Temporal-spatial expression of VEGF, angiopoietins-1 and 2, and Tie-2 during tumor angiogenesis and their functional correlation with tumor neovascular architecture, The

temporal-spatial expression of VEGF, angiopoietins-1 and 2, and Tie-2 during tumor angiogenesis and their functional correlation with tumor neovascular architecture, The

Tse, Victor

Angiopoietins play a pivotal role In tumor angiogenesis by modulating vascular endothelial proliferation and survival. The expression of angiopoietins 1 and 2 (Ang-1 and Ang-2) and vascular endothelial growth factor (VEGF) has been documented in human malignant glioma. The expression of Ang-1, Ang-2, VEGF, and Tie-2, a member of the receptor tyrosine kinases and the natural receptor for both Ang-1 and Ang-2, follows a distinct transcriptional profile in vivo. Ang-2 and VEGF were expressed early in tumor formation and their levels increased throughout tumor growth. Their expression coincided with the expansion of the tumor mass and the formation of the vascular tree. There was no significant change in the expression of Tie-2 and Ang-1. The expression of Ang-1 and Tie-2 was more noticeable at the periphery of the tumor. The expression of Ang-2 was more robust at the periphery and within the tumor mass, and VEGF was more concentrated within the center of the tumor. This distinct expression profile may explain the morphology of the newly formed vessels at various times and regions of the tumor. The lack of concomitant expression of Ang-1 may underscore the unopposed endovascular induction by Ang-2 and VEGF resulting in the chaotic appearance and fragility of tumor vessels. [Neurol Res 2003; 25: 729-738]

Keywords: Angiopoietin-1; angiopoietin-2; VEGF; Tie-2; glioblastoma; tumor vasculature


Malignant gliomas are highly vascularized tumors whose pathological hallmarks are vascular proliferation and necrosis1. The establishment of a malignant glioma is partially dependent on its ability to parasitize preexisting blood vessels2. As the tumor enlarges, it recruits vessels from the adjacent brain. These new tumor vessels proliferate in the tumor and are modified for its use3,4. This process, termed angiogenesis, is tightly modulated by a number of factors whose interaction determines the final architecture and function of the neovascular tumor network5,6. Among these tumor angiogenic factors are vascular endothelial growth factor (VEGF) and the angiopoietins (Ang-1 and Ang-2). The role of Ang-2 in directing endothelial migration and tube formation7, i.e., sprouting, has been shown to be essential in glioma angiogenesis, and is moderated by the presence of VEGF7-9. In its presence, Ang-2 induces vascular endothelial proliferation, causes disintegration of basal matrix, and promotes cellular migration. In its absence, Ang-2 destabilizes endothelial tubule formation and causes regression of the newly formed vessels10,11.

The role of Ang-1 in tumor angiogenesis is less certain12,13. Ang-1 seems essential to the maintenance of the integrity of newly formed blood vessels. Ang-1 decreases leakage of plasma from mature vasculature by mediating the interaction between endothelial cells and their pericytic elements in the formation of the basal membrane4,15. It regulates endothelial cell survival through the phosphatidylinositol 3′-kinase/Akt signal transduction pathway. The absence of Ang-1 or the overexpression of Ang-2 may lead to endothelial apoptosis and involution of the neovasculature16. Both Ang-1 and 2 act upon the same transmembrane receptor tyrosine kinase (Tie-2) but with opposite effects. Ang-2 mediates inhibition of Tie-2 functions on vascular endothelial cells by loosening adhesion between perivascular support cells and endothelial cells, which may enhance accessibility to angiogenic inducers such as VEGF. This facilitates endothelial expansion and migration. The balance of this seemingly opposing effect dictates the stability and maturity of newly formed vessels17.

In this study, a rodent brain tumor resulting from intracerebral injection of cells of the rat glioma line, RT-2 is used to demonstrate the temporal and spatial distributions of the in vivo expression of VEGF, Ang-1 and 2, and their receptor Tie-2 during tumor growth. Their expression correlates with the morphological changes in the tumor vessels and provides insight into their individual influence in shaping the architecture of the tumor vasculature.


Cells and cell culture

A rat glioblastoma (RT-2) cell line was used in all of the following experiments. RT-2 cells were grown in Dulbecco-Vogt’s Modified Eagle Medium (DMEM + high glucose) with 10% fetal bovine serum (Hyclone, Logan, Utah, USA). This tumor cell line has a 95% plating efficiency with a generation time of 8-12 h. The cells are spindle-shaped at log phase growth and become epitheloid at confluency. They are GFAP and vimentin positive. RT-2 cells carry a wildtype p53 gene and are capable of transactivating a wildtype p53 specific pg13CAT promoter construct18.

