SEARCH

SEARCH BY CITATION

Keywords:

  • angiogenesis;
  • brain tumors;
  • endothelial cells;
  • matrix metalloproteinases;
  • multidrug resistance

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

Endothelial cells (ECs) are new targets for tumor therapy. In this work, we purified endothelial cells from intracerebral and subcutaneous experimental gliomas as well as from normal brain in order to define some of the phenotypical differences between angiogenic and quiescent brain vasculature. We show that the multidrug resistance genes encoding drug efflux pumps at the brain endothelium are expressed differently in normal and tumoral vasculature. We also show that ECs from gliomas present increased activity of gelatinase B (MMP9), key enzyme in the angiogenic process. Importantly, we observe a different phenotype between ECs in the intracerebral and subcutaneous models. Our results provide molecular evidence of phenotypic distinction between tumoral and normal brain vasculature and indicate that the EC phenotype depends on interactions both with tumor cells and also with the microenvironment.

Abbreviations used
CSF

cerebrospinal fluid

DMEM

Dulbecco's modified Eagle's medium

EC

endothelial cells

eNOS

endothelial nitric oxide synthase

MACS

magnetic cell sorting

MMP

matrix metalloproteinase

PBS

phosphate-buffered saline

P-gp

P-glycoprotein

SDS–PAGE

sodium dodecyl sulfate–polyacrylamide gel electrophoresis

TEM

tumor-specific endothelial marker

Tumor angiogenesis, the formation of a blood vessel network within a tumor, has become one of the most promising therapeutic target in cancer medicine (Matter 2001). Development of this approach will require determining the molecular differences between healthy and tumoral endothelium. In vitro, tumor-specific endothelial markers have been identified in endothelial cells exposed to tumor-conditioned media (Wang et al. 2000). In vivo, tumor-specific endothelial markers (TEMs) have been identified in endothelial cells (ECs) isolated from colorectal tissue (St Croix et al. 2000). Similar identification would be of particular utility for brain tumors as they are among the solid tumors most characterized by a high degree of neovascularity (Leon et al. 1996).

Brain ECs forming the blood–brain barrier (BBB) are sealed by tight junctions, have a low number of pinocytic vesicles and express specialized transport systems (Pardridge 1999). Those specific EC properties allow the BBB to maintain a precise control over substances entering or leaving the brain. The brain ECs' phenotype results from sustained interactions with their microenvironment, particularly with the surrounding astrocytes and their foot processes but also with the pericytes as well as with the extracellular matrix (Risau 1998). The growth of most primary brain tumors is accompanied by brain edema. A disregulation of the BBB junctional complex at the blood–tumor barrier has been associated to this phenomenon (Papadopoulos et al. 2001). Some BBB-specific transporters, such as the GLUT-1 transporter, have also been shown to be down-regulated in tumoral brain vasculature (Boado et al. 1994). However other specific transport systems at the BBB such as the Lutheran glycoprotein have been shown in brain tumor neovasculature (Boado et al. 2000). A molecular comparison working with purified populations of normal brain and glioma ECs remains to be established. We showed recently some phenotypical differences between brain, lung and kidney ECs (Demeule et al. 2001) using a magnetic cell sorting approach. In the present study, we used this approach to compare the phenotype of ECs isolated from normal brain, orthotopic or ectotopic gliomas. We assessed the expression of some important actors in brain tumors pathology such as the drug efflux pump P-glycoprotein (P-gp) which is implicated in the brain tumors resistance to chemotherapy and the matrix metalloproteinases (MMPs), involved in the degradation of a variety of extracellular matrix components for their important role in tumor progression. The results obtained provide already differences in expression or activity between intracerebral glioma ECs compared to normal brain ECs for the studied markers. Striking differences were also found between intracerebral and subcutaneous glioma ECs suggesting that the peritumoral environment remains determinant for the establishment of angiogenic ECs phenotype.

