Eur. J. Biochem. 204,705-712 (1992)
0FEBS 1992
Priming of human polymorphonuclear leukocytes with granulocyte-macrophage colony-stimulating factor involves protein kinase C rather than enhanced calcium mobilisation Monika SCHATZ-MUNDING and Volker ULLRICH Faculty of Biology, University of Konstanz, Federal Republic of Germany
(Received August 30/0ctober 25, 1991) - EJB 91 1165
Pretreatment of human polymorphonuclear leukocytes with the recombinant human granulocytemacrophage colony-stimulating factor (rhGM-CSF) enhances leukotriene biosynthesis in response to a receptor agonist (e.g. N-formyl-methionyl-leucyl-phenylalanine, fMLP) or a Ca2+-ionophore (e.g. ionomycin). This priming effect could be traced back to an elevated release of arachidonic acid from the phospholipid pools and hence an increased leukotriene biosynthesis by 5-lipoxygenase. Preincubation of polymorphonuclear leukocytes with GM-CSF did not influence the basal intracellular Ca2+ level and does not enhance cytosolic free calcium after stimulation with fMLP or ionomycin. Only a small increase in the second Ca2+ phase after receptor agonist stimulation was found. However, the Ca2+-threshold level necessary for the liberation of arachidonic acid by phospholipase A2 was decreased from 350-400 nM calcium in untreated cells to about 250 nM calcium in primed cells. This allows phospholipase A2 to be activated by a release of calcium from intracellular stores and by ionomycin concentrations which are ineffective in untreated cells. Protein biosynthesis inhibitors like actinomycin D (10 pg/ml) and cycloheximide (50 pg/ml) had no effect on the enhanced leukotriene biosynthesis in primed cells after stimulation with ionomycin. However staurosporine (200 nM), an inhibitor of protein kinase C totally abolished the priming effect of GM-CSF after stimulation with ionomycin. The priming effect of GM-CSF could be mimicked by phorbol myristate acetate (PMA; 1 nM) and no additive or synergistic effect was found on leukotriene biosynthesis by simultanous pretreatment with PMA and GM-CSF and stimulation with either fMLP or ionomycin. These results provide evidence that the enhanced arachidonic acid release in GM-CSF-primed polymorphonuclear leukocytes after stimulation with either fMLP or ionomycin involves activation of protein kinase C which, by a still unknown mechanism, reduces the Caz+ requirement of phospholipase A l .
neutrophils include inhibition of migration, changes in receptor expression and effects on the cytoskeleton [5, 81. Again, different from the immediately observed effects, GM-CSF is known to 'prime' human neutrophils to an enhanced functional state in response to physiological agonists, e.g. N-formyl-methionyl-leucyl-phenylalanine (fMLP), platelet-activating factor (PAF), complement factor C5a or a Ca2'ionophore, e.g. ionomycin. Preincubation of the cells with GM-CSF will lead to an enhanced generation of superoxide [9 - 111, phagocytosis [12], liberation of arachidonic acid [I 31, PAF and leukotriene biosynthesis [14- 181. GM-CSF exerts its priming effect through a single class of Correspondence fa V. Ullrich, Fakultat fur Biologie, Universitat Konstanz, Postfach 5560, W-7750 Konstanz, Federal Republic of high-affinity receptors with an apparent molecular mass of 84 kDa on human cells [19, 201. Human neutrophils, Germany Ahhrrviations. A,Ach, arachidonic acid; fMLP, N-formyl- macrophages and eosinophils express the largest number of methionyl-lcucyl-phenylalanine; rhGM-CSF, recombinant human high-affinity receptors (Kd = 37 pM, 300 - 1000 sites/cell [I 91. granulocyte-macrophage colony-stimulating factor; HO-A,Ach, The subsequent signal transduction involves a guaninehydroxyeicosatetraenoic acid; LTB,, leukotriene Bq;PAF, platelet- nucleotide-binding protein [21, 221 which is pertussis-toxinactivating factor; PMA, phorbol myristate acetate; wOH-LTB,, osensitive. hydroxy-LTB,; wCOOH-LTB,, o-carboxy-LTB,. We have investigated the priming effect of GM-CSF on Enzymes. Arachidonate 5-lipoxygenase (EC 1.13.11.34); phosleukotriene biosynthesis in human PMN since leukotrienes pholipase Az (EC 3.1.1.4); phospholipase C (EC 3.1.4.3); phospholipase D (EC 3.1.4.4); protein kinase C (EC 2.7.1.37). play an important role in inflammation [23] and the GM-CSF
The human cytokine granulocyte-macrophage colonystimulating factor (GM-CSF) is a 22-kDa glycoprotein produced by activated T-lymphocytes, endothelial cells, fibroblasts, monocytes and macrophages [l- 31. It stimulates the proliferation and differentiation of erythroid and myeloid stem cells [4-71. GM-CSF is not found in the circulation at detectable levels, suggesting a paracrine action. Apart from its differentiating activity on the precursor cells, it is also capable of acting directly on mature effector cells like eosinophils, neutrophils and monocytes. Such immediate effects on human
706 effect has roused considerable interest [13, 14, 161. Although it is well documented that GM-CSF exerts its priming effect by enhancing the liberation of arachidonic acid after stimulation [4], the underlying mechanism is not known. Leukotriene biosynthesis involves the liberation of arachidonic acid by phospholipases, especially phospholipase A2 124,251 and the oxygenation by 5-lipoxygenase [26]. Both steps are Ca2+-dependent processes [27]. Free intracellular arachidonic acid levels are very low [28] and controlled not only by the liberation but also by the reacylation of this fatty acid. The regulation of 5-lipoxygenase is complex. Besides calcium, it needs ATP for maximal activity and undergoes translocation to a membrane site [29]. The recently identified '5-lipoxygenase activating protein' in membrane seems to play an essential role in activation of 5-lipoxygenase in whole cells [301. In this report we show that GM-CSF primes human polymorphonuclear leukocytes such that subsequent activation of phospholipase A2 by either ionomycin or fMLP requires a smaller rise in intracellular calcium compared to unprimed cells. The effect of GM-CSF appears to be mediated by protein kinase C.
HPLC were stored in 70 pl methanol at -20°C and mixed with 45 pl injection buffer immediately before analysis.
Measurements of intracellular Ca2 levels +
The Fura-2 method was performed as described [31]; lo7 polymorphonuclear leukocytes/ml buffer A containing 2 mM glucose and 1 mM CaC12 were loaded with 2 pM Fura-2 for 30 min. Calibration of intracellular calcium was performed by the method of Grynkiewicz et al. [32]. Maximal fluorescence (Fmax) was achieved by completely lysing the cells with 100 pM digitonin and Fmin by complexing calcium with 10 mM EGTA. For GM-CSF-priming, Fura-2-loaded cells were incubated for 30 rnin with 2 nM GM-CSF at only 25°C to avoid further intracellular distributions of the fluorescence dye. The influx of extracellular calcium was measured by using the manganese quenching method [33]. lo7 polymorphonuclear leukocytes/ml 10 mM Hepes, 150 mM NaCl, 5 mM KCI pH 7.4, 280 mOsm (buffer B) containing 2 mM glucose and 0.5 mM CaCl, were loaded with 2 pM Fura-2 for 30 rnin as described [31]. After addition of 0.2 mM MnCl,, the fluorescence emission above 450 nm after excitation at 364.5 nm was recorded.
MATERIALS AND METHODS Chemicals fMLP, dextran ( M , 506 000), digitonin, cycloheximide, actinomycin D, phorbol myristate acetate (PMA) and EGTA were purchased from Sigma Chemical Co. (Deisenhofen, FRG). Ardchidonic acid was obtained from Larodan (Malmo, Sweden) and [l-'4C]arachidonic acid from New England Nuclear (Dreieich, FRG). PAF and prostaglandin B2 were purchased from Paesel (Frankfurt, FRG), Fura-Z/AM (pentaacetoxymethylester) from Boehringer Mannheim (Mannheim, FRG) and Ficoll-paque from Pharmacia GmbH (Freiburg, FRG). Ionomycin and staurosporine were purchased from Calbiochem (Frankfurt, FRG). Genistein and tyrphostin were products from Gibco (Berlin, FRG). rhGM-CSF was a kind gift from Dr Seiler and Dr Krumwieh (Behringwerke Marburg, FRG). All other chemicals and solvents utilized were of HPLC or analytical grade.
Polymorphonuclear leukocytes and incubations The preparation of human polymorphonuclear leukocytes was performed by dextran sedimentation and subsequent centrifugation on Ficoll-paque as described [31].The viability, measured by trypan-blue exclusion, was greater than 96% and purity of the neutrophil granulocytes was about 95%. After isolation the granulocytes were stored at 4°C in buffer A (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na2HP04, 1.5 mM KH2P04, pH 7.4,280 mOsm). lo7 polymorphonuclear leukocytes in buffer A containing 1 mM CaClz were incubated either with or without rhGMCSF (2 nM) by permanent stirring conditions for 30 min at 30°C. Then the cells were prewarmed to 37°C for 5 min and stimulated for further 5 rnin at 37 "C.
Reverse-phase HPLC analysis Extraction and HPLC analysis of total arachidonate metabolites (oCOOH- and oOH-LTB,, 5-HO-d4Ach, LTB4 and its isomers) were performed as described [31]. Probes for
Measurement of phospholipase A2 activity For these experiments lo7polymorphonuclear leukocytes/ ml in buffer A, 1 mM CaCl,, were incubated with 2 pM [l14C]arachidonic acid (55 Cilmol) for 15 min at 35 -C, then washed twice with prewarmed buffer A, 1 mM CaCl,, and kept at room temperature until use. lo7 labeled cells/ml were preincubated for 5 rnin at 3 7 T , then stimulated with different concentrations of ionomycin. After 5 min, the suspension was centrifuged for 10 s at 10000 x g and radioactivity in the supernatant was measured in a liquid scintillation counter (PN 4700, Isomess, Strdubenhardt, FRG).
