Fish cast NETs: Neutrophil extracellular traps are released from fish neutrophils

ARTICLE IN PRESS

Developmental and Comparative Immunology 31 (2007) 805–816

Developmental & Comparative Immunology www.elsevier.com/locate/devcompimm

Fish cast NETs: Neutrophil extracellular traps are released from fish neutrophils Dusˇ an Palic´a,d,, Jelena Ostojic´b, Claire B. Andreasenc, James A. Rotha a

Department of Veterinary Microbiology and Preventive Medicine, College of Veterinary Medicine, Iowa State University, Ames, IA, USA b Department of Genetics, Development and Cell Biology, College of Agriculture, Iowa State University, Ames, IA, USA c Department of Veterinary Pathology, College of Veterinary Medicine, Iowa State University, Ames, IA, USA d Department of Natural Resource Ecology and Management, College of Agriculture, Iowa State University, Ames, IA, USA Received 10 August 2005; received in revised form 22 October 2006; accepted 20 November 2006 Available online 29 December 2006

Abstract Neutrophil extracellular traps (NETs), which are extracellular DNA structures released from neutrophils, are described and characterized for the first time in fish using fluorescent confocal microscopy. Confocal images of fish neutrophil suspensions stained with 60 -diamino-2-phenylindole, dihydrochloride DNA fluorescent stain (DAPI) revealed the presence of NETs which appeared as fibrous structures connecting several cells. Co-localization of NETs with neutrophil granular proteins and actin was investigated using specific antibodies and probes. Double staining of neutrophils with SYTOX green and DAPI revealed that SYTOX stain applied to living cells stained extracellular DNA, but not nuclei. NETs are actively released from stimulated living cells, associated with granular proteins, but not with cytoskeleton, and are not a product of nuclear degradation seen in late apoptotic stages. Additionally, a fluorometric microtiter plate assay to quantify the release of NETs was adopted for use with fish neutrophils, and the effect of stress on NETs release was studied. This assay detected the inhibition of DNA release during stress conditions. In summary, NETs were released from living fish kidney neutrophils upon stimulation, characterized using fluorescence DNA-binding dyes, specific antibodies and probes, and quantified using a microtiter plate fluorometric assay that can rapidly measure a large number of samples. Detection of NETs can be used as an additional assay to an existing battery of functional tests, and as a new research model to study the effects of stress, immunomodulators, and diseases. r 2006 Elsevier Ltd. All rights reserved. Keywords: Fathead minnows; NETs release assay; Stress; Fluorescent confocal microscopy; Neutrophil extracellular traps; Fish; Degranulation

1. Introduction Abbreviations: NETs, neutrophil extracellular traps; MPO, myeloperoxidase; DIC, differential interference contrast; DAPI, 40 , 60 -diamino-2-phenylindole, dihydrochloride. Corresponding author. Department of Veterinary Microbiology and Preventive Medicine, College of Veterinary Medicine, Iowa State University, Ames, IA, USA. Tel.: +1 515 2947661; fax: +1 515 2947874. E-mail address: [email protected] (D. Palic´).

Neutrophils are an important component of host defense against many bacterial, viral and fungal infections, and the evaluation of neutrophil function is valuable for assessment of the health status of individuals and animal populations [1]. In response to inflammatory stimuli, neutrophils migrate from

0145-305X/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.dci.2006.11.010

ARTICLE IN PRESS 806

D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

the circulating blood to infected tissues, where they efficiently bind, engulf, and kill bacteria by proteolytic enzymes, antimicrobial proteins, and reactive oxygen species [1,2]. Neutrophils also degranulate, releasing antimicrobial factors into the extracellular medium [2]. Fish neutrophils have morphological, histochemical, and functional similarities to mammalian neutrophils [3–5]. Fish neutrophils have phagocytic, chemotactic and bactericidal functions, an intense respiratory burst, peroxidase (myeloperoxidase—MPO) activity, and are capable of degranulation of primary granules [6–8]. A previously unknown neutrophil defense mechanism has been recently described in mammalian neutrophils, whereby stimulated neutrophils generate extracellular fibers composed of granular and nuclear constituents that form neutrophil extracellular traps (NETs) which are complex structures responsible for trapping and extracellular killing of different bacteria [9]. Major structural components of NETs are DNA, histones (H1, H2A, H2B, H3 and H4), and proteins from granules (elastase, MPO, cathepsin G, lactoferrin, and gelatinase), but NETs are not associated with granule membrane protein CD63, and cytoplasmic markers such as actin, tubulin, or annexin I [9]. NETs are not the result of leakage during cellular disintegration since activated neutrophils exclude vital dyes at least 2 h after the release of NETs, and only mature neutrophils are capable of producing NETs [9,10]. The membrane-impermeable DNA dye (SYTOX Green) is rigorously excluded from a variety of prokaryotic and eukaryotic cells and is often used to distinguish dead mammalian cells from live cells in mixed populations, or necrotic from apoptotic cells by flow cytometry, based on its exclusion from apoptotic cells and effective staining of necrotic cells. SYTOX–DNA complex exhibits strong fluorescence, while unbound SYTOX dye has very low fluorescence [11]. Exclusion of SYTOX dye from living cells allows for staining and quantification of released extracellular DNA from NETs using spectrofluorometry [9,10]. The DNA quantification assay was adapted for use with fathead minnow (Pimephales promelas Rafinesque, [1820]) kidney neutrophils using a microtiter plate method described for detection of degranulation [12]. Existence of NETs has not been demonstrated in fish. The fathead minnow (P. promelas, Rafinesque, 1820) is widely used as a laboratory model due to their aquacultural and ecological relevance, and the availability of functional assays for immunological

