Selective Eosinophil Leukocyte Recruitment by Transendothelial Migration and Not by Leukocyte-Endothelial Cell Adhesion Clare M. Morland, Susan J. Wilson, Stephen T. Holgate, and William R. Roche Departments of Pathology and Medicine I, University of Southampton, Southampton, United Kingdom

Eosinophil infiltration is the hallmark of allergic inflammatory events. However, the mechanisms governing the influx of eosinophils into the tissue at a site of an allergic reaction remains unclear. We have examined the interactions of eosinophils and neutrophils isolated from the same atopic donor with cultured human umbilical vein endothelial cell (EC) monolayers in the search for a mechanism for this selective eosinophil recruitment. First, the adherence of eosinophils and neutrophils to ECs stimulated with lipopolysaccharide, interleukin (IL)-Ia, and tumor necrosis factor-a were compared. Each mediator induced a similar dosedependent enhancement of eosinophil adhesiveness for both eosinophils and neutrophils. Thus, although cytokine activation of ECs in the vasculature adjacent to an inflammatory site probably serves as an important focusing mechanism for the extravasation of inflammatory cells at this site, there does not appear to be any selective EC-dependent mechanism for eosinophil recruitment. Little or no effect on eosinophil and neutrophil adherence was observed with IL-3, IL-5, granulocyte/macrophage colony-stimulating factor, platelet-activating factor (PAF), leukotriene B4 , or histamine. Second, the migration of eosinophils and neutrophils through an EC monolayer in response to chemoattractants was examined. PAF was found to selectively enhance eosinophil transendothelial migration at doses of 10-7 to 10- 10 M, with optimal effect at 10-8 M. This effect was gradient dependent and could be inhibited by WEB 2086, a specific PAF inhibitor. These results suggest that localized production of PAF may be a prime factor in the events leading to eosinophil accumulation at allergic inflammatory sites, and that selectivity for eosinophil recruitment occurs at the stage of transendothelial cell migration under the influence of cell-specific chemoattractants.

Understanding of the pathogenesis of allergic diseases has advanced from the simple model of repeated mast cell degranulation to embrace the concept of clinical events occurring on a background of allergen-mediated inflammation (1). Eosinophil leukocytes are regarded as important effector cells in the inflammation and tissue damage of allergic disease (2, 3). Eosinophil infiltration of the bronchial mucosa is a characteristic histologic feature of asthma (4), and eosinophil levels in blood, sputum, and bronchial mucosa have been found to correlate. with indices of disease severity (5-7).

Late-phase responses to allergen are a useful model of

(Received in original form July 11, 1991 and in revised form October 16,

1991) Address correspondence to: Dr. Clare M. Morland, Department of Pathology, Level E, South Block, Southampton General Hospital, Tremona Road, Southampton S09 4XY, United Kingdom. Abbreviations: 3-amino-l,2,4-triazole, AMT; endothelial cell, EC; electron microscopy, EM; fetal calf serum, FCS; formylmethyionylleucylphenylalanine, FMLP; granulocyte/macrophage colony-stimulating factor, GMCSF; Hanks' balanced salt solution, HBSS; interleukin, IL; lipopolysaccharide, LPS; leukotriene B4 , LTB4 ; orthophenylenediamine, OPD; phosphate-buffered saline, PBS; PBS containing 0.1% gelatin and 30 mg/ml DNase, PGD; polymyxin B sulphate, PMBS; tumor necrosis factor, TNF. Am. J. Respir. Cell Mol. BioI. Vol. 6. pp, 557-566, 1992

persisting airway obstruction in asthma and are accompanied by an influx of eosinophils into the bronchoalveolar lumen (8, 9). Degranulation of eosinophils causes release of cytotoxic proteins (10), including major basic protein, which has been localized to damaged respiratory epithelium in fatal asthma (11). Major basic protein has been shown to damage bronchial epithelium in vitro (12), and intratracheal instillation has been shown to induce bronchial hyperreactivity in primates (13). Despite this mounting evidence for the effector role of the eosinophil in allergic disease, the mechanisms governing the selective recruitment of eosinophils to sites of allergic reactions remain unclear. The influx of inflammatory cells to the site of a local inflammatory reaction in response to locally generated signals involves a sequence of events: margination along the walls of the microvasculature, adhesion to the endothelium, emigration through the vessel walls, and migration along a chemotactic gradient within the extravascular compartment (14). A crucial stage in this process is the attachment of the.leukocytes to the vascular endothelium. The development of in vitro assay systems using cultured vascular endothelial cells (ECs) has enabled the study of mechanisms involved at this adherence stage. After stimulation with tumor necrosis factor (TNF)-a and interleukin (IL)-Ia, vascular ECs have been shown to become more adhesive for both eosinophils (15) and neutro-

558

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992

phils (16, 17). This cytokine-induced increase in EC adhesiveness is due to the increased or novel expression of EC surface adhesion molecules. ECs are activated by cytokines to express the de novo adhesion receptor ELAM-1 (18, 19), which interacts with the sialyl Lewis X antigen on neutrophils (20,21). Cytokines also upregulate basal expression on ECs ofICAM-1 (22), which interacts with the leukocyte integrin molecule CDlla/CD18 on neutrophils (23) and may be an endothelial ligand for neutrophil CDllb/CD18 (24, 25). Both ELAM-1 and ICAM-1 appear to be involved in the adherence of eosinophils and neutrophils to cytokine-stimulated ECs (26). Recently, VCAM-1, another cytokine-inducible EC ligand, thought to be monocyte-specific, has also been shown to bind eosinophils and basophils but not to neutrophils (27). It has been suggested that this provides a possible mechanism for selective eosinophil adherence but coexpression of other EC adhesion molecules makes selective eosinophil binding unlikely. Equally, IL-5 has been shown to selectively activate peripheral blood eosinophils for EC adhesion (28), but this is unlikely to explain the localization of eosinophils to defined anatomic sites, such as the bronchial mucosa. Thus, the mechanism of the selective tissue eosinophilia in asthma remains unclear. In this report, we have performed comparative studies of the adherence of eosinophils and neutrophils to ECs stimulated by a number of cytokines, and their migration through the endothelium, in order to determine the method of selective recruitment of eosinophils. We have developed a sensitive assay to quantify adherent and migrated eosinophils and neutrophils by the measurement of their intracellular peroxidase content. The differential action of 3-amino-1,2,4-triazole (AMT), previously reported to inhibit purified eosinophil peroxidase activity with little effect on purified myeloperoxidase (29), was used to discriminate between the two types of granulocytes.

