Effects of Zymosan-activated Human Granulocytes on Isolated Human Airways1-3



In atopic asthmatic individualsexposure to allergen produces an early bronchoconstriction that may be followed by a late broncho-obstructive phase (1, 2). This late asthmatic reaction (LAR) is characterized by an influx of inflammatory cells into the airway wall (3). Increases in the number of eosinophils, neutrophils, and other inflammatory. cellshave been described in bronchial lavage fluid obtained 6 to 48 h after allergen challenge (4-6) in conjunction with a transient increase in bronchial responsiveness (7, 8). Thus there is a temporal association between the influx of granulocytes into the airway wall during the LAR and an influx of granulocytes that may be associated with an increase in bronchial responsiveness. Therefore we tested the hypothesis that activated human granulocytes constrict isolated human airways and increase the sensitivity of these airways to a cholinergic stimulus. Furthermore, to assess whether activated granulocytes affect the function of autonomic nerves in the airway wall, we studied the effect of activated granulocytes on the responses of central human bronchus to electrical field stimulation (EFS), which selectivelyactivates intramural postganglionic nerves (9). Methods Patients Human lung tissue was obtained from 18male and 5 female patients with a mean age of 64.9 yr (range 41 to 81yr) who underwent a thoracotomy for lung cancer: 20 were smokers, 14 had chronic obstructive pulmonary disease (COPD) according to the criteria of the American Thoracic Society (10), and none had characteristics of asthma. Preoperative lung function showed mean values for inspiratory vital capacity (YO) of 93.9 ± 4.2OJoofthepredicted value and FEV 1 as a percentage of VCI of72.7 ± }.6OJo ofthe predicted value. Medication during anesthesia was the same for all patients: atropine, thiopentone, fentanyl, O,/N,O, halothane, and pancuronium. Be-

SUMMARY In asthma a temporal association exists between the late allergic reaction (lAR), the Influx of granulocytes Into the airway wall, and an Increase in bronchial responsiveness. We therefore tested the hypothesis that activated human granulocytes constrict isolated human airways and increase their sensitivity to cholinergic stimuli. Bronchial rings ware dissected from 23 lung tissue specimens collected at thoracotomy and studied Isotonically in organ baths. Airways were Incubated with 1, 2, 5, 10, or 20 x 10· granUlocytes from normal or atopic donors. Activation of the cells with serum-treated zymosan (STZ, 0.2 mg/ml), which Itself did not alter baseline airway caliber, resuhed In a bronchoconstrlctlon proportional to the number of zymosan-activated granulocytes (ZAG) present (rs = 0.79,P < 0.001). This contraction was reduced by about 70% with the Ieukotriene C./O. receptor antagonist FPl 55712(11.5 11M; p < 0.001)or with the llpoxygenase Inhibitor nordlhydrogualaretlc acid (10 11M; P < 0.001). The scavengers of activated oxygen molecules superoxlde dlsmutase (300 U/ml) and bovine catalase (5,000 Ulml), the cyclooxygenase Inhibitor Indomethacin (1011M), or the histamine (H,) receptor antagonist mepyramlne (2.8 11M) had no effect. Granulocyte suspensions from atopic donors contained more eoslnophlls (p < 0.001),and the magnitUde of the contraction to 10 x 10· ZAG was related to the proportion of eoslnophlls (rs = 0.66, P < 0.01).The sensitivity of the airways to methacholine was unchanged In the presence of 1, 2, or 5 x 10· ZAG and decreased with 10 or 20 x 10· ZAG (p < 0.05). The cholinergic twitch and the nonadrenergic relaxation to near maximal electrical field stimulation (EFS) were unchanged In the presence of 5 x 10· ZAG, but the slow contractile response to EFS was slightly potentiated (p < 0.01)." during the LAR the numuer of granUlocytes and their activation resemble those used In these in vitro experiments the results suggest that during the LAR granulocytes do not Increase airway muscle sensltlvhy but may contribute to muscle contraction. Most of this muscle contraction seems secondsry to the release of lTC., which Is relatad to the proportion of eoslnophlls among the granUlocytes. AM REV RESPIR DIS 1991; 143:553-560

