INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ANGIOGENESIS, formation of new blood vessels, occurs under various pathological conditions (8). Especially in inflammatory diseases such as wound healing, chronic inflammation, solid tumor formation, and diabetic retinopathy, angiogenesis has been shown to be involved in maintenance of the inflammatory state by transporting inflammatory cells, nutrients, and oxygen to the site of inflammation (15). In fact, inflammatory tissue contains an abundance of inflammatory cells, angiogenic blood vessels, and inflammatory mediators (17, 18). Although the mechanisms of angiogenesis during inflammation remain unclear, monocytes and macrophages activated by inflammatory stimuli have been shown to induce angiogenesis through production of growth factors and cytokines (19, 33). In addition, we recently found (38) that activated polymorphonuclear leukocytes (PMNs) can also stimulate angiogenesis. Thus not only activated monocytes and macrophages but also activated PMNs seem to have important roles in stimulating angiogenesis in inflammatory diseases.

Ets-1 is a transcription factor that regulates the gene expression of proteases such as urokinase-type plasminogen activator (u-PA), matrix metalloproteinase (MMP)-1, MMP-3, and MMP-9 (11, 14, 27, 34). Many studies have shown that Ets-1 mediates angiogenesis. Iwasaka et al. (14) reported that vascular endothelial growth factor (VEGF) and basic fibroblast growth factor (bFGF) induce Ets-1 expression and Ets-1 stimulates angiogenesis by inducing the expression of u-PA and MMP-1. Moreover, Oda et al. (27) reported that overexpression of Ets-1 in vascular endothelial cells induced angiogenesis in vitro. Thus Ets-1 seems to play a central role in angiogenesis.

PMNs activated during inflammation adhere to endothelial cells (2, 36). The adherence of PMNs to endothelial cells is mediated by adhesion molecules such as E-selectin and ICAM-1 expressed in endothelial cells (9, 12, 39). We previously demonstrated (38) that ICAM-1 and E-selectin are involved in the induction of angiogenesis by PMNs because anti-ICAM-1 and anti-E-selectin antibodies inhibited PMN-induced angiogenesis. Recently, adhesion molecules have been reported to act as the signaling receptors that mediate changes in intracellular Ca²⁺ concentration (24) and tyrosine phosphorylation (5). Interestingly, the activation of tyrosine kinase has been reported to be involved in the induction of ets-1 in endothelial cells stimulated by VEGF (30). Therefore, it is possible that the signal transduction from adhesion molecules induces Ets-1 and then stimulates angiogenesis. Alternatively, adhesion molecules may have roles in cell-cell adhesion between endothelial cells in the process of PMN-induced angiogenesis. However, the roles of ICAM-1 and E-selectin in the process of PMN-induced angiogenesis have yet to be elucidated.

In the present study, we found the participation of Ets-1 in PMN-stimulated angiogenesis in bovine aortic endothelial cells (BAECs). Therefore, we investigated the roles of adhesion molecules in the induction of angiogenesis using Ets-1 expression and stimulation of angiogenesis with PMNs.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell culture. BAECs were obtained by scraping the luminal surface with a razor blade and cultured as described previously (37). Endothelial cells were characterized by microscopic observation and incorporation of acetylated low-density lipoprotein labeled with 1,1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanine perchlorate (13). Cells at passages 3-8 were used for the experiments.

Preparation of PMNs. PMNs were collected from male Wistar rats (6-8 wk old; Saitama Animal Supply, Saitama, Japan) as previously described (38). Each rat was injected intraperitoneally with 5 ml of 0.5% oyster glycogen in saline. After 4 h, the rats were injected intraperitoneally with 4 ml of 100 U/ml heparin. The cells infiltrating the abdominal cavity were collected with 50 ml of phosphate-buffered saline (PBS) containing 10% fetal bovine serum (FBS). After centrifugation (170 g) for 10 min at 4°C, the supernatant was discarded and the remaining red pellet was subjected to hypotonic lysis by addition of 0.2% NaCl. After 30 s, the lysate was made isotonic by addition of an equal volume of 1.6% NaCl solution and centrifuged at 170 g for 10 min. The supernatant was discarded, and the residual pellet was washed twice with 10 ml of PBS containing 0.1% FBS. The pellet was then suspended in 2 ml of minimum essential medium (MEM) containing 0.1% FBS. The purity of PMNs was confirmed by May Grünwald-Giemsa staining (>95%).

