Methods for Analysis of Matrix Metalloproteinase Regulation of Neutrophil-Endothelial Cell Adhesion.

Recent evidence indicates novel role for matrix metalloproteinases (MMPs), in particular gelatinase A (MMP-2), in the regulation of vascular biology that are unrelated to their well-known proteolytic breakdown of matrix proteins. We have previously reported that MMP-2 can modulate vascular reactivity by cleavage of the Gly32-Leu33 bound in big endothelin-1 (ET-1) yielding a novel vasoactive peptide ET-1[1-32]. These studies were conducted to investigate whether gelatinolytic MMPs could affect neutrophil-endothelial cell attachment. ET-1[1-32] produced by MMP-2 up-regulated CD11b/CD18 expression on human neutrophils, thereby promoted their adhesion to cultured endothelial cells. ET-1[1-32] evoked release of gelatinase B (MMP-9), which in turn cleaved big ET-1 to yield ET-1[1-32], thus revealing a self-amplifying loop for ET-1[1-32] generation. ET-1[1-32] was rather resistant to cleavage by neutrophil proteases and further metabolism of ET-1[1-32] was not a prerequisite for its biological actions on neutrophils. The neutrophil responses to ET-1[1-32] were mediated via activation of ET(A)receptors through activation of the Ras/Raf-1/MEK/ERK signaling pathway. These results suggest a novel role for gelatinase A and B in the regulation of neutrophil functions and their interactions with endothelial cells. Here we describe the methods in detail as they relate to our previously published work.

Introduction

Matrix metalloproteinases (MMPs) are a class of secreted enzymes with major functions in the degradation and remodeling of all components of the extracellular matrix (1). Gelatinase A (MMP-2) and gelatinase B (MMP-9) cleave denaturated collagens and type IV collagen, which is present in the basement membranes. This latter action is important for the mobilization of stem cells (2) and migration of lymphocytes and tumor cells (3). An increasing body of evidence indicates novel roles for gelatinases in the innate and adaptive immunity that are unrelated to matrix remodeling. For instance, gelatinase B was found to process cytokines and chemokines, resulting in skewed immune functions (3,4); while gelatinase A was reported to mediate platelet aggregation (5). Gelatinase A can also modulate vascular reactivity by reducing the vasodilatory potency of calcitonin gene-related peptide through cleavage of the Gly14-Leu15 bound (6) and by cleavage of the Gly32-Leu33 bound in big endothelin-1 (ET-1) to yield a novel vasoconstrictor peptide ET-1[1-32] (7). On a molar basis, ET-1[1-32] appears to be a more active peptide than the 21-amino acid ET-1 (7). Of particular interest, increased levels of gelatinase activity can often be detected with simultaneous increases in big ET-1 at sites of inflammation associated with myocardial ischemia (8,9), neointima formation (10) and atherosclerosis (11,12). Since these pathological conditions are also characterized by increased adhesiveness of leukocytes to the vascular endothelium (13) and ET-1 may function as an autocrine/paracrine modulator of leukocyte functions (14-17), we investigated whether ET-1[1-32] formed by gelatinase A and gelatinase B could affect neutrophil adhesion to endothelial cells and studied the underlying molecular mechanisms.

Materials and Methods

Activation of MMPs

Highly purified human gelatinases A or B are commercially available as pro-enzymes as well as active enzymes (Chemicon International, Mississauga, ON, Canada). Preparations of pro-MMP are often contaminated with small quantities of the corresponding active MMP. Larger quantities of active MMP can be obtained in the laboratory by incubating the pro-MMP (140 nM) with 4-aminophenylmercuric acetate (APMA, Sigma, 1 mM). APMA was prepared freshly by dilution from a 10 mM stock in NaOH with 1M Tris-HCl, pH 7.5. APMA activates pro-MMPs by disrupting a cystein switch. The activation reaction was allowed to proceed for 2 hrs at room temperature. If needed, unreacted APMA was scavenged by addition of BSA to a final concentration of 50 µM. Activity of MMP-2 and MMP-9 was tested against an extracellular matrix protein (e.g., Collagen type IV, Calbiochem) and by zymography (see below). Activated MMPs were also able to cleave some small vasoactive hormone peptides. We have shown that both MMP-2 and MMP-9 can cleave big endothelin-1[1-38] to yield two peptides, ET-1[1-32] and ET-1[33-38] (7,18). Only ET-1[1-32] has known biological activities, which were first discovered using an in vitro arterial system upon the preparation of the peptide in vitro (7).

