| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Institute of Brain Research, University of Tuebingen Medical School, D-72076 Tuebingen, Germany; and
Department of Internal Medicine I, Medical University Clinic Tubingen, Germany
1Correspondence: Institute of Brain Research, University of Tuebingen Medical School, Calwer Str. 3, D-72076 Tuebingen, Germany. E-mail: martin.deininger{at}uni-tuebingen.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
challenge. Endostatin expression and release was reduced by the nitric oxide synthase inhibitors aminoguanidine and L-NAME and induced by the NO synthase-independent NO donors sodium nitroprusside (SNP) and spermine-NONO-ate. SNP-mediated endostatin induction was abrogated by the soluble guanylate cyclase inhibitor 1H-(1.2.4) oxadiazolo (4,3-A) quinoxalin-1-one. Adenoviral endostatin transduction resulted in the release of endostatin from endothelial cells and in down-regulation of iNOS (NOS2) and eNOS (NOS3), and surprisingly in a 10% induction of PCNA. These results describe the modulation of endostatin release by the NO signaling cascade and provide important new pharmacological information for the systemic induction of endogenous endostatin release by common NO donor pharmacotherapy.—Deininger, M. H., Wybranietz, W. A., Graepler, F. T. C., Lauer, U. M., Meyermann, R., Schluesener, H. J. Endothelial endostatin release is induced by general cell stress and modulated by the nitric oxide/cGMP pathway.
Key Words: anti-angiogenic therapy • glioblastoma pathology • nitric oxide synthase
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
. Endostatin administration is nontoxic, does not induce acquired drug resistance, and is considered a potent new therapy strategy in solid neoplasias (4)
. Although alternatively spliced endostatin variants have been described in a wide range of human tissues, including liver, heart, kidney, placenta, prostate, ovaries, skeletal muscle, small intestine, and others, their biological function remains unresolved (5
6
7
8)
. Biochemical studies revealed that the ability of endostatin to inhibit neoangiogenesis at least in part is mediated by Zn2+ binding and elastase processing (9
10
11)
, and widespread endostatin expression was found in elastic fibers in vessel walls (12)
. Accordingly, a global role of endostatin in angiogenesis control has been suggested (13)
. Endostatin expression in endothelial cells of brain neoplasia and signaling cascades involved in endostatin up-regulation and function remain largely unknown. Recent reports revealed that endostatin binds to cell surface glypicans (14)
and tropomyosin (15)
. Exogenously added endostatin induces down-regulation of a wide range of immediate early genes, cell cycle-related genes, and genes regulating apoptosis inhibitors, mitogen-activated protein kinases, focal adhesion kinase, G-protein-coupled receptors, and cell structure components (16)
. Among others, exogenously added endostatin elicits Ca2+ release, a common intracellular signaling molecule that itself is capable to trigger gene expression mechanisms (17)
. In a previous study, we observed endostatin immunoreactivity in 10% of all endothelial cells in human glioblastoma (18)
.
To investigate mechanisms that influence endostatin release from endothelial cells, in vitro experiments were used to analyze endostatin expression and release in the four endothelial cell lines (human SVHCEC, mouse MS1, mouse SVR, and rat YPEN-1) by RT-PCR, flow cytometry, and Western blot after the common pathophysiological stimulations hydrogen peroxide (H2O2) and hypoxia (CoCl2). We hypothesized that the nitric oxide signaling cascade is involved in the regulation of endostatin expression. We therefore analyzed the effect of the chemical NO donors sodium nitroprusside (SNP) and spermine-NONO-ate, inhibitors of nitric oxide synthase function aminoguanidine (AG) and N(G)-nitro-L-arginine methyl ester (L-NAME) and of the NO inducer interferon- (IFN-) on endothelial endostatin expression and release. Involvement of the soluble guanylate cyclase pathway was analyzed by preincubation of endothelial cells with the inhibitor 1H-(1.2.4) oxadiazolo (4,3-A) quinoxalin-1-one (ODQ) before sodium nitroprusside challenge. Adenoviral endostatin transduction was used to identify feedback loops and reveal functional properties of intracellular endostatin overexpression.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
Cloning and production of recombinant mouse endostatin and control peptides
cDNA encoding the carboxyl-terminal endostatin fragment of mouse collagen XVIII was amplified by PCR from liver RNA and subcloned into a pET expression vector (Angewandte Gentechnologie Systeme, Heidelberg, Germany). Recombinant endostatin was produced in BL21(DE3) Escherichia coli cells (AGS, Heidelberg, Germany) by induction with IPTG for 4 h. Recombinant endostatin was purified by nickel-chelate chromatography. Protein concentration was determined by the Bradford assay with bovine serum albumin (BSA) as a standard (Bio-Rad, Munich, Germany). Recombinant control peptides were produced and described previously (19)
.
