Eosinophil cationic protein stimulates major basic protein inhibits airway mucus secretion

and

Jens D. Lundgren, MD,* Richard T. Davey, Jr, MD,** Bettina Lundgren, MD,* Joaquim Mullol, MD,*, *** Zvi Marom, MD,**** Carolea Logun, MSc,* James Baraniuk, MD,*** Michael A. Kaliner, MD,*** and James H. Shelhamer, MD* Bethesda, Md.. and Ne\tl York, N.Y. Possible roles of eosinophil (EO) products in modulating the release of mucu.sfrom airway explants were investigated. Cell- and membrane-free lysates from purijied human EOs (I to 20 x 10’) caused a dose-dependent release qf respiratory g1ycoconjugate.s (RGC) from cultured ,feline tracheul explants. Crude extracts from isolated EO granules also stimulated RGC release, suggesting that a granular protein might be responsible. Three proteins derived from EO granules, EO-derived neurotoxin, EO cationic protein (ECP). and major basic protein (MBP) were separated by sequential sizing and ajj?nity c,hrorncltoKrtrphv. ECP (0.025 to 25 pglml) caused a dose-dependent increase in RGC release,from both feline and human airwq explants and also stimulated the release of the serou.s cell-marker, Iactofc’rrin. from human bronchial explants. EO-derived neuroto.rin (0.025 to SO pglmml)fctiled to aflect RGC release, whereas MBP (50 p+glml) .signi$cantly inhibited RGC release from feline e.rp1ant.s.Thus, ECP stimulates RGC und lactoferrin release from airway e.~plants, whereas MBP inhibits RGC release. (J ALLERGYCLIN IMMUNOL 199/:87x589-98.1

One of the predominant features of the asthmatic attack is hypersecretion of mucus from the airway mucosa, leading to airflow obstruction and even total occlusion of the peripheral airways.’ Possible mechanisms behind mucus hypersecretion include abnormal neuronal regulation of the submucosal glands and the production of proinflammatory lipids and proteins from airway epithelium,’ mast cells,’ neutrophils,j macrophages ,4 and EOs . A hallmark of asthma is the accumulation of EOs in the airways. ‘. ’ A number of lines of evidence suggest that Eos contribute to the pathogenesis of the

From the *Critical Care Medicine Department. Clinical Center, **Division of Parasitology, and ***Allergic Diseases Section, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Md.; and ****Mount Sinai Medical Center, New York, N.Y. Parts of this article have been published as an abstract in Am Rev Respir Dis 1988;137(4 p2):ll. Received for publication May 8. 1990. Revised Nov. I, 1990. Accepted for publication Nov. 2, 1990. Reprint requests: Jens D. Lundgren, MD, Department of Infectious Diseases (144), Hvidovre Hospital, University of Copenhagen. DK-2650 Hvidovre. Denmark. l/1/26700

Abbreviations used

CRML 1066: ECP: EDN: EO: EPO: MBP: PBS: RGC: SI: Pi:

P2: PI: PT: MW: SDS-PAGE:

Type of culture medium Eosinophil cationic protein Eosinophil-derived neurotoxin Eosinophil Eosinophil peroxidase Major basic protein Phosphate-bufferedsaline Respiratory glycoconjugates Secretory index Period I Period 2 Isoelectric point Buffer of PBS with 0.05% Tween 80 Molecular weight Sodium dodecyl sulfatepolyacrylamide gel electrophoresis

asthmatic attack. First, there is a temporal relationship between EO accumulation in the airways and the development of the late-phase reaction and subsequent increases in airway reactivity.’ Second, the EOs in the airways release granular constituents, as reflected by increased amounts of ECP and MBP in bronchoalveolar lavage fluid8 or sputum’ during asthmatic at689

