Effect Growth Factors onEscherichia coli Alpha-Hemolysin ... · 2086 KONIG AND KONIG PMN, GM-CSFactivates the respiratory burst and increases arachidonate release (11, 12, 14, 28).

INFECrION AND IMMUNITY, May 1994, p. 2085-2093 Vol. 62. No. 50019-9567/94/$04.00+0Copyright © 1994, American Society for Microbiology

Effect of Growth Factors on Escherichia coli Alpha-Hemolysin-Induced Mediator Release from Human Inflammatory Cells:

Involvement of the Signal Transduction PathwayB. KONIG AND W. KONIG*

Medizinische Mikrobiologie und Immunologie, AG Infektabwehr, Ruhr-Univ'ersitdt Bochluni,D-44780 Bochum, Germany

Received 13 July 1993/Returned for modification 13 August 1993/Accepted 25 February 1994

Previously, we have shown that Escherichia coli alpha-hemolysin represents a potent stimulus for inflam-matory mediator release (02- release, ,-glucuronidase release, and leukotriene generation) from humanpolymorphonuclear granulocytes (PMN) as well as for histamine release from a human lymphocyte-monocyte-basophil cell suspension (LMB). In contrast, the E. coli alpha-hemolysin leads to a downregulation of cytokinerelease (interleukin 6 [IL-6], tumor necrosis factor alpha, and IL-1il) from human LMB. This study wasundertaken (i) to analyze the priming efficacy of growth factors (granulocyte-macrophage colony-stimulatingfactor [GM-CSF] and granulocyte CSF [G-CSF]) on inflammatory mediator release from human PMN andLMB challenged with hemolysin-producing E. coli bacteria as well as with cell-free E. coli alpha-hemolysin and(ii) to identify major components involved in GM-CSF and G-CSF priming. GM-CSF pretreatment led to anincreased chemiluminescence response from human PMN by up to 100%, leukotriene B4 generation wasenhanced up to fivefold, and histamine release from human LMB increased from 45% ± 15% to 75% ± 5%(mean ± standard distribution) of the total histamine content. G-CSF priming induced an increase in thechemiluminescence response by up to 50% ± 5% from human PMN and an increase in histamine release fromhuman LMB by 20% ± 5%. The growth factors, GM-CSF and G-CSF, modulated neither 0-glucuronidaserelease from human PMN nor IL-8 release from human PMN and LMB challenged with the E. colialpha-hemolysin. GM-CSF and G-CSF pretreatment increased the fluoride (NaF)-induced chemiluminescenceresponse by up to 10-fold; the serine/threonine phosphatase inhibitor okadaic acid inhibited GM-CSF- andG-CSF-induced priming. NaF-induced histamine release was enhanced up to 60 and 30% by GM-CSF andG-CSF priming, respectively. GM-CSF and G-CSF pretreatment did not modulate phorbol 12-myristate13-acetate-induced chemiluminescence response or histamine release. GM-CSF by itself induced an increasein 5-lipoxygenase-specific mRNA expression within 5 min. Our results indicate that (i) GM-CSF and G-CSFinteract with inflammatory cells via distinct cellular signalling, (ii) the signal transduction pathway isdependent on the cellular mediator, and (iii) the use of growth factors may be a potent tool to influence theclinical outcome in infectious diseases.

Hemolysin-producing (Hly+) Escherichia coli strains areisolated from patients with extraintestinal infections such asurinary tract infections, bacteremia, and septicemia (15). Thepathogenic relevance of the alpha-hemolysin has been provenin several animal as well as in vitro models (3, 21, 23). Afterinteraction with soluble hemolysin or with hemolysin-produc-ing E. coli bacteria, human neutrophils produce reactiveoxygen species, generate the chemotactically active leukotrieneB4 (LTB4) and release cytoplasmic and lysosomal enzymes,e.g., P-glucuronidase and lysozyme; human basophils releasehistamine (23). In contrast, E. coli alpha-hemolysin inhibitscytokine (interleukin 6 [IL-6], tumor necrosis-factor alpha[TNF-cx], and IL-11) release from human monocytes (22). Thecellular responses to E. coli alpha-hemolysin are mediated by acomplex signal transduction cascade which is dependent on thecell type (21). In this regard, substantial evidence for the factthat protein kinase C and guanine nucleotide-binding proteins(G proteins) play an essential role in the cellular activationprocess leading to reactive oxygen metabolites and mediatorsof inflammation such as leukotrienes from human neutrophilswas obtained (21, 24).

* Corresponding author. Mailing address: Med. Mikrobiologie undImmunologie, AG Infektabwehr, Ruhr-Universitiit Bochum, Univer-sitatsstr. 150, D-44780 Bochum, Germany.

IL-8, a recently described 6- to 10-kDa protein known for itsneutrophil chemotactic activity, is another potent mediator ofhost response released during injury and infection (1). IL-8 isproduced by a variety of cells in vitro including peripheralblood leukocytes and represents a cytokine that induces che-motaxis, degranulation, respiratory burst, adherence, shapechange, Ca2+ mobilization, and upregulation of CD1 1b/CD 18glycoprotein in human polymorphonuclear granulocytes (PMN)(2, 10, 29, 33).

In the past, several cytokines such as interferons, IL-2, andcolony-stimulating factors (CSFs) were used in the treatmentof immunological disorders and cancer. Granulocyte-macro-phage CSF (GM-CSF) and granulocyte CSF (G-CSF) belongto the CSFs, which are primarily concerned with hematopoie-sis. GM-CSF is a growth factor involved in the proliferationand maturation of progenitor cells of the hematopoietic sys-tem, e.g., eosinophils, monocytes/macrophages, and PMN.G-CSF preferentially stimulates the development of PMNfrom the appropriate precursor cell populations (28). In addi-tion to their role in hematopoiesis, GM-CSF and G-CSF playan important role in host defense by enhancing the functionalactivities of mature leukocytes, in particular for PMN. In thisregard, GM-CSF and G-CSF promote increased phagocytosis,oxygen metabolism to a number of stimuli, e.g., calciumionophore A23187, formyl-Met-Leu-Phe (FMLP), and C5a. In

2085

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

2086 KONIG AND KONIG

PMN, GM-CSF activates the respiratory burst and increasesarachidonate release (11, 12, 14, 28).Few data exist as to the effect of GM-CSF on the inflam-

rmatory process induced by microorganisms (19). Production ofgrowth factors (GM-CSF and G-CSF) is switched on or is moreextensively expressed after treatment of the cells with bacterialproducts such as lipopolysaccharide or after damage to thecells with chemical or physical agents. Quite recently, evidencethat GM-CSF and G-CSF may prime various effector cells fora subsequent agonist has been obtained (11).

