Deactivation of primed respiratory burst response of goldfish macrophages by leukocyte-derived macrophage activating factor(s)

Devclqmcdal

and Chmparatiw Immunology, Vol. 20, No. 6, pp. 427439, 1996 CopyrightQ 1996 Elscvitr Science J&l. All rights resend Printed in Great Britain 0145-305X/96 $15.00+0.00

PIk s0145-305x(96)ooo29-8

DEACTIVATION OF PRIMED RESPIRATORY BURST RESPONSE OF GOLDFISH MACROPHAGES BY LEUKOCYTE-DERIVED MACROPHAGE ACTIVATING FACTOR(S) Norman F. Neumann*

and Miodrag Belosevict

*Departments of Biological Sciences; and tMedical Microbiology and Immunology, Universityof Alberta, Edmonton, AlbertaTGG 2E9,Canada (Received May 1996; Accepted July 1996)

nAbstract-Macrophago activation factors (MAF), induced maximal priming of the respiratory burst response in GMCL after 6 h of stimulus, but by 24 or 48 h no priming effect was observed. Bacterial lipopolysaccharide (LPS) also primed the respiratory burst of gold5sh macrophages, but the kinetics of priming were different from that induced by MAE LPS induced a gradual increase in prlming potential over 48 h of cultivation. Co-stimulation of macrophages with MAP and LPS resulted in enhanced priming of respiratory burst activity compared to either factor alone; however, the kinetics of priming were similar to those induced by MAF only. The MAF antagonized the ability of LPS to prime the respiratory burst over extended cultivation. The priming kinetics of the respiratory burst induced by MAF and/or LPS were not unique to GMCL, but were also similar for primary cultures of IVDKM. Respiratory burst deactivated macrophages-mounted potent nitric oxide response, indicating that this deactivation event was selective for respiratory burst activity. Autocrine factors produced by MAF-actlvated macrophages augmented priming of the respiratory burst, suggesting that deactivation of primed respiratory burst responses was not due to cytokine mediators produced

by activated macrophages, but was most likely an intracellular deactivation event. Furthermore, production of reactive intermediates by activated fish macrophages was biphasic; with maximal ROI production occuring 6 h after stimulus, and maximal RN1 occuring 72 h after stimulus. Our results indicate that activated fish macrophages mount sequential antimicrobial responses that are selectively deprogrammed once maximal induction has occured. The ability to selectively deactivate ROI production without affecting subsequent RN1 production may play an important role in host defense: regulating the duration of ROI production, and thus minimizing host tissue damage in an otherwise futile attempt to eliminate ROI resistant pathogens. Copyright 0 1996 Elsevier Science Ltd.

qlKeywords-Macrophage; Priming; Respiratory burst; Activation; IQsh; Nitric oxide. Nomenclature MAF GMCL ROI RNI MXM LPS NO IFN TNF M-CSF GM-CSF

Address correspondence to Dr Miodrag Belosevic, Associate Professor, CW-405 Biological Sciences Building, Department of Biological Sciences, University of Alberta, Edmonton, Alberta Canada T6G 2E9. 427

macrophageactivation factors goldlish macrophagecell line reactiveoxygen intemediates reactivenitrogen intermediates in vitro-derivedkidney macrophages lipopolysaceharide nitric oxide interferon tumor necrosis factor macrophage colony stimulating factor granulocyte/macrophage colony stimulating factor

428

NBT PMA NB-MMLA

N. F. Neumann and M. Belosevic

nitroblue tetrazolium phorbol my&ate acetate W-monomethyl L-arginine

Introduction Phagocytes, such as neutrophils, macrophages and eosinophils, possess the unique ability to increase oxidative metabolism in response to receptormediated phagocytosis of a pathogen: a phenomenon known as the respiratory burst (reviewed in Ref. (1)). This burst in oxygen consumption is not a result of increased mitochondrial respiration required for phagocytosis, but represents a unique antimicrobial mechanism of phagocytes (2). Induction of the respiratory burst results in the production of numerous toxic oxygen intermediates; including superoxide (0~~ ), hydrogen peroxide (HzOZ), hydroxyl radical (OH.), and peroxynitrite (ON00 - ), the antimicrobial actions of which have been well documented (1). Macrophages can be programmed for an enhanced respiratory burst response following stimulation with cytokines. Cytokines do not trigger the respiratory burst, but induce the upregulation of enzymatic components necessary for an enhanced respiratory burst response; a phenomenon known as priming. Numerous cytokines have the ability to prime the respiratory burst of mammalian macrophages; these include EN-y, TNF-u, IL3, IL-4, M-CSF and GM-CSF (313). Triggering of the respiratory burst can be accomplished not only through receptor mediated phagocytosis (14), but also by compounds such as phorbol esters (15). The respiratory burst has been demonstrated in the phagocytes of lower vertebrates (16), and more recently in several invertebrates (17-19), indicating that it is a conserved antimicrobial killing mechanism. However, there is little information regarding the cytokine-induced regulation of the respiratory burst re-