Materials and Methods Reagents Lipopolysaccharide (LPS) , Escherichia coli, chromatographically purified, leukotriene B4 (LTB4 ) , and histamine were obtained from Sigma Chemical Co. (Poole, Dorset, UK). Recombinant human IL-1a, recombinant human TNF-a, recombinant human IL-3, and recombinant human granulocyte/macrophage colony-stimulating factor (GMCSF) were purchased from British Biotechnology (Oxford, UK). Platelet-activating factor (PAF) C16 and C18 were obtained from Bachem (Saffron Walden, Essex, UK). Recombinant human IL-5 (affinity-purified) was generously provided by Dr. 1. Tavernier (Roche Research, Ghent, Belgium). WEB 2086 was kindly provided by Boehringer Ingelheim Ltd. Endotoxin Testing of Reagents All media, reagents, and the fetal calf serum (FCS) batch used were screened for low endotoxin levels « 0.1 ng/ml) using the sensitive Toxicolor assay kit (Seikagaku Kogyo Co. Ltd., Tokyo, Japan). All stock solutions of cytokines were also assessed for endotoxin levels, and IL-1 and TNF-a were found to contain < 0.1 ng/ml endotoxin, giving negligible levels at working concentrations. Any glassware used was

baked for 2 hat 180° C to eliminate endotoxin contamination. The effect of polymyxin B sulphate (PMBS) (Sigma) on adherence was examined as a control for endotoxin contamination in the adherence assay. Human EC Culture ECs were isolated from human umbilical cord veins by digestion with 0.1% collagenase (type IV; Sigma) in Dulbecco's phosphate-buffered saline (PBS) (GIBCO, Paisley, Scotland) for 15 min at 37° C, according to Jaffe (30). The isolated cells were grown to confluence in RPMI-1640 (Flow Laboratories, Irvine, Scotland) supplemented with 20 % FCS (GIBCO) in 25-cm 2 tissue culture flasks (GIBCO) in a 37° C, 5 % CO2 humidified incubator. Confluent cells were passaged by treatment with 0.05% trypsin (type II; Sigma) in Dulbecco's PBS containing 0.02% EDTA (BDH, Poole, Dorset, UK) for 5 min at 37° C. Passaged cells were maintained in RPMI containing 15% FCS and EC growth supplement (100 J-tg/ml; Flow Laboratories). ECs up to second passage were used for all experiments. ECs were identified by their typical cobblestone morphology and the presence of Factor VIII antigen as detected by indirect immunofluorescence. Final mono layers for adhesion assays were prepared by seeding 96-well tissue culture plates (Nunc; GIBCO) with 2 X 1Q4 ECs/well, followed by incubation for 24 to 48 h to ensure confluence. All culture surfaces were coated with fibronectin (10 J-tg/ml; GIBCO) to enhance adhesion and growth of the ECs. Purification of Neutrophils and Eosinophils Peripheral blood neutrophils and eosinophils were isolated from the whole blood of mildly atopic volunteers by the method of Vadas and associates (31). None of the donors were currently receiving corticosteroid treatment. A total of 50 ml of blood was collected into preservative-free sodium heparin (20 U/ml blood; Weddel Pharmaceuticals, Wrexham, UK) and erythrocytes were removed by dextran sedimentation (l vol dextraven 110 [CP Pharmaceuticals, Wrexham, UK], 6% wt/vol in 0.9% saline: 2 vol blood, 45 min, 20° C). The upper leukocyte-rich buffy coat layer was removed and centrifuged (400 x g, 5 min). The cell pellet was resuspended in Tris-ammonium chloride (pH 7.2) for 10 min to lyse any remaining red blood cells, then the leukocytes were washed once in Hanks' balanced salt solution (HBSS) and resuspended to a concentration of 1 X lOs cells/ml in PBS containing 0.1% gelatin and 30 mg/ml DNase (PGD). Gradient solutions of 18, 20, 22, 23, and 24% metrizamide (Nycomed, Birmingham, UK) were diluted from a freshly prepared 30% stock solution in PGD buffer, and gradients were prepared by carefully layering 2-ml volumes of decreasing density into l l-ml conical centrifuge tubes (Sterilin, Feltham, Middlesex, UK). One-milliliter aliquots of the leukocyte suspension were layered onto the gradients, and cells were separated by centrifugation at 1,200 X g for 45 min at 20° C. The band of cells at each interface was then carefully aspirated and washed 2 times with HBSS. Eosinophils were harvested from the lower cell bands (22 to 23 % and 23 to 24 % interfaces) and neutrophils from the 20 to 22 % interface. Eosinophil and neutrophil purity were found to be 84.0 ± 6.6% and 89.0 ± 7.1 %, respectively, as determined by

Morland, Wilson, Holgate et al.: PAF-induced Selective Eosinophil Transendothelial Migration

differential cell counts according to Kimura and colleagues (32). Viability of both cell preparations was> 98 % by trypan blue exclusion. Adherence Assay The ability of various stimuli to induce adhesiveness in ECs was determined as follows. Confluent monolayers of ECs, as determined by phase-contrast microscopy, were gently washed once with 200 J.tl of assay buffer (RPMI-5 % FCS) followed by preincubation with 100-J.tl aliquots of test stimuli or assay buffer at 37° C for varying time periods (typically 4 h proved optimal). Six replicate wells were used for each experimental variable. Test stimuli were then removed by two further washes, and 50-J.tl aliquots of neutrophils (1 x lQ6 cells/ml in HBSS) or eosinophils (2 x 105 cell/ml) were added to each well and incubated for 30 min at 37° C. After incubation, the supernatants containing nonadherent cells were gently aspirated, cleared of cells by centrifugation, and saved for peroxidase analysis. Any further nonadherent cells were removed by gentle washing 3 times with PBS at 20° C. After removal of nonadherent cells, the EC monolayers were again checked by phase-contrast microscopy to confirm leukocyte-EC adhesion and the integrity of the EC monolayers. Intracellular peroxidase content was used to quantify adherent granulocytes. After lysis of the adherent cells with 100 J.tl of 1% non-ionic detergent NP40 (Sigma) per well, peroxidase activity was measured using a spectrophotometric assay. The chromogenic peroxidase substrate orthophenylenediamine (OPD) (0.4 mg/ml; Sigma) was freshly prepared in phosphate-citrate buffer, pH 5.0, containing 1.2% vol/vol of hydrogen peroxide. One hundred-milliliter aliquots of OPD alone or OPD containing 10 mM AMT (see RESULTS) were added to wells containing lysed adherent eosinophils or neutrophils, respectively. After incubation for 30 min at 37° C, the color reaction was stopped by addition of 50 J.tl of 2.5 M H2S04/well. Absorbance was measured at 490 nm on a plate-reading spectrophotometer (Titertek Multiscan; Flow Laboratories). Background levels of absorbance determined on wells containing NP40, 0 PD, and H2S04 were subtracted from all absorbance values. The peroxidase content of the supernatant was also determined to establish whether peroxidase was released from the granulocytes during the adhesion assay. Percentage adherence and percentage peroxidase release were calculated by comparison with peroxidase content of the purified granulocyte suspensions. Transendothelial Cell Migration The upper chamber polycarbonate filter inserts (3-J.tm pore size) of Transwell plates (Northumbria Biologicals Ltd., Northumberland, UK) were fibronectin-coated and seeded with 2 X 104 ECs and cultured for 24 h to ensure confluence. These inserts fit into the wells of a 24-well tissue culture plate to produce an upper chamber separated by a polarized EC layer from a lower chamber into which chemoattractive agents can be directly introduced, abluminal to the ECs. The EC monolayers were gently washed once with assay medium (RPMI-5 % FCS), and the medium was replaced with 100 J.tl of eosinophil or neutrophil suspensions