690 iundgren et ai

t.r,:ir IO minutes). The granule-enriched supernatant ~a., ;entt: fuged (10,000 R for 20 minutes). and the EC>granules ~VI-L‘ collected from the central part of the pellet ,rnc! h&~eJ in vials (NUNC, Roskilde, Denmarhi under l~quiti nttrogcn until use. A crude extract from the granule> ti.a- made hi dissolving the sonicated pellet in 0.01 N HCI CMRI~ 1066 medium was added to achieve a concentratmn .)t extrac’t corresponding to granules from IO” EOs per mlilihter I ai‘ stock solution. EO grmulr prorein sepwation. As an inihal izparation procedure of the various EO granule proteins. EO pranulcs were sonicated, extracted in 0.01 N HCI. and centrifuged to remove debris. The supernatant was applied IO a Sephadex G-50 sizing column ( 1.2 by 47 cm) equilibrated Mith 0.025 moi/L of sodium acetate. 0.15 mol/L of NaCt. pH J 5. Fractions of 0.5 ml were collected. and elution of pmteln was monitored at 277 nm in a spectrophotomctcr ” ’ The effluent was pooled into four fractions (El F!) E3 I. Each pooled fraction was concentrated by ultrafiltration (PM I(! hlter) and either reconstituted with CRMI IO66 medium (corresponding to granules from lo” EOs per mtlliliter of stock solution, assuming no loss during the purilkation procedure) or subjected to additional separation procedures. The fraction E3 was additionally processed to separate EDN from ECP. after diluting the preparation I : I with water. The mixture of ECP and EDN was applied to a heparin Sepharose 6B column (0.5 by I2 cm), cyuihbrated with PBS diluted 1: I volivol with distilled water ione half times PBS. pH 7.4). The column was eluted with a continuous gradient of NaCl from 0. I5 to I .S mol!!. -’ The non. adherent (i.e.. EDN) and the adherent (i.c. 1 ECP) peaks were collected, concentrated by ultrafiltration (PM IO tilter). washed twice with one half times PBS, and str)red tn liquid nitrogen. When these preparations were added to the airway cultures (see below), the quantity of one half times PBS added to the cultures was < I% of the total voiumc. Protein chemisrr?;. ‘The protem concentration of EO granule proteins was measured by absorbance spectroscopy at 277 nm with extinction coefficients as previously defined.“’ Identification of the individual protein!, and assessment of purity of the protein preparations were performed by MW analysis and with SDS-PAGE with Coomassie blue staining. As an additional precaution, identity of the individual proteins was also confirmed on the basis of specific reactivity on ELISA with murine monoclonal antibodies previously prepared against each of the four major human EC) granule proteins (data not presented). These data were confirmed

VOLUME NUMBER

Eosinophils

87 3

+50

r ***

** 1.

and airway

mucus secretion

691

* L 3. **

n=4

r

n=4

n=4

n=4

0.1 1 0.4 2 LYSATE FROM NUMBER OF EOSINOPHILS (xl06 cells/ml)

FIG. 1 Effect of crude extracts of lysed EOs (lo5 to 2 x IO6 EOs per milliliter of culture media) on the release of RGC from feline tracheal explants after 1 hour of incubation; n indicates the *p < 0.05; number of separate experiments. EOs were purified by a metrizamide gradient; **p < 0.02; ***p < 0.01.

with the Western blot technique (data not presented). The monoclonal antibodies were raised by one of the authors (R. T. D.) in mice against the various EO granule proteins purified by standard methods. The specificity of the four monoclonal antibodies was determined. The monoclonal antibodies for EPO and for MBP did not cross-react with each other or with ECP or EDN . The monoclonal antibodies for ECP and EDN cross-reacted with each other but not with MBP or EPO. That the monoclonal antibodies for ECP and EDN cross-reacted with each other is not surprising, since the two proteins have a 50% amino acid homology.21 Airway organ culture system to measure ‘H-RGC release. Airway explants were cultured, and RGC were radiolabeled as previously described.2.22Briefly, for feline airway studies, cats were sedated with intramuscular injection of ketamine, killed by intravenous pentobarbital, and the airways were resected. For human airway studies, macroscopically normal human bronchi were isolated from surgical specimens obtained after pulmonary lobectomy, primarily obtained because of bronchogenic carcinoma. For both airway preparations, two 3 by 3 mm airway fragments were placed on gel foam in Petri dishes together with ‘H-glucosamine (1 @i/ml) and CMRL 1066 culture medium with added penicillin (100 kg/ml), streptomycin (100 kg/ml), and amphotericin B (0.5 pg/ml). The cultures were kept in a controlled atmosphere of 45% 0,, 50% N,, and 5% CO, at 37” C on a rocker platform. After 42 hours, the cultures were treated according to the following protocols. Cultures were maintained for up to 10 days.