research [13,14]. In this study, confocal and light microscopy were used to demonstrate presence of NETs in activated fathead minnow neutrophils, and investigate co-localization of NETs with granular and cytoskeletal proteins, and a rapid, direct extracellular DNA detection assay was used to measure NET release from fish neutrophils. Also, markedly elevated, as well as chronically increased concentrations of cortisol, act as inhibitors of neutrophil function, and a significant decrease in degranulation of fish neutrophil primary granules have been observed after handling and crowding stress, which is often encountered in aquaculture [14]. Acute stress effects on NET release have not been previously reported, and in this study, the degranulation assay and DNA release assay were used to measure the effects of acute crowding and handling stress on neutrophils, and to examine the potential use of a DNA release assay in evaluation of neutrophil function. 2. Materials and methods 2.1. Fish Adult fathead minnows with an average weight of 3 g were maintained in the Department of Natural Resource Ecology and Management, Iowa State University, Ames, IA, USA. Fish were held in 300–1000 L tank recirculation system supplied with dechlorinated tap water at 20 1C and fed daily with dried flake food (Aquatoxs, Ziegler Bros Inc., PA, USA). Fathead minnows were cared for in accordance with approved Iowa State University animal care guidelines. 2.2. Reagents Reagents for the degranulation assay were prepared as previously described [12]. In short, calcium ionophore A23187 (CaI, Sigma; final concentration of 5 mg mL1), 3,30 , 5,50 -tetramethylbenzidine hydrochloride (TMB, Sigma; 2.5 mM in water), and hydrogen peroxide (H2O2, 5 mM in water) were prepared. Sulfuric acid (H2SO4, Fisher; 2 M) was used as a reaction stop solution. The detergent cetyltrimethylammonium bromide (CTAB, Sigma; 0.02% in water) was used as a lysing agent for determining total MPO content of neutrophils (100%). For the DNA release assay, stock solutions of cytochalasin B (cyto B, Sigma; 1 mg mL1), phorbol

ARTICLE IN PRESS D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

myristate acetate (PMA, Sigma; 1 mg mL1), and CaI (1 mg mL1) were prepared in dimethyl sulfoxide (DMSO, Sigma), and stored at 70 1C. A stock solution of yeast particulate b-glucan (MacroGard Feed Ingredient, MGFI, Biotec Pharmacon ASA, Tromso, Norway; 10 mg mL1) and lypopolysaccharide from Escherichia coli (LPS, Sigma 2 mg mL1) was prepared in Hanks balanced salt solution with Ca2+, Mg2+, and without phenol red (HBSS, Meditech-CellGro, AK, USA), and stored at 70 1C (MGFI) and 4 1C (LPS). The SYTOX green nucleic acid stain work solution (SYTOX, Molecular Probes, Eugene, OR; stock 5 mM in DMSO) was prepared immediately before use. Aliquots of the reagents were diluted in HBSS for each assay and preliminary titrations were used to determine optimal reagent concentrations. The final concentrations of reagents used in the assay after addition of HBSS and resuspended neutrophils were: cyto B 5 mg mL1; CaI 0.156, 0.312, 0.625, 1.25, 2.5, and 5 mg mL1; PMA 0.062, 0.125, 0.25, 0.5, 1, and 5 mg mL1; MGFI 5, 10, 50, 100, 200, and 400 mg mL1; LPS 0.05, 0.1, 0.5, 1, 5, and 10 mg mL1, and SYTOX 5 mM. 2.3. Neutrophil separation and slide preparation Kidney tissue from individual fish was aseptically collected, and neutrophils were separated using a previously described technique [12]. Briefly, four kidneys were pooled in Hank’s balanced salt solution without Ca2+, Mg2+ and phenol red (HBSSCMF, Mediatech-CellGro, AK, USA), homogenized in a 15 mL tissue grinder (Wheaton, USA) and pelleted for 15 min at 250g. Cells were separated on a medium with a specific gravity of 1.078 g mL1 (lymphocyte separation medium 1078, MediatechCellGro, AK, USA), viability was determined, and cell suspensions were adjusted to a standard concentration of 2.5  107 cells mL1. The neutrophil to non-neutrophil ratio was determined by differential leukocyte counts on Hemacolor (Harleco, EM Science, NJ, USA) stained cytospin preparations of cell isolates [8]. The standard neutrophil suspension was diluted 1:4 in HBSSCMF, two 25 mL drops of suspension were placed 2 cm apart on poly-L-lysine coated glass slide (Fisher, USA), and cells were incubated in humidified chamber for 20 min at 30 1C to allow for attachment. Immediately after incubation, work solutions of stimulants or HBSS (control) were gently added to the attached cells in final concen-

