In highly vascular malignant glioma, glioma cells themselves may express angiogenic factors and induce angiogenesis. Recent studies have shown that novel angiogenic factors, angiopoietin-1 (Ang1) and -2 (Ang2), play important roles in the modulation of vasculogenesis and angiogenesis. In this study, we determined Ang2 mRNA expression in cultured human malignant glioma cells (U105, U251, and U373 MG) by reverse transcriptase-PCR. Western blot analysis and immunocytochemical analysis with antihuman Ang2 antibody revealed that Ang2 protein was expressed and secreted by these cells. Furthermore, hypoxia increased the Ang2 protein level in cultured glioma cells. Serial sections of 32 human glioma tissues (14 glioblastomas, eight anaplastic astrocytomas, seven astrocytomas, and three pilocytic astrocytomas) were immunostained against Ang2, vascular endothelial growth factor, Tie2, von Willebrand factor, and α smooth muscle actin. The immunoreactivity of each angiogenic factor was higher in malignant gliomas than in low-grade gliomas. Ang2 protein was detected not only in endothelial cells but also in glioma cells, and its expression was prominent in both the area surrounding the necrosis and the periphery of glioblastomas. In the area surrounding necrosis, Ang2 was highly expressed and tumor vessels showed regression. In the tumor periphery, Ang2 was highly expressed and many small vessels stained positively for von Willebrand factor but not for α smooth muscle actin, suggesting angiogenesis. Statistical analysis revealed that the Ang2 expression was negatively correlated with vessel maturation in malignant gliomas and that vascular endothelial growth factor expression was positively correlated with vessel maturation in low-grade gliomas (P < 0.05). These results suggest that glioma cells themselves express Ang2 and that expression may be induced by hypoxic stimulation and may play a crucial role in the vessel maturation, angiogenesis, and vessel regression in malignant glioma.

Angiogenesis is an important phenomenon in physiological events, such as female reproductive processes (1), and in pathological events underlying the growth of solid tumors, such as rheumatoid arthritis and in proliferative diabetic retinopathy (2, 3, 4). Various angiogenic activators and inhibitors have been characterized for their roles in angiogenesis. These findings suggest that changes in the relative balance of activators and inhibitors can switch angiogenesis on or off, leading to the growth of new blood vessels (4). Angiogenic activators include VEGF,3 basic fibroblast growth factor, acidic fibroblast growth factor, and so on (4). Ang1 and Ang2 have been identified as novel angiogenic factors and as ligands of the EC-specific tyrosine kinase receptor Tie2 (5, 6). Ang1 and Ang2 exhibited considerable sequence homology and similar binding affinity for Tie2 (6). Although Ang1 induces autophosphorylation of Tie2, Ang2 does not. Rather, it competitively inhibits Ang1 binding to Tie2, acting like a natural antagonist (6). Gene targeting and transgenic studies have shown that both Angs and Tie2 were critical for normal vascular development (5, 6, 7, 8, 9). With respect to angiogenesis in adult tissues or tumor vessels, both Angs act as synergistic factors of VEGF. Ang1 and Ang2 enhanced VEGF-induced angiogenesis in vivo in a corneal micropocket assay, but neither Ang1 nor Ang2 alone stimulated angiogenesis (10). In vitro, in the presence of VEGF, Ang1 synergistically stimulates blood vessel sprouting and branching (11). In the rat ovary, Ang2 promotes vessel sprouting in the presence of abundant VEGF, whereas in its absence Ang2 contributes to vessel regression (6). Ang1 contributes to recruitment and maintenance of the association of periendothelial supporting cells and the matrix, resulting in the stabilization of vessel structures (10, 12). On the other hand, Ang2 initially disrupts the interactions between ECs and periendothelial support cells (13, 14, 15). These findings suggest that Ang2 promotes angiogenesis via regulation of Tie2 phosphorylation on ECs in a stage-specific manner. Thus, it may convert ECs and periendothelial support cells into more active phenotypes and initialize the Ang1/VEGF-induced angiogenesis.

Gliomas in the form of astrocytomas, anaplastic astrocytomas, and glioblastomas are the most common brain tumors in humans (16). Malignant gliomas such as anaplastic astrocytomas and glioblastomas are characterized by endothelial proliferation and prominent vascularization. VEGF plays a crucial role in glioma neovascularization, and other known angiogenic factors are commonly expressed by glioma cells (17, 18). This suggests that glioma cells themselves express Ang1 and Ang2 and contribute to neovascularization. The expressions of Ang1 and Ang2 in human glioblastoma tissues have been studied (13, 14) by in situ hybridization, and Ang1 mRNA but not Ang2 mRNA was detected in glioma cells. However, we hypothesized that glioma cells express Ang2 and explored the expression of Ang2 mRNA and protein in glioma cells using RT-PCR and immunohistochemical analysis with antihuman Ang2 antibody. We did indeed find Ang2 expression by human glioma cells and now report the relationship between angiogenic factors and vessel maturation in human gliomas.

Cell Culture.

Human malignant glioma cell lines (U105 MG, U251 MG, and U373 MG) were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 2 mml-glutamine, 100 IU/ml penicillin G, and 100 μg/ml streptomycin at 37°C in a humidified 5% CO2 atmosphere.

RT-PCR.