Animal surgery and microdissection

All procedures performed on rats were approved by the Administrative Panel on Laboratory Animal Care at Stanford University. Fisher rats (180-200g) were anesthetized with an intramuscular cocktail injection of ketamine (22-44 mg kg^sup -1^), xylazine (2.5 mg kg^sup -1^), and acepromazine (0.75 mg kg^sup -1^). Animals were then placed in a stereotactic apparatus (David Kopf, Tujunga, CA, USA), and a small frontal burr hole was drilled. The entry was 2 mm off midline and 1 mm posterior to the coronal suture with a 180g rat. Ten microliters of RT-2 (1 x 10^sup 5^ cells) were injected slowly and steadily over a 2-min period into the striatum at a depth of 5.0 mm through a 23-gauge syringe needle attached to a Hamilton microinjector. The needle was then retracted over a 2-min period. The burr hole was sealed with bone wax prior to closure of the incision. Rat brains harboring the tumor were harvested on specified days after injection. The brains were washed in cold buffer, sliced into 2-mm slabs and placed onto an illuminated cold dissection platform. Under magnification, the tumor and the peritumoral region (a 1-mm annulus) were dissected out using microsurgical techniques. Brain tissue from the contralateral basal ganglia and cortex was also saved. The specimens were quickly frozen with liquid nitrogen and kept at -80[degrees]C until their RNA could be isolated.

Semi-quantitative RT-PCR

The frozen brains were pulverized in liquid nitrogen and poured into the extraction buffer. Total RNA was prepared with the SNAP extraction system (Invitrogen, Carlsbad, CA, USA) and subsequently quantified by spectrophotometer. Two micrograms of total RNA was routinely used to synthesize cDNA with the Superscript II system (Invitrogen). Oligo-dT primers were used in the reverse transcription step. The yield of cDNA was measured according to the PCR signal generated from the internal standard house-keeping gene [beta]-actin. Specific primer sets for VEGF, Tie-2, Ang-1, and Ang-2 were used for performing PCR with the predetermined cycle numbers. Primer sequences were designed to span intron regions. In addition, all PCR primers were tested using murine genomic DNA as a negative control. The linear amplification range for each gene was determined empirically by carrying out the amplification between 15 to 45 cycles. The primer pairs for RT-PCR included the following: Ang-1: Forward: 5′-CAG TGG CTG GAA AAA CTT G-3′. Reverse: 5′-CAT TTG TCT GTT GGA GAA GC-3′. Ang-2: Forward: 5′-GGA GAG TAT TGG CTG GGC AAC GAG T-3′. Reverse: 5-GCA GAT GCA TTT GTC ATT GTC CGA ATC CTT-3′. Tie-2: Forward: 5′-ATT GAC GTG AAG ATC AAG AAT GCC ACC-3′. Reverse: 5′-ATC CGG ATT GTT TTT GGC CTT CCT GTT-3′. VEGF: Forward: 5′-GTG CAC TGG ACC CTG GCT TTA-3′. Reverse: 5′-AGT GAT TTT CTG GCT TTG TTC TAT-3′. The settings for the thermal cycling were as follows: Ang-1: 95[degrees]C for 5 min; then 95[degrees]C for 30 sec, 52[degrees]C for 1 min, and 72[degrees]C for 45 sec for 36 cycles; followed by 5 min at 72[degrees]C for final extension. Ang-2: 95[degrees]C for 5 min; then 95[degrees]C for 30 sec, 550C for 1 min and 72[degrees]C for 45 sec for 30 cycles; foliowed by 5 min at 72[degrees]C for final extension. VEGF: 98[degrees]C for 10 min, then hot start at 80[degrees]C; 94[degrees]C for 30 sec, 45[degrees]C for 30 sec, and 72[degrees]C for 1 min for 30 cycles; followed by 72[degrees]C for final extension. Tie-2: 94[degrees]C for 7 min; then 94[degrees]C for 45 sec, 55[degrees]C for 45 sec, and 72[degrees]C for 90 sec for 35 cycles; followed by 72[degrees]C for 7 min for final extension. The amplicons for Ang-1, Ang-2, Tie-2, VEGF, and [beta]-actin are 256 bp, 257 bp, 375 bp, and 201 bp, respectively. The PCR products were gel purified, and their base sequences were verified to be that of their target genes using the dye-terminated automated sequencing method (PAQN Facility, Beckmen Center, Stanford, CA, USA), [beta]-Actin primers were added to corresponding samples in the specified genes’ linear amplification range. Resulting amplicons were electrophoresed on a 1% agarose gel. The gel was stained with ethidium bromide and photographed on top of a 280 n m UV light box, and gel images were captured with the Kodak digital gel imaging system. Data were analyzed by the accompanying gel analysis software. The quantity and base pair size of the amplicons were estimated relative to DNA ladder standards. RT-PCR values were expressed as a ratio of the specific gene’s signal in selected linear amplification cycle divided by the [beta]-actin signal (relative units of mRNA expression)19-21. Data sets for the spatial distribution were pooled from results of three different experiments; three animals were used in each experiment (n=9 animals). Statistical analysis in the temporal distribution study was done by pooling data from three different experiments with three animals in each time point (n=63 animals). The RT-PCR ratio values were analyzed using GLM Frequency and Correlation procedures. Significance was calculated by using Student’s t-test.