fore and during the operation 10 patients received steroids andlor theophylline. A irway Preparations A macroscopically normal part of the resected tissue was immersed 30 to 60 min after surgical resection in Krebs-Henseleit buffer at room temperature (composition in mM: NaCI 118,KC14.7, CaCl,2.5, MgSO. 1.2, KH,PO. 1.2, NaHCO. 25, and glucose 5.55) that had been aerated with carbogen (9501001 and 5010 CO,) to produce a pH of 7.35, a Pco, of 4.7 kPa, and a Pol of 71.8 kPa. The tissue remainedin fresh aerated buffer throughout the dissection procedure and the experiments. On the cut surface airways were identified, cannulated, taken out, and dissected free from parenchyma and vesselsunder a stereomicroscope (x 20 magnification) using iris scissors and forceps. The cleaned airway was cut into segments of 4 to 5 mm length. For experiments with electrical field stimulation, which were done on the day of operation, central cartilaginous airways (generation 3 to 6) with a diameter Of 3 to 6 mm were used. For pharmacologic experiments segments from periph-

eral airways (generation 6 to II) with a diameter of 2 to 3 mm were used. These were stored overnight in a slow flow of aerated buffer of 4 0 C containing penicillin (3 x 10-' giL) and tobramycin (5 x 10-' giL). We formerly demonstrated that this storage procedure does not affect the contractility of the 'preparations to methacholine (II). The next day the segments were mounted between two

(Received in original form February 28, 1990 and in revised form August 15, 1990) 1 From the Department of Pediatrics, Subdivision of Pediatric Respiratory Medicine, and the Department of Pharmacology, Erasmus University and UniversityHospital Rotterdam/Sophia Children's Hospital, Rotterdam, The Netherlands. 1 Supported by The Netherlands Asthma Foundation. 3 Correspondence and requests for reprints should be addressed to J.C. de Jongste, Sophia Children's Hospital, Division of Pediatric Respiratory Medicine,Gordelweg 160,3038GE Rotterdam, The Netherlands.






-Log EC,. 6.4 ± 0.14 6.9 ± 0.18 - t 6.4 ± 0.28

10-°-10- M 10-'-5 x 10-0 M 10-'-5 x 10-' M 0



7 6 3 5

Maximal Contraction' (% Max)

10-'-10-' M 0.01-1.0 mglml

69.5 66.2 19.0 28.5 No effect

± ± ± ±

12.1 8.3 10.0t 6.3

Definition of abbreviations: n = number of praparations from different lungs; A23187 = calcium ionophore A23187; FMLP = N·formylmethionylleucylphenylalanine; PMA = phorbol-tz-mynstate-ta-acetate; ConA = concanavalin A; STZ = serum-treated zymosan. • The maximal contraction is expressed as a percentage of the maximal contraction to methacholine. t The -log EC" could not be computed. PMA slightly relaxed the airway up to 10-' M and caused contraction at higher concentrations.

small polished stainless steel hooks (diameter 0.3 mm) and placed between a glass hook at the bottom of a double-jacketed organ bath of 10 ml and a high-precision isotonic angular position transducer (3810/60; Penny and Giles Potentiometers Ltd., Christchurch, UK). The bath was siliconized(dimethyldichlorosilane, BDH, United Kingdom) to prevent adherence of the granulocytes. The signal from the transducer was monitored with a digital voltmeter (Fluke 73 multimeter; John Fluke Mfg. Co. Inc.,Everett, WA)and a pen recorder (BD 40; Kipp & Sons Scientific Instruments, Delft, The Netherlands). This method wasdescribed in detail previously (12). The preparations contracted against an isotonic load of 250 mg, which has been shown to be optimal for human bronchial segments (12). In all experiments the preparations were contracted twice with methacholine (10-' and 10-4 M) during an equilibration period of 2 h to assess smooth muscle function. Between stimulations the preparations were washed four times with intervals of 2 min and every 15 min afterwards until stabilization of resting length, which took 20 to 30 min.