When the cells had grown to subconfluency, they were washed with ice-cold PBS and harvested by scraping. For the isolation of polyadenylated mRNA from glioma cell lines U105 MG, U251 MG, and U373 MG and for the generation of first-strand cDNA we used the Micro-Fast Track Kit and cDNA Cycle Kit (Invitrogen) according to the manufacturer’s protocols. The PCR mixture contained 2 μl of cDNA template, 200 μm deoxynucleotide triphosphate, 2.5 μl of 10 × PCR buffer (Life Technologies, Inc., Rockville, MD), 2.5 mm MgCl2, and 10 pmol of each oligonucleotide primer in a final volume of 25 μl. The reaction was hot-started at 94°C for 5 min before the addition of LA Taq or Taq polymerase (Takara, Otsu, Japan). Amplification was performed for 35 cycles under the following conditions: 1 min of denaturing at 94°C, 1 min of annealing at 62°C, and 1 min of extension at 72°C. At the end of the last cycle, a prolonged extension step was carried out for 10 min. The primer sequences used for human Ang2 were as follows: primer 1 (sense), 5′-GGATCTGGGGAGAGAGGAAC-3′; and primer 2 (antisense), 5′-CTCTGCACCGAGTCATCGTA-3′. The expected amplified fragment for human Ang2 was 535 bp (19). The PCR products were electrophoretically separated on 2% agarose gels, and bands were visualized by ethidium bromide staining. Furthermore, the PCR products were extracted using QIAquick Gel Extraction Kit (Qiagen), and the sequence was checked for consistency with human Ang2.

Development of Antihuman Ang2 Antibody.

The polyclonal antibody against human Ang2 was developed by use of the synthetic peptide NH2-NFRKSMDSIGKKQYQVQHGS-COOH corresponding to the amino acids at positions 21–40 in human Ang2 (6). The synthetic peptide was coupled to KLH. Male Japanese white rabbits received injections at multiple intradermal sites with 1 mg of synthetic peptide conjugated to KLH emulsified in complete Freund’s adjuvant. Then the rabbits were further immunized at 3-week intervals with 1 mg of synthetic peptide conjugated to KLH emulsified with incomplete Freund’s adjuvant. The antisera obtained after the fourth booster injection were affinity-purified on a column with N-hydroxysuccimide (NHS)-activated Sepharose 4 Fast Flow (Amersham Pharmacia Biotech, Tokyo, Japan) conjugating the synthetic peptide. The affinity-purified antihuman Ang2 antibody was dialyzed with PBS/10% glycerol and was designated KM202.

Western Blot Analysis of Ang2 Protein Expression in Malignant Glioma Cell Lines and Its Media.

Subconfluent human malignant glioma cell lines U105 MG, U251 MG, and U373 MG were washed three times with ice-cold PBS, harvested by scraping, and homogenized in an appropriate amount of the homogenizing buffer (20 mm HEPES buffer containing 5 mm EGTA, 5 mm EDTA, 1 mm phenylmethylsulfonyl fluoride, 1 mm DTT, 0.1 mm leupeptin, 75 μm pepstatin A, 150 mm NaCl, and 0.1% Triton X-100). The homogenates were centrifuged at 15,000 × g for 15 min at 4°C, and the supernatants were obtained. The homogenates of human glioblastoma and placenta tissues obtained from operation were prepared by the same procedure. After determination of the protein concentration by the method of Bradford (20), an equal amount of protein (30 μg) from each sample was subjected to 10% SDS-PAGE (21) and electrophoretically transferred to a nitrocellulose membrane (22). After blocking with 5% nonfat dry milk/PBS with 0.1% Tween 20 for 1 h at room temperature, the membrane was incubated with KM202 (dilution, 1:1000) at 4°C overnight. After incubation with the secondary antibody (sheep antirabbit IgG absorbed horseradish peroxidase labeled; American Qualex; 1:10,000) for 1 h at room temperature, immunoreactive bands were detected using the enhanced chemiluminescence Western blotting analysis system (Amersham).

The medium from established U251 MG cultures was replaced with serum-free defined medium. After 6 and 24 h of cell culture, this conditioned medium was harvested and concentrated 50-fold using a SpeedVac concentrator (Savant Instruments, Inc., Farmingdale, NY). An equal quantity of conditioned medium was subjected to Western blot analysis as described above.

Ang2 Expression in Human Glioma after Hypoxia.

Subconfluent U251 MG cells were cultured in serum-free medium for 24 h to remove the influence of various growth factors contained in the serum. The cells were then exposed to hypoxic stimulation by the O2 replacement with 95% N2/5% CO2 gas mixture 12 h, according to the method by Mandriota and Pepper (23). After hypoxic stimulation, the medium was replaced by serum-free medium. Hypoxic U251 MG cells were harvested at 0 and 24 h after hypoxic stimulation. Pre-control U251 MG cells were harvested just before the stimulation. Other U251 MG cells were cultured and harvested in parallel under the same conditions except for O2 replacement (Sham-Operation). These samples were subjected to Western blot analysis for Ang2 protein expression. The blots were exposed to Hyperfilm (enhanced chemiluminescence; Amersham) at room temperature. The images were scanned and analyzed semiquantitatively using the NIH image program.

Immunohistochemical Analysis of Glioma Tissues.

Of the 32 glioma specimens investigated, 14 were glioblastomas (WHO grade 4), 8 were anaplastic astrocytomas (WHO grade 3), 7 were astrocytomas (WHO grade 2), and 3 were pilocytic astrocytomas (WHO grade 1; Ref. 16). All of them were surgical specimens obtained during the last 3 years at the Department of Neurosurgery, Kumamoto University School of Medicine, Kumamoto, Japan. The specimens were fixed in 6% formaldehyde and embedded in paraffin. Serial 4-μm sections were prepared on albumin-coated slides. The sections were deparaffinized and incubated in 3% H2O2 for 10 min to block endogenous peroxidase activity. Nonspecific protein binding was blocked with 3% BSA/PBS for 1 h. The sections were incubated with primary antibodies for 1 h at room temperature. The primary antibodies used were anti-Ang2 (1:100; KM202), anti-HIF-1α (1:100; Santa Cruz Biotechnology), anti-VEGF (1:100; Santa Cruz Biotechnology), anti-Tie2 (1:200; Santa Cruz Biotechnology), anti-vWF (1:200; DAKO), and anti-αSMA (1:400; Sigma) antibodies. The specimens for Ang2 were run in parallel with preabsorbed primary antibody using a sufficient amount of the synthetic Ang2 peptide as negative controls. In subsequent steps, we used the Vectastain ABC kit using DAB as a chromogen (Vector, Burlingame, CA). The sections were then counterstained with methyl green (DAKO). In every immunohistochemical staining, we performed additional staining without primary antibody in parallel and ascertained that no staining was seen.