Experimental designs. The experiment was started with a Pl , 4 hours, to measure baseline release of RGC from the explants. Pl was followed by P2, 1 hour, in which various preparations of EOs were added to some of the cultures, whereas other cultures were kept as controls. In time-course experiments, P2 varied from l/z hour to 4 hours. At the end of each period, the culture media was collected, and the explants were washed. The original culture media and the washings were pooled and assayed for ‘H-RGC, as described below. ‘H-RGC assay. The ‘H-RGC in the collected culture media was precipitated overnight in 10% trichloroacetic acid and 1% phosphotungstic acid at 4” C. The precipitate was harvested by centrifugation, washed twice, redissolved in 0.1 mol/L of NaOH, and the amount of radioactivity was measured by scintillation counting (Beckman LS 8100, Beckman Instruments, Inc., Fullerton, Calif.). An SI was established by comparing the disintegrations per minute for P2 to Pl . For each manipulation, three to four explants were used, and the average SI for these plates was calculated. By comparing the SI of chemically stimulated cultures with parallel control cultures, a percent change from control was calculated. EUSA for lactoferrin. In experiments testing the effect of ECP on human bronchial explants, the lactoferrin concentrations in culture media from Pl and P2 were determined by a noncompetitive ELISA.23 Microtiter plates were coated with 50 ~1 of rabbit antihuman lactoferrin diluted 1: 1000 in 0.1 mol/L of sodium carbonate buffer (pH 9.6) and

692

Lundgren

et al.

+100 +90 =I 2

E zs 5E JO u.k a

O& c.32 0 8

+80 E

*** I n=6

+70 -

*** r

+60 -

n=4

+50 +40-

n=4 **

+30 -

n=4 n.s.

+20 +10 0t

so z-

DJLL 2.5 SOURCE OF GRANULES-NUMBER (x 10Vml)

0.25 OF CELLS

FIG. 2. Effect of EO granule extracts on the release of RGC from feline tracheal explants after 1 hour of incubation. On the x axis, the number of EOs from which granule extracts were prepared are indicated. For the experiments with 2.5 x 105, 2.5 x 106, and 5 x IO7 EOs per milliliter of culture media, n = 6, whereas for the experiments with 5 x IO7 EOs per milliliter of culture media, n = 4.

incubated at 4” C overnight. The wells was washed with four volumes of a buffer (PT) (pH 7.4). After nonspecific binding sites were blocked with 1% goat serum in PT for 30 minutes at room temperature, 50 pl of sample or standard (diluted in PT) was added to each well and incubated at 37” C for 90 minutes. After wells were washed, rabbit antihuman lactoferrin conjugated to horseradish peroxidase (50 L) was added and incubated at 37” C for 90 minutes. The color reaction was developed with o-phenylcnediamine di-HCl, and the optical density at 490 nm was determined spectrophotometrically. The ratios of optical densities from the supematants collected during P2 to that for PI (SI) were calculated for each treatment, and the mean ( ? SEM) percent change in SIs from control values were calculated. Indirect immunohistochemistry for localization of lactoferrin. Lactoferrin-reactive material was located in human bronchial tissue with previously described techniqu& with polyclonal rabbit antihuman lactoferrin serum. Primary antibody localization was visualized with colloidal goldlabeled goat antirabbit antibodies with silver enhancement. Slides were counter stained with alcian blue to stain mucous glycoprotein containing mucus and goblet cells. Statistics. An unpaired Student’s t test was used to compare the percent change from control (average 2 SEMs). When two data points in parallel experiments were compared. a paired Student’s t test was used. A p value CO.05 was considered significant. On Figs. 1 to 5, asterisks are used to designate degree of significance with the following

scale: *p < 0.05; **p < 0.02; ***p < 0.01; n. s., not significant.

RESULTS EOs were isolated with the metrizamide gradient technique to >98% purity. Debris-free, crude extracts, obtained after freeze-thaw lysis of EOs, contained an activity capable of releasing RGC from feline tracheal organ cultures in a dose-dependent manner (Fig. 1). The RGC release was increased by 43% as compared to release from control cultures when cultures were exposed to the supematant from 2 x lo6 EOs per milliliter. The lysate from as few as 0.1 x lo6 EOs also caused a significant increase in RGC release. To determine if the RGC-secretagogue activity was localized in the granules of EOs, the effects of extracts from isolated EO granules were studied. Granule extracts (dose range, granules from 2.5 x IO” to 5 x 10’ EOs per milliliter of culture media) caused

a dose-dependent release of RGC, with a nearly 90% increase in RGC release as compared to that of controls after incubation with granules from 5 x IO’ EOs per milliliter of culture media (Fig 2). Granules from as few as 2.5 x lo6 cells per milliliter