Chemicals

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

Magnetic cell sorting (MACS) was performed using microbeads, a MidiMacs separation unit and positive selection MACS columns from Miltenyi Biotec (Auburn, CA, USA). Mouse anti-PECAM-1 antibody that was linked to microbeads was from Cedarlane Laboratories (Hornby, ON, Canada) whereas goat anti-PECAM-1 antibody, used in western blotting analysis was from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The monoclonal C219 antibody was purchased from ID Laboratories (London, ON, Canada) and the p170/MDR Ab-2 was from NeoMarkers, Inc (Fremont, CA, USA). The monoclonal antibody directed against the endothelial nitric oxide synthase (eNOS) was from Transduction Laboratories (Mississauga, ON, Canada). Antibody directed against Glial Fibrillary Acidic Protein (GFAP) was from Sigma-Aldrich (Oakville, ON, Canada). Electrophoresis reagents were purchased from Bio-Rad (Mississauga, ON, Canada). IgG horseradish peroxidase-linked whole antibody were from Jackson ImmunoResearch Laboratories (West Grove, PA, USA) and the chemiluminescence (ECL) reagent was purchased from Amersham-Pharmacia Biotech (Baie-d'Urfé, QC, Canada). Others reagents were from Sigma Chemical (Oakville, ON, Canada).

Glioma cell lines

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

The CNS-1 murine glioma cell line, kindly provided by Dr W. F. Hickey (Hanover, NH, USA), was cultivated in RPMI 1640 supplemented with 10% fetal calf serum. The C6 glioma cell line obtained from ATCC (American Type Culture Collections) was cultivated in Dulbecco's modified Eagle's medium (DMEM) low glucose supplemented with 15% fetal calf serum.

Migration assay

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

A modified Boyden chamber assay was used. C6 or CNS-1 cells (5 × 104) in DMEM medium were seeded in the upper chamber of a 8-µm pore size polycarbonate filter (Corning Transwell, Fisher Scientific, ON, Canada) coated with 0.5% gelatin. DMEM medium containing 5% serum was placed in the lower chamber as a chemoattractant. The chambers were incubated for 4 h at 37°C with 5% CO2 and 95% humidity. The medium was then removed, the filters were fixed in 4% formaldehyde in phosphate-buffered saline (PBS) and the cells were colored with 1% crystal violet in 10% methanol. Non-migrating cells in the upper chamber were removed. An image of the migrated cells was obtained using a scanner (Snapscan, AGFA), the cells were then lyzed with 0.1% acetic acid and the absorbance was measured using a Thermomax microplate reader (Molecular Devices, Sunnyvale, CA, USA).

Tumor implantation

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

All animal experiments were evaluated and approved by the Institutional Comity for Good Animal Practices (UQAM, Montréal, QC, Canada). For intracerebral tumor implantation, anesthetized 250–280 g male Lewis rats (Charles River, QC, Canada) were placed in a stereotaxic frame (Stoelting, Xymotec Biosystems, North York, ON, Canada) and a burr hole was drilled into the right frontal cortex at a point 3 mm lateral to midline and 2 mm anterior to the bregma. Viable CNS-1 glioma cells (5 × 104) suspended in 5 µL of RPMI 1640 (containing 1% methylcellulose) were injected at a 4-mm depth using a Hamilton syringe. The syringe was removed after 5 min and the wound was closed with sutures. Animals were killed 25 days after infusion; the brains were recovered and either fixed in 10% buffered formalin for histopathological evaluation or dissected for endothelial cell isolation. Subcutaneous injection of CNS-1 cells (5 × 106) was performed in the right flank of Lewis rats. Animals were killed 25 days postinjection and the tumors were dissected for endothelial cell isolation.

Isolation of EC fractions

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

The isolation of ECs was performed as previously described (Demeule et al. 2001). Briefly, meninges-free cortex was homogenized in Ringer solution, mixed with Dextran T-70 and centrifuged to remove myelin. Intracerebral and subcutaneous glioblastomas were dissected and minced. Samples were then incubated with collagenase A (1 mg/mL) for 45 min at 37°C with agitation and the cell suspensions were passed through a Nitex filter (180 µm) and then another Nitex filter (30 µm). The pellets were washed three times in PBS containing 0.5% bovine serum albumin and 2 mm ethylenediamine-tetraacetic acid (EDTA). The final pellets were resuspended in 80 µL of PBS with 0.5% bovine serum albumin and 2 mm EDTA per 107 cells. Microbeads linked to the anti-PECAM-1 antibody (20 µL/107 cells) were added to the cell suspension for 45 min at 4°C. The cells were washed by centrifugation (600 g, 10 min) and resuspended in a volume of 500 µL. The ECs, bound by the magnetic microbeads linked to the anti-PECAM-1 antibody, were selected with the separation unit. The cell fractions retained by the column were washed with PBS followed by centrifugation at 600 g for 5 min at room temperature. The final pellets containing ECs were kept at −80°C until used.