The immunoreactive pattern and distribution of angiogenic factors were carefully recorded by independent investigators. To quantify the expression levels of Ang2 and VEGF, we designed two parameters, the whole staining intensity of each glioma tissue specimen (−, negative; +, faint; ++, moderate; and +++, strong) and a LEI. Using photomicrographs, the LEI was determined by calculating the ratio of positive cells:total cell number in a neovascularized area of 0.5 mm2. In addition, a VMI (24, 25) was calculated using the ratio of αSMA-positive vessels:vWF-positive vessels in the same area in which the LEI value was determined.

In Situ Hybridization.

The pCR Vector (Invitrogen) containing the Ang2 PCR product described above was produced by a TA Cloning kit (Invitrogen), and a DIG-labeled RNA probe was generated by a DIG RNA Labeling Kit (Roche Diagnostics), according to the manufacturer’s protocol. The other vector containing the Ang2 PCR product inserted in the opposite direction was produced for control probe. Surgical specimens from the patients with glioblastoma were quickly frozen with OCT compound, and 5-μm cryosections were prepared. For in situ hybridization, sections were fixed with 4% paraformaldehyde for 15 min, incubated with 20 μg/ml of proteinase K for 20 min, and fixed again with 4% paraformaldehyde for 15 min. Acetylation was performed with 0.1 m triethanolamine for 1 min, followed by 0.25% acetic anhydride for 10 min at room temperature. Prehybridization solution containing 50% deionized formamide, 10 mm Tris-HCl (pH 7.6), 200 μg/ml tRNA, 1 × Denhardt’s solution, 10% dextran sulfate, 600 mm NaCl, 0.25% SDS, and 1 mm EDTA was applied for 1 h at 50°C. The RNA sense/antisense probe concentration was 0.5 μg/ml of the initial transcript and was incubated in a humidified chamber at 50°C for 16 h. Posthybridization stringency washes at 60°C included 2 × SSC plus 50% formamide for 15 min four times, followed by 1 × SSC at room temperature for 30 min twice. RNase A treatment (5 μg/ml) was performed. After blocking by 1% gout serum, the hybridized probe was detected by anti-DIG antibody conjugated to alkaline phosphatase (Roche Diagnostics) diluted 1:750 for 1 h at room temperature. Nitro blue tetrazolium (450 μg/ml)/5-bromo-4-chloro-3-indolyl phosphate (175 μg/ml) was used as the color substrate.

Serial cryosections were air-dried, acetone-fixed, and then immunostained using anti-Ang2 antibody by the method described above.

Statistical Assessment.

Statistical analysis of quantitative Ang2 expression was performed using a one-way ANOVA and post hoc Fisher’s protected least significant difference test. Values were expressed as mean ± SD. LEI/VMI values were statistically evaluated by checking Pearson’s correlation coefficient. The criterion for statistical significance was P < 0.05.

Establishment of Anti-Ang2 Antibody.

Western blot analysis using KM202 antibody (Fig. 1 A) detected a single immunoreactive band with Mr 65,000 in the homogenates of human glioblastoma (two cases; Lanes 1 and 2), human placenta at 16-week gestation (Lane 3), and human glioma cell line U251 MG (Lane 4). These immunoreactive bands in the glioblastoma homogenate (same as Lane 2) were not detected in the absence of first antibody (Lane 5) or when preabsorbed antibody, which had been preincubated with Ang2 synthetic peptide, was used (Lane 6).

Expression of Ang2 in Human Glioma Cell Lines.

Human Ang2 mRNA expression was analyzed in human malignant glioma cell lines U105 MG, U251 MG, and U373 MG by RT-PCR. As shown in Fig. 1 B, single 535-bp bands corresponding to the expected amplified fragment for human Ang2 were detected in of the all cell lines. The sequences of these PCR products extracted from the gel were completely consistent with human Ang2. These data suggest that human glioma cells express Ang2 mRNA.

To investigate whether glioma cells could produce the Ang2 protein, Western blot analysis using KM202 was performed. A single immunoreactive band with Mr 65,000, corresponding to the molecular weight of human Ang2 (5, 26), was detected in each glioma cell line. Immunoreactive bands were absent when KM202 was preabsorbed with the synthetic Ang2 peptides used as the antigen for immunization (Fig. 1,C). Western blot analysis of the conditioned medium obtained from U251 MG cells demonstrated that immunoreactive bands of Ang2 protein were enhanced during 24 h of culture (Fig. 1 D). These findings confirmed that cultured glioma cells themselves produce and secrete Ang2 protein.

Ang2 Expression Was Induced by Hypoxia in U251 MG Cells.

Hypoxia is a major inducer of angiogenesis. We posited that hypoxic stimulation induces Ang2 expression in glioma cells. To examine the effect of hypoxia on Ang2 expression, U251 MG cells were cultured for 12 h in serum-free medium in the hypoxia condition and then subjected to Western blot analysis at 0 and 24 h after stimulation. Ang2 expression in these hypoxic cells was clearly increased at 0 h after hypoxia (265.8 ± 28.1%; with statistical significance, P < 0.05). On the other hand, sham-operated cells showed no obvious change (Fig. 2, A and B). These findings confirm the conclusion that hypoxia induces the expression of Ang2 in glioma cells.

Immunohistochemical Findings in Human Glioma Tissues.

We used immunohistochemical methods to determine whether human glioma cells express Ang2. Ang2 protein was detected not only in vascular ECs (Fig. 3,D, arrow) but also in glioma cells (Fig. 3, B–D). The level of Ang2 immunoreactivity was high in malignant glioma cells (Fig. 3, B and D) and weak in low-grade glioma (Fig. 3,C). By in situ hybridization of malignant glioma, we could detect the Ang2 mRNA in both glioma cells (Fig. 3,I) and ECs (data not shown). Furthermore, its distribution was similar to that of Ang2 protein (Fig. 3,H, serial sections). Tie2 immunostaining was observed in vascular ECs (Fig. 3,E), and VEGF was mainly expressed in glioma cells (Fig. 3,F). The results of Ang2 and VEGF immunohistochemical studies are summarized in Table 1. Malignant gliomas (glioblastoma and anaplastic astrocytoma) expressed Ang2 and VEGF more strongly than did low-grade gliomas (astrocytoma and pilocytic astrocytoma).

