The ability of polymorphonuclear leukocytes (PMNs) to modulate endothelial cell (EC) activation was investigated. Adding PMNs to cultured HUVECs resulted in a release of IL-6 (888 ± 71 pg/ml, a 35-fold increase over release by the two cell types alone) and IL-8 (45.2 ± 14.5 ng/ml, a 6.4-fold over PMN release alone and a 173-fold increase over EC release alone). In contrast, the release of TNF-α, IL-1β, and platelet-derived growth factor was not affected by the EC-PMN coculture. Neutralizing mAbs to ICAM-1 or β2 integrins or a physical segregation of PMNs and ECs did not reduce EC stimulation. In contrast, cell-free supernatants of PMNs recapitulated EC activation with an 18-fold up-regulation of EC IL-6 mRNA. The filtration of PMN supernatant or PMN pretreatment with metabolic antagonists or membrane cross-linking agents all suppressed EC activation. By flow cytometry, PMNs released in the supernatant, heterogeneous membrane-derived microparticles containing discrete proteins of 28 to 250 kDa as resolved by SDS-PAGE. PMN microparticle formation was enhanced by inflammatory stimuli, including formyl peptide and phorbol ester, and was time-dependent, reaching a plateau after a 1-h incubation from stimulation. Purified PMN microparticles induced EC IL-6 release in a reaction that was quantitatively indistinguishable from that observed with unfractionated PMN supernatant and unaffected by a neutralizing Ab to soluble IL-6R. These findings demonstrate that membrane microparticles released from stimulated PMNs are competent inflammatory mediators to produce EC activation and cytokine gene induction.

The interaction and functional cross-talk between leukocytes and endothelium is essential for vascular homeostasis and competent immune-inflammatory responses in vivo (1, 2). This process is contributed by an “adhesion cascade” mediated by a variety of membrane receptors on both cell types (1, 2, 3). In addition to physical docking and intercellular adhesion, these receptor-counterreceptor interactions involving integrins, selectins, and members of the Ig gene superfamily participate in transmembrane signal transduction and modulation of cell activation (4). Second, cytokines/chemokines locally released by activated leukocytes and endothelial cells (ECs)3 have been shown to contribute a complementary signaling mechanism to regulate leukocyte and vascular cell gene expression during inflammatory responses (5, 6, 7, 8). Finally, the ability of endotoxin-stimulated monocytes to release heterogeneous membrane-derived microparticles has been postulated as a third potential mechanism to regulate, albeit indirectly, vascular cell effector functions. This possibility was exemplified by the ability of microparticles to promote prothrombinase complex assembly, thus facilitating intravascular generation of thrombin and enhanced procoagulant activity (9). Alternatively, the observation that monocyte-derived microparticles contained organized membrane receptors including tissue factor, CD14, and β2 integrins suggested their potential contribution to alternative signaling pathways in vascular cells (10).

In this study, we sought to reinvestigate the potential participation of polymorphonuclear leukocytes (PMNs) in EC stimulation as reflected in the release of inflammatory and chemotactic cytokines and the up-regulation of leukocyte-EC adhesion molecules. We found that coculturing PMNs with endothelium induced EC activation and inflammatory gene induction. In addition, this pathway was mediated by membrane microparticles released from activated PMNs acting as competent inflammatory mediators on the endothelium.

PMNs were isolated from acid-citrate-dextrose anticoagulated blood drawn after informed consent from normal healthy volunteers by differential centrifugation on a Ficoll-Hypaque gradient and dextran sedimentation as described previously (11). ECs were prepared by collagenase treatment and used between passages 2 and 4.

Anti-ICAM-1 mAbs 2D5, 6E6, and 3D6 were generated and characterized for function and epitope recognition in previous studies (12). Neutralizing anti-CD18 mAbs 60.3 and IB4 and anti-CD11a mAb TS1/22 were obtained from American Type Culture Collection (Manassas, VA). Anti-VCAM-1 (E16.15) and anti-E-selectin (H4/18) mAbs were a kind gift of Dr. J. Bender (Yale University). A neutralizing polyclonal anti-IL-6R Ab was purchased from Endogen (Cambridge, MA). Nonbinding mAb 14E11 was used as a negative control.