It was the purpose of our study to analyze the interactions ofGM-CSF and G-CSF with inflammatory cells (PMN, ba-sophils, and monocytes) with regard to the E. coli alpha-hemolysin-induced response pattern (chemiluminescence,LTB4, ,B-glucuronidase, lysozyme, histamine, and IL-8). Toelucidate the mode of action of GM-CSF and G-CSF, definedstimuli (NaF and phorbol 12-myristate 13-acetate [PMA]) as

well as inhibitors (okadaic acid, lavendustin, and genistein) ofdistinct elements of the signal transduction cascade were

introduced into our studies.

MATERIALS AND METHODS

Materials. Brain-heart infusion medium was obtained fromOxoid, Basingstoke, Hampshire, England. Synthetic leuko-trienes were a generous gift from Merck-Frosst, Pointe Claire,Quebec, Canada. The IL-8 antibody was kindly provided by M.Ceska, Sandoz, Vienna, Austria. The solvents used for high-pressure liquid chromatography (HPLC) were obtained fromlocal suppliers. Additional chemicals were obtained fromSigma, Deisenhofen, Federal Republic of Germany.

Bacterial strains. The hemolysin-positive (Hly+) strain E.coli 5K(pANN202-312) and the isogenic hemolysin-negative(Hly- ) strain E. coli 5K were used in our studies. Thehemolysin-negative E. coli 5K (Smr lacYl tonA21 thr-] supE44thi rK MK+) was transformed with the plasmid pANN202-312to the hemolysin-positive E. coli 5K(pANN202-312). Theplasmid pANN202-312 carrying the hemolysin determinant ofpHlyl52 was described previously (15). The cloning was per-formed at the Institut fur Genetik und Mikrobiologie, Univer-sitat Wiirzburg, Wurzburg, Federal Republic of Germany. Forcell stimulation, the washed bacteria (hly- and hly+) as well as

the culture supernatants (hly- and hly+) were used at theindicated concentrations.

Bacterial growth. Brain-heart infusion broth (10 ml) was

inoculated with 100 pul of an overnight culture. Bacterialgrowth proceeded for 3.5 h at 37°C on a shaker (150 rpm).Bacterial cultures were centrifuged at 4,000 x g (HeraeusChrist Minifuge T; Heraeus, Osterode, Federal Republic ofGermany) for 20 min at 4°C. The bacterial cell numbers usedwere adjusted to 109 bacteria per ml of phosphate-bufferedsaline (PBS).Hemolysin assay. The production of hemolysin was tested

on sheep blood agar plates. A quantitative hemolysin assay wasperformed as described previously (23).

Buffer. The buffer used for washing the cells and formediator release consisted of 137 mM NaCl, 8 mM Na2HPO4,3 mM KCl, and 3 mM KH2PO4 (pH 7.4; modified Dulbecco'sPBS).

Preparation of cells. Human PMN and a human lympho-cyte-monocyte-basophil cell (LMB) suspension were separatedfrom heparinized blood (15 U/ml) on a Ficoll-metrizoategradient and then subjected to dextran sedimentation (5). TheLMB fraction contained 84.6% ± 4.6% (mean ± standarddistribution [SD]) lymphocytes, 14.2% ± 4.1% monocytes, and

1.2% ± 0.5% basophilic granulocytes. The PMN and LMBwere adjusted to a concentration of 2 x 107/ml.

Cell viability. Cell viability was studied by trypan blueexclusion as well as by analysis of lactate dehydrogenase(LDH) release from stimulated and nonstimulated cells. Anal-ysis of LDH (EC 1.1.1.27) was carried out as describedpreviously (23).

Chemiluminescence. Chemiluminescence was measured at37°C in a Lumacounter M2080 (Lumac, The Netherlands).Samples for chemiluminescence were obtained by adding a

PMN suspension (50 pl, 106 cells) to polypropylene tubescontaining PBS (300 ,ul) and luminol (20 [l, 0.25 mM) and theappropriate stimulus.Measurement of histamine release. LMB (107/00 [lI of

PBS) were stimulated in the presence of 0.6 mM Ca2+ and 1mM Mg2+. The histamine content of the supernatants was

measured fluorophotometrically with a Technicon Autoana-lyzer (Bad Vilbel, Federal Republic of Germany). Histaminedihydrochloride dissolved in 2% HCl04 served as a control.For the determination of the total cellular histamine content,the cells were disrupted by the addition of 2 ml of HCl04 (2%).

Analysis of leukotriene release from human neutrophils.PMN (2 x 107/ml) were stimulated in the presence of 0.6 mMCa2+ and 1 mM Mg2 . Cell supernatants were analyzed forleukotrienes as described elsewhere (20, 23). Methanol-aceto-nitrile (2 ml, 50:50 [vol/vol]) was added to the culture super-natants. After centrifugation at 1,900 x g for 15 min (Cryofuge6-4; Heraeus Christ), the supernatants were evaporated todryness by lyophilization (Modulyo; Edwards-Kniese, Mar-burg, Federal Republic of Germany). The residues were

dissolved in 600 pLI of methanol-water (30:70), and 200 ,ul wasanalyzed by reversed-phase HPLC. The column (4.6 by 200mm) was packed with Nucleosil C18 (particle size, 5 Rm;Macherey-Nagel, Duren, Federal Republic of Germany).HPLC equipment consisted of a CM4000 pump, a SM4000detector (both Laboratory Data Control-Milton Roy, Hassel-roth, Federal Republic of Germany), and an automatic sampleinjector (WISP 710B, Waters, Eschborn, Federal Republic ofGermany). Leukotrienes were analyzed with a mobile phaseconsisting of methanol-water-acetonitrile-phosphoric acid (48:24:28:0.03 [vol/vol]) including 0.04% EDTA and 0.15%K2HPO4, pH 5.0. The flow rate was maintained at 0.9 ml/min.The A280 of the column effluent was determined. Quantifica-tion and identification of leukotrienes were performed withsynthetic standard solutions. LTB4 generation was calculatedas the combined amounts of LTB4 and the LTB4 omega-oxidation products (20-hydroxy-LTB4 and 20-carboxy-LTB4).

IL-8 assay. IL-8 release was determined with a sandwichELISA performed according to Bazzoni et al. (2). Each well ofa 96-well plate (Nunc Maxisorb) was precoated overnight at40C with 100 pI of PBS-Tween 20 (0.1%) containing anti-IL-8antibodies at a concentration of 5 pKg/ml. The plates were

washed three times with PBS-Tween, the appropriate samplesor IL-8 standard was added, and incubation proceeded for 2 hat 37°C. Thereafter, alkaline phosphatase-linked anti-IL-8antibody was added. After addition ofp-nitrophenylphosphate(15 mg/ml) for quantification, an ELISA reader and, forcalculation, Mikrotek software (SLT Labinstruments, Crail-sheim, Federal Republic of Germany) were used.

Detection of 5-LO-specific mRNA. Total cellular RNA was

prepared by a single-step isolation method with guanidinium-thiocyanate-phenol-chloroform (9). Total RNA was reverse

transcribed with poly(dT)12_18. Amplification was performedwith primers specific for 5-lipoxygenase (5-LO). Aliquots ofeach resulting reaction mixture were analyzed by 2% agarosegel electrophoresis. The sense primer was 5'atcaggacgttcacg

INFECT. IMMUN.

