sign in sign up
<<
Home , Basophil, Neutrophil

NADPH Oxidase−Independent Formation of Extracellular DNA Traps by Mahbubul Morshed, Ruslan Hlushchuk, Dagmar Simon, Andrew F. Walls, Kazushige Obata-Ninomiya, Hajime This information is current as Karasuyama, Valentin Djonov, Alexander Eggel, Thomas of September 26, 2021. Kaufmann, Hans-Uwe Simon and Shida Yousefi J Immunol 2014; 192:5314-5323; Prepublished online 25 April 2014;

doi: 10.4049/jimmunol.1303418 Downloaded from http://www.jimmunol.org/content/192/11/5314

Supplementary http://www.jimmunol.org/content/suppl/2014/04/25/jimmunol.130341 Material 8.DCSupplemental http://www.jimmunol.org/ References This article cites 43 articles, 14 of which you can access for free at: http://www.jimmunol.org/content/192/11/5314.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 26, 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 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 © 2014 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

NADPH Oxidase–Independent Formation of Extracellular DNA Traps by Basophils

Mahbubul Morshed,* Ruslan Hlushchuk,† Dagmar Simon,‡ Andrew F. Walls,x Kazushige Obata-Ninomiya,{ Hajime Karasuyama,{,‖ Valentin Djonov,† Alexander Eggel,# Thomas Kaufmann,* Hans-Uwe Simon,* and Shida Yousefi*

Basophils are primarily associated with a proinflammatory and immunoregulatory role in allergic diseases and parasitic . Recent studies have shown that basophils can also bind various both in the presence and the absence of opsonizing Abs. In this report, we show that both human and mouse basophils are able to produce mitochondrial and to form extracellular DNA traps upon IL-3 priming and subsequent activation of the complement factor 5 a receptor or Fc«RI. Such extracellular traps (BETs) contain mitochondrial, but not nuclear DNA, as well as the basogranulin and mouse mast 8. BET formation occurs despite the absence of any functional NADPH oxidase in basophils. BETs can be Downloaded from found in both human and mouse inflamed tissues, suggesting that they also play a role under in vivo inflammatory conditions. Taken together, these findings suggest that basophils exert direct innate immune effector functions in the extracellular space. The Journal of Immunology, 2014, 192: 5314–5323.

asophils make up ,0.5% of all peripheral leuko- types at sites of allergic inflammation. IL-3 is not only an im- http://www.jimmunol.org/ cytes but share with mast cells the unique capacity to portant activation factor for basophils, it is also important for their B synthesize various hallmark mediators of allergic in- differentiation and survival (4). In addition, IL-3 primes basophils flammation and to express high-affinity IgE receptors (FcεRI) (1). to produce mediators, such as C4 (LTC4)inan Aggregation of FcεRI bound to IgE by multivalent Ag leads to -independent manner in response to stimulation with , granule exocytosis, and mediator release. complement factor 5 (C5a) (4). It also has been shown that thymic Besides functioning as effector cells, basophils also may play an stromal lymphopoietin (TSLP) can act as a key regulator of ba- important role in the initiation of an allergic inflammatory re- sophil hematopoiesis, independent of IL-3 (5). sponse (e.g., by recruiting other inflammatory cells) (2, 3). A role for basophils in innate is suggested by their

Basophils are also a prime source for the Th2 IL-4 and expression of the functional TLRs 2 and 4 as well as by the by guest on September 26, 2021 IL-13 (1). The phenotype and functions of human basophils are possibility of activating these cells by multiple , including tightly regulated by IL-3, a produced by various cell Der p1 and hookworm, in a non–IgE-dependent manner (6, 7). Basophils also can recognize microbes in an IgE-dependent manner (8) and release antibacterial factors, such as b- *Institute of Pharmacology, University of Bern, CH-3010 Bern, Switzerland; and (9). Previously published work has demonstrated †Institute of Anatomy, University of Bern, CH-3012 Bern, Switzerland; ‡Department of Dermatology, Inselspital, Bern University Hospital, CH-3010 Bern, Switzerland; that both and are capable of killing bac- x Immunopharmacology Group, University of Southampton, Southampton General teria extracellularly by forming DNA traps, so-called Hospital, Southampton SO16 6YD, United Kingdom; {Department of Immune Regulation, Tokyo Medical and Dental University Graduate School of Medical and extracellular traps (NETs) and extracellular traps (EETs) Dental Sciences, Tokyo 113-8518, Japan; ‖Core Research for Evolutional Science (10, 11). In this study, we have investigated whether basophils also and Technology, Japan Science and Technology Agency, Tokyo 113-8518, Japan; can form extracellular DNA traps. and #Institute of Immunology, University of Bern, Inselspital, CH-3010 Bern, Switzerland Received for publication December 23, 2013. Accepted for publication March 31, Materials and Methods 2014. Reagents This work was supported by the Swiss National Science Foundation. M.M. is a Ph.D. Recombinant human IL-3, rat IgG2a, rabbit IgG, as well as recombinant student of the Graduate School of Cellular and Biomedical Sciences of the University of Bern. Images were acquired on equipment supported by the Microscopy Imaging human eotaxin and thymic stromal lymphopoietin (TSLP) were from R&D Centre of the University of Bern. Systems Europe (Abingdon, U.K.). GM-CSF was provided by Novartis Pharma (Nurnberg,€ Germany). Human and mouse C5a were purchased Address correspondence and reprint requests to Dr. Shida Yousefi, Institute of Phar- from Hycult Biotech (Uden, The Netherlands). LPS, dihydrorhodamine- macology, University of Bern, Friedbuehlstrasse 49, CH-3010 Bern, Switzerland. E-mail address: shida.yousefi@pki.unibe.ch 123 (DHR), NBT, and fMLF, as well as Luria–Bertani and BHI media were obtained from Sigma-Aldrich (Buchs, Switzerland). Lipoteichoic The online version of this article contains supplemental material. acid (LTA) was a gift from Dr. T. Hartung (University of Konstanz, Abbreviations used in this article: BET, basophil extracellular trap; BG, basogranu- Konstanz, Germany). -activating factor (PAF) was from Millipore lin; C5a, complement factor 5a; DHR, dihydrorhodamine-123; DNase, deoxyribonu- (Zug, Switzerland). Diphenylene iodonium chloride (DPI) and PMA were clease; DPI, diphenylene iodonium chloride; EET, eosinophil extracellular trap; from Calbiochem (Merck, Darmstadt, Germany). Mitoquinone (MitoQ) EthD-1, ethidium homodimer-1; i.d., intradermal(ly); LTA, lipoteichoic acid; MitoQ, was from Biotrend Chemicals (Zurich, Switzerland). Propidium iodide mitoquinone; mMCP-8, mouse protease 8; MMP-9, matrix metallopepti- (PI), Hoechst 33342, SYTO13, MitoSOX Red, mounting medium con- dase 9; mtDNA, mitochondrial DNA; nDNA, nuclear DNA; NET, neutrophil extra- cellular trap; NOX, NADPH oxidase; PAF, platelet-activating factor; phox, taining DAPI, as well as calcein-AM and ethidium homodimer-1 (EthD-1), phagocytic oxidase; PI, propidium iodide; qPCR, quantitative PCR; ROS, reactive HBSS, and RPMI 1640 medium were from Invitrogen (Paisley, U.K.). oxygen species; TNP, 2,4,6-trinitrophenyl; TSLP, thymic stromal lymphopoietin. X-VIVO 15 medium was from Lonza (Verviers, Belgium). Anti-FcεRIa Ab (CRA1) was from Miltenyi Biotec (Bergisch Gladbach, Germany). APC- Copyright Ó 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00 conjugated anti-human FcεR1a (CRA1) and FITC-conjugated anti-human www.jimmunol.org/cgi/doi/10.4049/jimmunol.1303418 The Journal of Immunology 5315