Production of monoclonal antibody
For generation of monoclonal antibodies, Balb/c mice were immunized with 50 µg of recombinant endostatin and hybridoma cell lines were established by standard procedures. Cell culture supernatants were screened by ELISA, Western blot, and immunohistology and positive clones were subcloned. The antibody was affinity purified (BMA Biomedicals, Augst, Switzerland), again tested for specificity, adapted to a concentration of 1 µg/µL, and used in all consecutive experiments.
Immunocytochemistry
Glioblastoma tumor specimens were fixed in 4% buffered paraformaldehyde, pH 7.5 overnight, dehydrated, and embedded in paraffin. Approximately 4 µm sections were cut and rehydrated. Rehydrated sections were boiled (in a 600W microwave oven) three times for 5 min in citrate buffer (2.1 g sodium citrate/L, pH 6). Endogenous peroxidase was inhibited with 1% H2O2 in methanol for 15 min. Sections were incubated with 10% nonimmune porcine serum (Biochrom, Berlin, FRG) to block nonspecific binding of immunoglobulins. The primary mouse anti-endostatin antibody or the mouse anti-CD31 antibody (endothelial cells) was added both at a concentration of 10 µg/mL in 10% BSA/TBS (tris-balanced saline, pH 7.5) and applied overnight at 4°C. Antibody binding was detected by biotinylated rabbit anti-mouse IgG F(ab)2 antibody fragment (1:400 for 30 min, DAKO, Hamburg, FRG), followed by incubation with a peroxidase-conjugated streptavidin-biotin complex (DAKO, Hamburg, FRG). The enzyme was visualized with diaminobenzidine as a substrate (Fluka, Neu-Ulm, FRG). Sections were counterstained with Mayer’s hemalaun. Controls included absence of the primary antibody, irrelevant mAbs, and blocking experiments.
All probes were double labeled with endostatin and CD31. In double labeling experiments, slices were pretreated as above. Then the antibody directed against CD31 (macrophages; Dakopatts, Glostrup, Denmark) was added at a concentration of 10 µg/mL in BSA-TBS and applied overnight at 4°C. Visualization was achieved by biotinylated rabbit anti-mouse IgG diluted, alkaline phosphatase-conjugated AB complex, and Fast Blue BB salt chromogen substrate solution development. Slices were once more irradiated in a microwave, incubated with nonspecific porcine serum, and the primary mouse anti-endostatin antibody was labeled as above.
The total number of endostatin/CD31-label vessels was counted and compared with the total number of CD31-label vessels in three 200x magnification fields of randomly chosen areas of solid tumor growth.
Cell culture and stimulation
The human SVHCEC brain endothelial cell line derived from brain endothelial cells was a kind gift from Professor A. Muruganandam and has been described before (20)
. Mouse MS1, mouse SVR, and rat YPEN-1 endothelial cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). MS1 is a pancreatic islet endothelial cell line transduced with a temperature-sensitive SV40 large T antigen (tsA-58-3) construct. The line retains many properties of endothelial cells, including uptake of acetylated LDL and expression of both factor VIII-related antigen and VEGF receptor. SVR is a derivative of MS1. MS1 cells were transduced with a retrovirus encoding H-ras and hygromycin resistance. The line causes death by hemorrhage and anemia with thrombocytopenia in nude mice. The YPEN-1 cell line was derived from prostate cells of 8-wk-old Copenhagen male rats. Cells were immortalized by infection with adenovirus12 SV40 hybrid virus. YPEN-1 cells express integrin a6 1 and integrin 3 on their plasma membrane and demonstrate tube formation in Matrigel.