807

trations of CaI 5 mg mL1, PMA 1 mg mL1, and MGFI 200 mg mL1, and cells were incubated in a humidified chamber for 20 min at 30 1C. 2.4. Immunocytochemical staining Prepared slides were subjected to three different staining procedures, and all slides were done in triplicate. For fluorescent microscopy, immediately after incubation, cells were carefully fixed in 4% paraformaldehyde for 25 min, and mounted with antifade reagent with 40 , 60 -diamino-2-phenylindole, dihydrochloride DNA fluorescent stain (ProLong Gold Antifade reagent with DAPI; Molecular Probes, Eugene, OR). For double-labeling with SYTOX green, SYTOX green was added to incubated cells for 3 min to stain extracellular DNA; cells were fixed in 4% paraformaldehyde for 25 min, and mounted with ProLong Gold with DAPI. For light microscopy, incubated cells were stained with Hemacolor procedure [8] and mounted with Permount (Fisher, USA). For immunocytochemical staining, immediately after incubation, cells were carefully fixed in 4% paraformaldehyde for 25 min and washed in potassium phosphate buffered solution (KPBS). Slides were then incubated for 30 min in Image-iT FX signal enhancer solution (Molecular Probes, Eugene, OR), followed by 90-min incubation in blocking solution (2% normal donkey serum, Jackson Immunoresearch, West Grove, PA; 2% normal goat serum; Sigma; 1% bovine serum albumin, BSA, Sigma; 0.4% Triton X-100, Fisher Scientific Inc., Fairlawn, NJ). Slides were doublelabeled with a primary antibody cocktail containing 20 mg mL1 of mouse monoclonal anti-elastase (Calbiochem, San Diego, CA) and 0.2 mg mL1 of rabbit polyclonal anti-MPO (Abcam Inc., Cambridge, MA), and incubated overnight at room temperature in a sealed humid chamber. After incubation, slides were washed in KPBS with Triton X-100, incubated in the secondary antibody cocktail (goat anti-mouse Alexa 488 antibody, Molecular Probes; and donkey anti-rabbit Cy5 antibody, Jackson ImmunoResearch; diluted to 1:200), washed in KPBS with Triton X-100, and incubated in Rhodamine Phalloidin (Molecular Probes) diluted to 1:150. Finally, slides were washed in KPBS, and mounted with ProLong Gold with DAPI. Negative controls were carried out in parallel during all processing by the omission of the primary antibodies.

ARTICLE IN PRESS D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

808

2.5. Confocal and light microscopy After treatment with stimulants, neutrophil suspensions labeled with DAPI, SYTOX green, antibodies, and phalloidin were visualized and images captured using a Leica confocal scanning laser microscope (TCS-NT; Leica Microsystems Inc., Exton, PA). Neutrophil suspensions stained with Hemacolor stain were examined with a Zeiss Axioplan 2 upright microscope (Carl Zeiss MicroImaging, Inc, Thornwood, NY) using differential interference contrast (DIC). Images were captured with an AxioCam MRc color camera (Carl Zeiss MicroImaging, Inc, Thornwood, NY). All figures were prepared using Photoshop (ver. 7.0, Adobe, San Jose, CA) and Freehand (ver.10.0, Macromedia, San Francisco, CA). 2.6. Functional assays Neutrophil degranulation was measured with a previously described assay where release of MPO from neutrophil granules was measured in response to CaI stimulation in a microtiter plate. All assays were run on four samples of neutrophils pooled from four fish, each sample being run in duplicate wells (n ¼ 4). Test wells received 75 mL of HBSS and 50 mL of CaI; control (background) wells received 125 mL of HBSS; and total MPO content (lysed cells) wells received 125 mL of CTAB. 25 mL of cell suspension (2.5  107 cells mL1) was added to each well and incubated at 30 1C for 20 min. After incubation, 50 mL of TMB was added, followed immediately with 50 mL of H2O2. The color change reaction was allowed to proceed for 2 min, and 50 mL of 2 M sulfuric acid was added to stop the reaction. Test plates were centrifuged at 600g for 15 min, 200 mL of supernatant from each well was transferred to another plate and optical density (OD) in each well was determined at 405 nM using a microtiter plate spectrophotometer (V-Max, Molecular Devices, USA) with SOFTmax PRO 4.0 software. The percent release of MPO was calculated using the following formula: % release ¼ ½ðODstimulated  ODbackground Þ= ðODlysed  ODbackground Þ  100: The release of DNA from neutrophils in response to CaI, MGFI, PMA (with and without cyto B), and LPS, was determined using 96-well flat bottom microtiter plates (Fisher, USA). All assays were run

on four samples, each sample being run in duplicate wells (n ¼ 4). Test wells received 75 mL of cyto B or HBSS (if release without cyto B was measured) and 50 mL of stimulants, and control (background) wells received 125 mL of HBSS. Background values for each trial were determined concurrently with neutrophils exposed to stimuli. Plates containing reagents were pre-warmed to 30 1C, 25 mL of cell suspension containing 2.5  107 cells mL1 was added to each well and incubated at 30 1C for 20 min. After incubation, 50 mL of SYTOX was added, the color change reaction was allowed to proceed for 5 min, and fluorescence in each well was determined as arbitrary fluorescence units (AFU, excitation 488 nm, emission 527 nm) using a microtiter plate spectrofluorometer (SpectraMax Gemini XS, Molecular Devices, USA) with SOFTmax PRO 4.0 software. To study the kinetics of DNA release, stimulated neutrophils were incubated for times varying from 0 to 60 min. DNA release for fathead minnow kidney neutrophil suspensions was normalized using the following correction formula: AFU ¼ ðNCW  MAFUÞ=½ð%N  NCWÞ=100; where AFU is the normalized arbitrary fluorescence unit value, NCW the number of cells per well, MAFU the measured arbitrary fluorescence unit value, %N the percent of neutrophils in cell suspension. 2.7. Stress effects on degranulation and DNA release To determine ability of the assay to measure decreases in neutrophil activity in stressed fish, they were exposed to handling and crowding stress as previously described, with modifications [12]. A total of 112 fish were acclimatized for 2 weeks in a stock tank and used in the experiment. Fish were fed daily and water quality parameters were monitored two times per week for the duration of the acclimation and experiment: water temperature was 2071 1C, pH was 7.870.3, dissolved O2 was 771 mg L1, total ammonia nitrogen was o1.0 mg L1 and total nitrite nitrogen was below detection limit (HACH spectrophotometer 2000NR). At the beginning of the experiment (day 0), 64 fish were quickly netted from the stock tank and randomly divided into a stressed group (48 fish) and control group (16 fish). The fish from the stressed group were immediately transferred to a 4 L beaker with 2 L tank water buffered with sodium

