- Email: [email protected]
Engagement of adenosine receptors inhibits hydrogen peroxide (H2O2−) release by activated human neutrophils
CLINlCALIMMUNOLOGYANDIMMUNOPATHOLOGY
42, 76-85
(1987)
Engagement of Adenosine Receptors Inhibits Hydrogen Peroxide (H,O,) Release by Activated Human Neutrophilsl BRUCE
N. CRONSTEIN,STEVEN
M. KUBERSKY.GERALD ANDROCHELLE HIRSCHHORN
WEISSMANN,
New York Universiry Medical Center. Depurtrnent of Medicine, Divisions of Rheumutology nnd Genetics Adenosine and its analogs, acting at specific cell surface receptors, inhibit generation of superoxide anion by neutrophils. Since it has been suggested that hydrogen peroxide (H,O,) release may not be contingent upon superoxide anion release, we studied the effects of 2-chloroadenosine, a potent adenosine receptor agonist, on the formation of H,O, by neutrophils exposed to various stimuli: n-formyl-methionyl-leucyl-phenylalanine (FMLP). concanavalin A. phorbol myristate acetate (PMA). serum-treated zymosan particles (STZ). and immune complexes. 2-Chloroadenosine (O.Ol- IO p&f) inhibited formation of H,Oz by neutrophils exposed to FMLP, concanavalin A, and STZ particles. As we have found with 0; generation, 2chloroadenosine failed to inhibit H,O, release by neutrophils stimulated by either phorbol myristate acetate or immune complexes. The data show that whereas adenosine and its analogs inhibit neutrophil release of H,Oz and superoxide anion in response to most ligands, they fail to inhibit activation of neutrophils by immune complexes. Nor do they inhibit neutrophil activation by PMA, an agent which bypasses cell surface receptors by direct activation of protein kinase C. Surprisingly, we found that adenosine deaminase activity was adsorbed onto zymosan particles during opsonization and enhanced release of H,Oz by neutrophils exposed to STZ. These studies with yeast cell walls suggest that if microorganisms adsorb adenosine deaminase from serum, then the intracellular microbicidal activity of neutrophils is enhanced. C 1987 Academic Prer~. Inc.
INTRODUCTION
Stimulated neutrophils undergo a respiratory burst in the course of which they release a variety of potentially toxic oxygen metabolites, including superoxide anion (05) and hydrogen peroxide (H,O,). We have recently demonstrated that adenosine and its analogs inhibit 0, generation by human neutrophils exposed to a variety of soluble and particulate stimuli (1, 2). Furthermore, adenosine and its analogs affect neutrophil function via interaction with an adenosine A, receptor on the surface of neutrophils (3). These findings have since been confirmed by others (4). Reports from several laboratories have indicated that H,O, and other oxygen metabolites are derived, either spontaneously or catalytically, from 0; (5). However, recent studies have suggested that the release of 0; and H,O, into the t This work was supported by grants from the American Lung Association, the Lupus Foundation of New York, the New York Arthritis Foundation and by grants from the USPHS, National Institutes of Health (AM-11949, HL-19721, Al-17365, and AI-10343). Dr. Cronstein is a fellow of the Arthritis Foundation and is the recipient of a Clinical Investigator Award (Kl l-AM-01490-01). 76 0090-1229187 $1.50 Copyright C 1987 by Academic Press. Inc. All nght3 of reproduction I” any form reerved.
ADENOSINE
INHIBITS
HYDROGEN
PEROXIDE
RELEASE
77
supernatant medium may not occur in parallel (6, 7). We therefore studied the effects of the engagement of adenosine receptors on the generation of H,O,. We now report that 2-chloroadenosine, a poorly metabolized agonist at the adenosine receptor, inhibits generation of H,O, by cells exposed to n-formyl-methionylleucyl-phenylalanine (FMLP), to concanavalin A (Con A), and to serum-treated zymosan particles (STZ). However, 2-chloroadenosine does not inhibit H,O, generation by cells exposed to phorbol myristate acetate (PMA) or to immune complexes (BSA-anti-BSA); these findings are similar to those we have reported measuring 0; release. We also report that zymosan particles adsorb adenosine deaminase activity from serum. We have previously reported that adenosine deaminase. by metabolizing endogenous adenosine, enhances release of 0; from stimulated neutrophils. Therefore the observation that zymosan particles adsorb adenosine deaminase suggests that some microorganisms (e.g., yeast) may contribute to their own destruction by removing endogenous adenosine from the phagocytic vacuole in which the oxidase resides. METHODS
AND MATERIALS
Materials. Cytochalasin B was purchased from Aldrich Chemical Company, Milwaukee, Wisconsin. Adenosine, 2-chloroadenosine, FMLP, adenosine deaminase (type VIII in ammonium sulfate), scopoletin, bovine serum albumin (essentially fatty acid free), and horseradish peroxidase (type VI, salt-free powder) were purchased from Sigma Chemical Company, St. Louis, Missouri. Zymosan was obtained from ICN Nutritional Biochemicals, Cleveland, Ohio. Superoxide dismutase was obtained from Miles Laboratories, Inc., Elkhart, Indiana. Erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA) was obtained from Burroughs Wellcome & Company, Research Triangle Park, North Carolina. [8-14C]Adenosine (54 mCi/mmol) was obtained from Schwarz/Mann Division, Becton-Dickinson, Spring Valley, New York. Deoxycoformycin (DCF) was a gift from Dr. J. Douros of the National Cancer Institute. Ethyl H,O, was supplied by Polysciences, Inc., Warrington, Pennsylvania, and DEAE cellulose chromatography paper (DE-81) was obtained from Whatman, Inc., Clifton, New Jersey. Rabbit anti-BSA (IgG fraction) was obtained from Cappel Company, Cochraneville, Pennsylvania. Preparation of cell suspensions. Heparinized blood was obtained from normal volunteers. Purified preparations of neutrophils were isolated by means of hypaque/ficoll gradients followed by dextran sedimentation and hypotonic lysis of red blood cells. This procedure allowed studies of cell suspensions containing 98 t 2% neutrophils with few contaminating red cells or platelets. The cells were suspended in a buffered salt solution consisting of Na+ (150 mM), Kf (5 mM), Ca’+ (1.3 mM), Mg*+ (1.2 mM>, Cl- (155 mM), and Hepes (10 mM), pH 7.45 (8). Preparation of immune complexes. The specific antibody from 5 ml of serum was diluted with 5 ml of Hepes buffer, titrated with BSA (2 mg/ml) to maximal absorbance at OD,,O. washed, and resuspended in 5 ml of Hepes buffer (9). Immune complexes were added at a final 1: 10 dilution. Preparation of serum-treated zymosan. Zymosan was suspended in serum at a concentration of 10 mg/ml and incubated at 37°C for 30 min with frequent vor-
78
CRONSTEIN
ET AL.
texing. The particles were washed twice in buffer, resuspended, final concentration of 1 mg/ml (10).