Reacylation of free arachidonic acid After incubation with 10 pM arachidonic acid (1 pM [l14C]arachidonic acid), 1 ml of the cell suspension was transferred to 2.4 ml chloroform/methanol (1 : 1, by vol.) and lipids were extracted at pH 3 according to Bligh and Dyer [34]. The chloroform phase was evaporated under nitrogen and dissolved in 70 p1 chloroform/methanol (9 : 1, by vol.). The radioactivity in 20 pl was measured while 50 pl was analyzed by thin-layer chromatography in a solvent system of petrol ether/diethyl ether/acetic acid (90:25: 1, by vol.) [31].
Statistical analysis Results are expressed as mean f SEM of the numbers of experiments indicated.
RESULTS Stimulation of human polymorphonuclear leukocytes with the receptor agonist fMLP results in the generation of only low amounts of HPLC-detectable 5-lipoxygenase metabolites after an incubation time of 5 min. Preincubation with different concentrations of GM-CSF for 30 rnin and subsequent stimulation with fMLP increases the synthesis of leukotrienes by
707
human granulocytes about fivefold (Fig. 1).For further experiments 2 n M GM-CSF were used for an optimal priming effect and a preincubation time of 30 rnin also proved to give maximal yields (data not shown). GM-CSF alone does not result in any HPLC-detectable amounts of 5-lipoxygenase metabolites. The concentration- and time-dependent priming effect of GM-CSF on leukotriene biosynthesis in polymorphonuclear leukocytes was also found by using the Ca2+ionophore ionomycin (0.05 pM) instead of fMLP (data not shown). No significant difference was found by using 25 "C instead of 30'C for priming with GM-CSF (data not shown). Our special interest referred to the effect of GM-CSF on intracellular Ca2+ levels, since we recently showed that the liberation of arachidonic acid by phospholipase A2 in human neutrophils after stimulation by the Ca2'4onophore ionomycin needs a permanently elevated intracellular CaZ level above a threshold concentration of about 350-400 nM calcium [27]. This Ca2+ level was exceeded with ionomycin concentrations above 0.05 pM. The same experiments were performed with GM-CSF-primed polymorphonuclear leuko-
cytes. Fura-2-loaded cells were incubated for 30 rnin with 2 nM GM-CSF and, after stimulation with different concentrations of ionomycin, the intracellular Ca2+ levels were monitored. After 5 rnin of stimulation, the amount of 5lipoxygenase metabolites were measured by HPLC. Fig. 2 shows that the threshold Ca2+ level necessary for the generation of 5-lipoxygenase metabolites was decreased in GMCSF-treated cells to about 250 nM intracellular calcium (depending on the blood donor) compared to 350-400 nM calcium in untreated cells [27. This Ca2+-threshold level was exceeded with even 0.01 pM ionomycin (Figs 5 and 6). Priming with GM-CSF had no effect on either the basal CaZ' levels nor the enhanced Ca2+ levels after stimulation with different concentrations of ionomycin (data not shown).
+
[Caf'l (nM)
= !J J J
LOO
1
300
-
200
-
[lonomycin] [ 1 5 - L o x Met (ng)
0.05
265 180
0.02
101
0 01
11
0.002
/-J
loo1 50
9
m
L
0
0
1 0.01
0.1 [GM-CSF]
L 10
nd
H
lmin
lonomycin
(nM)
Fig. 1. Dose/response relationship of the priming effect of GM-CSF on fMLP-induced leukotriene biosynthesis. Polymorphonuclear leukocytes (lo7 cells/ml in buffer A, 1 mM CaCI2) were preincubated for 30 rnin at 3O'C with diffcrent concentrations of GM-CSF as indicated and stimulated with 0.1 pM fMLP; 5 min after stimulation, aliquots of the incubations were analyzed for 5-lipoxygenase (5-Lox) metabolites by HPLC. The results are expressed as mean f SEM from four different experiments.
Fig. 2. Determination of the Ca2+-threshold level for the generation of 5-lipoxygenase metabolites in GM-CSF-primed cells after stimulation with ionomycin. Fura-2-loaded polymorphonuckdr leukocytes ( lo7 cells/ml in buffer A containing 2 mM glucose and 1 mM CaCI2) were pretreated with 2 nM GM-CSF at 25°C for 30 min. The cells were incubated for 3 min at 37°C and then stimulated with different concentrations of ionomycin. After 5 rnin, 1-ml aliquots were transferred to ethyl acetate and analyzed for 5-lipoxygenase metabolites (C5Lox Met.: 5-OH-d4Ach, wCOOH-LTB4, wOH-LTB4, LTB4 and its isomers). The figure shows one typical experiment out of four.
n
n -GM~CSF .."..
. i
t
]
t
fMLP
c c r A
H
lmin
GM-CSF p....._...._/ +
t fMLP
..._ ....._
-GM-CSF
Fig. 3. Effect of GM-CSF on intracellular Ca" levels. Fura-2-loaded polymorphonuclear leukocytes (lo7cells/ml in buffer A containing 2 mM glucose and 0.5 mM CaC1,) were preincubated with or without GM-CSF for 30 rnin at 25°C. Then cells were prewarmed to 37°C for 3 rnin and stimulated with 0.1 pM fMLP. To complex the extracellular calcium, 1 mM EGTA was added 2 rnin before fMLP stimulation. The monitored Ca2+ curves show one typical experiment out of four.
708
GM - CSF '\\ 8.
'+
GM- CSF
'+ G M - C S F
Fig. 4. CaZ+ influx in GM-CSF primed-cells with (A) fMLP and (B) ionomycin measured by MnZ+influx. Fura-2-loaded polymorphonuclear leukocytes (lo' cells/ml in buffer B containing 2 mM glucose and 0.5 mM CaC12) were preincubated with or without 2 nM GM-CSF for 30 min at 25°C. Then 0.2 mM MnClz was added and the cells were prewarmed to 37°C for 3 min and stimulated with (A) 0.1 pM fMLP or (B) 0.05 pM ionomycin. The monitored fluorescence curves show one typical experiment out of three.
By using the receptor agonist fMLP we also found no significant change in the amount of the intracellular Ca2+ signal in GM-CSF-primed neutrophils, in contrast to unprimed cells after stimulation with 0.1 pM fMLP. Fig. 3 shows the intracellular Ca2+ transient in the presence and absence of extracellular calcium. In the presence of extracellular calcium. a small increase in the second phase of the Ca2+ signal was always observed either due to influx of extracellular calcium or inhibition of the Ca2+-ATPase. When extracellular calcium was complexed by a molar excess of EGTA, no increase in the intracellular CaZ transient (calcium released from the intracellular Ca2+ stores) in GM-CSF-primed cells compared to unprimed cells was detectable. Only a somewhat slower decline to the basal Ca2+level was always observed in GM-CSF-treated cells. Measurements of the Ca2+ influx by the method of manganese quenching of the Fura-2 signal shows the Ca2 influx mediated by the receptor agonist fMLP or ionomycin. We also found no significant difference in this Ca2+influx after stimulation with either fMLP or ionomycin in GM-CSF-primed cells in comparison to unprimed cells (Fig. 4). These results suggest that one effect of GM-CSF is to slightly reduce Ca2' efflux by the Ca2+-ATPase. In unprimed human polymorphonuclear leukocytes the extracellular calcium was absolutely necessary for the generation of 5-lipoxygenase metabolites after stimulation with either fMLP or ionomycin [27]. As Table 1 shows, extracellular calcium had no influence on leukotriene biosynthesis in GM-CSF-primed cells after stimulation with the receptor agonist fMLP. Most striking, and a qualitative difference to untreated cells, was that after stimulation with ionomycin significant amounts of 5-lipoxygenase metabolites were found in the absence of extracellular calcium in GM-CSF-primed cells. The release of arachidonic acid was measured with [l'"Clarachidonic-acid-labeled neutrophils as described in Materials and Methods. Fig. 5 shows the appearance of free radioactivity (arachidonic acid and its metabolites) in the supernatant of GM-CSF-primed and unprimed granulocytes. The dose/response curve with ionomycin started at much lower concentrations of ionomycin (0.01 pM) than in unprimed cells (0.05 pM ionomycin) although the correspond-
Table 1. Influence of extracellular calcium on leukotriene biosynthesis in control and GM-CSF-primed cells. Polymorphonuckdr lcukocytes (PMN; 107/ml in buffer A, 0.5 mM CaCI,) were pretreated with or without 2 nM GM-CSF and stimulated with either 0.1 WMfMLP or 2 pM ionomycin. To complex the extracellular calcium, 1 mM EGTA was added 2 min before stimulation; 5 min after stimulation, incubations were stopped and analyzed by HPLC for 5-lipoxygenase (5Lox) metabolites. Results are expressed as mean SEM from thrcc different experiments.
Agonist
CaZ
+
GM-CSF
5-Lox mctabolitcs
+
+
ng/107 P M N fMLP
++ -
Ionomycin
+ + -
-
+ + -
+ +
6+ I_+ 33f34+
2 1 5 8
299 k 49 18+ 1 349 59 144k 7
*
ing intracellular Ca2+ levels were not changed (Fig. 2). GMCSF alone does not lead to any liberation of arachidonic acid. The same results were obtained when the 5-lipoxygenase metabolites were measured by HPLC after stimulation with different concentrations of ionomycin in GM-CSF-primed and untreated cells. In granulocytes preincubated for 30 min with 2 nM GM-CSF 5-lipoxygenase metabolites were found with wen 0.01 pM ionomycin, whereas in cells preincuhated without GM-CSF at least 0.05 pM ionomycin was necessary (Fig. 6). These results substantiate an increase in the maximum response as well as a shift to the left in the dose, response curve in GM-CSF-primed cells by ionomycin-induced arachidonic acid liberation and leukotriene biosynthesis. In human polymorphonuclear leukocytes the intracellular concentration of free arachidonic acid is also controlled by the reacylation of the fatty acid into the membranes. Therefore one possible effect of GM-CSF could have been an inhibition
709 Table 2. Reacylation of free arachidonic acid. Polymorphonuclear leukocytes (PMN; 107/ml in buffer A, 1 mM CaCI,) were preincubated with or without 2 nM GM-CSF and stimulated for 5 min with either 30 BM arachidonate. (d4Ach; 1 pM [l-'4C]arachidonate, 51 850 cpm) alone or together with 0.1 pM fMLP or 2 pM ionomycin. After lipid extraction, the incubations were analyzed by thin-layer chromatography. Values arc expressed as mean SEM from three different experiments.
t GM-CSF
Agonist 0
-
-
/?