All cells were raised in RPMI 1640 medium with Glutamax II (Gibco BRL, Paisley, UK) containing 10% fetal calf serum (FCS, Gibco BRL), 100 units/mL penicillin, and 100 µg/mL streptomycin (Fluka, Buchs, Switzerland) at 37°C and 5% CO2.
In stimulation experiments, cells were washed twice and resuspended in serum-free medium. Two mM cobalt chloride (CoCl2, Sigma, Deisenhofen, Germany) was added for 24 h to mimic hypoxia (21)
, or 20 mM H2O2 (Merck, Darmstadt, Germany) for 15 min to mimic reactive oxygen challenge (22)
. In screening experiments, cells were incubated for 24 h in serum-free medium in time-dependent up-regulation experiments for the indicated times. Two mM sodium nitroprusside (SNP, Sigma), a NO synthase-independent NO donor (23)
, was added to the cells in serum-free medium for 4 h to evaluate a causal relationship between NO and endostatin expression. In inhibition experiments, cells were incubated before stimulation with 5 mM aminoguanidine (AG, Sigma; 24) or 1 mM N(G)-nitro-L-arginine methyl ester (L-NAME, Sigma) for 24 h (25)
. Titration experiments were used before the shown results to test for superior concentrations. Cells were stimulated with 20 mM H2O2 as described above and left in serum-free medium for 2 h. Specificity of 2 mM SNP-mediated endostatin induction was once more tested at a concentration of 100 µM SNP and by spermine-NONOate (Sigma-Aldrich, Darmstadt, Germany) incubation at a concentration of 0.2 µM. cGMP dependence of the observed endostatin up-regulation was analyzed using preincubation of the cells for 24 h with the soluble guanylate cyclase inhibitor ODQ (Sigma-Aldrich) at a concentration of 20 µM and consequent 100 µM SNP challenge. Furthermore, 400 U/mL IFN- (Peprotech, Rocky Hill, NJ, USA) was applied for 4 h in serum-free medium. For further analyses, cells were trypsinized, centrifuged, and processed as indicated below.
Flow cytometry
Cells were trypsinized, washed twice, and either permeabilized with methanol for 15 min at -20°C or analyzed unfixed and unpermeabilized. Cells were stained with mouse-anti endostatin monoclonal antibody at a concentration of 10 µg/mL in PBS-BSA for 1 h at 4°C. Visualization was achieved by adding FITC-conjugated rabbit-anti mouse IgG (Serotech, Oxford, UK) for 30 min. Cells were washed and analyzed using a FACScan Cytometer (Becton Dickinson, Überlingen, Germany) and CellQuest software. Controls included preparations lacking the primary antibody or irrelevant primary antibodies.
RNA extraction and reverse transcriptase polymerase chain reaction (RT-PCR)
Cells were collected and total RNA extraction was performed using the RNeasy® Mini Kit protocol as suggested by the manufacturer (Qiagen GmbH, Hilden, Germany). cDNA was prepared using random hexamers, RNAsin® RNase inhibitor and Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI, USA). Samples were incubated 30' at 37°C and 5' at 95°C. Polymerase chain reaction was then performed using Thermus aquaticus (Taq) DNA Polymerase Mini Kit (Promega) according to the manufacturer’s instructions. The forward and reverse primers for endostatin were 5'-TGCCCAGCTCCTGGCCCGCCGCTT-3' and 5'-GTGCATCAACACAGGCGCCTCTTC-3' and gave a single 306 bp fragment of human endostatin cDNA. ß-Actin forward and reverse primers 5'-TCACCCTGAAGTACCCCATCGAG-3' and 5'-TTGGCCTTGGGGTTCAGGGGGG-3', respectively, yielding a 150 bp fragment in human ß-actin cDNA, were used as controls. Reaction conditions were denaturation at 94°C for 2 h, 30 cycles of denaturation at 94°C for 1 h, annealing at 60°C (endostatin), 55°C (ß-actin), elongation at 72°C for 1 h. Finally, the reaction was incubated at 72°C for 10 h. Reactions were analyzed on agarose gels and sublconed for sequencing. The cDNA concentrations were normalized to yield equivalent actin PCR products to allow for comparison of endostatin mRNA expression. Controls included experiments lacking the polymerase and revealed no detectable bands.