ARTICLE IN PRESS D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

bicarbonate to pH of 8.0, and constant aeration. The stress conditions were maintained for 20 min, and fish were moved to a 1000 L recirculation tank supplied with the same water as stock tank. This procedure provided a crowding density of approximately 60 g L1 of fish. Fish from the control group were immediately killed with an overdose of MS222 (1 g L1), randomly divided in four samples of four fish each and assayed for neutrophil function as day 0. Four samples (randomly selected four fish/ sample) from stressed group, and four samples from control group (remaining 48 fish from stock tank) were assayed for neutrophil function on days 1, 3 and 7 after treatment (n ¼ 4). The mean % of neutrophil MPO release, and mean % of neutrophil DNA release (determined as described above) from control group was used as 100% of neutrophil activity for each day. Stressed group neutrophil activity was compared to control group activity using the following formula: % of control group neutrophil activity ¼ (% MPO or AFU from S/% MPO or AFU from control)  100. 2.8. Statistical analysis Data are presented as means7standard error of the mean (SEM) unless otherwise indicated. The differences in kinetics of DNA release, differences in DNA release due to the various stimulants, correlation of DNA release and concentration of stimulants, and inhibition of MPO and DNA release due to handling and crowding stress were examined using regression plots, and ANOVA followed by Student’s t-test (GraphPad Prism 3.00, 1999). Po0.05 was considered significant. 3. Results Mean neutrophil purity from the kidney cell suspensions was 74.172.8%; the remaining cells were lymphocytes and thrombocytes, with less than 1% monocytes/macrophages. Viability of cells before and after 60 min exposure to working concentrations of reagents was 495%. 3.1. Confocal and light microscopy images Confocal images of DAPI stained fish neutrophil cell suspensions stimulated with CaI, MGFI, or PMA revealed the presence of extracellular DNA structures or fish NETs (Fig. 1(A–C)). NETs were

809

not detected in unstimulated (HBSS treated) neutrophil cell suspensions (Fig. 1(D)). NETs appeared as fibrous structures, surrounding single cells or connecting several cells (Fig. 1(A–C)). Light microscopy of stimulated, Hemacolor stained fish neutrophils had structures similar to the DAPI stained NETs (Fig. 1(E–G)). Stimulated cells were surrounded with fibrous structures connecting several cells, while such structures were not detected in control (unstimulated) cell preparations (Fig. 1(H)). Double staining of the cells with SYTOX green (live cells) and DAPI (fixed cells) revealed that SYTOX stain applied to living cells stained extracellular DNA, but not nuclei of the stimulated cells (Fig. 2(B)). DAPI stain co-localized with SYTOX on extracellular DNA, and in addition labeled nuclei of fixed cells (Fig 2(A,C)). Confocal images of the cells triple stained with anti-MPO, Rhodamine Phalloidin (marker for actin) and DAPI, revealed presence of neutrophil primary granule contents associated with NETs (Fig. 3(D)). MPO co-localized with DNA fibers forming a complex of small (o0.2 mM) enzymatic granules connected with fiber-like structures (Fig. 3(B)). NETs are not associated with cytoskeletal protein actin, and cellular integrity has been preserved after stimulation and release of NETs (Fig. 3(C)). 3.2. DNA release assay Release of DNA from fathead minnow kidney neutrophils exposed to different concentrations of CaI, PMA, MGFI, and LPS is shown in Fig. 4. Cells without SYTOX green, and stimulants with and without SYTOX green had o50 AFU. Cells exposed to SYTOX green in HBSSCMF had o1000 AFU, and cells in HBSS (control) had o1500 AFU. DNA release from cells exposed to stimulants in media devoid of Ca2+ (HBSSCMF) was not significantly different from unstimulated control cells (o1200 AFU, not shown). In media containing Ca2+, the addition of different concentrations of CaI significantly correlated with an increase in DNA release (2200 AFU for 0.31 mg mL1, to 4500 AFU for 5 mg mL1 of CaI; compared to HBSS treated controls; Po0.001) (Fig. 4.). When LPS was used as stimulant, significant difference from controls (2600 AFU, 0.5 mg mL1) was observed, although increase in LPS concentrations caused less DNA release stimulation. Stimulation with PMA

ARTICLE IN PRESS 810

D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

Fig. 1. Neutrophil extracellular traps from the fathead minnow kidney neutrophils. A–D: confocal images of extracellular DNA fibrous structures, or NETs (DNA labeled with DAPI; blue). All stimulants caused release of NETs in extracellular space; NETs were surrounding and connecting cells. A: calcium ionophore (CaI) stimulated neutrophils; B: phorbol myristate acetate (PMA) stimulated cells; C: yeast particulate b-glucan (MGFI) stimulated cells; D: control, HBSS with Ca2+, Mg2+ did not provoke significant release of extracellular DNA fibrous material (scale bar A–D 10 mM). E–H: differential interference contrast (DIC) images of fibrous structures released by stimulated neutrophils and stained with rapid leukocyte staining procedure (Hemacolor). E: CaI; F: PMA; G: MGFI; H: control, HBSS with Ca2+, Mg2+ did not cause significant release of such structures (scale bar E–H 5 mM).