Brains harboring tumors were removed on specified days after implantation of tumor cells. Prior to removing the brain, rats were sacrificed by lethal injection according to approved institutional guidelines and perfused with 10% formalin and acetic acid. Their brains were cut into 2-mm slabs prior to paraffin embedding. Sections 6-8 [mu]m thick were stained with hematoxylin and eosin for histological and volumetric studies. Images were accrued using the NIH Imaging program coupled with frame grabber software. For immunostaining of paraffin sections, the sections were first dewaxed and then quenched with 4:1 methanol/peroxide. For the angiogenesis study, the neovasculature was highlighted by staining with polyclonal antibody against von Willebrand factor (vWF; DAKO A0082, 1:100 dilution; DAKO Corp., Carpinteria, CA, USA), and subsequently developed and visualized with DAB substrate and the Vectastain(R) Elite ABC system (Vector Labs, Burlingame, CA, USA). The slides were pre-treated with 0.3% pepsin to optimize the exposure of the antigen. Volumetric analysis of tumor growth was done using paraffin sections and re-constructed using MRC image analysis software. The data were pooled from three different experiments in which three animals were used for each time point. Duplicate sections per animal were used. Vascular density analysis in Figure 2 was performed on H+E slides prepared from paraffin sections. Five random fields at 40x original magnification were examined, this was done by using the MicroBrightfield StereoInvestigator software. Data were expressed as mean + or – standard deviation/mm^sub 2^. Significance was calculated using Student’s t-test.

For immunofluorescence staining and confocal microscopy, the animals were perfused with cold saline buffer followed by a 4% neutral paraformaldehyde phosphate buffer (PFPB). The harvested brains were immersed and kept at 40C in PFPB, and later were immersed in a 30% sucrose solution prior to cutting into 40-??? serial sections on a freezing microtome. The staining of rat vascular endothelial cells was done using RECA (mouse anti-rat RECA-1, Serotec HIS52, 1 :50 dilution; Serotec, Raleigh, NC, USA) and vWF. For dual immunofluorescence, FITC- and Texas Red-labeled secondary anti-bodies were used to visualize the bound antibodies (Jackson Immunoresearch, USA). The cells were counter-stained with DAPI nuclear stain. The stained cells were viewed with a Zeiss epifluorescence microscope (Axioskop) and the images were digitally recorded with AxioVision software. The specimens were examined under low (100x) and high (400x) power fields. Over six random fields were examined per slide. Each field was divided into four quadrants. Two replicate slides per animal were examined. The areas defined for the analyses were adjusted slightly between sections to obtain the best representation of blood vessel distribution within individual sections while avoiding artifacts that were incidentally identified as blood vessels by the software during test count on each image. Photomicrograms were digitally captured for calculation of vascular density and vessel size within these areas. Statistical analysis was carried out and expressed as mean + or – SD. Diameter of vessels was defined as the smallest vector measurement between two edges of an outlined blood vessel. Statistical significance was calculated by using Kruskal-Wallis test and expressed as medians, means + or – SD. Selected sections were further documented by using a laser scanning confocal microscope (MCR-1024E, BioRad, USA). Laser pre-tuned to 488n m was used to excite the FITC-conjugated secondary antibodies and 543 nm was used to excite Texas Red-labeled secondary antibody. Band pass 505-550 nm and 565-615nm filters were used with beam splitter to obtain images. Sections were scanned at 1 [mu]m intervals, compressed by the accompanying software, and averaged to optimize fluorescent signals and decrease noise. Images shown in Figures 2 and 3 were digitally assembled with Adobe Photoshop.