0.01
1
0.1 [ Ionomycin]
(FM)
Fig. 5. Priming effect of GM-CSF on arachidonic acid liberation. [I 14C]Arachidonic-acid-labelledpolymorphonuclear leukocytes (lo7/ ml in buffer A, 1 mM CaCl,) were preincubated with or without 2 nM GM-CSF and stimulatcd with different concentrations of ionomycin; 5 min after stimulation, aliquots of the supernatant were analyzed in a liquid scintillation countcr for radioactivity. The values are given as mcan f SEM (cpm/107PMN) from three different experiments.
FMLP
+ A4Ach
GM-CSF
Free arachidonic acid
-
cpm/107PMN 11348 f 1809
+
13619 & 2161
Ionomycin + A,Ach
-
6275f 510 5893+ 381
A4Ach
-
17588 f 4857 18974 f 3714
+ +
Table 3. Effect of staurosporine on GM-CSF-enhanced leukotriene biosynthesis after stimulation with ionomycin. Polymorphonuclear leukocytes (PMN; 107/ml in buffer A, 1 mM CaC12) were preincubated with or without 200 nM staurosporine at 20°C for 5 min, thcn incubated with or without 2 nM GM-CSF and stimulated with 0.05 pM ionomycin; 5 rnin after stimulation,an aliquot was analyzed by HPLC for total amount of 5-lipoxygenase metabolites. Results are expressed as mean 5 SEM from three different experiments. Staurosporine
5-Lipoxygenase metabolites - GM-CSF
+ GM-CSF
ng/107 PMN 0 X
? In
0.01
0.1
1
[Ionomycin] (PM) Fig. 6. Priming effect of GM-CSF on leukotriene biosynthesis. lo7 polymorphonuclear leukocytes/ml in buffer A, 1 mM CaCl,, were preincubated with or without 2 nM GM-CSF and stimulated by different concentrations of ionomycin; 5 min after stimulation, aliquots were transferred into ethyl acetate and analyzed for 5-lipoxygenase (5-Lox) metabolites (5-OH-A4Ach, toCOOH-LTB4, wOH-LTB4, LTR, and its isomcrs) by HPLC. Values are given as mean & SEM from three different experiments.
of the reacylation so that the levels of free arachidonic acid are elevated in GM-CSF-primed cells. To our knowledge this possibility has never been investigated. To test the hypothesis that GM-CSF increases the intracellular concentration of arachidonic acid by inhibition of the reacylation reaction, we measured the incorporation of exogenous [l-'4C]arachidonate into the phospholipids after priming with GM-CSF compared to untreated cells (Table 2). No significant difference in unesterified levels of arachidonic acid 5 rnin after stimulation with either 2 pM ionomycin or 0.1 pM fh4LP plus 10 pM arachidonate or 10 pM arachidonate alone was found. However the total amount of free arachidonic acid differed between the two stimuli due to activation of the reacylation by ionomycin. An explanation for the prolonged time period necessary for the priming effect of GM-CSF could be that new protein are synthesized. However, GM-CSF-enhanced leukotriene
-
+
124 f 57 109 f 45
202 f 56 107 f 43
biosynthesis after stimulation with ionomycin or fMLP was not affected by preincubation with either actinomycin D, an inhibitor of transcription, or cycloheximide, a n inhibitor of translation. Polymorphonuclear leukocytes were incubated for 5 min with either actinomycin D (10 pg/ml) or cycloheximide (50 pg/ml) and then GM-CSF was added for a further 30 min. The cells were stimulated with 0.05 pM ionomycin or 0.1 pM fMLP and 5-lipoxygenase metabolites were measured by HPLC. Actinomycin D o r cycloheximide alone had no direct effect nor did they inhibit leukotriene biosynthesis in GM-CSF-treated cells (data not shown). Also a preincubation time of 30 rnin with either actinomycin D or cycloheximide did not suppress the priming effect of GM-CSF on ionomycin-induced leukotriene biosynthesis. Addition of staurosporine, known as an inhibitor of protein kinase C, 5 rnin before priming the neutrophils with GMCSF totally abolished the increased leukotriene biosynthesis in ionomycin-stimulated cells (Table 3). In untreated neutrophils, staurosporine had no effect on leukotriene biosynthesis after stirnulation with ionomycin. The concentration necessary for complete inhibition of the GM-CSF effect was, however, much higher than reported in the literature for the inhibition of protein-kinase-C-mediated events [35],so that other phosphorylation reactions had also to be taken into account. To test whether tyrosine phosphorylation is involved in the
710
C
PMA GM-CSF PMA
GM-CSF
K
.?I 0
[14- IS]. GM-CSF by itself does not directly exert this effect but primes the cells for subsequent enhanced responses to a second stimulating agent like the receptor agonist fMLP or the Ca'+-ionophore ionomycin. Our results clearly demonstrate and support previous reports [13] that preincubation with GM-CSF enhances arachidonic acid liberation in polymorphonuclear leukocytes after stimulation with ionomycin. This increased substrate availability was often discussed as a result of GM-CSF-mediated increased calcium release after receptor agonist stimulation. Recent publications show an increase in the intracellular Ca" transient in GM-CSF-preincubated neutrophils after stimulation with the chemotactic factors fMLP, PAF and LTB4 [lS, 381. In contrast, Sullivan et al. found no Ca2+ increase in GM-CSF-treated cells after stimulation with fMLP [39]. By our experiments we could exclude significant effects of GM-CSF on the fMLP- or ionomycin-induced release of calcium from intracellular stores or on the increased influx of calcium from the extracellular side. Only a slightly reduced Ca2+ efflux via Ca2+-ATPase was observed in GM-CSFprimed cells after stimulation with fMLP. However, the threshold level necessary for the release of arachidonic acid after stimulation with ionomycin was clearly decreased from 350 - 400 nM intracellular calcium in unprimed cells [27] to about 250 nM intracellular calcium in GM-CSF-primed cells. As an important consequence of the release of intracellular calcium, the cells now became able to activate phospholipase A2 and hence to synthesise leukotrienes. Similar results were reported for activation of the oxidative burst in human neutrophils. In GM-CSF-treated cells elevation of intracellular calcium by a subthreshold concentration of ionomycin is sufficient to activate the oxidative burst [40]. Priming of human polymorphonuclear leukocytes with tumor necrosis factor also enhances phospholipase A2 activity without any effect on intracellular calcium levels [41]. Since ionomycin does not directly act on phospholipase C [24], the liberation of arachidonic acid should be mediated by phospholipase A2 which seems to be the target of GM-CSF action. In human cells the level of free intracellular arachidonic acid is very low [28] and controlled by hydrolysis and reacylation. The increased level of free arachidonic acid in GM-CSFprimed polymorphonuclear leukocytes after stimulation with ionomycin was not due to an inhibition of reacylation of free arachidonic acid into the lipid membrane. No difference in concentrations of exogenously submitted arachidonic acid was found in GM-CSF-treated cells compared to control cells after stimulation with either ionomycin or fMLP. A likely candidate mediating the GM-CSF effect is protein kinase C. Although staurosporine, an inhibitor of protein kinase C, totally abolished the priming effect of GM-CSF, an inhibitory effect of staurosporine on other protein kinases at the concentration used can not be excluded because of the lack of absolute specifity for protein kinase C. GM-CSF produced only a small increase in basal diacylglycerol levels but markedly enhanced diacylglycerol in response to fMLP [42] which could also be related to stimulation of phospholipase D [43]. However, Corey and Rosoff found no increase in basal or fMLP-induced levels of diacylglycerol or inositol 1,4,5trisphosphate in response to GM-CSF [44]; thus they conclude that protein kinase C is not involved in the action of GMCSF. An increase in leukotriene biosynthesis was also found by pretreatment of neutrophil granulocytes with PMA and stimulation with either ionomycin or fMLP. These data are consistent with the hypothesis that activation of protein kinase
500B
C
PMN GM-CSF PMN
GM-CSF Fig.7. Priming effect of PMA and GM-CSF on leukotriene biosynthesis. l o 7 polymorphonuclear leukocytes/ml in buffer A, 1 mM CaCI,, were pretreated for 30 min at 30 ' C with either 1 nM PMA or 2 nM GM-CSF alone or together and stimulated with (A) 0.1 pM fMLP or (B) 0.05 pM ionomycin; 5 min after stimulation, the 5lipoxygenase (5-Lox) metabolites were analyzed by HPLC. Values are expressed as mean f SEM from three or four independent experiments (C = control).
enhanced leukotriene biosynthesis in GM-CSF-primed cells, we used two inhibitors for tyrosine kinases, genistein and tyrphostin (10 pg/ml) [36, 371. No answer could be obtained from these experiments since leukotriene biosynthesis was totally abolished by the two inhibitors even in the control cells (data not shown). Priming of human granulocytes could also be achieved by pretreatment of the cells with PMA and stimulation with either fMLP or ionomycin (Fig. 7). PMA at 1 nM had no influence on intracellular Ca2+ levels, while higher concentrations changed intracellular free calcium after stimulation. Also with PMA no or only minor amounts of 5-lipoxygenase metabolites could be detected by HPLC analysis. However, neither a synergistic nor an additive increase in 5-lipoxygenase metabolites was found after priming with a combination of PMA and GM-CSF and stimulation with ionomycin or fMLP. It therefore seems that a protein kinase C is involved in the priming effect of GM-CSF.