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

EFFECT OF GROWTH FACTORS ON MEDIATOR RELEASE 2087

gccgagg3', and the antisense primer was 5'ccaggaacagctcgttttcctg3'.

Statistical analysis. If not stated otherwise, all data aremeans ± SDs of at least three individual experiments withcells from different donors. Significance was examined byStudent's t test for independent means.

RESULTS

Effect of GM-CSF and G-CSF on the E. coli alpha-hemoly-sin-induced chemiluminescence response from human PMN.Previously, we have shown that E. coli alpha-hemolysin inducesa characteristic and dose-dependent chemiluminescence re-sponse from human PMN (23). To analyze the effects ofGM-CSF and G-CSF on the chemiluminescence response,human PMN (10") were preincubated with buffer, GM-CSF, orG-CSF in various amounts (5, 10, and 50 ng) for 30 min at37°C; the subsequent incubation proceeded in the presence ofhemolysin-positive E. coli bacteria [E. coli 5K(pANN202-312)]at an optimal concentration for chemiluminescence response(hemolytic activity, 50 to 70%; 107 to 108 bacteria) (23) for upto 20 min. GM-CSF pretreatment resulted in an increasedbasal chemiluminescence response (buffer) in a dose-depen-dent manner (from 2,000 cpm up to 3,500 cpm with 10 ng ofGM-CSF); G-CSF treatment did not modulate the basalchemiluminescence response from human PMN (data notshown). As is apparent from Fig. 1, in which the results of atypical experiment are shown (mean values for one experimentperformed in triplicate), pretreatment (30 min) with 5 ng ofGM-CSF led to an increase in the E. coli alpha-hemolysin-induced chemiluminescence response of up to twofold. GM-CSF amounts greater than 5 ng did not further increase thechemiluminescence response from human PMN. A preincuba-tion time of 10 to 60 min was similarly effective with regard tothe priming efficacy (data not shown). In the presence ofG-CSF, the results are clearly less pronounced; G-CSF pre-treatment enhanced the chemiluminescence response maxi-mally by up to 30% ± 10% (10 ng of G-CSF) (data notshown); a preincubation time of 10 to 30 min was mosteffective with regard to the priming efficacy (data not shown).

Effect of GM-CSF and G-CSF on enzyme release (,I-gluc-uronidase, lysozyme, and LDH) from human neutrophilsinduced by the E. coli alpha-hemolysin. In the next series ofexperiments, we analyzed the effects of GM-CSF and G-CSFon enzyme release (3-glucuronidase, lysozyme, and LDH)from human PMN challenged with hemolysin-producing E.coli bacteria. In the first set of experiments, human PMN werepreincubated in the absence or presence of GM-CSF or G-CSF(0, 1, 10, or 50 ng) for 30 min at 37°C; the incubationproceeded for a further 30 min at 37°C in the presence of anoptimal concentration of the Hly+ bacteria, which lead to3-glucuronidase and lysozyme release but not to LDH release

(hemolytic activity, 55% ± 15%). As is apparent from Table 1,for an incubation time of 30 min the E. coli alpha-hemolysin-induced 3-glucuronidase release increased to 24% ±+ 6% ofthe total cellular enzyme activities. Similar results were ob-tained when human PMN were challenged with hemolysin-positive E. coli bacteria for a longer time (up to 90 min) orwere expressing a hemolytic activity between 10 and 100%.Independently of the GM-CSF and G-CSF concentrations, therelease of P-glucuronidase was only slightly enhanced. Withregard to lysozyme release, GM-CSF and G-CSF induced amaximal lysozyme release of 11% ± 2% after an incubationtime of 90 min; no LDH release was observed. As is apparentfrom Table 1, with a GM-CSF or G-CSF pretreatment of 30min and an incubation time of 30 min with Hly+ E. coli

70 -

601

50 -

40 -

E3-

oI

200

10

0

5 10 15 20

Time [min]FIG. 1. Effects of GM-CSF on the hemolysin-induced chemilumi-

nescence response from human PMN. Neutrophils (2 x 10") were

preincubated in the absence (buffer control) or presence of GM-CSFat the indicated concentrations (nanograms per 107 cells) for 30 min at37°C, and then incubation proceeded in the presence of Hly+ E. coli5K(pANN202-312) (hemolytic activity, 55%l ± 15`1c; 1(7 to 1(8bacteria) for the indicated times at 37C. Results of one of three typicalexperiments from three different donors are shown.

bacteria [SK(pANN202-312); hly+ activity, 55c ±- 15%] a

decrease in E. coli alpha-hemolysin-induced lysozyme releasewas observed. The inhibitory effects were maximal with 10 ng

of GM-CSF and were reversed with greater amounts (Table 1).G-CSF pretreatment led to a decrease in lysozyme release in a

dose-dependent manner (Table 1). GM-CSF and G-CSF didnot affect LDH release (less than 10% of total LDH content)from human PMN challenged with Hly+ E. coli bacteria (datanot shown). Hly E. coli bacteria of strain 5K did not induce3-glucuronidase, lysozyme, or LDH release from human PMN;

neither GM-CSF nor G-CSF showed any modulatory effects.Effects of GM-CSF and G-CSF on E. coli alpha-hemolysin-

induced LTB4 generation. Human PMN were preincubated inthe presence of PBS (buffer control) or in the presence ofGM-CSF or G-CSF at different concentrations (1, 5, 10, and 50ng) for 30 min at 37°C; incubation then proceeded for a further2, 5, 10, or 20 min at 37°C in the presence of buffer or of theHly+ E. coli bacteria at an optimal concentration for LTB4generation (Hly+, 55% ± 15%) (22). GM-CSF by itselfinduced only low amounts of LTB4 (up to 1.7 ± 0.5 ng/107cells) after a total incubation time of 50 min even when theamount was increased to 50 ng; G-CSF failed to do so. Humanneutrophils were primed for an enhanced LTB4 generationafter challenge with Hly+ E. coli bacteria [SK(pANN202-312)](Fig. 2). However, depending on the donor, both GM-CSF and

GM-CSF [ng]0 - 5

VOL. 62, 1994

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

2088 KONIG AND KONIG

TABLE 1. Effects of GM-CSF and G-CSF on 13-glucuronidase and lysozyme release from human PMN challengedby E. coli alpha-hemolysina

Release of compound (% of total enzyme activities) with indicated amt (ng) of factor"Factor and stimulus 0 1 10 50

,B-Glucoronidase Lysozome ,B-Glucoronidase Lysozome ,-Glucoronidase Lysozome 3-Glucoronidase Lysozome