Tube formation assay. Tube formation was measured in 24-well culture plates with the three-dimensional culture method described in our previous report (38). Collagen gel solution (0.5 ml) consisting of a mixture of 8 volumes of type I collagen solution (Koken, Tokyo, Japan), 1 volume of 10-fold concentrated MEM, 1 volume of 0.05 N NaOH, 200 mM HEPES, and 260 mM NaHCO₃ was poured into each well of the culture plates and incubated for 60 min at 37°C. The BAEC suspension (5 × 10⁵ cells/ml) in 1 ml of MEM containing 10% FBS was added to the wells and cultured. When the cultures reached confluence, the medium was replaced with MEM containing 0.1% FBS. After 48 h, various numbers of PMNs with or without 1 µM N-formylmethionyl-leucyl-phenylalanine (FMLP) were added and incubated for 3 days at 37°C. Mouse anti-human ICAM-1 (CD54) monoclonal antibody (50 µg/ml; Immunotech, Marseille, France) and mouse anti-human E-selectin (CD62E) monoclonal antibody (50 µg/ml; Pharmingen, San Diego, CA) were added 15 min before PMN treatment. The cultures were washed three times with PBS and fixed with 2.5% glutaraldehyde in PBS. Randomly selected fields measuring 0.86 × 1.3 mm were photographed in each well under phase-contrast microscopy. Tube formation was quantified from three randomly selected fields per experiment by measuring the total additive length of all cellular structures including all branches with a computer-assisted image analyzer (MCID; Imaging Research).

Diffusion chamber assay. To examine whether activated PMNs stimulate in vivo angiogenesis, we used a diffusion chamber assay system modified to assess in vivo angiogenesis as previously described (35). The diffusion chamber was made from a chamber kit purchased from Millipore (Bedford, MA). A cellulose membrane filter (0.45 µm, 14-mm diameter) was glued to each side of the ring chamber with MF (Millipore) cement. Male Wistar rats (200-250 g) were anesthetized by intraperitoneal injection of pentobarbital sodium (10 mg/rat). Before chamber implantation, the backs of the animals were depilated and disinfected with tincture of iodine. The chambers containing PMNs or vehicle were implanted into a subcutaneous pocket in the back of the rats. Seven days after implantation, the chambers were removed from the animals and fixed with 10% formalin solution.

Northern blot hybridization. BAECs were grown to 90% confluence in MEM containing 10% FBS and antibiotics, and then the cultures were starved in MEM containing 0.1% FBS for 48 h. PMNs stimulated with or without FMLP were added to the cultures and incubated for various periods. Total RNA was extracted from BAECs by a modified guanidinium isothiocyanate method with ISOGEN (Nippon Gene, Tokyo, Japan). Aliquots of 20 µg of total RNA were separated by electrophoresis through 1% agarose-formaldehyde gels. The RNA was transferred onto Hybond-N nylon membranes (Amersham Pharmacia Biotech, Little Chalfont, UK) and hybridized with the indicated random prime-labeled cDNA probes (Amersham Life Sciences). The rat ets-1 probe was a 1.4-kb BamHI fragment of ets-1 cDNA cloned into the pLXSN plasmid vector. Hybridization was carried out for 1 h at 68°C in ExpressHyb hybridization solution (Clontech, Palo Alto, CA). The membranes were finally washed in a solution containing 1.7 mM NaCl, 1.7 mM sodium citrate, and 0.1% SDS at 50°C for 40 min and exposed to BioMax film (Kodak, Rochester, NY) at 80°C for 48 h. The membranes were stripped and rehybridized with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA, a constitutively expressed gene. The cDNA probe for GAPDH was prepared by reverse transcription-PCR as described previously (32). The primer pairs used for amplification of GAPDH were 5'-TCCACCACCCTGTTGCTGTA-3' and 5'-ACCACAGTCCATGCCATCAC-3'. The PCR product was electrophoresed through a 1.5% agarose gel, and the GAPDH-specific band was extracted with a Qiaex II gel extraction kit (Qiagen K. K., Tokyo, Japan). The signal intensity was quantified with an imaging analyzer (Image Hyper II; DigiMo, Osaka, Japan).