caused a sig-

nificant increase in RGC release. When cultures,

VOLUME NUMBER

87 3

which had been previously challenged with crude granule extract, were rechallenged on subsequent days, the same increased RGC response was observed as noted with cultures not previously exposed to crude extract (data not presented). Cultures previously challenged with crude granule extract also had the same response to methacholine (increased RGC release) as unexposed cultures (data not presented). EO granules contain four well-defined proteins, EPO, EDN, ECP and MBP.24 These granule proteins can be separated from each other by molecular sizing, followed by ion exchange chromatography. The first purification step separates EPO, ECP/EDN, and MBP by use of a Sephadex G-50 column. Application of EO granule extract onto a Sephadex G-50 column resulted in elution of three peaks when the eluent was monitored for changes in absorbance at 277 nm (Fig. 3, top). The first peak corresponded to the void volume of the column (designated E2, see Fig. 3) and contained a variety of bands as observed on SDSPAGE gel. The three major bands in this fraction were identified as native EPO and two subunits thereof.25 The second peak (designated E3) contained two slightly separated bands, each with an MW of approximately 17,000 (mixture of ECP and EDN),‘9-21 whereas the third and final peak (E4) contained one major band with an MW of 9, whereas heparin has a pI >3. A preparation of EO granules corresponding to 5 x lo7 EOs per milliliter of culture media was added to feline airways and resulted in an increased RGC release (90.0% ? 11.7% compared to control, n = 6; p < 0.01). If, however, the preparation was mixed with heparin (25 p,g/ml) and then added to the airways, a significantly reduced increase in RGC release occurred (27.3% + 2.5% compared to control, p < 0.05, Student’s paired t test). Heparin alone had no effect on RGC release (4.2% + 9.4%

Eosinophils

and airway

mucus secretion

693

ELUTION PROFILE OF EOSINOPHIL GRANULE PROTEIN EXTRACT ON G-50 SUPERFINE COLUMN

26

36

FRACTION

48

80

NUMBER

EFFECT OF VARIOUS FRACTIONS COLLECTED FROM G-50 SIZING COLUMN

E3

***

FIG. 3. Top: Spectrophotometric (277 nm) elution profile of a Sephadex G-50 sizing column after application of EO granule extract. The first peak (E2) corresponds to the elution of EPO, the second peak (E3) to ECP and EDN, whereas the third peak (E4) corresponds to MBP (see METHOD for details). The fractions pooled to generate El to E4 are indicated on the figure; VV, void volume; CV, column volume. Bottom: Effect of El, E2, E3, and E4 on the release of RGC from feline tracheal explants (n = 8; 1 hour of incubation). El to E4 were concentrated to a concentration equivalent to granules from 5 x 10’ EOs per milliliter of culture media, assuming no loss during the purification procedure. Comparing E2 to E3 by a Student’s paired t test elicits a p value cO.02.

compared to control). These data suggested that the major secretagogue activity could be partially inhibited by heparin and suggested that the effect could be due to either ECP or EDN. A lower concentration of heparin (2.5 kg/ml) failed to significantly inhibit the RGC-releasing effect of EO granule extracts (n = 5; data not presented). EDN can be separated from ECP by use of the higher affinity of ECP for heparin. Thus, the E3 fraction was applied to a hepatin Sepharose affinity column, and the nonadherent effluent, as well as the

694

Lundgren

et al.

;. ALLERGY

EFFECT OF PURIFIED ON FELINE TRACHEAL

%

+80

i.L!N.

IMMUNOL MARCH 7497

ECP AND EON RGC RELEASE

=2

lug 0 rn

n=5 .--

0

T

7

ECP

m EDN n=4 n.s.

I 2.5 ’ 0.25 ’ 0.025 ’ 50 25 CONCENTRATION (microgram/ml) FIG. 4. Dose response of ECP (dose range, 0.025 to 25 ~giml) 50 *g/ml) on the release of RGC from feline tracheal explants