IgE (clone Ige21) Abs were from eBioscience (Vienna, Austria). FITC- formaldehyde-fixed and paraffin-embedded human and mouse skin sec- conjugated anti-mouse FcεRIa (clone MAR-1), rat anti-mouse mast cell tions, respectively, using appropriate secondary Abs labeled with Alexa protease 8 (mMCP-8, TUG8), APC-conjugated anti-mouse c- (CD117, Fluor 488 (Invitrogen). After washing with PBS, the specimens were clone 2B8), APC-conjugated anti-mouse Gr-1 (Ly-6G/Ly-6C, clone mounted in a drop of fluorescence mounting medium (DakoCytomation, RB6-8C5), and APC-conjugated goat anti-rat Abs were from BioLegend Glostrup, Denmark). Slides were analyzed by confocal laser scanning (Lucerna-Chem, Luzern, Switzerland). Anti-rabbit matrix metallopeptidase microscopy (LSM 510; Carl Zeiss Micro Imaging, Jena, Germany). 9 (MMP-9) Ab was purchased from Abcam (Cambridge, U.K.). Anti-mouse gp91phox (clone 53/gp91[phox]), APC-conjugated anti-human CCR3 Electron microscopy (CD193, clone 5E8), and PE-conjugated anti–IL-3Ra (CD123, clone 9F5) Abs were obtained from BD Biosciences (Allschwil, Switzerland). Anti- For transmission electron microscopy, freshly isolated and activated human histone Ab recognizing the histones H1, H2A, H2B, H3, and H4 basophils were fixed in 2.5% (v/v) glutaraldehyde solution buffered with was from Millipore (clone H11-4). Polyvalent human IgG was a gift from sodium cacodylate (pH 7.4, 540 mOsm), postfixed in 1% OsO4 (buffered CSL Behring (Bern, Switzerland). with 0.1 M sodium cacodylate [pH 7.4, 340 mOsm]), dehydrated in eth- anol, and embedded in epoxy resin. Thick sections (70–80 nm) were Cells prepared and mounted on copper grids coated with Formvar (polyvinyl formal; Fluka, Buchs, Switzerland). Cells were then stained with lead Written informed consent was obtained from all blood donors, and the citrate and uranyl acetate prior to viewing in a Philips EM-400 transmis- Ethics Committee of the Canton of Bern approved this study. Human sion electron microscope. basophils were purified from healthy donors and atopic patients by negative For scanning electron microscopy, cells were fixed and processed on the selection from peripheral blood using the EasySep human basophil en- coverslips at all times. The fixative was 2.5% (v/v) glutaraldehyde solution richment kit (StemCell Technologies, Grenoble, France). Unwanted cells buffered with sodium cacodylate (pH 7.4, 540 mOsm). After fixation, cells were targeted for removal by tetrameric Ab complexes recognizing CD2, were postfixed in 1% OsO4 (buffered with 0.1 M sodium cacodylate [pH

CD3, CD14, CD15, CD16, CD19, CD24, CD34, CD36, CD45RA, CD56, 7.4, 340 mOsm]), dehydrated in ethanol, and underwent critical-point Downloaded from glycophorin A, and dextran-coated magnetic particles. Purity of the isolated drying. The coverslips were glued onto aluminum stabs, sputtered with basophils was .95% as assessed by Diff-Quik (Medion, Dudingen, gold and examined in a Philips XL-30 FEG scanning electron microscope. Switzerland) and light microscopy as well as flow cytometry (Supplemental Fig. 1A). Determination of cell death Human blood neutrophils were purified from healthy individuals by Ficoll–Hypaque centrifugation (12). Briefly, PBMC were separated by Cell death of activated basophils was assessed by uptake of 450 nM centrifugation on Ficoll–Hypaque (Seromed–Fakola, Basel, Switzerland). calcein-AM and 450 nM EthD-1 (both from Invitrogen) using live cell The lower layer consisting mainly of and erythrocytes was imaging techniques under confocal laser scanning microscopy. To analyze http://www.jimmunol.org/ basophil viability over longer periods of time, cells were cultured for the treated with erythrocyte lysis solution (155 mM NH4Cl, 10 mM KHCO3, and 0.1 mM EDTA [pH 7.3]). The resulting cell populations contained indicated times, and cell death was assessed by uptake of ethidium bromide m .95% neutrophils as assessed by Diff-Quik staining and light microscopy. and flow cytometric analysis (11, 16). PMA (50 nM, 100 nM, and 10 M) Mouse basophils were generated from IL-3–dependent, conditional was used as a positive control. Hoxb8-immortalized myeloid progenitors derived from wild-type and NADPH oxidase NOX2-deficient mice (13). NOX2-deficient mice were Reactive oxygen species measurements purchased from The Jackson Laboratory and provided by Dr. K.-H. Krause Cells were seeded in clear glass bottom, black wall 96-well microplates (Geneva, Switzerland). A total of 2 3 105 cells/ml were passaged in RPMI m (Greiner Bio-One, Frickenhausen, Germany), primed with IL-3 for 15 min, 1640/GlutaMAX with 10% FCS, 100 U/ml penicillin/100 g/ml strepto- and subsequently activated with C5a or anti-FcεRIa Ab for 30 min or for mycin, 50 mM 2-ME, mouse IL-3 (added as 10% of WEHI-3B cell–condi- the indicated times. During cell activation, incubations were with 1 mM by guest on September 26, 2021 tioned medium as a source for mouse IL-3), and 100 nM 4-hydroxytamoxifen DHR (Sigma-Aldrich) or with 5 mM MitoSOX Red. A total of 50 nM 3 4 for 3–4 d. To initiate differentiation into basophils, 2.5 10 cells/ml were PMA was used as control. To block ROS generation, the NADPH oxidase washed and then cultured in the same medium in the absence of 4-hydrox- inhibitor DPI (5 mM, except in concentration-dependence experiments) ytamoxifen. At day 6, cells were washed twice with PBS and used for sub- and the flavoprotein inhibitor MitoQ (1 mM) were added to basophils 10 sequent experiments. Mouse Hoxb8 neutrophils were generated as described min prior to stimulation. ROS generation was measured at different time previously (13, 14). Hoxb8 basophils and Hoxb8 neutrophils were character- points using a SpectraMax M2 plate reader (Bucher Biotec, Zurich, ized by flow cytometry (Supplemental Fig. 1B). Switzerland) at 488 nM excitation and 530 nM emission for DHR and 510 nM excitation and 580 nM emission for MitoSOX Red (17, 18). Activation of basophils Cells were seeded on 12-mm glass coverslips in X-VIVO 15 medium NBT assay (Lonza, Walkersville, MD), primed with 10 ng/ml recombinant human IL-3 To directly analyze NADPH oxidase activity, granulocytes were seeded on for 15 min, and subsequently stimulated with 10 mg/ml anti-FcεR1a Ab or glass coverslips, and the NBT assay was performed as described previously 10 nM C5a for 30 min. Other stimuli were used as follows: 100 nM fMLF, (19). Briefly, cells were stimulated as described above and 0.5 mg/ml NBT 20 ng/ml TSLP, 100 nM eotaxin, 100 nM PAF, 100 ng/ml LPS, and 30 mg/ was added during the last 20 min of activation. Cells were then washed ml LTA. Therefore, the total basophil activation time was 45 min. For 6 with PBS and stained with Diff-Quik to label the nuclei. The conversion of mouse basophils, the 6-d differentiated Hoxb8 basophils (10 3 10 /ml) NBT to insoluble dark purple formazan deposits was acquired by bright- were washed with PBS and resuspended in RPMI 1640 medium with 0.2% field microscopy. To determine whether formazan deposits are associated BSA. They were then sensitized by mouse monoclonal anti–2,4,6-trini- with DNA release by single cells, 5 mM MitoSOX Red was added during trophenyl (TNP) IgE (IgE1-b4; a gift from Dr. G. Nilsson, Stockholm, the last 10 min of activation. After cell fixation with 4% formaldehyde, Sweden), added as 15% hybridoma supernatant, for 90 min at 37˚C. After 1 mg/ml Hoechst 33342 was added, and samples were mounted on glass sensitization, cells were washed, resuspended, and counted for subsequent coverslips. Fluorescence images are shown next to the corresponding use. A total of 100 ng/ml TNP-BSA (Biosearch Technologies, Novato, Nomarski images in which black spots represent formazan deposits. CA) were added for 20 min for basophil activation. Immunoblotting Confocal laser scanning microscopy Electrophoresis was conducted in Run Blue 4–20% SDS polyacrylamide Cells were either fixed with 4% formaldehyde, freshly prepared from gels (Expedeon, Cambridgeshire, U.K.), and the separated proteins were paraformaldehyde in PBS, or observed live using confocal laser scanning transferred onto polyvinylidene difluoride membranes (Millipore). The microscopy. The following dyes were used for DNA detection: 0.5 mg/ml membranes were incubated with blocking solution (PBS containing 0.1% PI, 0.5 mM SYTO13, 5 mM MitoSOX Red, and 1 mM Hoechst 33342 (all Tween 20 and 5% nonfat dry milk) overnight at 4˚C. The membranes were Invitrogen, Paisley, U.K.). Possible colocalization of granule and nuclear then incubated with anti-mouse gp91phox Ab (1:1000), in blocking solution proteins with extracellular DNA was determined using immunofluores- overnight at 4˚C. As a loading control, stripped filters were incubated with cence staining with mouse anti-human basogranulin (BB1; 1:100) (15), rat anti-GAPDH mAb (1:10,000; Millipore). Filters were washed in PBS– anti-mouse TUG8 (mMCP-8; 1:100), and anti-human histone (clone H11-4; Tween (0.1%) for 30 min and incubated with HRP-conjugated secondary 1:10,000) Abs. Mitochondrial DNA (mtDNA) was visualized by MitoSOX Ab (GE Healthcare, Buckinghamshire, U.K.) for 1 h at room temperature. Red (Invitrogen), added 10 min before fixation (16). Additional immu- The blots were developed by an ECL technique (ECL kit; GE Healthcare), nofluorescence analyses with the same Abs were performed on 5-mm according to the manufacturer’s instructions. 5316 EXTRACELLULAR DNA TRAP FORMATION BY BASOPHILS