SDS-PAGE and Western blotting. Confluent BAECs in 10-cm culture dishes were starved of serum for 48 h and treated with PMNs stimulated with 1 µM FMLP. The cells were washed twice with ice-cold PBS and lysed in lysis buffer [20 mM Tris · HCl (pH 7.4), 0.1% SDS, 1% Triton X-100, 1% sodium deoxycholate, and 1 mM p-amidinophenylmethanesulfonyl hydrochloride] for 30 min on ice. The cell lysates were centrifuged at 12,000 rpm for 5 min at 4°C. After the supernatants were collected, the protein concentration was determined with a D_C protein assay kit (Bio-Rad Laboratories, Hercules, CA). Samples containing equal amounts of protein (40 µg) were separated on 10% SDS-polyacrylamide gels under reducing conditions and transferred onto Trans-Blot nitrocellulose membranes (Bio-Rad). Nonspecific binding was blocked with 0.2% Aurora blocking reagent (ICN Biomedicals, Costa Mesa, CA) in PBS containing 0.1% Tween 20 for 60 min. The membranes were incubated for 1 h with a 1:1,000 dilution of rabbit polyclonal anti-human Ets-1 (Santa Cruz Biotechnology, Santa Cruz, CA), a 1:1,000 dilution of mouse anti-human ICAM-1 (Zymed Laboratories, San Francisco, CA), or a 1:1,000 dilution of mouse anti-human E-selectin (Pharmingen, San Diego, CA) antibodies and developed with an enhanced chemiluminescence Western blotting detection system (ECL, Amersham Pharmacia Biotech) with horseradish peroxidase (P)-conjugated second antibodies. As the second antibody, a 1:5,000 dilution of P-conjugated goat anti-rabbit IgG (Bio-Rad) for the anti-Ets-1 antibody or a 1:5,000 dilution of P-conjugated goat anti-mouse IgG (Zymed Laboratories) for anti-ICAM-1 and anti-E-selectin antibody was used. The membranes were exposed to chemiluminescence-sensitive film (Hyperfilm, Amersham) for 3-30 s. Densities of signals on the blots were measured with an image analyzer (ImageHyper II).

Statistical analysis. Results are expressed as means ± SE of n observations for each experiment. Statistical analysis was performed with the Bonferroni-Dunn procedure after ANOVA. Differences between means were considered significant at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In vivo angiogenesis induced by PMNs. We previously reported (38) that PMNs stimulate in vitro angiogenesis. To determine whether PMNs induce in vivo angiogenesis, diffusion chambers containing PMNs (1 × 10⁵ cells/ml) were implanted in the backs of rats for 7 days. Typical morphology of PMN-induced angiogenesis is shown in Fig. 1. In the surrounding tissues of control chambers containing saline, newly formed vessels were not observed (Fig. 1A). Implantation of the chamber containing activated PMNs induced the formation of a forestlike network of neomicrovascular vessels. Moreover, membrane hyperplasia and bleeding from the periphery of neovascular vessels were observed (Fig. 1B), suggesting that PMNs can stimulate angiogenesis not only in vitro but also in vivo.

View larger version (86K): [in this window] [in a new window]   Fig. 1.   Typical morphology of polymorphonuclear leukocyte (PMN)-induced angiogenesis formed on the diffusion chambers in the backs of rats. Diffusion chambers containing sterile saline as a control (A and C) and 1 × 105 PMNs (B and D) were put into the backs of rats surgically, and after 7 days the chambers were removed as described in METHODS. The framed areas of tissues in A and B are magnified in C and D, respectively. Black arrows, vessels; white arrow, neovascular tissue; black arrowhead, membrane hyperplasia; white arrowheads, bleeding.