Protein preparation and Western blot
Cells were lysed in a buffer containing 125 mM Tris base, 20% glycerol, 2% SDS, 1% bromophenol blue, 2% 2-ME, and protease inhibitors. Supernatants were precipitated by acetone and resuspended in running buffer. All samples were sonicated and boiled. Approximately 30 µg of total protein was loaded per lane, electrophoresed on a 12% SDS-polyacrylamide gel, and transferred to PVDF membranes (Bio-Rad, Munich, Germany) by semi-dry blot. Membranes were blocked with FCS and the primary monoclonal antibody directed against endostatin or antibodies directed against iNOS (NOS2), eNOS (NOS3), or nNOS (NOS1) (Santa Cruz Biotech, Santa Cruz, CA, USA, 26) were visualized using HRP-conjugated avidin-biotin complex and ECL visualization. Controls included blocking experiments lacking primary antibody and irrelevant antibodies.
Construction of a recombinant adenovirus expressing murine endostatin.
The adenoviral endostatin and control vectors have been described before (27)
. The cDNA of murine endostatin was PCR cloned from 2-wk-old C57BL/6 mouse liver. The 18 amino acid E3/19K signal sequence (MRYMILGLLALAAVCSAA) was inserted upstream from the endostatin sequence by PCR. Plasmid DNA was amplified in DH5 cells (Life Technologies) and the signal sequence murine endostatin (ss-mEndo) sequence was confirmed (ABI Prism 310 autosequencer; PE Applied Biosystems, Foster City, CA, USA). The ss-mEndo construct was digested with EcoRI and cloned by blunt-end ligation into the multiple cloning site of the adenoviral shuttle plasmid pAd/CMV.1. The resulting plasmid was recombined with type 5 E1A/B-deleted Ad2 and used to infect 293 cells (American Type Culture Collection, Manassas, VA). Plaque DNA was extracted using proteinase K digestion, phenol extraction, and ethanol precipitation and screened for ss-mEndo by PCR. The resulting virus, Ad-ss-mEndo, was amplified in 293 cells. A similar strategy was used to create control recombinant viruses containing the genes for GFP (green fluorescent protein) and a null virus without endostatin or GFP sequences. Viruses were titered using a standard plaque-forming assay in 293 cells.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Endostatin/collagen XVIII was observed only in singular disseminated (mean=3.45, SE=2.069) endothelial cells (Fig. 1A-D
). Endostatin/collagen XVIII immunoreactivity was observed in macrophages/microglial, astrocytic, immunological, and neuronal cell populations. Endothelial endostatin/collagen XVIII immunoreactivity was observed in disseminated vessels with no focal accumulation in areas of infiltrative, solid, or necrotic tumor growth in arterioles and venules of the neovasculature, a common pathophysiological hallmark of glioblastoma multiforme, and in the vasculature of adjacent areas of infiltrative tumor growth, which did not show pathological alterations. We observed that endothelial cells of vessels surrounded by an accentuated inflammatory infiltrate were more frequently labeled by endostatin/collagen XVIII. RT-PCR with specific primers was used to ensure that endostatin/collagen XVIII mRNA is expressed in four glioblastoma patients in vivo and not merely taken up from the bloodstream (Fig. 1E
).
|
To determine whether the in vivo-labeled molecule is an endostatin-containing collagen XVIII fragment or the 20 kDa endostatin, we analyzed four previously characterized endothelial cell lines (human SVHCEC, mouse MS1 and SVR, and rat YPEN-1 cells) with flow cytometry (upper), RT-PCR (below), and Western blot of cellular lysates (below) and supernatants (lower) (Fig. 1F
). We observed endostatin/collagen XVIII mRNA only in SVR cells. Western blot of endothelial cellular lysates revealed no immunoreactive bands. In SVR supernatants, however, we detected release of the 20 kDa endostatin. Some endostatin release was observed in SVHCEC supernatants.