In vivo tumor growth

Implantation of RT-2 cells into the basal ganglia of syngeneic rats resulted in tumor formation. The engrafted cells, initially concentrated along the injection tract (Figure 1A), subsequently dispersed and accumulated around pre-existing blood vessels and at sites close to the subependymal zone. By the fourth day after implantation, a solid tumor nodule had become apparent. Blood vessels could be recognized around and within the tumor. Mean tumor volume increased from 0.52 + or -0.63 mm^sup 3^ on the fourth post-implant day to 1.05 + or – 0.13 mm^sup 3^, 1.37 + or – 0.04 mm^sup 3^, 4.61 + or – 1.31 mm^sup 3^, 8.658 + or – 1.31 mm^sup 3^, 10.16 + or – 1.38 mm^sup 3^ on the sixth, eighth, tenth, twelfth, and the fourteenth days, respectively (Figure 1B). The growing tumors became infiltrative (Figure 1; Day 8) and took on a whorled appearance (Figure 2A). As the tumor expanded, it became more necrotic and hemorrhagic, The hemorrhagic area was less than 1 % on Day 6 and increased to about 20% of the tumor volume by Day 14 (Figure IA). The necrotic and hemorrhagic areas were concentrated at the center of the tumor and surrounded by an area of highly proliferative cells as judged by cells with mitotic figures that were readily noticeable under high magnification. These actively dividing cells represented the growing and invading margin of the tumor. Vascular growth and distribution within the tumor Vascular formation within the tumor mass and its surrounding brain was visualized using histological and immunodetection methods. Light microscopy, fluorescence, and laser scanning confocal microscopy were used to optimize the visualization and to address vascular density, vascular dimension, and the expression of vWF and RECA (Figures 2 and 3). There were noticeable changes in the tumor vasculature throughout the period of tumor expansion. At the early stages of tumor formation (within six days of implantation), blood vessels within the tumor were small in caliber, 13.0 + or – 1.7 microns (Figures 2B and 3). Although the tumor appeared to be densely populated with vessels, vWF staining was light and the overall density of new vessels was low (Figures 2B and 3). A progressive increase in neovascular density occurred from Day 6 to Day 10 (Figures 2C and 3), from 4.2 + or – 3.0 vessels to 18.3 + or – 1.5 vessels per mm unit area. There was a concomitant increase in the arborization and caliber of the tumor vessels (Figure 3). The arborization of sprouting of vessels seemed to have slowed down after Days 10-12. By Days 10 and 12 there were significant increases in vessel caliber relative to Days 6 and 8, from 13.0 + or – 1.7 microns to 37 + or – 19.9 microns (with significant scatter in the vascular caliber, p

Topographical distribution of growth factors in the established tumor and its perimeter

The expression of Ang-1 and Ang-2 together with VEGF was examined by semi-quantitative RT-PCR. The tumor mass was isolated from its surrounding brain by microdissection (Figure 4A). The perimeter of the tumor was defined as the peripheral region of the tumor where there were identifiable tumor cells separated from the major bulk of the tumor. Measurement of mRNA in tumor and brain harvested on Day 12 (Figure 4B,C) showed that Ang-2 was more prominent at the tumor’s perimeter (1.10 + or – 0.25 relative expression units) and within the tumor mass (1.07 + or – 0.15 relative expression units) than in the overlying cortex (0.11 + or – 0.04 relative expression units) and contralateral hemisphere (0.12 + or – 0.01 relative expression units). Conversely, the expression of VEGF was more pronounced within the tumor (1.59 + or – 0.35 relative expression units) than at its perimeter (0.65 + or -0.35 relative expression units), or in the contralateral hemisphere (0.32 + or -0.14 relative expression units). The expression of Tie-2 was more pronounced at the perimeter of the tumor (0.38 + or -0.21 relative expression units). The expression of Ang-1 was of interest. It was higher at the tumor’s perimeter (0.98 + or – 0.04 relative expression units) than in the tumor mass (0.73 + or – 0.07 relative expression units) or in the surrounding cortex (0.62 + or – 0.02 relative expression units) and contralateral hemisphere. However, the differences of its topographical distribution were not as statistically significant as those calculated for VEGF, Ang-2, and Tie-2. Nevertheless, by having a higher level of Ang-1, Ang-2, and Tie-2 within the tumor perimeter might explain the discrepancy of the vascular density and appearance of blood vessels between different areas of the tumor. It was noteworthy that during this period of tumor growth, the majority of the vessels were found in the periphery of the tumor. The center of the tumor mass was dominated by necrotic cells and hemorrhagic areas. Furthermore, at the center of the tumor the diameters of the vessels were persistently larger and with a chaotic appearance (Figure 3). In contrast, vessels at the perimeter of the tumor were of smaller caliber and of a more normal appearance. The geographically distinct expression of these angiogenic modulators within and around the tumor further suggests their specific roles in the formation of the growing tumor’s vascular tree. In the presence of VEGF and Ang-2 there would be vascular proliferation but the integrity of these newly formed vessels might not be maintained without Ang-1.