Induction of Ets-1 expression by PMNs. We examined whether PMNs stimulated ets-1 mRNA and/or protein expression in endothelial cells. As shown in Fig. 2A, PMNs (1 × 10⁵ cells/ml) induced ets-1 mRNA expression in BAECs and the activation of PMNs by FMLP additionally increased the ets-1 mRNA expression compared with PMNs alone. However, addition of FMLP to BAECs in the absence of PMNs did not affect ets-1 mRNA expression (Fig. 2A). The induction of ets-1 mRNA expression by activated PMNs was dependent on PMN number at 1 × 10⁴ and 1 × 10⁵ cells (Fig. 2B). To determine the time course of ets-1 mRNA expression, BAECs were exposed to activated PMNs for various periods (0-12 h). The induction of ets-1 mRNA expression started from 1 h after addition of activated PMNs, and the peak was observed at 3 h after addition (Fig. 2C). To further clarify the induction of Ets-1 in BAECs stimulated by PMNs, the level of Ets-1 protein was also examined by Western blotting. The increase in Ets-1 protein was also observed at 3 and 6 h after stimulation with activated PMNs (Fig. 3).

View larger version (28K): [in this window] [in a new window]   Fig. 2.   Induction of ets-1 mRNA expression in bovine aortic endothelial cells (BAECs) stimulated by PMNs. A: BAECs were starved of serum for 48 h and then treated with or without 1 × 105 PMNs/ml in the presence or absence of N-formylmethionyl-leucyl-phenylalanine (FMLP; 1 µM) for 3 h before RNA extraction. B: BAECs were starved of serum for 48 h and then treated with 0, 1 × 104, or 1 × 105 PMNs/ml stimulated with 1 µM FMLP for 3 h before RNA extraction. C: BAECs were starved of serum for 48 h and then treated with 1 × 105 PMNs/ml stimulated with 1 µM FMLP for various periods (0-12 h) before RNA extraction. After electrophoresis of 20 µg total RNA/sample and transfer onto nylon membranes, the blots were sequentially hybridized with 32P-labeled ets-1 cDNA (top) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA (bottom) probes in each assay. Each column indicates the mean ± SE ratio of ets-1 mRNA expression to GAPDH mRNA from 2-4 independent experiments.

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Western blotting analysis of Ets-1 expression stimulated by PMNs. BAECs were starved of serum for 48 h and then treated with 1 × 105 PMNs/ml of stimulated with FMLP (1 µM) for various periods (0-24 h). Aliquots of 40 µg of protein from the BAEC lysate were fractionated by SDS-PAGE and immunoblotted with anti-Ets-1 polyclonal antibody. Each column indicates the mean ± SE density of bands in 2 independent experiments.

Effects of ets-1 antisense oligonucleotide on PMN-stimulated angiogenesis and Ets-1 expression. To investigate whether ets-1 plays a role in PMN-induced angiogenesis, the effects of ets-1 antisense oligonucleotide were examined (Fig. 4). Typical morphological changes of BAECs are shown in Fig. 4, A-C. BAECs cultured with 0.1% FBS formed some tubelike structures (Fig. 4A). Addition of activated PMNs by treatment of BAECs with FMLP markedly enhanced the formation of tubelike structures with a network of branching cellular cords beneath the surface of the monolayer (Fig. 4B). The activated PMN-induced tube formation was inhibited by 3 µM ets-1 antisense oligonucleotide (Fig. 4C). The effects of ets-1 antisense oligonucleotide on activated PMN-induced angiogenesis are summarized in Fig. 4D. Activated PMNs stimulated angiogenesis in BAECs, and the angiogenesis was significantly blocked by ets-1 antisense but not by sense or mismatch oligonucleotides (Fig. 4D). Moreover, the activated PMN-induced ets-1 mRNA and Ets-1 protein expression were significantly decreased by treatment with 3 µM ets-1 antisense oligonucleotide but not by sense or mismatch oligonucleotides (Fig. 5, A and B).

