| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Infectious Diseases Section, Veterans Affairs Medical Center, Building 45, 500 West Fort Street, Boise, ID 83702, USA
2 University of Idaho, Moscow, ID, USA
3 Defence Science and Technology Laboratory, Porton Down, Salisbury, UK
4 University of Washington, Seattle, WA, USA
Correspondence
A. E. Bryant
amy.bryant{at}mindspring.com
Received 29 October 2005
Accepted 11 January 2006
Abbreviations: MFI, mean fluorescence intensity; PE, phycoerythrin; PLC, phospholipase C; PMNL, polymorphonuclear leukocytes.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We have recently shown that intramuscular injection of PLC caused a rapid and irreversible decline in skeletal muscle blood flow due to toxin-induced intravascular aggregates of activated platelets, leukocytes and fibrin (Bryant et al., 2000a, b). Neutrophils in these aggregates were often marginated along the endothelium but all remained intravascular (Bryant et al., 2000a). These findings suggested that the large heterotypic aggregates of platelets and leukocytes induced by PLC also contributed to impairment of the tissue inflammatory response.
In vitro, PLC-induced platelet/PMNL aggregate formation was largely gpIIbIIIa dependent (Bryant et al., 2003, 2000a). Yet blockade of gpIIbIIIa did not completely prevent PLC-induced platelet/neutrophil binding, suggesting that other adherence molecules play a role in these critical heterotypic cellular interactions. For most physiologic agonists (e.g. thrombin, ADP), induction of platelet/leukocyte aggregates involves an initial P-selectin (CD62P)-dependent tethering step, followed by a gpIIbIIIa-dependent stabilization step (Blockmans et al., 1995).
Thus, the present study investigated (1) the contribution of CD62P to PLC-induced platelet/platelet and platelet/neutrophil aggregate formation, and (2) the functional ability of PMNL, with and without adherent platelets, to migrate across an endothelial cell layer. Results demonstrate that P-selectin and gpIIbIIIa each uniquely contribute to PLC-induced platelet/leukocyte complex formation and that the adherence of large platelet aggregates significantly reduces PMNL diapedesis.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cellular complex formation in whole blood. PLC-induced platelet/platelet and platelet/neutrophil complex formation was followed by flow cytometry as previously described (Bryant et al., 2000a). Briefly, heparinized whole blood was obtained from healthy human volunteers who gave signed informed consent and who denied taking any medication for the previous 10 days. Whole blood (100 µl) was mixed with 10 µl each of FITC-conjugated anti-human CD42b (a pan platelet marker, green; Coulter) and phycoerythrin (PE)-conjugated anti-human CD1lb (a neutrophil marker, red; Coulter). Stock PLC was diluted in PBS containing 2·0 mM CaCl2 and 100 µM ZnCl2. These ions are necessary for PLC binding and enzymic activities, respectively. Blood was stimulated with either PLC (0·03–1 unit) or PBS vehicle for 5 min at 37 °C. Red blood cells were removed by formic acid lysis and remaining cells were fixed in 2 % (v/v) paraformaldehyde in PBS for flow cytometric analysis as described below.
A modification of this protocol was followed to prepare aggregates for use in transendothelial cell migration assays (Fig. 1
). In this case, 100 µl heparinized whole blood was stimulated with either 1 unit of PLC or PBS vehicle for 10 min at 37 °C (Fig. 1, step 1). To stop the reaction, anti-PLC mAb (20 µg ml–1) was added to all samples and red blood cells were removed by formic acid lysis. Cells were sedimented by gentle centrifugation (300 g, 5 min) in the presence of anti-P-selectin antibody (10 µg ml–1; Pharmingen) to inhibit spontaneous platelet binding induced by soft pelleting of the cells (Fig. 1, step 2). This treatment had no effect on complexes formed prior to centrifugation. Cells were resuspended in cell isolation buffer (CIB) (10 mM HEPES, 129 mM NaCl, 9·9 mM NaHCO3, 2·8 mM KCl, 0·8 mM KH2PO4, 0·8 mM MgCl2, 5·6 mM D-glucose, pH 7·4 supplemented with 1·5 % human serum albumin) to the original 100 µl volume and used in the transendothelial cell migration assay (see next section). Using this isolation procedure, the number of leukocytes obtained from 100 µl whole blood and subsequently added to the inserts was routinely 4–5x105. These values did not change with PLC treatment.
|
Transendothelial cell migration assays.