The temporal expression of VEGF, Ang-1, Ang-2, Tie-2

Semi-quantitative RT-PCR was used to measure changes in mRNA expression of these pro-angiogenic factors at different time points during tumor growth. Expression was presented relative to that of [beta]-actin. In vitro, RT-2 cells have a barely detectable level of Ang-2, but have a more noticeable level of Ang-1, Tie-2, and VEGF (results not shown). In vivo, Ang-2 expression increased starting at Day 4 post-implantation (0.09 + or – 0.02 relative expression units, Figure 5B), and reached a plateau at Day 12 (1.25 + or – 0.16 relative expression units). VEGF showed a gradual increase over a similar time course. There was no significant change in levels of Tie-2 in this study, and for Ang-1 there was a slight decline at Day 10, but at Day 14 no significant difference was detected (Figure 5B). The lack of persistent expression of Ang-1 in the midst of tumor growth was of interest especially in the context of the increases in vessel caliber and in the static expression of vWF in the neovasculature particularly at the center of the tumor. These changes coincided with the increase in hemorrhage noted within the tumor and could be the cause of increased fragility of the newly formed vasculature. The increase of Ang-1 expression in the later phase of tumor growth correlated with the increase in neovascular expansion at the periphery of the tumor and with the re-appearance of smaller caliber vessels.


Rat RT-2 glioblastoma cells implanted in the basal ganglia of rat brains usually formed a 2-mm diameter tumor by the sixth day after implantation (Figure 1A). The tumor was created by the coalescence of tumor cells, previously forming perivascular cuffs around preexisting vessels, into a tumor mass. At the beginning, most of the tumor vessels were stained lightly with vWF but heavily with RECA suggesting that there was active ramification of pre-existing venules at this phase of tumor growth. As the tumor continued to expand there was an increase in endothelial proliferation and sprouting in which new vessels appeared within the stroma of the tumor, reflected by the increase in staining of antibodies to vWF and RECA (Figures 2 and 3). These new vessels initially displayed normal morphology but as the tumor grew, they increased in caliber and decreased in number of branching points. Concomitantly, hemorrhage and necrosis appeared within the tumor mass, suggesting that these vessels were fragile and leaky22. Eventually, areas of the tumor presumed to be avascular zones appeared. These areas were surrounded by the proliferative tumor margin. The vessels in this region of growth were of normal-looking, small caliber vessels intermixed with a few abnormal looking large lumen vessels. The small caliber vessels most likely represented newly formed vessels recruited from the surrounding brain, and vessels of the larger lumen represented the involuting vessels. These morphological changes (the expansion of pre-existing vessels, involution of these vessels, and subsequent recruitment of vessels) underscores the fact that tumor angiogenesis is a dynamic process and is tightly coupled with the proliferation of the tumor cells.

The sequential and geographical segregation of these developments may argue that central necrosis results from vascular involution, as opposed to our previous belief that the tumor outstrips its own blood supply23. The invasion and/or migration are, in essence, a type of natural selection process that propels the tumor toward a more favorable environment. Previous reports had shown tumor angiogenesis follows a well-defined scheme orchestrated by the temporal and spatial expression of Ang-1, Ang-2, and VECF. Ang-2 prompts co-opting of pre-existing vessels and later initiates tumor angiogenesis24, while VEGF and Ang-1 promote vessel sprouting12,25. The spatial expression of these factors within the tumor mass and its surrounding brain was measured using semi-quantitative RT-PCR. Expression of Ang-2 and VEGF was concentrated within the tumor and the immediate surrounding brain (Figure 4) as reported for other types of tumors26. The prominence of expression of Ang-2 at the perimeter and at the center of the tumor is reasonable given the hypothesized involvement of Ang-2 in the recruitment and ramification of host vessels. The greater expression of VEGF within the tumor mass than at its periphery suggests that VEGF may be more important to promotion of intratumoral growth of parasitized vessels than to their initial recruitment27. The greater expression of VEGF in this region may also be the result of it being up-regulated by hypoxia28-30. The expression of Tie-2 and Ang-1 was more robust at the perimeter of the tumor. Its co-expression with Ang-2 at the perimeter emphasizes the significance of the balance of these two factors in the creation and maintenance of a functional vasculature. It was not surprising to find a lower level of expression of Ang-2 within the core of the tumor since its expression and that of Tie-2 is negatively influenced by hypoxia31. The co-expression of Ang-1, Ang-2, and Tie-2 at the perimeter of the tumor at this stage of tumor growth may explain the relative normal looking vessels of a smaller caliber in this region.