View larger version (64K): [in this window] [in a new window]   Fig. 4.   Effects of ets-1 antisense oligonucleotide on PMN-induced angiogenesis in BAECs. Endothelial cells were cultured on collagen gel in 24-well plates to confluence, and then minimum essential medium (MEM) containing 0.1% FBS and 1 × 105 PMNs/ml stimulated with or without 1 µM FMLP were added to the cells and incubated for 72 h. ets-1 sense, antisense, or mismatch oligonucleotide (all at 3 µM) was added to the BAECs 6 h before addition of PMNs. The sequences of the oligonucleotides of ets-1 were as follows: ATG AAG GCG GCC GTC GAT CT (sense), AGA TCG ACG GCC GCC TTC AT (antisense), and ATG CAC AGC TCC GCC AGG TT (mismatch). The cultures were fixed with 0.25% glutaraldehyde and photographed (original magnification ×100). Photomicrographs show control (A), treatment with activated PMNs (B), and effects of ets-1 antisense oligonucleotide on activated PMN-induced angiogenesis (C). The tubelike structures formed were quantified by measuring the total additive length of all cellular structures including all branches with a computer-assisted image analyzer (D). Results are expressed as the means ± SE of 3 experiments. P <0.05 vs. BAECs alone; *P < 0.05 vs. PMNs with FMLP.

View larger version (33K): [in this window] [in a new window]   Fig. 5.   Effects of ets-1 antisense oligonucleotide on the induction of Ets-1 expression in BAECs stimulated with PMNs. BAECs were starved of serum for 48 h and pretreated with ets-1 sense, antisense, or mismatch oligonucleotide (all at 3 µM) for 6 h. The BAECs were then treated with 1 × 105 PMNs/ml stimulated with 1 µM FMLP for 3 h before total RNA and protein extraction. The sequences of the sense, antisense, and mismatch oligonucleotides of ets-1 are shown in Fig. 4. A: Northern blotting analysis of ets-1 mRNA expression in BAECs. Each column indicates the mean ± SE ratio of ets-1 mRNA expression to GAPDH mRNA from 4 independent experiments. *P < 0.05 vs. PMN with FMLP. B: Western blotting analysis of Ets-1 protein expression in BAECs. Each column indicates the mean ± SE density of bands in 2 independent experiments.

Effects of antibodies to adhesion molecules on ets-1 mRNA expression. We previously reported (38) that FMLP treatment enhanced adhesion of PMNs to BAECs and the adhesion was inhibited by treatment with 1 µM anti-E-selectin and anti-ICAM-1 antibodies. Furthermore, we showed (38) that PMN-induced angiogenesis was strongly inhibited by anti-ICAM-1 and anti-E-selectin antibodies. To confirm the expression of ICAM-1 and E-selectin expression in endothelial cells, immunoblotting for ICAM-1 and E-selectin was performed (Fig. 6). Weak ICAM-1 expression was observed in BAECs under basal conditions, and the addition of activated PMNs to BAECs enhanced ICAM-1 expression from 1 to 6 h after addition (Fig. 6A). E-selectin expression was also enhanced by activated PMNs from 18 h after addition (Fig. 6B).

View larger version (46K): [in this window] [in a new window]   Fig. 6.   Expression of ICAM-1 and E-selectin in BAECs. BAECs were starved of serum for 48 h, and then FMLP (1 µM)-stimulated PMNs were added for various periods (0-24 h). The obtained proteins (40 µg) were fractionated by SDS-PAGE and then immunoblotted with anti-ICAM-1 monoclonal antibody (A) or anti-E-selectin monoclonal antibody (B). Each column indicates the mean ± SE density of bands in 2 independent experiments.

View larger version (72K): [in this window] [in a new window]   Fig. 7.   Effects of antibodies to adhesion molecules on PMN-induced ets-1 mRNA expression in BAECs. BAECs were starved of serum for 48 h and pretreated with 0.01-1 µg/ml anti-E-selectin or anti-ICAM-1 antibody. Subsequently, the BAECs were stimulated with 1 × 105 PMNs/ml stimulated with 1 µM FMLP for 3 h before RNA extraction. After electrophoresis of 20 µg RNA/sample and transfer onto nylon membranes, the blots were sequentially hybridized with 32P-labeled ets-1 cDNA (top) and GAPDH cDNA (bottom) probes. Each column indicates the mean ± SE ratio of ets-1 mRNA expression to GAPDH mRNA from 4 independent experiments. *P < 0.05 vs. control.