Effect of Cell Activators on Isolated Peripheral Airways To find a cell activator that would not change airway caliber cumulative concentration-response curves(CCRC) wereobtained with the

following stimulants: calcium ionophore A23187,N-formylmethionylleucylphenylalanine (FMLP), phorbol-12-myristate-13-acetate, concanavalin A, and serum-treated zymosan (STZ). As shown in table 1STZ (0.01 to 1.0 mg/ml) was the only compound that did not alter the contractile state of the airways.The highest concentration of STZ that would remain in suspension was 0.2 mg/ml, and therefore this concentration was used to activate the granulocytes. It gives maximal HzOz production by neutrophils and eosinophils (13) and induces about 40070 of the maximal LTC4 formation from purified human eosinophils (14). STZ was prepared by homogenizing zymosan at 4 mg/ml in fresh human serum with a Potter-Elvejhem homogenizer (B. Braun, Melsungen, Germany), followed by incubation of the mixture for 30 min at 37° C. After centrifugation and two washings the STZ was suspended in 0.9% NaCI at 20 mg/ml (15).

Donors and Isolation of Granulocytes Granulocytes wereisolated either from laboratory personnel (n = 5) or subjects under control for hemostasis (n = 5) who denied symptoms of allergyor asthma. Granulocytes from atopic donors (n = 3) were also used to see if these had similar effects. The characteristics of the atopic donors are given in table 2. The number of observations with


Age 28 22 30



Asthma, Eczema, Rhinitis



a, e a, r a

PD,.FEV, Histamine (IlY)'



100 120 78

5.6 7.5 5.8

None None None

Definition of abbreviations: RAST = strongly positive (> 3 +) radioallergosorbent test; PD"FEV, ~ provocative dose that caused a 20"Al fall in FEV, from baseline; C = cat dander; HDM = house dust mite; P = grass pollens; a = asthma defined as paroxysms of wheezing; e = eczema; r = rhinitis; "Aleo = eosinophils as a percentage of leukocytes in peripheral blood. • Dosimeter method (42). t Within 24 h before donation.

granulocytes from atopic donors is indicated with the results. The granulocytes were isolated according to the method by Roos and de Boer (16). Briefly, heparinized blood (20 to 30 ml) was diluted two times with phosphate-buffered saline (PBS) at pH 7.4 containing 13 mM trisodium citrate and centrifuged (1,000 x g) at room temperature over a pyrogen-free Ficoll-Metrizoate gradient of 1.077 g/ml density (Nycomed, Norway) for 20 min. The erythrocytes in the pellet were lysed by repeated incubation (15 min) with an excess of ice-cold isotonic ammonium chloride. After centrifugation the cells were washed twice and suspended in PBS containing 0.5070 (wt/wt) bovine albumin (Sigma Chemical Co., St. Louis, MO). Total cell number was counted in a Burker cell chamber. Viability was tested with the trypan blue exclusion test. Cytocentrifuge smears of each isolate were prepared from 10 IIIcell suspension, and differentialcounts wereobtained after staining according to May-Griinwald-Giemsa.

Influence of Granulocytes Activated with STZ on Peripheral Airways The influence of ZAG on the contractile state and sensitivity to methacholine of the airway preparations was investigatedas follows. Half an hour after a first CCRC to methacholine (10-8 to 10-4 M) 1,2,5, 10, or 20 x 106 granulocytes were added to the organ bath. Timeparallel controls receivedno granulocytes.After stabilization, usually 10 min later, STZ was added (0.2 mg/ml), which elicited a contractile response. In figure 1 (top) is illustrated the response to addition and activation of 10 x 106 granulocytes. Once this response reached a plateau a second CCRC was obtained in the presence of the ZAG (not shown in figure 1). After washing and stabilization a third CCRC to methacholine was made. Mechanism of Bronchoconstriction by ZAG Toelucidate whether eicosanoids or histamine contributed to the bronchoconstriction evoked by ZAG portions of 10 x 106 granulocytessuspended in 0.5 ml PBS wereincubated at 37° C with the lipoxygenase inhibitor nordihydroguaiaretic acid (NDGA, 10-5 M), the cyclooxygenase inhibitor indomethacin (10-5 M), the LTC4/D4 receptor antagonist FPL 55712(11.5 x 10-6 M), or the histamine receptor antagonist mepyramine (2.8 x 10-6 M). At this concentration NDGA totally inhibits the formation of Iipoxygenase products and indomethacin totally blocks the formation of cyclooxygenase but not lipoxygenase products by human granulocytes stimulated with calcium ionophore (17). FPL 55712 at this concentration totally prevents the response to 10-7 M LTC4 , which causes maximal contraction of human trachealis (18), and mepyramine at this concentration blocks the contraction of isolated human bronchi to 10-4 M histamine (19).Sipiultaneously, bronchial segments were incubated with the same concentrations of these drugs. After 15 min of incubation the granulocytes were trans-