Fig. 2. Neutrophil extracellular traps are released from living fathead minnow kidney neutrophils. A: living cells were impermeable for SYTOX, and DNA from NETs was readily stained prior to cell fixation (green). B: DAPI labeled cells and corresponding NETs after cell fixation (blue). C: overlay of the DAPI and SYTOX staining revealed co-localization on a single NET and exclusion of SYTOX green from nuclei of the living cells (scale bar 10 mM).

(without cyto B) caused an increase in DNA release (1800 AFU, Po0.05), but there was no correlation between an increase in PMA concentration and DNA release. MGFI stimulated cells showed a significant difference from unstimulated controls when 50 and 400 mg mL1 were used.

The kinetic study of DNA release demonstrated that CaI caused significantly higher NET release than HBSS treated controls as soon as 10 min after stimulation (2000 AFU, Po0.001, Fig. 5.), and a significantly higher DNA release was observed when CaI was used as stimulant than when PMA,

ARTICLE IN PRESS D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

811

Fig. 3. Myeloperoxidase (MPO), but not actin, co-localize with NETs. A: NETs and cell nuclei stained with DAPI (blue); B: MPO labeling (green), note punctate immunoreactivity (arrows) localized extracellularly and connected with thin fibers; C: phalloidin (Phal, red) labeled actin; D: MPO (open arrows), but not actin immunoreactivity co-localizes with DNA labeling on NETs.

MGFI and LPS were used (Po0.001). DNA release positively correlated with the duration of incubation time. When LPS was used as stimulant, increase in DNA release was significantly different from the HBSS-treated control (Po0.001) as early as 20 min (2600 AFU). When PMA with and without cyto B was used as a stimulant, DNA release significantly higher than controls was detectable after 50 min

(2100 AFU without and 2400 AFU with cyto B, Po0.05). The stimulation of DNA release from fathead minnow neutrophils with MGFI was not significantly different from non-stimulated control (HBSS) at any time point. The maximum DNA release was recorded at 60 min post stimulation, with CaI causing a maximum of 7100 AFU to be released. LPS (2900 AFU) and PMA (2500 AFU

ARTICLE IN PRESS D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

812

4500

___ _

--- HBSS with Ca2+, Mg2+ CaI MGFI * PMA PMA + Cyto B LPS *

4000 Adjusted fluorescence [arbitrary units]

*

3500 3000

*

HBSS w/o Ca2+, Mg2+ *

*

2500 * *

2000

*

*

*

*

*

*

*

1500

5.0

10

5.0 400

1.0

5.0

2.5 200

0.5

1.0

1.25 100

0.25

0.5

0.62 50

0.12

0.1

0.31 10

0.06

0.05

0.15 5.0

1000 0 Dose[µg mL-1] Fig. 4. Release of NETs (DNA) from neutrophils stimulated with different concentrations of stimulants after 20 min incubation: CaI (0.15–5 mg mL1), PMA (0.06–5 mg mL1, without or with Cyto B 2.5 mg mL1), MGFI (5–400 mg mL1), and LPS (0.05–10 mg mL1). Stimulants (with or without SYTOX) and cells without SYTOX did not have significant fluorescence (o50 AFU). Increase in fluorescence was observed in CaI, PMA (without cyto B), and LPS treated cell suspensions and was significantly different (*Po0.05) than unstimulated control with HBSS. Unstimulated control cells and stimulated cells in media devoid of Ca2+ and Mg2+ (HBSSCMF) showed significantly lower fluorescence than cells in HBSS, indicating that presence of Ca2+ is necessary for DNA release (not shown). Data is presented as means7SEM (n ¼ 4).

with cyto B and 2200 AFU without cyto B) caused significantly less maximal levels of exocytosis than CaI at 60 min post stimulation (Po0.001; Fig. 5). 3.3. Decrease of exocytosis in stressed fish The effect of handling and crowding stress, on neutrophil activity measured as percent of MPO release and percent of DNA release is shown in Fig. 6. Fathead minnows were exposed to handling and crowding stress for 20 min and MPO and DNA assays were performed on 0, 24, 72 h and 7 days post exposure using CaI as the stimulant. There was no significant difference in percentage of neutrophils isolated from control and stressed group fish throughout the experiment. The overall decrease in DNA release and MPO exocytosis to 80% of control activity in the stressed group was observed during the period of 7 days after exposure and is shown in Fig. 6. A significant decrease of DNA release and MPO exocytosis in the stressed group (80% of control group DNA release;

79% of control group MPO exocytosis, Po0.05) was first detected 24 h after exposure (Fig 6.). DNA release in the stressed group returned to control levels, and MPO exocytosis in the stressed group remained significantly lower than control (84%, Po0.05; Fig. 5.) at day 7 post stress.