Because the distribution of Ang2 was not uniform, we examined the regional relationship between Ang2 and tumor vessels (Fig. 4). Ang2 expression in glioblastoma was high in the area surrounding necrosis (Fig. 4, region-1, D) and in the tumor periphery (Fig. 4, region-3, F). It was low in the intermediate zone between the necrosis and the tumor periphery (Fig. 4, region-2, E). In the region surrounding necrosis, tumor vessels had thick hyperplastic walls with faint αSMA staining, suggesting vascular regression (Fig. 4, G and J). In the tumor periphery, many small vWF-positive vessels without αSMA were noted, suggesting angiogenesis (Fig. 4, I and L). In the intermediate zone, small vessels were rarely found, and most vessels were positive for both vWF and αSMA (Fig. 4, H and K). Thus, Ang2 expression was correlated with both vascular regression and angiogenesis. Ang2 expression in ECs was weak only in regressed vessels (Fig. 4, region-1, A, D, G, and H). In other regions, ECs expressed Ang2, especially in the tumor periphery (Fig. 4, E and F).

Because Ang2 expression was induced by hypoxia in cultured glioma, we performed immunohistochemistry of HIF-1α to evaluate the grade of hypoxia in tumor tissue. HIF-1α is a transcriptional factor that is unstable in cells exposed to oxygen but becomes stable and increases in ischemic conditions (27, 28, 29). Thus, HIF-1α immunoreactivity is considered to be a marker of ischemia. HIF-1α immunoreactivity is highest in the region surrounding necrosis and relatively higher in the tumor periphery than in the intermediate zone (Fig. 4). The blood supply may be insufficient in the periphery of tumor, where glioma cells are proliferating, resulting in mild hypoxia and Ang2 expression. Thus, Ang2 expression is correlated to the grade of hypoxia in human glioma tissue.

Correlation between the Vessel Maturation and Angiogenic Factors.

To examine the role of Ang2 in tumor angiogenesis, we analyzed the relationship between local expression of angiogenic factors and vessel maturation. Serially sectioned specimens were immunostained with Ang2, vWF, and αSMA. We randomly selected five regions with abundant vWF-positive vessels (over 20 vessels in 0.5 mm2) in each specimen. Regions surrounding the necrosis were not included. We evaluated the LEI of Ang2 and VEGF and VMI in the same region and calculated the mean values. Mature vessels were surrounded by abundant periendothelial support cells expressing αSMA (high VMI; Fig. 4, H and K). Immature vessels were surrounded by few or no periendothelial support cells (low VMI; Fig. 4, I and L). We performed statistical analysis using Pearson’s correlation coefficient to assess the correlation between VMI and LEI in both low-grade and malignant gliomas (Fig. 5). The LEI of VEGF was correlated positively with the VMI in low-grade gliomas (Fig. 5,B; P < 0.05, r = 0.708). In malignant gliomas, Ang2 LEI was correlated inversely with the VMI with statistical difference (Fig. 5 D; P < 0.05, r = 0.466). Thus, we concluded that vessel maturation in low-grade gliomas depends on the level of VEGF expression and that Ang2 contributes to immature vessel formation in malignant gliomas.

Ang2 mRNA is expressed in cultured ECs and smooth muscle cells (19, 23) and in the human artery, vein, ovary, placenta, and uterus (6, 19), all of which are predominant sites of vascular remodeling. Ang2 mRNA is detected in human hepatocellular carcinoma cells (30) but not in other malignant tumors (23, 31). In human glioblastoma tissues, in situ hybridization revealed Ang2 mRNA in neovascular EC but not in glioma cells (13, 14). However, in this study we detected Ang2 protein and mRNA in both glioma cells and ECs. Furthermore, glioma cells expressed Ang2 protein at higher levels than ECs in some specimens. It is likely that there is a difference in the stability of Ang2 protein and mRNA between glioma and ECs. Furthermore, the distribution of Ang2 is not uniform, and Ang2 expression is weak or absent in the main lesions of tumor tissues (intermediate zone; Fig. 4, region-2). These factors may explain the discrepancy between other in situ hybridization studies and our study.

Because hypoxia has been shown to be an inducer of Ang2 expression in cultured EC lines (23, 26), we assessed the effect of hypoxic stimulation on cultured glioma cells. Western blot analysis revealed that the expression of Ang2 protein increased after hypoxic stimulation. Because VEGF is known to induce Ang2 expression (26), there is a question as to whether glioma Ang2 is induced by growth factors such as VEGF or directly by hypoxia. Glioma cells were cultured in serum-free medium during hypoxia, and we could not detect any increase of VEGF by Western blot analysis (data not shown) in this study. These results suggest that hypoxic stimulation directly induces the Ang2 expression in glioma cells.

Several experiments (32, 33, 34, 35) have shown that platelet-derived growth factor, transforming growth factor β, and VEGF are involved in vessel maturation. Moreover, recent evidence suggested that Ang1 and Ang2 play important roles in angiogenesis and vessel maturation. It is now thought that Ang1 binding to Tie2 recruits periendothelial support cells and maintains the vessel integrity. On the other hand, Ang2, in antagonizing the actions of Ang1 by competitively binding to Tie2, induces the vessel structures to become loosened by reducing EC contacts with the matrix and dissociating the periendothelial support cells (36). EC contact with periendothelial support cells is a critical determinant of blood vessel maturation (32, 37). It has been suggested (13, 15) that glioma angiogenesis is initiated by EC-derived Ang2, which induces weakening of EC interactions with the microenvironment, including the EC-pericyte interactions. Our immunohistochemical study revealed that glioma cells expressed Ang2, and it is possible that glioma cells themselves, and not only EC, regulate angiogenesis.