The functional status of neutrophils when complexed with platelets was evaluated using a transendothelial cell migration assay. Human umbilical vein endothelial cells (HUVEC; Cambrex Bio Science) were maintained in tissue culture using complete endothelial cell growth medium (cEGM-2; Cambrex) following the manufacturer's instructions. Cells of passage 3–4 were plated on gelatin and fibronectin-coated Transwell polycarbonate filters with 5 µm pore size (Corning) at 2·5x104 cells/insert in 100 µl cEGM. Filter inserts were suspended in the wells of a 24-well tissue culture plate containing 600 µl cEGM and HUVEC were cultured to confluence (
6 days). At the end of this period, visual light microscopic inspection of the insert revealed the presence of a bilayer of endothelial cells – one monolayer on the apical surface and one on the basal surface of the insert. This finding has been reported by others (Ebisawa et al., 1992). The integrity of the endothelial cell barrier in each insert was verified before conducting an experiment by a transendothelial electrical resistance (TEER) volt/ohm meter (EndOhm-6; World Precision Instruments). Intact layers routinely gave a mean total unit area resistance value of 39·6 
cm2 compared to 16·5 cm2 for an insert without cells.
One hundred microlitres of prepared platelet/neutrophil aggregates was placed into the top chamber of the insert (Fig. 1
, step 3). A chemoattractant, n-formylmethionylleucylphenylalanine (5x10–7 M final concentration), was then added to the lower chamber, and the plate returned to the tissue culture incubator. After 2 h, the plate was removed and endothelial cell integrity was again evaluated by electrical resistance as described above. The upper and lower layers of the inserts were gently washed with PBS, then the insert was processed for electron microscopy as described below (Fig. 1, step 4). Leukocytes that had crossed the endothelial cell layer into the lower chamber were collected and stained with fluorochrome-conjugated anti-CD11b and anti-CD42b antibodies, as described above, for 20 min at 37 °C, then fixed in 1 % paraformaldehyde in PBS (Fig. 1, step 5) for flow cytometry analysis.
The transmigration experiment was performed in replicates of three to six on four separate occasions using blood from different donors.
Electron microscopy. Transwell inserts were prepared for electron microscopy using standard procedures. Briefly, after the 2 h incubation period, the upper and lower faces of the inserts were gently washed with prewarmed EGM-2. The filter was left attached to the plastic insert throughout all steps, and reagents exchanged by passing the insert among the wells of a 24-well tissue culture plate containing the appropriate solutions. After washing, the cells were fixed in two changes of 4C 2·5 % glutaraldehyde in 0·1 M cacodylate buffer, pH 7·2–7·4 (Electron Microscopy Sciences), for 2 h each, then post-fixed in 1 % osmium tetroxide (Ted Pella), stained with uranyl acetate (Ernst Fullam) and dehydrated in a series of ethanol dilutions. Inserts were infiltrated stepwise with EMBED (Electron Microscopy Sciences) to 100 %, following the manufacturer's instructions for the hard resin. The inserts were then suspended in a small flexible embedding cup containing 100 % resin, the upper surface of the insert was covered with additional EMBED, and the entire unit polymerized for 48 h at 60 °C. The exposed plastic insert was cut away from the hardened resin with a jeweller's saw, and the blocks were sent to the University of Idaho's core electron microscopy facility. Blocks were cut bisecting the filter along its diameter, and faced using a diamond knife to remove any marks left by the jeweller's saw. The cut faces were given a thin coat of carbon via vacuum evaporation, placed in a specimen vice and viewed with a LEO SUPRA 35VP field emission scanning electron microscope in backscatter electron detection mode. Digital images were captured as TIFF files directly onto the LEO computer.