DISCUSSION Our results confirm recent findings that preincubation of human polymorphonuclear leukocytes with rhGM-CSF results in enhanced leukotriene biosynthesis after stimulation
711 C enhances phospholipase A2 activity in human polymorphonuclear leukocytes [45,46]. Interestingly, no synergistic or additive effect on leukotriene biosynthesis after stimulation with fMLP or ionomycin was found by priming with PMA and GM-CSF at the same time. So we conclude that PMA and GM-CSF act in the same way to lower the Ca2 requirement for the arachidonic acid release by phospholipasc A2 through an activation of protein kinase C. One possible action of GM-CSF may result from the tyrosine-kinasc-dependent phosphorylation of several neutrophil proteins which is in part mediated by a pertussis-toxin-sensitive G-protein [18,47]. With the two tyrosine kinase inhibitors tested, it was not possible to derive any conclusion since leukotriene biosynthesis was also inhibited in unprimed cells. From the amino acid sequence of the GM-CSF receptors it was concluded that they do not contain intrinsic tyrosine kinase activities [48]. Other cytokines, such as G-CSF, MCSF, interleukins 1, 4 and 6 or tumor necrosis factor, do not induce tyrosine phosphorylation [48]. So GM-CSF seems to trigger different postreceptor events compared to other cytokines. Protein biosynthesis was reported to take place in GMCSF-treated cells [49-511. From our results with the protein biosynthesis inhibitors cycloheximide and actinomycin D, we conclude that, for the enhanced arachidonic acid release in GM-CSF-treated human neutrophils, protein biosynthesis seems not to be necessary. Similar results were found by DiPersio et al. [13]. However, recently it was shown that GMCSF-enhanced fMLP-induced PAF production in human polymorphonuclear leukocytes could be inhibited by cycloheximide and actinomycin D [51]. Taken together, the presented data provide good evidence that GM-CSF exerts its priming effect on human neutrophil leukotriene biosynthesis by activation of a phospholipase A2 which requires less calcium than under normal conditions. Since protein biosynthesis is not involved, GM-CSF seems to trigger post-translational modifications or translocation of phospholipase A2 perhaps by a phosphorylation step which is inhibitable by staurosporine. PMA and GM-CSF act in the same way so that protein kinase C is a likely mediator of this signal transduction. Whether GM-CSF exerts its activation of phospholipase A2 by modulation of inhibitory proteins (e.g. phosphorylation of lipocortin) or activating proteins (e.g. phosphorylation of the phospholipase-A,-activating protein) or by activation of a phospholipasc A2 isoenzyme, can not be answered at present. +
The present studies were supported by grants of the Deutsche Forschungsgemeinschaft. The authors thank Miss R. Vahle for expert technical assistance.
REFERENCES 1. Cline, M. J. & Golde, D. W. (1974) Nature 248, 703-704. 2. Munker, R., Gasson, J., Ogawa, M. & Koeffler, H. P. (1986) Nature 323, 79 - 82. 3. Chervenick, P. A . & LoBuglio, A. F. (1972) Science 178, 164166. 4. Stieff, C. A,, Emerson, S. G., Donahue, R. E., Nathan, D. G., Wang, E. A,, Wong, G. G. & Clark, S. C. (1985) Science 239, 1171-1176. 5. Gasson, J. C., Weisbart, R. H., Kaufman, S. E., Clark, S. C., Hewick, R. M., Wong, G. G. & Golde, D. W. (1984) Science 226,1339-1342. 6. Metcalf. D. (1985) Science 229, 16-22. 7. Clark, S. C. & Kamen, R. (1987) Science 236,1229- 1237.
8. Weisbart, R. H., Golde, D. W., Clark, S. C . & Gasson, J. C. (1986) J. Immunol. 137, 3584-3587. 9. Weisbart, R. H., Golde, D. W., Clark, S. C., Wong, G. G. & Gasson, J. C. (1985) Nature 314, 361 -363. 10. Weisbart, R. H., Kwan, L., Golde, D. W. & Gasson, J. C. (1987) Blood 69,18 - 21. 11. McColl, S. R., Beauseigle, D., Gilbert, C. & Naccache, P. H. (1990) J. Immunol. 145,3047-3053. 12. Fleischmann, J., Golde, D. W., Weisbart, R. H. & Gasson, J. C. (1986) Blood 68,708-711. 13. DiPersio, J. F., Billing, P., Williams, R. & Gasson, J . C. (1988) J. Immunol. 140,4315-4322. 14. Dahinden, C. A., Zingg, J., Maly, F. E. & Dc Weck, A. (1988) .I. Exp. Med. 167, 1281 -1295. 15. Wirthmueller, U., De Weck, A. L. & Dahinden, C. A . (1989) J . Immunol. 142,3213-3218. 16. McColl, S. R.,Krump, E., Naccache, P. H. & Borgeat, P. (1989) Agents Actions 27,466 -468. 17. DiPersio, J . F., Naccache, P. H., Borgeat, P., Gasson, J. C., Nguycn, M.-H. & McColl, S. R. (1988) Prostaglandins 36, 673 - 691. 18. McColl, S. R., Krump, E., Naccache, P. H., Poubellc, P. E., Braquet, P., Braquet, M. & Borgedt, P. (1991) J . Immunol. 146. 1204- 1211. 19. DiPcrsio, J., Billing, P., Kaufman, S., Egtesady, P., Williams, R. E. & Gasson, J. C. (1988) J . Biol. Chem. 263, 1834- 1841. 20. Cannistra, S. A., Groshek, P., Garlick, R., Miller, J. &Griffin, J. D. (1990) Proc. Natl Acad. Sci. USA 87,93 -97. 21. Gomez-Cambronero, J., Yamazaki, M., Metwally, F., Molski, T. F. P., Bonak, V. A., Huang, C.-K., Becker, E. L. & Sha'afi, R. I. (1989) Proc. Natl Acad. Sci. U S A 86, 3569-3573. 22. McColl, S. R., Kreis, C., DiPersio, J. F., Borgeat, P. & Naccache, P. H. (1989) Blood 73, 588-591. 23. Samuelsson, B., Dahlen, S. E., Lindgren, J. A., Rouzer, C. A. & Serhan, C. N. (1987) Science 237, 1171-1176. 24. Takcnawa, T., Homma, Y . & Nagai, Y. (1983) J . hnmunal. 130, 2849 - 2855. 25. Godfrey, R. W., Manzi, R. M., Clark, M. A. & Hoffstein, S. T. (1987) J . Cell. B i d . 104, 925-932. 26. Rouzer, C. A., Matsumoto, T. & Samuelsson, B. (1986) Proc. Natl Acad. Sci. USA 83, 857 - 861. 27. Schatz-Munding, M., Hatzelmann, A. & Ullrich, V. (1991) Eur. J . Biochem. 197,487-493. 28. Irvine, R. F. (1982) Biochem. J . 204, 3 - 16. 29. Rouzer, C. A. & Kargman, S. (1988) J . Biol. Chem. 263, 10980 10988. 30. Dixon, R. A. F., Diehl, R. E., Opas, E., Rands, E., Vickers, P., Evans, J. F., Gillard, J. W. & Miller, D. K. (1990) Nature 343. 282 -284. 31. Hatzelmann, A. & Ullrich, V. (1987) Eur. J . Biochem. 169, 175 184. 32. Grynkiewicz, G.. Poenie, M. & Tsien, R. Y. (1985) J . Biol. Chem. 260,3440 - 3450. 33. Merrit, J. E., Jacob, R. & Hallam, T. J. (1989) J . Biol. Chem. 264, 1522- 1527. 34. Bligh, E. G. & Dyer, W. J. (1959) Can. J . Biochem. Physiol. 37, 911 -917. 35. Dewald, B., Thelen, M., Wymann, M. & Baggiolini, M. (1989) Biochem. J. 264, 879 - 884. 36. Akiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Walandke, S. I., Ioh, N., Shibuya, M. & Fukami, J . (1987) J . Biol. Chem. 262,5592 - 5595. 37. Lane, P. J. L., Ledbetter, J. A,, McConnell, F. M., Draves, K., Deans, J., Schieven, G. L. & Clark, E. A. (1991) J . Immunoi. 146, 71 5 - 722. 38. Naccache, P. H., Faucher, N., Borgeat, P., Gasson, J. C. & DiPersio, J. F. (1988) J . Immunol. 140, 3541 -3546. 39. Sullivan, R., Griffin, J. D., Simons, E. R., Schafer, A. I., Meshulam, T., Fredette, J. P., Maas, A. K . , Gadenne. A,-S., Leavitt, J. L. & Melnick, D. A. (1987) J . Immunol. 13Y,34223430. 40. Sullivan, R., Fredette, J. P., Griffin, J. D., Leavitt, J. L., Simons, E. R. & Melnick, D. A. (1989) J . Biol. Chem. 264, 6302 -6309.
712 41. Bauldry, S. A., McCall, C. E., Cousart, S. L. & Bass, D. A. (1991) J. lmmunol. 146, 1277-1285. 42. Tyagi. S. R., Winton, E. F. & Lambeth, J. D. (1989) FEES Lett. 257,188-190. 43. Bourgoin, S., Plante, E., Gaudry, M., Naccache, P. H., Borgeat, P. & Poubelle, P. E. (1990) J . Exp. Med. 172, 767 - 777. 44. Corey, S. J. & Rosoff, P. M. (1989) J . Biol. Chem. 264, 14165 14171. 45. McColl, S. R., Hunt, N. P. & Cleland, L. G. (1986) Biochem. Biophys. Res. Commun. 141, 399-404. 46. Liles, W. C., Meier, K. E. & Henderson, W. R. (1987) J. lmmunol. 138,3396 - 3402.
47. Gomez-Cambronero, J., Huang, C.-K., Bonak, V. A., Wang, E., Casnellie, J. E., Shiraishi, T. & Sha’afi, R . I . (1989) Biochem. Biophys. Res. Commun. 162, 1278-1485. 48. Kanakura, Y., Druker, B., Cannistra, S. A., Furukawa, Y., Torimoto, Y. &Griffin, J. D. (1990) Blood 76, 706-715. 49. McColl, S. R., Paquin, P. & Beaulieu, A. D. (1990) Biochem. Biophys. Res. Commun. 172, 1209- 1216. 50. Edwards, S. W., Holden, C. S., Humphreys, J. M. & Hart, C. A. (1989) FEBS Lett. 256, 62-66. 51. Wirthmueller, U., De Weck, A. L. & Dahinden, C. A. (1990) Biochem. Biophys. Res. Commun. 170, 556- 562.