PMNs were suspended at 3 × 106 cells/ml in medium 199 (BioWhittaker, Walkersville, MD) supplemented with 20% heat-inactivated FBS (BioWhittaker), penicillin (100 U/ml)-streptomycin (100 μg/ml), and l-glutamine (2 mM) (pH 7.4). Cells were stimulated in the presence or absence of Ca2+ ions (2.5 mM), FMLP (10 μM), Con A (5 μg/ml), or PMA (10 ng/ml) and added to monolayers of ECs grown to confluency in 96-well plates for 10 h at 37°C. In control experiments, ECs were also directly stimulated with the same concentrations of the stimuli for 10 h at 37°C. At the end of the incubation under the various conditions tested, the cell-free supernatants were collected, centrifuged for 10 min at 200 × g, and analyzed for released IL-6, IL-8, TNF-α, IL-1β, or platelet-derived growth factor by ELISA (Endogen). In another set of experiments, FMLP (10 μM)-stimulated PMNs and resting ECs were preincubated with 20 μg/ml of control nonbinding mAb 14E11 or mAbs to CD18 (60.3), CD11a (TS1/22), or ICAM-1 (2D5, 3D6, or 6E6) for 20 min at 22°C. PMNs were then added to ECs for 10 h at 37°C before the determination of IL-6 release by ELISA. In contact inhibition experiments, confluent ECs were grown on insert Transwell membranes (8 μm, Costar, Cambridge, MA), whereas FMLP-stimulated PMNs were incubated in the lower Transwell compartment. After a 10-h culture at 37°C, IL-6 release was determined by ELISA. In other experiments, cell-free supernatants from FMLP-stimulated PMNs (3 × 106/ml) or 10 ng/ml of TNF-α-containing medium were collected and sterile-filtered through a 0.2-μm pore-sized membrane of disposable Millex-VV filter units (Millipore, Bedford, MA). Filters were rinsed by flushing 2 ml of PBS through the pores and retrieved in 1 ml of reversed flow. Aliquots of filtered supernatants or retrieved particles were added to confluent EC monolayers, and the level of IL-6 secretion was measured after a 10-h incubation as described previously. In another series of experiments, FMLP-stimulated PMNs (3 × 106/ml) were incubated in 96-well tissue culture plates for 2 to 6 h at 37°C. Cells were centrifuged at 200 × g, and the cell-free supernatant (300 μl) was collected and added to confluent EC monolayers for an additional 10 h of incubation at 37°C before the determination of released IL-6 and IL-8 as described above. In other experiments, cytokine release was determined after a 0- to 43-h stimulation of ECs at 37°C. In another series of experiments, cytokine release from ECs was determined following heat-denaturation of FMLP-stimulated PMN supernatants at 100°C for 5 min or pretreatment of the PMN suspension with 1% paraformaldehyde/PBS for 2 h at 4°C. Alternatively, FMLP-stimulated PMNs (3 × 106/ml) were incubated with the metabolic inhibitor 2-deoxy-d-glucose (2DG) (50 mM; Sigma, St. Louis, MO) alone or in combination with cycloheximide (CX) (10 μg/ml; Sigma) and sodium azide (0.02%; Fisher Scientific, Pittsburgh, PA) for 2 h at 37°C. Cell-free supernatants under the various experimental conditions were collected and analyzed for induction of EC release of IL-6 following a 10-h incubation as described above. In another series of experiments, FMLP-stimulated PMN suspensions or their derived cell-free supernatants were incubated with an anti-IL-6R neutralizing Ab, anti-CD18 mAb IB4, or control mAb 14E11 (all at 25 μg/ml) for 20 min at 22°C before addition to EC monolayers and determination of IL-6 release. In other experiments, FMLP-stimulated suspensions of PMNs were incubated with confluent EC monolayers for 10 h at 37°C. After washes, ECs were fixed in cold (−20°C) methanol for 15 min, washed, and incubated in PBS containing 5% FBS and 0.05% Tween-20 for 15 min at 22°C. ECs were then incubated in RPMI 1640 containing 10% FBS and mAbs to ICAM-1 (2D5, 20 μg/ml), VCAM-1 (E16.15, 1:2), E-selectin (H4/18, 1:10), MHC class I (W6/32, 20 μg/ml), or control nonbinding IgG (14E11, 20 μg/ml) for 1 h at 22°C. After three washes at 22°C, cells were incubated with peroxidase-conjugated goat anti-mouse IgG for 30 min at 22°C followed by the addition of tetramethylbenzidine and H2O2 as substrates and determination of OD at λ = 450 nm.

ECs (passage 2–3) seeded in 75-cm2 cell culture flasks to 90% confluence were incubated in the presence of untreated or 0.2 μm-filtered cell-free supernatant from FMLP-stimulated PMNs (5 × 106/ml) for 10 h at 37°C. At the end of the incubation, ECs were washed, and total RNA was extracted by the RNAzol B method (Tel-Test, Friendswood, TX). Samples (15 μg) were separated by electrophoresis on denaturing agarose formaldehyde gels and transferred to nylon membranes (GeneScreen, New England Nuclear Life Science Products, Boston, MA) by overnight capillary transfer in 20× SSC (1× SSC is 150 mM NaCl and 15 mM sodium citrate (pH 7.0)). Following UV cross-linking, filters were sequentially hybridized with [32P]deoxyCTP (Amersham, Arlington Heights, IL), random-primed (Boehringer Mannheim, La Jolla, CA), labeled IL-6 or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) probes. Hybridization was performed in 5× SSC, 10× Denhardt’s solution, 1% SDS, and 100 μg/ml of salmon sperm DNA for 14 h at 60°C, with washes in 2× SSC and 1% SDS at 60°C. Blots were exposed to a Kodak phosphorimaging screen, and signals were quantified using equipment and software from the same manufacturer. Probe stripping was conducted in 0.5% SDS at 90°C.