However, Ang2 expression was prominent in the regions with vessel regression as well as in regions manifesting angiogenesis. It is reported that Ang2 acts to induce not only angiogenesis but also vessel regression (6, 31). The mechanisms underlying these contrary effects of Ang2 remain unclear. We consider that they may be dependent on the Ang1 expression level in malignant glioma. The region around necrosis in glioblastoma is exposed to hypoxia, and many cytokines are secreted (38). Ang1 is known to be down-regulated by hypoxia and interleukin-1β (39, 40). Ang1 expression was thus considered to decrease around necrosis. In contrast, Ang2 is induced by hypoxia and some cytokines such as VEGF and transforming growth factor β1(23, 26). Furthermore, Tie2 receptor expression is stimulated by hypoxia, tumor necrosis factor-α, and interleukin-1β (41). Thus, Ang1 expression decreased, whereas Ang2 and Tie2 expressions increased in the perinecrotic areas of malignant gliomas. This obvious decrease of the Ang1 with increased Ang2/Tie2 expression may contribute to vessel regression. However, it remains unclear whether Ang2 directly induces vessel regression in the region surrounding necrosis. In contrast to the necrotic region, Ang1 expression is maintained in the tumor periphery (31), whereas Ang2 signal increases. It is possible that Ang2 acts to induce angiogenesis, rather than vessel regression, in the periphery of tumor.

In the intermediate region where most of the vessels were mature, Ang2 expression was not particularly high. These findings were obvious, and statistical analysis showed an inverse correlation between vessel maturation and Ang2 expression. With a decrease of Ang2 expression, vessels may mature by preserved Ang1 activity in malignant glioma. In contrast to malignant glioma, the expression index of VEGF but not of Ang2 was significantly positively correlated with vessel maturation in low-grade gliomas. Because low-grade gliomas expressed angiopoietins at a lower level than did malignant gliomas, VEGF may replace angiopoietins as an effector of tumor vessel maturation.

To define the critical role of angiopoietins in glioma angiogenesis, it is necessary to examine the signal pathway and the transcriptional mechanisms of Ang2 expression in glioma cells. Such studies may contribute to the development of new treatment strategies, e.g., anti-angiogenic therapy, in patients with glioma (42, 43, 44, 45, 46).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

      
1

Supported by Grants-in-aid for Scientific Research from the Ministry of Education, Sports, Science and Culture of Japan.

            
3

The abbreviations used are: VEGF, vascular endothelial growth factor; Ang, angiopoietin; EC, endothelial cell; RT-PCR, reverse transcriptase-PCR; KLH, keyhole limpet hemocyanin; HIF-1α, hypoxia inducible factor-1α; vWF, von Willebrand factor; αSMA, α smooth muscle actin; DAB, 3,3′-diaminobenzidine; LEI, local expression index; VMI, vessel maturation index; DIG, digoxigenin.

Fig. 1.

Establishment of anti-Ang2 antibody KM202. A, Western blot analysis against human tissue homogenates: glioblastoma (20 μg; Lanes 1 and 2), placenta at 16-week gestation (20 μg; Lane 3), and U251 MG human glioblastoma cell (20 μg; Lane 4). These immunoreactive bands (glioblastoma, same as Lane 2) disappeared in the absence of first antibody KM202 (Lane 5) and when using preabsorbed antibody (Lane 6). Cultured human malignant glioma cells express Ang2 mRNA and protein. B, RT-PCR analysis of Ang2 mRNA in U105 MG, U251 MG, and U373 MG cell lines showed a single band corresponding to the expected PCR product size. C, Western blot analysis using antihuman Ang2 antibody KM202 showed a single immunoreactive band in glioma cell lysates. These immunoreactive bands disappeared when we used the preabsorbed KM202 antibody. D, after subconfluent U251 MG cells were medium-changed, the media at 0, 6, and 24 h after were collected and concentrated 50-fold. Western blot analysis of these media demonstrated an increase of Ang2 immunoreactivity with time.

Fig. 1.

Establishment of anti-Ang2 antibody KM202. A, Western blot analysis against human tissue homogenates: glioblastoma (20 μg; Lanes 1 and 2), placenta at 16-week gestation (20 μg; Lane 3), and U251 MG human glioblastoma cell (20 μg; Lane 4). These immunoreactive bands (glioblastoma, same as Lane 2) disappeared in the absence of first antibody KM202 (Lane 5) and when using preabsorbed antibody (Lane 6). Cultured human malignant glioma cells express Ang2 mRNA and protein. B, RT-PCR analysis of Ang2 mRNA in U105 MG, U251 MG, and U373 MG cell lines showed a single band corresponding to the expected PCR product size. C, Western blot analysis using antihuman Ang2 antibody KM202 showed a single immunoreactive band in glioma cell lysates. These immunoreactive bands disappeared when we used the preabsorbed KM202 antibody. D, after subconfluent U251 MG cells were medium-changed, the media at 0, 6, and 24 h after were collected and concentrated 50-fold. Western blot analysis of these media demonstrated an increase of Ang2 immunoreactivity with time.

Close modal
Fig. 2.

Change in Ang2 expression in U251 MG cells after hypoxia. U251 MG cells were exposed to hypoxic stimulation by O2 replacement with a 95% N2/5% CO2 gas mixture for 12 h. The expression of Ang2 protein was analyzed and evaluated by Western blotting (n = 5) at 0 and 24 h after the cessation of hypoxic stimulation. Ang2 expression increased after hypoxic stimulation (265.8 ± 28.1%; with statistical significance, P < 0.05) and decreased gradually 24 h after (154 ± 12.5%; without statistical significance), whereas that of sham-operated cells showed no obvious change.

Fig. 2.

Change in Ang2 expression in U251 MG cells after hypoxia. U251 MG cells were exposed to hypoxic stimulation by O2 replacement with a 95% N2/5% CO2 gas mixture for 12 h. The expression of Ang2 protein was analyzed and evaluated by Western blotting (n = 5) at 0 and 24 h after the cessation of hypoxic stimulation. Ang2 expression increased after hypoxic stimulation (265.8 ± 28.1%; with statistical significance, P < 0.05) and decreased gradually 24 h after (154 ± 12.5%; without statistical significance), whereas that of sham-operated cells showed no obvious change.