Inhibition of PLC-induced cellular complex formation by anti-platelet strategies. To investigate the molecular mechanisms of PLC-induced platelet/neutrophil complex formation, studies used cells that were cytoplasmically, rather than surface, labelled. This system has several advantages over the whole blood system. First, by marking the cell with a cytoplasmic label, it ensures that anti-adhesin strategies will not be hindered by surface-bound fluorochrome-labelled antibodies. Second, aliquots of these samples can be directly taken to the cytospin for visual evaluation. Third, antibodies used to block cell–cell interactions are not taken up by irrelevant cell types, thus less reagent can be used in a more specific manner. The disadvantages of this system are twofold: (1) there is a higher background adherence of platelets to leukocytes compared to the whole blood system using surface labelling (approx. 15 vs 5 % in whole blood) – a likely consequence of the more extensive isolation and staining procedures applied to both platelets and neutrophils in the purified cell system, and (2) there is a lower percentage of PMNL binding platelets (35–45 vs 85–95 % in whole blood). This latter observation suggests that soluble or cellular elements of whole blood contribute to complex formation induced by PLC.
For these studies, purified platelets and neutrophils were prepared as follows. Platelets in platelet-rich plasma were cytoplasmically labelled with the green fluorescent dye, BCECF-AM (2 µg ml–1; Molecular Probes). After 30 min, platelets were separated from unincorporated dye and plasma proteins by gel filtration as previously described (Bryant et al., 2003). Briefly, platelet-rich plasma (5–6 ml) was loaded onto a column packed with Sepharose 2B that was pre-equilibrated with CIB. Fractions containing BCECF-labelled, fibrinogen-free platelets were pooled and platelets were quantified by flow cytometry using 3·6 µm fluorobeads as an internal standard. The platelet concentration was adjusted to 1x108 cells ml–1 with CIB, and the final volume was supplemented to contain 2·0 mM CaCl2 and 100 µM ZnCl2 [complete CIB (cCIB)]. This isolation technique consistently yielded high numbers of purified resting platelets as determined by the absence of P-selectin surface staining. Further, these platelets were fully responsive (i.e. expressed high levels of P-selectin) to known activators such as thrombin (data not shown).
Neutrophils were isolated by differential density-gradient centrifugation as previously described (Bryant et al., 1993). Isolated neutrophils were labelled intracellularly with the red fluorescent dye, dihydroethidium [8 µg (1x107 cells)–1; Molecular Probes]. After 30 min, unincorporated dye was removed by a single wash in CIB, and cells were resuspended to a concentration of 2x106 ml–1 in cCIB.
For each test, fluorescently labelled isolated platelets and neutrophils were combined in a total volume of 1 ml (platelet : neutrophil ratio, 50 : 1), and the solution reconstituted to contain fibrinogen (100 µg ml–1). Cells were then preincubated for 10 min with the following agents, either individually or in combination: gpIIbIIIa inhibitor (eptifibatide, Integrilin, 2 µg ml–1, a commercially available synthetic heptapeptide that is a high-affinity inhibitor of gpIIbIIIa; Schering-Plough), a monoclonal anti-CD62P blocking antibody (8 µg ml–1, murine IgG1; Pharmingen) or an isotype-matched control antibody (8 µg ml–1, murine IgG1; Pharmingen). Recombinant PLC [0·25 U (1x106 platelets)–1] or PBS vehicle was then added for 10 min at 37 °C. The reaction was terminated by the addition of an equal volume of 2 % paraformaldehyde in PBS and samples immediately analysed by flow cytometry. A portion of each sample (150 µl) was fixed to a microscope slide using a cytospin, air-dried and stained first with Hansel stain then with Wright's stain. Images were viewed with a Zeiss Axiophot light microscope and photographed at x100 magnification. The extent of platelet/platelet aggregation in each specimen was scored over the range of 0 to 4+ by an observer who examined a minimum of three fields per slide in a blinded study.