PMN-derived microparticles were quantitated by flow cytometry using a fluorescent lipid intercalating dye, PKH26-GL (Sigma). This aliphatic chromophore partitions into lipid bilayers and confers a red fluorescence. PMNs (5 × 106 cells/ml) were labeled with PKH26-GL (4 μM) according to the manufacturer’s specifications. The labeled cells were incubated in the presence or absence of CaCl2 (2.5 mM), serum, and 10 μM FMLP. At various time intervals between 3 min and 6 h at 37°C, the cell-containing supernatant, cell-free supernatant containing PMN-derived vesicles, and 0.2-μm pore-filtered cell-free supernatant were isolated and analyzed by flow cytometry. In other experiments, increasing concentrations of FMLP (0–10 μM) were used to stimulate PMN suspensions in medium alone for 1 h before analysis of microparticle release by flow cytometry. Samples were analyzed for forward and side scatter parameters using a FACS (Becton Dickinson, Mountain View, CA). Each sample was analyzed for a total of 25,000 events or a 10-s interval. A gate was chosen to include particles distinctly positive for red fluorescence. In parallel experiments, membrane microparticles purified from the supernatant of FMLP-stimulated PMNs by ultracentrifugation at 60,000 rpm for 2 h were lysed in 1% Triton X-100 and 0.05% SDS plus protease inhibitors. Extracts were separated by electrophoresis on a 6% SDS polyacrylamide gel with visualization of protein bands by Coomassie blue staining under nonreducing conditions.

A 10 h-coculture of FMLP (10 μM)-stimulated PMNs with ECs resulted in a dramatic increase in the release of cytokines IL-6 and IL-8 (Fig. 1,A). Under these experimental conditions, IL-6 release (888 ± 71 pg/ml) increased by 35-fold over values observed with either cell types alone; IL-8 release (45.2 ± 14.5 ng/ml) increased by 6.4- and 173-fold over values observed with PMNs or ECs alone, respectively (Fig. 1,A). The potential requirement of PMN stimulation and/or divalent ion in EC activation was investigated. First, optimal induction of EC release of IL-6 by PMN coculture was observed in the presence of Ca2+ ions (Table I), and serum (see below). Second, PMN stimulation with inflammatory agonists, including the chemoattractant FMLP, phorbol ester (PMA), or Con A, all resulted in increased EC release of IL-6 as compared with unstimulated cultures or ECs directly stimulated with FMLP, PMA, or Con A under the same experimental conditions (Table I). The addition of PMNs to EC monolayers was also associated with an up-regulation of inducible leukocyte-EC adhesion molecules, ICAM-1, VCAM-1, and E-selectin (Fig. 1,B). As determined by ELISA, the magnitude of this response was quantitatively indistinguishable from that observed in control incubation reactions with TNF-α-activated ECs (Fig. 1,B). In contrast, the release of IL-1β, platelet-derived growth factor, or TNF-α or the modulation of MHC class I molecules was not affected under the same experimental conditions of PMN-EC coculture (Fig. 1, A and B).

FIGURE 1.

EC activation by PMN coculture. A, PMNs were cocultured with resting EC monolayers for 10 h at 37°C before determination of the indicated cytokine release by ELISA. B, The experimental procedures are essentially as described in A, except that the modulation of the various indicated adhesion molecules was determined by ELISA. mAb 14E11 was used as a nonbinding control mAb. EC stimulation was carried out in the presence of 10 ng/ml of TNF-α. Data are the mean ± SEM of three independent experiments.

FIGURE 1.

EC activation by PMN coculture. A, PMNs were cocultured with resting EC monolayers for 10 h at 37°C before determination of the indicated cytokine release by ELISA. B, The experimental procedures are essentially as described in A, except that the modulation of the various indicated adhesion molecules was determined by ELISA. mAb 14E11 was used as a nonbinding control mAb. EC stimulation was carried out in the presence of 10 ng/ml of TNF-α. Data are the mean ± SEM of three independent experiments.

Close modal
Table I.