adherentmaterials, was collected.‘“, 2’ On SDS-PAGE gels, the nonadherent fraction of the column (i.e., EDN) contained a single protein band, whereas the adherentfraction (i.e., ECP) consistedof two slightly separatedbands, as has been noted previously.” The MW of the molecules obtained after heparin Sepharosechromatographydid not differ from MW observed after SDS-PAGEof E3 (see above). A dose-response experiment examining the effectsof purified EDN and ECPon the releaseof feline tracheal RGC is presented in Fig. 4. ECP caused a dose-dependentincrease in RGC release when ECP was incubated in concentrations ranging from 0.025 to 25 pg/ml, whereasEDN (range, 0.025 to 50 pg/ml) had no effect. The time course of the response to ECP (2.5 pg/ml) on RGC release from feline tracheal organ cultures after l/2, 1, 2, and 4 hours of incubation is illustrated in Fig. 5. Maximum stimulation was observed after 1 hour, but a significant effect of ECP persisted through 2 and 4 hours. Since ECP appearedto be the predominant RGC secretagogueresponsible for the stimulatory effect of crude EO extracts in feline tracheal organ cultures, it was of interest to investigate if ECP also had RGC releasingcapabilities in human respiratory tissue. Lactofetrin was localized immunohistochemically to the submucosalglands in human bronchial mucosa(Fig. 6). MUCUScells, which were identified by alcian blue

and EDN (dose range, 0.025 to (1 hour of incubation).

staining, did not contain lactoferrin, whereas serous cells stain darkly. The epithelium did not contain significant lactoferrin immunoreactive cell populations. The addition of exogenouslactoferrin ablatedthe binding of the antiseraasreflectedin an absenceof positive staining. Therefore, releaseof lactoferrin into the culture media was used as a marker of serous cell secretion, much as was observed in lactoferrin release from human nasal mucosa.23The dose-responseeffect of ECP on the release of RGC and lactoferrin from human bronchial organ cultures is presentedin Table 1. ECP in a concentration of 2.5 pg/ml of culture mediacaused18.2% increasein RGC releaseand 50% increase in lactoferrin release. Diluting ECP to 0.25 p.g/ ml resulted in the failure to releaseRGC, whereas lactoferrin releasestill was significantly increased. To investigate further the inhibitory effect of MBP (i.e., E4, Fig. 3), adose-responseexperiment of MBP on feline tracheal organ cultures was performed. RGC release was significantly inhibited by - 23.7% i I .6% comparedto control cultures (n = 6; p < 0.0 1) when 50 Fg/ml of MBP was incubated with the cultures for 1 hour, whereas5 u.g/ml of MBP failed to causealterations in basehneRGC release( - 0.6% k 8.7%; n = 3). In three experiments, 50 pg/ml oi MBP caused a significant decreasein RGC release ( - 21.7% ? 2%) after 1 hour of incubation, but in the subsequentl-hour period after MBP was-removed

VOLUME NUMBER

Eosinophils

87 3

and airway

mucus secretion

695

TIME COURSE EFFECT OF ECP ON FELINE TRACHEAL RGC RELEASE

HOURS OF INCUBATION FIG. 5. Time course of ECP (2.5 pg/ml) on RGC release from feline tracheal explants. ECP was incubated for %, 1, 2, and 4 hours; n = 6 for all time points except for the l-hour time point where n = 5.

TABLE 1. Effect of ECP on RGC and lactoferrin ECP (pglmll

2.5 0.25 0.025

release from human

RGC release (% change from control)

+ 18.2 +- 3.4 +2.2 k 2.0 +6.3 iz 2.5

Lactoferrin release (% change from control)

P

co.05 ns

ns

bronchi

-l-49.9 2 13.0 + 19.4 t 2.6 + 13.5 + 5.5

P

0.05).

DISCUSSION EOs and their secretory products have been proposed to play a pathogenetic role in asthma because

of their ciliostatic, cytotoxic, and histamine-releasing abilities.26 The highly basic EO granule protein, MBP, has been implicated in these respects. However, another granule protein, ECP, appears to enhance RGC and lactoferrin release from feline and human airway organ cultures, whereas MBP inhibits release of RGC from feline tracheal cultures. Thus, ECP and MBP may have counterbalancing roles in influencing the homeostasis of the airways. The RGC stimulator-y response observed after incubation with EO lysates, EO granule extracts, and

FIG. 6. lmmunohistochemical localization of lactoferrin (LF) in human bronchial mucosa. LF was identified as the black precipitate in serous cells of submucosal glands (upper insert, indicated by arrow). The slides were counterstained with alcian blue, which stains mucous glycoprotein containing mucus cells and goblet cells. These cells appear gray (see control (C), lower insert). Therefore, the LF immunoreactive serous cells are distinct from the alcian blue-staining mucus cells. No epithelial cells possessed LF immunoreactive material. The addition of exogenous LF to the anti-LF serum ablated the binding (C, lower insert). The bar represents 100 km.