and used at a
Concentrations of soluble stimuli, cytochalasin B, adenosine deaminase, and deoxycoformycin. FMLP was added at a final concentration of 0.1 @l, Con A
was added to cell suspensions at a final concentration of 100 Fg/ml, and PMA was added to suspensions at a final concentration of I &ml. In those experiments in which cytochalasin B was studied, the final concentration of this agent was 5 pg/ml. Adenosine deaminase was added to cell concentrations at a final concentration of 0.25 IU/ml and deoxycoformycin was added at a final concentration of 1 FM. Measurement of hydrogen peroxide generation. Hydrogen peroxide generation was assessed by measurement of horseradish peroxidase-catalyzed oxidation of scopoletin as reported by loss of fluorescence. Duplicate reaction mixtures containing neutrophils (I -2 x 106/ml), scopoletin (6.24 mM), and horseradish peroxidase (0.03 mg/ml) were incubated with agents in the presence or absence of cytochalasin B (5 pg/ml) for 5 min at 37°C. Cells were then exposed to stimuli and incubated for 5 min at 37°C. The reaction was terminated by centrifugation of the reaction mixtures at IOOOg for 5 min at 4°C. Supernates were collected and diluted I:3 in 0.05 M sodium phosphate buffer and fluorescence was measured with excitation at 350 nm and emission at 460 nm. Duplicate tubes containing standards alone (ethyl hydrogen peroxide, 0.2-2 FM) were prepared with scopoletin (2.08 mM final concentration) and horseradish peroxidase (0.01 mg/ml final concentration) in a final volume of 3 ml. Fluorescence of scopoletin decreases in a linear fashion with increasing concentrations of hydrogen peroxide; experimental values were calculated from the standard curve and are expressed as nanomoles of H20, released/lo6 PMN/S min (11). None of the reagents used in these experiments affected measurement of H,Oz by this method. Determination ofadenosine deaminase activity. Conversion of [i4C]adenosine to [14C]inosine and [i4C]hypoxanthine was determined by a modification of the method of Coleman and Hutton (12, 13). Briefly, serum-treated zymosan particles (1 mg/ml final concentration) were incubated with 0.09 mM [i4C]adenosine in 24 mM Tris, 2.4 mM EDTA buffer, pH 7.5, in the presence and absence of 50 PM EHNA. The reaction was terminated at 120 min of incubation by addition of lo-p.1 aliquots of ethanol to 50+ aliquots of the reaction mixture. Aliquots were cochromatographed with standards on DE-81 paper in 1 mM ammonium formate. The relevant areas were identified under uv light, the chromatogram sectioned, and radioactivity of appropriate sections determined. Adenosine is widely separated from hypoxanthine and inosine under these conditions. Activity was calculated by determining the percentage of the total counts recovered which were present as inosine and hypoxanthine. Assays were performed in triplicate. Calculation
of ZC,, of 2-chloroadenosine for hydrogen peroxide generation.
Data were fit to a curve by the ALLFIT program (Dr. Carl Johnson, Department of Pharmacology, University of Cincinnati, College of Medicine, Cincinnati, Ohio) on an Apple IIe desktop computer and the IC,, calculated for the appropriate curves by means of the computer program (14).
ADENOSINE
INHIBITS
HYDROGEN
PEROXIDE
79
RELEASE
RESULTS The effect of 2xhloroadenosine
on H202 release by unstimulated neutrophils.
Unstimulated neutrophils (in the absence of cytochalasin B) released H,O, (Table 1). 2-Chloroadenosine, at a concentration previously shown to maximally inhibit 0; generation (10 t~J4), inhibited H,O, release by unstimulated neutrophils (Table 2). Cytochalasin B significantly reduced the amount of H,O, released by unstimulated neutrophils (Table 1). In the presence of cytochalasin B so little H,O, was released by unstimulated neutrophils that we could not determine whether 2-chloroadenosine inhibited H,O, release from unstimulated cells in the presence of cytochalasin B. The effect of 2-chloroadenosine on H202 release by neutrophils stimulated by soluble stimuli. We have previously reported that 2-chloroadenosine inhibits O,-
release in response to a variety of soluble stimuli (1). We therefore evaluated the effect of 2-chloroadenosine on H,02 release by neutrophils exposed to the soluble stimuli tz-formyl-methionyl-leucyl-phenylalanine (FMLP), concanavalin A (Con A), and phorbol myristate acetate (PMA). After exposure to FMLP, neutrophils released H20, (Table 1). 2-Chloroadenosine inhibited H,O, release by neutrophils exposed to FMLP (Fig. 1). Inhibition depended on the concentration of 2-chloroadenosine present with KS0 = 0.048 ~fr 0.020 t&Y and maximal inhibition of 49 + 1%. Cytochalasin B slightly enhanced H,O, release from neutrophils stimulated by FMLP (Table 1). 2-Chloroadenosine inhibited H,O, release in the presence of cytochalasin B. The IC,, and maximal inhibition of H,O, release by 2-chloroadenosine were not affected by the addition of cytochalasm B (ICsO = 0.094 + 0.013 PM, maximal inhibition = 49 k 1%: Fig. 1 and Table 2). TABLE 1 RELEASEOF H,O, BYNEUTROPHILSINTHEPRESENCEANDABSENCEOFCYTOCHALASINB(~
kg/ml)
H,O, release (nmol/106 PMN/S min) Stimulus None Soluble stimuli FMLP Concanavalin A Phorbol myristate acetate Particulate stimuli Serum-treated zymosan particles Immune complexes
+ Cytochalasin B
(n)
- Cytochalasin B
(n)
0.32
+- 0.10
(7)
0.90
t 0.14*
(7)
4.50 2.82 7.63
-+ 0.60 ‘- 0.27 + 0.83
(4) (4) (41
3.80 1.12 7.30
2 0.60** + 0.27* ? 1.03
(4) (4) (4)
(6)
1.95 4.19
+ 0.20 2 0.20
(6)
1.54 2 0.35 f 0.36
4.60
(3)
(5)
* P < 0.001, + cytochalasin B vs - cytochalasin B, paired t test. ** P < 0.05, + cytochalasin B vs - cytochalasin B, paired t test. Note. 2 x IO6 neutrophils were incubated for 5 min at 37°C in the presence of cytochalasin B (5 kg/ml) or buffer. Stimulus was then added and the cells incubated for 5 min at 37°C. The H,O, concentration in the supernatant medium was then determined as described under Methods and Materials. With the exception of those experiments in which immune complexes were used as the stimulus for H,O, release, the results reported in the presence and absence of cytochalasin B were obtained using PMNs from the same donor in the same experiment.
80 INHIBITIONOF
CRONSTEIN
ET AL.
TABLE H,O, RELEASEBY~-CHLOROADENOSINE(~~ OF~YTOCHALASIN
2 urn) INTHEPRESENCEANDABSENCE B (5 &ml) % Inhibition ( t-SEMI
Stimulus
+ Cytochalasin B
None Soluble stimuli FMLP Concanavalin A Phorbol myristate acetate Particulate stimuli Serum-treated zymosan particles Immune complexes
In)
- Cytochalasin B
(n)
56 2 6*
(8)
52 ‘- 2** 53 -+ 13** -3 k 2
(3) (4) (4)
45 k 5** 36 -+- II*** -2 -’ 2
(4) (4) (4)
50 5 12** 313
(6)
26 t 6** 7_t5
(4) (5)
(3)
* P < 0.001 vs control, Student’s t test. ** P < 0.01 vs control. Student’s t test. *** P < 0.02 vs control, Student’s t test. Note. 2 x IO6 neutrophils were incubated for 5 min at 37°C in the presence of 2-chloroadenosine, buffer, or cytochalasin B (5 ug/ml). Stimulus was then added and the cells incubated for 5 min at 37°C. The H,02 concentration in the supernatant medium was then determined as described under Methods and Materials. With the exception of those experiments in which immune complexes were used as the stimulus for H,O, release. the results reported in the presence and absence of cytochalasin B were obtained using PMNs from the same donor in the same experiment.