GM-CSFBuffer only 7.5 ± 3.2 10.0 ± 2.3 9.3 ± 4.2 8.8 ± 3.1 10.7 ± 3.1 7.7 ± 3.1 10.8 ± 4.5 11.0 ± 2.8E. coli 5K(pANN202-312) 24.6 ± 6.5 52.0 ± 4.3 33.1 ± 7.3 41.6 ± 5.8 34.3 ± 6.5 37.5 ± 3.7 34.2 ± 5.8 52.1 ± 5.3E. coli 5K 7.8 ± 0.9 10.3 ± 2.5 8.4 ± 1.2 10.4 ± 2.3 8.0 ± 2.5 10.2 ± 3.0 9.3 ± 2.5 11.8 ± 2.1

G-CSFBuffer only 7.5 ± 3.2 10.0 ± 2.3 4.8 ± 1.5 10.0 ± 2.1 5.9 ± 2.4 10.0 ± 2.3 8.8 ± 4.8 11.0 ± 2.3E. coli SK(pANN202-312) 24.6 ± 6.5 52.0 ± 4.8 32.6 ± 7.0 47.9 ± 2.5 31.3 ± 4.8 43.7 ± 4.0 31.5 ± 5.2 41.6 ± 3.8E. coli 5K 7.8 ± 0.9 10.3 ± 2.5 6.8 ± 2.8 10.4 ± 2 7.3 ± 1.2 9.8 ± 1.0 8.4 ± 1.9 10.4 ± 0.8

Hemolytic activity was 55% ± 15% (see Materials and Methods).b Data are means ± SDs for three independent experiments.

G-CSF primed human PMN for an enhanced LTB4 generationafter challenge with Hly+ E. coli bacteria (Table 2). Nonethe-less, the priming efficacy of G-CSF at 10 ng was less than thatof GM-CSF. A preincubation time of 10 min for GM-CSF andfor G-CSF was necessary to show the priming effects (data notshown). Hly- bacteria of strain E. coli 5K failed to inducesignificant leukotriene generation from human PMN; neitherGM-CSF nor G-CSF showed modulatory effects on leuko-triene generation from PMN treated with E. coli 5K. Theexperiments were performed under noncytotoxic conditions(trypan blue exclusion; LDH release, 5%).

Effects of GM-CSF and G-CSF on histamine release fromhuman basophils. The E. coli alpha-hemolysin is a potentinducer for histamine release from human basophils (23). Tostudy the influence of GM-CSF and G-CSF on E. coli alpha-hemolysin-induced histamine release, an LMB suspension wasused. In a first set of experiments, the optimal preincubationtime and the optimal concentrations for GM-CSF and G-CSFpretreatment were determined. LMB were preincubated in thepresence of GM-CSF (0, 1, 5, 10, or 50 ng) or G-CSF (0, 1, 5,10, or 50 ng) for 0, 30, 45, or 60 min at 37°C. The incubationthen proceeded in the absence (buffer control only) or pres-ence of Hly+ E. coli bacteria expressing hemolysin at anoptimal concentration for histamine release (hemolytic activ-ity, 55% ± 15%). GM-CSF pretreatment primed humanbasophils similarly to human neutrophils over the concentra-tion range tested; maximal priming effects were obtained at aGM-CSF amount of 10 ng (data not shown). The primingeffects were quite similar when a preincubation time between30 and 60 min at 37°C was chosen (data not shown). G-CSFpretreatment did not significantly increase E. coli alpha-hemolysin-induced histamine release (data not shown). Wethen studied the priming effects of GM-CSF (10 ng) whenhuman LMB were stimulated with Hly+ E. coli 5K(pANN202-312) at different hemolysin concentrations. Human LMB werepreincubated in the absence (PBS buffer alone) or in thepresence of GM-CSF (10 ng) for 30 min at 37°C; the incuba-tion proceeded in the absence or presence of Hly+ E. colibacteria (hemolytic activities, 80, 64, and 50%) for a further 60min at 37°C. As is apparent from Table 3, the priming effect ofGM-CSF was observed at all tested hemolysin concentrations.Hly- E. coli bacteria of strain 5K failed to induce histaminerelease from human LMB; neither GM-CSF nor G-CSFshowed modulatory effects on histamine release from LMBtreated with E. coli 5K.

Effect of GM-CSF and G-CSF on IL-8 release from humanPMN and LMB. First, we studied the effects of GM-CSF and

G-CSF by themselves on IL-8 release from human PMN andLMB. As is apparent from Table 4, GM-CSF and G-CSF atconcentrations of 10 ng induced a time-dependent increase inIL-8 release from human PMN. The GM-CSF-induced maxi-mal IL-8 release was observed after an incubation time of 100min; thereafter, IL-8 release declined. A similar activationpattern was obtained for G-CSF. For LMB, identical resultswere obtained (data not shown). E. coli alpha-hemolysin(hemolytic activity, 55% ± 15%) led to an increase (overvalues obtained with the buffer control) in IL-8 release fromhuman PMN after an incubation time of 60 min; thereafter, theIL-8 release declined to baseline levels (Table 4). In humanLMB, Hly+ E. coli bacteria expressing hemolytic activitiesfrom 14.2 to 46% led to a downregulation of IL-8 releasecompared with values obtained with the buffer control (Fig. 3).To study the effects of GM-CSF and G-CSF on hemolysin-

14 GM-CSF [ng]No12 W

6--

2 5 10 20Time [min]

FIG. 2. Effects of GM-CSF on hemolysin-induced LTB4 generationfrom human PMN. Neutrophils (107/500 I.l) were preincubated in theabsence (buffer) or presence of GM-CSF (1, 5, 10, or 50 ng/107 cells)for 30 min at 37°C, and incubation then proceeded in the absence orpresence of Hly+ E. coli 5K(pANN202-312) (hemolytic activities, 55%± 15%) bacteria for a further 2, 5, 10, or 20 min at 37°C. Data aremeans + SDs (bars) for three independent experiments.

INFEcr. IMMUN.

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

EFFECT OF GROWTH FACTORS ON MEDIATOR RELEASE 2089

TABLE 2. Effects of GM-CSF and G-CSF on LTB4 generation by the indicated E. coli strains

LTB4 generation (ng/1l(7 neutrophils) by the indicated E. coli strain after the indicated time (min)Preincubation 10 2)

with:5K(pANN202-312) 5K 5K(pANN202-312) 5K 5K(pANN20I2-312) 5K

Buffer only 18.12 ± 4.3 18.2 ± 5.8 34.58 ± 7.3 19.4 ± 3.1 15.84 ± 3.4 17.7 ± 2.7GM-CSF (10 ng) 121.70 ± 13.1 19.2 ± 4" 109.79 ± 13.8" 17.6 ± 7.9 71.74 ± 8.5" 18.9 ± 4.2G-CSF (10 ng) 41.0 ± 7.3" 18.4 ± 2.3 60.40 ± 10.1" 19.4 ± 5.7 18.80 ± 4.5 16.8 ± 6.4

" Significant increase (P < 0.t)5) in LTB4 generation compared with the buffer control.

induced IL-8 release, human PMN or LMB were preincubatedwith PBS (buffer control), with GM-CSF (10 ng), or withG-CSF (10 ng) for 40 min at 37°C, and then incubationproceeded for up to 90 min in the presence of Hly+ E. colibacteria (hemolytic activities, 55% + 15%) at 37°C. In humanPMN as well as in human LMB, neither GM-CSF nor G-CSFinfluenced E. coli alpha-hemolysin-induced IL-8 release.