Effect of PMN stimulation on EC activationa

Culture ConditionsIL-6 Release (pg/ml)\E
Unstimulated PMNs,− Ca2+/ECs 385 ± 90\E 
Unstimulated PMNs,+ Ca2+/ECs 434 ± 38\E 
FMLP-stimulated ECs 90 ± 14\E 
FMLP-stimulated PMNs/ECs 731 ± 117\E 
PMA-stimulated ECs 78 ± 22\E 
PMA-stimulated PMNs/ECs 2263 ± 382\E 
Con A-stimulated ECs 30 ± 14\E 
Con A-stimulated PMNs/ECs 848 ± 182 
Culture ConditionsIL-6 Release (pg/ml)\E
Unstimulated PMNs,− Ca2+/ECs 385 ± 90\E 
Unstimulated PMNs,+ Ca2+/ECs 434 ± 38\E 
FMLP-stimulated ECs 90 ± 14\E 
FMLP-stimulated PMNs/ECs 731 ± 117\E 
PMA-stimulated ECs 78 ± 22\E 
PMA-stimulated PMNs/ECs 2263 ± 382\E 
Con A-stimulated ECs 30 ± 14\E 
Con A-stimulated PMNs/ECs 848 ± 182 
a

Unstimulated PMNs in serum-containing medium were equilibrated in the presence or absence of 2.5 mM CaCl2 or stimulated with the various indicated agonists in the presence of Ca2+ ions before addition to EC monolayers. Alternatively, ECs were directly incubated with FMLP, PMA, or Con A. EC release of IL-6 under the various conditions tested was determined by ELISA after a 10-h incubation at 37°C. Data are the mean ± SD of two independent determinations.

Preincubating PMNs or ECs with neutralizing mAbs to ICAM-1 or β2 integrins CD18 or CD11a failed to reduce EC release of IL-6 under these experimental conditions (Fig. 2,A). Similarly, a physical separation of ECs and PMNs in Transwell chambers only partially reduced the IL-6 generation by ECs observed in coculture experiments (Fig. 2,B). At variance with these conditions, cell-free supernatants of FMLP-stimulated PMNs induced EC release of IL-6 in a reaction that was quantitatively indistinguishable from that obtained in coculture experiments (Fig. 3). This response was abrogated by heat denaturation of the PMN supernatant or by prior treatment of the PMN suspension with paraformaldehyde or with metabolic inhibitor 2DG alone or in combination with sodium azide and CX (Fig. 3). Filtration of the cell-free supernatant from FMLP-stimulated PMNs through a 0.2-μm filter or ultrafiltration with a 100-kDa cut-off range completely abolished the stimulatory effect of this supernatant on EC release of IL-6 (Fig. 3). In control experiments, filtration of a TNF-α-containing supernatant did not reduce EC release of IL-6 (Fig. 3). Finally, cell-free supernatants of FMLP-stimulated PMNs induced EC release of IL-6 and IL-8 in a time-dependent manner, reaching a maximum secretion at 10 to 18 h poststimulation (Fig. 4, A and B).

FIGURE 2.

EC activation by PMN coculture is contact-independent. A, FMLP-stimulated PMNs and resting EC monolayers were preincubated with 20 μg/ml of control mAb 14E11, anti-CD18 mAb 60.3, anti-CD11a mAb TS1/22, anti-ICAM-1 mAbs 2D5 or 3D6 (both mapped to domain 1), or anti-ICAM-1 mAb 6E6 (mapped to domain 2) for 20 min at 22°C. PMNs were then added to ECs for 10 h at 37°C before determination of IL-6 release by ELISA. B, In another set of experiments, ECs were grown to confluency on polycarbonate membrane inserts of a Transwell plate. FMLP-stimulated PMNs were intercepted from direct contact with ECs by plating PMNs in the lower compartment of the Transwell plate for 10 h at 37°C before determination of IL-6 release by ELISA.

FIGURE 2.

EC activation by PMN coculture is contact-independent. A, FMLP-stimulated PMNs and resting EC monolayers were preincubated with 20 μg/ml of control mAb 14E11, anti-CD18 mAb 60.3, anti-CD11a mAb TS1/22, anti-ICAM-1 mAbs 2D5 or 3D6 (both mapped to domain 1), or anti-ICAM-1 mAb 6E6 (mapped to domain 2) for 20 min at 22°C. PMNs were then added to ECs for 10 h at 37°C before determination of IL-6 release by ELISA. B, In another set of experiments, ECs were grown to confluency on polycarbonate membrane inserts of a Transwell plate. FMLP-stimulated PMNs were intercepted from direct contact with ECs by plating PMNs in the lower compartment of the Transwell plate for 10 h at 37°C before determination of IL-6 release by ELISA.

Close modal
FIGURE 3.

EC activation mediated by PMN-released microparticles. EC monolayers were incubated with FMLP-stimulated PMNs or cell-free supernatants of FMLP-stimulated PMNs for 10 h at 37°C before determination of IL-6 release by ELISA. In the indicated experiments, PMN supernatant was heat-inactivated; treated with 1% paraformaldehyde or with the metabolic inhibitors 2DG (50 mM), sodium azide (SZ) (0.02%), and CX (10 μg/ml); or subjected to filtration or ultrafiltration before determination of IL-6 release. Filtered or unfiltered medium containing 10 ng/ml of TNF-α was used as a control. Data are the mean ± SEM of three independent experiments.