|
RESULTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
(a). In this instance, the CD11b-positive neutrophil population was clearly distinguished and was largely devoid of adherent platelets as indicated by the low percentage of dual colour positivity (6 %) in this population. Among the platelet-positive neutrophils, each white cell bound only a few platelets as indicated by the low MFI of the platelet signal (i.e. MFICD42b=1·9). A second discrete population was also visible (Fig. 2a, arrow) and represented 4 % of the total events collected. Cytometric analysis showed that this population consisted solely of aggregates of platelets greater than or equal to 10 µm in size and that the apparent CD11b positivity of this population was artefact created by the bleed-over of the green (FITC) fluorescence into the red (PE) channel (not shown). No electronic compensation was used to reposition this population. Instead, it was gated and analysed separately, allowing us to study platelet/platelet interactions in the same histogram used to evaluate platelet/neutrophil complexes.
|
). Further, each neutrophil bound multiple platelets as indicated by an increase in MFICD42b to 5·7 (Fig. 2b). In addition, the separate population of platelet/platelet aggregates became much more prominent (16 % of total events, Fig. 2b, arrow).
Transendothelial cell migration of platelet/leukocyte complexes
The functional capability of neutrophils complexed with platelets was evaluated using a transendothelial cell migration assay. Platelet/neutrophil complexes isolated from PLC-treated whole blood, as shown in Fig. 1, were used for this functional assay. One representative experiment of four is presented in detail here.
PLC stimulated 82 % of neutrophils in whole blood to become platelet (i.e. CD42b) positive compared to 26 % in unstimulated whole blood (Table 1). The increased background complex formation and the higher CD11b surface expression in these experiments were attributable to the added processing of cells (i.e. lysis of red blood cells without subsequent fixation, followed by centrifugation to recover aggregates). Neutrophils from either control or PLC-stimulated samples (n=6 per group) were placed on the endothelial cell-covered insert filters and allowed to migrate toward a chemoattractant in the lower chamber. After 2 h, the mean number of transmigrating neutrophils from the untreated whole blood was 16·4 % of those added to the upper chamber (Table 1). Transmigrated neutrophils remained virtually platelet free (MFICD42b=1·4), but had increased size and reduced granularity (not shown). Surface CD11b expression was within 10 % of pre-transmigration levels. In contrast only 6·4 % of neutrophils from PLC-treated samples migrated across the endothelial cell layer (Table 1) – a 61 % reduction in transmigration compared to the control. Interestingly, the transmigrated neutrophils from the PLC-treated samples were virtually platelet-free (CD42b MFI=1·5) and had a 22·5 % reduction in surface CD11b expression. No large platelet aggregates and only a few individual platelets were detected in this lower chamber (not shown).
|
a). Most neutrophils in these samples bound none, one or two platelets each. Platelets were generally not aggregated and were round and smooth, suggesting they were in a non-activated state. Neutrophils were rarely seen between the filter and the lower endothelial cell layer, suggesting they had successfully completed their transmigration to the lower chamber. In contrast, large aggregates of platelets and neutrophils were seen adjacent to the upper endothelial cell layer in PLC-treated samples (Fig. 3b). Neutrophils in PLC-treated samples were also seen within pores, but appeared to lack the large platelet aggregates seen on the surface of the insert. When an occasional free neutrophil was seen below the lower endothelial cell layer, it was devoid of adherent platelets (not shown). Platelets themselves formed large aggregates and displayed morphological changes typical of cellular activation, such as irregular shapes and ruffled membranes. In some areas of intense platelet/neutrophil aggregates, destruction of the upper endothelial cell layer was observed (not shown) and likely accounted for the slight reduction in transendothelial electrical resistance in these samples (27·1 vs 34·0 cm–2 in PLC-treated vs control wells, respectively).
|
). Light microscopy of the unstimulated combination of purified cells demonstrated that platelets were sometimes associated with neutrophils (Fig. 4a, arrow). Addition of PLC stimulated both an increase in the percentage of neutrophils binding platelets and an increase in the number of platelets bound per neutrophil (Fig. 4b), resulting in large aggregates composed of many platelets and neutrophils (Fig. 4b, large arrow). Large aggregates of pure platelets were also seen (Fig. 4b, small arrow).
|
) and decreased the number of platelets bound per neutrophil. However, this treatment slightly increased the percentage of neutrophils binding platelets. This unexpected finding was explained by light microscopic analysis of the complexes formed. Specifically, by blocking gpIIbIIIa, large platelet aggregates were not formed; therefore, relatively more individual platelets were available to bind singly or in small numbers to neutrophils (Fig. 4c, arrowheads), thereby increasing the percentage of dual positive events.