purified ECP and MBP were probably not caused by a cytotoxicity, since the changes in RGC releasing effect persisted in cultures challenged on subsequent days. Furthermore, cultures previously challenged with EO products were as responsive as unstimulated cultures to methacholine. In addition, previous studies have suggested that the toxic effect of the EO granule proteins on airway epithelium is not apparent until several hours after incubation with the epithelium,” whereas the effects we were studying occurred within the first hour of incubation. Last, the highest concentration of ECP used in our dose-response experiments of the purified protein failed to cause cell exfoliation or ciliostasis of guinea pig tracheal epithelial cells after incubation as long as 24 hours. ‘* Thus, the effects of ECP and MBP on RGC release in feline airways most likely are nontoxic.

The mucus stimulatory effect of LOS rnc~ l~keiv resides in the granules, since the extract of whoh~ ly~ti EOs and extract from purified EO granule> >tintuiat~c.l RGC release. Approximately the same RCii: .~umulatory effect was observed after incubating- lvbalc i ~OIV ” 0.1 X 10” whole EOs (Fig. I) and granule iysate from 2.5 x 1O’EOs (Fig. 2). This tinding mohi lrkel> is due to loss of granules or solubilized granu!~: proteins during purification. KY. Evidence that the fractions believed to COJ~UI EDN, and MBP actually do contain these proteins falls along two lines. The methods used to separarc the EO granule proteins in this study have previously been described in detail. “‘K’ Furthermore, the identity of the granule proteins was verified by gel electrophoresis and confirmed by selective immunoreactivity against murine monoclonal antibodies prepared against each of the four human EO granulc proteins, and this was confirmed with the Western blot technique. Amino acid sequence analysis ha> rccently been reported for some of the EO granule proteins.?‘.” Since increasing quantities of ECP and MBP become available through recombinant DNA methodology, it should become increasingly less c~umbersome to test these observations in other modei systems. E2 included EPO as well as a fraction ot the E3 shoulder, that is, ECPIEDN. That E2 (Fig. 3) stimulated RGC release can be interpreted in two ways. Either EPO stimulated RGC release or it was the content of ECP in E2 that caused the increase in RGC release. Thus, a role of EPO in causing mucus release can not be determined in the experiments included in this article. The evidence supporting the suggestion that ECP stimulates mucus-producing cells also falls along several lines. First, the secretory activity appeared to localize to the granules. Second. heparin (MW, 15,000) with its acidic pl, partly blocked the RGCreleasing effect of EO granule extracts, presumably by binding to molecules with a basic pl (e.g., ECP). Conversely, ECP has previously been demonstrated to neutralize the anticoagulant activity of heparin.‘” Third, the E3 fraction of sequentially eluted EO granule proteins (which contains ECP and EDN) caused a significantly larger increase in RGC release than E2 that contains EPO. Fourth, ECP causes release of the nonspecific marker of mucus, RGC, from both feline and human airway explant cultures. The effect of ECP in human bronchial explants was approximately half that observed in feline tracheal explants, possibly because of differences in penetration of ECP to the active site in the tissue or because of differences in the amounts of proteolytic enzymes capable of degrading ECP. Finally, ECP stimulates the release of iactoferrin

VOLUME87 NUMBER3

from human bronchial explants, a specific marker of serous cell secretion. The time course of ECP-stimulated RGC release from feline tracheal explants follows a similar time course, as observed in response to methacholine.29 In addition to submucosal serous cells, the identification of which other cell types are stimulated by ECP is not known. ELISAs with monoclonal antibodies directed against various epitopes from the different mucus-producing cells may eventually become useful in this respect. The specificity of the secretagogue response to ECP was examined by the experiments with EDN, which has a 50% amino acid homology with ECP.2’ EDN and ECP have approximately the same pI and differ only slightly in MW. ‘9-2’.26EDN failed to alter baseline RGC release even when EDN was used at twice the highest concentration of ECP. MBP is also a basic protein, but it actually inhibited RGC release. Thus, the stimulatory effect of ECP cannot be attributed to electrostatic interactions between this granule protein and the very acidic RGC. A direct role of the EO on mucus production has not previously been suggested. In particular, the mucus secretagogue action of ECP has not previously been observed. Both ECP and MBP cause the release of histamine from rat peritoneal mast cells” and are toxic to guinea pig tracheal epithelium. ” We found that ECP and MBP had opposing effects on RGC release. Thus, the net effect of EO granule proteins on mucus release may depend on the relative quantity of proteins released or the relative reactivity of the secretory cells. These observations provide a new facet to the appreciation of the potential role of EOs in asthma, also termed by some as “chronic eosinophilic bronchitis.“’ Mucus hypersecretion is believed to play a major role in airway obstruction observed in this disease entity’ and may be the major cause of death. It appears that EOs may play an important, and previously unsuspected, role in causing mucus hypersecretion in asthma since (1) on activation, EOs produce lipid mediators such as eicosanoids’4 and platelet-activating factor,15both of which cause RGC release in feline and human explants,16. ” and (2) EOs may cause airwaysurface epithelial shedding,13, 26 possibly leading to easier access of inhaled allergens to the mucosal mast cells, causing, when these are activated, increased RGC release.* Combination of mast cell mediators and EO products may also activate sensory nerves leading to axon reflex-mediated stimulation of submucosal glands through release of neuropeptides,30 such as the tachykinins.3’- Q As suggested in this article, ECP, a secretory product of activated EOs, directly stimulates the release of submucosal gland products.