FIGURE 3.

EC activation mediated by PMN-released microparticles. EC monolayers were incubated with FMLP-stimulated PMNs or cell-free supernatants of FMLP-stimulated PMNs for 10 h at 37°C before determination of IL-6 release by ELISA. In the indicated experiments, PMN supernatant was heat-inactivated; treated with 1% paraformaldehyde or with the metabolic inhibitors 2DG (50 mM), sodium azide (SZ) (0.02%), and CX (10 μg/ml); or subjected to filtration or ultrafiltration before determination of IL-6 release. Filtered or unfiltered medium containing 10 ng/ml of TNF-α was used as a control. Data are the mean ± SEM of three independent experiments.

Close modal
FIGURE 4.

Time-course induction of EC IL-6 and IL-8 by PMN-derived supernatants. EC monolayers were incubated with medium containing FMLP and 20% serum or supernatants of FMLP-stimulated PMNs (containing FMLP and 20% serum) for different time intervals at 37°C before determination of IL-6 (A) or IL-8 (B) by ELISA. For each panel, the basal level of IL-6 and IL-8 in the PMN supernatants is also shown.

FIGURE 4.

Time-course induction of EC IL-6 and IL-8 by PMN-derived supernatants. EC monolayers were incubated with medium containing FMLP and 20% serum or supernatants of FMLP-stimulated PMNs (containing FMLP and 20% serum) for different time intervals at 37°C before determination of IL-6 (A) or IL-8 (B) by ELISA. For each panel, the basal level of IL-6 and IL-8 in the PMN supernatants is also shown.

Close modal

By Northern hybridization, incubating resting EC monolayers with cell-free supernatants of FMLP-stimulated PMNs resulted in an ∼18-fold induction of IL-6 mRNA as compared with background levels of untreated ECs (Fig. 5). Consistent with the data presented above, filtration of the PMN supernatant significantly reduced the increase in IL-6 mRNA expression under the same experimental conditions (Fig. 5). In control hybridization studies, PMN-derived supernatants failed to modulate the mRNA levels of GAPDH under the same experimental conditions (Fig. 5).

FIGURE 5.

Induction of EC IL-6 mRNA by PMN-derived supernatant. EC monolayers were incubated with supernatants of unstimulated or FMLP-stimulated PMNs before or after a 0.2-μm filtration. Total RNA was extracted, separated by denaturing agarose-formaldehyde gels, transferred to nylon membranes, and hybridized with IL-6 or GAPDH random-primed, labeled cDNA probes. A densitometric analysis of the radioactive bands is shown at the bottom of the figure.

FIGURE 5.

Induction of EC IL-6 mRNA by PMN-derived supernatant. EC monolayers were incubated with supernatants of unstimulated or FMLP-stimulated PMNs before or after a 0.2-μm filtration. Total RNA was extracted, separated by denaturing agarose-formaldehyde gels, transferred to nylon membranes, and hybridized with IL-6 or GAPDH random-primed, labeled cDNA probes. A densitometric analysis of the radioactive bands is shown at the bottom of the figure.

Close modal

To determine a potential mechanism of leukocyte microparticle release, PMNs were fluoresceinated with PKH26-GL, which is a lipid intercalating membrane dye, and analyzed by flow cytometry for microparticle formation. Analysis of PMN suspension under these conditions revealed two distinct populations identified by different forward and side scatter parameters and fluorescence intensities (Fig. 6, top left quadrant and inset). PMN stimulation with FMLP resulted in an ∼4-fold increase in the population with smaller forward and side scatter parameters (Fig. 6, top right quadrant), which is consistent with its potential identity with released membrane microparticles. This population was heterogeneous in size and was maintained in cell-free PMN supernatants, whereas the larger population corresponding to the cellular fraction was entirely depleted under these experimental conditions (Fig. 6, lower left quadrant). However, filtration of cell-free PMN supernatant through a 0.2-μm filter resulted in removal of the microparticle population by flow cytometry (Fig. 6, lower right quadrant). Microparticle release of PMN suspensions in medium alone occurred in an FMLP concentration-dependent manner, reaching a plateau at 100 nM of added stimulus (Fig. 7,A). In time-course experiments, microparticle release occurred in a time-dependent reaction, reaching a plateau at 1 h after stimulation, with no further increases for ≤6 h incubation (Fig. 7,B). Consistent with the data presented above, optimal microparticle release required the presence of serum and was enhanced by FMLP stimulation of PMNs (Fig. 7 B).

FIGURE 6.