In contrast, incubation with blocking anti-CD62P antibody reduced the percentage of neutrophils binding platelets to baseline levels but did not significantly reduce platelet/platelet aggregates (Fig. 4d, small arrow). Occasionally neutrophils bound these large platelet aggregates likely via a CD11b/fibrinogen/gpIIbIIIa axis (Spangenberg et al., 1993) such that the MFICD42b of these few dual-colour events remained elevated. By blocking both adherence molecules, the distribution of cells returned to the untreated background profile (Fig. 4e).
|
DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
As evidence that this hypothesis has merit, we have recently demonstrated that intramuscular injection of PLC induced a rapid and irreversible decline in skeletal muscle perfusion due to the formation of occlusive intravascular aggregates of activated platelets, leukocytes and fibrin (Bryant et al., 2000b). Further, we have shown that PLC directly activates gpIIbIIIa (Bryant et al., 2003, 2000a) by mobilization of calcium stores (Bryant et al., 2003) and that this activation contributes to the formation of both platelet/platelet and large platelet/neutrophil aggregates (Bryant et al., 2000a). The results of the present work also demonstrate a role for P-selectin in PLC-induced platelet/leukocyte complex formation. Thus, as a working model (Fig. 5), we propose that PLC-induced activation of gpIIbIIIa mediates the formation of large platelet/platelet aggregates that are tethered to the leukocyte principally via toxin-mediated upregulated CD62P.
|
Translating these molecular mechanisms into a clinical application suggests that, as an adjunct to traditional antibiotic therapy, immunotherapeutic agents that prevent platelet aggregation could maintain vessel patency and ensure adequate perfusion of tissues in patients at risk for gas gangrene, just as they do in patients at risk for acute coronary artery events. However, as the present work suggests, targeting a single adherence molecule in gas gangrene may be less than optimal. For instance, blocking gpIIbIIIa would eliminate large platelet/platelet aggregates and prevent their binding to leukocytes. In the absence of these large complexes, occlusion of vessels would be prevented. Neutrophils, however, would retain the ability to bind multiple activated platelets in a rosette pattern. Such interactions are known to hyperstimulate neutrophil respiratory burst activity (Peters et al., 1999; Ruf et al., 1992) and to facilitate chemical cross-talk between these cells, resulting in direct endothelial cell injury and local vascular dysfunction (Ward et al., 1986), and hyperproduction of thromboxanes (Maugeri et al., 1992), leukotrienes (Maugeri et al., 1994) and cytokines (Neumann et al., 1997). Overproduction of these potent immunomodulators could also contribute to local tissue injury and to the systemic manifestations of gas gangrene. Addition of a strategy targeting platelet P-selectin (CD62P) would prevent these latter complications (Neumann et al., 1997).
In summary, we conclude that rapid tissue destruction in gas gangrene is mediated by the pleiotropic effects of PLC on platelets, leukocytes and endothelial cells, and that a combined immunotherapeutic strategy targeting both gpIIbIIIa and CD62P may restore the tissue inflammatory response, prevent vascular occlusion, maintain tissue viability, and reduce the need for radical amputation in patients with clostridial gas gangrene. Studies to evaluate these therapeutic approaches in experimental gas gangrene are currently under way in our laboratory.
|
ACKNOWLEDGEMENTS |
|---|
A portion of this work was presented in abstract form at the 7th biennial meeting of the Anaerobe Society of the Americas, abstract P1-2, Annapolis, MD, July 19–21, 2004.
The authors wish to thank Michael Wyett for the schematic illustrations, Franklin Bailey for electron microscopy and Ronald Bach for helpful discussions.
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Rafii, M. Park, A. E. Bryant, S. J. Johnson, and R. D. Wagner Enhanced Production of Phospholipase C and Perfringolysin O (Alpha and Theta Toxins) in a Gatifloxacin-Resistant Strain of Clostridium perfringens Antimicrob. Agents Chemother., March 1, 2008; 52(3): 895 - 900. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