Close modal
Fig. 3.

Immunohistochemistry of Ang2 in human glioma of anaplastic astrocytoma (A, D, E, and F; case-16), glioblastoma (B; case-4), and astrocytoma (C; case-29). Serial sections of anaplastic astrocytoma (D–F) were stained by Ang2 (D), Tie2 (E), and VEGF (F). Preabsorbed KM202 was used as the primary antibody (A). Serial frozen sections (case-8) of H&E staining (G), immunohistochemistry of Ang2 (H), and in situ hybridization using Ang2 probe (I) and control probe (J). Immunostaining procedures were carried out according to instructions in the Vectastain ABC kit, using DAB as a chromogen, and the immunostained sections were counterstained with methyl green (A–F). Case numbers correspond to Table 1. Arrow, vascular EC staining; scale bar, 50 μm.

Fig. 3.

Immunohistochemistry of Ang2 in human glioma of anaplastic astrocytoma (A, D, E, and F; case-16), glioblastoma (B; case-4), and astrocytoma (C; case-29). Serial sections of anaplastic astrocytoma (D–F) were stained by Ang2 (D), Tie2 (E), and VEGF (F). Preabsorbed KM202 was used as the primary antibody (A). Serial frozen sections (case-8) of H&E staining (G), immunohistochemistry of Ang2 (H), and in situ hybridization using Ang2 probe (I) and control probe (J). Immunostaining procedures were carried out according to instructions in the Vectastain ABC kit, using DAB as a chromogen, and the immunostained sections were counterstained with methyl green (A–F). Case numbers correspond to Table 1. Arrow, vascular EC staining; scale bar, 50 μm.

Close modal
Fig. 4.

Immunohistochemical analysis of the relationship between Ang2 expression and vessel maturation. The top schematic drawing shows the three regions examined in the glioblastoma tissue (case-8) including the area of necrosis. Each region was immunostained by HIF-1α. Region-1, area surrounding the necrosis; region-2, intermediate zone between the necrosis and the tumor periphery; region-3, tumor periphery. The four photographs in each column are of serial sections. These serial sections were stained by H&E. (A, B, and C), Ang2 (D, E, and F), vWF (G, H, and I), and αSMA (J, K, and L). Immunostaining procedures were carried out according to instructions in the Vectastain ABC kit, using DAB as a chromogen, and the immunostained sections were counterstained with methyl green. Circle, necrotic area; arrow, tumor vessel near necrosis; arrow head(s), tumor vessels stained by vWF but not by αSMA in the tumor periphery; scale bar, 100 μm.

Fig. 4.

Immunohistochemical analysis of the relationship between Ang2 expression and vessel maturation. The top schematic drawing shows the three regions examined in the glioblastoma tissue (case-8) including the area of necrosis. Each region was immunostained by HIF-1α. Region-1, area surrounding the necrosis; region-2, intermediate zone between the necrosis and the tumor periphery; region-3, tumor periphery. The four photographs in each column are of serial sections. These serial sections were stained by H&E. (A, B, and C), Ang2 (D, E, and F), vWF (G, H, and I), and αSMA (J, K, and L). Immunostaining procedures were carried out according to instructions in the Vectastain ABC kit, using DAB as a chromogen, and the immunostained sections were counterstained with methyl green. Circle, necrotic area; arrow, tumor vessel near necrosis; arrow head(s), tumor vessels stained by vWF but not by αSMA in the tumor periphery; scale bar, 100 μm.

Close modal
Fig. 5.

Statistical analysis (Pearson’s correlation coefficient analysis) of the correlation between angiogenic factor expression and vessel maturation in low-grade glioma (A, B, and C) and malignant glioma (D, E, and F). VEGF LEI in low-grade glioma (B) and Ang2 LEI in malignant glioma (D) are statistically correlated with maturation index. Values of P < 0.05 were considered significant (*). N. S., not significant.

Fig. 5.

Statistical analysis (Pearson’s correlation coefficient analysis) of the correlation between angiogenic factor expression and vessel maturation in low-grade glioma (A, B, and C) and malignant glioma (D, E, and F). VEGF LEI in low-grade glioma (B) and Ang2 LEI in malignant glioma (D) are statistically correlated with maturation index. Values of P < 0.05 were considered significant (*). N. S., not significant.

Close modal
Table 1

Immunoreactive intensity of Ang2 and VEGF in 32 human glioma tissues

The samples were evaluated independently by two investigators.
Case no.AgeSexAng2aVEGFa
Glioblastoma     
17 
23 ++ 
25 ++ 
33 +++ +++ 
38 ++ 
51 ++ 
54 ++ 
55 +++ ++ 
57 − 
10 64 +++ +++ 
11 65 ++ ++ 
12 67 ++ ++ 
13 71 ++ ++ 
14 75 ++ 
Anaplastic astrocytoma     
15 31 ++ +++ 
16 38 +++ +++ 
17 40 +++ 
18 45 +++ 
19 50 
20 54 
21 42 ++ ++ 
22 73 
Astrocytoma     
23 
24 23 
25 35 
26 38 +++ 
27 50 ++ ++ 
28 50 ++ ++ 
29 60 
Pilocytic astrocytoma     
30 
31 
32 20 
The samples were evaluated independently by two investigators.
Case no.AgeSexAng2aVEGFa
Glioblastoma     
17 
23 ++ 
25 ++ 
33 +++ +++ 
38 ++ 
51 ++ 
54 ++ 
55 +++ ++ 
57 − 
10 64 +++ +++ 
11 65 ++ ++ 
12 67 ++ ++ 
13 71 ++ ++ 
14 75 ++ 
Anaplastic astrocytoma     
15 31 ++ +++ 
16 38 +++ +++ 
17 40 +++ 
18 45 +++ 
19 50 
20 54 
21 42 ++ ++ 
22 73 
Astrocytoma     
23 
24 23 
25 35 
26 38 +++ 
27 50 ++ ++ 
28 50 ++ ++ 
29 60 
Pilocytic astrocytoma     
30 
31 
32 20 
a

−, negative; +, faint; ++, moderate; +++, strong immunoreactivity.