Endothelial endostatin release is up-regulated by hydrogen peroxide (H2O2) and hypoxia (CoCl2)
We then analyzed endothelial endostatin/collagen XVIII expression and release after preincubation of line cells with an in vitro model of reactive oxygen challenge, H2O2 (Fig. 2A
), and hypoxia, CoCl2 (Fig. 2B
), by flow cytometry (upper) and Western blot of cellular lysates (middle) and supernatants (lower). After H2O2 challenge, we observed prominent endostatin/collagen XVIII induction by flow cytometry in all analyzed cell lines. Western blot of cellular lysates shows predominantly induction of endostatin in YPEN-1 cells. In the supernatants, induction of endostatin release was most prominently observed from SVHCEC and YPEN-1 cells.
|
After CoCl2 challenge, we again detected endostatin/collagen XVIII induction in all analyzed cell lines by flow cytometry. Western blot revealed induction of the 20 kDa endostatin in YPEN-1 cellular lysates and weaker but in part heavier bands in SVR cellular lysates. Only weak induction of endostatin release was observed in endothelial cell supernatants.
To provide more information about the kinetics of H2O2-induced endostatin/collagen XVIII up-regulation, sequential flow cytometry (upper), RT-PCR (below), and Western blot of SVHCEC cellular lysates (below) and supernatants (lower) were performed 1, 2, 3, and 4 h after H2O2 challenge (Fig. 3A
). We observed a time-dependent up-regulation of endostatin/collagen XVIII up to 4 h poststimulation using flow cytometry. Induction of endostatin/collagen XVIII mRNA peaked at t = 2 h poststimulation. In SVHCEC cellular lysates, only weak bands of 
50 kDa were observed throughout the stimulation period. In SVHCEC supernatants; in contrast, we observed prominent induction of endostatin release that peaked at t = 3 h poststimulation.
|
Endostatin release is blocked by NO synthase inhibitors and induced by a NO donor in SVHCEC endothelia
We then analyzed the effect of two inhibitors of NO synthase and one chemical, NO synthase-independent NO donor, on endostatin/collagen XVIII expression and release using flow cytometry (upper) and Western blot of SVHCEC cellular lysates (below) and supernatants (lowest). SVHCEC cells were incubated with inhibitors of NO up-regulation—aminoguanidine (AG), an incomplete inhibitor of iNOS (NOS2), and L-NAME, a NO synthase inhibitor that is considered to inhibit all three NO synthase isoforms—for 24 h prior to H2O2 challenge or with sodium nitroprusside for 4 h (Fig. 3B
). Flow cytometry revealed prominent reduction of endostatin/collagen XVIII labeling after L-NAME, but not aminoguanidine treatment and induction by sodium nitroprusside. Western blot of cellular lysates revealed even more accentuated endostatin/collagen XVIII after L-NAME and aminoguanidine preincubation. However, in SVHCEC supernatants, L-NAME reduced and sodium nitroprusside induced release of 20 kDa endostatin. Sodium nitroprusside challenge resulted in the release of 40 kDa endostatin/collagen XVIII proteins. Parallel Western blot of SVHCEC cellular lysates with antibodies directed against the three nitric oxide synthase isoforms revealed prominent reduction of nNOS (NOS1) by aminoguanidine and L-NAME pretreatment and weaker iNOS (NOS2) and eNOS (NOS3) expression (Fig. 3B
). After preincubation with both NO synthase inhibitors, considerable eNOS (NOS3) expression was still observed, probably accounting for the incomplete inhibition of endostatin release. In sodium nitroprusside challenged cells, no NO synthase immunoreactive bands were detected.
