the 24. Movitz, C., C. Sjo¨lin, and C. Dahlgren. 1999. Cleavage of annexin I in human neutrophils is mediated by a membrane-localized metalloprotease. Biochim. treatment of inflammatory and autoimmune diseases with known Biophys. Acta 1416: 101–108. associations to defects in dying cell clearance. 25. Hundhausen, C., D. Misztela, T. A. Berkhout, N. Broadway, P. Saftig, K. Reiss, D. Hartmann, F. Fahrenholz, R. Postina, V. Matthews, et al. 2003. The disintegrin-like metalloproteinase ADAM10 is involved in constitutive cleavage Acknowledgments of CX3CL1 (fractalkine) and regulates CX3CL1-mediated cell-cell adhesion. We thank Andreas Ludwig for providing the ADAM inhibitors GI 254023X Blood 102: 1186–1195. by guest on September 25, 2021 26. Ernst, S., C. Lange, A. Wilbers, V. Goebeler, V. Gerke, and U. Rescher. 2004. An and GW 280264X and Robert Pick for valuable advice concerning the annexin 1 N-terminal peptide activates leukocytes by triggering different live cell tracking experiments. members of the formyl peptide receptor family. J. Immunol. 172: 7669–7676. 27. Debret, R., H. El Btaouri, L. Duca, I. Rahman, S. Radke, B. Haye, J. M. Sallenave, and F. Antonicelli. 2003. Annexin A1 processing is associated Disclosures with caspase-dependent apoptosis in BZR cells. FEBS Lett. 546: 195–202. The authors have no financial conflicts of interest. 28. Waldhauer, I., D. Goehlsdorf, F. Gieseke, T. Weinschenk, M. Wittenbrink, A. Ludwig, S. Stevanovic, H. G. Rammensee, and A. Steinle. 2008. Tumor- associated MICA is shed by ADAM proteases. Cancer Res. 68: 6368–6376. 29. Kirkin, V., N. Cahuzac, F. Guardiola-Serrano, S. Huault, K. Lu¨ckerath, References E. Friedmann, N. Novac, W. S. Wels, B. Martoglio, A. O. Hueber, and 1. Gerke, V., C. E. Creutz, and S. E. Moss. 2005. Annexins: linking Ca2+ signalling M. Zo¨rnig. 2007. The Fas ligand intracellular domain is released by ADAM10 to membrane dynamics. Nat. Rev. Mol. Cell Biol. 6: 449–461. and SPPL2a cleavage in T-cells. Cell Death Differ. 14: 1678–1687. 2. Perretti, M., and F. D’Acquisto. 2009. Annexin A1 and glucocorticoids as 30. Schulz, B., J. Pruessmeyer, T. Maretzky, A. Ludwig, C. P. Blobel, P. Saftig, and effectors of the resolution of inflammation. Nat. Rev. Immunol. 9: 62–70. K. Reiss. 2008. ADAM10 regulates endothelial permeability and T-Cell trans- 3. Liemann, S., and A. Lewit-Bentley. 1995. Annexins: a novel family of calcium- migration by proteolysis of vascular endothelial cadherin. Circ. Res. 102: 1192– and membrane-binding proteins in search of a function. Structure 3: 233–237. 1201. 4. Liemann, S., and R. Huber. 1997. Three-dimensional structure of annexins. Cell. 31. Anders, A., S. Gilbert, W. Garten, R. Postina, and F. Fahrenholz. 2001. Regu- Mol. Life Sci. 53: 516–521. lation of the alpha-secretase ADAM10 by its prodomain and proprotein con- 5. Wang, W., and C. E. Creutz. 1992. Regulation of the chromaffin granule ag- vertases. FASEB J. 15: 1837–1839. gregating activity of annexin I by phosphorylation. Biochemistry 31: 9934–9939. 32. Lopez-Perez, E., Y. Zhang, S. J. Frank, J. Creemers, N. Seidah, and F. Checler. 6. Rescher, U., V. Goebeler, A. Wilbers, and V. Gerke. 2006. Proteolytic cleavage of 2001. Constitutive alpha-secretase cleavage of the beta-amyloid precursor pro- annexin 1 by human leukocyte elastase. Biochim. Biophys. Acta 1763: 1320–1324. tein in the furin-deficient LoVo cell line: involvement of the pro- con- 7. Vong, L., F. D’Acquisto, M. Pederzoli-Ribeil, L. Lavagno, R. J. Flower, vertase 7 and the disintegrin metalloprotease ADAM10. J. Neurochem. 76: V. Witko-Sarsat, and M. Perretti. 2007. Annexin 1 cleavage in activated neu- 1532–1539. trophils: a pivotal role for proteinase 3. J. Biol. Chem. 282: 29998–30004. 33. Le Gall, S. M., P. Bobe´, K. Reiss, K. Horiuchi, X. D. Niu, D. Lundell, 8. Gerke, V., and S. E. Moss. 2002. Annexins: from structure to function. Physiol. D. R. Gibb, D. Conrad, P. Saftig, and C. P. Blobel. 