sponse across diverse taxa. Recent evidence suggests that the cytokine regulation of respiratory burst responses in fish macrophages may be similar to that of mammals (20-22). Previous reports have demonstrated that fish macrophages can be primed for an enhanced respiratory burst response following stimulation for 48 h with supernatants from mitogen stimulated kidney leukocytes (20-22). The cytokine(s) responsible for this priming effect is (are) produced by Tlymphocytes, which themselves require the presence of macrophages for optimal production (22). The ability of this T-cell derived lymphokine to prime the respiratory burst, and its potential ability to induce nitric oxide production by fish macrophages (23), has led some authors to speculate that fish may possess an IFNy-like molecule (21, 23). In the present study, we show that goldfish macrophages undergo a rapid priming of the respiratory burst response after only 6 h of stimulation with MAF. Maximal priming of this antimicrobial response was followed by a gradual decline in the priming potential of MAF. Macrophage activation factors also abrogated the ability of LPS to prime the respiratory burst response over extended cultivation periods. Our results indicate that MAF induced a rapid priming of the respiratory burst in fish macrophages, and possibly a delayed deactivation of this response, which did not affect the ability of macrophages to mount subsequent nitric oxide reactions.

Materials and Methods Fish Goldfish (Carassius uuratus) from 10 to 15 cm long were purchased from either Ozark Fisheries Inc. (Southland, MI) or Grassy Forks Fisheries (Martinsville, IN) and maintained at the Aquatic Facility of the Department of Biological Sciences,

Deactivation of cytokine-primed

macrophages respiratory burst

429

University of Alberta. The fish were maintained at 20°C in a flow-through water system on a simulated natural photoperiod (Edmonton, Alberta). Fish were fed to satiation daily with trout pellets, and acclimated to this environment for at least 3 weeks prior to use.

balanced salt solution. The remaining adherent cell layer was given fresh media containing serum (10% calf serum, 2.5% goldfish serum) and incubated for 72 h. Macrophage activation factor-containing supematants were subsequently removed, filter sterilzed, and stored at - 20°C until used in assays.

Production of MAF from Goldfish Kidney Leukocytes

Establbhment of IVDKA4 cultures. Previous experiments done in our laboratory indicated the presence of a soluble autocrine growth factor(s) required for proliferation of a spontaneously growing GMCL (23, 24). The addition of supernatants from actively growing GMCL to primary kidney leukocyte cultures, resulted in selective proliferation of macrophages, with >99% of the cells being non-specific esterase positive; > 93% showing typical macrophage morphology using a Modified Wright’s Stain; and the ability to produce nitric oxide after treatment with MAF and/or LPS (23). Using this methodology we established and propagated a number of primary fish macrophage cultures for several months. Kidney leukocytes were isolated using 51% continuous Percoll gradient centrifugation. Cells found at the Percollmedium interface were removed, washed twice with serum-free medium and subsequently cultured in 25 cm2 tissue culture flasks. Leukocytes were cultured at a concentration 2 x lo6 cells/ml in medium containing serum, antibiotics, and 25% alter sterilized growth factor supematants obtained from lo-day-old GMCL cultures. Primary macrophage cultures were grown in flasks for approximately 5 days and non-adherent cells and debris removed by gently washing flasks once with pre-warmed (20°C) HBSS. Remaining adherent cells were fed fresh medium containing serum, antibiotics, and growth factor. Cells were incubated for 2-3 days before being transfered to 75 cm2 tissue culture flasks, and were fed with fresh medium containing growth factor when required (as assessed by

Leukocytes were isolated from the kidneys of goldfish using procedures described elsewhere (23). Briefly, fish were anesthetized with MS222, and the kidneys aseptically removed and placed in ice-cold medium. Kidneys were pressed through sterile stainless steel screens into medium containing antibiotics (50 pg/mL gentamicin, 100 U/mL penicillin, 100 pg/ mL streptomycin) and heparin (50 U/ mL). Kidney cells from three fish were pooled into a single tube for purification of kidney leukocytes for production of MAF. For purification of leukocytes, the pooled cell suspension was layered on 51% Percoll (Sigma) and centrifuged at 400 x g for 30 min. Cells at the medium51% Percoll interface were transferred into clean tubes and washed twice by centrifugation at 200 x g for- 10 min at 4°C. Purified leukocytes were enumerated using a haemocytometer and their number adjusted to 4 x lo6 viable cells per mL (as determined using Trypan Blue). Leukocytes were seeded in tissue culture flasks (25 or 75 cm’), and incubated overnight at 20°C prior to the addition of mitogens. Macrophage activation factor was produced by stimulating mixed leukocyte cultures with 10 &mL concanavalin A (Boehringer Manheim), 10 ng/mL PMA (Sigma), and 100 ng/mL Ca” ionophore A23187 (Sigma). Cultures were stimulated with these mitogens for 6 h, after which the mitogens were removed by washing the adherent cell layer with several changes of Hanks