Flow cytometry of extracellular DNA nets in a concentration-dependent manner The purity of human basophil preparations was controlled by expression (Fig. 1A, left panel). Basophils primed with IL-3 only, or alterna- analysis of lineage-associated markers using anti-FcεRIa, anti-IgE, anti– tively, stimulated with anti-FcεRIa Ab in the absence of IL-3 priming IL-3Ra, and anti-CCR3 Abs. For cell surface expression analysis of Hoxb8 showed no DNA release (Fig. 1A, right panels). Similarly, we ob- basophils and Hoxb8 neutrophils, anti–c-kit, anti-FcεRIa, anti–Gr-1 (Ly- served DNA release from IL-3–primed basophils following stimula- 6G/Ly-6C), and appropriate -matched control Abs were used. In- tion with C5a, eotaxin, or TSLP but not in the absence of IL-3 tracellular staining of these cells was performed according to the BD cell fix and permeabilization kit protocol (BD Biosciences). Anti-TUG8, anti– priming (Fig. 1B, left panel). Quantitative analysis revealed that 60– MMP-9, or isotype-matched control Abs were added. Cells were incu- 80% of the basophils demonstrated evidence for DNA release and no bated, followed by staining with appropriate secondary Abs (both 1 h, 4˚C). differences in efficacy were observed between anti-FcεRIa Ab, C5a, Cells were washed twice and analyzed by flow cytometry (FACSCalibur) eotaxin, and TSLP when optimal concentrations were used (Fig. 1B, and quantitated using FlowJo software (Ashland, OR). left panel). Interestingly, stimulation with fMLF, PAF, LPS, and LTA Nippostrongylus brasiliensis also was followed by the formation of extracellular net-like DNA- containing structures that, however, did not require IL-3 priming Third-stage larvae (L3) of N. brasiliensis were washed in PBS (37˚C) and injected intradermally (i.d.) (500 organisms) into mice. After 18 d, one half (Fig. 1B). DNase treatment removed the extracellular structures, in- of the mice received a second i.d. inoculation of N. brasiliensis, and were dicating that they indeed contain DNA (Supplemental Fig. 2A). sacrificed 2 d later; the other half were sacrificed 18 d after the first in- To further visualize the DNA released by human basophils, we oculation without the second inoculation (20). Skin tissues were removed, applied electron microscopy. Using transmission electron micros- fixed with 4% formaldehyde, and embedded in paraffin. Subsequently, copy, we observed multiple mitochondria in basophils (Fig. 1C, sections were prepared for staining analysis. Skin tissues from three in- dividual mice were analyzed for each group. upper panels). Upon activation of high-affinity IgE receptors, mi- Downloaded from tochondria and granules translocated toward the leading edge of the Quantitative PCR cells and nuclei remained intact during this process (Fig. 1C, lower To remove the extracellular DNA, stimulated cells were briefly vortexed, panels). Moreover, when analyzed by scanning electron micros- centrifuged at 1400 rpm for 2 min, and the cellular DNA then isolated with copy, such IL-3 primed and IgE receptor–activated basophils the genomic DNA extraction kit from Macherey–Nagel (Duren,€ Germany). showed evidence of DNA release, which was not seen in control Quantitative PCR (qPCR) assays were performed using iQ SYBR Green Ab treated cells (Fig. 1D). http://www.jimmunol.org/ Supermix (Bio-Rad Laboratories). Primers were as follows: for human mitochondrial cytochrome oxidase 1 (COX1)(59-GCC CCC GAT ATG Similar to NETs and EETs (10, 11), we expected colocalization GCG TTT CC- 39 and 59-GTT CAA CCT GTT CCT GCG CC-39) (21), for of the extracellular DNA with granule proteins following basophil mouse mitochondrial cytochrome oxidase I (Cox1)(59-GCC CCA GAT activation. We stained basophils with anti-basogranulin Ab (23, ATA GCA TTC CC-39 and 59-GTT CAT CCT GTT CCT GCT CC-39), for 24) and PI. Using confocal microscopy, we could demonstrate that 9 9 9 human 18S rDNA (5 -ATC CCT GAA AAG TTC CAG CA-3 and 5 -GCT basogranulin, a granule specifically expressed by baso- TAT GTC TTC ACA ACC CA-39), and for mouse 18s rDNA (59-ATC CCT GAG AAG TTC CAG CA-39 and 59-CCT GGG TCA TCA TCA GGC TT-39). phils, is released following activation of these cells, colocalizing Primers were synthesized by Microsynth (Balgach, Switzerland). Dupli- with the extracellular DNA (Fig. 1E, upper panels). In contrast, cate real-time PCRs were performed in the iQ5 Multicolor Real-time PCR histones were observed within the nucleus but not in association Detection System (Bio-Rad) with 25 ng extracted genomic DNA and 200 with the extracellular DNA (Fig. 1E, lower panels). Taken to- by guest on September 26, 2021 nM of each primer in 25 ml final PCR mix. Cycling variables were 3 min at 95˚C followed by 40 cycles of 10 s at 95˚C and 30 s at 55˚C. gether, these data strongly suggest that activated human basophils are able to form extracellular traps in vitro and that they contain RT-PCR DNA and basogranulin and, perhaps, additional proteins. For the NOX and phox family of proteins, RT-PCR was performed as Human basophils release mtDNA in a cell death–independent described previously (22). Reverse transcription of 2 mg RNA was per- formed using the Superscript II kit, according to the manufacturer’s manner instructions (Invitrogen). PCR was then performed using Taq DNA Poly- There have been conflicting reports regarding the source of DNA merase (Roche, Mannheim, Germany). Primers were as follows: NOX1 used for extracellular trap formation and whether cell death of (59-TTA ACA GCA CGC TGA TCC TG-39 and 59-CAC TCC AGT GAG ACC AGC AA-39) (353 bp), NOX2 (59-AAA TGG TGG CAT GGA TGA neutrophils and eosinophils accompanies trap formation (16, 25). To TT-39 and 59-GGG ATT GGG CAT TCC TTT AT-39) (475 bp), NOX3 (59- determine the source of released DNA, we performed quantitative CCA GGG CAG TAC ATC TTG GT-39 and 59-CCG TGT TTC CAG GGA PCR experiments on cellular DNA. We amplified COX1 as a unique GAG TA-39 (432 bp), NOX4 (59-AAC CGA ACC AGC TCT CAG AA-39 mitochondrial gene and 18S rDNA as a nuclear encoded gene and 59-CCC AAA TGT TGC TTT GGT TT-39) (417 bp), NOX5 (59-TAC CTC CTC GTG TGG CTT CT-39 and 59-GCT CAG AGG CAA AGA TCC (21). The qPCR data revealed that after IL-3 priming and sub- TG-39) (535 bp), p22phox (59-GTT TGT GTG CCT GCT GGA GT-39 and 59- sequent activation of IgE or C5a receptors, the mtDNA but not TGG GCG GCT GCT TGATGG T-39 (298 bp), p47phox (59-CCC AGC CAG the nuclear DNA (nDNA) content of basophils is reduced (Fig. 2A). CAC TAT GTG TA-39 and 59-CCC AGG TCT TCT CGT AGT CG-39)(453 Similarly, GM-CSF priming and activation of human neutrophils phox 9 9 9 bp), p67 (5 -CGA GGG AAC CAG CTG ATA GA-3 and 5 -CAT GGG with C5a also resulted in a reduced cellular content of mtDNA AAC ACT GAG CTT CA-39)(608bp),andGAPDH (59-CCC CTT CAT TGA CCT-39 and 59-GAG TCC TTC CAG GAT ACC-39 (300 bp). but not nDNA compared with unstimulated cells (Supplemental Fig. 2B). In contrast, PMA activation of human neutrophils was Statistical analysis not associated with a reduction of the mtDNA content, at least Data are presented as mean levels 6 SEM. Statistical analysis was performed not within a 45-min time period (Supplemental Fig. 2B). Similar by two-tailed Student t or ANOVA with Bonferroni’s multiple comparison data also were observed using mouse neutrophils (Supplemental tests. All statistical analysis was done using Graph Pad Prism 5 software Fig. 2C). (La Jolla, CA). A p value , 0.05 was considered statistically significant. A possible role for cell death in DNA release was investigated by time-lapse video microscopy following calcein-AM and EthD-1 Results staining as well as analyzing ethidium bromide uptake by flow Both IgE-dependent and -independent stimulations of human cytometry. None of the physiological stimuli causing mtDNA basophils result in extracellular trap formation release led to an EthD-1 and ethidium bromide uptake, respectively, Activation of IL-3–primed human basophils through cross- at least not within a time period of 48 h (Fig. 2B). Such uptake, linking of high-affinity IgE receptors resulted in the formation however, was observed following exposure to high concentrations The Journal of Immunology 5317 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 1. Release of DNA from human basophils following physiological stimulation. (A) Confocal microscopy. DNA release after anti-FcεRIa stimulation requires IL-3 priming and is concentration dependent (n = 3). Right, Representative original data. Original magnification 3630. Bars, 10 mm. (B) Confocal microscopy. DNA release following non–IgE-dependent stimulation in the presence and absence of IL-3 priming (n = 4–7). Right, Representative original data. Original magnification 3630. Scale bars, 10 mm. (C) Transmission electron microscopy. Upper panels, Freshly isolated human blood basophils contain multiple mitochondria (arrowheads). Lower panels, Following IgE receptor activation of human basophils, granules and mitochondria are found in close proximity, near the plasma membrane. The nucleus is completely intact. (D) Scanning electron microscopy. DNA fibers (arrowheads) are visualized in IgE receptor–activated basophils. (E) Confocal microscopy. Upper panels, C5a-activated IL-3–primed basophils release DNA together with basogranulin (arrows). Lower panels, In contrast, the extracellular DNA does not bind detectable amounts of histones (arrows). Histones, however, are clearly seen in the nucleus. Representative images of three independent experiments are shown. Original magnification 3630. Scale bars, 10 mm. of PMA (10 mM) (Fig. 2B). In fact, PMA led to ∼50% cell death Human basophils release mtDNA in a ROS-dependent manner within 45 min. Lower PMA concentrations (100 nM), however, ROS have been shown to be essential for both NET and EET did not induce basophil death within 45 min, but only at later time formation (11, 17, 25). Therefore, we investigated whether baso- points, and IL-3 could, at least partially, protect basophils from phil extracellular trap (BET) formation is also ROS dependent. As PMA-induced cell death (Fig. 2B, lower right panel). with neutrophils and eosinophils, DPI completely blocked the 5318 EXTRACELLULAR DNA TRAP FORMATION BY BASOPHILS Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 2. BETs contain mtDNA and are formed by viable human basophils. (A) qPCR. Human basophils were activated as indicated and cellular DNA extracted for analysis (n =3).**p , 0.01, ***p , 0.001. For comparison, we performed the same experiments in human neutrophils (see Supplemental Fig. 2B). (B) Time-lapse microscopy and flow cytometry. No effect on human basophils viability was observed following short-term (45 min) and long-term (48 h) stimulation with the indicated agonists. Only high concentrations of PMA (10 mM) caused uptake of EthD-1 within 45 min. Representative images of three independent experiments are shown. Original magnification 3630. Scale bars, 10 mm. Flow cytometric viability data are presented as means 6 SEM (n =3). release of DNA from activated human basophils (Fig. 3A), sug- sequent experiments (Supplemental Fig. 3A, 3B). Taken together, gesting that ROS are also required for BET formation. We, these data suggest that basophils are able to generate ROS fol- therefore, measured ROS generation following activation of lowing their activation and subsequently allow BET formation. basophils in the presence and absence of DPI. Basophils primed with In contrast to neutrophils and eosinophils, basophils have been IL-3 and subsequently activated through high-affinity IgE or C5a reported to lack NADPH oxidase activity (26, 27). DPI is widely receptors showed increased ROS production compared with unsti- used as a general inhibitor of ROS production and blocks not only mulated or DPI-treated cells (Fig. 3B). The blocking effect of DPI on NADPH oxidase activity, but also the enzymatic activity of several ROS production by basophils was concentration dependent. Com- flavoproteins that are part of the mitochondrial oxidation pathway plete inhibition of activation-induced ROS production consistently (28). We analyzed mRNA expression of the NADPH oxidase was observed at 5 mM DPI, a concentration that was used in all sub- (NOX) family of proteins as well as p22phox, p47phox, and p67phox. The Journal of Immunology 5319