4. Discussion NETs have been described recently in mammals [9,10,15,16]. Limited literature on this newly discovered neutrophil functional mechanism suggests that: NETs consist of DNA fibers, histones, and granular proteins, but membrane proteins are not detected on NETs; the potential role of NETs is immobilization and killing of bacteria and yeasts; neutrophils responsive to cytokine stimuli are capable of producing NETs; stimulated neutrophils actively release NETs; and NETs are not a product of cellular degradation or release of nuclear material from dead cells [9,10,15,17]. The presence of NETs

ARTICLE IN PRESS D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

Adjusted fluorescence [arbitrary units]

8000

HBSS with Ca2+, Mg2+

*

*Stimulants + Sytox

7000

*

*Cells w/o Sytox

*

CaI

6000

MGFI

813

*

PMA

5000

PMA + Cyto B 4000

LPS *

3000

*

*

*

*

2000

* * *

* * *

1000 0 0

10

20

30 Time [min]

40

50

60

Fig. 5. Release of NETs (DNA) from neutrophils stimulated with different stimulants over time: CaI (5 mg mL1), PMA (1 mg mL1, without or with 2.5 mg mL1 of cyto B), MGFI (200 mg mL1), and LPS (0.5 mg mL1). Cells stimulated with CaI, LPS and PMA showed significant increase in fluorescence as soon as 10, 20 and 50 min post stimulation, respectively (compared to unstimulated control in HBSS, *Po0.05). Data are presented as means7SEM (n ¼ 4).

Control MPO release Control DNA release 110

Stressed MPO release Stressed DNA release

DNA and MPO release [% of control]

105 100 95 90 * *

85 *

80

*

75

0 Day0

Day1

Day3

Day7

Fig. 6. The reduction of DNA and MPO release from neutrophils in adult fathead minnows exposed to handling and crowding stress for 20 min. At 24 and 72 h post exposure significant decrease in activity was observed in DNA and MPO release (*Po0.05), and DNA release returned to control levels by day 7. There was no significant difference between DNA and MPO release, except on day 7 (Po0.05). Data is presented as means7SEM (n ¼ 4).

ARTICLE IN PRESS 814

D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

in fish, and effects of stress on release of NETs from neutrophils have not been previously described. Neutrophils isolated from fathead minnow kidneys released NETs when stimulated with CaI, PMA and particulate yeast b-glucan (MGFI), as seen in Fig. 1(A–C). NETs were stained with fluorescent DNA dye (DAPI) and visualized with confocal microscopy. Similar structures (NETs) were observed in mammalian neutrophils stimulated with different stimulants (IFN-g, IL-8, PMA or LPS), stained with different DNA staining procedures (Hoechst, propidium iodide), and visualized with confocal and electronic microscopy [9,10]. Structures resembling the location and overall appearance of NETs as seen during fluorescent microscopy were observed during light microscopic examination of stimulated fathead minnow neutrophils stained with a rapid leukocyte staining procedure (Fig. 1(E–G)). Rapid staining procedures (Hemacolor) have not been previously used in detection of NETs. Such structures were not seen in unstimulated neutrophils (control), and the percentage of cells associated with NET-like structures was not significantly different from the neutrophil to non-neutrophil ratio on cytospin slides of the same cell suspension (data not shown). Exclusion of intranuclear DNA fluorescent staining was not previously used to demonstrate that NETs are released from living cells. Stimulated neutrophils were incubated with SYTOX green stain, that is not capable of penetrating membranes of living cells [11], and carefully washed before cells were fixed and stained with DAPI. Double-labeling of NETs with two DNA specific fluorescent dyes demonstrated that SYTOX was only associated with extracellular DNA in NETs, and was not associated with nuclear material of the living cells, while DAPI stained both NETs, and nuclei of the fixed cells (Fig. 2.). Therefore, fish NETs are cast from stimulated, living neutrophils, and can be visualized with DNA fluorescent staining and confocal microscopy, or with rapid leukocyte staining procedures and light microscopy. Immunochemistry performed on NETs released from human neutrophils showed co-localization of DNA, histones (H1, H2A, H2B, H3 and H4), and proteins from granules (elastase, MPO, cathepsin G, lactoferrin, and gelatinase), but did not demonstrate association with granule membrane protein CD63, and cytoplasmic markers such as actin, tubulin, or annexin I [9]. Triple labeling immunocytochemistry performed on stimulated fish neutrophil suspensions

demonstrated that DNA fibers co-localize with MPO immunoreactivity. Furthermore, actin did not co-localize with MPO or DNA on NETs, indicating that fish NETs are not associated with cytoskeleton (Fig. 3.). Extracellular DNA release and formation of fibrous structures in association with granular and histone proteins with antimicrobial properties [2,18] is an essential step for the role of NETs in neutrophil killing, since application of DNase significantly reduced extracellular killing of bacteria and yeasts [9,15,17], and DNase producing bacterial strains show increased survival when exposed to NETs producing neutrophils [19,20]. Assays for detecting extracellular DNA release from stimulated neutrophils were recently described for human neutrophils [9,10], but modifications of such assays for use with fish neutrophils have not been reported. Fathead minnow kidney neutrophils responded to stimulation with CaI, b-glucans, PMA and LPS, and released DNA in dose and time dependent manner (Figs. 4 and 5). In fish neutrophils, unstimulated control cells and stimulated cells in media devoid of Ca2+ and Mg2+ (HBSSCMF) had significantly lower fluorescence than cells in HBSS (not shown), indicating that presence of Ca2+ is necessary for DNA release. Human neutrophils have been stimulated for NET release with IL-8, PMA, C5a (with and without pretreatment of cells with IFN-a or IFN-g, and LPS) [9,10]. It has been reported that use of CaI on leukocyte membranes selectively increased concentration of intracellular Ca2+, leading to effects similar to activation of the leukocyte signaling pathways with cytokines [21,22]. Increase of intracellular Ca2+ after application of CaI was observed in fish neutrophils using a calcium imaging system (unpublished data). Therefore, it is possible that effects of CaI on fathead minnow neutrophil membrane initiated cellular responses, including degranulation [12]. PMA is a known activator of protein kinase C (PKC), a potent stimulator of leukocyte functions [2], and stimulation of human neutrophils with PMA can cause release of NETs [9]. PMA induced release of NETs was detectable, but significantly lower than CaI induced DNA release in fathead minnow neutrophils. Use of LPS was reported to release NETs in human neutrophils [9], and release of NETs with LPS from E. coli has been observed in fish (Fig 4). Particulate yeast b-glucans (MGFI) have been used as potent stimulators of fish neutrophil degranulation, and their activity is probably related to the presence of