A small proportion of isolated cell fractions was plated out on collagen I coated six-well plates in Ham's F-10 medium containing 15% heat-inactivated fetal calf serum, 2 mm glutamine, 50 µg/mL gentamicin and 75 µg/mL endothelial cell growth supplement. Cultures were maintained in humidified 5% CO2/95% air at 37°C. Medium was changed every other day.

Western blot analysis

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

Proteins from tissue homogenates and from EC fractions were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE). Protein samples were suspended in sample buffer composed of 62.5 mm Tris-HCl pH 6.8, 10% glycerol, 2% SDS, 5%β-mercaptoethanol and 0.00625% bromophenol blue. Samples were heated for 5 min at 95°C, except for P-gp detection. After electrophoresis, proteins were transferred electrophoretically to a polyvinylidene difluoride membrane. P-gp, eNOS, PECAM-1 and GFAP were detected on western blots using specific antibodies. Mab C219 (1 : 500) and Mab p170 Ab-2 (1 : 500), directed against P-gp, and antibodies against eNOS (1 : 1000), PECAM-1 (1 : 500) and GFAP (1 : 500) were used as previously described (Demeule et al. 2001). Horseradish peroxidase-conjugated anti-IgGs were used as secondary antibodies and P-gp, eNOS, PECAM-1 and GFAP were revealed with ECL reagents according to the manufacturer's instructions. Protein concentration was determined with the micro BCA protein assay kit from Pierce (Rockford, IL, USA).

RT-PCR analysis

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

Total RNA was isolated from normal rat brain, intracerebral and subcutaneous tumors and from EC fractions using TRIzol Reagent according to the manufacturer's instructions (Gibco BRL, Gaithersburg, MD, USA). Subsequently, mRNA was amplified with the Master Amp™ RT-PCR kit (MasterAMP, Madison, WI, USA). The following oligonucleotides for rat cDNAs were used, as previously reported (Demeule et al. 2001): mdr1a sense primer 5′-GATGGAATTGATAATGTGGACA-3′, mdr1a antisense primer, 5′-AAGGATCAGGAACAATAAA-3′, mdr1b sense primer, 5′-GAAATAATGCTTATGAATCCCAAAG-3′; mdr1b antisense primer, 5′-GGTTTCATGGTCGTCGTCTCTTGA-3′; GFAP sense primer 5′-CGTTTACCAGGCAGAACTTCGG-3′; GFAP antisense pri-mer 5′-TGGCGGCGATAGTCATTAGC-3′; GAPDH sense primer, 5′-CCATCACCATCTTCCAGGAG-3′; GAPDH antisense primer 5′-CCTGCTTCACCACCTTCTTG-3′. The expected sizes of the PCR products of mdr1a, mdr1b, GFAP and GAPDH are 351 bp, 325 bp, 522 bp and 540 bp, respectively. For each primer pairs, PCR amplifications were first performed for varying number of cycles (25, 30, 35). A number of cycles corresponding to the linear range of amplification was then applied. Control for genomic contamination of RNA preparations was performed by omitting the reverse transcriptase step. The PCR products were separated by 2% agarose gel electrophoresis and visualized by UV light in the presence of ethidium bromide.

Zymographic analysis

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

Substrate gel zymography of normal brain, homogenates and isolated ECs from normal brain, intracerebral or subcutaneous gliomas were performed as previously described (Beaulieu et al. 1999). The same amount of protein (25 µg) from each sample was denaturated in sample buffer as described for western blotting but without β-mercaptoethanol. The samples were agitated for 20 min at room temperature and loaded without boiling on a 7.5% SDS–acrylamide/bisacrylamide gel containing 0.1% (w/v) gelatin. After electrophoresis, the gels were soaked in 2.5% Triton X-100 (2 × 15 min) and rinsed five times in Nanopure™ water. The gels containing gelatin were incubated at 37°C for 20 h in a buffer containing 50 mm Tris-HCl, 20 mm NaCl, 5 mm CaCl2 and 0.02% Brij-35, pH 7.6. In control experiments, 1,10-phenantroline (0.5 mm final concentration) was added to the incubation buffer. Gels were then stained in 0.1% Coomassie blue in 30% methanol and 10% acetic acid and destained in the same solution without Coomassie blue.