Eosinophils

and airway

mucus secretion

697

The kind provision of blood from patients under the care of Drs. A. Leavitt and A. Fauci, from the National Institute of Allergy and Infectious Diseases,and Drs. M. Lotze and S. Rosenberg,from the Surgical Branch, National Cancer Institute, is gratefully appreciated. Dr. Lundgren was sponsoredin part by the Danish Medical ResearchCouncil, a travel grant from Draco, and the Danish National Committeeon the Preventionof Lung Diseases. Dr. Baraniuk was sponsoredby a grant from the Procter and Gamble Co., Inc. Dr. Mull01 was supported in part by a grant from the Glaxo Co. REFERENCES I Lundgren JD, Shelhamer JH. Pathogenesis of mucus hypersecretion. J ALLERGY CLIN IMMUNOL 1990;85:399-417. 2. Shelhamer JH, Marom Z, Kaliner M. Immunological and neuropharmacologic stimulation of mucous glycoprotein release from human airways in vitro. J Clin Invest 1980;66: 1400-8. 3. Logun C, Rieves RD, Lundgren JD, Marom 2, Kaliner M, Shelhamer JH. Activated human neutrophils release a high molecular weight protein which stimulates respiratory glycoconjugate release from human airways in vitro. Am Rev Respir Dis 1988;137(4 p2):A14. 4. Marom Z, Shelhamer JH, Kaliner M. Human pulmonary macrophage-derived mucus secretagogue. J Exp Med 1984; 159:844-60. 5. Barnes PJ. New concepts in the pathogenesis of bronchial hyperresponsiveness and asthma. J ALLERGY CLIN IMMUNOL 1989;83:1013-26. 6. Kaliner M, McFadden ER. Bronchial asthma. In: Samter M, ed. Immunological diseases. 4th ed. Boston/Toronto: Little, Brown, 1988:1067-l 118. 7. Hutson PA, Church MK, Clay TP, Miller P, Holgate ST. Early and late-phase bronchoconstriction after allergen challenge of nonanesthetized guinea pigs. Am Rev Respir Dis 1988;37:54857. 8. de Monchy JGRD, Kauffman HF, Venge P, et al. Bronchoalveolar eosinophilia during allergen-induced late asthmatic reactions. Am Rev Respir Dis 1985;131:373-6. 9. Dor PJ, Ackerman SJ, Gleich GJ. Charcot-Leyden crystals and eosinophil granule major basic protein in sputum of patients with respiratory diseases. Am Rev Respir Dis 1984;130: 1072-7. 10. Wardlaw AJ, Dunnette S, Gleich GJ, Collins JV, Kay AB. Eosinophils and mast cells in bronchoalveolar lavage in subjects with mild asthma. Am Rev Respir Dis 1988;137:62-9. 11 Zheutlin LM, Ackerman SJ, Gleich GJ, Thomas LL. Stimulation of basophil and rat mast cell histamine release by eosinophil granule-derived cationic proteins. J Immunol 1984: 133:2180-5. 12 Motojima S, Frigas E, Loegering DA, Gleich GJ. Toxicity of eosinophil proteins for guinea pig tracheal epithelium in vitro. Am Rev Respir Dis 1989;139:801-5. 13. Laitinen LA, Heins M, Laitinen A, Kawa T. Hoehtela T. Damage of the airway epithelium and bronchial reactivity in patients with asthma. Am Rev Respir Dis 1985; 13 1:599-606. 14. Henderson WR, Harley JB, Fauci AS. Arachidonic acid metabolism in normal and hypereosinophilic syndrome human eosinophils: generation of leukotriene B,, C,, and 15lipoxygenase products. Immunology 1984;5 1:679-86. 1.5. Lee T, Lenihan DJ, Malone B, Roddy LL, Wasserman SI. Increased biosynthesis of platelet-activating factor in activated human eosinophils. J Biol Chem 1984;259:5526-30.