Molecular aspects of GM-CSF and G-CSF effects. In subse-quent experiments, we focused on distinct steps of cellularsignalling to explain the above described effects of GM-CSFand G-CSF pretreatment on E. coli alpha-hemolysin-inducedcellular responses (chemiluminescence, LTB4 response, hista-mine release, and IL-8 release). For this purpose, two well-defined stimuli were introduced into our studies, the direct Gprotein activator fluoride (NaF) and PMA, a direct proteinkinase C activator. Furthermore, the serine/threonine phos-phatase inhibitor okadaic acid and the tyrosine kinase inhibi-tors genistein and lavendustin were used.

(i) Chemiluminescence. To elucidate the key elements in-volved in GM-CSF and G-CSF effects on the chemilumines-cence response from human PMN, we first used the direct Gprotein activator fluoride (NaF) and the direct protein kinaseC activator PMA as stimuli. The experiments were performedas described above. As is apparent from Fig. 4, GM-CSF (10ng) and G-CSF pretreatment (10 ng) enhanced the NaF (15mM)-induced chemiluminescence response up to 10-fold (Fig.4), while the PMA (10-8 M)-induced chemiluminescenceresponse was only slightly modulated (by up to 20%) (data notshown). Preincubation of human PMN with okadaic acid(10-8 M), a serine/threonine phosphatase inhibitor, abolishedthe priming effects of GM-CSF (10 ng) (Fig. 5) and G-CSF (10ng) (data not shown) on the NaF (15 mM)-induced (Fig. 5) andPMA (10' M)-induced as well as the E. coli alpha-hemolysin(107 to 108 bacteria)-induced chemiluminescence responses

(data not shown). Tyrosine kinase inhibitors (genistein and

lavendustin) over the tested concentration range (10-" to10 M) failed to influence GM-CSF and G-CSF priming.

(ii) LTB4 generation. 5-LO represents one major elementinvolved in LTB4 generation. Therefore, we analyzed theeffects of GM-CSF and G-CSF (1, 5, and 10 ng) on theexpression of 5-LO-specific mRNA in human neutrophils up toan incubation time of 90 min (5, 15, 30, 60, and 90 min) at37°C. As is apparent from Fig. 6, for a GM-CSF or G-CSFamount of 10 ng an increase in 5-LO-specific mRNA becameapparent after an incubation time of 2 min. UnstimulatedPMN as well as PMN challenged with Hly+ or Hly bacteriashowed no 5-LO-specific mRNA over the tested time period.

(iii) Histamine release. Again, NaF and PMA were intro-duced into our study. The experiments were performed as

described above. As is apparent from Table 3, for a NaFconcentration of 15 mM and GM-CSF or G-CSF at 10 ng, theNaF-induced histamine release increased from 9.5% ± 2.5%to 74% ± 7% (release from total histamine content [GM-CSF]) or maximally up to 42% + 4% (G-CSF). The PMA-induced histamine release (9.5% 2.5%) of total histaminecontent was not modulated by the growth factors. Whenhuman LMB were preincubated with tyrosine kinase inhibitors(genistein and lavendustin) or with okadaic acid prior toGM-CSF or G-CSF priming and subsequent stimulation by E.coli alpha-hemolysin, NaF, or PMA, the priming efficacies ofboth growth factors were not affected (data not shown).

DISCUSSION

Our results clearly show that GM-CSF and G-CSF pretreat-ment influence intracellular signalling upon stimulation with E.coli alpha-hemolysin. The distinct effects of GM-CSF andG-CSF are dependent on the types of inflammatory mediators.

In the past decade, the role of cytokines as immunoregula-tory molecules has been well established (6, 7, 12, 22). Several

TABLE 3. Effects of GM-CSF and G-CSF on histamine release from human LMB challenged with Hly+ and Hly E. colibacteria, NaF, and PMA

Stimulus Hemolytic activity Histamine release (% of total content) after preincubation with":

('') Buffer only GM-CSF G-CSF

Buffer only 11.8 + 2.5 12.70 ± 3.6 9.5 ± 2.5

E. coli SK(pANN202-312) 80 64.1 ± 5.8" 97.7 ± 13.5" 68.4 ± 16.864 53.4 ± 4.7" 81.5 ± 12.7" 65.5 ± 12.35(t 35.9 ± 3.9" 67.4 ± 4.8 36.4 ± 6.8

E. coli 5K 0 10.8 ± 2.5 9.4 ± 4.0 9.8 ± 4.8NaF (15 mM) 0 9.5 2.5 74 7 42 4PMA (10"-- M) 0 9.5 ± 2.5 9.4 ± 4.0 1(1.1 ± 4

"GM-CSF and G-CSF at (l ng. Data are means ± SDs."Significant increase (P < 11.115) in histamine release as compared with the buffer control.

VOL. 62, 1994

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

2090 KONIG AND KONIG

TABLE 4. Time- and dose-dependent IL-8 release from humanPMN by GM-CSF, G-CSF, and Hly+ E. coli 5K(pANN202-312)

Time IL-8 release (pg/107 cells)"(min) Buffer only GM-CSF G-CSF E. coli*

1 99± 9 114 ± 5 96± 3 NDC5 122 ± 13 108 ± 7 92 ± 6 ND

15 152 ± 12 130 ± 12 162 ± 14 800 ± 56d30 190 ± 15 346 ± 18 431 ± 35 1,400 + 110"60 120 ± 6 400 ± 212d 600 ± 39" 4,800 ± 700"90 400 ± 18 2,540 ± 53" 740 ± 56d 900 ± 60"120 400 ± 16 1,280 ± 18d 1,120 ± 65d ND

" Data are means ± SDs for three independent experiments. The amount ofGM-CSF and of G-CSF was 10 ng.

"Hemolytic activity, 55% ± 15%.'ND, not determined."Significant increase (P < 0.05) in IL-8 release compared with the buffer

control.

cytokines, e.g., interferons, IL-2, and CSFs, were used in thetreatment of immunological disorders and cancer. GM-CSF issecreted by a large variety of cells. The principal sources aremacrophages, endothelial cells, and activated T cells. GM-CSFwas originally described as a factor that stimulated the growthof colonies from immature bone marrow progenitor cells (28).However, it became apparent that, in addition to its activity asa mitogen for granulocyte precursors, GM-CSF has numerouseffects on mature PMN (14). To date, enhanced functionalactivities on mature neutrophils, eosinophils, basophils, andmononuclear phagocytes have been described (14, 28). In thisregard, it has been shown that GM-CSF primes neutrophils forenhanced oxidative metabolism upon stimulation with FMLP,C5a, and LTB4 (25). Our results support and extend thesefindings. GM-CSF pretreatment and, to a lesser degree, G-CSF pretreatment increased the E. coli alpha-hemolysin-in-duced chemiluminescence response from human PMN. The E.coli alpha-hemolysin presents a physiological stimulus releasedfrom E. coli bacteria during the logarithmic growth phase.