The temporal expression of VEGF, Ang-1, Ang-2, and Tie-2 was monitored throughout tumor growth in an attempt to correlate the histological changes in the tumor with the morphological changes in its vascular architecture. Although one would expect expression of the Tie-2 receptor, known to be involved in the formation and differentiation of new vascular networks32,33, to increase during tumor angiogenesis, Tie-2 receptor expression did not change as the tumor grew. There are several possible reasons for this finding. First, some of the new tumor vessels may be composed of endothelial cells that do not yet express Tie-2. Second, the new tumor vessels are often mosaics of endothelial cells and tumor cells. The small differences in Tie-2 may be beyond the sensitivity of this detection method. In fact, this mosaicism may also explain the abnormal permeability and defective vasoactive responses to hypoxia and pressure changes known to occur34.

The expression of Ang-2 and VEGF increased as the tumors grew (Figure 5A). Since RT-2 cells in vitro express VEGF, Ang-1, and Tie-2 but not Ang-2, the increased Ang-2 expression in vivo may be attributed to local tissue effects35. In vivo, the time course of VEGF and Ang-2 expression were similar, reaching plateaus by the tenth day. Ang-2 expression increased 14-fold (1.25 + or – 0.16 relative expression units on Day 12 versus 0.09 + or – 0.02 on Day 4), while VEGF expression increased 6-fold. This enormous increase in Ang-2 level reflects its pivotal role in tumor angiogenesis. It plays a role in the dissolution of the basal matrix of pre-existing vessels and it allows sprouting to occur. The concomitant increases in VEGF with Ang-2 underscore their roles in promoting endothelial proliferation and migration and are reflected in the increase in diameter of the vessels36 and the stacking of endothelial cells within the vessel wall (Figure 2B). The expression of Ang-1 declined during tumor growth and by the tenth day after implantation was lower than that at basal level. On the fourteenth day after implantation, expression of Ang-1 within the tumor gradually returned to its baseline. The lack of persistence in the expression of Ang-1 in the midst of tumor growth was of interest. What caused the fall of Ang-1 expression is not clear, but in the ebbs of its expression and in the wake of Ang-2 expression, there would be an imbalance in the effects of these two angiopoietins on the Tie-2 receptors. This may cause a shift in the vascular tree from a senescent state to a more active state. Furthermore, a concomitant increase in the expression of VEGF and Ang-2 would cause fragmentation of the basal matrix and migration of proliferating endothelial cells. This imbalance may underlie the dysmorphism (dilated lumen, gaps in the endothelium, and intraluminal thrombosis) and fragility of the newly formed vasculature. These, in turn, may cause the increase in vascular permeability and hemorrhage evident in the center of the tumor.

The bimodal nature of the expression of Ang-1 may signify two phases of tumor angiogenesis as proposed by Holash et al.26. The inital phase occurs within the tumor mass where pre-existing blood vessels are ramified to support the expansion of the tumor. Ang-2 and VEGF are the dominant modulators responsible for this change37. In the absence of Ang-1, these vessels may not reach maturity and would subsequently dissipate; this is followed by a robust development of the tumor vasculature in support of the migrating/invading edge of the tumor, leaving the necrotic and hemorrhagic center so typical of a malignant glioma.


This study is supported in part by the Arjay and Frances Miller Brain Tumor Research Fund (to V.T.), ECS Brain Tumor Research Fund (to V.T. and Y.Y.), and Donald E. and Delia B. Baxter Foundation grants (to V.T.). K. Fabel was supported by the Deutsch Forschungsgemeinschaft (DFG). We thank George Yancopoulos, MD, PhD, and Eric Shooter, PhD, for their comments on the paper, B. Hoyte for help preparing the figures, and D. Schaal, PhD, for editing the manuscript. Some of the results presented in this paper were presented at the annual meeting of the Congress of Neurosurgeons 2002 (USA). The authors declare that they have no competing financial interests.