Flow cytofluorometric determination of PMN release of membrane microparticles. PMNs were treated with a membrane-intercalating fluorescent dye and analyzed for forward and side scatter parameters by flow cytometry before or after a 0.2-μm filtration or following FMLP stimulation. The insets indicate the fluorescence intensity of the labeled populations detected under the various conditions tested.

FIGURE 6.

Flow cytofluorometric determination of PMN release of membrane microparticles. PMNs were treated with a membrane-intercalating fluorescent dye and analyzed for forward and side scatter parameters by flow cytometry before or after a 0.2-μm filtration or following FMLP stimulation. The insets indicate the fluorescence intensity of the labeled populations detected under the various conditions tested.

Close modal
FIGURE 7.

FMLP- and time-dependence of PMN microparticle release. A, PMN suspensions in medium alone were stimulated with the indicated concentrations of FMLP and analyzed for microparticle release by flow cytometry. B, Unstimulated or FMLP (10 μM)-stimulated PMN suspensions in the presence or absence of serum were analyzed for microparticle release at the indicated timepoints. Data are from a representative experiment of two independent determinations.

FIGURE 7.

FMLP- and time-dependence of PMN microparticle release. A, PMN suspensions in medium alone were stimulated with the indicated concentrations of FMLP and analyzed for microparticle release by flow cytometry. B, Unstimulated or FMLP (10 μM)-stimulated PMN suspensions in the presence or absence of serum were analyzed for microparticle release at the indicated timepoints. Data are from a representative experiment of two independent determinations.

Close modal

To directly investigate the potential role of PMN-derived membrane particles in EC activation and inflammatory gene induction, microparticles were purified from supernatants of FMLP-stimulated PMNs by ultracentrifugation at 60,000 rpm for 2 h. Analysis of these purified microparticles by SDS-PAGE revealed a pattern of limited complexity, with the appearance of a prominent ∼85-kDa component and fainter bands ranging in relative molecular mass between 28 and 250 kDa (Fig. 8). In EC activation experiments, the addition of purified PMN microparticles to EC monolayers resulted in prominent IL-6 release in a reaction that was quantitatively indistinguishable from that observed in coculture conditions or with the unfractionated, cell-free, PMN-derived supernatant (Fig. 9). Under these experimental conditions, preincubating FMLP-stimulated PMNs with neutralizing mAbs to β2 integrins or to soluble IL-6Rα failed to decrease EC release of IL-6 stimulated by PMN microparticles (Fig. 9).

FIGURE 8.

Biochemical analysis of membrane-derived PMN microparticles. PMN-derived microparticles were purified from serum-free supernatants of FMLP-stimulated cells by ultracentrifugation at 60,000 rpm for 2 h, lysed, and separated on a 6% SDS polyacrylamide gel under nonreducing conditions. Protein bands were visualized by Coomassie blue staining. Molecular mass markers in kilodaltons are shown on the left.

FIGURE 8.

Biochemical analysis of membrane-derived PMN microparticles. PMN-derived microparticles were purified from serum-free supernatants of FMLP-stimulated cells by ultracentrifugation at 60,000 rpm for 2 h, lysed, and separated on a 6% SDS polyacrylamide gel under nonreducing conditions. Protein bands were visualized by Coomassie blue staining. Molecular mass markers in kilodaltons are shown on the left.

Close modal
FIGURE 9.

EC activation by purified PMN-derived microparticles is not mediated by soluble IL-6R. Left panel, Membrane microparticles were purified from FMLP-stimulated PMN supernatants by reversed flow through a 0.2-μm filter and incubated with resting EC monolayers for 10 h at 37°C before determination of IL-6 release by ELISA. Unfractionated PMN supernatant was used as a control. Right panels, FMLP-stimulated PMN suspensions or cell-free supernatants were preincubated with 20 μg/ml of control mAb 14E11, anti-CD18 mAb IB4, or functionally blocking anti-IL-6R Ab (IL-6Rα) for 20 min at 22°C and added to EC monolayers before determination of IL-6 release by ELISA. Data are the mean ± SEM of three independent experiments.

FIGURE 9.

EC activation by purified PMN-derived microparticles is not mediated by soluble IL-6R. Left panel, Membrane microparticles were purified from FMLP-stimulated PMN supernatants by reversed flow through a 0.2-μm filter and incubated with resting EC monolayers for 10 h at 37°C before determination of IL-6 release by ELISA. Unfractionated PMN supernatant was used as a control. Right panels, FMLP-stimulated PMN suspensions or cell-free supernatants were preincubated with 20 μg/ml of control mAb 14E11, anti-CD18 mAb IB4, or functionally blocking anti-IL-6R Ab (IL-6Rα) for 20 min at 22°C and added to EC monolayers before determination of IL-6 release by ELISA. Data are the mean ± SEM of three independent experiments.