Concanavalin A was a poor stimulus of H,O, release from neutrophils (Table 1). Cytochalasin B dramatically enhanced H20, release from neutrophils exposed to concanavalin A (Table 1). 2-Chloroadenosine (10 pM) inhibited H,02 release both in the presence and in the absence of cytochalasin B (Table 2). In previous experiments we have found that adenosine does not inhibit O,-
”
0.01 0.1 1 [Z-Chloroadenos~ne]
10
100
i,uMi
1. 2-Chloroadenosine inhibits H,O, release from neutrophils stimulated by FMLP (0.1 JLM in the presence or absence of cytochalasin B (5 kg/ml). Neutrophils (1-2 x IOVml) were incubated with the indicated concentrations of 2-chloroadenosine or buffer in the presence or absence of cytochalasin B (5 t&ml) for 5 min at 37°C before stimulation with FMLP. After stimulation the cells were incubated for 5 min at 37°C and the reaction was terminated by centrifugation at 800g for 5 min at 4°C. H,O, was then quantitated as described. All data points represent the mean of at least four separate determinations using cells from different donors. Control H,O, release in these experiments was 3.8 + 0.60 nmol/l06 PMN in the absence of cytochalasin B and 4.50 k 0.60 nmol/106 PMN in the presence of cytochalasin B. FIG.
ADENOSINE
INHIBITS
0.
HYDROGEN
.r’
PEROXIDE
RELEASE
81
1-e
0.01
0.1
1
10
100
[ 2- Chloroadenosine] (I/M)
FIG. 2. 2-Chloroadenosine inhibits H,O, release from neutrophils stimulated by STZ (1 mgiml) in the presence and absence of cytochalasin B (5 &ml). Neutrophils (I -2 x 10Vml) were incubated for 5 min at 37°C in the presence of 2xhloroadenosine at the indicated concentrations or buffer in the presence or absence of cytochalasin B. After stimulation with STZ particles the cells were incubated for 5 min at 37°C and the reaction was terminated by centrifugation at 8OOg for 5 min at 4°C. H,O, was then quantitated as described. All data points represent the results of at least four separate determinations on cells from different donors. Control H,O, release was 1.95 2 0.20 nmol/106 PMN in the absence of cytochalasin B and 1.54 k 0.35 nmol/106 PMN in the presence of cytochalasin B.
generation in response to PMA (1). Similarly, H,O, generation by neutrophils exposed to PMA was not inhibited by 2-chloroadenosine (Table 2). Cytochalasin B did not affect H,O, release by neutrophils exposed to PMA (Table l), nor did cytochalasin B enable 2-chloroadenosine (10 @4) to inhibit H,O, release (Table 2). The effect of 2-chloroadenosine on H,O, release by neutrophils stimulated by particulates. We next examined the ability of 2-chloroadenosine to inhibit H,O, release in response to two phagocytic stimuli, serum-treated zymosan particles and immune complexes. Serum-treated zymosan particles were modest stimuli of H,O, release from neutrophils (Table 1). 2-Chloroadenosine inhibited H,O, generation by neutrophils exposed to STZ (Fig. 2). Inhibition of H,O, generation by 2-chloroadenosine depended on its concentration with an IC,, of 3.70 k 9.61 PM and maximal inhibition of 49 + 18%. Cytochalasin B either did not affect or minimally inhibited H,Oz release in response to STZ (Table 1). As reflected by the significantly lower IC,,, 2-chloroadenosine inhibited neutrophil function more dramatically when phagocytosis was inhibited by cytochalasin B (Table 2 and Fig. 2; IC,, = 0.303 + 0.361 PM, maximal inhibition = 60 k 1 I%, F(l.12) = 79.96, P < 0.01). We have previously reported that 2-chloroadenosine does not inhibit O,- generation stimulated by immune complexes (2). Similarly, 2-chloroadenosine only minimally inhibits H,O, release stimulated by immune complexes (Table 2). Cytochalasin B did not affect H,O, generation in response to immune complexes (Table 1) nor did it enable 2-chloroadenosine to inhibit H,O, release (Table 2). Serum-treated zymosan particles possess adenosine deaminase activity which enhances H,O, release. We have previously demonstrated that removal of endogenously released adenosine by addition of adenosine deaminase enhances superoxide anion generation (1). We therefore examined the effect of adenosine deami-
82
CRONSTEIN
ET AL.
nase on HzO, release by neutrophils stimulated by both STZ and FMLP. Adenosine deaminase significantly enhanced H,O, release from neutrophils stimulated by FMLP (Table 3). Surprisingly, adenosine deaminase did not enhance H,O, release by neutrophils exposed to STZ (Table 3). The finding that adenosine deaminase failed to enhance H,O, release from neutrophils stimulated by STZ suggests that these yeast cell wall particles may either retain endogenous adenosine deaminase activity or adsorb adenosine deaminase activity from serum. To test this hypothesis we studied the effect of preincubating STZ with a potent inhibitor of adenosine deaminase, deoxycoformycin (DCFSTZ), on H,O, release provoked by STZ. Deoxycoformycin alone did not affect H,O, release by PMN exposed to FMLP (2 ? 9% inhibition, fz = 5). By contrast, STZ preincubated with deoxycoformycin stimulated release of less H,O, than did STZ (Table 3). Determination of the adenosine deaminase activity of serum-treated zyrnosan particles. Therefore, it appeared likely that adenosine deaminase may have been adsorbed onto the particles from serum. We therefore determined the adenosine deaminase activity of STZ particles. Untreated zymosan particles did not possess detectable adenosine deaminase activity (<3 nmolimg zymosan/hr). STZ particles displayed significant adenosine deaminase activity (Table 4). Adenosine deaminase activity was partially inhibited by the competitive inhibitor of adenosine deaminase EHNA. DISCUSSION
These experiments show that 2-chloroadenosine, an agent previously shown to affect neutrophil function via an adenosine A, receptor, inhibits release of H,O, by unstimulated neutrophils and neutrophils stimulated by FMLP, concanavalin A, and STZ particles. The IC,, for 2-chloroadenosine inhibition of H,O, release and the maximal inhibition of H,O, release by 2-chloroadenosine are similar to TABLE EFFECT OF ADDED ADENOSINE DEAMINASE DEAMINASE ON STZ PARTICLES (DCF-STZ)
3
(ADA) AND INACTIVATION OF ABSORBED ADENOSINE ON O,_ AND H,O, GENERATION BY NEUTROPHILS
Stimulus
Agent
N
H,O, generation (‘3 control ? SEM)
FMLP STZ DCF-ST2
ADA ADA
4 4 4
139 k 12* 97 i 4 83 k 4**
* P < 0.05 vs control, Student’s t test. ** P < 0.01 vs control, Student’s t test. Note. Adenosine deaminase activity absorbed onto STZ particles was inactivated by incubation with deoxycoformycin for 30 min at room temperature. The STZ particles were washed twice before use. 2 x 106 neutrophils were incubated with ADA (0.25 IUlml) or buffer for 5 min at 37°C before addition of stimuli. The cells were then incubated for a further 5 min at 37°C and the reaction terminated by centrifugation at 800g at 4°C for 5 min. Control H,O, generation stimulated by FMLP (0.1 (LM) was 1.955 2 0.150 nmoVl06 PMN/S min. Serum-treated zymosan particles stimulated 1.950 + 0.250 nmol H,0,/106 PMN.