w t

MChlO- 5M


Fig. 1. Recording of a typical response of human bronchus to 10 x 10· ZAG in the absence (top) and presence (bottom) of the lipoxygenase inhibitor NDGA (10-S M), preceded by a graded contraction to methacholine. At the points indicated by the arrows methacholine (Mch), NDGA, 10 x 10· granulocytes (cells), or serumtreated zymosan, 0.2 mg/ml (STZ), were added to the organ bath. Washing is indicated by w.

ferred to the corresponding organ baths. STZ (0.2 mg/ml) was added after stabilization. Time-parallel control preparations and granulocyte suspensions werealso incubated without a drug at 37° C for 15 min. To see whether oxygen metabolites wereinvolved in the contraction of airway smooth muscle to ZAG a combination of superoxide dismutase (SOD, 300 V/ml) and bovine catalase (5,000V/ml) wasadded to the organ bath just before addition of 10 x 106 granulocytes. At these concentrations SOD immediately metabolizes 0,- into H,O, (20) and catalase degrades H,O, into H,O and 0, (21) so that the formation of free 0,- or H,O, is inhibited. After stabilization STZ (0.2 mg/ml) was added.


Fig. 2. Tracing of a typical response of a fresh human bronchus to electrical field stimulation (EFS) in vitro. At the point indicated by the arrow a 30-s tetanus of supramaximal voltage (50 V), short pulse duration (0.3 ms), and a frequency of 30 Hz was given. A cholinergic contraction is followed by a deflection below baseline (nonadrenergic relaxation) and a slow contraction (9).

cholinergic excitatory nerves. The relaxation phase is nonadrenergic and partly inhibited by the neurotoxin tetrodotoxin (9, 22). In addition EFS gives rise to a tetrodotoxininsensitive slow contractile response that results from the release of metabolites of arachidonic acid in the bronchial wall (9). A typical response is shown in figure 2. First, a control response to EFS was obtained in each of two airway segments from 11 tissue specimens.After 30 min 5 x 10" granulocytes, the highest number that would not significantly alter smooth muscle sensitivity to methacholine (see table 3), wereadded to one of the two organ baths and activated with STZ (0.2 mg/ml). The control preparation received methacholine to match the increase in tone induced by 5 x 106 ZAG because earlier findings indicated that the response to EFS depended on baseline airway tone (23). Half an hour later a third response to EFS was obtained after contracting both the control and the ZAG-treated airways with methacholine to a level similar to that before the second

Effect of ZAG on Responses to EFS in Central Airways Tho segments of central cartilaginous airways from each tissue specimen were mounted in an organ bath under a 500-mg isotonic load and stimulated twice with methacholine (10-5 and 10-4 M). As described in detail before (9) EFS was applied with a custom-made tissue stimulator that produced voltage-constant rectangular pulses of alternating polarity via platinum plate electrodes positioned parallel to the preparation. To produce near maximal responses 30-s tetani of supramaximal voltage (50 V), a short pulse duration (0.3 ms), and a frequency of 30 Hz were used. We previously showed that these stimuli lead to a triphasic contraction-relaxation-contraction response of the airwaymuscle.The initial contractile phase results from activation of

EFS stimulation. The effect of ZAG on the nonadrenergic relaxation response to EFS was examined in separate experiments. Two airway segments from each of seven tissue specimens were incubated with a combination of FPL 55712 (11.5 x 10-6 M), atropine (1.2 x 10-6 M), and indomethacin (6 x 10-6 M) for 20 min before stimulation to inhibit the cholinergic peak and the slow contractile phase of the response to EFS (9). Next, one preparation received 5 x 106 granulocytes, which werestimulated with STZ (0.2 mg/ml), and the other remained in normal buffer. The segments werethen precontracted with histamine (5 x 10-6 M). After stabilization a graded pulse train was applied (0.3 ms, 50 V, and 1,2,5, 10,20, and 50 Hz) to obtain frequency response curves of EFS-induced relaxation (22).