ARTICLE IN PRESS D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816

a glucan-specific receptor [23,24], but MGFI did not cause DNA release that significantly differed from the controls. Glucans are primarily found in fungal cell walls, and are known for the stimulation of neutrophils during fungal infections [25]. The role of NETs against bacterial and fungal infections was described [9,15,17], and although glucan stimulation does not provoke fathead minnow neutrophils to cast NETs, it is a potent stimulator of MPO exocytosis, leading to higher concentrations of hypochloric acid whose presence may be of higher importance in destroying fungal cell walls in fish mucosal surfaces [25]. The release of NETs in fish was induced via Ca++ dependent pathway, but decreased NET release observed during PMA activation of PKC, as well as LPS and glucan receptor stimulation, appeared to be significantly lower than in CaI stimulated neutrophils. The potential connection between degranulation of neutrophil primary (azurophilic) granules and NETs release has been suspected [9], and histone proteins have been observed in membrane fraction of bovine neutrophils [26] but no direct connection between DNA release and degranulation has been confirmed. In this study, neutrophils stimulated with CaI released NETs, but potent stimulators of degranulation in fathead minnows, such as particulate yeast b-glucans [27], showed significantly lower (3–4 fold) stimulation of DNA release than CaI, although each stimulant caused visible release of NETs in immunocytochemistry experiments. It is possible that DNA release from fathead minnow kidney neutrophils has a different mechanism than degranulation, and additional research is necessary to determine cellular and molecular mechanisms for NETs release from stimulated cells and determine pathways involved in the release of NETs in fish and mammalian systems. It has been recognized that handling and crowding can induce an increased stress response in fish, with an increase in blood cortisol concentrations, and a correlation with decreased degranulation [12]. The kinetic study of DNA release in fish exposed to handling and crowding stress demonstrated a decrease of DNA exocytosis in response to stimulation by CaI detected at 24 h after exposure, which is comparable to a decrease in degranulation (80% of non-stressed control). Nevertheless, neutrophil DNA release returned to levels not significantly different from controls after 7 days post stress, while degranulation remained significantly lower than

815

controls (Fig. 5). A significant change in neutrophil to non-neutrophil ratio in kidney cell suspensions was not observed in control and stressed fish. The NETs are extracellular DNA fibrous structures released from living fish kidney neutrophils upon stimulation. For the first time, the presence of NETs was detected in fish neutrophils using confocal fluorescent microscopy, rapid leukocyte staining methods, and quantified using a microtiter plate fluorometric assay. The DNA release assay is direct, quantitative, can rapidly measure a large number of samples, and is capable of detecting inhibition of DNA release in neutrophils from stressed fish. Detection of NETs in fish can be used as an addition to the existing neutrophil functional tests in fathead minnows, and as a new research model to study effects of stress, immunomodulators, toxicants, and diseases on neutrophil biology. Acknowledgments This research was supported by a grant from National Water Resources Institute No. 2002IA25 under 104g National Research Grant funding program. MacroGard Feed Ingredients was generously provided by Maja Johnsen (Biotec Pharmacon ASA, Tromso, Norway). Drs. G.J. Atchison, R.B. Bringolf, and J.E. Morris helped in founding and maintaining the fathead minnow colony. Dr. E.C. Powell and Dr. D.S. Sakaguchi provided supportive expertise. Thomas Skadow and Dr. M. Joksimovic´ provided technical assistance.

References [1] Densen P, Mandell GL. Granulocytic phagocytes. In: Mandell GL, Douglas RG, Bennet JE, editors. Principles and practices of infectious diseases. New York: Churchill Livingstone; 1990. p. 81. [2] Smolen JE. Characteristics and mechanisms of secretion by neutrophils. In: Hallet MB, editor. The neutrophil: cellular biochemistry and physiology. Boca Raton: CRC Press; 1989. p. 23. [3] Ellis AE. The immunology of teleosts. In: Roberts RJ, editor. Fish pathology, third ed. London/Edinburgh/New York/Philadelphia/St. Louis/Sydney/Toronto: WB Saunders; 2001. p. 133. [4] Secombes CJ. The nonspecific immune system: cellular defenses. In: Iwama G, Nakanishi T, editors. The fish immune system: organism, pathogen and environment. San Diego: Academic Press; 1996. p. 63. [5] Rowley AF. Collection, separation and identification of fish leucocytes. In: Stolen JS, Fletcher TC, Anderson DP, Roberson BS, van Muiswinkel WB, editors. Techniques in