ADENOSINE
INHIBITS
HYDROGEN TABLE
ADENOSINE
DEAMINASE
83
RELEASE
4
ACTIVITY OF SERUM-TREATED AND ABSENCE OF EHNA
Condition
PEROXIDE ZYMOSAN (50 km)
Experiment
1
PARTICLES
IN THE PRESENCE
Experiment 2
Specific activity (nmolimg STZ/hr) STZ STZ + EHNA Physiologic ADA activity
22.9 2 2.3 6.4 k 0.4 16.5
10.8 2 1.2 5.3 + 0.7 5.5
Note. Adenosine deaminase activity of untreated or serum-treated zymosan particles was assayed by determining conversion of [14C]adenosine to [14C]inosine and [‘4C]hypoxanthine after a 2-hr incubation at 37°C. All determinations were performed in triplicate. Physiologic ADA activity is the difference between total and EHNA-inhibitable adenosine deaminase activity. The adenosine deaminase activity of untreated zymosan particles was ~3 nmol/mg STZ/hr.
the ICso and maximal inhibition of 0; release by this compound (I -3). As with 0; generation, 2-chloroadenosine failed to inhibit H,O, release by neutrophils stimulated by either phorbol myristate acetate or immune complexes (1, 2). Thus, as is the case for 0; generation, 2-chloroadenosine is a stimulus-specific inhibitor of H,O, release from neutrophils. Curnutte and Tauber (6) and Hoffstein et al. (7) have presented evidence which supports the hypothesis that release of 0; and H,O, may not be contingent. Our experiments with 2-chloroadenosine do not support this contention: indeed, we have found that adenosine receptors modulate release of 0; and H,O, in parallel. The mechanism(s) by which engagement of adenosine receptors inhibits release of 0; or H,O, in response to these soluble and particulate agents is currently unknown. Stimulus-response coupling in the neutrophil is frequently accompanied by turnover of membrane phospholipids in consequence of which diacylglycerol is formed. This diacylglyceride, in the presence of calcium, activates protein kinase C (IS- 18). Since PMA, like diacylglycerol, directly activates protein kinase C (19-20), engagement of adenosine receptors must inhibit 0; and H,O, release at a step proximal to activation of protein kinase C. Schimmel et al. have reported that adenosine affects phospholipid turnover in rat adipocytes via interaction with specific adenosine receptors on adipocytes (21). Our data suggest that stimulus-response coupling via F, receptors differs from other receptor-ligand interactions at one or more adenosine-insensitive steps proximal to activation of protein kinase C. 2-Chloroadenosine inhibited H,O, release in response to FMLP equally well in the presence or absence of cytochalasin B. By contrast, 2-chloroadenosine is an even more effective inhibitor of H,O, release (Fig. 2) when phagocytosis of STZ particles is partially inhibited by cytochalasin B (10). In addition, we observed greater inhibition by 2-chloroadenosine of FMLP-stimulated than of STZ-stimulated H,O, release. These findings indicate that 2-chloroadenosine inhibits H,Oz release more effectively when phagocytosis either is inhibited or does not occur. As opposed to neutrophils phagocytosing particles, neutrophils stimulated by soluble agents do not internalize their cell membrane and may, in fact, add external
84
CRONSTEIN
ET AL.
membrane in response to some stimuli (22, 23). These observations suggest that adenosine receptors are internalized in the course of phagocytosis and are no longer available to mediate the inhibitory action of adenosine. The adsorption of adenosine deaminase from serum by ST2 particles was intriguing. We have previously shown that serum contains two isozymes of adenosine deaminase which differ with respect to both the K, at which they metabolize adenosine and the genetic loci which code for their synthesis (24). Only one of these isozymes metabolizes adenosine at physiologic concentrations and this enzyme can be identified functionally by the ability of EHNA to inhibit enzymatic activity (25). Our data indicate that zymosan particles adsorb both isozymes from serum, but predominantly that isozyme which metabolizes adenosine at physiologic concentrations. Since 2-chloroadenosine is poorly metabolized by adenosine deaminase, this adenosine analog inhibits H,O, release stimulated by STZ particles despite the presence of the enzymatic activity. The observation that adenosine deaminase is adsorbed onto yeast cell wall particles suggests a novel role for this enzyme in host defense. The H,O,-myeloperoxidase-halide system is one of the mechanisms by which neutrophils kill microorganisms (5, 26, 27). Our studies with yeast cell walls suggest that if microorganisms adsorb adenosine deaminase from serum, then the intracellular microbicidal activity of neutrophils is enhanced. ACKNOWLEDGMENT The authors wish to acknowledge the excellent technical assistance provided by Mr. Stephen Buck.
REFERENCES 1. Cronstein. B. N.. Kramer, S. B., Weissmann, G.. and Hirschhorn, R., J. Exp. Med. 158, 1160, 1983. 2. Cronstein, B. N., Kubersky, S. M., Weissmann, G., and Hirschhorn, R., C/in. Res. 33, 337A. 198.5. 3. Cronstein, B. N., Rosenstein, E. D., Kramer, S. B., Weissmann, G., and Hirschhorn, R., J. Immunol. 135, 1366, 1985. 4. Roberts, P. A., Newby, A. C., Hallett, M. B.. and Campbell, A. K., B&hem. J. 227, 669, 1985. 5. Babior, B. M., Nenl Engl. .I. Med. 298, 659, 1978. 6. Curnutte, J. T.. and Tauber, A. I.. Pediatr. Res. 17, 281. 1983. 7. Hoffstein, S. T., Gennaro, D. E., and Manzi, R. M.. Injlammation 9, 425. 198.5. 8. Boyum. A., Stand. J. C/in. Lab. Invest. 21 (Suppl.). 77, 1968. 9. Ward, P. A., and Zvaifler, N. J.. J. Immunol.111, 1771. 1973. 10. Goldstein, I. M., Roos, D.. Kaplan, H. B., and Weissmann, G., J. C/in. Invest. 56, 1155. 1975. 11. Root, R. K., Metcalf, J., Oshino. N., and Chance, B., J. Clin. Invest. 55, 945, 1975. 12. Coleman, M. S., and Hutton, J. J., Biochem. Med. 13, 46, 1975. 13. Hirschhom, R., Roegner, V., Jenkins, T., Seaman, C.. Piomelli. S.. and Borkowsky, W., J. C/in. Invest. 64, 1130, 1979. 14. De Lean, A., Munson, P. J., and Rodbard. D.. Amer. J. Physiol. 235, E97. 1978. 15. McPhail, L. C., Clayton, C. C., and Snyderman, R., Science 224, 622, 1984. 16. Fujita. I., Irita, K., Takeshige, K., and Minakami. S., Biochem. Biophys. Rev. Commun. 120, 318, 1984. 17. Dewald, B., Payne, T. G., and Baggiolini. M., Biochem. Biophys. Res. Common. 125, 367. 1984. 18. Korchak, H. M., Vienne. K., Rutherford. L. E.. and Weissmann. G., Fed. Proc. 43, 2749. 1984.