Data Analysis Contractions were expressed as a percentage of the second response to 10-4 M methacholine obtained at the beginning of each experiment (seeprevious paragraphs on patients and airway preparations). Relaxations were expressed as a percentage ofthe maximal relaxation in the presence of isoprenaline (10-4 M) and ethylenediaminetetraacetic acid (EDTA, 4 x 10-3 M), determined at the end of the experiment. The sensitivity to methacholine was expressed as the negative logarithm of the methacholine concentration that caused 50070 of the maximal narrowing (-log EC so) . The sensitivity to EFS was defined as the effective frequency that causes half-maximal relaxation to EFS (EF so) . The -log EC so and EF so were calculated using the BMDP software module for nonlinear regression (24). The curves were fitted to a four-parameter logistic function (25). Contractions, relaxations, and EF so were compared with paired t tests. Valuesof -log EC 5 0 wereanalyzed with repeated-measurements analysis of variance (ANOVA) using the MANOVA program of SPPsx (26). Correlations were examined by calculating Spearman's rank correlation


- Log EC•• Methacholine

n' 7 6 6 7 6 4

(3) (3) (3) (3) (3) (2)

Before Granulocytes 5.94 6.00 6.05 5.98 6.03 6.12

± ± ± ± ± ±

0.10 0.13 0.10 0.05 0.06 0.12

Granulocytes Present 6.00 6.07 6.01 5.97 5.88 5.76

± ± ± ± ± ±

0.10 0.12 0.18 0.15

o.nt 0.14:/:

After Granulocytes 5.95 5.99 5.95 5.97 6.00 6.07

± ± ± ± ± ±

0.09 0.14 0.11 0.04 0.07 0.08

• Number of preparations from different lungs. Numbers in parentheses indicate the number of observations with ZAG from atopic donors. Of each preparation the first CCRC was obtained before ZAG, the second in the presence of ZAG, and the third after washout. Statistical significance is indicated. The presence of ZAG is a significant factor in the ANOVA. t p < 0.05.

t p < 0.001.


556 70











:c o ::> "''-

50 40




t ~0

100 1












20 20


10 10




1 2 5 10 20 number of grarulocytes Ix 1.000.000)

Fig. 3. Mean contractileresponsesto increasing numbers of granulocytes. The open bars show the contractile responseto unstimulated granulocytes. The hatched bars showthe contractioncausedby the samegranulocytesafter stimulation withSTZ(0.2mglml).Thehorizontal axis gives the logarithm of the number of granulocytes.The vertical axis represents the contractile state of the preparations expressed as a percentage of the maximal contractionto methacholine(%Mch); n = 4-7. Statistical significance comparedto the responsewith unstimulatedgranulocytes: *p < 0.05; **p < 0.Q1. Open bars = granulocytes; hatchedbars = granulocytesplus zymosan.

coefficient (rs). P values < 0.05 (two-sided) wereconsidered significant. All data are presented as mean ± SEM.

Drugs Methacholine, histamine(both Janssen Pharmaceutica, Belgium), atropine (Brocacef,The Netherlands), mepyramine (Rhone-Poulenc, France), bovine catalase (Boehringer Mannheim, WestGermany), superoxidedismutase, EDTA (both Sigma Chemical Co.) and concanavalin A (Pharmacia, Uppsala, Sweden) were dissolved in saline. Indomethacin (Duchefa, The Netherlands),nordihydroguaiaretic acid (Sigma Chemical Co.) and FPL 55712 (Fisons, United Kingdom) were dissolved in methanol. Phorbol-12-myristate-13-acetate, calcium ionophore A23187, and N-formylmethionylleucylphenylalanine (all from Sigma Chemical Co.)were dissolvedin dimethylsulfoxide. L-isoproterenol sulfate (Janssen Pharmaceutica, Belgium) wasdissolved in water containing ascorbic acid (88 mg/L), Preliminary experiments showedthat ascorbate, methanol, and dimethylsulfoxide in theseconcentrations have no effect on airway muscle function. Results