ARTICLE IN PRESS 816

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

D. Palic´ et al. / Developmental and Comparative Immunology 31 (2007) 805–816 fish immunology—fish immunology technical communications no. 1. Fair Haven, NJ, USA: SOS Publications; 1990. p. 113. Lamas J, Ellis AE. Atlantic salmon (Salmo salar) neutorphil responses to Aeromonas salmonicida. Fish Shellfish Immunol 1994;4(3):201. Rodrigues A, Esteban MA, Meseguer J. Phagocytosis and peroxide release by seabream (Sparus aurata L.) leucocytes in response to yeast cells. Anatom Record Part A 2003; 272A:415. Palic´ D, Andreasen CB, Frank ED, Menzel BW, Roth JA. Gradient separation and cytochemical characterisation of neutrophils from kidney of fathead minnow (Pimephales promelas Rafinesque, 1820). Fish Shellfish Immunol 2005; 18:263. Brinkmann V, Reichard U, Goosmann C, Fauler B, Uhlemann Y, Weiss DS, et al. Neutrophil extracellular traps kill bacteria. Science 2004;303:1532. Martinelli S, Urosevic M, Daryadel A, Oberholzer PA, Baumann C, Fey MF, et al. Induction of genes mediating interferon-dependent extracellular trap formation during neutrophil differentiation. J Biol Chem 2004;279:44123. Jones JL, Singer LV. Fluorescence microplate-based assay for tumor necrosis factor activity using SYTOX green stain. Anal Biochem 2001;293:8. Palic´ D, Andreasen CB, Menzel BW, Roth JA. A rapid, direct assay to measure degranulation of primary granules in neutrophils from kidney of fathead minnow (Pimephales promelas Rafinesque, 1820). Fish Shellfish Immunol 2005; 19:217. Russom CL, Bradbury SP, Broderus SJ, Hammermeister DE, Drummond RA. Predicting modes of toxic action from chemical structure: acute toxicity in fathead minnows (Pimephales promelas). Environ Toxicol Chem 1997;16:948. Palic´ D, Herolt DM, Andreasen CB, Menzel BW, Roth JA. Anesthetic efficacy of tricaine methanesulphonate, metomidate and clove oil: effects on plasma cortisol concentration and neutrophil function in fathead minnows (Pimephales promelas Rafinesque, 1820). Aquaculture 2006;254:675. Sumby P, Barbian KD, Gardner DJ, Whitney AR, Welty DM, Long RD, et al. Extracellular deoxyribonuclease made by group A Streptococcus assists pathogenesis by enhancing evasion of the innate immune response. In: Proceedings of the national academy of sciences of the United States of America, vol. 102, 2005. p. 1679.

[16] Alghamdi AS, Foster DN. Seminal DNase frees spermatozoa entangled in neutrophil extracellular traps. Biol Reprod 2005;73:1174. [17] Urban CF, Reichard U, Brinkmann V, Zychlinsky A. Neutrophil extracellular traps capture and kill Candida albicans yeast and hyphal forms. Cell Microbiol 2006;8:668. [18] Patat SA, Carnegie RB, Kingsbury C, Gross PS, Chapman R, Schey KL. Antimicrobial activity of histones from hemocytes of the Pacific white shrimp. Eur J Biochem/ FEBS 2004;271:4825. [19] Beiter K, Wartha F, Albiger B, Normark S, Zychlinsky A, Henriques-Normark B. An endonuclease allows Streptococcus pneumoniae to escape from neutrophil extracellular traps. Curr Biol 2006;16:401. [20] Buchanan JT, Simpson AJ, Aziz RK, Liu GY, Kristian SA, Kotb M, et al. DNase expression allows the pathogen group A Streptococcus to escape killing in neutrophil extracellular traps. Curr Biol 2006;16:396. [21] Cronstein BN, Haines KA. Stimulus-response uncoupling in the neutrophil. Adenosine A2-receptor occupancy inhibits the sustained, but not the early, events of stimulus transduction in human neutrophils by a mechanism independent of actin-filament formation. Biochem J 1992; 281:631. [22] Andreas V, Petkovic M, Arnhold J. Involvement of phosphatidic acid in both degranulation and oxidative activity in fMet–Leu–Phe stimulated polymorphonuclear leukocytes. Cell Physiol Biochem 2003;13:165. [23] Reid DM, Montoya M, Taylor PR, Borrow P, Gordon S, Brown GD, et al. Expression of the beta-glucan receptor, dectin-1, on murine leukocytes in situ correlates with its function in pathogen recognition and reveals potential roles in leukocyte interactions. J Leukocyte Biol 2004;76:86. [24] Ainsworth AJ. A beta-glucan inhibitable zymosan receptor on channel catfish neutrophils. Vet Immunol Immunopathol 1994;41:141. [25] Faurschou M, Borregaard N. Neutrophil granules and secretory vesicles in inflammation. Microb Infect 2003; 5(14):1317. [26] Lippolis JD, Reinhardt TA. Proteomic survey of bovine neutrophils. Vet Immunol Immunopathol 2005;103:53. [27] Palic´ D, Andreasen CB, Herolt DM, Menzel BW, Roth JA. Immunomodulatory effects of b-glucan on neutrophil function in fathead minnows (Pimephales promelas Rafinesque, 1820). Dev Compar Immunol 2006;30:817.