Close modal

In this study, we have shown that PMNs stimulated with inflammatory agonists release a heterogeneous population of membrane microparticles. In addition, released PMN microparticles acted on endothelium as competent inflammatory agonists, stimulating inflammatory gene expression, release of cytokines IL-6 and IL-8, and prominent up-regulation of leukocyte-EC adhesion molecules ICAM-1, VCAM-1, and E-selectin.

The possibility that cellular-derived microparticles could contribute to vascular cell responses has been postulated earlier. In the model of the platelet, activation by disparate stimuli including thrombin and the complement membrane attack complex C5b-9 has been characterized previously for its ability to induce membrane microvesiculation and particle release (13, 14). This process was shown to require a calpain-dependent dissociation of membrane proteins from the submembrane cytoskeleton (15) and potential intracellular signaling by the activity of one or more platelet protein kinases (i.e., myosin light chain kinase and Ca2+-calmodulin complex) (14). Functionally, released platelet microparticles contributed a favorable, negatively charged, phosphatidylserine environment for prothrombinase complex assembly and amplification of a procoagulant response. In this context, blood samples from patients with a generalized activation of coagulation were shown to contain platelet-derived microparticles, thus suggesting their potential role in disseminated amplification of coagulation in vivo (16). Alternatively, the expression of functional β3 integrins on platelet-derived microparticles suggested their potential involvement in adhesion and intracellular signaling mechanisms (17). Recently, a similar paradigm of microvesiculation has been extended to the leukocyte, with the demonstration that endotoxin-stimulated monocytes released heterogeneous phosphatidylserine-containing membrane microparticles, potentially contributing to prothrombinase activity and integrin-dependent adhesion reactions (9, 10).

In expanding these earlier observations, we now show that PMN-derived microparticles not only provide a negatively charged microenvironment for potential amplification of the coagulation cascade but also act as potent proinflammatory agonists competent to initiate a broad pathway of signal transduction and gene expression in ECs. This finding was reflected in the dramatic de novo up-regulation of endothelial IL-6 mRNA, which was associated with prominent release of IL-6 and IL-8, and in the induction of multiple EC-leukocyte adhesion molecules. Consistent with this paradigm, depletion of the microparticle fraction from PMN supernatants abrogated the inflammatory response of the endothelium, whereas prevention of cell-to-cell contact or neutralizing Abs to ICAM-1 or β2 integrins was ineffective. Although the mechanism of the observed leukocyte vesiculation remains to be elucidated, previous studies have suggested a role for cytoskeletal rearrangements (18) and/or a protein synthesis- and energy-dependent response to stimulation in this process (9). Consistent with this view, PMN microparticle release required divalent cations and serum and was significantly enhanced by stimulation with various inflammatory agonists in a rapid reaction that was completed 1 h after stimulation. By flow cytometry and initial biochemical characterization, PMN microparticles were heterogeneous in size and of limited complexity, containing a prominent ∼85-kDa band and additional proteins ranging in relative molecular mass between 28 and 250 kDa. It is unlikely that the stimulatory effect on the endothelium described here reflects endotoxin contamination of the culture medium of leukocyte microparticles. First, comparable experimental conditions in the absence of PMN supernatant were not associated with endothelial cytokine release. Second, ultrafiltration and heat treatment of PMN supernatant completely abrogated this pathway of EC activation, whereas endotoxin stimulation of ECs was unaffected.

The molecular basis of EC stimulation by PMN-derived microparticles is currently unknown. A potential working hypothesis for this pathway may involve a direct physical association of membrane microparticles with the EC surface followed by transmembrane signal transduction and de novo gene expression in ECs. Regardless of the underlying mechanism(s), the pathway described here appears entirely unrelated to the process of retrograde activation of endothelium by PMN-released soluble IL-6Rα described by Modur et al. (19). In this context, the same neutralizing Ab to IL-6Rα used by Modur et al. failed to reduce EC release of IL-6 stimulated by PMN supernatants.

The in vivo existence of the leukocyte-derived microparticles described here has not yet been investigated. However, we hypothesize that this pathway of leukocyte stimulation of endothelium may have potentially significant implications for general inflammatory responses and the pathogenesis of vascular injury in vivo (20). Microparticle release from activated leukocytes may provide a mechanism to amplify the local concentration of inflammatory and chemotactic cytokines and inducible adhesion molecules, facilitating intercellular communication and cross-signaling pathways between leukocytes and ECs (1). Alternatively, this process may also contribute in the exacerbation of aberrant leukocyte activation, prothrombin activation (9), and intercellular adhesion and migration during the initial phases of vascular injury and the atherosclerotic disease (20).

In summary, we have described an alternative mechanism of leukocyte-EC cross-talk that is of potential relevance for the earliest aspects of inflammation and vascular cell responses. Elucidation of the complex signal transduction pathway initiated by leukocyte microparticles in endothelium should provide important insights into the potential impact of this mechanism for inflammatory conditions in vivo.