Granulocytes Granulocyte suspensions from healthy volunteers consisted of93.8 ± 0.8070 neutrophils, 4.0 ± 0.6070 eosinophils, and 2.2 ± 0.4070 lymphocytes and those of the atopic donors of 81.4 ± 2.30/0 neutrophils, 14.0 ± 1.9070 eosinophils, and 3.3 ± 0.9070 lymphocytes (p < 0.001 for difference in percentage neutrophils and eosinophils). Viability exceeded 96070.





Fig. 4. Influence of different drugs on the contractile response to 10 x 108 granulocytes. Theopen barsshow the contractile responseafter incubationwith unstimulatedgranulocytes in the presence of thedifferentdrugs. The hatched bars show the contractioncaused by the same granulocytes after stimulation with STZ (0.2 mg/ml). The foiiowing drugs were used: NDGA(10-5 M, n = 16),FPL55712(11.5 x 1O-8M,n = 16),indomethacin (10-5 M, n = 7), mepyramine(2.8 x 10-8 M, n = 5), and a combination of superoxide dismutase (300 units/ml) and bovine catalase(5,000 units/ml, n = 5). Controlpreparations(n =16) received granulocytesbut no drugs. The vertical axis represents the contractile state of the preparationsexpressed as a percentageof the maximal contraction to methacholine (%Mch). Statistical significance comparedto control: *p < 0.05, ***p < 0.001. Open bars = granulocytes; hatched bars = granulocytes plus zymosan.

Influence of ZAG on Peripheral Airways Figure 3 shows that activation of 1, 2, 5, 10, or 20 x 106 granulocytes with STZ (0.2 mg/ml) produced significantly more bronchoconstriction than unactivated granulocytes (n = 4 to 7, p < 0.05). The bronchoconstriction was proportional to the number of zymosan-activated granulocytes present (rs = 0.79, p < 0.001). Table 3 shows that in these experiments the sensitivity (-log EC so) of the airways to methacholine was unchanged in the presence of 1, 2, or 5 x 106 ZAG but was significantly lower with 10 or 20 x 106 ZAG (p < 0.05). ZAG from atopic donors were used in two to three experiments.

Mechanism of Bronchoconstriction by ZAG Figure 4 shows the contraction induced by 10 x 106 granulocytes in the presence of different drugs compared to the effect on paired control preparations of the same number of granulocytes from the same donors without drugs. The lipoxygenase inhibitor NDGA reduced both the contraction after addition of granulocytes (0.4 ± 3.5 versus 10.2 ± 2.7070; n = 16, p < 0.05) and the contraction after activation of the granulocytes (16.2 ± 7.2 versus 49.6 ± 5.6070; n = 16, p < 0.001). This is illustrated in figure 1. FPL 55712 did not reduce the contraction to unstimulated granulocytes (8.6 ± 5.1 ver-





nLlTlber of eosinophils (x 1OO.OOOl

Fig. 5. Correlationbetweenthe numberofeosinophils present in 10 x 10" granulocytes stimulated with STZ (0.2mglml)andthe magnitUde olthe contraction induced by these granulocytes. The logarithmic horizontalaxis gives the number of eosinophils present in 10 x 108 granulocytes. The vertical axis represents the contraction inducedby these granulocytesexpressedas a percentage of the maximal contraction to methacholine (%Mch).The closed circles representthe mean results from granulocytesfrom normaldonors(n = 10)and the open circles mean results with granUlocytes from atopic donors(n = 6; see text). rs is Spearman's rank correlation coefficient. rs = 0.66; P < 0.Q1.

sus 10.2 ± 5.7070; n = 16, p > 0.05) but significantly diminished the contraction to activated granulocytes (49.6 ± 5.6 versus 15.2 ± 8.3OJo;n = 16,p

Effects of zymosan-activated human granulocytes on isolated human airways.

In asthma a temporal association exists between the late allergic reaction (LAR), the influx of granulocytes into the airway wall, and an increase in ...
982KB Sizes 0 Downloads 0 Views