ADENOSINE
INHIBITS
HYDROGEN
PEROXIDE
RELEASE
85
19. Castagna. M., Takai, Y.. Kaibuchi, K.. Sano, K., Kikkawa, U.. and Nishizuka, Y., J. Bio/. Chern. 257, 7847. 1982. 20. Wolfson, M., McPhail, L. C., Nasrallah, V. N., and Snyderman, R.. J. Immunol. 135, 2057, 1985. 21. Schimmel, R. J., Honeyman, T. W., and McMahon, K. K., Biochem. J. 212, 499, 1983. 22. Rich, A. M., Cristello, P., Vienne, K., Vosshall, L. B., Haines, K. A., Korchak. H. M.. and Weissmann. G., J. Cell Viol. 101, 234a, 1985. 23. Hoffstein, S. T.. Friedman, R. T., and Weissmann, G., J. Cell Bio/. 95, 234. 1982. 24. Ratech, H., and Hirschhorn, R., C/in. Chim. Acfa 115, 341, 1981. 25. Schrader, W. P., Pollara, B., and Meuwissen, H. J., Proc. Natl. Acad. Sci. USA 75, 446. 1978. 26. Klebanoff, S. J., J. Exp. Med. 126, 1063, 1967. 27. Klebanoff, S. J., J. Bacterial. 95, 2131, 1968. Received May 12. 1986; accepted with revision August 19, 1986
42, 76-85
(1987)
Engagement of Adenosine Receptors Inhibits Hydrogen Peroxide (H,O,) Release by Activated Human Neutrophilsl BRUCE
N. CRONSTEIN,STEVEN
M. KUBERSKY.GERALD ANDROCHELLE HIRSCHHORN
WEISSMANN,
New York Universiry Medical Center. Depurtrnent of Medicine, Divisions of Rheumutology nnd Genetics Adenosine and its analogs, acting at specific cell surface receptors, inhibit generation of superoxide anion by neutrophils. Since it has been suggested that hydrogen peroxide (H,O,) release may not be contingent upon superoxide anion release, we studied the effects of 2-chloroadenosine, a potent adenosine receptor agonist, on the formation of H,O, by neutrophils exposed to various stimuli: n-formyl-methionyl-leucyl-phenylalanine (FMLP). concanavalin A. phorbol myristate acetate (PMA). serum-treated zymosan particles (STZ). and immune complexes. 2-Chloroadenosine (O.Ol- IO p&f) inhibited formation of H,Oz by neutrophils exposed to FMLP, concanavalin A, and STZ particles. As we have found with 0; generation, 2chloroadenosine failed to inhibit H,O, release by neutrophils stimulated by either phorbol myristate acetate or immune complexes. The data show that whereas adenosine and its analogs inhibit neutrophil release of H,Oz and superoxide anion in response to most ligands, they fail to inhibit activation of neutrophils by immune complexes. Nor do they inhibit neutrophil activation by PMA, an agent which bypasses cell surface receptors by direct activation of protein kinase C. Surprisingly, we found that adenosine deaminase activity was adsorbed onto zymosan particles during opsonization and enhanced release of H,Oz by neutrophils exposed to STZ. These studies with yeast cell walls suggest that if microorganisms adsorb adenosine deaminase from serum, then the intracellular microbicidal activity of neutrophils is enhanced. C 1987 Academic Prer~. Inc.
INTRODUCTION
Stimulated neutrophils undergo a respiratory burst in the course of which they release a variety of potentially toxic oxygen metabolites, including superoxide anion (05) and hydrogen peroxide (H,O,). We have recently demonstrated that adenosine and its analogs inhibit 0, generation by human neutrophils exposed to a variety of soluble and particulate stimuli (1, 2). Furthermore, adenosine and its analogs affect neutrophil function via interaction with an adenosine A, receptor on the surface of neutrophils (3). These findings have since been confirmed by others (4). Reports from several laboratories have indicated that H,O, and other oxygen metabolites are derived, either spontaneously or catalytically, from 0; (5). However, recent studies have suggested that the release of 0; and H,O, into the t This work was supported by grants from the American Lung Association, the Lupus Foundation of New York, the New York Arthritis Foundation and by grants from the USPHS, National Institutes of Health (AM-11949, HL-19721, Al-17365, and AI-10343). Dr. Cronstein is a fellow of the Arthritis Foundation and is the recipient of a Clinical Investigator Award (Kl l-AM-01490-01). 76 0090-1229187 $1.50 Copyright C 1987 by Academic Press. Inc. All nght3 of reproduction I” any form reerved.
ADENOSINE
INHIBITS
HYDROGEN
PEROXIDE
RELEASE
77
supernatant medium may not occur in parallel (6, 7). We therefore studied the effects of the engagement of adenosine receptors on the generation of H,O,. We now report that 2-chloroadenosine, a poorly metabolized agonist at the adenosine receptor, inhibits generation of H,O, by cells exposed to n-formyl-methionylleucyl-phenylalanine (FMLP), to concanavalin A (Con A), and to serum-treated zymosan particles (STZ). However, 2-chloroadenosine does not inhibit H,O, generation by cells exposed to phorbol myristate acetate (PMA) or to immune complexes (BSA-anti-BSA); these findings are similar to those we have reported measuring 0; release. We also report that zymosan particles adsorb adenosine deaminase activity from serum. We have previously reported that adenosine deaminase. by metabolizing endogenous adenosine, enhances release of 0; from stimulated neutrophils. Therefore the observation that zymosan particles adsorb adenosine deaminase suggests that some microorganisms (e.g., yeast) may contribute to their own destruction by removing endogenous adenosine from the phagocytic vacuole in which the oxidase resides. METHODS
AND MATERIALS
Materials. Cytochalasin B was purchased from Aldrich Chemical Company, Milwaukee, Wisconsin. Adenosine, 2-chloroadenosine, FMLP, adenosine deaminase (type VIII in ammonium sulfate), scopoletin, bovine serum albumin (essentially fatty acid free), and horseradish peroxidase (type VI, salt-free powder) were purchased from Sigma Chemical Company, St. Louis, Missouri. Zymosan was obtained from ICN Nutritional Biochemicals, Cleveland, Ohio. Superoxide dismutase was obtained from Miles Laboratories, Inc., Elkhart, Indiana. Erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA) was obtained from Burroughs Wellcome & Company, Research Triangle Park, North Carolina. [8-14C]Adenosine (54 mCi/mmol) was obtained from Schwarz/Mann Division, Becton-Dickinson, Spring Valley, New York. Deoxycoformycin (DCF) was a gift from Dr. J. Douros of the National Cancer Institute. Ethyl H,O, was supplied by Polysciences, Inc., Warrington, Pennsylvania, and DEAE cellulose chromatography paper (DE-81) was obtained from Whatman, Inc., Clifton, New Jersey. Rabbit anti-BSA (IgG fraction) was obtained from Cappel Company, Cochraneville, Pennsylvania. Preparation of cell suspensions. Heparinized blood was obtained from normal volunteers. Purified preparations of neutrophils were isolated by means of hypaque/ficoll gradients followed by dextran sedimentation and hypotonic lysis of red blood cells. This procedure allowed studies of cell suspensions containing 98 t 2% neutrophils with few contaminating red cells or platelets. The cells were suspended in a buffered salt solution consisting of Na+ (150 mM), Kf (5 mM), Ca’+ (1.3 mM), Mg*+ (1.2 mM>, Cl- (155 mM), and Hepes (10 mM), pH 7.45 (8). Preparation of immune complexes. The specific antibody from 5 ml of serum was diluted with 5 ml of Hepes buffer, titrated with BSA (2 mg/ml) to maximal absorbance at OD,,O. washed, and resuspended in 5 ml of Hepes buffer (9). Immune complexes were added at a final 1: 10 dilution. Preparation of serum-treated zymosan. Zymosan was suspended in serum at a concentration of 10 mg/ml and incubated at 37°C for 30 min with frequent vor-
78
CRONSTEIN
ET AL.
texing. The particles were washed twice in buffer, resuspended, final concentration of 1 mg/ml (10).