FIGURE 3. Human basophil extracellular DNA trap release is dependent on mitochondrial ROS. (A) Confocal microscopy. Preincubation with DPI completely blocked DNA release from basophils. Representative images of three independent experi- ments are shown. Original magnification 3630. Bars, 10 mm. (B) Total ROS production is in- creased in IL-3–primed basophils following high- affinity IgE-receptor- or C5a receptor activation. Increases in ROS can be blocked by DPI (n = 4). (C) RT-PCR. Although basophils express NOX2, the p47phox and p67phox subunits are missing. Freshly isolated human basophils and neutrophils served as sources for mRNA preparation. Data are representative of three independent experiments. (D) NBT assay. Purified cells were allowed to ad- Downloaded from here to glass coverslips and activated as indicated. Slides were counterstained with Diff-Quik stain. Although activated neutrophils demonstrate dark formazan deposits (arrows), activated basophils from the same donor show no evidence for respi- ratory burst activity. Original magnification 3630. Scale bars, 10 mm. (E) Mitochondrial ROS pro- http://www.jimmunol.org/ duction that can be blocked by MitoQ is increased in IL-3–primed basophils following high-affinity IgE-receptor and C5a receptor activation (n = 3). *p , 0.05, **p , 0.01. (F) Confocal microscopy. Thirty-minute preincubation with MitoQ completely blocked DNA release from basophils. Represen- tative images of three independent experiments are shown. Original magnification 3630. Scale bars, 10 mm. by guest on September 26, 2021