Effects of antibodies to adhesion molecules on H₂O₂-induced angiogenesis. We previously reported (37) that addition of H₂O₂ to BAECs enhanced angiogenesis. To determine the roles of ICAM-1 and E-selectin in the induction of angiogenesis by stimulation of endothelial cells without PMNs, the effects of anti-ICAM-1 and anti-E-selectin antibodies on H₂O₂-induced angiogenesis were examined (Fig. 8). H₂O₂-induced angiogenesis was inhibited in a concentration-dependent manner by treatment with anti-E-selectin antibody but not by anti-ICAM-1 antibody (Fig. 8, A and B). Moreover, the expression of ets-1 mRNA induced by H₂O₂ was not inhibited by either antibody (Fig. 9).

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Effects of antibodies to adhesion molecules on angiogenesis induced by H2O2. BAECs were preincubated with or without anti-ICAM-1 (0.01-1 µg/ml; A) or anti-E-selectin (0.01-1 µg/ml; B) monoclonal antibodies for 30 min. After incubation, H2O2 (1 µM) was added to the cultures and incubated for 3 days. Results are expressed as means ± SE of 3 experiments. *P < 0.05 vs. H2O2-stimulated BAEC without antibodies.

View larger version (71K): [in this window] [in a new window]   Fig. 9.   Effects of antibodies to adhesion molecules on H2O2-induced ets-1 mRNA expression in BAECs. BAECs were starved of serum for 48 h and pretreated with 0.01-1 µg/ml anti-E-selectin or anti-ICAM-1 antibody. Subsequently, the BAECs were stimulated with 1 µM H2O2 for 3 h before RNA extraction. After electrophoresis of 20 µg RNA/sample and transfer onto nylon membranes, the blots were sequentially hybridized with 32P-labeled ets-1 cDNA (top) and GAPDH cDNA (bottom) probes. Each column indicates the mean ± SE ratio of ets-1 mRNA expression to GAPDH mRNA from 4 independent experiments.

Effects of superoxide dismutase or catalase on ets-1 mRNA expression in BAECs stimulated with PMNs. To investigate the role of H₂O₂ released from PMNs in stimulation of Ets-1 expression, the effects of catalase and superoxide dismutase (SOD) on ets-1 mRNA expression stimulated by PMN were studied. Activated PMN-induced ets-1 mRNA expression was inhibited by catalase but not by SOD (Fig. 10).

View larger version (53K): [in this window] [in a new window]   Fig. 10.   Effects of superoxide dismutase (SOD) and catalase on ets-1 mRNA expression in BAECs stimulated with PMNs. BAECs were serum-starved for 48 h, and 1 or 10 U/ml SOD or catalase was added. Subsequently, the BAECs were stimulated with 1 × 105 PMNs/ml with 1 µM FMLP for 3 h before RNA extraction. Each column indicates the mean ± SE ratio of ets-1 mRNA expression to GAPDH mRNA from 3 independent experiments. *P < 0.05 vs. control.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our previous study (38) showed that PMNs stimulate angiogenesis in BAECs. However, the mechanisms underlying induction of PMN-induced angiogenesis remained unclear. The initiation of angiogenesis requires digestion of the extracellular matrix via induction of protease activities for endothelial cell migration into the interstitial space (4). Recently, the transcription factor Ets-1, which regulates the gene expression of proteases such as u-PA, MMP-1, MMP-3, and MMP-9, was shown to mediate angiogenesis induced by VEGF and epidermal growth factor (EGF) (14, 27, 34). In the present study, we found that Ets-1 expression in endothelial cells was stimulated by activated PMNs and both PMN-induced angiogenesis and Ets-1 expression were strongly reduced by ets-1 antisense oligonucleotide. Thus Ets-1 also seems to play a central role in PMN-induced angiogenesis in addition to angiogenic growth factor-induced angiogenesis.