1

This work was supported by National Institutes of Health Grants HL43773 and HL54131 and was completed during the tenure of an American Heart Association Established Investigatorship Award to D.C.A.

3

Abbreviations used in this paper: EC, endothelial cell; PMN, polymorphonuclear leukocyte; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; 2DG, 2-deoxy-d-glucose; CX, cycloheximide.

1
Carlos, T. M., J. M. Harlan.
1994
. Leukocyte-endothelial adhesion molecules.
Blood
84
:
2068
2
Dustin, M. L., T. A. Springer.
1991
. Role of lymphocyte adhesion receptors in transient interactions and cell locomotion.
Annu. Rev. Immunol.
9
:
27
3
Butcher, E. C..
1991
. Leukocyte-endothelial cell recognition: three (or more) steps to specificity and diversity.
Cell
67
:
1033
4
Schwartz, M. A., M. D. Schaller, M. H. Ginsberg.
1995
. Integrins: emerging paradigms of signal transduction.
Annu. Rev. Dev. Biol.
11
:
549
5
Rot, A..
1992
. Endothelial cell binding of NAP-1/IL-8: role in neutrophil emigration.
Immunol. Today
13
:
291
6
Schall, T. J..
1991
. Biology of the RANTES/SIS cytokine family.
Cytokine
3
:
165
7
Pober, J. S., R. S. Cotran.
1990
. Cytokines and endothelial cell biology.
Physiol. Rev.
70
:
427
8
Tanaka, Y., D. H. Adams, S. Hubscher, H. Hirano, U. Siebenlist, S. Shaw.
1993
. T-cell adhesion induced by proteoglycan-immobilized cytokine MIP-1β.
Nature
361
:
79
9
Robinson, R. A., L. Worfolk, P. B. Tracy.
1992
. Endotoxin enhances the expression of monocyte prothrombinase activity.
Blood
79
:
406
10
Satta, N., F. Toti, O. Feugeas, A. Bohbot, J. Dachary-Prigent, V. Eschwege, H. Hedman, J.-M. Freyssinet.
1994
. Monocyte vesiculation is a possible mechanism for dissemination of membrane-associated procoagulant activities and adhesion molecules after stimulation by lipopolysaccharide.
J. Immunol.
153
:
3245
11
Mesri, M., J. Plescia, D. C. Altieri.
1998
. Dual regulation of ligand binding by CD11b I domain: inhibition of intercellular adhesion and monocyte procoagulant activity by a factor X-derived peptide.
J. Biol. Chem.
273
:
744
12
Duperray, A., L. R. Languino, J. Plescia, A. McDowall, N. Hogg, A. G. Craig, A. R. Berendt, D. C. Altieri.
1997
. Molecular identification of a novel fibrinogen binding site on the first domain of ICAM-1 regulating leukocyte-endothelium bridging.
J. Biol. Chem.
272
:
435
13
Sims, P. J., T. Wiedmer, C. T. Esmon, H. J. Weiss, S. J. Shattil.
1989
. Assembly of the platelet prothrombinase complex is linked to vesiculation of the platelet plasma membrane.
J. Biol. Chem.
264
:
17049
14
Wiedmer, T., P. J. Sims.
1991
. Participation of protein kinases in complement C5b-9-induced shedding of platelet plasma membrane vesicles.
Blood
78
:
2880
15
Fox, J. E. B., C. D. Austin, J. K. Boyles, P. K. Steffen.
1990
. Role of the membrane skeleton in preventing the shedding of procoagulant-rich microvesicles from the platelet plasma membrane.
J. Cell Biol.
111
:
483
16
Holme, P. A., N. O. Solum, F. Brosstad, M. Roger, M. Abdelnoor.
1994
. Demonstration of platelet-derived microvesicles in blood from patients with activated coagulation and fibrinolysis using a filtration technique and Western blotting.
Thromb. Haemostasis
72
:
666
17
Gemmell, C. H., M. V. Sefton, E. L. Yeo.
1993
. Platelet-derived microparticle formation involves glycoprotein IIb-IIIa.
J. Biol. Chem.
268
:
14586
18
Maercklein, P. B..
1977
. Plasma membrane vesiculation: correlation between macrophage spreading and the shedding of cell surface vesicles.
Lab. Invest.
37
:
430
19
Modur, V., Y. Li, G. A. Zimmerman, S. M. Prescott, T. M. McIntyre.
1997
. Retrograde inflammatory signaling from neutrophils to endothelial cells by soluble interleukin-6 receptor α.
J. Clin. Invest.
100
:
2752
20
Ross, R..
1993
. The pathogenesis of atherosclerosis: a perspective for the 1990s.
Nature
362
:
801