Aminoguanidine, L-NAME, and sodium nitroprusside modulate endostatin/collagen XVIII labeling in MS1, SVR, and YPEN-1 cells
To show that endostatin/collagen XVIII modulation by aminoguanidine, L-NAME, and sodium nitroprusside is not confined to SVHCEC cells, we used flow cytometry to analyze endostatin/collagen XVIII labeling in MS1, SVR, and YPEN-1 cells (Fig. 4A
). Compared with H2O2-stimulated cells (Fig. 2A
), we observed prominent reduction of endostatin/collagen XVIII labeling in all analyzed cell lines by aminoguanidine. L-NAME resulted in reduced endostatin/collagen XVIII labeling only in SVR and YPEN-1 cells, not in the MS1 endothelial cell line. Compared with unstimulated endothelial cell lines (Fig. 1F
), we observed endostatin/collagen XVIII induction by the NO donor sodium nitroprusside only in SVR cells.
|
Endostatin release is induced by IFN-, another NO donor spermine-NONOate, and inhibited after preincubation with the soluble guanylate cyclase inhibitor ODQ
To provide more evidence for the data described above, we incubated SVHCEC cells with IFN-, another NO donor, spermine-NONOate, and lower concentrations of sodium nitroprusside (100 µM). To evaluate the involvement of the soluble guanylate cyclase in endostatin up-regulation, we used preincubation of ODQ before SNP challenge (Fig. 4B
). Compared with unstimulated SVHCEC cells (Fig. 1F
), Western blot of SVHCEC cellular lysates (upper) and their supernatants (lower) revealed induction of intracellular endostatin/collagen XVIII and extracellular endostatin release by IFN-. These findings are not surprising because IFN--induced NO synthase induction is a well-known consequence that might account for the induction of endostatin release. Induction of endostatin release was also observed after spermine-NONOate preincubation, a less toxic NO donor than sodium nitroprusside, providing evidence for the specificity of the above-demonstrated data. Preincubation with 100 µM sodium nitroprusside also showed the induction of endostatin release, providing evidence for the above-named data. Preincubation with ODQ before sodium nitroprusside challenge resulted in the reduction of both, intracellular expression of endostatin/collagen XVIII and release of endostatin, suggesting that soluble guanylate cyclase is prominently involved in endostatin up-regulation.
Adenoviral transduction causes release of endostatin, down-regulation of iNOS (NOS2) and eNOS (NOS3), and induction of PCNA in SVHCEC endothelia
Adenoviral endostatin transduction of SVHCEC cells was used to identify possible alterations in NO synthase and PCNA expression by intracellular endostatin expression; 48 h after endostatin transduction (Fig. 5A
), flow cytometry (left) revealed prominent GFP (green fluorescent protein) and endostatin labeling. Western blot confirmed accentuated endostatin release to the endostatin but not the null transduced SVHCEC supernatants (right).
|
The number of cells expressing the proliferating cell nuclear antigen (PCNA) increased after endostatin transduction (Fig. 5B
). To analyze a possible feedback loop in endostatin up-regulation, we analyzed eNOS (NOS3) and iNOS (NOS2) expression in endostatin and null transduced cells and observed a reduction of eNOS (NOS3) expressing cells from 15.8% to 2.33% and of iNOS (NOS2) expressing cells from 36.24% to 2.63% after endostatin transduction.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Endostatin is an anti-angiogenic 20 kDa carboxyl-terminal fragment of collagen XVIII that specifically inhibits endothelial proliferation and migration in vitro and potently inhibits angiogenesis and tumor growth in vivo by inducing apoptosis in endothelial cells (1
, 2
, 4)
. Using a monoclonal antibody directed against endostatin, we observed that endothelial endostatin/collagen XVIII expression in glioblastoma patients is rare. Limited endostatin expression is of note because both induction of endothelial endostatin expression and endostatin gene therapy, shown to be effective in inhibiting tumor growth and formation of metastases in mouse models of neoplasia, might have beneficial effects on tumor growth in these patients (29)
. A major drawback of adenoviral gene therapy strategies in humans are adverse reactions with live-threatening consequences. In turn, addition of exogenously produced endostatin to patients does not induce tumor regression as observed in animal models. Therefore, we searched for alternative methods of induction of endostatin release from endothelial cells. Cerebral endothelial cells play a pivotal role in the neovascularization process during glioblastoma growth. Moreover, endothelial neoplasia is a hallmark alteration in glioblastoma patients. We observed that endostatin is up-regulated in mouse MS1, human SVHCEC, mouse SVR, and rat YPEN-1 endothelial cells by H2O2, a model of reactive oxygen species generation, and CoCl2, a model of hypoxia. These data show a novel and so far unrecognized signaling cascade involved in endostatin up-regulation that might apply not only to glioblastoma, but also for other diseases of the brain and other organs, and provide important new data that support the hypothesis of the previously suggested close opposing effects of endostatin and VEGF.