2009. ADAMs 10 and 17 Rev. 82: 331–371. represent differentially regulated components of a general shedding machinery 9. Rescher, U., A. Danielczyk, A. Markoff, and V. Gerke. 2002. Functional acti- for membrane proteins such as transforming growth factor alpha, L-selectin, and vation of the formyl peptide receptor by a new endogenous ligand in human lung tumor necrosis factor alpha. Mol. Biol. Cell 20: 1785–1794. A549 cells. J. Immunol. 169: 1500–1504. 34. Orrenius, S., B. Zhivotovsky, and P. Nicotera. 2003. Regulation of cell death: the 10. Solito, E., I. A. Romero, S. Marullo, F. Russo-Marie, and B. B. Weksler. 2000. calcium-apoptosis link. Nat. Rev. Mol. Cell Biol. 4: 552–565. Annexin 1 binds to U937 monocytic cells and inhibits their adhesion to mi- 35. Rosengarth, A., and H. Luecke. 2003. A calcium-driven conformational switch crovascular endothelium: involvement of the alpha 4 beta 1 integrin. J. Immunol. of the N-terminal and core domains of annexin A1. J. Mol. Biol. 326: 1317– 165: 1573–1581. 1325. 11. Byrne, A., and D. J. Reen. 2002. Lipopolysaccharide induces rapid production of 36. Smith, S. F., T. D. Tetley, A. Guz, and R. J. Flower. 1990. Detection of lipocortin IL-10 by monocytes in the presence of apoptotic neutrophils. J. Immunol. 168: 1 in human lung lavage fluid: lipocortin degradation as a possible proteolytic 1968–1977. mechanism in the control of inflammatory mediators and inflammation. Environ. 12. Arur, S., U. E. Uche, K. Rezaul, M. Fong, V. Scranton, A. E. Cowan, W. Mohler, Health Perspect. 85: 135–144. and D. K. Han. 2003. Annexin I is an endogenous ligand that mediates apoptotic 37. Perretti, M., A. Ahluwalia, J. G. Harris, N. J. Goulding, and R. J. Flower. 1993. cell engulfment. Dev. Cell 4: 587–598. Lipocortin-1 fragments inhibit neutrophil accumulation and neutrophil-dependent The Journal of Immunology 145

edema in the mouse. A qualitative comparison with an anti-CD11b monoclonal 42. Gavins, F. N., P. Sawmynaden, B. E. Chatterjee, and M. Perretti. 2005. A twist in antibody. J. Immunol. 151: 4306–4314. anti-inflammation: annexin 1 acts via the lipoxin A4 receptor. Prostaglandins 38. Gastardelo, T. S., A. S. Damazo, J. Dalli, R. J. Flower, M. Perretti, and Leukot. Essent. Fatty Acids 73: 211–219. S. M. Oliani. 2009. Functional and ultrastructural analysis of annexin A1 and its 43. Hayhoe, R. P., A. M. Kamal, E. Solito, R. J. Flower, D. Cooper, and M. Perretti. receptor in extravasating neutrophils during acute inflammation. Am. J. Pathol. 2006. Annexin 1 and its bioactive peptide inhibit neutrophil-endothelium 174: 177–183. interactions under flow: indication of distinct receptor involvement. Blood 39. Facio, F. N., Jr., A. A. Sena, L. P. Araujo, G. E. Mendes, I. Castro, M. A. Luz, L. Yu, 107: 2123–2130. S. M. Oliani, and E. A. Burdmann. 2011. Annexin 1 mimetic peptide protects 44. Movitz, C., L. Brive, K. Hellstrand, M. J. Rabiet, and C. Dahlgren. 2010. The against renal ischemia/reperfusion injury in rats. J. Mol. Med. (Berl). 89: 51–63. annexin I sequence gln(9)-ala(10)-trp(11)-phe(12) is a core structure for inter- 40. Scannell, M., M. B. Flanagan, A. deStefani, K. J. Wynne, G. Cagney, C. Godson, action with the formyl peptide receptor 1. J. Biol. Chem. 285: 14338–14345. and P. Maderna. 2007. Annexin-1 and peptide derivatives are released by apo- 45. D’Acquisto, F., M. Perretti, and R. J. Flower. 2008. Annexin-A1: a pivotal ptotic cells and stimulate phagocytosis of apoptotic neutrophils by macrophages. regulator of the innate and adaptive immune systems. Br. J. Pharmacol. 155: J. Immunol. 178: 4595–4605. 152–169. 41. Walther, A., K. Riehemann, and V. Gerke. 2000. A novel ligand of the formyl 46. Yang, Y., M. Leech, P. Hutchinson, S. R. Holdsworth, and E. F. Morand. 1997. peptide receptor: annexin I regulates neutrophil extravasation by interacting with Antiinflammatory effect of lipocortin 1 in experimental arthritis. Inflammation the FPR. Mol. Cell 5: 831–840. 21: 583–596. Downloaded from http://www.jimmunol.org/ by guest on September 25, 2021