430

color change in media and overcrowding of cells). In vitro-derived kidney macrophage cultures were assayed for respiratory burst responses on day 14 of cultivation.

Respiratory Burst Assay Macrophages were assayed for respiratory burst responses as follows: 2 x lo5 GMCL were seeded into wells of a 96-well culture plate in 100 pL of medium containing serum and antibiotics. The macrophages were cultured overnight in order to stabilize respiratory activity. Macrophage activation factor @al dilution 1:3) was subsequently added to macrophage cultures in the presence or absence of LPS (final concentration, 5 pg/ mL). Kinetics of respiratory burst priming induced by MAF and/or LPS on GMCL were determined by assaying for respiratory activity 1, 3, 6, 12,24, or 48 h after stimulation. For dose-response experiments, different dilutions of MAF (1:4-l: 16) and/or concentrations of LPS ((F40 pg/mL) were added to macrophage cultures, and the respiratory burst activity measured 6 h after incubation (a time point at which maximal MAF-induced priming occured). Two times 10’ IVDKM were seeded in individual wells of a 96-well culture plate, and incubated overnight prior to the addition of MAF (1:3 tinal dilution) and/or LPS (final concentration, 10 c(g/ mL). Respiratory burst response was measured in parallel cultures 6 and 48 h after treatment. Respiratory burst responses were assayed using the nitroblue tetraxolium (NBT, Sigma) method (25). Nitroblue tetrazolium was dissolved in dimethyl sulfoxide (2O%w/v) to which prewarmed (SOC) serum-free medium was gradually added to completely dissolve NBT. This NBT solution was diluted in pre-warmed (SOY!) serum free media to give a 8nal concentration of 2 mg/mL and

N. F. Neumann and M. Belosevic

heated at 50°C for 10 min. Medium containing NBT was filter sterilized and PMA was added to a final concentration of 200 ng/mL for triggering the respiratory burst of GMCL and 400 ng/mL for triggering IVDKM. Fifty microlitres of the NBT/PMA solution was added to macrophage cultures and incubated for 25 min at room temperature. Supematants were subsequently removed and macrophage layers fixed by adding 200 pL of 70% methanol for 1 min. Unreduced NBT was removed by washing cells several times with 70% methanol. Reduced NBT was dissolved by adding 120 pL of 2 A4 KOH to individual wells and pipetting vigorously. One hundred and forty microlitres of dimethyl sulfoxide was then added to wells and absorbances read at 630 nm.

Nitric Oxide Assay In addition to producing ROIs, macrophages in representative cultures were also assayed for their ability to produce RNIs. GMCL (1 x 10’ cells) in parallel cultures were assayed for ROI and RNI production 0, 6, 12, 24, 72 or 96 h after stimulation with MAF (1:4) and LPS (5 pg/mL), in the presence or absence of Ng-MMLA, a potent inhibitor of nitric oxide production in fish macrophages (23, 24). Production of ROI was assayed for using the protocols described above. Production of RNI was done by quantifying nitrite content from the supematants of activated macrophages using the Griess reaction (26). Briefly, 75 l.tL of macrophage supematants were removed from individual wells and placed in a separate 96-well microtitre plate. One-hundred microlitres of 1% sulfanilamide (Sigma) in 2.5% phosphoric acid was added to each sample, followed by 100 pL of 0.1% N-naphthylethylenediamine (Sigma) in 2.5% phosphoric acid. Optical density was determined using a spectrophotometer (Riotek) at 540 nm. The molar

Deactivation of cytokine-primed macrophages respiratoryburst

concentration of nitrite in samples was determined from standard curves generated using known concentrations of sodium nitrite.

Production of Autocrine Mediators from Activated Macrophages In experiments examining the kinetics of priming macrophage respiratory burst responses by MAF, it was clear that macrophages underwent a MAFdependent deactivation of their primed response (Fig. 1). In order to determine whether autocrine mediators produced by MAF activated macrophages were able to deactivate primed respiratory burst responses, macrophages were pulsed with MAF and potential autocrine deactivating cytokines removed at a later time point. Two times lo5 GCLM were incubated with MAF for 6 h, after which medium was removed and cells washed 4xin serum-free medium, and once with medium containing serum in order to remove MAF. Cells were

431

incubated in 100 FL of medium with serum for an additional 18 h and the supematants removed and used to inhibit the rapid priming of respiratory burst activity normally induced by MAF (i.e. after 6 h). In separate 96well culture plates, 2 x lo5 GMCL were pre-treated with supematants for 0, 6, or 12 h before the addition of MAF. Respiratory burst responses were measured 6 h after the addition of MAF using the protocols outlined previously.