559

(105 cells/well). Potential chemotactic agents were then added to the lower chambers such that medium levels in the two chambers of each well were equal. The transwells were incubated for 1 hat 37° C, then 25 J.tl of 2.5 M EDTA was added to the lower chambers to remove any migrated cells from the lower surface of the filters, and the upper chamber inserts were removed from each well. Migrated cells in the lower chambers were deposited onto the well surface by centrifugation (10 min, 1,000 x g) and the supernatants removed. The cells were lysed by addition of 1% NP40 and quantified by measurement of their intracellular peroxidase. All mediators were tested in triplicate, and the data presented are representative of five experiments using different granulocyte donors and EC preparations. Electron Microscopy (EM) Techniques For transmission EM, the EC-coated filters were removed after the transmigration of eosinophils or neutrophils through the monolayers and fixed by immersion in 2.5 % glutaraldehyde in 0.1 M sodium cacodylate at pH 7.2..The filters were then washed in the sodium cacodylate buffer,postfixed with 1.0% osmium tetroxide, stained en bloc with 2.5% uranyl acetate, dehydrated, and embedded in Spurr resin. Sections perpendicular to the ECs were cut at 1 J.tm and stained with toluidine blue for orientation, and then 90-nm sections were cut and stained with lead citrate. Sections were examined using a Hitachi H7000 microscope. For scanning EM, the whole filters were fixed and postfixed as for transmission EM, dehydrated, and then critical point dried, mounted on aluminium stubs, and splutter-coated with gold. Statistics One-way ANOVA and the Student's t test were used to compare stimulated granulocyte-EC adherence to unstimulated baseline levels and for transmigration results.

Results Quantitation of Adherent Eosinophils and Neutrophils A calibration curve was constructed relating peroxidase activity, as measured by optical density, to cell number using graded numbers of cells determined visually, to verify the measurement of cellular peroxidase activity as an index of granulocyte numbers. High-purity eosinophil (95 %) and neutrophil (98 %) preparations from hypereosinophilic and normal donors, respectively, were used. Within the ranges indicated, highly significant linear correlations were obtained (r = 0.98, P < 0.001 for eosinophils; r = 0.986, P < 0.001 for neutrophils) (Figure 1, upper panel). Eosinophils had approximately 3 times greater peroxidase activity than did neutrophils on a cell-for-cell basis. The peroxidase activity of the supernatants collected after the adherence incubation period was determined in order to establish whether there was release of peroxidase from the granulocytes during their adhesion to the ECs. Table 1 shows that for the mediators used in this assay, low levels of peroxidase were detectable in the supernatants removed after adhesion and depleted of nonadherent cells by centrifugation. Adherence to ECs did not result in peroxidase release greater than background levels of spontaneous release.

560

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992

TABLE 1

3.5

Percentage of peroxidase released*

Ec: 0 0> ...,

3.0 EC Pretreatment

2.5

Unstimulated LPS treated (0.1 fJ.g/ml) IL-I treated (I U/ml) TNF treated (l U/ml) Spontaneous release

II

~ 2.0 Q)

o c

1.5

CQ

.c "-

0

1.0

II)

.c

«

0.5 0 0

0.5

1.25

2.5

Cell Number

5

3.75

Ix 10 4 cells)

Eosinophils

Neutrophils

(%)

(%)

1.95 4.6 3.9 8.7 4.85

o 3.6 3.3 7.0 12.5

Definition of abbreviations: EC = endothelial cell; LPS = lipopolysaccharide; IL-I = interleukin-I; TNF = tumor necrosis factor. * ECs were pretreated for 4 h with LPS, IL-l, TNF, or assay medium. After removal of the stimuli, the ECs were incubated for 30 min with eosinophils or neutrophils. Spontaneous peroxidase release refers to aliquots of eosinophil or neutrophil suspensions incubated for 30 min in wells containing no ECs. Results are presented as the percentage of the total peroxidase content of cells added. Means of quadruplicate determinations are shown.

100

80

----~I

Q)