696

Lundgren

et al.

16. Marom Z, Shelhamer JH, Kaliner M. The effect of arachidonic acid, monohydroxyeicosatetraenoic acid, and prostaglandins on the release of mucous glycoproteins from human airways in vitro. J Clin Invest 1981;67:1695-1702. 17. Lundgren JD, Kaliner M, Logun C, Shelhamer JH. Plateletactivating factor and tracheobronchial respiratory glycoc&tjugate release in feline and human explants: involvement of the lipoxygenase pathway. Agents Actions 1990;30:329-37. 18. Vadas MA, David JR, Butterworth A, Pisani NT, Siongok TA. A new method for the purification of human eosinophils and neutrophils, and a comparison of the ability of these cells to damage schistosomula of Schisrosoma mansoni. J Immunol 1979;122:1228-36. 19. Ackerman SJ, Loegering DA, Venge P, et al. Distinctive cationic proteins of the human eosinophil granule: major basic protein, eosinophil cationic protein, and eosinophil-derived neurotoxin. J Immunol 1983;131:2977-86. 20. Ackerman SJ, Durack DT, Gleich GJ. The Gordon phenomenon revisited: purification, localization and biochemical characterization of the human eosinophil-derived neurotoxin. In: Yoshida T, Torisu M, eds. Immunobiology of the eosinophil. New York: Elsevier Biomedical, 1983:181-99. 21. Gleich GJ, Loegering DA, Bell MP, Checkel JL, Ackerman SJ, McKean DJ. Biochemical and functional similarities between human eosinophil-derived neurotoxin and eosinophil cationic protein: homology with ribonuclease. Proc Nat1 Acad Sci USA 1986;83:3146-50. 22. Lundgren JD, Hirata F, Marom 2, et al. Dexamethasone inhibits respiratory glycoconjugate secretion from feline airways in vitro by the induction of lipocortin (lipomodulin) synthesis. Am Rev Respir Dis 1988;137:353-7.

J. ALLERGY

CLIN. IMMUNOL. MARCH 1991

23. Raphael GD, Druce HM, Baraniuk JN, Kaliner M. Pathophysiology of rhinitis. I. Assessment of the sources of protein in methacholine-induced nasal secretions. Am Rev Respir Dis 1987;138:413-20. 24. Gleich GJ, Adolphson CR. The eosinophil leukocyte: structure and function. Adv Immunol 1986;39:177-253. 25. Carlson MGCH, Peterson GB, Venge P. Human eosinophil peroxidase: purification and characterization. J Immunol 1985; 134: 1875-9. 26. Frigas E, Gleich GJ. The eosinophil and the pathophysiology of asthma. J ALLERGYCL~ IMMUNOL1986;77:527-35. 27. Wasmoen TL, Bell MP, Loegering DA, Gleich GJ, Prendengast FG, McKean DJ. Biochemical and amino acid sequence analysis of human eosinophil granule major basic protein. J Biol Chem 1988;263:12559-63. 28. Venge P, Dahl R, Ha&ten R, Olsson I. Cationic proteins of human eosinophils and their role in the inflammatory reactions. In: AAF Mahmound, Austen KF, eds. The eosinophil in health and disease, New York: Grune & Stratton, 1980:131-44. 29. Lundgren JD, Baraniuk JN, Ostrowski NL, Kaliner MA, Shelhamer JH. Gastrin-releasing peptide stimulates glycoconjugate release from feline trachea. Am J Physiol 1990;258:L68-L74. 30. Barnes PJ. Asthma as an axon reflex. Lancet 1986;1:242-4. 31. Coles SJ, Neil KH, Reid LM. Potent stimulation of glycoprotein secretion in canine trachea by substance P. J Appl Physiol 1984;57:1323-7. 32. Lundgren JD, Wiederman CJ, Logun C, Plutchok I, Kaliner M, Shelhamer JH. Substance P receptor-mediated secretion of respiratory glycoconjugate from feline airways in vitro. Exp Lung Res 1989;15:17-29. .‘;