Statistics The data were analyzed using SuperANOVA software for the Power Macintosh. One- and two-way analysis of variance were employed to determine differences between experimental groups.

Resulh

or

To assess the kinetics of cytokine- and/ endotoxin-induced priming of the

P

-t

mscrophsgss only

+

LPS (5 pgrml)

-a-

MAF + LPS (6 pgrml)

O.l_ 1

3

6

12

24

46

Time (h) Figure 1. Kinetics of MAF (1:3dilution) and/or LPS (5 pg/mL)-induced priming of the respiratory burst response of GMCL. Macrophages were stimulated with MAF and/or LPS for the specified time periods before being triggered for respiratoryburst activityusing 50 ng/mL PMA.The data are from a representative experiment out of three that were performed. Each pointrepresentsthe mean optical densityof six cultures ( f SE.?&),blanked against KOH and DMSO.

432

respiratory burst response of fish macrophages, we treated GMCL with MAF and/or LPS and assayed for PMAtriggered respiratory burst activity at regularly timed intervals. Macrophage activation factors primed the respiratory burst of macrophages as early as 1 h after treatment, with maximal priming observed 6 h after stimulation (Fig. 1). Maximal priming of the respiratory burst response plateaued over the next 6 h, but declined precipitously by 24 h, to levels similar to those of unstimulated macrophages. By 48 h respiratory burst responses were significantly lower in MAF-stimulated macrophages than in macrophages unstimulated control (P
N. F. Neumann and M. Belosevic

basal respiratory burst levels of unprimed macrophages after 48 h of incubation (Fig. 1). The kinetics of priming were similar for four different MAFs. Different MAFs varied only in their ability to completely antagonize the priming effect of LPS (Figs 1 and 4). However, of all MAFs tested, co-stimulated macrophages had significantly lower respiratory burst responses compared to LPS only stimulated counterparts by 48 h (data not shown). Macrophage activation factors could also augment the respiratory burst of macrophages that had been pre-treated with LPS 18 h before the addition of MAF (Fig. 3). Macrophage activation factor-enhanced priming of the respiratory burst of LPS pre-stimulated macrophages when cells were incubated for an additional 6 h with MAF. However, 24 h later co-stimulated macrophages had significantly lower respiratory burst responses than those co-stimulated for only 6 h (P
433

Deactivation of cytokine-primed macrophages respiratory burst

-H-

noMAF

4-

MAF 1:4

+

YAF 1:lb

Figure 2. Dose-response of MAF- and/or LPS-induced priming of the respiratory burst of GMCL. Macrophages were stimulated with various concentrationsof MAF and/or LPS and PMA-triggered respiratory burst response measured 6 h later.The data are from a representattveexperiment out of two that were performed. Each point represents the mean optical density of quadruplicate cultures ( fS.E.M.) blanked against the basal respiratoryburst level of unprimed PMA stimulated macrophages.

tive for the respiratory burst response, and that cytokine- and LPS-activated fish macrophages appear to mount temporally spaced cytotoxic responses. Macrophage activation factor by itself did not trigger respiratory burst activity, ruling out the possibility that this decline in priming potential was a result of an exhausted respiratory burst response (data not shown). In addition, the deactivating event could not be attributed to cell death, since macrophages contin6 24 ued to increase in number in the presence Time (h) of MAF over 24 h, albeit to a lesser Figure 3. Ability of MAF to enhance and subse- extent (10% fewer) than unstimulated quently deactivate the respiratory burst response macrophages (data not shown). of LPS-pretreated GMCL. Macrophageswere preContrary to our hypothesis, that MAF treated with 5 pg/mL LPS for 18 h, and then treamacrophages may produce ted with either medium (LPS group) or MAF (1:4 activated final dilution. MAF + LPS group). PMA-triggered autocrine mediators which deactivate respimtory burst response was measured 6 and primed respiratory burst responses, acti24 h after the addition of medlum or h4AF.The data am from a repmsentattve experfment out of vated macrophages secreted factors which two that were performed. Each point represents enhanced priming of MAF-activated the mean optical density ( iSE.MJ of quadrupli- macrophages, regardless of pretreatment cate cuttures, blanked against the basal respimtory burst level of unprimed PMA stimulated condition (Fig. 5). It appears, therefore, macrophages at these time points. that the macrophagederived autocrine