and used at a
Concentrations of soluble stimuli, cytochalasin B, adenosine deaminase, and deoxycoformycin. FMLP was added at a final concentration of 0.1 @l, Con A
was added to cell suspensions at a final concentration of 100 Fg/ml, and PMA was added to suspensions at a final concentration of I &ml. In those experiments in which cytochalasin B was studied, the final concentration of this agent was 5 pg/ml. Adenosine deaminase was added to cell concentrations at a final concentration of 0.25 IU/ml and deoxycoformycin was added at a final concentration of 1 FM. Measurement of hydrogen peroxide generation. Hydrogen peroxide generation was assessed by measurement of horseradish peroxidase-catalyzed oxidation of scopoletin as reported by loss of fluorescence. Duplicate reaction mixtures containing neutrophils (I -2 x 106/ml), scopoletin (6.24 mM), and horseradish peroxidase (0.03 mg/ml) were incubated with agents in the presence or absence of cytochalasin B (5 pg/ml) for 5 min at 37°C. Cells were then exposed to stimuli and incubated for 5 min at 37°C. The reaction was terminated by centrifugation of the reaction mixtures at IOOOg for 5 min at 4°C. Supernates were collected and diluted I:3 in 0.05 M sodium phosphate buffer and fluorescence was measured with excitation at 350 nm and emission at 460 nm. Duplicate tubes containing standards alone (ethyl hydrogen peroxide, 0.2-2 FM) were prepared with scopoletin (2.08 mM final concentration) and horseradish peroxidase (0.01 mg/ml final concentration) in a final volume of 3 ml. Fluorescence of scopoletin decreases in a linear fashion with increasing concentrations of hydrogen peroxide; experimental values were calculated from the standard curve and are expressed as nanomoles of H20, released/lo6 PMN/S min (11). None of the reagents used in these experiments affected measurement of H,Oz by this method. Determination ofadenosine deaminase activity. Conversion of [i4C]adenosine to [14C]inosine and [i4C]hypoxanthine was determined by a modification of the method of Coleman and Hutton (12, 13). Briefly, serum-treated zymosan particles (1 mg/ml final concentration) were incubated with 0.09 mM [i4C]adenosine in 24 mM Tris, 2.4 mM EDTA buffer, pH 7.5, in the presence and absence of 50 PM EHNA. The reaction was terminated at 120 min of incubation by addition of lo-p.1 aliquots of ethanol to 50+ aliquots of the reaction mixture. Aliquots were cochromatographed with standards on DE-81 paper in 1 mM ammonium formate. The relevant areas were identified under uv light, the chromatogram sectioned, and radioactivity of appropriate sections determined. Adenosine is widely separated from hypoxanthine and inosine under these conditions. Activity was calculated by determining the percentage of the total counts recovered which were present as inosine and hypoxanthine. Assays were performed in triplicate. Calculation
of ZC,, of 2-chloroadenosine for hydrogen peroxide generation.
Data were fit to a curve by the ALLFIT program (Dr. Carl Johnson, Department of Pharmacology, University of Cincinnati, College of Medicine, Cincinnati, Ohio) on an Apple IIe desktop computer and the IC,, calculated for the appropriate curves by means of the computer program (14).
ADENOSINE
INHIBITS
HYDROGEN
PEROXIDE
79
RELEASE
RESULTS The effect of 2xhloroadenosine
on H202 release by unstimulated neutrophils.
Unstimulated neutrophils (in the absence of cytochalasin B) released H,O, (Table 1). 2-Chloroadenosine, at a concentration previously shown to maximally inhibit 0; generation (10 t~J4), inhibited H,O, release by unstimulated neutrophils (Table 2). Cytochalasin B significantly reduced the amount of H,O, released by unstimulated neutrophils (Table 1). In the presence of cytochalasin B so little H,O, was released by unstimulated neutrophils that we could not determine whether 2-chloroadenosine inhibited H,O, release from unstimulated cells in the presence of cytochalasin B. The effect of 2-chloroadenosine on H202 release by neutrophils stimulated by soluble stimuli. We have previously reported that 2-chloroadenosine inhibits O,-
release in response to a variety of soluble stimuli (1). We therefore evaluated the effect of 2-chloroadenosine on H,02 release by neutrophils exposed to the soluble stimuli tz-formyl-methionyl-leucyl-phenylalanine (FMLP), concanavalin A (Con A), and phorbol myristate acetate (PMA). After exposure to FMLP, neutrophils released H20, (Table 1). 2-Chloroadenosine inhibited H,O, release by neutrophils exposed to FMLP (Fig. 1). Inhibition depended on the concentration of 2-chloroadenosine present with KS0 = 0.048 ~fr 0.020 t&Y and maximal inhibition of 49 + 1%. Cytochalasin B slightly enhanced H,O, release from neutrophils stimulated by FMLP (Table 1). 2-Chloroadenosine inhibited H,O, release in the presence of cytochalasin B. The IC,, and maximal inhibition of H,O, release by 2-chloroadenosine were not affected by the addition of cytochalasm B (ICsO = 0.094 + 0.013 PM, maximal inhibition = 49 k 1%: Fig. 1 and Table 2). TABLE 1 RELEASEOF H,O, BYNEUTROPHILSINTHEPRESENCEANDABSENCEOFCYTOCHALASINB(~
kg/ml)
H,O, release (nmol/106 PMN/S min) Stimulus None Soluble stimuli FMLP Concanavalin A Phorbol myristate acetate Particulate stimuli Serum-treated zymosan particles Immune complexes
+ Cytochalasin B
(n)
- Cytochalasin B
(n)
0.32
+- 0.10
(7)
0.90
t 0.14*
(7)
4.50 2.82 7.63
-+ 0.60 ‘- 0.27 + 0.83
(4) (4) (41
3.80 1.12 7.30
2 0.60** + 0.27* ? 1.03
(4) (4) (4)
(6)
1.95 4.19
+ 0.20 2 0.20
(6)
1.54 2 0.35 f 0.36
4.60
(3)
(5)
* P < 0.001, + cytochalasin B vs - cytochalasin B, paired t test. ** P < 0.05, + cytochalasin B vs - cytochalasin B, paired t test. Note. 2 x IO6 neutrophils were incubated for 5 min at 37°C in the presence of cytochalasin B (5 kg/ml) or buffer. Stimulus was then added and the cells incubated for 5 min at 37°C. The H,O, concentration in the supernatant medium was then determined as described under Methods and Materials. With the exception of those experiments in which immune complexes were used as the stimulus for H,O, release, the results reported in the presence and absence of cytochalasin B were obtained using PMNs from the same donor in the same experiment.
80 INHIBITIONOF
CRONSTEIN
ET AL.
TABLE H,O, RELEASEBY~-CHLOROADENOSINE(~~ OF~YTOCHALASIN
2 urn) INTHEPRESENCEANDABSENCE B (5 &ml) % Inhibition ( t-SEMI
Stimulus
+ Cytochalasin B
None Soluble stimuli FMLP Concanavalin A Phorbol myristate acetate Particulate stimuli Serum-treated zymosan particles Immune complexes
In)
- Cytochalasin B
(n)
56 2 6*
(8)
52 ‘- 2** 53 -+ 13** -3 k 2
(3) (4) (4)
45 k 5** 36 -+- II*** -2 -’ 2
(4) (4) (4)
50 5 12** 313
(6)
26 t 6** 7_t5
(4) (5)
(3)
* P < 0.001 vs control, Student’s t test. ** P < 0.01 vs control. Student’s t test. *** P < 0.02 vs control, Student’s t test. Note. 2 x IO6 neutrophils were incubated for 5 min at 37°C in the presence of 2-chloroadenosine, buffer, or cytochalasin B (5 ug/ml). Stimulus was then added and the cells incubated for 5 min at 37°C. The H,02 concentration in the supernatant medium was then determined as described under Methods and Materials. With the exception of those experiments in which immune complexes were used as the stimulus for H,O, release. the results reported in the presence and absence of cytochalasin B were obtained using PMNs from the same donor in the same experiment.