1 Scherer H. Gliomstudien III Angioplasticche Glioma. Virchow Arch 1935; 294: 823-826

2 Folkman J. How is blood vessel growth regulated in normal and neoplastic tissue? G.H.A. Clowes Memorial Award lecture. Cancer Res 1986; 46: 467-473

3 Holash J, Maisonpierre PC, Compton D, Boland P, Alexander CR, Zagzag D, et al. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VECF. Science 1999; 284: 1994-1998

4 Plate KH, Mennel HD. Vascular morphology and angiogenesis in glial tumors. Exp Toxicol Pathol 1995; 47: 89-94

5 Plate KH, Breier G, Risau W. Molecular mechanisms of developmental and tumor angiogenesis. Brain Pathol 1994; 4: 207-218

6 Folkman J, D’Amore PA. Blood vessel formation: What is its molecular basis? Cell 1996; 87: 1153-1155

7 Mochizuki Y, Nakamura T, Kanetake H, Kanda S. Angiopoietin 2 stimulates migration and tube-like structure formation of murine brain capillary endothelial cells through c-Fes and c-Fyn. J Cell Sci 2002; 115: 175-183

8 Stratmann A, Risati W, Plate KH. Cell type-specific expression of angiopoietin-1 and angiopoietin-2 suggests a role in glioblastoma angiogenesis. Am I Pathol 1998; 153: 1459-1466

9 Zagzag D, Friedlander DR, Margolis B, Grumet M, Semenza GL, Zhong H, et al. Molecular events implicated in brain tumor angiogenesis and invasion. Pediatr Neurosurg 2000; 33: 49-55

10 Vajkoczy P, Farhadi M, Gaumann A, Heidenreich R, Erber R, Wunder A, et al. Microtumor growth initiates angiogenic sprouting with simultaneous expression of VEGF, VEGF: receptor-2, and angiopoietin-2. J Clin Invest 2002; 109: 777-785

11 Yu Q, Stamenkovic I. Angiopoietin-2 is implicated in the regulation of tumor angiogenesis. Am J Pathol 2001; 158: 563-570

12 Koblizek TI, Weiss C, Yancopoulos GD, Deutsch U, Risau W. Angiopoietin-1 induces sprouting angiogenesis in vitro. Curr Biol 1998; 8: 529-532

13 Tian S, Hayes AJ, Metheny-Barlow LJ, Li LY. Stabilization of breast cancer xenograft tumour neovasculature by angiopoietin-1. Br J Cancer’2002; 86: 645-651

14 Thurston G, Rudge JS, Ioffe E, Zhou H, Ross L, Groll SD, et al. Angiopoietin-1 protects the adult vasculature against plasma leakage. Nat Med 2000; 6: 460-463

15 Audero E, Cascone I, Zanon I, Previtali SC, Piva R, Schiffer D, et al. Expression of angiopoietin-1 in human glioblastomas regulates tumor-induced angiogenesis: In vivo and in vitro studies. Arterioscler Thromb Vase Biol 2001; 21: 536-541

16 Kim I, Kim HG, So JN, Kim JH, Kwak HJ, Koh GY. Angiopoietin-1 regulates endothelial cell survival through the phosphatidylinositol 3′-Kinase/Akt signal transduction pathway. Circ Res 2000; 86: 24-29

17 Hanahan D. Signaling vascular morphogenesis and maintenance. Science 1997; 277: 48-50

18 Hswiao M, Tse V, Carmel J, Tsai Y, Feigner PL, Haas M, et al. Intracavitary liposome-mediated p53 gene transfer into glioblastoma with endogenous wild-type p53 in vivo results in tumor suppression and long-term survival. Biochem Biophys Res Commun 1997; 233: 359-364

19 Mandriota SJ, Pepper MS. Regulation of angiopoietin-2 mRNA levels in bovine microvascular endothelial cells by cytokines and hypoxia. Circ Res 1998; 83: 852-859

20 Lindgren M, Johansson M, Sandstrom J, Jonsson Y, Bergenheim AT, Henriksson R. VEGF and tPA co-expressed in malignant glioma. Acta Oncol 1997; 36: 615-618

21 Lin TN, Wang CK, Cheung WM, Hsu CY. Induction of angiopoietin and Tie receptor mRNA expression after cerebral ischemia-reperfusion. J Cereb Blood Flow Metab 2000; 20: 387-395