o

c: CQ

.c

60

"-

0

en

.c

«

40

~

~~~-----... Eosinophils

20

.1.

AMT, in all neutrophil quantifications. To optimize the concentration of AMT required for neutrophil quantification, the peroxidase activities of lysed preparations of pure eosinophils (95 % purity, 2 x l()5 cells/ml) and neutrophils (98% purity, 106 cells/ml) were determined using substrate solutions containing different concentrations of AMT. AMT selectively inhibited eosinophil-derived peroxidase in a dose-dependent manner (Figure 1, lower panel), with> 80% inhibition in eosinophil peroxidase activity and < 20% inhibition of neutrophil peroxidase at 10 mM. This concentration of AMT was then included in all measurements of adherent neutrophil numbers.

O-+--~-~---.,..-----------,

o

2.5

5.0

10

20

AMT (mM) Figure 1. Upperpanel: Standard curve relating peroxidase activity to cell number. A linear relationship is seen for both neutrophils (r = 0.986, P < 0.001) and eosinophils (r = 0.98, P < 0.001). Means ± SD of six replicates are plotted (SD bars omitted if too small). Lower panel: The differential effect of 3-amino-l,2,4-triazole (AMT) on the activity of intracellular neutrophil and eosinophil peroxidase, as indicated by percent residual peroxidase activity plotted against concentrations of AMT. Means ± SD of six replicates are plotted.

Use of AMT to Discriminate Eosinophil Peroxidase from Neutrophil Peroxidase Previous studies of eosinophil and neutrophil adherence to ECs have relied upon radiolabeling of cell suspensions isolated from different donors (atopic and normal, respectively). For this comparative study, we have used eosinophils and neutrophils isolated from the same atopic donor. Neutrophil preparations obtained from metrizamide gradient separation of mildly hypereosinophilic, atopic subjects were contaminated with hypodense eosinophils to varying degrees. Because eosinophil peroxidase has much greater activity than neutrophil peroxidase (29), quantification of neutrophils by measurement of their peroxidase content would be significantly influenced by even a small « 10%) contamination of eosinophils. This problem has been resolved by the inclusion of the specific eosinophil peroxidase inhibitor,

Adherence of Eosinophils and Neutrophils to ECs The ability of LPS, IL-1a, TNF-a, IL-3, IL-5, and GM-CSF to stimulate EC adhesiveness for eosinophils and neutrophils was compared. The BCs were pretreated for 4 h with each mediator and washed twice before addition of granulocytes for 30 min. LPS, IL-1, and TNF-a were found to increase the adhesiveness of ECs for both eosinophils and neutrophils, with similar dose-related response curves for each cell population (Figure 2). The addition of PMBS to the ECs immediately before the addition of mediators completely blocked the LPS-induced increase in adherence of both neutrophils and eosinophils (P < 0.001, two-way ANOVA, data not shown), whereas IL-1- and TNF-stimulated adherence was unaffected, indicating that their effects were not due to endotoxin contamination. The kinetics of enhanced EC adhesiveness for eosinophils was compared with that for neutrophils, and similar time courses were observed with each stimulus (Figure 3). A maximal EC response to LPS, IL-l, and TNF was observed at 6 h, with decreased levels by 24 h. The kinetics of adherence of eosinophils and neutrophils to ECs were also examined, and both cell types showed similar increases in adherence with time of incubation (data not shown). IL-3, IL-5, and GM-CSF pretreatment ofECs for 4 h produced negligible increases in EC adhesiveness, without any preferential enhancement of eosinophil or neutrophil adhesion (Table 2). Similarly, preincubation of PAF, LTB4 , and histamine with ECs for 5 to 15 min or for 4 h, followed after washing by a 30-min adherence reaction with eosinophils

Neutrophils

90

60

80 Q)

70

50

c:

(1)

40

60

Q;

50

~ 30

40

~ 20

30

10

Q) ~

Q)

.s:

(1)

"'0

«

~EOS

o

oc

.s:

LPS

70

Eosinophils

Nphils

~

20

O+----,----.---,----,---~--ll---l

o

10

o

10 lD

.1 Dl 001.0001

o

10 lD

.1

0.5

2

4

6

24

Time (Hours)

Dl 001.0001

40

LPS (ug/ml) Neutrophils

100

Eosinophils

30

90

Q)

o

c: Q)

80

~

] Q,)

o

c

70

...

60

.c.

50

Q,)

«

~

(1)

"'0

-c ~

20

"tl

10

Nphils

40

O-t---,...----r----r-----r---...,.---tl---l

30

o

0.5

20

2

4

6

24

Time (Hours)

10

o

100 10

1.0 0.1

.01

o

100 10

60

IL-1 (U/ml) Neutrophils

100

TNF a

70

1.0 0.1 .01

Eosinophils

Q)

c:

... Q)

90

50

o

40

Q)

.c:

(1)

o

c

80

~ 30

70

~ 20

...

60

.s:

50

Q,)

10

(1)

"'0

«

~

O~--..------r-----r--""T'"""--....-----ll---l

40

o

0.5

2

4

6

24

Time (Hours)

30

20

10

o

100 10

1.0 0.1 .01

o 100 10 1.0 0.1 .01

TNF-a Iu/mll Figure 2. Enhancement of endothelial cell (EC) adhesiveness for eosinophils and neutrophils. ECs were incubated with lipopolysaccharide (LPS) (upper panel) recombinant human interleukin-l (rhIL-l) (middle panel), and recombinant human tumor necrosis factor-a (rhTNF-a) (lower panel) for 4 h, then washed, and eosinophils/neutrophils were allowed to adhere for 30 min. Means + SD of six replicates are shown for one of six experiments. The solid bars indicate adherence levels for unstimulated control ECs, and

Figure 3. Kinetics of LPS-, IL-l-, and TNF-stimulated EC adhesiveness for eosinophils and neutrophils. ECs were incubated with LPS (0.1 ILg/ml) (upper panel), IL-l (1 U/ml) (middle panel), and TNF (l U/ml) (lower panel) for the time periods indicated, then washed 2 times before incubation of eosinophils and neutrophils for 30 min. Values represent mean ± SD of four observations in one representative experiment.

the hatched bars indicate mediator-stimulated adherence above control levels. Significance values for stimulated adherence compared with unstimulated controls are *** P < 0.001, ** P < 0.01, and * P < 0.05.

562

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992

500

TABLE 2

Influence of LTB 4 , PAF, histamine, IL-3, IL-5, and GM-CSF on Ee adhesiveness for neutrophils and eosinophils* Eosinophils

Neutrophils

(%)

(%)

EC Stimulus

15 min Untreated LTB4 (10-8 M) PAF (10- 8 M) Histamine (10- 8 M) 4h Untreated IL-3 (1 ng/ml) IL-5 (1 U/ml) GM-CSF (10 U/ml)

c: 0

';:;

~

400

0)

1*

~

:§

300

**

Q)

.c

100 83.7 94.1 92.2

± 9.8 ± 19.9 ± 7.0 ± 7.0

100 113.5 110.0 106.7

± 13.2 ± 20.0 ± 11.6 ± 11.6

100 111.4 110.5 106.8

± 3.3 ± 20.9 ± 13.8 ± 5.9

100 104.7 116.2 106.2

± 3.5 ± 17.8 ± 13.9 ± 16.2

0

"0

c:

.:---l*

200

(J)

II)

c: ~

~

,~----- ~ --- -- "±'-----~

100

I

N

Definition of abbreviations: LTB4 = leukotriene B4 ; PAF = plateletactivating factor; IL-3 = interleukin-3; IL-5 = interleukin-5; GM-CSF = granulocyte/macrophage colony-stimulating factor; BC = endothelial cell. * Test stimuli were incubated for 15 min or 4 h as indicated on the ECs (mean ± SD shown). No statistical significance was reached.