434

N. F. Neumann and M. Belosevic

-m+

0

6

Respiratory bunt Nitric oxide

12

24

72

96

Time (h) Figure 4. Biphasic ROI and RNI production by GMCL. Macrophages were treated with MAF (1:4 final dilution) and LPS (5 pa/ml) and super-oxideand nitrite productionmonitored over 96 h at the specified time intervals. The data are representative of two separate experiments that were performed. Each point represents the mean optical density or pM nitrite ( fS.E.M.) of quadruplicate cultures, blanked agalnst basal levels of unstimuiated macrophages.

factors by themselves, primed the respiratory burst response of goldfish macrophages (Fig. 6). The kinetics of MAF-induced priming of IVDKM were similar to those observed using GMCL. In vitro-derived kidney 0.16

g

cl 1

0

0.1

n @I H

P f U

macrophages underwent rapid priming of respiratory burst responses after 6 h of treatment with MAF or MAF and LPS, but had significantly reduced respiratory burst responses by 48 h in all cultures treated with MAF (Fig. 7). These results

mecrophrgeo only MAFonly MAF + control eupernrtants MAF + lutoertne supematants

0.06

I! 5

Flgun 6. Enhancement of MAF-induced priming of respiratory burst response of GMCL by autocrine factor(s) produced by activated GMCL. Macrophages were pulsed with MAF for 6 h, and washed to remove the MAF stimulus.Supernatants were collected from MAF-pulsed macrophages after an addltional18 h of cultivation.Controlsupernatants were removedimmediately after washing to ensure complete removal of MAE In separate wells, macrophages were pre-treated with autocrine factor(s) contalnlng supematants for 0,6, or 12 h prior to the addltlon of MAE PMA-triggered respiratory burst response was measured 6 h after the addition of MAE The data are representative of three separate experiments that were performed. Each point represents the mean optical den&y (&S.E.M.) of quadruplicate cultures, blanked against the basal resplratory burst levels of the macrophage only group In the ‘0 h pre-treatment’.

Deactivation of cytokine-primed macrophages respiratory burst

435

indicate that this deactivation phenomenon observed in GMCL was not an artifact of the cell line, but represents an important functional response in vivo.

Discussion In the present study, we report on a novel phenomenon of macrophage activation: a selective deactivation of primed respiratory burst responses of activated Time (h) fish macrophages. This deactivating event Figure 6. Priming of the respiratory burst of GMCL by autocrine factor(s). Macrophages were stimu- was selective for the ROI and did not affect the ability of activated macrophages lated with autocrine supernatants for the specified times, and then assayed for PMA-triggered to mount potent RN1 responses. Interestrespiratory burst activity.The data are represenof cytotoxic reactive tative of two separate experiments that were per- ingly, production was formed. Each point represents the mean optical intermediates by fish macrophages density (fS.E.M.) of quadrupficate cultures, biphasic; with maximal ROI production blanked against basal respiratory burst level of unprimed PMA-stimulated macrophages at occuring after only 6 h of stimulus and maximal RNI production occurring 72 h these time points. after stimulation. The results obtained in this study suggest that MAF acts as an important regulatory signal in the activation of fish macrophage antimicrobial mechanisms; initially ‘programming’ macrophages for the induction of temporally spaced anti“‘“1 6houm microbial responses, and possibly initiating delayed deactivation of early induced maximized antimicrobial responses. This selective programming and deprogramming of sequential macrophage antimicrobial armamentarium could be of upmost significance in host defence against pathogens. Deprogramming of macrophage-primed respiratory burst responses may play an important role in preventing unnecessary tissue destruction incurred by prolonged production or inadvertent triggering of ROI production by macrophages. In addition, pathogenic microbes which survive this initial oxidaTreatment tive assault by macrophages (as a result of Ffgure 7. Ability of MAF (1:3 dilution) and/or LPS (10 pg/mL) to prime the respiratory burst of possessing detoxifying enzymes such as IVDKM. In vitro-derived kidney macrophages superoxide dismutase or catalase), may be were treated for 6 or 46 h MAF and/or LPS, before being triggered for respiratory burst susceptible to subsequent antimicrobial activity wlth 100 ng/mL PM. Numbers above the attacks mounted by activated macrobars indicate the number of fish used for each phages (i.e. nitric oxide, peroxynitrite, group. Bars represent the mean optical density tryptophan degradation, iron depriva(fS.E.M.).