Recent evidence has shown that cytokines can stimulate theproduction of 5-LO products, e.g., LTB4 (7). LTB4 is a majormediator of leukocyte activation in acute inflammatory reac-tions that produces chemotaxis, lysosomal enzyme release, andcell aggregation. Other investigators have shown that GM-CSFprimes PMN to produce LTB4 when treated with the Caionophore A23187, with platelet activating factor, or with thechemotactic peptide FMLP (11, 27). LTB4 induces polymor-phonuclear leukocyte migration in vivo and is a potent activa-tor of human leukocytes (24). It thus represents a potentmediator of inflammation. We showed that GM-CSF dramat-ically increased E. coli alpha-hemolysin-induced generation ofLTB4 from human neutrophils, which are the cells of the firstline of defense against invading microorganisms. Previously,we have shown that E. coli alpha-hemolysin is also a potentstimulus for the release of histamine, which is responsible forvasodilation (23). GM-CSF pretreatment also enhanced E. colialpha-hemolysin-induced histamine release. These results in-dicate that GM-CSF pretreatment primes effector cells for anexaggerated inflammatory event. G-CSF pretreatment primedinflammatory cells for cellular responses to a lesser degreethan did GM-CSF.

In contrast to inflammatory mediators such as 02- metab-olites, leukotrienes, and histamine, the release of cytokines(IL-6, TNF-oa, and IL-1,) from human LMB was downregu-lated in the presence of the E. coli alpha-hemolysin (22). Inaddition to LTB4, IL-8 evokes a chemotactic response from

A0,25 -

0,2 -

r-m

> 0,1 5 -

0

0)v-C 0,1 -

CD

incubation time 30 min

0,05

buffer 46 44 42 40 27 14

hemolytic activities of

E. coli5K(pANN202-312)

B1,4 -

1,2 -

-Jm 1

-00)r_ 0,6 -

_ 0,4 -

0,2 -

incubation time 90 min

buffer 46 44 42 40 27 14

hemolytic activities of

E. coliK(pANN202-312)

FIG. 3. IL-8 release from human LMB challenged with Hly+ E. coliSK(pANN202-312) bacteria. Human LMB (106/ml) were incubatedwithout (buffer control) or with Hly+ SK(pANN202-312) bacteria for30 min (A) or 90 min (B) at 37°C. The cell supernatants were analyzedfor IL-8 by ELISA. The data are means + SDs for three independentexperiments. An asterisk indicates a significant difference from thePBS control (P < 0.05).

PMN and therefore may be important in recruiting PMN intosites of inflammation (1). IL-8 has also been implicated in alarge number of inflammatory diseases in humans includingpsoriasis, rheumatoid arthritis, and gram-negative sepsis (32).Concomitantly with other investigators, we showed that humanPMN and LMB constitutively produced IL-8 mRNA andprotein (29). Our data show that up to 12 ng/ml of IL-8 proteinis released into the supernatants by human peripheral bloodPMN (2 x 107 cells/ml) stimulated with GM-CSF (5 ng/ml) orG-CSF. This event is donor dependent and may be due toindividual cell reactivity. In contrast to LTB4 generation fromhuman LMB, E. coli alpha-hemolysin, down to a hemolyticactivity of 10% led to a decrease in basal IL-8 level and adecrease in IL-8-specific mRNA. When human PMN werechosen as target cells, E. coli alpha-hemolysin induced a slightincrease in IL-8 release in a time-dependent manner; a down-regulation was not observed up to an incubation time of 120min. Evidence that IL-8 interacts with TNF and GM-CSF aswell as with G-CSF has been obtained (1, 33). In contrast tothe above-described effects of GM-CSF and G-CSF on inflam-

INFECT. IMMUN.

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

EFFECT OF GROWTH FACTORS ON MEDIATOR RELEASE

- PBS/GM-CSF

PBS/G-CSF

-x- PBS/PBS/

E 400-

0

'- 300-

200

100

0 10 20 30 40 50

Time [min]FIG. 4. Effects of GM-CSF and G-CSF on NaF-induced chemilu-

minescence responses from human PMN. Neutrophils (2 x 106) werepreincubated in the absence (buffer control) or presence of GM-CSF(10 ng) or G-CSF for 30 min at 37°C; thereafter, incubation proceededin the presence of buffer or NaF (15 mM) for the indicated times at37°C. Data are means + SDs for three independent experiments.

matory mediator release (02- generation, LTB4 generation,and histamine release) from human PMN and LMB in re-

sponse to E. coli alpha-hemolysin, the effects of E. colialpha-hemolysin on IL-8 release were not influenced either byGM-CSF or by G-CSF pretreatment, regardless of the cell typeused (PMN or LMB). Thus, we conclude that depending onthe mediator, distinct signal transduction pathways which aredifferently influenced by GM-CSF or G-CSF are involved.Up to now, the specific mechanisms by which GM-CSF and

G-CSF activate and prime various neutrophil functions includ-ing the generation of LTB4, the release of IL-8 and ofsuperoxide anions, and the degranulation of neutrophils orbasophils are not clear. The findings of Bourgoin et al. excludethe involvement of the Ptd1ns(4,5)P2-specific phospholipaseC-diacylglycerol pathway in neutrophil priming by GM-CSF(4). Gomez-Cambronero et al. suggested a role of guaninenucleotide-regulatory proteins (G proteins) in GM-CSF prim-ing (17). We observed that GM-CSF pretreatment (with resultssimilar to those of McColl et al.) and to a lesser degree G-CSFpretreatment enhanced chemiluminescence response inducedvia direct G-protein activation through NaF by sixfold (GM-CSF) or fourfold (G-CSF), while PMA-induced chemilumines-cence response via direct activation of protein kinase C wasonly slightly enhanced (25). Similarly, the NaF-induced but notthe PMA-induced histamine release was enhanced by GM-CSF and G-CSF (by 60% ± 5% and 29% ± 6%, respectively).

600

500

okadaic acid/GM-CSF

okadaic acid/PBS

- PBS/GM-CSF

-x- PBS/PBS

E 400-

0

r-300 -

200

100

0 10 20 30 40 50

Time [min]FIG. 5. Influence of okadaic acid on the GM-CSF priming efficacy

for the NaF-induced chemiluminescence response from human PMN.Neutrophils (2 x 106) were preincubated in the absence (buffercontrol) or presence of okadaic acid (10-8 M) for 10 min at 37°C, andthen incubation proceeded in the presence of buffer or GM-CSF (10ng) for 30 min at 37°C; thereafter, PMN were stimulated with NdF (15mM) for a further 40 min at 37°C. Data are means + SDs for threeindependent experiments.