and neutrophils as above, had no effect on EC adhesiveness (Table 2). Transendothelial Migration In order to examine the effect of potentially chemoattractive mediators on eosinophil and neutrophil migration through ECs, the mediators were added to the lower transmigration chambers. In the absence of stimuli, spontaneous migration was very low, i.e., < 10% of the granulocytes present in the upper chamber migrated through the EC layer. Formylmethionylleucylphenylalanine (FMLP) and LTB4 were found to be potent but nonselective inducers of leukocyte migration across the endothelium (Figure 4). In contrast, PAF induced the migration of eosinophils only, with no effect upon neutrophils over a 60-min time period. This selective 80

***

0

0

-7

10

-8

10

-9

10

PAF C18 (MOLES)

-10

10

-8

10

+1M3:3 2086

Figure 5. Dose-response curve of PAF-stimulated transendothelial migration of eosinophils (closed triangles) and neutrophils (open triangles) (* P < 0.05, ** P < 0.01, respectively). Results are expressed as the percentage increase in migration over background migration levels in the absence of any chemotactic stimulus. The addition of the specific PAF inhibitor WEB 2086 (10- 4 M) at the PAF concentration inducing maximum eosinophil migration can be seen to completely eliminate this effect.

action was independent of either the concentration of PAF from 10-7 to 10-10 M (Figure 5) or of the length of incubation period (10 to 60 min). The maximal effect of PAF on eosinophil migration was found at a concentration of 10-8 M. The two molecular species of PAF, CI6 and CI8, were of similar potency and selectivity. This specific recruitment of eosinophils required a concentration gradient as it was not seen when PAF was added to the eosinophils on the luminal surface of the ECs. The addition of the selective inhibitor WEB 2086 (10-4 M) (33) completely abolished the PAF effect (Figure 5), indicating that this effect was not mediated by a contaminant of the preparation. The EC monolayers were found to be tightly confluent by scanning EM before and after transendothelial migration, indicating that PAF did not induce eosinophil-mediated lysis of the EC monolayer. Transmission EM on sections of EC-coated filters after PAF-induced eosinophil transmigration revealed the adherence of eosinophils to the ECs (Figure 6A) and their subsequent migration through the intercellular junctions of the ECs and the pores of the underlying filter (Figure 6B).

Discussion PAF

LTB.

FMLP

Neutrophils

Figure 4. Platelet-activating factor (PAF) enhances the transendothelial migration of eosinophils (*** P < O.Oot, one-way ANOVA and Student's t test) but not neutrophils. Leukotriene B4 (LTB4) and formylmethionylleucylphenylalanine (FMLP) enhance transendothelial migration of both eosinophils and neutrophils (*** P < 0.001, ** P < 0.01, respectively). Means + SD of triplicate observations are shown. Cell migration is expressed as a percentage of the total number of cells added per well.

We have developed a simple microwell method for quantifying granulocyte adherence to ECs that enables comparative studies of eosinophil and neutrophil populations isolated from the same donor. This method is based on the spectrophotometric measurement of intracellular peroxidase after cell disruption as a measure of cell number. The assay has the advantages of minimal manipulation of the granulocytes before the adhesion assay, avoidance of radiolabeling that may interfere with cell functions, speed, and ease of execution. After lysis of the adherent cells, the plates can be stored at -20° C until performance of the peroxidase assay, without any detrimental effects..Furthermore, by measuring

Morland, Wilson, Holgate et al.: PAF-induced Selective Eosinophil Transendothelial Migration

2um

Figure 6. Transmission electron microscopy of EC-coated filters fixed after PAF-induced eosinophil transendothelial migration. Panel A shows an eosinophil attachedto an EC covering the 3-l4m pores of the polycarbonate filter. Panel B shows an eosinophil migrating through the EC layerand an underlying pore in the filter.

peroxidase activity to quantify cell number, it is possible to discriminate between eosinophil and neutrophil populations by exploiting the different properties of eosinophil peroxi dase and neutrophil peroxidase. The assay is extremely sensitive for eosinophils because of their higher peroxidase activity compared with neutrophils (29), enabling the use of lower eosinophil numbers than is possible using radiolabeled cells . The report by Kroegel and colleagues (34) that mediators can activate eosinophils and induce the release of eosinophil peroxidase raises a potential problem with the use of intracellular peroxidase measurements as an indicator of cell number in adherence and migration reactions. This was avoided in adherence studies as test mediators were removed by washing before the addition of granulocytes. In the migration studies, migrated cells were collected by centrifugation,

563

supernatants containing test mediators and any secreted peroxidase were removed, and the cells were lysed and their peroxidase content was determined. Therefore, if anything, the estimation of the number of migrated cells would be lower than the true number because of the presence of partially degranulated cells. Eosinophil rather than neutrophil infiltration is characteristic of inflammatory reactions occurring in allergic and parasitic diseases, indicating that specific mechanisms for this eosinophil recruitment exist. The initial localized adherence of inflammatory cells to the vascular endothelium adjacent to an inflammatory site, and their subsequent transendothelial cell migration, is crucial to their infiltration of the extravascular compartment. In this study in vitro, we have compared the adherence of eosinophils to cytokine-stimulated ECs with that of neutrophils, and the transendothelial migration of both cell types in response to chemotactic agents. Differences in such interactions of eosinophils and neutrophils may indicate an important mechanism for the selective recruitment of eosinophils during allergic responses. Altered adhesive properties of the endothelium close to an inflammatory site would appear to be essential for local leukocyte adherence, since an increase in leukocyte adhesiveness alone would result in random interactions along the blood vessel. Evidence for active endothelial involvement in leukocyte adherence is indicated by the expression of EC surface adhesion molecules upon exposure to inflammatory stimuli, and their apparent involvement in adherence reactions (35). These cytokine-induced changes in surface adhesion receptor expression on ECs also occur in vivo (36), and similar changes occur in vascular endothelium after cutaneous allergen challenge (26, 37). In this study, we have shown that LPS, IL-Ia, and TNF-a all induce a dose-dependent enhancement of EC adhesiveness for both eosinophils and neutrophils. The time course of this upregulation in EC adhesiveness was similar for both cell types. No differences were found in the kinetics of the adherence reaction for eosinophils and neutrophils to ECs . Our results are consistent with an earlier study of eosinophil adherence to ECs using radiolabeled eosinophils (15) that suggested that either EC or eosinophil activation can induce increased