n

436

tion). Deactivation of primed respiratory burst responses would minimize tissue damage in an otherwise futile attempt to eliminate reactive oxygen insensitive pathogens. The co-ordination of temporally arranged antimicrobial mechanisms, such as reactive oxygen and nitrogen formation, would maximize the potency of each individual antimicrobial response, allowing macrophages to divert their metabolic energy to a single, highly effective antimicrobial pathway. The enzyme responsible for respiratory burst activity, the respiratory burst oxidase, is similar between the different phagocytic cell types (l), but regulation of its activity (i.e. affinity for substrate), or initiation of transcription of its multiple gene products is differentially regulated (4). Whereas maximal priming effects are observed after 48 h of stimulation of murine macrophages with IFN-y, neutrophils are primed for maximal respiratory burst activity within hours of exposure to this cytokine (4). Likewise, mRNA transcription products of the gp9lphox subunit of the respiratory burst oxidase can be detected much earlier in neutrophils than in macrophages after exposure to IFN-y (4). This raises the possibility that the cells used in our assays may be of the neutrophilic or granulocytic lineage in fish. However, the GMCL have been conclusively identified as macrophagelike, possessing the following characteristics: (1) non-specific esterase activity; (2) morphologic similarity to resident kidney macrophages at both the light and electron microscope level; (3) chemotactic towards endotoxin activated fish serum; (4) able to phagocytose sheep red blood cells and Letihmania major, an obligate intracellular pathogen of mammalian macrophages; (5) capable of producing reactive nitrogen and oxygen intermediates in response to both LPS and/or MAF stimulation; and (6) adherent to glass or plastic within minutes of contact (23, 24). In vitro-derived kidney macrophages also show morphologic and functional simila-

N. F. Neumann and M. Belosevic

rities to GMCL, in that the vast majority are non-specific esterase positive staining cells, and produce reactive nitrogen and oxygen intermediates in response to LPS and/or MAF stimulation (23) (and present manuscript). Thus, goldfish macrophages can be programmed for rapid respiratory burst mediated cytotoxicity against invading pathogens. Comparatively, mammalian macrophages usually require longer exposure periods to cytokines for the initiation of antimicrobial functions (24-118 h). Interestingly, murine macrophage respiratory burst potential can be maximally primed after only 2 h of stimulation with zymosan, suggesting that under certain conditions mammalian macrophages can also be programmed for rapid respiratory burst mediated cytotoxicity (27). The crude MAF preparation used in our experiments made it impossible to attribute the pattern of priming observed to a single cytokine, and as such, other factors found in the crude MAF preparations may play a role in deactivating primed macrophage respiratory burst potentials. However, crude preparations of mammalian MAF, prepared from mitogen-stimulated murine lymphocytes, displays similar respiratory burst priming kinetics as rIFN-y, suggesting that IFN-y is the main mediator of mammalian lymphokine MAF activity (3, 28). The use of crude MAF preparations for the induction of macrophage antimicrobial mechanisms in vitro may mimic more naturally the responses that occur in vivo, in which a complex network of cytokines interact, resulting in a fine tuned but highly effective immune response. We are currently attempting to purify the molecule(s) responsible for priming goldfish macrophage respirtory burst activity. The results obtained in this study contrast those previously described regarding MAF-induced priming of the respiratory burst response of fish macrophages. Graham and Secombes (20-22) demonstrated that MAF primed the

Deactivation of cytokine-primed macrophages respiratory burst

respiratory burst response of rainbow trout resident kidney macrophages after 48 h of cultivation (20-22). They demonstrated that the absolute differences between MAF-stimulated and unstimulated cells were relatively small but significant (i.e. approximately 15%) and that maximal priming occurred 48 h after treatment. More recent experiments have demonstrated a more pronounced effect of priming after 48 h (29-31). This effect has also been demonstrated by others (31). We also found that in some cultures of IVDKM, respiratory burst responses were higher in MAF-induced macrophages than in unstimulated counterparts at 48 h. However, macrophages in parallel cultures showed much greater respiratory burst responses at 6 h than 48 h, suggesting that they were undergoing deactivation of their primed state. There are several plausible explanations for the apparent discrepancy in results: (a) In our study, GCLM and IVDKM were used as effector cells to measure MAFinduced priming of respiratory burst whereas resident kidney responses, macrophages have been used by others. It is well established in mammalian systems, that different sub-populations of macrophages show variable responses to priming and triggering agents of respiratory burst activity (11); (b) Growth factor(s) or autocrine mediators produced by GCLM may cause an acceleration of MAF-induced priming of respiratory burst activity, an effect that also occurs in IVDKM cultures. It is known that GM-CSF, a product of mammalian macrophages, is capable of