In vitro preparation and characterization of ET-1[1-32]

ET-1[1-32] was prepared as described previously by cleaving synthetic human big ET-1 (Sigma-Aldrich, Oakville, ON, Canada) with activated MMPs. The cleavage reaction was conducted for16 h at 37°Cin HEPES-phosphate saline solution (in mM: NaCl 142, KCl 4.7, MgSO41.17, CaCl21.56, HEPES 10, KH2PO41.18; pH 7.4).

HPLC analysis: The incubation mixture was separated on an HPLC-chromatograph (Waters) using a 12.5 cm x 4 mm C-18 column (LiChrospher, Merck) with a (1% per minute)-gradient of 5% CH3CN in 0.1% aqueous TFA against 0.1% TFA in acetonitrile at 0.5 ml/min flow rate.

Mass analysis: HPLC resolved peaks were collected and the mass of peptides was determined with a Voyager Elite matrix assisted laser desorption ionization (MALDI) mass spectrometer (Applied Biosystems, Framingham, MA) equipped with delayed extraction and a reflectron. The instrument was run in the reflectron mode using 20 kV acceleration. External calibrations were completed using a mixture of known peptides.Samples were analysed using a two-layer preparation method (19). The first layer was 0.7 µL of 5 mg/ml sinapinic acid (Aldrich) in 3:2 acetone: methanol. The second layer was 0.4 µl of a 1:1 mixture of the sample solution and sinapinic acid solution. The second layer sinapinic acid solution was prepared at a concentration of 10 mg/ml in 1:1 acetonitrile: 0.1% trifluoroacetic acid (aq.) (Supelco). After the second layer dried, the sample spot was rinsed to remove excess salts by placing a 5 µl drop of 0.1% TFA on top of the sample spot and then removing the drop using a Kimwipe.

For the studies, ET-1[1-32] stock solution (10-4M) was diluted in phosphate buffered saline, containing 0.1% bovine serum albumin, pH 7.4 immediately before use. Endotoxin concentration of ET-1[1-32] preparations was below the detection limit (0.125 EU/ml) of the Limulus amebocyte lysate assay (E-Toxate, Sigma).

Isolation and treatment of neutrophil granulocytes

Neutrophil granulocytes were isolated from the peripheral venous blood obtained from non-smoking healthy volunteers by dextran sedimentation, centrifugation through Ficoll-Hypaque gradients (Pharmacia Diagnostics AB) and hypotonic lysis of red blood cells (17). Neutrophils (107cells/ml, purity >97%) were re-suspended in a modified Hanks’ balanced salt solution (in mM: NaCl 140, KCl 5, K2PO410, CaCl21.4, MgCl21.2, glucose 5, and 0.1% bovine serum albumin, pH 7.4), pre-incubated with enzyme inhibitors, as indicated, then challenged with ET-1[1-32] for the indicated times at 37ºC. The cells were then pelleted, and pellets and supernatants were collected for further analysis.

Flow cytometry analysis of adhesion molecule expression

Direct immunofluorescence labeling of control or treated neutrophils was performed as described previously (20). In brief, leukocytes were stained with a saturating concentration of FITC-labeled anti-human L-selectin monoclonal antibody (mAb) DREG-56 (IgG1, PharMingen) or R-phycoerythrin-labeled anti-human CD11b mAb Leu-185 (IgG1, BD Biosciences). Appropriately labeled, class-matched irrelevant mouse IgG1 was used as a negative control for each staining. Single or double-color immunofluorescence was analyzed by a FACScan flow cytometer (Becton Dickinson Immunocytometry Systems) with Lysis II software.