eosinophil-EC adhesion. Similar conclusions have been reached by Kyan-Aung and co-workers (26) using eosinophils and neutrophils from different donors . This lack of discrimination for eosinophil and neutrophil adherence by cytokine-stimulated ECs suggests similar EC-dependent mechanisms of adherence for the two cell types. ELAM-I and ICAM-l have both been reported to mediate eosinophil and neutrophil adherence to TNF-stimulated ECs (26). There is also evidence that eosinophils and neutrophils express similar distributions of COlVCD18 receptors (38), suggesting that they share COl I/COl 8 mechanisms of adhesion . This is supported by the observation that cytokineinduced EC adhesion reactions with eosinophils are partially COl8-dependent (15). Recently, it has been suggested that VCAM-I expres sed on ECs may mediate selective adhesion of eosinophils compared with neutrophils via its interaction with VLA-4a on eosinophils (27). However, this is disputed by Kyan-Aung and co-workers (26), who have found no enhancement of adherence of eosinophils or neutrophils to ECs stimulated with IL-4, which is known to upregulate VCAM-I

564

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992

expression and induce VCAM-l-dependent adherence reactions (39). The influence of IL-3, IL-5, and GM-CSF on EC adhesiveness for eosinophils and neutrophils was also examined, since GM-CSF can be locally secreted by activated ECs themselves (40), and IL-3 and IL-5 are reported to be secreted by sensitized mast cells (41) and may therefore be produced locally during an allergic inflammatory reaction. These three cytokines are hemopoietic stimulation factors known to support eosinophil maturation, stimulate eosinophil functions, and prolong survival of mature blood eosinophils in vitro (42-44). Thus, localized production of these cytokines may have an important influence upon eosinophils, further enhancing their accumulation at allergic sites. Because cytokines released at an inflammatory site are likely to encounter and therefore influence ECs before diffusing into the bloodstream, we examined their effect on EC adherence properties. We found negligible increases in EC adhesiveness for both eosinophils and neutrophils after incubation of the ECs with IL-3, IL-5, and GM-CSF. However, it has been reported that IL-5, but not IL-3 or GM-CSF, can selectively enhance eosinophil-EC adherence by its action on eosinophils (28). Recently, Lawrence and Springer (45) described a novel mechanism for the initiation of neutrophil adherence to endothelium under conditions of physiologic flow, involving the binding to and rolling along a lipid bilayer containing the selectin GMPI40, which is expressed on endothelial membranes after activation by inflammatory mediators such as thrombin (46, 47). Such an interaction may be important in the margination of leukocytes observed in postcapillary venules adjacent to an inflammatory site and may bring the leukocytes into close contact with the endothelium for sufficient time to enable chemoattractants diffusing out between these ECs to influence leukocyte integrin-mediated adherence. Thus, it is possible that IL-5 may have a role in selective eosinophil recruitment by acting on eosinophils marginating at the inflammatory site. The short-acting, low-molecular-weight mediators PAF, LTB4 , and histamine were found to have no EC-dependent enhancement of adherence of neutrophils and eosinophils, in keeping with previous observations for PAF (15). However, PAF and LTB4 are reported to enhance eosinophil and neutrophil adherence to ECs when both leukocytes and ECs are stimulated simultaneously (15,48-51). Thus, although these mediators may be involved in rapid granulocyte-dependent adherence reactions, there appears to be no selectivity in this process to explain eosinophil recruitment. The adhesion stage is a necessary prerequisite for the migration of inflammatory cells through the endothelium, since leukocytes are only capable of responding to a chemotactic agent when attached to a surface (52). The absence of any selective EC-dependent recruitment for eosinophils at this stage led us to examine the subsequent step of transendothelial migration of eosinophils and neutrophils in response to chemotactic agents. The development of in vitro models oftransendothelial migration ofleukocytes (53) provides a means of examining the action of single agents in isolation. The EC monolayer/filter system used here retains EC polarity and has the advantage over chemotaxis assays of enabling study of cell migration through a relevant cellular barrier in response to abluminal EC chemotactic agents.

Background levels of migration over a 60-min incubation period for both eosinophils and neutrophils were low, indicating that few cells migrate across untreated ECs unless a chemotactic agent is present beneath the monolayers. LTB4 (10-8 M) and FMLP (l0- 8 M) induced rapid migration of both eosinophils and neutrophils, with 64 ± 4.2 % (mean ± SD) of eosinophils and 59.1 ± 3.2 % neutrophils migrating in response to LTB4 , and 35 ± 4.2 % of eosinophils and 41.7 ± 5.9% of neutrophils in response to FMLP, over a 60min incubation period. Previous reports of LTB4 chemotactic activity have been conflicting, Wardlaw and colleagues (54) finding LTB4 to be chemotactic for neutrophils only, whereas Uden and associates (55) found it to act on both eosinophils and neutrophils. In contrast to LTB4 and FMLP, PAF was found to induce trans endothelial migration of eosinophils only, with no increase over background migration for neutrophils. This is the first indication of a specific signal for eosinophil recruitment using migration through an EC monolayer in vitro and suggests that PAF may be a crucial mediator in events leading to the tissue eosinophilia characteristic of inflammation that accompanies allergic and parasitic disease. PAF is a potent phospholipid mediator released from various cell types involved in allergic inflammation reactions, with a range of biologic activities that suggest it has a role as a mediator in asthma (56). It is has been shown previously to be a chemotactic agent for eosinophils (54) and to induce their activation (34, 57). Inhalation of PAF has been shown to induce bronchoalveolar eosinophilia (58), and intradermal injection of PAF has been shown to induce the migration of eosinophils into the injection site (59), but only in atopic patients. This may indicate that cytokine priming of eosinophils in atopic patients may influence their responsiveness to PAF. We have addressed the possibility of this affecting our results by performing all comparative experiments with eosinophils and neutrophils from the same donor. Our findings that PAF induces specific eosinophil migration through the endothelium but not specific adherence are in keeping with our EM findings in endobronchial biopsies from patients with asthma, exhibiting close association of a variety of leukocytes with the ECs of the bronchial subepithelial vascular plexus but showing a predominance of eosinophils within the extravascular space with few neutrophils evident (60). Although eosinophils are capable of EC damage, scanning and transmission EM revealed no disruption of the EC monolayer during migration. This indicated that the migratory process in these experiments occurs by a physiologically relevant mechanism rather than by cell disruption. A complex set of interactions may be involved in the selective PAF-induced migration of eosinophils. PAF is known to activate eosinophils, inducing LTC4 production and degranulation (57). ECs may themselves synthesize and release PAF when stimulated with TNF and IL-l (61), and it has been demonstrated that LTC4 and LTD4 stimulate ECs in culture to synthesize PAF (62). In addition, PAF has been shown to induce fully reversible cytoskeletal changes in ECs, resulting in their retraction (63). This effect plus direct effects of PAF on eosinophils may enable their rapid migration through the endothelium to occur. In summary, adhesion reactions between granulocytes