PMNs adhere to endothelial cells via adhesion molecules such as ICAM-1 and E-selectin. Adhesion molecules were initially thought to function only in cell adhesion between vascular endothelial cells and leukocytes (3, 6, 16). However, adhesion of PMNs to endothelial cells was reported recently to trigger various physiological changes including an increase in intracellular Ca²⁺ concentration and activation of transcription factor nuclear factor-B (1, 7, 22, 25, 28). Our previous study (38) showed that anti-ICAM-1 and anti-E-selectin antibodies, which inhibited adhesion between PMNs, prevented PMN-induced angiogenesis by endothelial cells. In fact, the expression of ICAM-1 and E-selectin was confirmed on BAECs stimulated by PMNs. Thus both ICAM-1 and E-selectin seem to be essential factors for PMN-induced angiogenesis. Importantly, the activated PMN-induced increase in ets-1 mRNA expression was inhibited by anti-ICAM-1 antibody but not by anti-E-selectin antibody. ICAM-1 but not E-selectin might act as a signaling receptor for the induction of Ets-1. We previously reported (37) that H₂O₂ stimulates angiogenesis through the induction of Ets-1. Interestingly, H₂O₂-induced angiogenesis was inhibited by anti-E-selectin antibody but not by anti-ICAM-1 antibody. Nguyen et al. (26) previously reported that formation of tubelike structures by BAEC cultured on fibronectin-coated plates was inhibited by antibodies to sialyl Lewis^X/A and E-selectin. E-selectin seems to function in capillary morphogenesis via endothelial cell-cell interaction during angiogenesis. These findings indicate that although ICAM-1 and E-selectin are essential factors, they have a different roles in PMN-induced angiogenesis, i.e., ICAM-1 might act as a signaling receptor for induction of Ets-1 expression, and E-selectin might act in formation of tubelike structures via endothelial cell-cell adhesion.

The activated PMN-induced ets-1 mRNA expression was further stimulated by treatment with anti-E-selectin antibody. There are several possible mechanisms that could account for these observations. First, the signal from E-selectin by cell-cell adhesion between endothelial cells during formation of tubelike structures may negatively regulate ets-1 mRNA expression induced by activated PMNs. However, this possibility was excluded by the lack of stimulatory effect of anti-E-selectin antibody on H₂O₂-induced ets-1 mRNA expression, although H₂O₂ induces the formation of tubelike structures. Second, the signal from E-selectin by the interaction between PMN and endothelial cells may negatively regulate ets-1 mRNA expression induced by activated PMNs. In fact, H₂O₂-induced ets-1 mRNA expression was not affected by treatment with E-selectin antibody. Thus future studies are needed to determine the role of E-selectin in PMN-induced ets-1 mRNA expression.

Activated PMNs have been shown to release reactive oxygen species (ROS) including H₂O₂ (11, 21, 23). Our previous studies indicated that H₂O₂ (0.1-10 µM) stimulates angiogenesis via induction of Ets-1 (37) and that PMN-stimulated angiogenesis was inhibited by catalase but not by SOD (38). PMN-induced ets-1 mRNA expression was also inhibited by catalase. Thus H₂O₂ released from PMNs seems to be involved in the stimulation of angiogenesis through the induction of Ets-1 expression. In the present study, we used nonstimulated endothelial cells to investigate the mechanisms underlying activated PMN-induced angiogenesis, although the activation of endothelial cells is also necessary for the interaction with PMNs. Importantly, H₂O₂ has been shown to stimulate the expression of adhesion molecules including ICAM-1 (23, 29). In fact, leukocyte accumulation under inflammatory conditions seems to be mediated by ROS such as H₂O₂ and superoxide (20, 31). The increase of ICAM-1 protein level was observed ~2 h before stimulation of Ets-1 protein level by treatment with activated PMNs. It is possible that PMN-induced Ets-1 expression is mediated by stimulation of ICAM-1 expression induced by H₂O₂ released from PMNs. Future studies are needed to determine the role of H₂O₂ in the regulation of adhesion molecule expression during PMN-induced angiogenesis.

In conclusion, our findings suggest that ets-1, ICAM-1, and E-selectin have critical roles in PMN-induced angiogenesis. ICAM-1 may act as a signaling receptor to induce Ets-1 induction, whereas E-selectin seems to be involved in the formation of tubelike structures via cell-cell interactions between endothelial cells.