Basophils did not express two essential components of NADPH blocked by the flavoprotein inhibitor MitoQ (29, 30) (Fig. 3E). cytoplasmic ROS production pathway, namely p47phox and p67phox Furthermore, these data support our assumption that basophils are (Fig. 3C). able to generate mitochondrial ROS upon activation. Moreover, We also measured the enzymatic activity of the NADPH oxidase ROS generated by mitochondria seem to be required for BET in basophils and neutrophils by NBT assay and observed intracel- formation, as MitoQ also blocked DNA release from IL-3 primed lular dark precipitates indicating activity in acti- and C5a-stimulated basophils (Fig. 3F). vated neutrophils, but not basophils. Even stimulation with PMA did not induce any detectable respiratory burst activity (Fig. 3D). To BETs are present in human skin undergoing an inflammatory directly demonstrate the release of DNA in the absence of any de- response tectable respiratory burst activity in single basophils, we combined We investigated whether BETs can be detected in skin tissues as the NBT assay with cellular DNA staining using the fluorescent dyes a possible consequence of allergic, autoimmune, or other in- MitoSOX Red and Hoechst 33342. Combining confocal laser flammatory conditions. We first counted basophil numbers using scanning and Nomarski (DIC) microscopy allowed the visualization anti-basogranulin immunofluorescence analysis. Bullous pemphi- of activated basophils releasing DNA in the absence of any detect- goid, folliculitis, and Wells’ syndrome skin biopsies able respiratory burst activity. In contrast, with activated neutrophils showed higher basophil numbers within inflammatory infil- releasing DNA, the formation of formazan crystals was clearly ev- trates compared with or urticarial skin lesions ident (Supplemental Fig. 3C). Together, these data strongly suggest (Fig. 4A). In these tissues, we subsequently searched for BETs. that ROS generation and DNA release by activated basophils is Extracellular DNA was detected in association with basogranulin NADPH oxidase independent. in specimens of and Wells’ syndrome, but We subsequently measured mitochondrial ROS activity using also in atopic dermatitis, urticaria, and eosinophilic folliculitis. MitoSOX Red, a fluorescent dye that specifically reflects mito- The fraction of DNA + basogranulin - releasing basophils usually chondrial ROS activity (18). We observed that the increased ranged between 15 and 40% (Fig. 4B, left panel). Similar to EETs fluorescent intensity in IL-3–primed human basophils following in these diseases (31), the long, thin DNA structures often seemed stimulation of high-affinity IgE or C5a receptors was completely to connect the releasing basophils with other cells, including other 5320 EXTRACELLULAR DNA TRAP FORMATION BY BASOPHILS

IL-3 priming was required. On the other hand, these mouse basophils have been constantly cultured in the presence of IL-3 during the in vitro differentiation process. Furthermore, extra- cellular DNA released by activated mouse basophils colocalized with the granule protein TUG8 (Fig. 5B). In addition, we dem- onstrated, by determining the amounts of the Cox1 and 18s rDNA genes by qPCR before and after basophil stimulation, that the source of DNA release was the mitochondria. Similar to the data observed in the human system (Fig. 2A), activated mouse basophils demonstrated strong evidence for release of mtDNA, but not nDNA (Fig. 5C). We next analyzed the release of DNA in Hoxb8 basophils and Hoxb8 neutrophils derived from NOX2 (p91phox)-deficient mice. Immunoblot analysis showed that indeed both wild-type and NOX2- deficient Hoxb8 basophils express no detectable NOX2 protein. In contrast, wild-type but not NOX2-deficient Hoxb8 neutrophils ex- press NOX2 (Supplemental Fig. 4A). In addition, mouse basophils, like human basophils, did not show any respiratory burst activity

upon activation (Supplemental Fig. 4B). Whereas the lack of NADPH Downloaded from oxidase completely prevented DNA release from activated neu- trophils, genetic lack of NOX2 did not prevent either IgE-dependent or -independent DNA release from basophils (Fig. 5D). We also demonstrate the release of DNA in mouse basophils lacking any respiratory burst activity, a finding in contrast to the result with mouse