437

priming the respiratory burst of murine macrophages much sooner than IFN-y (10); (c) Secretory products of contaminating cell types (albeit to a small extent), cellular debris, or an artifact of isolation, may also affect the kinetics of MAFinduced respiratory burst responses of resident kidney macrophages. The use of GMCL minimixes these potential confounding effects; and (d) Resident kidney macrophages may be undergoing deactivation of their primed state by 24-48 h (the time point at which most assays are done), and if assayed earlier (i.e. at 6 h) may have significantly greater antimicrobial response. Our results support the recent finding that macrophage-derived autocrine factors can enhance teleost macrophage respiratory burst response (31). In addition, our findings on the kinetics of priming induced by LPS are similar to those reported by others, where maximal priming occured after extended cultivation periods, 48 h to several days (32, 33). The results of the present study raise interesting questions regarding cytokineinduced regulation of teleost macrophage cytotoxic mechanisms, and provide insight for examining the sequential regulation of macrophage antimicrobial mechanisms across all classes of vertebrates. Acknowledgements work was supported by Natural Sciences and Engineering Council of Canada (NSERC) and Alberta Heritage Foundation for Medical Research (AHFMR). We thank Dr C. 3. Seccmbes, University of Aberdeen, for his helpful suggestions.

References 1. Dinauer,

hi. C. The respiratory burst oxidase and tbe molecular genetics of chronic gramdomatoua disease. Crit. Rev. Clin. Lab. Sci. 30~329-369; 1993. 2. Iyer. G. Y. N.. Islam, M. F.. Quastcl, J. H. Biocbcmical aspects of pbagocytosis. Nature 19253~541; 1993.

3. Nathan, C. F., Murray, H. W., Wiebc, M. E., Rubin, B. Y. Identilkation of interferon-y as the Iymphokine that actintcs hmmm macropbage oxidative metabolism and antimicrobii activity. J. Exp. Med. 158:670-689; 1983. 4. Cassatella, M. A., Bazzoni,F., Flynn, R. M., Dusi, S., Trincbieri, G., Rossi, F. Molccula~

430

basis of interferon-y and lipopolymccharide enhancement of phagccyte respiratory burst capabiltiy. J. Biol. Chem. 265~20241-20246; 1990. 5. Green, S. P., Hamilton, J. A., Uhlinger, J. D., Phillips, W. A. Expression of p47-phox and p67phox proteins in murine bone marrowderived macrophages: enhancement by lipopolysaccharide and tumor necrosis factor a but not colony stimulating factor 1. J. Let&. Biol. 55530-535; 1994. 6. Ding, A. H., Nathan, C. F., Stuehr, D. J. Release of reactive nitrogen intermediates and reactive oxygen intermediates from mouse peritoneal macrophages: comparison of activating cytokine and evidence for independent uroduction. J. Immunol. 141:2407-2412: 1988. I. ~Camistra, S. A., Vallenga. E., Groshek, P., Rambaldi. A., GritBn. J. D. Human aranulocytamonocyte colony stimulating factor and interleukin3 stimulate monocyte cytotoxicity through a tumor necrosis factor-dependent mechanism. Blood 71:672-676; 1988. 8. Ho, J. L., Reed, S. J., Sobel, J., Arruda, S., He, S. H.. Wick. E. A.. Grabstein. K. H. Interleukin-3 induces antimicrobial activity against L&hmania amozonenrh and Trypanosomacnui and tumoricidal activity in human peripheral blood-derived macrophages. Infect. Immun. 60:198&1993; 1992. 9. Laursen, A. L., Moller. B., Rungby, J., Petersen, C. M., Anderson, P. L. Pneumocysti.r carinii-induced activation of the respiratory burst in human monocytes and macrophages. Clin. Exp. Immunol. 98:196202; 1994. IO. Reed, S. G., Nathan, C. F., P&l, D. L.. Rodericks, P., Shanebeck, K., Conlon, P. J., Grabstein. K. H. Recombinant aranulocvte/ macrophage colony-stimulating factor act&es macrophages to inhibit Trypanosomacruzi and release hydrogen peroxide. J. Exp. Med. l&1734-1746, 1987. 11 Phillips, W. A., Hamilton, J. A. Phorbol esterstimulated superoxide production by murine bone marrowderived macropbages requires preexposure to cytokines. J. Immunol. 142:244>2449; 1989. 12. Phillips, W. A., Croatto, M., Hamilton, I. A. Priming the macrophage respiratory burst with IL-4: enhancement with TNF-a but inhibition by IFN-1. Immunology 70:498-503; 1990. 13. Wing, E. J., Ampel, N. M., Waheed, A., Shadduck, R. K. Macrophage colony-stimulating factor (M-CSF) enhances the capacity of murine macrophages to secrete oxygen reduction products. J. bnmunol. 1352052-2056; 1985. 14. Cbanock, S. J., Benna, J. E., Smith, R. M., Babior, B. M. The respiratory burst oxidase. J. Biol. Chem. 26924519-24522; 1994. 15. Green, S. P., Phillips, W. A. Activation of the macrophage respiratory burst by phorbol myristatc acetate: evidence for both tyrosine-kinasedependent and independent pathways. B&hem. Biophys. Acta 1222:241-248; 1994.