These results indicate that signal transduction elements up-stream of protein kinase C, e.g., G proteins, are involved inchemiluminescence and histamine priming by GM-CSF andG-CSF. Thus, our new data support our previous findings thatG proteins are involved in E. coli alpha-hemolysin-inducedcellular signalling (21).

Receptors for the hematopoietic growth factors erythropoi-etin, IL-3, and GM-CSF are members of a structurally relatedreceptor superfamily. Interestingly, while none of these recep-tors encodes tyrosine kinase activities, tyrosine phosphoryla-tion has been observed in various responsive cells stimulatedwith each factor, indicating the involvement of tyrosine kinasesin the priming effects of GM-CSF and G-CSF (13, 16, 18, 26).Indeed, the tyrosine kinase inhibitor lavendustin inhibitedGM-CSF-induced IL-8 release and inhibited GM-CSF-primedIL-8 release induced by FMLP. The FMLP-induced histaminerelease from GM-CSF-primed basophils was not influenced bylavendustin (data not presented). In contrast, tyrosine kinaseinhibitors, e.g., lavendustin and genistein, failed to modulatethe priming effects of GM-CSF or of G-CSF in our describedtest systems. Moreover, as described in Results, the threonine/serine phosphatase inhibitor okadaic acid diminished thepriming efficacy of GM-CSF and G-CSF. Pretreatment ofhuman PMN with okadaic acid abolished GM-CSF and G-CSFpriming with regard to the NaF- and E. coli alpha-hemolysin-

VOL. 62, 1994 2091

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

2092 KONIG AND KONIG

1 2 3 4 5 6 7 8 9 10 11 12

FIG. 6. 5-LO-specific mRNA expression by GM-CSF and G-CSF.PMN (2 x 107/ml) were incubated in the presence of GM-CSF (10 ng)(lanes 1 to 6) or G-CSF (10 ng) (lanes 7 to 12) for 2, 5, 10, 30, 60, or90 min at 37°C (lanes 1 and 7, 2 and 8, 3 and 9, 4 and 10, 5 and 11, and6 and 12, respectively). Total RNA was reverse transcribed witholigo(dT)12-18, and the product was amplified by PCR (35 cycles) asdescribed in Materials and Methods. The amplified product was of theappropriate size (375 bp). A 100-bp ladder (Gibco BRL) was used asa molecular size marker (right lane). The figure represents a typicalresult out of four experiments.

induced chemiluminescence response. The priming effects ofGM-CSF and G-CSF on E. coli alpha-hemolysin-inducedhistamine release were not affected by okadaic acid orgenistein (data not presented). The possible involvement of athreonine/serine kinase, e.g., RAF-1, in GM-CSF priming hasalso been suggested by other investigators (8). Thus, GM-CSFmay act via different signal-transducing elements depending on

the inflammatory mediator and/or the cell type.For LTB4 generation, 5-LO and the 5-LO-activating protein

play central roles (24, 30). Consistent with results obtained byother investigators, GM-CSF alone did not stimulate therelease of detectable amounts of LTB4 or its omega products(12, 27). McColl et al. excluded a direct effect of GM-CSF on5-LO activity (27). Our results have shown that GM-CSF andG-CSF in a concentration range from 1 to 50 ng/107 cellsupregulate 5-LO-specific mRNA expression in human neutro-phils within 2 to 5 min, thus offering a higher availability for5-LO. Recently, a concentration-dependent effect of GM-CSFon 5-LO-activating protein de novo synthesis was reported(31). Nonetheless, maximal effects on mRNA level wereobserved after 4 h; the total 5-LO-activating protein concen-tration increased only minimally. Furthermore, neutrophilsexpress the receptor for LTB4, which is downregulated as aconsequence of GM-CSF treatment (11). Thus, some of theGM-CSF priming events, e.g., the enhanced LTB4 generationfrom human neutrophils, may be mediated by direct autocrinestimulation of neutrophils by LTB4.Whether pretreatment predominantly with GM-CSF, lead-

ing to an exaggerated response pattern from inflammatorycells, will enhance the host defense against microorganismsand their products or will cause inappropriate inflammationand tissue damage has to be further elucidated. Thus, further

studies are directed to analysis of the signal transductionpathway involved in the GM-CSF and G-CSF priming effectsand its role in microbial pathogenicity.

ACKNOWLEDGMENT

W.K. was supported in this work by a grant from the DeutscheForschungsgemeinschaft (Ko 427/7-5).

REFERENCES1. Baggiolini, M., A. Walz, and S. L. Kunkel. 1989. Neutrophil-

activating peptide-1/interleukin 8, a novel cytokine that activatesneutrophils. J. Clin. Invest. 84:1045-1049.

2. Bazzoni, F., M. A. Cassatella, F. Rossi, M. Ceska, B. Dewald, andM. Baggiolini. 1991. Phagocytosing neutrophils produce and re-lease high amounts of the neutrophil-activating peptide 1/interleu-kin 8. J. Exp. Med. 173:771-774.

3. Bhakdi, S., M. Muhly, S. Korom, and G. Schmidt. 1990. Effects ofEscherichia coli hemolysin on human monocytes-cytocidal actionand stimulation of interleukin-1 release. J. Clin. Invest. 85:1746-1753.

4. Bourgoin, S., E. Plante, M. Gaudry, P. H. Naccache, P. Borgeat,and P. E. Poubelle. 1990. Involvement of a phospholipase D in themechanism of action of granulocyte-macrophage colony-stimulat-ing factor (GM-CSF): priming of human neutrophils in vitro withGM-CSF is associated with accumulation of phosphatidic acid anddiacylglycerol. J. Exp. Med. 172:767-777.

5. Boyum, A. 1976. Isolation of lymphocytes, granulocytes and mac-rophages. Scand. J. Immunol. 5:9-15.

6. Brom, J., and W. Konig. 1992. Signal transduction and priming ofhuman neutrophils. Int. Arch. Allergy Immunol. 99:387-389.

7. Brom, J., and W. Konig. 1992. Cytokine-induced (interleukin-3, -6,and -8 and tumour necrosis factor-beta) activation and deactiva-tion of human neutrophils. Immunology 75:281-285.

8. Carroll, M. P., L. I. Clark, U. R. Rapp, and W. S. May. 1990.Interleukin 3 and granulocyte-macrophage colony-stimulating fac-tor mediate rapid phosphorylation and activation of cytosolic c-raf.J. Biol. Chem. 265:19812-19817.

9. Chomezynski, P., and N. Sacchi. 1987. Single step method of RNAisolation by acid guanidium thiocyanate-phenol-chloroform-ex-traction. Anal. Biochem. 162:156-159.

10. Colditz, I. G., R. D. Zwahlen, and M. Baggiolini. 1990. Neutrophilaccumulation and plasma leakage induced in vivo by neutrophil-activating peptide-1. J. Leukocyte Biol. 48:129-137.

11. Dahinden, C. A., J. Zingg, and F. E. Maly de Wal. 1988. Leuko-triene production in human neutrophils primed by recombinanthuman granulocyte/macrophage colony-stimulating factor andstimulated with the complement component C5A and FMLP assecond signals. J. Exp. Med. 167:1281-1295.