neutrophils (Supplemental Fig. 4C). Taken together, these data con- http://www.jimmunol.org/ firm our results with human basophils suggesting that BET formation occurs independent of NADPH oxidase–derived ROS. To explore whether mouse basophils can also form extracellular traps under in vivo conditions, we used an infection model with N. brasiliensis, which is known to induce basophil infiltration (20). All mice received an i.d. inoculation with N. brasiliensis. One set of mice received a second inoculation and were sacrificed 2 d later. The other set was left without a second inoculation. We stained the skin tissues with PI and anti-TUG8 Ab, and observed by guest on September 26, 2021 strong basophil infiltration in tissues that received two i.d. inoc- ulations with N. brasiliensis compared with the tissues that were FIGURE 4. Basophil infiltration and extracellular trap formation in infected one time only (Fig. 6A). Moreover, following two hel- A inflammatory skin diseases. ( ) Immunofluorescence. Basophils were minth infections, .30% of the infiltrating basophils demonstrated identified using anti-basogranulin Ab and their infiltration in the indicated evidence for the formation of BETs, which, furthermore, showed inflammatory skin diseases quantified (n = 4–11). *p , 0.05. Right, Representative original data. Original magnification 3630. Scale bars, 10 colocalization of DNA with mMCP-8 (Fig. 6B). mm. (B) Confocal microscopy. The relative fractions of basophils forming BETs were determined in the indicated inflammatory skin diseases (n = 3–4). Discussion *p , 0.05; **p , 0.01; ***p , 0.001. Right, Representative original data. Extracellular DNA traps generated by neutrophils and eosinophils Arrows point to BETs. Original magnification 3630. Bars, 10 mm. have been implicated in innate immune defense mechanisms against (10, 11, 17, 32, 33). Our study demonstrates for the first basophils (Fig. 4B, right panels). In normal skin, we identified time, to our knowledge, that physiological activation of human and a few basophils, but no BET formation. mouse basophils can also lead to the rapid formation of extracellular DNA traps, both in vitro and in vivo. We identified cytokine, IgE, Activated mouse basophils also form extracellular traps Toll-like, , and lipid mediator receptors that, upon activa- in vitro and in vivo tion, induced BET formation within minutes. Some of the receptor- We used conditional mouse Hoxb8-immortalized progenitor cells mediated signaling mechanisms appeared to be effective in IL-3 to generate large amounts of mouse basophils in vitro (13). For primed basophils only, others required no priming. Because BETs comparison, we established another Hoxb8-immortalized pro- appear to contain basophil granule proteins, it seems unsurprising genitor cell line to produce large numbers of mouse neutrophils as that most physiological agonists able to trigger BET formation are previously described (13, 14). Upon differentiation, we charac- known triggers of basophil (15, 34–36). terized the immunophenotype of both Hoxb8 basophils and Using microscopic (staining of mtDNA, live cell imaging) and Hoxb8 neutrophils by flow cytometry. Mature mouse basophils, molecular biological techniques (PCR amplification of mito- but not neutrophils, expressed the FcεRI (FcεRIa) as well as TUG8 chondrial and nuclear genes, in situ hybridization), we have pre- (mMCP-8) (Supplemental Fig. 1B). In contrast, neutrophils, but not viously shown that both viable neutrophils and eosinophils released basophils, expressed Gr-1 (Ly-6G/Ly-6C) (Supplemental Fig. 1B). mtDNA rather than nDNA (11, 16, 17). In this paper, we show by MMP-9 was expressed by both (Supplemental Fig. 1B). electron microscopy that basophils contain mitochondria that, As with human basophils, 60–80% mature mouse Hoxb8 baso- upon cell activation, accumulate at the leading edge in close phils released DNA upon activation of C5a or high-affinity IgE proximity to the plasma membrane in the absence of any visible receptors (Fig. 5A). In contrast to human basophils, however, no nuclear DNA involvement. Moreover, we observed the projection The Journal of Immunology 5321 Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021

FIGURE 5. Activated mouse basophils form functional extracellular traps in vitro and in vivo. (A) Confocal microscopy. Hoxb8 basophils release DNA following activation of high-affinity IgE or C5a receptors in the absence of IL-3 priming (n = 3–4). **p , 0.01; ***p , 0.001. Right, Representative original data. Original magnification 3630. Scale bars, 10 mm. (B) Confocal microscopy. Hoxb8 basophils were activated as indicated (total 45 min). Activated basophils release DNA together with MCP-8 (TUG8). Representative images of three independent experiments are shown. Original magnifi- cation 3630. Scale bars, 10 mm. (C) Quantitative PCR. Hoxb8 basophils were activated as indicated and cellular DNA extracted for analysis (n = 4). **p , 0.01. For comparison, we performed the same experiments in Hoxb8 neutrophils (Supplemental Fig. 2C). (D) Confocal microscopy. DNA release from NOX2-deficient Hoxb8 basophils (knockout [KO] NOX2 basophils) following high-affinity IgE or C5a receptor stimulation. For comparison, we performed similar experiments in NOX2-deficient Hoxb8 neutrophils (KO NOX2 neutrophils), which were unable to release DNA, even after 100 nM PMA stimulation (n = 3). Right, Representative original data. Original magnification 3630. Scale bars, 10 mm. of thin DNA fibers of activated basophils. The exact molecular To compare the amounts of cellular mtDNA and nDNA before mechanisms responsible for the generation of these DNA fibers and after activation, we applied a quantitative PCR method. We remain to be investigated. detected a significant depletion of mtDNA, but not of nDNA, within

FIGURE 6. Basophil infiltration and ex- tracellular trap formation in mouse inflamed skin. (A) Immunofluorescence. Basophils were identified in N. brasiliensis (Nb)–infec- ted skin using anti-mMCP-8 (TUG8) Ab and the infiltration quantified over time (n =3). ***p , 0.001. (B) Confocal microscopy. Relative fraction of basophils forming BETs was determined in Nb-infected skin (n =3–4). ***p , 0.001. Right, Representative original data. Arrows point to BETs. Original mag- nification 3630. Scale bars, 10 mm. 5322 EXTRACELLULAR DNA TRAP FORMATION BY BASOPHILS

45 min after basophil activation. Moreover, although the majority functional NADPH oxidase. The subsequent signaling events of the basophils demonstrated evidence of mtDNA release, 85–90% initiated by ROS, which then lead to extracellular trap formation, of the mtDNA was still detectable in the whole cell population remain to be determined. even after activation. Similar results were obtained using human It has previously been reported that the formation of extracel- and mouse neutrophils. This suggests that the majority of gran- lular traps contributes to antibacterial defense in vitro and in vivo ulocytes still contain functional mitochondria after BET forma- (10, 11). Since basophils express TLRs (6, 36) and release anti- tion, explaining, at least partially, why the formation of extracellular microbial factors (9), they had been implicated as an antibacterial traps does not limit the life span of the cells responsible for their defense. Whether BETs are able to mediate an antibacterial ac- generation (11). Thus, we report in this paper a sensitive method to tivity in the extracellular space remains to be investigated. detect both the amount and source of DNA released in association We recently developed a technique for generating mouse baso- with extracellular trap formation. phils in nearly unlimited numbers (13). With this method, myeloid There have been conflicting reports regarding the mechanism(s) progenitor cells are conditionally immortalized using Hoxb8 in the of extracellular trap formation in both neutrophils and eosinophils. presence of IL-3. Outgrowing cell lines were selected for their Although some authors assume that a special form of cell death is potential to differentiate into basophils upon shutdown of Hoxb8 required (NETosis in neutrophils (25); EETosis for eosinophils (37), expression. Owing to this technical development, we were enabled other groups have either demonstrated that cell death is not required to study the mechanisms required for functional BET formation for or have not seen cell death at all in association with extracellular also in the mouse system. The results of our study support the trap formation (11, 16, 17). The reason for these discrepancies may assumption that Hoxb8 basophils closely reflect the biological lie in the different systems used for cell activation. From our point characteristics of primary mouse basophils. Downloaded from of view, non-physiological activation with PMA and/or ionomycin Previously published work, performed in mice, suggested that should be avoided in studying the formation of extracellular traps, basophils mediate adaptive immunity against helminthes (3). The and hence, conclusions based on such in vitro approaches are un- identification of BETs in association with N. brasiliensis infection convincing. Although we agree that within an inflammatory re- under in vivo conditions suggests that basophils might be able to sponse both extracellular trap formation and cell death can occur exert direct innate immune effector functions against pathogens.