Activation of the Ras/Raf-1/MEK/ERK signaling pathway

Neutrophils were lysed in ice-cold lysis buffer (in mM: Tris 20, EGTA 1, Na3VO42, NF 25, 0.5% Triton X-100, PMSF 2, aprotinin 40 μg/ml and 10 μg/ml each of pepstatin A, chymostatin and leupeptin, pH 7.4) for 15 min on ice, and centrifuged at 4ºC for 10 min at 14,000 g. Western blot analysis of phosphorylated MEK and ERK 1/2 was performed using the Phospho Plus MEK 1/2 and ERK 1/2 MAP kinase antibody kits (New England Biolabs, Beverly, MA) as described (21). ERK 1/2 activity was assayed following immunoprecipitation of ERK 1/2 from the cell lysates with an immobilized anti-phospho-p44/p42 MAP kinase antibody (New England Biolabs) using Elk-1 fusion protein, as a specific target for ERK 1/2 (22). Phosphorylation of ERK 1/2 at Ser383 was quantified by densitometry following immunoblotting by using an affinity purified anti-phospho-Elk-1 polyclonal antibody (New England Biolabs) and chemiluminescence detection (Lumi-Light Western Detection Substrate, Roche Diagnostics).

Raf-1 kinase assay: Raf-1 activity was determined by using the method of Garner et al. (23) with modifications. In brief, Raf-1 was immunoprecipitated from the neutrophil lysates with an anti-Raf-1 antibody (C-12, Santa Cruz Biotechnology, Santa Cruz, CA), antigen-antibody complexes were then isolated by protein A-Sepharose CL-4B (Sigma), and Raf-1 activity was measured using the Raf-1 Kinase Cascade Assay kit (Upstate Biotechnology, Lake Placid, NY) in accordance with the manufacturer’s protocol.

Ras activation assay: Activated p21Ras(Ras-GTP) from neutrophil lysates was affinity purified by using GST-Ras binding domain of Raf-1 (residues 1-149) fusion protein conjugated to agarose beads (Upstate Biotechnology) (24). The beads were washed extensively and boiled in reducing sample buffer. The eluted proteins were resolved on a 10% SDS-polyacrylamide gel, transferred to a PVDF membrane, probed with a mouse anti-Ras mAb (clone RAS10, Upstate Biotechnology), and visualized using a goat anti-mouse secondary antibody conjugated to horseradish peroxidase (BioRad) and a chemiluminescence detection system (Lumi-Light Western Detection Substrate, Roche Diagnostics).

Metabolism of ET-1[1-32]

Measurements were performed as previously described for ET-1[1-38] and ET-1[1-21] (25). Briefly, ET-1[1-32] (100 pmol) was incubated with preparations of neutrophil cytosol (0.5 - 5mg protein) or membrane (0.2 - 2mg protein) at 37°C for 2 hrs. The incubation mixture was separated on HPLC and the resolved peaks were collected and further characterized on a MALDI mass spectrometer, as described above.

Measurement of gelatinase release

Gelatinolytic activity released from unchallenged and challenged neutrophils was determined using [3H]-acetylated gelatin as a substrate following activation of the latent gelatinase by 1.7 mM phenylmercuric acetate (26). Gelatinase released was expressed as the percentage of total enzyme units released from neutrophils treated with 0.1 % Triton X-100.

Zymographic analysis of MMPs

Thegelatinolytic activity of neutrophil MMPs was measured essentially as described before (7). The samples of conditioned supernatants and neutrophil homogenate were mixed with electrophoresis loading buffer (10:1, vol:vol, 20ml per well) and subjected to electrophoresis, which was carried out on a 7.5% SDS-PAGE co-polymerized with gelatin (2 mg/ml). Gelatin was added from a 10 fold concentrated stock containing pre-boiled gelatin type A (20 mg/ml, Sigma, Cat. # G 8150) and 1% SDS. Following electrophoresis, gels were washed with 2.5% Triton X-100 for 1 hr (3 times, 20 min each) and incubated for 48 hrs in enzyme assay buffer (25 mM Tris, pH 7.5, 5 mM CaCl2, 0.9% NaCl, 0.05% Na3N) for the development of enzyme activity bands. After incubation, the gels were stained with 0.05% Coomassie brilliant blue G-250 in a mixture of methanol: acetic acid: water (2.5: 1: 6.5) and de-stained in 4% methanol with 8% acetic acid. The gelatinolytic activities were detected as transparent bands against the background of Coomassie brilliant blue-stained gelatin. If more sensitivity was necessary, the gels were further destained in a solution of 1% Triton x-100 (1–2 hrs). This procedure increased the signal-to-noise ratio, allowing the visualization of faint gelatinase bands at 92, 125 and > 200 kDa. To measure the activities of the detected enzymes, zymograms were read using a ScanJet 3c scanner (Hewlett Packard, Boise, ID). The intensity of the bands was expressed as arbitrary units and analyzed using SigmaGel measurement software (Jandel Corporation, San Rafael, CA).