Densitometric analysis

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

The intensities of the bands obtained from western blots analysis were estimated with a Personal densitometer SI (Molecular Dynamics, Sunnyvale, CA, USA). Enrichment factors are ratios of relative band density of proteins detected by western blots in isolated ECs and tissue homogenates. Results shown are means ± SE from three independent isolations of ECs.

Growth characteristics of CNS-1 model

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

We first compared the in vitro invasive characteristics of the CNS-1 glioma cell line to those of the widely used C6 glioma cell line. The migration capacity of CNS-1 cells is twofold higher than that of C6 cells (Fig. 1a). Histopathological examination of rat brain sections bearing CNS-1 glioma shows morphological characteristics of human glioblastoma multiforme, including necroses with palisading cells in the tumor center area (Fig. 1b), perivascular spreading of tumor cells and tumoral cell permeation of the adjacent parenchyma at the periphery (Fig. 1c). EC hyperplasia is observed in various sections of the tumors (Fig. 1d).

image

Figure 1. In vitro and in vivo characteristics of CNS-1 glioma cells. (a) Comparison of CNS-1 and C6 glioma cell migration capacity in a modified Boyden chamber assay. (b) Histological characteristics of CNS-1 intracerebral gliomas. Hematoxylin–eosin staining shows that the tumor center presents necrotic areas with palisading cells at the borders (× 250). (c) Perivascular spreading of tumor cells with single cell permeation of the adjacent parenchyma are found at the tumor border (× 250). (d) The vascular endothelial cells within the tumor present a tumescent aspect characteristic of a proliferative state. Left panels (× 250). Right panels are magnified.

Download figure to PowerPoint

EC markers expression in isolated cell fractions

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

ECs are positively selected with a monoclonal antibody directed against the endothelial marker platelet-endothelial cell adhesion molecule-1 (PECAM-1, or CD31). The two EC markers, PECAM-1 and eNOS, were immunodetected in the sample homogenates and in the final isolated EC fractions. Both eNOS and PECAM-1 are present in the three isolated EC fractions at levels much higher than in homogenates (Fig. 2a). eNOS expression is enriched 42-fold in the EC fraction from normal brain and by 52 (± 10)- and 41 (± 8)-fold in intracerebral and subcutaneous EC fractions, respectively. PECAM-1 expression is enriched by 32(± 7)-fold in the EC fraction from normal brain and by 29(± 5)- and 25(± 4)-fold in intracerebral and subcutaneous EC fractions, respectively. The astrocyte-derived glial fibrillary acidic protein GFAP is undetected in the final isolated cell fractions whereas it is detected in CNS-1 cells and in the three tissue homogenates (Fig. 2a).

image

Figure 2. Marker expression analysis of isolated ECs. (a) Immunodetection of endothelial (eNOs, PECAM-1) and glial (GFAP) markers in homogenates (H) and endothelial cells (EC) isolated from normal brain and intracerebral or subcutaneous CNS-1 gliomas. Protein samples (25 µg) were loaded on SDS–PAGE and transferred to a polyvinylidene difluoride membrane. (b), Isolated cell fractions were plated on collagen I coated plates and grow without further purification. Pictures were taken using a MicroCam (Polaroid). ECs isolated from normal brain using one of three different experiment are shown (× 100).

Download figure to PowerPoint

In culture, the isolated cell fractions show the typical growth characteristics of primary ECs in culture (Fig. 2b). They first form small islets and further grow as a continuous monolayer without the need of any further purification to remove contaminants. Those primary cultures were estimated to be 100% pure by morphological criteria. No morphological difference was observed between normal and tumoral ECs.