We thank Masayo Obata for her technical assistance.

1
Shweiki D., Itin A., Neufeld G., Gitay-Goren H., Keshet E. Patterns of expression of vascular endothelial growth factor (VEGF) and VEGF receptors in mice suggest a role in hormonally regulated angiogenesis.
J. Clin. Investig.
,
91
:
2235
-2243,  
1993
.
2
Ferrara N. The role of vascular endothelial growth factor in pathological angiogenesis.
Breast Cancer Res. Treat.
,
36
:
127
-137,  
1995
.
3
Folkman J. Angiogenesis in cancer, vascular, rheumatoid and other disease.
Nat. Med.
,
1
:
27
-31,  
1995
.
4
Hanahan D., Folkman J. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis.
Cell
,
86
:
353
-364,  
1996
.
5
Davis S., Aldrich T. H., Jones P. F., Acheson A., Compton D. L., Jain V., Ryan T. E., Bruno J., Radziejewski C., Maisonpierre P. C., Yancopoulos G. D. Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning.
Cell
,
87
:
1161
-1169,  
1996
.
6
Maisonpierre P. C., Suri C., Jones P. F., Bartunkova S., Wiegand S. J., Radziejewski C., Compton D., McClain J., Aldrich T. H., Papadopoulos N., Daly T. J., Davis S., Sato T. N., Yancopoulos G. D. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis.
Science (Wash. DC)
,
277
:
55
-60,  
1997
.
7
Suri C., Jones P. F., Patan S., Bartunkova S., Maisonpierre P. C., Davis S., Sato T. N., Yancopoulos G. D. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis.
Cell
,
87
:
1171
-1180,  
1996
.
8
Sato T. N., Qin Y., Kozak C. A., Audus K. L. Tie-1 and tie-2 define another class of putative receptor tyrosine kinase genes expressed in early embryonic vascular system.
Proc. Natl. Acad. Sci. USA
,
90
:
9355
-9358,  
1993
.
9
Sato T. N., Tozawa Y., Deutsch U., Wolburg-Buchholz K., Fujiwara Y., Gendron-Maguire M., Gridley T., Wolburg H., Risau W., Qin Y. Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation.
Nature (Lond.)
,
376
:
70
-74,  
1995
.
10
Asahara T., Chen D., Takahashi T., Fujikawa K., Kearney M., Magner M., Yancopoulos G. D., Isner J. M. Tie2 receptor ligands, angiopoietin-1 and angiopoietin-2, modulate VEGF-induced postnatal neovascularization.
Circ. Res.
,
83
:
233
-240,  
1998
.
11
Koblizek T. I., Weiss C., Yancopoulos G. D., Deutsch U., Risau W. Angiopoietin-1 induces sprouting angiogenesis in vitro.
Curr. Biol.
,
8
:
529
-532,  
1998
.
12
Folkman J., D’Amore P. A. Blood vessel formation: what is its molecular basis?.
Cell
,
87
:
1153
-1155,  
1996
.
13
Stratmann A., Risau W., Plate K. H. Cell type-specific expression of angiopoietin-1 and angiopoietin-2 suggests a role in glioblastoma angiogenesis.
Am. J. Pathol.
,
153
:
1459
-1466,  
1998
.
14
Zagzag D., Hooper A., Friedlander D. R., Chan W., Holash J., Wiegand S. J., Yancopoulos G. D., Grumet M. In situ expression of angiopoietins in astrocytomas identifies angiopoietin-2 as an early marker of tumor angiogenesis.
Exp. Neurol.
,
159
:
391
-400,  
1999
.
15
Lauren J., Gunji Y., Alitalo K. Is angiopoietin-2 necessary for the initiation of tumor angiogenesis? Am.
J. Pathol.
,
153
:
1333
-1339,  
1998
.
16
Kleihues P., Burger P. C., Scheithauer B. W. The new WHO classification of brain tumours.
Brain Pathol.
,
3
:
255
-268,  
1993
.
17
Plate K. H., Breier G., Weich H. A., Risau W. Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo.
Nature (Lond.)
,
359
:
845
-848,  
1992
.
18
Plate K. H., Breier G., Weich H. A., Mennel H. D., Risau W. Vascular endothelial growth factor and glioma angiogenesis: coordinate induction of VEGF receptors, distribution of VEGF protein and possible in vivo regulatory mechanisms.
Int. J. Cancer
,
59
:
520
-529,  
1994
.
19
Witzenbichler B., Maisonpierre P. C., Jones P., Yancopoulos G. D., Isner J. M. Chemotactic properties of angiopoietin-1 and -2, ligands for the endothelial-specific receptor tyrosine kinase Tie2.
J. Biol. Chem.
,
273
:
18514
-18521,  
1998
.
20
Bradford M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding.
Anal. Biochem.
,
72
:
248
-254,  
1976
.
21
Laemmli U. K. Cleavage of structural proteins during the assembly of the head of bacteriophage T4.
Nature (Lond.)
,
227
:
680
-685,  
1970
.
22
Towbin H., Staehelin T., Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications.
Proc. Natl. Acad. Sci. USA
,
76
:
4350
-4354,  
1979
.
23
Mandriota S. J., Pepper M. S. Regulation of angiopoietin-2 mRNA levels in bovine microvascular endothelial cells by cytokines and hypoxia.
Circ. Res.
,
83
:
852
-859,  
1998
.
24
Goede V., Schmidt T., Kimmina S., Kozian D., Augustin H. G. Analysis of blood vessel maturation processes during cyclic ovarian angiogenesis.
Lab. Investig.
,
78
:
1385
-1394,  
1998
.
25
Benjamin L. E., Golijanin D., Itin A., Pode D., Keshet E. Selective ablation of immature blood vessels in established human tumors follows vascular endothelial growth factor withdrawal.