A recent report demonstrated that exogenously added endostatin elicits an increase in intracellular Ca2+ (17)
. These results are of note because Ca2+ is a downstream mediator of G-protein signaling that leads to the induction of NO synthase. Little is known about signaling cascades involved in endostatin up-regulation. Previous work showed that endostatin causes a block at one or more steps in vascular endothelial growth factor (VEGF) -induced cellular migration, and VEGF in turn can cause a block of the inhibition by endostatin of VEGF-induced migration of endothelial cells (3)
. VEGF is up-regulated in hypoxic, perinecrotic regions of glioblastoma growth, stimulated by H2O2 and hypoxia, and modulated by nitric oxide (30
31
32)
. Interactions of endostatin and reactive oxygen species (ROS) have been suggested. Endostatin was found to induce tyrosine kinase activity and the formation of multiprotein signaling complexes in endothelial cells (33)
. These findings are important because tyrosine kinase-induced NO formation is well characterized (34)
. H2O2 induces iNOS (NOS2) and cellular H2O2 production is NO dependent (35)
. NO is produced by a wide variety of cell types by both constitutive and inducible NO synthases and has many physiological functions ranging from regulation of vascular tone to neurotransmission and modulation of inflammatory processes.
We observed prominent nNOS (NOS1) and eNOS (NOS3) and lesser iNOS (NOS2) expression after H2O2 challenge in an endothelial cell line. Accordingly, inhibition of endostatin up-regulation was partly promoted by the NO synthase inhibitor aminoguanidine, a more powerful inhibitor of iNOS (NOS2) than of eNOS (NOS3), and completely abrogated by L-NAME, a more powerful inhibitor of eNOS (NOS3) than of iNOS (NOS2). These findings are significant because they provide first evidence for the involvement of the nitric oxide synthase pathway in endostatin regulation. Moreover, detailed pharmacological insights for clinical therapy are provided. However, considerable but not complete inhibition of endostatin induction was observed, suggesting additional signaling cascades to be involved in endostatin modulation. SVHCEC endothelial cells still expressed considerable amounts of eNOS (NOS3) after aminoguanidine and L-NAME pretreatment, which might account for the incomplete endostatin down-regulation after NO synthase inhibitor preincubation. To further detail the relationship between NO generation and endostatin up-regulation, we applied the NO synthase-independent NO donors sodium nitroprusside and spermine-NONO-ate and observed prominent endostatin up-regulation. These data demonstrate endostatin induction downstream of NO generation and provide convincing evidence for the induction of endostatin by NO.
We then detected complete abrogation of the observed NO-induced endostatin up-regulation by preincubation of the cells with ODQ, an inhibitor of soluble guanylate cyclase, the primary receptor for intracellular NO. Guanylate cyclase is involved in signal transduction pathways in which cGMP acts as a second messenger in several types of cells (36
, 37)
. In endothelial cells, sGC is activated by NO produced by constitutively expressed nitric oxide synthase. The binding of NO induces a several 100-fold increase of the sGC catalytic activity, leading to cGMP production from GTP. The second messenger cGMP then triggers the downstream events leading to smooth muscle relaxation. cGMP is considered one of the main mediators of anti-angiogenic signaling (38)
. The activation of VEGF by cGMP described previously is of considerable interest; taken together with the above-demonstrated data, it provides more evidence for the close interactions of VEGF and endostatin (39)
.