Expression of P-glycoprotein

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

P-gp is a membrane transport protein that confers the multidrug resistance phenotype by which a cancer cell, exposed to a single anticancer drug, becomes simultaneously resistant both to that drug and to other drugs of unrelated structure and function. Immunodetection of P-gp in CNS-1 glioma cells, homogenates and isolated ECs from normal brain as well as from intracerebral or subcutaneous glioma was performed by western blot analysis using two different antibodies (Fig. 3a). In CNS-1 cells, P-gp is detected only with the C219 antibody. In normal brain and brain tumor homogenates P-gp is detected with the C219 antibody when a long exposure of the film is performed whereas a very short exposure is sufficient to detect P-gp with both antibodies in isolated ECs from normal brain and intracerebral glioma. Thus, P-gp expression is highly enriched in ECs isolated from normal brain as well as intracerebral glioma, whereas no expression was found in isolated ECs from subcutaneous glioma. The enrichment levels of the ECs isolated from intracerebral glioma are similar to those obtained with ECs from normal brain [64(± 11)- and 83(± 15)-fold, respectively]. In rodents, P-gp is encoded by three genes (mdr1a, mdr1b and mdr2). Both mdr1a and mdr1b have been associated with multidrug resistance. The expression of mdr transcripts in CNS-1 cells and in homogenates and isolated ECs from normal brain, intracerebral and subcutaneous glioma was determined by RT-PCR analysis (Fig. 3b). CNS-1 cells in culture express only the mdr1b transcript as mdr1a was undetectable even after 40 cycles of PCR. In tissue homogenates, mdr1a is expressed in normal brain and intracerebral glioma but not in subcutaneous glioma. Expression of the individual mdr isoforms were very different in ECs isolated from the three samples, showing that the high level of P-gp detected by western blot in isolated normal brain ECs is due solely or primarily to the expression of the mdr1a-encoding gene, whereas both mdr1a and mdr1b are present in intracerebral glioma ECs. The expression of the multidrug resistance-associated (MRP1), a membrane transporter related to-non-P-gp multidrug resistance, was also determined in all samples by RT-PCR analysis (Fig. 3b). The Mrp1 transcript was amplified in CNS-1 cells in culture, in the three tissue homogenates as well as in the three ECs isolated fractions. Expression of the glial marker GFAP was determined in all samples by RT-PCR analysis (Fig. 3b). Whereas the GFAP transcript was strongly amplified in the three tissue homogenates, it remains undetectable in the three EC isolated fractions even after 35 cycles.

image

Figure 3. Expression of P-gp in isolated ECs. (a) Immunodetection of P-gp in homogenates (H) and endothelial cells (EC) isolated from normal brain, i.c. or s.c. CNS-1 gliomas. Protein samples (10 µg) were loaded on SDS–PAGE and transferred to a polyvinylidene difluoride membrane. (b) RT-PCR analysis of rat mdr1a, mdr1b, and mrp1 gene expression in RNA samples purified from homogenates and from isolated ECs from normal brain, intracerebral or subcutaneous CNS-1 gliomas. RT-PCR analysis of GFAP and GAPDH genes are used as controls. Results shown are from a single experiment using one of three different EC isolations (n = 3).

Download figure to PowerPoint

Matrix metalloproteinases activity

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

We used the gelatin-zymography assay to measure the levels of pro- and activated MMP activity (Fig. 4a). No gelatinase activity is seen with normal brain homogenate. Intracerebral tumors show a MMP-9 activity which is undetectable in subcutaneous tumors. Both intracerebral and subcutaneous tumor homogenates present both pro- and actived forms of MMP-2, which is the major MMP activity expressed by CNS-1 cells. Considerable differences in gelatinase activity are seen between ECs isolated from the three tissues. The most significant difference is the high increase in MMP-9 activity in glioma ECs as compared to normal brain ECs. Phenantrolin control shows that all the activities detected in zymography belonged to MMPs. We investigated the appearance of the basal lamina surrounding vessels in intracerebral gliomas compared to vessels in the parenchyma. In normal parenchyma adjacent to the tumor, reticulin staining permits visualization of the basal lamina at the edge of vessels whereas it is absent around vessels in the tumor (Fig. 4b).

image

Figure 4. MMPs activity in isolated ECs. (a) Zymography-gelatin of homogenates and of ECs isolated from normal brain, intracerebral or subcutaneous CNS-1 glioma. A representative experiment from three different EC isolations is shown. (b) Reticulin staining of the basal lamina around vessels in normal brain parenchyma versus intracerebral glioma. In the normal parenchyma the vessels are surrounded by a basal lamina visualized by the black silver impregnation (top panel) (× 400) (arrows). In the tumor center the presence of red blood cells permits visualization of the vessels, but no basal lamina is found (bottom panel) (× 400) (arrows).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

The most commonly used experimental rat brain glioma cell lines, such as 9L gliosarcoma cells or C6 glioma cells, grow as well demarcated tumors with a limited infiltrative pattern (Barth 1998) whereas the CNS-1 cell line has been previously reported to grow with a more infiltrative pattern, closer to what is observed in human gliomas (Kruse et al. 1994). The high migration capacity of the CNS-1 cell line may explain the greater ability of intracranially implanted CNS-1 cells to invade adjacent normal brain. In experimental brain tumors, the pseudopalisading pattern and the concomitant development of necrosis have been previously associated with the presence of an angiogenic switch (Peoc'h et al. 1999). Moreover, endothelial cell hyperplasia in tumors has been shown to be an important indicator of occurring angiogenesis (Beranek 2002). Taken together, these observations show that the CNS-1 model presents anatomical and morphological characteristics, including induced angiogenesis, which validate its use to further investigate the molecular events associated with brain tumors.