J. Clin. Investig.
,
103
:
159
-165,  
1999
.
26
Oh H., Takagi H., Suzuma K., Otani A., Matsumura M., Honda Y. Hypoxia and vascular endothelial growth factor selectively up-regulate angiopoietin-2 in bovine microvascular endothelial cells.
J. Biol. Chem.
,
274
:
15732
-15739,  
1999
.
27
Semenza G. L., Nejfelt M. K., Chi S. M., Antonarakis S. E. Hypoxia-inducible nuclear factors bind to an enhancer element located 3′ to the human erythropoietin gene.
Proc. Natl. Acad. Sci. USA
,
88
:
5680
-5684,  
1991
.
28
Huang L. E., Arany Z., Livingston D. M., Bunn H. F. Activation of hypoxia-inducible transcription factor depends primarily upon redox-sensitive stabilization of its α subunit.
J. Biol. Chem.
,
271
:
32253
-32259,  
1996
.
29
Huang L. E., Gu J., Schau M., Bunn H. F. Regulation of hypoxia-inducible factor 1α is mediated by an O2-dependent degradation domain via the ubiquitin-proteasome pathway.
Proc. Natl. Acad. Sci. USA
,
95
:
7987
-7992,  
1998
.
30
Tanaka S., Mori M., Sakamoto Y., Makuuchi M., Sugimachi K., Wands J. R. Biologic significance of angiopoietin-2 expression in human hepatocellular carcinoma.
J. Clin. Investig.
,
103
:
341
-345,  
1999
.
31
Holash J., Maisonpierre P. C., Compton D., Boland P., Alexander C. R., Zagzag D., Yancopoulos G. D., Wiegand S. J. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF.
Science (Wash. DC)
,
284
:
1994
-1998,  
1999
.
32
Hirschi K. K., Rohovsky S. A., D’Amore P. A. PDGF, TGF-β, and heterotypic cell-cell interactions mediate endothelial cell-induced recruitment of 10T1/2 cells and their differentiation to a smooth muscle fate.
J. Cell Biol.
,
141
:
805
-814,  
1998
.
33
Lindahl P., Johansson B. R., Leveen P., Betsholtz C. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice.
Science (Wash. DC)
,
277
:
242
-245,  
1997
.
34
Antonelli-Orlidge A., Saunders K. B., Smith S. R., D’Amore P. A. An activated form of transforming growth factor β is produced by cocultures of endothelial cells and pericytes.
Proc. Natl. Acad. Sci. USA
,
86
:
4544
-4548,  
1989
.
35
Benjamin L. E., Hemo I., Keshet E. A plasticity window for blood vessel remodeling is defined by pericyte coverage of the preformed endothelial network and is regulated by PDGF-B and VEGF.
Development (Camb.)
,
125
:
1591
-1598,  
1998
.
36
Hanahan D. Signaling vascular morphogenesis and maintenance.
Science (Wash. DC)
,
277
:
48
-50,  
1997
.
37
Hirschi K. K., D’Amore P. A. Pericytes in the microvasculature.
Cardiovasc. Res.
,
32
:
687
-698,  
1996
.
38
Kleihues P., Burger P. C., Plate K. H., Ohgaki H., Cavenee W. K. Glioblastoma Kleihues P. Cavenee W. K. eds. .
Pathology and Genetics of Tumors of the Nervous System
,
:
16
-24, International Agency for Research on Cancer Lyon  
1997
.
39
Enholm B., Paavonen K., Ristimaki A., Kumar V., Gunji Y., Klefstrom J., Kivinen L., Laiho M., Olofsson B., Joukov V., Eriksson U., Alitalo K. Comparison of VEGF, VEGF-B, VEGF-C and Ang-1 mRNA regulation by serum, growth factors, oncoproteins and hypoxia.
Oncogene
,
14
:
2475
-2483,  
1997
.
40
Ristimaki A., Narko K., Enholm B., Joukov V., Alitalo K. Proinflammatory cytokines regulate expression of the lymphatic endothelial mitogen vascular endothelial growth factor-C.
J. Biol. Chem.
,
273
:
8413
-8418,  
1998
.
41
William C., Koehne P., Jurgensen J. S., Grafe M., Wagner K. D., Bachmann S., Frei U., Eckardt K-U. Tie2 receptor expression is stimulated by hypoxia and proinflammatory cytokines in human endothelial cells.
Circ. Res.
,
87
:
370
-377,  
2000
.
42
Kim K. J., Li B., Winer J., Armanini M., Gillett N., Phillips H. S., Ferrara N. Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumour growth in vivo.
Nature (Lond.)
,
362
:
841
-844,  
1993
.
43
Millauer B., Shawver L. K., Plate K. H., Risau W., Ullrich A. Glioblastoma growth inhibited in vivo by a dominant-negative Flk-1 mutant.
Nature (Lond.)
,
367
:
576
-579,  
1994
.
44
Millauer B., Longhi M. P., Plate K. H., Shawver L. K., Risau W., Ullrich A., Strawn L. M. Dominant-negative inhibition of Flk-1 suppresses the growth of many tumor types in vivo.
Cancer Res.
,
56
:
1615
-1620,  
1996
.
45
Lin P., Polverini P., Dewhirst M., Shan S., Rao P. S., Peters K. Inhibition of tumor angiogenesis using a soluble receptor establishes a role for Tie2 in pathologic vascular growth.
J. Clin. Investig.
,
100
:
2072
-2078,  
1997
.
46
Lin P., Buxton J. A., Acheson A., Radziejewski C., Maisonpierre P. C., Yancopoulos G. D., Channon K. M., Hale L. P., Dewhirst M. W., George S. E., Peters K. G. Antiangiogenic gene therapy targeting the endothelium-specific receptor tyrosine kinase Tie2.
Proc. Natl. Acad. Sci. USA
,
95
:
8829
-8834,  
1998
.