Retroviral endostatin transduction experiments showed the down-regulation of eNOS (NOS3) and iNOS (NOS2) and induction of PCNA expression in endothelial cells. It remains a matter of debate whether NO induces or reduces endothelial proliferation. The individual NO synthase isoforms have been attributed different roles in the promotion of endothelial survival. Although eNOS (NOS3) and iNOS (NOS2) are thought to be survival factors (40
, 41)
, induction of endothelial apoptosis by NO challenge is well known and the role of NO and nitric oxide isoforms in the regulation of endothelial survival is still a matter of debate (42)
. These observations most likely were derive from the downstream induction of highly specific proteins from endothelial cells, which in turn are both mediators and inhibitors of endothelial proliferation. Therefore, identification of factors that modulate either endothelial survival or apoptosis has come into focus. Our findings support the hypothesis of endothelial apoptosis downstream of NO formation, because endostatin has prominent proapoptotic effects on endothelial cells. These findings not only confirm the prominent role of the NO synthase signaling cascade in endostatin up-regulation, but show a negative feedback loop of endostatin to reduce its own expression (Fig. 5C
). Previous results had shown prominent NO synthase expression in glioblastoma. iNOS (NOS2) was observed in reactive glial cells, infiltrating tumor cells, and the endothelium of small blood vessels in edematous peritumoral tissue; more eNOS (NOS3) expressing endothelial cells were observed in glioblastomas than in low-grade astrocytomas (43)
.
Most notable is the induction of PCNA expression after adenoviral endostatin transduction. These data are in apparent conflict with data published earlier showing that extracellular admission of endostatin to endothelial cells results in decreased PCNA expression and even apoptosis. However, PCNA is a homotrimeric protein that forms a sliding clamp around DNA and functions as a DNA polymerase processive factor during replication and nucleotide excision repair. PCNA was discovered as an antigen found only in the nucleus of dividing cells, from which it derives its name (44)
. It was identified as a protein with elevated levels during S phase of the cell cycle (45)
. PCNA was then identified as an essential factor in DNA replication (46)
. Through its multiple protein–protein interactions, PCNA was later identified to coordinate events in replication, epigenetic inheritance, repair, and cell cycle control (47)
. Although intracellular endostatin has functions on cell cycle progression opposite from extracellular endostatin supplementation, the increased PCNA expression may as well reflect altered interactions of PCNA in the nucleus with a wide range of other proteins by intracellular endostatin overexpression. Another recent observation described the induction of migration in subsets of endothelial cells derived from hemangiomas (48)
. Previous results showed that a range of genes including PDGF-Rß, FGFR-4, and FLT-4 are mutated in hemangiomas (49)
. It was therefore suggested that alterations in signaling pathways downstream of the VEGF receptor as a result of a somatic mutation could convert the inhibitory effect of endostatin to one of stimulation, a hypothesis that might also attribute to SVHCEC cells, which are SV40 transformed (20)
. On the other hand, endostatin itself or its promotor may be mutated in these cells, resulting in abnormal proliferation in general and after addition of native endostatin, a condition that has been suggested as causative role in the development of Knobloch syndrome, recently (50)
.
In conclusion, our data reveal that endothelial endostatin expression and release are governed by the nitric oxide/cGMP pathway and provide convincing pharmacological evidence that its induction may be a promising alternative therapy strategy in patients with glioblastoma brain tumors using commonly applied NO donors.
|
ACKNOWLEDGMENTS |
|---|
Received for publication January 30, 2003. Accepted for publication March 10, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Suhr, K. Brixius, M. de Marees, B. Bolck, H. Kleinoder, S. Achtzehn, W. Bloch, and J. Mester Effects of short-term vibration and hypoxia during high-intensity cycling exercise on circulating levels of angiogenic regulators in humans J Appl Physiol, August 1, 2007; 103(2): 474 - 483. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Wenzel, A. Schmidt, K. Reimann, J. Hescheler, G. Pfitzer, W. Bloch, and B.K. Fleischmann Endostatin, the Proteolytic Fragment of Collagen XVIII, Induces Vasorelaxation Circ. Res., May 12, 2006; 98(9): 1203 - 1211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Thum and J. Bauersachs Spotlight on endothelial progenitor cell inhibitors: short review Vascular Medicine, May 1, 2005; 10(2_suppl): S59 - S64. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