0FEBS 1992
Priming of human polymorphonuclear leukocytes with granulocyte-macrophage colony-stimulating factor involves protein kinase C rather than enhanced calcium mobilisation Monika SCHATZ-MUNDING and Volker ULLRICH Faculty of Biology, University of Konstanz, Federal Republic of Germany
(Received August 30/0ctober 25, 1991) - EJB 91 1165
Pretreatment of human polymorphonuclear leukocytes with the recombinant human granulocytemacrophage colony-stimulating factor (rhGM-CSF) enhances leukotriene biosynthesis in response to a receptor agonist (e.g. N-formyl-methionyl-leucyl-phenylalanine, fMLP) or a Ca2+-ionophore (e.g. ionomycin). This priming effect could be traced back to an elevated release of arachidonic acid from the phospholipid pools and hence an increased leukotriene biosynthesis by 5-lipoxygenase. Preincubation of polymorphonuclear leukocytes with GM-CSF did not influence the basal intracellular Ca2+ level and does not enhance cytosolic free calcium after stimulation with fMLP or ionomycin. Only a small increase in the second Ca2+ phase after receptor agonist stimulation was found. However, the Ca2+-threshold level necessary for the liberation of arachidonic acid by phospholipase A2 was decreased from 350-400 nM calcium in untreated cells to about 250 nM calcium in primed cells. This allows phospholipase A2 to be activated by a release of calcium from intracellular stores and by ionomycin concentrations which are ineffective in untreated cells. Protein biosynthesis inhibitors like actinomycin D (10 pg/ml) and cycloheximide (50 pg/ml) had no effect on the enhanced leukotriene biosynthesis in primed cells after stimulation with ionomycin. However staurosporine (200 nM), an inhibitor of protein kinase C totally abolished the priming effect of GM-CSF after stimulation with ionomycin. The priming effect of GM-CSF could be mimicked by phorbol myristate acetate (PMA; 1 nM) and no additive or synergistic effect was found on leukotriene biosynthesis by simultanous pretreatment with PMA and GM-CSF and stimulation with either fMLP or ionomycin. These results provide evidence that the enhanced arachidonic acid release in GM-CSF-primed polymorphonuclear leukocytes after stimulation with either fMLP or ionomycin involves activation of protein kinase C which, by a still unknown mechanism, reduces the Caz+ requirement of phospholipase A l .
neutrophils include inhibition of migration, changes in receptor expression and effects on the cytoskeleton [5, 81. Again, different from the immediately observed effects, GM-CSF is known to 'prime' human neutrophils to an enhanced functional state in response to physiological agonists, e.g. N-formyl-methionyl-leucyl-phenylalanine (fMLP), platelet-activating factor (PAF), complement factor C5a or a Ca2'ionophore, e.g. ionomycin. Preincubation of the cells with GM-CSF will lead to an enhanced generation of superoxide [9 - 111, phagocytosis [12], liberation of arachidonic acid [I 31, PAF and leukotriene biosynthesis [14- 181. GM-CSF exerts its priming effect through a single class of Correspondence fa V. Ullrich, Fakultat fur Biologie, Universitat Konstanz, Postfach 5560, W-7750 Konstanz, Federal Republic of high-affinity receptors with an apparent molecular mass of 84 kDa on human cells [19, 201. Human neutrophils, Germany Ahhrrviations. A,Ach, arachidonic acid; fMLP, N-formyl- macrophages and eosinophils express the largest number of methionyl-lcucyl-phenylalanine; rhGM-CSF, recombinant human high-affinity receptors (Kd = 37 pM, 300 - 1000 sites/cell [I 91. granulocyte-macrophage colony-stimulating factor; HO-A,Ach, The subsequent signal transduction involves a guaninehydroxyeicosatetraenoic acid; LTB,, leukotriene Bq;PAF, platelet- nucleotide-binding protein [21, 221 which is pertussis-toxinactivating factor; PMA, phorbol myristate acetate; wOH-LTB,, osensitive. hydroxy-LTB,; wCOOH-LTB,, o-carboxy-LTB,. We have investigated the priming effect of GM-CSF on Enzymes. Arachidonate 5-lipoxygenase (EC 1.13.11.34); phosleukotriene biosynthesis in human PMN since leukotrienes pholipase Az (EC 3.1.1.4); phospholipase C (EC 3.1.4.3); phospholipase D (EC 3.1.4.4); protein kinase C (EC 2.7.1.37). play an important role in inflammation [23] and the GM-CSF
The human cytokine granulocyte-macrophage colonystimulating factor (GM-CSF) is a 22-kDa glycoprotein produced by activated T-lymphocytes, endothelial cells, fibroblasts, monocytes and macrophages [l- 31. It stimulates the proliferation and differentiation of erythroid and myeloid stem cells [4-71. GM-CSF is not found in the circulation at detectable levels, suggesting a paracrine action. Apart from its differentiating activity on the precursor cells, it is also capable of acting directly on mature effector cells like eosinophils, neutrophils and monocytes. Such immediate effects on human
706 effect has roused considerable interest [13, 14, 161. Although it is well documented that GM-CSF exerts its priming effect by enhancing the liberation of arachidonic acid after stimulation [4], the underlying mechanism is not known. Leukotriene biosynthesis involves the liberation of arachidonic acid by phospholipases, especially phospholipase A2 124,251 and the oxygenation by 5-lipoxygenase [26]. Both steps are Ca2+-dependent processes [27]. Free intracellular arachidonic acid levels are very low [28] and controlled not only by the liberation but also by the reacylation of this fatty acid. The regulation of 5-lipoxygenase is complex. Besides calcium, it needs ATP for maximal activity and undergoes translocation to a membrane site [29]. The recently identified '5-lipoxygenase activating protein' in membrane seems to play an essential role in activation of 5-lipoxygenase in whole cells [301. In this report we show that GM-CSF primes human polymorphonuclear leukocytes such that subsequent activation of phospholipase A2 by either ionomycin or fMLP requires a smaller rise in intracellular calcium compared to unprimed cells. The effect of GM-CSF appears to be mediated by protein kinase C.
HPLC were stored in 70 pl methanol at -20°C and mixed with 45 pl injection buffer immediately before analysis.
Measurements of intracellular Ca2 levels +
The Fura-2 method was performed as described [31]; lo7 polymorphonuclear leukocytes/ml buffer A containing 2 mM glucose and 1 mM CaC12 were loaded with 2 pM Fura-2 for 30 min. Calibration of intracellular calcium was performed by the method of Grynkiewicz et al. [32]. Maximal fluorescence (Fmax) was achieved by completely lysing the cells with 100 pM digitonin and Fmin by complexing calcium with 10 mM EGTA. For GM-CSF-priming, Fura-2-loaded cells were incubated for 30 rnin with 2 nM GM-CSF at only 25°C to avoid further intracellular distributions of the fluorescence dye. The influx of extracellular calcium was measured by using the manganese quenching method [33]. lo7 polymorphonuclear leukocytes/ml 10 mM Hepes, 150 mM NaCl, 5 mM KCI pH 7.4, 280 mOsm (buffer B) containing 2 mM glucose and 0.5 mM CaCl, were loaded with 2 pM Fura-2 for 30 rnin as described [31]. After addition of 0.2 mM MnCl,, the fluorescence emission above 450 nm after excitation at 364.5 nm was recorded.
MATERIALS AND METHODS Chemicals fMLP, dextran ( M , 506 000), digitonin, cycloheximide, actinomycin D, phorbol myristate acetate (PMA) and EGTA were purchased from Sigma Chemical Co. (Deisenhofen, FRG). Ardchidonic acid was obtained from Larodan (Malmo, Sweden) and [l-'4C]arachidonic acid from New England Nuclear (Dreieich, FRG). PAF and prostaglandin B2 were purchased from Paesel (Frankfurt, FRG), Fura-Z/AM (pentaacetoxymethylester) from Boehringer Mannheim (Mannheim, FRG) and Ficoll-paque from Pharmacia GmbH (Freiburg, FRG). Ionomycin and staurosporine were purchased from Calbiochem (Frankfurt, FRG). Genistein and tyrphostin were products from Gibco (Berlin, FRG). rhGM-CSF was a kind gift from Dr Seiler and Dr Krumwieh (Behringwerke Marburg, FRG). All other chemicals and solvents utilized were of HPLC or analytical grade.
Polymorphonuclear leukocytes and incubations The preparation of human polymorphonuclear leukocytes was performed by dextran sedimentation and subsequent centrifugation on Ficoll-paque as described [31].The viability, measured by trypan-blue exclusion, was greater than 96% and purity of the neutrophil granulocytes was about 95%. After isolation the granulocytes were stored at 4°C in buffer A (137 mM NaCl, 2.7 mM KCl, 8.1 mM Na2HP04, 1.5 mM KH2P04, pH 7.4,280 mOsm). lo7 polymorphonuclear leukocytes in buffer A containing 1 mM CaClz were incubated either with or without rhGMCSF (2 nM) by permanent stirring conditions for 30 min at 30°C. Then the cells were prewarmed to 37°C for 5 min and stimulated for further 5 rnin at 37 "C.
Reverse-phase HPLC analysis Extraction and HPLC analysis of total arachidonate metabolites (oCOOH- and oOH-LTB,, 5-HO-d4Ach, LTB4 and its isomers) were performed as described [31]. Probes for
Measurement of phospholipase A2 activity For these experiments lo7polymorphonuclear leukocytes/ ml in buffer A, 1 mM CaCl,, were incubated with 2 pM [l14C]arachidonic acid (55 Cilmol) for 15 min at 35 -C, then washed twice with prewarmed buffer A, 1 mM CaCl,, and kept at room temperature until use. lo7 labeled cells/ml were preincubated for 5 rnin at 3 7 T , then stimulated with different concentrations of ionomycin. After 5 min, the suspension was centrifuged for 10 s at 10000 x g and radioactivity in the supernatant was measured in a liquid scintillation counter (PN 4700, Isomess, Strdubenhardt, FRG).