22 Wesseling P, van der Laak JA, de Leeuw H, Ruiter DJ, Burger PC. Quantitative immunohistological analysis of the microvasculature in untreated human glioblastoma multiforme. Computer-assisted image analysis of whole-tumor sections. J Neurosurg 1994; 81: 902-909

23 Scherer H. Cerebral astrocytomas and their derivatives. Am J Cancer 1940; 40: 158-198

24 Lauren J, Cunji Y, Alitalo K. Is angiopoietin-2 necessary for the initiation of tumor angiogenesis? Am J Pathol 1998; 153: 1333-1339

25 Hansbury MJ, Nicosia RF, Zhu WH, Holmes SJ, Winkler JD. Production and characterization of a Tie2 agonist monoclonal antibody. Angiogenesis 2001; 4: 29-36

26 Holash J, Wiegand SJ, Yancopoulos GD. New model of tumor angiogenesis: Dynamic balance between vessel regression and growth mediated by angiopoietins and VEGF. Oncogene 1999; 18: 5356-5362

27 Plate KH, Breier G, Welch HA, Risau W. Vascular endothelial growth factor is a potential tumour angiogenesis factor in human gliomas in vivo. Nature 1992; 359: 845-848 28 Song HS, Son MJ, Lee YM, Kirn WJ, Lee SW, Kirn CW, ef ai. Oxygen tension regulates the maturation of the blood-brain barrier. Biochem Biophys Res Commun 2002; 290: 325-331

29 Ikeda E, Achen MG, Breier G, Risau W. Hypoxia-induced transcriptional activation and increased mRNA stability of vascular endothelial growth factor in C6 glioma cells. J Biol Chem 1995; 270: 19761-19766

30 Shweiki D, Itin A, Soffer D, Keshet E. Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis. Nature 1992; 359: 843-845

31 Abdulmalek K, Ashur F, Ezer N, Ye F, Magder S, Hussein SN. Differential expression of Tie-2 receptors and angiopoietins in response to in vivo hypoxia in rats. Am J Physiol Lung Cell Mol Physiol 2001 ; 281: L582-L590

32 Koblizek Tl, Runting AS, Stacker SA, Wilks AF, Risau W, Deutsch U. Tie2 receptor expression and phosphorylation in cultured cells and mouse tissues. Eur J Biochem 1997; 244: 774-779

33 Huang L, Turck CW, Rao P, Peters KG. GRB2 and SH-PTP2: Potentially important endothelial signaling molecules downstream of the TEK/TIE2 receptor tyrosine kinase. Oncogene 1995; 11: 2097-2103

34 Chang YS, di Tomaso E, McDonald DM, Jones R, Jain RK, Munn LL. Mosaic blood vessels in tumors: Frequency of cancer cells in contact with flowing blood. Proc Natl Acad Sci USA 2000; 97: 14608-14613

35 Mandriota SJ, Pyke C, Di Sanza C, Quinodoz P, Pittet B, Pepper MS. Hypoxia-inducible angiopoietin-2 expression is mimicked by iodonium compounds and occurs in the rat brain and skin in response to systemic hypoxia and tissue ischemia. Am J Pathol 2000; 156: 2077-2089

36 Lobov IB, Brooks PC, Lang RA. Angiopoietin-2 displays VEGFR-dependent modulation of capillary structure and endothelial cell survival in vivo. Proc Natl Acad Sci USA 2002; 99: 11205-11210

37 Peoch M, Farion R, Hiou A, Le Bas JF, Pasquier B, Remy C. Immunohistochemical study of VEGF, angiopoietin 2 and their receptors in the neovascularization following microinjection of C6 glioma cells into rat brain. Anticancer Res 2002; 22: 2147-2151

Victor Tse*[dagger], Lei Xu*, Yun C. Yung*, Justin G. Santarelli*, David Juan*, Klaus Fabel*, Gerald Silverberg* and Griffith Harsh IV*

* Department of Neurosurgery, Stanford Medical School, Stanford, CA [dagger] Department of Surgery, Santa Clara Valley Medical Center, San Jose, CA, USA

Correspondence and reprint requests to: Victor Tse, MD, PhD, Department of Neurosurgery, Stanford University School of Medicine, MSLS Building, Room P310, 1201 Welch Road, Stanford, CA 94305-5487, USA. [vkt@stanford.edu] Accepted for publication June 2003.

Copyright Forefront Publishing Group Oct 2003

Provided by ProQuest Information and Learning Company. All rights Reserved