Neutrophil-endothelial cell adhesion assay

Culture of endothelial cells: Human coronary artery endothelial cells (HCAEC, cryopreserved in passage 3) obtained from Clonetics/BioWhittaker were cultured in T25 flasks (26). HCAEC (passages 4 to 6) seeded into 96-well microplates (Sarstedt) and grown to confluence were used in the adhesion assays.

Adhesion assay: Monolayers of HCAEC in 96-well microplates were stimulated with bacterial lipopolysaccharide (1 μg/ml, E. coli O155:B4, Sigma) for 4 h at 37ºC in a 5% CO2atmosphere. The cells were then washed three times with PBS, and 2x10551Cr-labeled neutrophils in 100 μl were added. In some experiments, ET-1[1-32] was added with neutrophils to activated HCAEC. After incubation for 30 min at 37ºC on an orbital shaker at 90 rpm, loosely adherent or unattached neutrophils were washed three times, and the endothelial monolayer plus the adherent neutrophils were lysed in 150 μl of 0.1% Triton X-100 in water. The number of adherent neutrophils in each experiment was estimated from the radioactivity of a control sample.

Statistical analysis

Results are reported as means ± SEM. Statistical comparisons were made by ANOVA using ranks (Kruskal-Wallis test), followed by Dunn’s multiple contrast hypothesis test to identify differences when various treatments were compared to the same control or by the Mann-Whitney test for unpaired observations. P values < 0.05 were considered significant for all tests.

Results and Discussion

Neutrophil gelatinase cleaves big ET-1 to yield ET-1[1-32]

Neutrophils can metabolize big ET-1 into smaller peptides (25).When big ET-1 was incubated with neutrophil cytosol and membrane fractions and the resultant incubates were analyzed using mass spectrometry or HPLC/mass spectrometry, fragments of ET-1[1-38] including ET-1[23-38], ET-1[1-22], ET-1[1-20], ET-1[1-19] and ET-1[1-18] were detected Fig. 1A.Gelatin-zymography revealed the presence of gelatinase B, but not gelatinase A, in both cytosol and membrane preparations of human neutrophils (18). Since gelatinase A can cleave big ET-1 to ET-1[1-32] (7); we tested whether gelatinase B also possess a similar activity. Indeed, APMA-activated gelatinase B cleaved big ET-1 to yield ET-1[1-32], as demonstrated using HPLC and MALDI-TOF mass spectroscopy Fig. 1B.

Fig. 1

HPLC/MALDI-TOF analysis shows that big ET-1 (ET-1[1-38]) is cleaved to yield many smaller peptides by proteases in the cytosol and cell membrane of neutrophils. Based on their masses, these peptides were assigned the ET-sequences indicated in brackets. (A). Bar: peaks in the region of the HPLC chromatogram corresponding to neutrophil proteins, not ET peptides. Mass analysis on a MALDI TOF instrument reveals that MMP-9 (gelatinase B), cleaves big ET-1 to yield ET-1[1-32] (B).

It is not clear whether neutrophils are a major source of bioactive ET peptides, such as ET-1[1-21] and ET-1[1-32]. Neutrophils contain predominantly pro-MMP-9, not active MMP-9 (18). Therefore, the conversion of big ET-1 into ET-1[1-32] is likely regulated by a specific MMP-9 activation event (that occurs during or following release of granule contents) as well as by rapid peptide degradation. In line with this, only small quantities of ET-1 peptides were detected with our HPLC/mass spectrometry assays Fig. 1A.