Morland, Wilson, Holgate et al.: PAF-induced Selective Eosinophil Transendothelial Migration

and ECs are important as initiators of extravasation, with EC activation adjacent to an inflammatory site acting as a focus for inflammatory cell-EC interaction, but not apparently exerting any selective influence on the type of granulocyte adhering. Selection seems to occur at the transendothelial migration stage, with cell-specific chemoattractants recruiting specific cell populations into the tissues, resulting in the characteristic patterns of inflammation that are associated with different disease states. Acknowledgments: This work was supported by the National Asthma Campaign (United Kingdom).

References 1. Barnes, P. J. 1989. A new approach to the treatment of asthma. N. Engl. J. Med. 321:1517-1527. 2. Gleich, G. J., and C. R. Adolphson. 1986. The eosinophilic leukocyte: structure and function. Adv. Immunol. 39: 177-253. 3. Holgate, S. T., W. R. Roche, and M. J. Church. 1991. The role of the eosinophil in asthma. Am. Rev. Respir. Dis. 143:S66-S70. 4. Djukanovic, R., J. W. Wilson, K. M. Britten et al. 1990. Quantitation of mast cells and eosinophils in the bronchial mucosa of symptomatic atopic asthmatics and healthy control subjects using immunohistochemistry. Am. Rev. Respir. Dis. 142:863-871. 5. Taylor, J., and A. R. Luksza. 1987. Peripheral blood eosinophil counts and bronchial responsiveness. Thorax 42:452-456. 6. Wardlaw, A. J., S. Dunnette, G. L. Gleich, J. v. Collins, and A. B. Kay. 1988. Eosinophils and mast cells in bronchoalveolar lavage in subjects with mild asthma: relationship to bronchial hyperreactivity. Am. Rev. Respir. Dis. 137:62-69. 7. Bousquet, J., P. Chanez, J. Y. Lacoste etal. 1990. Eosinophilic inflammation in asthma. N. Engl. J. Med. 323:1033-1039. 8. Metzger, W. J., D. Zavala, H. B. Richerson et al. 1987. Local allergen challenge and bronchoalveolar lavage of allergic asthmatic lungs. Am. Rev. Respir. Dis. 135:433-440. 9. De Monchy, J. G. R., H. F. Kaufmann, P. Venge et al. 1985. Bronchoalveolar eosinophils during allergen-induced late asthmatic reactions. Am. Rev. Respir. Dis. 131:373-376. 10. Frigas, E., and G. J. Gleich. 1986. The eosinophil and the pathophysiology of asthma. J. Allergy Clin. Immunol. 77:527-537. 11. Filley, W. V., K. E. Holley, G. M. Kephart, and G. J. Gleich. 1982. Identification by immunofluorescence of eosinophil granule major basic protein in lung tissues of patients with bronchial asthma. Lancet i: 11-16. 12. Frigas, E., D. A. Loergering, and G. J. Gleich. 1980. Cytotoxic effects of the guinea pig eosinophil major basic protein on tracheal epithelium. Lab. Invest. 42:35-42. 13. Gundel, R. H., L. G. Letts, and G. J. Gleich. 1991. Human eosinophil major basic protein induces airway constriction and airway hyperresponsiveness in primates. J. Clin. Invest. 87: 1470-1473. 14. Harlan, J. M. 1985. Leukocyte-endothelial interactions. Blood 65:513525. 15. Lamas, A. M., C. M. Mulroney, and R. P. Schleimer. 1988. Studies on the adhesive interaction between purified human eosinophils and cultured vascular endothelial cells. J. Immunol. 140:1500-1505. 16. Bevilacqua, M. P., J. S. pober, M. E. Wheeler, R. S. Cotran, and M. A. Gimbrone, Jr. 1985. Interleukin 1 acts on cultured human vascular endothelium to increase the adhesion of polymorphonuclear leukocytes, monocytes, and related leukocyte cell lines. J. Clin. Invest. 76:20032011. 17. Gamble, J. R., J. M. Harlan, S. J. KIebanoff, and M. A. Vadas. 1985. Stimulation of the adherence of neutrophils to umbilical vein endothelium by human recombinant tumor necrosis factor. Proc. Natl. Acad. Sci. USA 82:8667-8671. 18. Pober, J. S., M. P. Bevilacqua, D. J. Mendrick, L. A. Lapierre, W. Fiers, and M. A. Gimbrone. 1986. Two distinct monokines, interleukin 1 and tumor necrosis factor, each independently induce biosynthesis and transient expression of the same antigen on the surface of cultured human vascular endothelial cells. J. Immunol. 136: 1680-1687. 19. Bevilacqua, M. P., J. S. Pober, D. L. Mendrick, R. S. Cotran, and M. A. Gimbrone. 1987. Identification of an inducible endothelial-leukocyte adhesion molecule. Proc. Natl. Acad. Sci. USA 84:9238-9242. 20. Phillips, M. L., E. Nudelman, F. C. A. Gaetaetal. 1990. ELAM-1 mediates cell adhesion by recognition of a carbohydrate ligand, Sialyl-Le-. Science 250: 1130-1132. 21. Walz, G., A. Aruffo, W. Ko1anus, M. Bevilacqua, andB. Seed. 1990. Recognition by ELAM-1 of the sialyl-le- determinant on myeloid and tumor cells. Science 250:1133-1135. 22. Pober, J. S., M. A. Gimbrone, L. A. Lapiere et al. 1986. Overlapping pat-

565

terns of activation of human endothelial cell by interleukin 1, tumor necrosis factor, and immune interferon. J. Immunol. 137:1893-1896. 23. Lo, S. K., G. A. V. Seventer, S. M. Levin, and S. D. Wright. 1989. Two leukocyte receptors (CDlla/CD18 and CD11 b/CD18) mediate transient adhesion to endothelium by binding to different ligands. J. Immunol. 143:3325-3329. 24. Smith, C. W., R. Rothlein, B. J. Hughes et al. 1988. Recognition of an endothelial determinant for CD 18-dependent human neutrophil adherence and transmigration. J. Clin. Invest. 82:1746-1756. 25. Barton, R. W., R. Rothlein, J. Ksiazek, and C. Kennedy. 1989. The effect of anti-intercellular adhesion molecule-Ion phorbol-ester-induced rabbit lung inflammation. J. Immunol. 142: 1278-1282. 26. Kyan-Aung, U., D. O. Haskard, R. N. Poston, M. H. Thornhill, and T. H. Lee. 1991. Endothelial leukocyte adhesion molecule-1 and intercellular adhesion molecule-1 mediate the adhesion of eosinophils to endothelial cells in vitro and are expressed by endothelium in allergic cutaneous inflammation in vivo. J. Immunol. 146:521-528. 27. Bochner, B. S., F. W. Luskinskas, M. A. Gimbroneetal. 1991. Adhesion of human basophils, eosinophils, and neutrophils to interleukin I-activated human vascular endothelial cells: contributions of endothelial cell adhesion molecules. J. Exp. Med. 173:1553-1556. 28. Walsh, G. M., A. Hartnell, A. J. Wardlaw, K. Kurihara, C. J. Sanderson, and A. B. Kay. 1990. IL-5 enhances the in vitro adhesion of human eosinophils, but not neutrophils, in a leukocyte integrin (CDll/18)dependent manner. Immunology 71:258-265. 