Concanavalin A was a poor stimulus of H,O, release from neutrophils (Table 1). Cytochalasin B dramatically enhanced H20, release from neutrophils exposed to concanavalin A (Table 1). 2-Chloroadenosine (10 pM) inhibited H,02 release both in the presence and in the absence of cytochalasin B (Table 2). In previous experiments we have found that adenosine does not inhibit O,-
”
0.01 0.1 1 [Z-Chloroadenos~ne]
10
100
i,uMi
1. 2-Chloroadenosine inhibits H,O, release from neutrophils stimulated by FMLP (0.1 JLM in the presence or absence of cytochalasin B (5 kg/ml). Neutrophils (1-2 x IOVml) were incubated with the indicated concentrations of 2-chloroadenosine or buffer in the presence or absence of cytochalasin B (5 t&ml) for 5 min at 37°C before stimulation with FMLP. After stimulation the cells were incubated for 5 min at 37°C and the reaction was terminated by centrifugation at 800g for 5 min at 4°C. H,O, was then quantitated as described. All data points represent the mean of at least four separate determinations using cells from different donors. Control H,O, release in these experiments was 3.8 + 0.60 nmol/l06 PMN in the absence of cytochalasin B and 4.50 k 0.60 nmol/106 PMN in the presence of cytochalasin B. FIG.
ADENOSINE
INHIBITS
0.
HYDROGEN
.r’
PEROXIDE
RELEASE
81
1-e
0.01
0.1
1
10
100
[ 2- Chloroadenosine] (I/M)
FIG. 2. 2-Chloroadenosine inhibits H,O, release from neutrophils stimulated by STZ (1 mgiml) in the presence and absence of cytochalasin B (5 &ml). Neutrophils (I -2 x 10Vml) were incubated for 5 min at 37°C in the presence of 2xhloroadenosine at the indicated concentrations or buffer in the presence or absence of cytochalasin B. After stimulation with STZ particles the cells were incubated for 5 min at 37°C and the reaction was terminated by centrifugation at 8OOg for 5 min at 4°C. H,O, was then quantitated as described. All data points represent the results of at least four separate determinations on cells from different donors. Control H,O, release was 1.95 2 0.20 nmol/106 PMN in the absence of cytochalasin B and 1.54 k 0.35 nmol/106 PMN in the presence of cytochalasin B.
generation in response to PMA (1). Similarly, H,O, generation by neutrophils exposed to PMA was not inhibited by 2-chloroadenosine (Table 2). Cytochalasin B did not affect H,O, release by neutrophils exposed to PMA (Table l), nor did cytochalasin B enable 2-chloroadenosine (10 @4) to inhibit H,O, release (Table 2). The effect of 2-chloroadenosine on H,O, release by neutrophils stimulated by particulates. We next examined the ability of 2-chloroadenosine to inhibit H,O, release in response to two phagocytic stimuli, serum-treated zymosan particles and immune complexes. Serum-treated zymosan particles were modest stimuli of H,O, release from neutrophils (Table 1). 2-Chloroadenosine inhibited H,O, generation by neutrophils exposed to STZ (Fig. 2). Inhibition of H,O, generation by 2-chloroadenosine depended on its concentration with an IC,, of 3.70 k 9.61 PM and maximal inhibition of 49 + 18%. Cytochalasin B either did not affect or minimally inhibited H,Oz release in response to STZ (Table 1). As reflected by the significantly lower IC,,, 2-chloroadenosine inhibited neutrophil function more dramatically when phagocytosis was inhibited by cytochalasin B (Table 2 and Fig. 2; IC,, = 0.303 + 0.361 PM, maximal inhibition = 60 k 1 I%, F(l.12) = 79.96, P < 0.01). We have previously reported that 2-chloroadenosine does not inhibit O,- generation stimulated by immune complexes (2). Similarly, 2-chloroadenosine only minimally inhibits H,O, release stimulated by immune complexes (Table 2). Cytochalasin B did not affect H,O, generation in response to immune complexes (Table 1) nor did it enable 2-chloroadenosine to inhibit H,O, release (Table 2). Serum-treated zymosan particles possess adenosine deaminase activity which enhances H,O, release. We have previously demonstrated that removal of endogenously released adenosine by addition of adenosine deaminase enhances superoxide anion generation (1). We therefore examined the effect of adenosine deami-
82
CRONSTEIN
ET AL.
nase on HzO, release by neutrophils stimulated by both STZ and FMLP. Adenosine deaminase significantly enhanced H,O, release from neutrophils stimulated by FMLP (Table 3). Surprisingly, adenosine deaminase did not enhance H,O, release by neutrophils exposed to STZ (Table 3). The finding that adenosine deaminase failed to enhance H,O, release from neutrophils stimulated by STZ suggests that these yeast cell wall particles may either retain endogenous adenosine deaminase activity or adsorb adenosine deaminase activity from serum. To test this hypothesis we studied the effect of preincubating STZ with a potent inhibitor of adenosine deaminase, deoxycoformycin (DCFSTZ), on H,O, release provoked by STZ. Deoxycoformycin alone did not affect H,O, release by PMN exposed to FMLP (2 ? 9% inhibition, fz = 5). By contrast, STZ preincubated with deoxycoformycin stimulated release of less H,O, than did STZ (Table 3). Determination of the adenosine deaminase activity of serum-treated zyrnosan particles. Therefore, it appeared likely that adenosine deaminase may have been adsorbed onto the particles from serum. We therefore determined the adenosine deaminase activity of STZ particles. Untreated zymosan particles did not possess detectable adenosine deaminase activity (<3 nmolimg zymosan/hr). STZ particles displayed significant adenosine deaminase activity (Table 4). Adenosine deaminase activity was partially inhibited by the competitive inhibitor of adenosine deaminase EHNA. DISCUSSION
These experiments show that 2-chloroadenosine, an agent previously shown to affect neutrophil function via an adenosine A, receptor, inhibits release of H,O, by unstimulated neutrophils and neutrophils stimulated by FMLP, concanavalin A, and STZ particles. The IC,, for 2-chloroadenosine inhibition of H,O, release and the maximal inhibition of H,O, release by 2-chloroadenosine are similar to TABLE EFFECT OF ADDED ADENOSINE DEAMINASE DEAMINASE ON STZ PARTICLES (DCF-STZ)
3
(ADA) AND INACTIVATION OF ABSORBED ADENOSINE ON O,_ AND H,O, GENERATION BY NEUTROPHILS
Stimulus
Agent
N
H,O, generation (‘3 control ? SEM)
FMLP STZ DCF-ST2
ADA ADA
4 4 4
139 k 12* 97 i 4 83 k 4**
* P < 0.05 vs control, Student’s t test. ** P < 0.01 vs control, Student’s t test. Note. Adenosine deaminase activity absorbed onto STZ particles was inactivated by incubation with deoxycoformycin for 30 min at room temperature. The STZ particles were washed twice before use. 2 x 106 neutrophils were incubated with ADA (0.25 IUlml) or buffer for 5 min at 37°C before addition of stimuli. The cells were then incubated for a further 5 min at 37°C and the reaction terminated by centrifugation at 800g at 4°C for 5 min. Control H,O, generation stimulated by FMLP (0.1 (LM) was 1.955 2 0.150 nmoVl06 PMN/S min. Serum-treated zymosan particles stimulated 1.950 + 0.250 nmol H,0,/106 PMN.