12. DiPersio, J. F., P. Billing, R. Williams, J. C. Gasson, and J. F.DiPersio. 1988. Human granulocyte-macrophage colony-stimulat-ing factor and other cytokines prime human neutrophils forenhanced arachidonic acid release and leukotriene B4 synthesis. J.Immunol. 140:43315-43318.

13. Evans, J. P., S. A. R. Mire, A. V. Hoffbrand, and R. G. Wickre-masinghe. 1990. Binding of G-CSF, GM-CSF, tumor necrosisfactor-alpha, and gamma-interferon to cell surface receptors onhuman myeloid leukemia cells triggers rapid tyrosine and serinephosphorylation of a 75 kd protein. Blood 75:88-92.

14. Fabian, I., Y. Kletter, S. Mor, C. Geller-Bernstein, M. Ben-Yaakov, B. Volovitz, and D. W. Golde. 1992. Activation of humaneosinophil and neutrophil functions by haematopoietic growthfactors: comparisons of IL-1, IL-3, IL-5 and GM-CSF. Br. J.Haematol. 80:137-143.

15. Goebel, W., and J. Hedgpeth. 1982. Cloning and functional char-acterization of the plasmid-encoded hemolysin determinant ofEscherichia coli. J. Bacteriol. 151:1290-1298.

16. Gomez-Cambronero, J., C. K. Huang, V. A. Bonak, E. Wang, E.Casnellie, T. Shiraishi, and R. I. Sha'afi. 1989. Tyrosine phosphor-ylation in human neutrophils. Biochem. Biophys. Res. Commun.162:1478-1485.

17. Gomez-Cambronero, J., M. Yamazaki, F. Metwally, T. F. P.Melski, V. A. Bonak, C. Huang, E. L. Becker, and R. I. Sha'afi.

INFECT. IMMUN.

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

EFFECIT OF GROWTH FACTORS ON MEDIATOR RELEASE 2093

1989. Granulocyte-macrophage colony-stimulating factor and hu-man neutrophils: role of guanine nucleotide regulatory proteins.Proc. Natl. Acad. Sci. USA 86:3569-3573.

18. Isfort, R. J., and J. N. Ihie. 1990. Multiple hematopoietic growthfactors signal trough tyrosine phosphorylation. Growth Factors2:213-216.

19. Kawakami, M., H. Tsutsumi, T. Kumakawa, H. Abe, M. Hirai, S.Kurosawa, M. Mori, and M. Fukushima. 1990. Levels of serumgranulocyte-stimulating factor in patients with infections. Blood76:1962-1964.

20. Knoller, J., W. Schonfeld, M. Koller, T. Hensler, and W. Konig.1988. Arachidonic acid metabolites from polymorphonuclear leu-kocytes of healthy donors, severely burned patients, and childrenwith cystic fibrosis. Routine monitoring by high-performanceliquid chromatography. J. Chromatogr. 427:199-208.

21. Konig, B., and W. Konig. 1991. Roles of human peripheral bloodleukocyte protein kinase C and G proteins in inflammatorymediator release by isogenic Escherichia coli strains. Infect. Im-mun. 59:3801-3810.

22. Konig, B., and W. Konig. 1993. Induction and suppression ofcytokine release (TNF-a; IL-6; IL-1p) by E. coli pathogenicityfactors (adhesins, a-hemolysin). Immunology 78:526-533.

23. Konig, B., W. Konig, J. Scheffer, J. Hacker, and W. Goebel. 1986.Role of Escherichia coli alpha-hemolysin and bacterial adherencein infection: requirement for release of inflammatory mediatorsfrom granulocytes and mast cells. Infect. Immun. 54:886-892.

24. Konig, W., W. Schonfeld, M. Raulf, M. Koller, J. Knoller, J.Scheffer, and J. Brom. 1990. The neutrophil and leukotrienes-role in health and disease. Eicosanoids 3:1-22.

25. McColl, S. R., D. Beauzeigle, C. Gilbert, and P. H. Naccache. 1990.Priming of the human neutrophil respiratory burst by granulocyte-macrophage colony-stimulating factor and tumor necrosis factor-alpha involves regulation and a post-cell surface receptor level.Enhancement of the effect of agents which directly activateG-proteins. J. Immunol. 145:3047-3053.

26. McColl, S. R., F. DiPersio, A. C. Caon, P. Ho, and P. H. Naccache.1991. Involvement of tyrosine kinases in the activation of humanperipheral blood neutrophils by granulocyte-macrophage colony-stimulating factor. Blood 78:1842-1852.

27. McColl, S. R, E. Krump, P. H. Naccache, P. E. Poubelle, P.Braquet, M. Braquet, and P. Borgeat. 1991. Granulocyte-mac-rophage colony-stimulating factor increases the synthesis of leu-kotriene B4 by human neutrophils in response to platelet-activat-ing factor. Enhancement of both arachidonic acid availability and5-lipoxygenase activation. J. Immunol. 146:1204-1211.

28. Metcalf, D. 1985. The granulocyte-macrophage colony stimulatingfactors. Cell 43:5-6.

29. Mielke, V., J. G. J. Baumann, M. Sticherling, T. Ibs, A. G.Zomershoe, K. Seligmann, H.-H. Henneicke, J.-M. Schroder, W.Sterry, and E. Christophers. 1990. Detection of neutrophil-acti-vating peptide NAP/IL-8 and NAP[IL-8 mRNA in human recom-binant IL-la and human recombinant tumor necrosis factor-a-stimulated human dermal fibroblasts. J. Immunol. 144:153-156.

30. Reid, G. K., S. Kargman, P. J. Vickers, J. A. Mancini, C. Lereille,D. Ethier, D. K. Miller, J. W. Gillard, R A. F. Dixon, and J. F.Evans. 1990. Correlation between expression of 5-lipoxygenase-activating protein, 5-lipoxygenase, and cellular leukotriene synthe-sis. J. Biol. Chem. 265:19818-19823.

31. Van Zee, K. J., L. E. DeForge, and E. Fischer. 1991. IL-8 in septicshock, endotoxemia, and after IL-1 administration. J. Immunol.146:3478-3482.

32. Willems, J., M. Joniau, S. Cinque, and J. van Damme. 1991.Human granulocyte chemotactic peptide (IL-8) as a specificneutrophil degranulator; comparison with other cytokines. Immu-nology 67:540-542.

33. Yoshimura, T. K, S. Matsushima, S. Tanaka, E. A. Robinson, E.Appela, J. Oppenheim, and E. J. Leonard. 1987. Purification of ahuman monocyte-derived neutrophil chemotactic factor thatshares sequence homology with other host defense cytokines.Proc. Natl. Acad. Sci. USA 84:9233-9237.

VOL. 62, 1994

on Novem

ber 28, 2020 by guesthttp://iai.asm

.org/D

ownloaded from