concurrently, it has never been convincingly demonstrated that these Therefore, our data provide additional evidence for a multifaceted http://www.jimmunol.org/ phenomena depend on each other (32, 33). It should be noted that and significant role for basophils in both innate and adaptive human basophils treated with high concentrations of monosodium immune responses to infectious agents. urate crystal (MSU) (200 mg/ml) for 4 h in serum-free medium showed evidence for nDNA release, but no details regarding the Acknowledgments molecular mechanism have been provided (38). We thank Evelyne Kozlowski and Werner Graber for excellent technical In contrast to a requirement for cell death, there seems to be assistance (confocal and electron microscopy). consensus among researchers in the field that the generation of ROS is required for both NET and EET formation (11, 39). Neutrophils and eosinophils derived from patients suffering from chronic Disclosures by guest on September 26, 2021 granulomatous disease are unable to form extracellular traps (11, The authors have no financial conflicts of interest. 39), suggesting that a functional NADPH oxidase is required. Using a pharmacological approach, we obtained data suggesting that the formation of a BET also represents a ROS-dependent References 1. Karasuyama, H., K. Obata, T. Wada, Y. Tsujimura, and K. Mukai. 2011. Newly process. On the other hand, basophils have been reported to lack appreciated roles for basophils in and protective immunity. Allergy 66: a functional NADPH oxidase (27). Perhaps, as a consequence, 1133–1141. basophils do not seem to be capable of intracellular bacterial 2. Obata, K., K. Mukai, Y. Tsujimura, K. Ishiwata, Y. Kawano, Y. Minegishi, N. Watanabe, and H. Karasuyama. 2007. Basophils are essential initiators of killing (26). For other basophil functions, however, a role for ROS a novel type of chronic allergic inflammation. Blood 110: 913–920. has been demonstrated (40, 41). In such a situation, we decided to 3. Ohnmacht, C., C. Schwartz, M. Panzer, I. Schiedewitz, R. Naumann, and reanalyze the expression of the NADPH oxidase in human baso- D. Voehringer. 2010. Basophils orchestrate chronic allergic dermatitis and pro- tective immunity against helminths. Immunity 33: 364–374. phils and observed that these cells, in contrast to human neu- 4. Valent, P., and C. A. Dahinden. 2010. Role of in the regulation of trophils, do not express p47phox and p67phox, two essential subunits basophil development and secretion. Curr. Opin. Hematol. 17: 60–66. phox 5. Siracusa, M. C., S. A. Saenz, D. A. Hill, B. S. Kim, M. B. Headley, of the (42, 43). Moreover, NOX2 (p91 ) deficiency T. A. Doering, E. J. Wherry, H. K. Jessup, L. A. Siegel, T. Kambayashi, et al. prevented the release of DNA from mouse neutrophils, but not 2011. TSLP promotes -3‑independent basophil and from mouse basophils. These data suggest that the formation of type 2 inflammation. Nature 477: 229–233. 6. Komiya, A., H. Nagase, S. Okugawa, Y. Ota, M. Suzukawa, A. Kawakami, BET depends on the generation of ROS, which, however, is not T. Sekiya, K. Matsushima, K. Ohta, K. Hirai, et al. 2006. Expression and generated by NADPH oxidase. function of Toll-like receptors in human basophils. Int. Arch. Allergy Immunol. Indeed, using MitoSOX Red, a mitochondrial ROS fluorescent 140(Suppl. 1): 23–27. 7. Stone, K. D., C. Prussin, and D. D. Metcalfe. 2010. IgE, mast cells, basophils, dye (18), we obtained evidence for mitochondrial and eosinophils. J. Allergy Clin. Immunol. 125(2, Suppl 2)S73–S80. production upon basophil activation. We, therefore, applied 8. Melioli, G., G. Passalacqua, C. E. Baena-Cagnani, and G. W. Canonica. 2012. and bacteria interaction in the induction of basophil activation: is this MitoQ, which has been characterized as a specific inhibitor of the lost ring between allergy and infections in pediatric patients? Curr. Opin. mitochondrial ROS generation (29, 30). Pretreatment of basophils Allergy Clin. Immunol. 12: 164–170. with MitoQ decreased ROS activity in stimulated basophils and 9. Chen, K., W. Xu, M. Wilson, B. He, N. W. Miller, E. Bengte´n, E. S. Edholm, P. A. Santini, P. Rath, A. Chiu, et al. 2009. enhances immune completely blocked agonist-induced DNA release. The question surveillance by activating , proinflammatory and -stimulating remains, why both neutrophils and eosinophils require NADPH programs in basophils. Nat. Immunol. 10: 889–898. oxidase activity for extracellular trap formation and basophils do 10. Brinkmann, V., U. Reichard, C. Goosmann, B. Fauler, Y. Uhlemann, D. S. Weiss, Y. Weinrauch, and A. Zychlinsky. 2004. Neutrophil extracellular traps kill not. We speculate at this point, that basophils contain higher bacteria. Science 303: 1532–1535. numbers of mitochondria compared with neutrophils and eosino- 11. Yousefi, S., J. A. Gold, N. Andina, J. J. Lee, A. M. Kelly, E. Kozlowski, I. Schmid, A. Straumann, J. Reichenbach, G. J. Gleich, and H. U. Simon. 2008. phils, allowing them to generate the relatively high concentrations Catapult-like release of mitochondrial DNA by eosinophils contributes to anti- of ROS required for BET formation even in the absence of a bacterial defense. Nat. Med. 14: 949–953. The Journal of Immunology 5323