ADENOSINE
INHIBITS
HYDROGEN TABLE
ADENOSINE
DEAMINASE
83
RELEASE
4
ACTIVITY OF SERUM-TREATED AND ABSENCE OF EHNA
Condition
PEROXIDE ZYMOSAN (50 km)
Experiment
1
PARTICLES
IN THE PRESENCE
Experiment 2
Specific activity (nmolimg STZ/hr) STZ STZ + EHNA Physiologic ADA activity
22.9 2 2.3 6.4 k 0.4 16.5
10.8 2 1.2 5.3 + 0.7 5.5
Note. Adenosine deaminase activity of untreated or serum-treated zymosan particles was assayed by determining conversion of [14C]adenosine to [14C]inosine and [‘4C]hypoxanthine after a 2-hr incubation at 37°C. All determinations were performed in triplicate. Physiologic ADA activity is the difference between total and EHNA-inhibitable adenosine deaminase activity. The adenosine deaminase activity of untreated zymosan particles was ~3 nmol/mg STZ/hr.
the ICso and maximal inhibition of 0; release by this compound (I -3). As with 0; generation, 2-chloroadenosine failed to inhibit H,O, release by neutrophils stimulated by either phorbol myristate acetate or immune complexes (1, 2). Thus, as is the case for 0; generation, 2-chloroadenosine is a stimulus-specific inhibitor of H,O, release from neutrophils. Curnutte and Tauber (6) and Hoffstein et al. (7) have presented evidence which supports the hypothesis that release of 0; and H,O, may not be contingent. Our experiments with 2-chloroadenosine do not support this contention: indeed, we have found that adenosine receptors modulate release of 0; and H,O, in parallel. The mechanism(s) by which engagement of adenosine receptors inhibits release of 0; or H,O, in response to these soluble and particulate agents is currently unknown. Stimulus-response coupling in the neutrophil is frequently accompanied by turnover of membrane phospholipids in consequence of which diacylglycerol is formed. This diacylglyceride, in the presence of calcium, activates protein kinase C (IS- 18). Since PMA, like diacylglycerol, directly activates protein kinase C (19-20), engagement of adenosine receptors must inhibit 0; and H,O, release at a step proximal to activation of protein kinase C. Schimmel et al. have reported that adenosine affects phospholipid turnover in rat adipocytes via interaction with specific adenosine receptors on adipocytes (21). Our data suggest that stimulus-response coupling via F, receptors differs from other receptor-ligand interactions at one or more adenosine-insensitive steps proximal to activation of protein kinase C. 2-Chloroadenosine inhibited H,O, release in response to FMLP equally well in the presence or absence of cytochalasin B. By contrast, 2-chloroadenosine is an even more effective inhibitor of H,O, release (Fig. 2) when phagocytosis of STZ particles is partially inhibited by cytochalasin B (10). In addition, we observed greater inhibition by 2-chloroadenosine of FMLP-stimulated than of STZ-stimulated H,O, release. These findings indicate that 2-chloroadenosine inhibits H,Oz release more effectively when phagocytosis either is inhibited or does not occur. As opposed to neutrophils phagocytosing particles, neutrophils stimulated by soluble agents do not internalize their cell membrane and may, in fact, add external
84
CRONSTEIN
ET AL.
membrane in response to some stimuli (22, 23). These observations suggest that adenosine receptors are internalized in the course of phagocytosis and are no longer available to mediate the inhibitory action of adenosine. The adsorption of adenosine deaminase from serum by ST2 particles was intriguing. We have previously shown that serum contains two isozymes of adenosine deaminase which differ with respect to both the K, at which they metabolize adenosine and the genetic loci which code for their synthesis (24). Only one of these isozymes metabolizes adenosine at physiologic concentrations and this enzyme can be identified functionally by the ability of EHNA to inhibit enzymatic activity (25). Our data indicate that zymosan particles adsorb both isozymes from serum, but predominantly that isozyme which metabolizes adenosine at physiologic concentrations. Since 2-chloroadenosine is poorly metabolized by adenosine deaminase, this adenosine analog inhibits H,O, release stimulated by STZ particles despite the presence of the enzymatic activity. The observation that adenosine deaminase is adsorbed onto yeast cell wall particles suggests a novel role for this enzyme in host defense. The H,O,-myeloperoxidase-halide system is one of the mechanisms by which neutrophils kill microorganisms (5, 26, 27). Our studies with yeast cell walls suggest that if microorganisms adsorb adenosine deaminase from serum, then the intracellular microbicidal activity of neutrophils is enhanced. ACKNOWLEDGMENT The authors wish to acknowledge the excellent technical assistance provided by Mr. Stephen Buck.
REFERENCES 1. Cronstein. B. N.. Kramer, S. B., Weissmann, G.. and Hirschhorn, R., J. Exp. Med. 158, 1160, 1983. 2. Cronstein, B. N., Kubersky, S. M., Weissmann, G., and Hirschhorn, R., C/in. Res. 33, 337A. 198.5. 3. Cronstein, B. N., Rosenstein, E. D., Kramer, S. B., Weissmann, G., and Hirschhorn, R., J. Immunol. 135, 1366, 1985. 4. Roberts, P. A., Newby, A. C., Hallett, M. B.. and Campbell, A. K., B&hem. J. 227, 669, 1985. 5. Babior, B. M., Nenl Engl. .I. Med. 298, 659, 1978. 6. Curnutte, J. T.. and Tauber, A. I.. Pediatr. Res. 17, 281. 1983. 7. Hoffstein, S. T., Gennaro, D. E., and Manzi, R. M.. Injlammation 9, 425. 198.5. 8. Boyum. A., Stand. J. C/in. Lab. Invest. 21 (Suppl.). 77, 1968. 9. Ward, P. A., and Zvaifler, N. J.. J. Immunol.111, 1771. 1973. 10. Goldstein, I. M., Roos, D.. Kaplan, H. B., and Weissmann, G., J. C/in. Invest. 56, 1155. 1975. 11. Root, R. K., Metcalf, J., Oshino. N., and Chance, B., J. Clin. Invest. 55, 945, 1975. 12. Coleman, M. S., and Hutton, J. J., Biochem. Med. 13, 46, 1975. 13. Hirschhom, R., Roegner, V., Jenkins, T., Seaman, C.. Piomelli. S.. and Borkowsky, W., J. C/in. Invest. 64, 1130, 1979. 14. De Lean, A., Munson, P. J., and Rodbard. D.. Amer. J. Physiol. 235, E97. 1978. 15. McPhail, L. C., Clayton, C. C., and Snyderman, R., Science 224, 622, 1984. 16. Fujita. I., Irita, K., Takeshige, K., and Minakami. S., Biochem. Biophys. Rev. Commun. 120, 318, 1984. 17. Dewald, B., Payne, T. G., and Baggiolini. M., Biochem. Biophys. Res. Common. 125, 367. 1984. 18. Korchak, H. M., Vienne. K., Rutherford. L. E.. and Weissmann. G., Fed. Proc. 43, 2749. 1984.
ADENOSINE
INHIBITS
HYDROGEN
PEROXIDE
RELEASE
85
19. Castagna. M., Takai, Y.. Kaibuchi, K.. Sano, K., Kikkawa, U.. and Nishizuka, Y., J. Bio/. Chern. 257, 7847. 1982. 20. Wolfson, M., McPhail, L. C., Nasrallah, V. N., and Snyderman, R.. J. Immunol. 135, 2057, 1985. 21. Schimmel, R. J., Honeyman, T. W., and McMahon, K. K., Biochem. J. 212, 499, 1983. 22. Rich, A. M., Cristello, P., Vienne, K., Vosshall, L. B., Haines, K. A., Korchak. H. M.. and Weissmann. G., J. Cell Viol. 101, 234a, 1985. 23. Hoffstein, S. T.. Friedman, R. T., and Weissmann, G., J. Cell Bio/. 95, 234. 1982. 24. Ratech, H., and Hirschhorn, R., C/in. Chim. Acfa 115, 341, 1981. 25. Schrader, W. P., Pollara, B., and Meuwissen, H. J., Proc. Natl. Acad. Sci. USA 75, 446. 1978. 26. Klebanoff, S. J., J. Exp. Med. 126, 1063, 1967. 27. Klebanoff, S. J., J. Bacterial. 95, 2131, 1968. Received May 12. 1986; accepted with revision August 19, 1986