Reacylation of free arachidonic acid After incubation with 10 pM arachidonic acid (1 pM [l14C]arachidonic acid), 1 ml of the cell suspension was transferred to 2.4 ml chloroform/methanol (1 : 1, by vol.) and lipids were extracted at pH 3 according to Bligh and Dyer [34]. The chloroform phase was evaporated under nitrogen and dissolved in 70 p1 chloroform/methanol (9 : 1, by vol.). The radioactivity in 20 pl was measured while 50 pl was analyzed by thin-layer chromatography in a solvent system of petrol ether/diethyl ether/acetic acid (90:25: 1, by vol.) [31].
Statistical analysis Results are expressed as mean f SEM of the numbers of experiments indicated.
RESULTS Stimulation of human polymorphonuclear leukocytes with the receptor agonist fMLP results in the generation of only low amounts of HPLC-detectable 5-lipoxygenase metabolites after an incubation time of 5 min. Preincubation with different concentrations of GM-CSF for 30 rnin and subsequent stimulation with fMLP increases the synthesis of leukotrienes by
707
human granulocytes about fivefold (Fig. 1).For further experiments 2 n M GM-CSF were used for an optimal priming effect and a preincubation time of 30 rnin also proved to give maximal yields (data not shown). GM-CSF alone does not result in any HPLC-detectable amounts of 5-lipoxygenase metabolites. The concentration- and time-dependent priming effect of GM-CSF on leukotriene biosynthesis in polymorphonuclear leukocytes was also found by using the Ca2+ionophore ionomycin (0.05 pM) instead of fMLP (data not shown). No significant difference was found by using 25 "C instead of 30'C for priming with GM-CSF (data not shown). Our special interest referred to the effect of GM-CSF on intracellular Ca2+ levels, since we recently showed that the liberation of arachidonic acid by phospholipase A2 in human neutrophils after stimulation by the Ca2'4onophore ionomycin needs a permanently elevated intracellular CaZ level above a threshold concentration of about 350-400 nM calcium [27]. This Ca2+ level was exceeded with ionomycin concentrations above 0.05 pM. The same experiments were performed with GM-CSF-primed polymorphonuclear leuko-
cytes. Fura-2-loaded cells were incubated for 30 rnin with 2 nM GM-CSF and, after stimulation with different concentrations of ionomycin, the intracellular Ca2+ levels were monitored. After 5 rnin of stimulation, the amount of 5lipoxygenase metabolites were measured by HPLC. Fig. 2 shows that the threshold Ca2+ level necessary for the generation of 5-lipoxygenase metabolites was decreased in GMCSF-treated cells to about 250 nM intracellular calcium (depending on the blood donor) compared to 350-400 nM calcium in untreated cells [27. This Ca2+-threshold level was exceeded with even 0.01 pM ionomycin (Figs 5 and 6). Priming with GM-CSF had no effect on either the basal CaZ' levels nor the enhanced Ca2+ levels after stimulation with different concentrations of ionomycin (data not shown).
+
[Caf'l (nM)
= !J J J
LOO
1
300
-
200
-
[lonomycin] [ 1 5 - L o x Met (ng)
0.05
265 180
0.02
101
0 01
11
0.002
/-J
loo1 50
9
m
L
0
0
1 0.01
0.1 [GM-CSF]
L 10
nd
H
lmin
lonomycin
(nM)
Fig. 1. Dose/response relationship of the priming effect of GM-CSF on fMLP-induced leukotriene biosynthesis. Polymorphonuclear leukocytes (lo7 cells/ml in buffer A, 1 mM CaCI2) were preincubated for 30 rnin at 3O'C with diffcrent concentrations of GM-CSF as indicated and stimulated with 0.1 pM fMLP; 5 min after stimulation, aliquots of the incubations were analyzed for 5-lipoxygenase (5-Lox) metabolites by HPLC. The results are expressed as mean f SEM from four different experiments.
Fig. 2. Determination of the Ca2+-threshold level for the generation of 5-lipoxygenase metabolites in GM-CSF-primed cells after stimulation with ionomycin. Fura-2-loaded polymorphonuckdr leukocytes ( lo7 cells/ml in buffer A containing 2 mM glucose and 1 mM CaCI2) were pretreated with 2 nM GM-CSF at 25°C for 30 min. The cells were incubated for 3 min at 37°C and then stimulated with different concentrations of ionomycin. After 5 rnin, 1-ml aliquots were transferred to ethyl acetate and analyzed for 5-lipoxygenase metabolites (C5Lox Met.: 5-OH-d4Ach, wCOOH-LTB4, wOH-LTB4, LTB4 and its isomers). The figure shows one typical experiment out of four.
n
n -GM~CSF .."..
. i
t
]
t
fMLP
c c r A
H
lmin
GM-CSF p....._...._/ +
t fMLP
..._ ....._
-GM-CSF
Fig. 3. Effect of GM-CSF on intracellular Ca" levels. Fura-2-loaded polymorphonuclear leukocytes (lo7cells/ml in buffer A containing 2 mM glucose and 0.5 mM CaC1,) were preincubated with or without GM-CSF for 30 rnin at 25°C. Then cells were prewarmed to 37°C for 3 rnin and stimulated with 0.1 pM fMLP. To complex the extracellular calcium, 1 mM EGTA was added 2 rnin before fMLP stimulation. The monitored Ca2+ curves show one typical experiment out of four.
708
GM - CSF '\\ 8.
'+
GM- CSF
'+ G M - C S F
Fig. 4. CaZ+ influx in GM-CSF primed-cells with (A) fMLP and (B) ionomycin measured by MnZ+influx. Fura-2-loaded polymorphonuclear leukocytes (lo' cells/ml in buffer B containing 2 mM glucose and 0.5 mM CaC12) were preincubated with or without 2 nM GM-CSF for 30 min at 25°C. Then 0.2 mM MnClz was added and the cells were prewarmed to 37°C for 3 min and stimulated with (A) 0.1 pM fMLP or (B) 0.05 pM ionomycin. The monitored fluorescence curves show one typical experiment out of three.
By using the receptor agonist fMLP we also found no significant change in the amount of the intracellular Ca2+ signal in GM-CSF-primed neutrophils, in contrast to unprimed cells after stimulation with 0.1 pM fMLP. Fig. 3 shows the intracellular Ca2+ transient in the presence and absence of extracellular calcium. In the presence of extracellular calcium. a small increase in the second phase of the Ca2+ signal was always observed either due to influx of extracellular calcium or inhibition of the Ca2+-ATPase. When extracellular calcium was complexed by a molar excess of EGTA, no increase in the intracellular CaZ transient (calcium released from the intracellular Ca2+ stores) in GM-CSF-primed cells compared to unprimed cells was detectable. Only a somewhat slower decline to the basal Ca2+level was always observed in GM-CSF-treated cells. Measurements of the Ca2+ influx by the method of manganese quenching of the Fura-2 signal shows the Ca2 influx mediated by the receptor agonist fMLP or ionomycin. We also found no significant difference in this Ca2+influx after stimulation with either fMLP or ionomycin in GM-CSF-primed cells in comparison to unprimed cells (Fig. 4). These results suggest that one effect of GM-CSF is to slightly reduce Ca2' efflux by the Ca2+-ATPase. In unprimed human polymorphonuclear leukocytes the extracellular calcium was absolutely necessary for the generation of 5-lipoxygenase metabolites after stimulation with either fMLP or ionomycin [27]. As Table 1 shows, extracellular calcium had no influence on leukotriene biosynthesis in GM-CSF-primed cells after stimulation with the receptor agonist fMLP. Most striking, and a qualitative difference to untreated cells, was that after stimulation with ionomycin significant amounts of 5-lipoxygenase metabolites were found in the absence of extracellular calcium in GM-CSF-primed cells. The release of arachidonic acid was measured with [l'"Clarachidonic-acid-labeled neutrophils as described in Materials and Methods. Fig. 5 shows the appearance of free radioactivity (arachidonic acid and its metabolites) in the supernatant of GM-CSF-primed and unprimed granulocytes. The dose/response curve with ionomycin started at much lower concentrations of ionomycin (0.01 pM) than in unprimed cells (0.05 pM ionomycin) although the correspond-
Table 1. Influence of extracellular calcium on leukotriene biosynthesis in control and GM-CSF-primed cells. Polymorphonuckdr lcukocytes (PMN; 107/ml in buffer A, 0.5 mM CaCI,) were pretreated with or without 2 nM GM-CSF and stimulated with either 0.1 WMfMLP or 2 pM ionomycin. To complex the extracellular calcium, 1 mM EGTA was added 2 min before stimulation; 5 min after stimulation, incubations were stopped and analyzed by HPLC for 5-lipoxygenase (5Lox) metabolites. Results are expressed as mean SEM from thrcc different experiments.
Agonist
CaZ
+
GM-CSF
5-Lox mctabolitcs
+
+
ng/107 P M N fMLP
++ -
Ionomycin
+ + -
-
+ + -
+ +
6+ I_+ 33f34+
2 1 5 8
299 k 49 18+ 1 349 59 144k 7
*
ing intracellular Ca2+ levels were not changed (Fig. 2). GMCSF alone does not lead to any liberation of arachidonic acid. The same results were obtained when the 5-lipoxygenase metabolites were measured by HPLC after stimulation with different concentrations of ionomycin in GM-CSF-primed and untreated cells. In granulocytes preincubated for 30 min with 2 nM GM-CSF 5-lipoxygenase metabolites were found with wen 0.01 pM ionomycin, whereas in cells preincuhated without GM-CSF at least 0.05 pM ionomycin was necessary (Fig. 6). These results substantiate an increase in the maximum response as well as a shift to the left in the dose, response curve in GM-CSF-primed cells by ionomycin-induced arachidonic acid liberation and leukotriene biosynthesis. In human polymorphonuclear leukocytes the intracellular concentration of free arachidonic acid is also controlled by the reacylation of the fatty acid into the membranes. Therefore one possible effect of GM-CSF could have been an inhibition
709 Table 2. Reacylation of free arachidonic acid. Polymorphonuclear leukocytes (PMN; 107/ml in buffer A, 1 mM CaCI,) were preincubated with or without 2 nM GM-CSF and stimulated for 5 min with either 30 BM arachidonate. (d4Ach; 1 pM [l-'4C]arachidonate, 51 850 cpm) alone or together with 0.1 pM fMLP or 2 pM ionomycin. After lipid extraction, the incubations were analyzed by thin-layer chromatography. Values arc expressed as mean SEM from three different experiments.
t GM-CSF
Agonist 0
-
-
/?