Metabolism of ET-1[1-32] by neutrophils

We next investigated whether further cleavage of ET-1[1-32] was a requirement for its bioactivity. ET-1[1-32] (100 pmol) was incubated with preparations of neutrophil cytosol (0.5-5mg protein) or membrane (0.2-2mg protein) and assessed for conversion into smaller peptides. There was little conversion of ET-1[1-32] into other smaller peptides such as ET-1[1-22], ET-1[1-21], ET-1[1-20] and ET-1[1-19] using a combination of HPLC with mass spectroscopy (18). Consistent with these observations, the effects of ET-1[1-32] on adhesion molecule expression were insensitive to phosphoramidon, an inhibitor of the metal-dependent endothelin converting enzymes (L-selectin expression was 35±6% and 31±7% of control, CD11b expression 178±17% and 184±16% of control in response to 100 nM ET-1[1-32] in the absence or presence of phosphoramidon, respectively, n=5, P>0.1) (18).

To further confirm that the biological effects of ET-1[1-32] were due to the inherent bioactivity of this peptide, neutrophils were challenged with big ET-1 (100 nM), the inactive precursor of ET-1 peptides, in the presence of phosphoramidon (an endothelin converting enzyme inhibitor) and o-phenanthroline (an MMP inhibitor). There were no detectable changes in adhesion molecule expression on neutrophils under these conditions (L-selectin and CD11b/CD18 expression were 110±6% and 102±3% of control, respectively, n=4, P>0.1) (18).

ET-1[1-32] induces gelatinase release from neutrophils

Next, we investigated whether ET-1[1-32] is capable of inducing gelatinase release from human neutrophils. Indeed, ET-1[1-32] evoked a concentration-dependent release of gelatinase (Fig. 2) without affecting cell viability (data not shown). Since neutrophils contain only gelatinase B, but not gelatinase A, and both gelatinase A and B are capable of cleaving big ET-1 to yield ET-1[1-32], a self-amplifying loop may exist for stimulation of ET-1[1-32] formation and gelatinase B release and consequently for neutrophil activation. Intriguingly, ET-1[1-32] appears to function as a partial neutrophil agonist, for it did not induce superoxide formation (18).

Fig. 2

ET-1[1-32] induces gelatinase release from human neutrophils. Isolated neutrophils were challenged with ET-1[1-32] for 30 min at 37°C. Gelatinase released into the culture medium was measured as described in Methods and is expressed as percentage of total cellular enzyme activity. N=4, P<0.05 vs unstimulated.

ET-1[1-32] induces L-selectin shedding and up-regulates CD11b/CD8 expression on neutrophils

ET-1[1-32] induced a rapid down-regulation of L-selectin and up-regulation of CD11b/CD18 on the surface of neutrophils (Fig. 3). These actions of ET-1[1-32] were concentration-dependent with EC50 values of 1 to 3 nM (18). Consistent with previous observations (7), on a molar basis ET-1[1-32] appeared to be a more potent neutrophil agonist than ET-1 (17,18). Our previous pharmacological and receptor binding studies showed that human neutrophils predominantly express ETAreceptors (17). Consistently, the actions of ET-1[1-32] on neutrophil expression of L-selectin and CD11b/CD18 as well as gelatinase release were almost completely inhibited by the selective ETAreceptor antagonist FR139317, but not by the ETBreceptor-selective antagonist BQ788 (18).

Fig. 3

Effects of ET-1[1-32] on surface expression of adhesion molecules on human neutrophils. Neutrophils were challenged with 100 nM ET-1[1-32] for 30 min at 37°C, then stained with fluorescein-labeled anti-L-selectin and anti-CD11b monoclonal antibodies. C, negative control of immunostaining with irrelevant antibodies. Results are representative of 6 experiments.