12. Mihalache, C. C., S. Yousefi, S. Conus, P. M. Villiger, E. M. Schneider, and proton channels in IgE-mediated activation of human basophils. Proc. Natl. H. U. Simon. 2011. Inflammation-associated autophagy-related programmed Acad. Sci. USA 105: 11020–11025. necrotic death of human neutrophils characterized by organelle fusion events. J. 28. Maghzal, G. J., K. H. Krause, R. Stocker, and V. Jaquet. 2012. Detection of Immunol. 186: 6532–6542. reactive oxygen species derived from the family of NOX NADPH oxidases. Free 13. Gurzeler, U., T. Rabachini, C. A. Dahinden, M. Salmanidis, G. Brumatti, Radic. Biol. Med. 53: 1903–1918. P. G. Ekert, N. Echeverry, D. Bachmann, H. U. Simon, and T. Kaufmann. 2013. 29. Kelso, G. F., C. M. Porteous, C. V. Coulter, G. Hughes, W. K. Porteous, In vitro differentiation of near-unlimited numbers of functional mouse basophils E. C. Ledgerwood, R. A. J. Smith, and M. P. Murphy. 2001. Selective targeting using conditional Hoxb8. Allergy 68: 604–613. of a redox-active ubiquinone to mitochondria within cells: antioxidant and 14. Wang, G. G., K. R. Calvo, M. P. Pasillas, D. B. Sykes, H. Ha¨cker, and antiapoptotic properties. J. Biol. Chem. 276: 4588–4596. M. P. Kamps. 2006. Quantitative production of or neutrophils ex 30. Bulua, A. C., A. Simon, R. Maddipati, M. Pelletier, H. Park, K.-Y. Kim, vivo using conditional Hoxb8. Nat. Methods 3: 287–293. M. N. Sack, D. L. Kastner, and R. M. Siegel. 2011. Mitochondrial reactive 15. Mochizuki, A., A. R. McEuen, M. G. Buckley, and A. F. Walls. 2003. The re- oxygen species promote production of proinflammatory cytokines and are ele- lease of basogranulin in response to IgE-dependent and IgE-independent stimuli: vated in TNFR1-associated periodic syndrome (TRAPS). J. Exp. Med. 208: 519– validity of basogranulin measurement as an indicator of basophil activation. J. 533. Allergy Clin. Immunol. 112: 102–108. 31. Simon, D., S. Hoesli, N. Roth, S. Staedler, S. Yousefi, and H. U. Simon. 2011. 16. Yousefi, S., C. Mihalache, E. Kozlowski, I. Schmid, and H. U. Simon. 2009. Eosinophil extracellular DNA traps in skin diseases. J. Allergy Clin. Immunol. Viable neutrophils release mitochondrial DNA to form neutrophil extracellular 127: 194–199. traps. Cell Death Differ. 16: 1438–1444. 32. Yousefi, S., D. Simon, and H. U. Simon. 2012. Eosinophil extracellular DNA 17. Morshed, M., S. Yousefi, C. Sto¨ckle, H. U. Simon, and D. Simon. 2012. Thymic traps: molecular mechanisms and potential roles in disease. Curr. Opin. Immu- stromal lymphopoietin stimulates the formation of eosinophil extracellular nol. 24: 736–739. traps. Allergy 67: 1127–1137. 33. Simon, D., H. U. Simon, and S. Yousefi. 2013. Extracellular DNA traps in al- 18. Robinson, K. M., M. S. Janes, and J. S. Beckman. 2008. The selective detection lergic, infectious, and autoimmune diseases. Allergy 68: 409–416. of mitochondrial superoxide by live cell imaging. Nat. Protoc. 3: 941–947. 34. Brunner, T., C. H. Heusser, and C. A. Dahinden. 1993. Human peripheral blood 19. Pollock, J. D., D. A. Williams, M. A. C. Gifford, L. L. Li, X. Du, J. Fisherman, basophils primed by (IL-3) produce IL-4 in response to immuno- S. H. Orkin, C. M. Doerschuk, and M. C. Dinauer. 1995. Mouse model of X- globulin E receptor stimulation. J. Exp. Med. 177: 605–611. linked chronic granulomatous disease, an inherited defect in super- 35. Ochensberger, B., G. C. Daepp, S. Rihs, and C. A. Dahinden. 1996. Human Downloaded from oxide production. Nat. Genet. 9: 202–209. blood basophils produce interleukin-13 in response to IgE-receptor-dependent 20. Obata-Ninomiya, K., K. Ishiwata, H. Tsutsui, Y. Nei, S. Yoshikawa, Y. Kawano, and -independent activation. Blood 88: 3028–3037. Y. Minegishi, N. Ohta, N. Watanabe, H. Kanuka, and H. Karasuyama. 2013. The 36. Bieneman, A. P., K. L. Chichester, Y. H. Chen, and J. T. Schroeder. 2005. Toll- skin is an important bulwark of acquired immunity against intestinal helminths. like receptor 2 ligands activate human basophils for both IgE-dependent and J. Exp. Med. 210: 2583–2595. IgE-independent secretion. J. Allergy Clin. Immunol. 115: 295–301. 21. Coˆte´, H. C. F., Z. L. Brumme, K. J. P. Craib, C. S. Alexander, B. Wynhoven, 37. Ueki, S., R. C. Melo, I. Ghiran, L. A. Spencer, A. M. Dvorak, and P. F. Weller. L. Ting, H. Wong, M. Harris, P. R. Harrigan, M. V. O’Shaughnessy, and 2013. Eosinophil extracellular DNA trap cell death mediates lytic release of free

J. S. Montaner. 2002. Changes in mitochondrial DNA as a marker of nucleoside secretion-competent eosinophil granules in humans. Blood 121: 2074–2083. http://www.jimmunol.org/ toxicity in HIV-infected patients. N. Engl. J. Med. 346: 811–820. 38. Schorn, C., C. Janko, M. Latzko, R. Chaurio, G. Schett, and M. Herrmann. 2012. 22. Li, B., K. Bedard, S. Sorce, B. Hinz, M. Dubois-Dauphin, and K. H. Krause. Monosodium urate crystals induce extracellular DNA traps in neutrophils, 2009. NOX4 expression in human leads to constitutive generation of eosinophils, and basophils but not in mononuclear cells. Front. Immunol. 3: 277. reactive oxygen species and to constitutive IL-6 expression. J. Innate Immun. 1: 39. Bianchi, M., A. Hakkim, V. Brinkmann, U. Siler, R. A. Seger, A. Zychlinsky, and 570–581. J. Reichenbach. 2009. Restoration of NET formation by gene therapy in CGD 23. McEuen, A. R., J. Calafat, S. J. Compton, N. J. W. Easom, M. G. Buckley, controls . Blood 114: 2619–2622. E. F. Knol, and A. F. Walls. 2001. Mass, charge, and subcellular localization of 40. Yoshimaru, T., Y. Suzuki, T. Matsui, K. Yamashita, T. Ochiai, M. Yamaki, and a unique secretory product identified by the basophil-specific BB1. J. K. Shimizu. 2002. Blockade of superoxide generation prevents high-affinity Allergy Clin. Immunol. 107: 842–848. receptor-mediated release of allergic mediators by rat mast 24. Agis, H., M. T. Krauth, A. Bo¨hm, I. Mosberger, L. Mullauer,€ I. Simonitsch- cell line and human basophils. Clin. Exp. Allergy 32: 612–618. Klupp, A. F. Walls, H. P. Horny, and P. Valent. 2006. Identification of basog- 41. Kepley, C. L., F. T. Lauer, J. M. Oliver, and S. W. Burchiel. 2003. Environmental

ranulin (BB1) as a novel immunohistochemical marker of basophils in normal polycyclic aromatic hydrocarbons, benzo(a)pyrene (BaP) and BaP-quinones, by guest on September 26, 2021 and patients with myeloproliferative disorders. Am. J. Clin. Pathol. enhance IgE-mediated release and IL-4 production in human baso- 125: 273–281. phils. Clin. Immunol. 107: 10–19. 25. Fuchs, T. A., U. Abed, C. Goosmann, R. Hurwitz, I. Schulze, V. Wahn, 42. El-Benna, J., P. M. C. Dang, and M. A. Gougerot-Pocidalo. 2008. Priming of the Y. Weinrauch, V. Brinkmann, and A. Zychlinsky. 2007. Novel cell death program neutrophil NADPH oxidase activation: role of p47phox phosphorylation and leads to neutrophil extracellular traps. J. Cell Biol. 176: 231–241. NOX2 mobilization to the plasma membrane. Semin. Immunopathol. 30: 279– 26. de Boer, M., and D. Roos. 1986. Metabolic comparison between basophils and 289. other leukocytes from human blood. J. Immunol. 136: 3447–3454. 43. de Mendez, I., M. C. Garrett, A. G. Adams, and T. L. Leto. 1994. Role of p67- 27. Musset, B., D. Morgan, V. V. Cherny, D. W. MacGlashan, Jr., L. L. Thomas, phox SH3 domains in assembly of the NADPH oxidase system. J. Biol. Chem. E. Rı´os, and T. E. DeCoursey. 2008. A pH-stabilizing role of voltage-gated 269: 16326–16332.