N. F. Neumann and M. Belosevic

16. Secombes. C. J., Fletcher, T. C. The role of phagocytes in the protective mechanisms of fish. AM. Rev. Fish Dis. 253-71; 1992. 17. Anderson, R. S. Hemocyte-derived reactive oxygen intermediate production in four bivalve mollusks. Dev. Comp. Immunol. 18:89-96; 1994. 18. Valembois, P., Lassegues, M. I. vitrogeneration of reactive oxygen species by free coelomic cells of the annelid Eisenkafetida andrei: an analysis by chemiluminescence and nitro blue tetraxolium reduction. Dev. Comp. Immunol. 19:19520+ 1995. 19. Song, Y. L., Hseih, Y. T. Immtmostimulation of tiger shrimp (Pencea.r monodon) hemocytes for generation of microbicidal substances: Analysis of reactive oxygen species. Dev. Comp. Immunol. 18:201-209; 1994. 20. Graham, S., Secombes, C. J. The production of a macrophage-activating factor from rainbow trout Safmo gairdneri leukocytes. Immunology 65:293291; 1988. 21. Graham, S., Secombes, C. J. Do fish lymphocvtes secrete IFN-r? J. Fish Biol. 36563-573: 1-990.

22. Graham, S., Secombes, C. J. Cellular requirements for lymphokine secretion by rainbow trout Salmo gaiheri leukocytes. Dev. Comp. Immunol. 145968; 1990. 23. Neumann, N. F., Fagan, D., Belosevic, M. Macrophage activating factor(s) secreted by mitogen stimulated goldfish kidney leukocytes synergixe with bacterial lipopolysaccharide to induce nitric oxide production in teleost macrophages. Dev. Camp. lmmunol. 19:473482; 1995. 24. Wang, R., Neumann, N. F., Belosevic, M. Establishment and characterization of a macrophage cell line from the goldfish. Fish Shellfish Immunol. 5:329-346; 1995. 25. Rook, G. A. W., Steele, J., Umar, S., Dockrell, H. M. A simule method for tbe solubilisation of reduced NBT, and its use as a calorimetric assay for activation of human macrophages by iinterferon. J. Immunol. Meth. 82161-167: 1985. 26. Green, L. C., Wagner, D. A., Glogowski, J., Skipper, P. L., Wishnok, J. S., Tannenbaum, S. R. Analysis of nitrate, nitrite., and 15nitrate in biological fluids. Analyt. Biochem. 126:131-138; 1982. 27. Assreuy, J., Cunha, F. Q., Epperlein, M., Noronha-Dutra, A., O’Donnell, C. A., Liew, F. Y., Moncada, S. Production of nitric oxide and superoxide by activated macrophages and killing of Leishmmriu major. Eur. J. Immunol. 24672676; 1994. 28. Nakagawara, A., DeSantis, N. M., Nogueira, N., Nathan, C. F. Lymphokines enhance the capacity of human monocytes to secrete reactive oxygen intermediates. J. Clin. Invest. 70:10421048; 1982. 29. Hardie, L. J., Chappell, L. H., Secombes, C. J. Human tumor necrosis factor a influences rainbow trout Oncorhynchur mykiss leukocyte responses. Vet. Imnunol. Immunopatb. 40:7384; 1994.

Deactivation of cytokine-primed macrophages respiratory burst

30. Jang, S. I., Hardie, L. J., Secombm, C. J. Elevation of rainbow trout oncoryJulchus my&i.w macrophage respiratory burst activity with macrophagwlerived supematams. J. Leuk. Biol. 57943-947; 1995. 31. Tahir, A., Secombes, C. J. Modulation of dab (Limrmdo lim&u, L.) macrophage respiratory burst activity. Fish Shell&h Immmol. 6:135146; 1996. 32. Francis, C. H., Ellis, A. E. Production of a

439

lymphokine (macrophage activating factor) by salmon (S&r0 s&r) leukocytes stimulated with outer membrane protein antigens fo Aeromonas sabnonicida. Fish Shelllkh Immunol. 4489-497; 1994. 33. Solem, S. T., Jorgensen, J. B., Robertse.n, B. Stimulation of respiratory burst and phagocytic activity in Atlantic salmon (Salmo salar L.) macrophages by lipopolysaccharide. Fish Shellfish Immunol. fk475-491; 1995.