0.01
1
0.1 [ Ionomycin]
(FM)
Fig. 5. Priming effect of GM-CSF on arachidonic acid liberation. [I 14C]Arachidonic-acid-labelledpolymorphonuclear leukocytes (lo7/ ml in buffer A, 1 mM CaCl,) were preincubated with or without 2 nM GM-CSF and stimulatcd with different concentrations of ionomycin; 5 min after stimulation, aliquots of the supernatant were analyzed in a liquid scintillation countcr for radioactivity. The values are given as mcan f SEM (cpm/107PMN) from three different experiments.
FMLP
+ A4Ach
GM-CSF
Free arachidonic acid
-
cpm/107PMN 11348 f 1809
+
13619 & 2161
Ionomycin + A,Ach
-
6275f 510 5893+ 381
A4Ach
-
17588 f 4857 18974 f 3714
+ +
Table 3. Effect of staurosporine on GM-CSF-enhanced leukotriene biosynthesis after stimulation with ionomycin. Polymorphonuclear leukocytes (PMN; 107/ml in buffer A, 1 mM CaC12) were preincubated with or without 200 nM staurosporine at 20°C for 5 min, thcn incubated with or without 2 nM GM-CSF and stimulated with 0.05 pM ionomycin; 5 rnin after stimulation,an aliquot was analyzed by HPLC for total amount of 5-lipoxygenase metabolites. Results are expressed as mean 5 SEM from three different experiments. Staurosporine
5-Lipoxygenase metabolites - GM-CSF
+ GM-CSF
ng/107 PMN 0 X
? In
0.01
0.1
1
[Ionomycin] (PM) Fig. 6. Priming effect of GM-CSF on leukotriene biosynthesis. lo7 polymorphonuclear leukocytes/ml in buffer A, 1 mM CaCl,, were preincubated with or without 2 nM GM-CSF and stimulated by different concentrations of ionomycin; 5 min after stimulation, aliquots were transferred into ethyl acetate and analyzed for 5-lipoxygenase (5-Lox) metabolites (5-OH-A4Ach, toCOOH-LTB4, wOH-LTB4, LTR, and its isomcrs) by HPLC. Values are given as mean & SEM from three different experiments.
of the reacylation so that the levels of free arachidonic acid are elevated in GM-CSF-primed cells. To our knowledge this possibility has never been investigated. To test the hypothesis that GM-CSF increases the intracellular concentration of arachidonic acid by inhibition of the reacylation reaction, we measured the incorporation of exogenous [l-'4C]arachidonate into the phospholipids after priming with GM-CSF compared to untreated cells (Table 2). No significant difference in unesterified levels of arachidonic acid 5 rnin after stimulation with either 2 pM ionomycin or 0.1 pM fh4LP plus 10 pM arachidonate or 10 pM arachidonate alone was found. However the total amount of free arachidonic acid differed between the two stimuli due to activation of the reacylation by ionomycin. An explanation for the prolonged time period necessary for the priming effect of GM-CSF could be that new protein are synthesized. However, GM-CSF-enhanced leukotriene
-
+
124 f 57 109 f 45
202 f 56 107 f 43
biosynthesis after stimulation with ionomycin or fMLP was not affected by preincubation with either actinomycin D, an inhibitor of transcription, or cycloheximide, a n inhibitor of translation. Polymorphonuclear leukocytes were incubated for 5 min with either actinomycin D (10 pg/ml) or cycloheximide (50 pg/ml) and then GM-CSF was added for a further 30 min. The cells were stimulated with 0.05 pM ionomycin or 0.1 pM fMLP and 5-lipoxygenase metabolites were measured by HPLC. Actinomycin D o r cycloheximide alone had no direct effect nor did they inhibit leukotriene biosynthesis in GM-CSF-treated cells (data not shown). Also a preincubation time of 30 rnin with either actinomycin D or cycloheximide did not suppress the priming effect of GM-CSF on ionomycin-induced leukotriene biosynthesis. Addition of staurosporine, known as an inhibitor of protein kinase C, 5 rnin before priming the neutrophils with GMCSF totally abolished the increased leukotriene biosynthesis in ionomycin-stimulated cells (Table 3). In untreated neutrophils, staurosporine had no effect on leukotriene biosynthesis after stirnulation with ionomycin. The concentration necessary for complete inhibition of the GM-CSF effect was, however, much higher than reported in the literature for the inhibition of protein-kinase-C-mediated events [35],so that other phosphorylation reactions had also to be taken into account. To test whether tyrosine phosphorylation is involved in the
710
C
PMA GM-CSF PMA
GM-CSF
K
.?I 0
[14- IS]. GM-CSF by itself does not directly exert this effect but primes the cells for subsequent enhanced responses to a second stimulating agent like the receptor agonist fMLP or the Ca'+-ionophore ionomycin. Our results clearly demonstrate and support previous reports [13] that preincubation with GM-CSF enhances arachidonic acid liberation in polymorphonuclear leukocytes after stimulation with ionomycin. This increased substrate availability was often discussed as a result of GM-CSF-mediated increased calcium release after receptor agonist stimulation. Recent publications show an increase in the intracellular Ca" transient in GM-CSF-preincubated neutrophils after stimulation with the chemotactic factors fMLP, PAF and LTB4 [lS, 381. In contrast, Sullivan et al. found no Ca2+ increase in GM-CSF-treated cells after stimulation with fMLP [39]. By our experiments we could exclude significant effects of GM-CSF on the fMLP- or ionomycin-induced release of calcium from intracellular stores or on the increased influx of calcium from the extracellular side. Only a slightly reduced Ca2+ efflux via Ca2+-ATPase was observed in GM-CSFprimed cells after stimulation with fMLP. However, the threshold level necessary for the release of arachidonic acid after stimulation with ionomycin was clearly decreased from 350 - 400 nM intracellular calcium in unprimed cells [27] to about 250 nM intracellular calcium in GM-CSF-primed cells. As an important consequence of the release of intracellular calcium, the cells now became able to activate phospholipase A2 and hence to synthesise leukotrienes. Similar results were reported for activation of the oxidative burst in human neutrophils. In GM-CSF-treated cells elevation of intracellular calcium by a subthreshold concentration of ionomycin is sufficient to activate the oxidative burst [40]. Priming of human polymorphonuclear leukocytes with tumor necrosis factor also enhances phospholipase A2 activity without any effect on intracellular calcium levels [41]. Since ionomycin does not directly act on phospholipase C [24], the liberation of arachidonic acid should be mediated by phospholipase A2 which seems to be the target of GM-CSF action. In human cells the level of free intracellular arachidonic acid is very low [28] and controlled by hydrolysis and reacylation. The increased level of free arachidonic acid in GM-CSFprimed polymorphonuclear leukocytes after stimulation with ionomycin was not due to an inhibition of reacylation of free arachidonic acid into the lipid membrane. No difference in concentrations of exogenously submitted arachidonic acid was found in GM-CSF-treated cells compared to control cells after stimulation with either ionomycin or fMLP. A likely candidate mediating the GM-CSF effect is protein kinase C. Although staurosporine, an inhibitor of protein kinase C, totally abolished the priming effect of GM-CSF, an inhibitory effect of staurosporine on other protein kinases at the concentration used can not be excluded because of the lack of absolute specifity for protein kinase C. GM-CSF produced only a small increase in basal diacylglycerol levels but markedly enhanced diacylglycerol in response to fMLP [42] which could also be related to stimulation of phospholipase D [43]. However, Corey and Rosoff found no increase in basal or fMLP-induced levels of diacylglycerol or inositol 1,4,5trisphosphate in response to GM-CSF [44]; thus they conclude that protein kinase C is not involved in the action of GMCSF. An increase in leukotriene biosynthesis was also found by pretreatment of neutrophil granulocytes with PMA and stimulation with either ionomycin or fMLP. These data are consistent with the hypothesis that activation of protein kinase
500B
C
PMN GM-CSF PMN
GM-CSF Fig.7. Priming effect of PMA and GM-CSF on leukotriene biosynthesis. l o 7 polymorphonuclear leukocytes/ml in buffer A, 1 mM CaCI,, were pretreated for 30 min at 30 ' C with either 1 nM PMA or 2 nM GM-CSF alone or together and stimulated with (A) 0.1 pM fMLP or (B) 0.05 pM ionomycin; 5 min after stimulation, the 5lipoxygenase (5-Lox) metabolites were analyzed by HPLC. Values are expressed as mean f SEM from three or four independent experiments (C = control).
enhanced leukotriene biosynthesis in GM-CSF-primed cells, we used two inhibitors for tyrosine kinases, genistein and tyrphostin (10 pg/ml) [36, 371. No answer could be obtained from these experiments since leukotriene biosynthesis was totally abolished by the two inhibitors even in the control cells (data not shown). Priming of human granulocytes could also be achieved by pretreatment of the cells with PMA and stimulation with either fMLP or ionomycin (Fig. 7). PMA at 1 nM had no influence on intracellular Ca2+ levels, while higher concentrations changed intracellular free calcium after stimulation. Also with PMA no or only minor amounts of 5-lipoxygenase metabolites could be detected by HPLC analysis. However, neither a synergistic nor an additive increase in 5-lipoxygenase metabolites was found after priming with a combination of PMA and GM-CSF and stimulation with ionomycin or fMLP. It therefore seems that a protein kinase C is involved in the priming effect of GM-CSF.