ET-1[1-32] activates ERK through the Ras/Raf-1/MEK pathway

Experiments with PD98059, which inhibit the activation of MEK and consequently of ERK, revealed that neutrophil responses to ET-1[1-32] were ERK-dependent (18). However, PD98059 was more effective in reversing ET-1[1-32]-induced up-regulation of CD11b/CD18 expression and gelatinase release (72±5% and 80±5% inhibition, respectively), than down-regulation of L-selectin expression (59±4% inhibition) (18). To provide direct evidence for the involvement of the Ras/Raf-1/MEK/ERK signaling pathway, we studied activation of this cascade. ET-1[1-32] induced rapid (within 2 min) association of GTP-bound active Ras with the glutathione-S-transferase-Ras binding domain of Raf-1 Fig. 4A and increased Raf-1 kinase activity Fig. 4B. Consequently, enhanced phosphorylation of MEK and ERK was also detected within 5 min of addition of ET-1[1-32] Fig. 4C.

Fig. 4

ET-1[1-32] activates the Ras/Raf-1 kinase/MEK/ERK 1/2 signaling pathway. A, Ras was detected by immunoblotting following affinity precipitation of GTP-bound active Ras from lysates of neutrophils. B, Raf-1 kinase activity was determined using a MAPK cascade system following immunoprecipitation of Raf-1. C, MEK and ERK 1/2 phosphorylation was detected by immunoblotting. Neutrophils were challenged with ET-1[1-32] for 2 min (A,B) or 5 min (C). Results are representative of 3 independent experiments.

These actions of ET-1[1-32] were concentration-dependent and was comparable to those evoked by 100 nM platelet-activating factor (18). ET-1[1-32]-induced ERK phosphorylation was inhibited by 75±6% with PD98059 (18). The similar degree of inhibition of ERK phosphorylation and CD11b/CD18 expression and gelatinase release observed with PD98059 indicates a tight correlation between ERK activation and neutrophil responses to ET-1[1-32]. It should be noted that the inhibition was incomplete, indicating the involvement of other signaling pathways.

ET-1[1-32] promotes neutrophil adhesion to human coronary artery cells in vitro

To assess the biological significance of ET-1[1-32]-induced changes in expression of adhesion molecules on the neutrophil surface, we performed a neutrophil-endothelial cell adhesion assay. To mimic blood flow, the assay was performed under non-static conditions (26). While ET-1[1-32] did not stimulate neutrophil adhesion to monolayers of unstimulated HCAEC, about 1.5-fold more neutrophils adhered to LPS-stimulated HCAEC when neutrophils were added to HCAEC together with ET-1[1-32] than in the absence of ET-1[1-32] (Fig. 5). Since neutrophils were incubated in the adhesion assays for 30 min with activated HCAEC, stimulation of neutrophil adhesion by ET-1[1-32] can be attributed primarily to the effects of this peptide on neutrophils. Indeed, much longer periods of culture of HCAEC with ET-1[1-32] were required to detect slight increases in expression of E-selectin and ICAM-1 on HCAEC (18).

Fig. 5

ET-1[1-32] enhances adhesion of neutrophils (PMN) to monolayers of human coronary artery endothelial cells (HCAEC). HCAEC were incubated with LPS (1 μg/ml) for 4 h, washed and PMN without or together with ET-1[1-32] (30 nM) were then added for 30 min at 37°C. Values are means ± SEM (n=5).

No adhesion experiments were performed with neutrophils alone pre-incubated with ET-1[1-32] since this peptide may induce neutrophil aggregation through up-regulation of CD11b/CD18 expression, thereby making interpretation of the results difficult. Our studies with function-blocking monoclonal antibodies revealed that ET-1[1-32]-induced neutrophil adherence to activated HCAEC were predominantly mediated through CD18 integrins and their endothelial counter-ligand, ICAM-1 (18).

Conclusions

Our results showed that two MMPs, gelatinase A and B, via cleavage of big ET-1 could affect expression of adhesion molecules on human neutrophils and consequently promote their adhesion to endothelial cell. Such actions may have relevance to pathological conditions where simultaneous over-expression of gelatinolytic MMPs and big ET-1 can be detected at the sites of tissue injury and inflammation. Thus, gelatinase A and B, often considered as pure effector molecules, may act as amplifiers of leukocyte recruitment, one of the key events of inflammation. Finally, gelatinases constitute a prototypic example of the regulation of innate immune reactions by proteolysis.