29. Cramer, R., M. R. Soranzo, P. Dri et al. 1984. A simple reliable assay for myeloperoxidase activity in mixed neutrophil-eosinophil cell suspensions: application to detection of myeloperoxidase deficiency. J. Immunol. Methods 70: 119-125. 30. Jaffe, E. A. 1984. Culture and identification of large vessel endothelial cells. In Biology of Endothelial Cells. E. A. Jaffe, editor. Martinus Nijhoff, Boston. 1-13. 31. Vadas, M. A., J. R. David, A. Butterworth, N. T. Pisani, and T. A. Siongok. 1989. A new method for the purification of human eosinophils and neutrophils, and a comparison of the ability of these cells to damage schistosomula of Schistosoma mansoni. J. Immunol. 122: 1228-1236. 32. Kimura, I., Y. Moritani, and Y. Tanizaki. 1973. Basophils in bronchial asthma with reference to reagin-type allergy. Clin. Allergy 3:195-202. 33. Casals-Stenzel, J., G. Muacevic, and K. Weber. 1987. Pharmacological actions of WEB 2086, a new specific antagonist of platelet-activating factor. J. Pharmacol. Exp. Ther. 241:974-981. 34. Kroegel, C., T. Yukawa, G. Dent, P. Chanez, K. F. Chung, and P. J. Barnes. 1988. Platelet-activating factor induces eosinophil peroxidase release from purified human eosinophils. Immunology 64:559-562. 35. Pober, J. S., andR. S. Cotran. 1990. The role of endothelial cells ininflammation. Transplantation 50:537-544. 36. Munro, J. M., J. S. Pober, and R. S. Cotran. 1989. Tumor necrosis factor and interferon gamma induce distinct patterns of endothelial activation and associated leukocyte accumulation in skin of Papio anubis. Am. J. Pathol. 135:121-133. 37. Leung, D. Y. M., J. S. Pober, and R. S. Cotran. 1991. Expression of endothelial-leukocyte adhesion molecule-I in elicited late phase allergic reactions. J. Clin. Invest. 87:1805-1809. 38. Hartnell, A., R. Moqbel, G. M. Walsh, B. Bradley, and A. B. Kay. 1990. Fe gamma and CDll/18 receptor expression on normal density and low density human eosinophils. Immunology 69:264-270. 39. Thornhill, M. H., S. M. Wellicome, D. Mahiouz, J. S. S. Lanchbury, U. Kyan-Aung, and D. O. Haskard. 1991. TNF combined with IL-4 or lFN gamma selectively enhance endothelial cell adhesion for T cells: the contribution of VCAM-1 dependent and independent binding mechanisms. J. Immunol. 146:592-598. 40. Sieff, C. A., S. Tsai, andD. V. Faller. 1987. Interleukin 1 induces cultured human endothelial cell production of granulocyte-macrophage colonystimulating factor. J. Clin. Invest. 79:48-51. 41. Plaut, M., J. H. Pierce, C. J. Watson, J. Hanley-Hyde, R. P. Nordan, and W. E. Paul. 1989. Mast cell lines produce lymphokines in response to cross-linkage of FceRI or to calcium ionophores. Nature 339:64-67. 42. Rothenberg, M. E., W. F. Owen, Jr., D. S. Silbersteinetal. 1988. Human eosinophils have prolonged survival, enhanced functional properties, and become hypodense when exposed to human interleukin 3. J. Clin. Invest. 81: 1986-1992. 43. Yamaguchi, Y., Y. Hayashi, Y. Sugama et al. 1988. Highly purified murine interleukin 5 (IL5) stimulates eosinophil function and prolongs in vitro survival. J. Exp, Med. 167:1737-1742. 44. Lopez, A. F., D. J. Williamson, J. R. Gamble et al. 1986. Recombinant human granulocyte-macrophage colony-stimulating factor stimulates in vitro mature human neutrophil and eosinophil function, surface receptor expression and survival. J. Clin. Invest. 78: 1220-1228. 45. Lawrence, M. B., and T. A. Springer. 1991. Leukocytes roll on a selectin at physiologic flow rates: distinction from and prerequisite for adhesion through integrins. Cell 65:859-873. 46. Hattori, R., K. K. Hamilton, R. D. Fugate, R. P. McEver, and P. J. Sims. 1989. Stimulation of secretion of endothelial cell von Willebrand factor

566

47.

48. 49.

50.

51. 52. 53.

54. 55.

AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 6 1992

is accompanied by rapid redistribution to the cell surface of the intracelLular granule membrane protein GMP-140. J. Bioi. Chem. 264:7768-7771. Geng, J., M. P. Bevilacqua, K. L. Moore et al. 1990. Rapid neutrophil adhesion to activated endothelium mediated by GMP-140. Nature 343: 757-760. Kimani, G., M. G. Tonnesen, and P. M. Henson. 1988. Stimulation of eosinophil adherence to human vascular endothelial cells in vitro by plateletactivating factor. J. Immunol. 140:3161-3166. Zimmerman, G. A., and T. M. McIntyre. 1988. Neutrophil adherence to human endothelium in vitro occurs by CDw18 (Mol, MAC-lILFA-l, GP 150,95) glycoprotein-dependent and -independent mechanisms. J. Clin. Invest. 81:531-537. Hoover, R. L., M. J. Karnovsky, K. Frank Austen, E. J. Corey, and R. A. Lewis. 1984. Leukotriene B4 action on endothelium mediates augmented neutrophil adhesion. Proc. Natl. Acad. Sci. USA 81:2191-2193. Gimbrone, M. A., A. F. Brock, and A. I. Schafer. 1984. Leukotriene B4 stimulates polymorphonuclear leukocyte adhesion to cultured vascular endothelial cells. J. CUn. Invest. 74: 1552-1555. Synderman, R., and E. J. Goetzl. 1981. Molecular and cellular mechanisms of leukocyte chemotaxis. Science 213: 830-837. Taylor, R. F., T. H. Price, S. M. Schwartz, and D. C. Dale. 1981. Neutrophil-endothelial cell interactions on endothelial monolayers grown on micropore filters. J. Clin. Invest. 67:584-587. Wardlaw, A. J., R. Mocbel, O. Cromwell, and A. B. Kay. 1986. Plateletactivating factor. A potent chemotactic and chemokinetic factor for human eosinophils. J. Clin. Invest. 78: 1701-1706. Uden, A. M., J. Palmblad, J. A. Lindgren, and C. Malmsten. 1983. Effects of novel Iipoxygenase products on migration of eosinophils and neutro-

phils in vitro. Int. Arch. Allergy Appl. Immunol. 72:91-93. 56. Braquet, P., T. Y. Shen, L. Touqui, and B. B. Vargaftig. 1987. Perspectives in platelet-activating factor research. Pharmacol. Rev. 39:97-145. 57. Bruynzeel, P. L., B. L. Koenderman, P. T. M. Kok, M. L. Hameling, and J. Verhagen. 1986. Platelet-activating factor (PAF-acether) induced leukotriene C 4 formation and luminol dependent chemiluminescence by human eosinophils. Pharmacol. Res. Commun. 18:61-69. 58. Arnoux, B., A. Denjean, C. P. Page, D. Nolibe, J. Morley, and J. Beneviste. 1988. Accumulation of platelets and eosinophils in baboon lung after Paf-acether challenge. Am. Rev. Respir. Dis. 137:855-860. 59. Henocq, E., and B. B. Vargaftig. 1986. Accumulation of eosinophils in response to intracutaneous PAF-acether and allergens in man. Lancet i: 1378-1379. 60. Beasley, R., W. R. Roche, J. A. Roberts, and S. T. Holgate. 1989. Cellular events in the bronchi in mild asthma and after bronchial provocation. Am. Rev. Respir. Dis. 139:806-817. 61. Bussolino, F., G. Camussi, and C. Baglioni. 1988. Synthesis and release of platelet-activating factor by human vascular endothelial cells treated with tumor necrosis factor or interleukin-I alpha. J. Bioi. Chem. 263: 11856-1l861. 62. Mclntyre, T. M., G. A. Zimmerman, and S. M. Prescott. 1986. Leukotriene C 4 and D4 stimulate human endothelial cells to synthesize platelet-activating factor and bind neutrophils. Proc. Natl. Acad. Sci. USA 83:2204-2208. 63. Bussolino, F., G. Camussi, M. Aglietta et al. 1987. Human endothelial cells are targets for platelet-activating factor. J. Immunol. 139:24392446.