DISCUSSION Our results confirm recent findings that preincubation of human polymorphonuclear leukocytes with rhGM-CSF results in enhanced leukotriene biosynthesis after stimulation
711 C enhances phospholipase A2 activity in human polymorphonuclear leukocytes [45,46]. Interestingly, no synergistic or additive effect on leukotriene biosynthesis after stimulation with fMLP or ionomycin was found by priming with PMA and GM-CSF at the same time. So we conclude that PMA and GM-CSF act in the same way to lower the Ca2 requirement for the arachidonic acid release by phospholipasc A2 through an activation of protein kinase C. One possible action of GM-CSF may result from the tyrosine-kinasc-dependent phosphorylation of several neutrophil proteins which is in part mediated by a pertussis-toxin-sensitive G-protein [18,47]. With the two tyrosine kinase inhibitors tested, it was not possible to derive any conclusion since leukotriene biosynthesis was also inhibited in unprimed cells. From the amino acid sequence of the GM-CSF receptors it was concluded that they do not contain intrinsic tyrosine kinase activities [48]. Other cytokines, such as G-CSF, MCSF, interleukins 1, 4 and 6 or tumor necrosis factor, do not induce tyrosine phosphorylation [48]. So GM-CSF seems to trigger different postreceptor events compared to other cytokines. Protein biosynthesis was reported to take place in GMCSF-treated cells [49-511. From our results with the protein biosynthesis inhibitors cycloheximide and actinomycin D, we conclude that, for the enhanced arachidonic acid release in GM-CSF-treated human neutrophils, protein biosynthesis seems not to be necessary. Similar results were found by DiPersio et al. [13]. However, recently it was shown that GMCSF-enhanced fMLP-induced PAF production in human polymorphonuclear leukocytes could be inhibited by cycloheximide and actinomycin D [51]. Taken together, the presented data provide good evidence that GM-CSF exerts its priming effect on human neutrophil leukotriene biosynthesis by activation of a phospholipase A2 which requires less calcium than under normal conditions. Since protein biosynthesis is not involved, GM-CSF seems to trigger post-translational modifications or translocation of phospholipase A2 perhaps by a phosphorylation step which is inhibitable by staurosporine. PMA and GM-CSF act in the same way so that protein kinase C is a likely mediator of this signal transduction. Whether GM-CSF exerts its activation of phospholipase A2 by modulation of inhibitory proteins (e.g. phosphorylation of lipocortin) or activating proteins (e.g. phosphorylation of the phospholipase-A,-activating protein) or by activation of a phospholipasc A2 isoenzyme, can not be answered at present. +
The present studies were supported by grants of the Deutsche Forschungsgemeinschaft. The authors thank Miss R. Vahle for expert technical assistance.
REFERENCES 1. Cline, M. J. & Golde, D. W. (1974) Nature 248, 703-704. 2. Munker, R., Gasson, J., Ogawa, M. & Koeffler, H. P. (1986) Nature 323, 79 - 82. 3. Chervenick, P. A . & LoBuglio, A. F. (1972) Science 178, 164166. 4. Stieff, C. A,, Emerson, S. G., Donahue, R. E., Nathan, D. G., Wang, E. A,, Wong, G. G. & Clark, S. C. (1985) Science 239, 1171-1176. 5. Gasson, J. C., Weisbart, R. H., Kaufman, S. E., Clark, S. C., Hewick, R. M., Wong, G. G. & Golde, D. W. (1984) Science 226,1339-1342. 6. Metcalf. D. (1985) Science 229, 16-22. 7. Clark, S. C. & Kamen, R. (1987) Science 236,1229- 1237.
8. Weisbart, R. H., Golde, D. W., Clark, S. C . & Gasson, J. C. (1986) J. Immunol. 137, 3584-3587. 9. Weisbart, R. H., Golde, D. W., Clark, S. C., Wong, G. G. & Gasson, J. C. (1985) Nature 314, 361 -363. 10. Weisbart, R. H., Kwan, L., Golde, D. W. & Gasson, J. C. (1987) Blood 69,18 - 21. 11. McColl, S. R., Beauseigle, D., Gilbert, C. & Naccache, P. H. (1990) J. Immunol. 145,3047-3053. 12. Fleischmann, J., Golde, D. W., Weisbart, R. H. & Gasson, J. C. (1986) Blood 68,708-711. 13. DiPersio, J. F., Billing, P., Williams, R. & Gasson, J . C. (1988) J. Immunol. 140,4315-4322. 14. Dahinden, C. A., Zingg, J., Maly, F. E. & Dc Weck, A. (1988) .I. Exp. Med. 167, 1281 -1295. 15. Wirthmueller, U., De Weck, A. L. & Dahinden, C. A . (1989) J . Immunol. 142,3213-3218. 16. McColl, S. R.,Krump, E., Naccache, P. H. & Borgeat, P. (1989) Agents Actions 27,466 -468. 17. DiPersio, J . F., Naccache, P. H., Borgeat, P., Gasson, J. C., Nguycn, M.-H. & McColl, S. R. (1988) Prostaglandins 36, 673 - 691. 18. McColl, S. R., Krump, E., Naccache, P. H., Poubellc, P. E., Braquet, P., Braquet, M. & Borgedt, P. (1991) J . Immunol. 146. 1204- 1211. 19. DiPcrsio, J., Billing, P., Kaufman, S., Egtesady, P., Williams, R. E. & Gasson, J. C. (1988) J . Biol. Chem. 263, 1834- 1841. 20. Cannistra, S. A., Groshek, P., Garlick, R., Miller, J. &Griffin, J. D. (1990) Proc. Natl Acad. Sci. USA 87,93 -97. 21. Gomez-Cambronero, J., Yamazaki, M., Metwally, F., Molski, T. F. P., Bonak, V. A., Huang, C.-K., Becker, E. L. & Sha'afi, R. I. (1989) Proc. Natl Acad. Sci. U S A 86, 3569-3573. 22. McColl, S. R., Kreis, C., DiPersio, J. F., Borgeat, P. & Naccache, P. H. (1989) Blood 73, 588-591. 23. Samuelsson, B., Dahlen, S. E., Lindgren, J. A., Rouzer, C. A. & Serhan, C. N. (1987) Science 237, 1171-1176. 24. Takcnawa, T., Homma, Y . & Nagai, Y. (1983) J . hnmunal. 130, 2849 - 2855. 25. Godfrey, R. W., Manzi, R. M., Clark, M. A. & Hoffstein, S. T. (1987) J . Cell. B i d . 104, 925-932. 26. Rouzer, C. A., Matsumoto, T. & Samuelsson, B. (1986) Proc. Natl Acad. Sci. USA 83, 857 - 861. 27. Schatz-Munding, M., Hatzelmann, A. & Ullrich, V. (1991) Eur. J . Biochem. 197,487-493. 28. Irvine, R. F. (1982) Biochem. J . 204, 3 - 16. 29. Rouzer, C. A. & Kargman, S. (1988) J . Biol. Chem. 263, 10980 10988. 30. Dixon, R. A. F., Diehl, R. E., Opas, E., Rands, E., Vickers, P., Evans, J. F., Gillard, J. W. & Miller, D. K. (1990) Nature 343. 282 -284. 31. Hatzelmann, A. & Ullrich, V. (1987) Eur. J . Biochem. 169, 175 184. 32. Grynkiewicz, G.. Poenie, M. & Tsien, R. Y. (1985) J . Biol. Chem. 260,3440 - 3450. 33. Merrit, J. E., Jacob, R. & Hallam, T. J. (1989) J . Biol. Chem. 264, 1522- 1527. 34. Bligh, E. G. & Dyer, W. J. (1959) Can. J . Biochem. Physiol. 37, 911 -917. 35. Dewald, B., Thelen, M., Wymann, M. & Baggiolini, M. (1989) Biochem. J. 264, 879 - 884. 36. Akiyama, T., Ishida, J., Nakagawa, S., Ogawara, H., Walandke, S. I., Ioh, N., Shibuya, M. & Fukami, J . (1987) J . Biol. Chem. 262,5592 - 5595. 37. Lane, P. J. L., Ledbetter, J. A,, McConnell, F. M., Draves, K., Deans, J., Schieven, G. L. & Clark, E. A. (1991) J . Immunoi. 146, 71 5 - 722. 38. Naccache, P. H., Faucher, N., Borgeat, P., Gasson, J. C. & DiPersio, J. F. (1988) J . Immunol. 140, 3541 -3546. 39. Sullivan, R., Griffin, J. D., Simons, E. R., Schafer, A. I., Meshulam, T., Fredette, J. P., Maas, A. K . , Gadenne. A,-S., Leavitt, J. L. & Melnick, D. A. (1987) J . Immunol. 13Y,34223430. 40. Sullivan, R., Fredette, J. P., Griffin, J. D., Leavitt, J. L., Simons, E. R. & Melnick, D. A. (1989) J . Biol. Chem. 264, 6302 -6309.
712 41. Bauldry, S. A., McCall, C. E., Cousart, S. L. & Bass, D. A. (1991) J. lmmunol. 146, 1277-1285. 42. Tyagi. S. R., Winton, E. F. & Lambeth, J. D. (1989) FEES Lett. 257,188-190. 43. Bourgoin, S., Plante, E., Gaudry, M., Naccache, P. H., Borgeat, P. & Poubelle, P. E. (1990) J . Exp. Med. 172, 767 - 777. 44. Corey, S. J. & Rosoff, P. M. (1989) J . Biol. Chem. 264, 14165 14171. 45. McColl, S. R., Hunt, N. P. & Cleland, L. G. (1986) Biochem. Biophys. Res. Commun. 141, 399-404. 46. Liles, W. C., Meier, K. E. & Henderson, W. R. (1987) J. lmmunol. 138,3396 - 3402.
47. Gomez-Cambronero, J., Huang, C.-K., Bonak, V. A., Wang, E., Casnellie, J. E., Shiraishi, T. & Sha’afi, R . I . (1989) Biochem. Biophys. Res. Commun. 162, 1278-1485. 48. Kanakura, Y., Druker, B., Cannistra, S. A., Furukawa, Y., Torimoto, Y. &Griffin, J. D. (1990) Blood 76, 706-715. 49. McColl, S. R., Paquin, P. & Beaulieu, A. D. (1990) Biochem. Biophys. Res. Commun. 172, 1209- 1216. 50. Edwards, S. W., Holden, C. S., Humphreys, J. M. & Hart, C. A. (1989) FEBS Lett. 256, 62-66. 51. Wirthmueller, U., De Weck, A. L. & Dahinden, C. A. (1990) Biochem. Biophys. Res. Commun. 170, 556- 562.