The isolation of ECs from solid tissues is possible using magnetic cell sorting. This approach allowed us previously to identify significant differences between brain, lung and kidney ECs (Demeule et al. 2001). In the present study, we used this approach to compare the phenotype of ECs isolated from intracerebral or subcutaneous gliomas, and from normal brain. eNOs expression was highly enriched in the final isolated cells fraction. In normal brain as well as in brain tumors, eNOS expression is restricted to ECs (Miyawaki et al. 1995; Iwata et al. 1999). PECAM-1 expression was also highly enriched in the three isolated cell fractions. PECAM-1 expression within the central nervous system is confined to ECs of the blood–brain barrier (Williams et al. 1996). In gliomas, expression of PECAM-1 has also been observed in all ECs (Aroca et al. 1999). GFAP mRNA and GFAP protein were undetectable in the isolated cell fractions with the methods and levels of detection used, indicating that this purified cell population is not contaminated by either glial cells or tumor cells. These results indicate that the final isolates from the three tissues are highly enriched in ECs. GFAP is expressed at a much lower level in subcutaneous glioma homogenate than in intracerebral glioma homogenate. This lower expression is observed both at the ARNm and protein levels. This shows that the intracerebral environment exerts a paracrine regulation over the tumor cells that maintains the expression of GFAP. This regulation is absent in the subcutaneous model.

The efflux pump P-gp first described for its role in cancer cells multidrug resistance is also expressed in normal brain capillaries. It plays an important role by preventing many hydrophobic molecules from crossing the blood–brain barrier (BBB) and reaching the brain parenchyma. In CNS-1 cells P-gp was detected only with the C219 antibody. By RT-PCR only the mdr1b gene was amplified in CNS-1 cells. This suggests that the p170/MDR Ab-2 raised against human MDR1 encoding P-gp, reacts only with the rat mdr1a encoding P-gp isoform. In accordance with this observation the C219 antibody which recognizes all P-gp isoforms allowed the detection of P-gp in normal brain and tumor homogenates, as it detects the protein expressed by the glial or tumor cells as well as by the ECs. No such detection was observed in tissue homogenates with p170/MDR Ab-2 as it detects only the mdr1a encoding P-gp expressed by ECs that are in very low proportion in homogenates. Those observations amplified again the level of enrichment obtained after the ECs isolation. We reported previously that, in brain, P-gp is predominantly expressed in the luminal membrane of ECs (Beaulieu et al. 1997; Demeule et al. 2001). In the current study we show for the first time that P-gp is also expressed by freshly isolated intracerebral glioma ECs. This is in accordance with immunohistochemical studies which showed strong expression of P-gp in almost all ECs of primary brain tumors (Becker et al. 1991; Toth et al. 1996; Sawada et al. 1999). Furthermore it has been shown that ECs of tumor blood vessels express P-gp at the same level as do their normal counterparts (Toth et al. 1996). The up-regulation of mdr1b also occurs in ECs cultured from brain capillaries (Régina et al. 1998; Seetharaman et al. 1998). This up-regulation has been associated with a de-differentiation of ECs in culture which are no longer subject to the paracrine regulation of the surrounding astrocytes. This up-regulation of the mdr1b gene, concomitant with expression of the brain endothelium specific mdr1a gene observed here in vivo, suggests that some important barrier properties are maintained in the angiogenic vessels which develop in brain tumors even if EC de-differentiation occurs. The fact that no GFAP mRNA amplification was obtained with ECs fractions strongly suggests that the mdr1b expression observed is not due to tumor or glial cells contaminating the EC fractions. Neither mdr1a nor mdr1b transcripts were amplified from subcutaneous glioma ECs, in agreement with our inability to immunodetect P-gp in these samples. A similar observation was made where the expression of GLUT-1 was found to be completely different in intracerebral versus subcutaneous gliomas (Arosarena et al. 1994). Mrp1 gene expression was found in all the samples analyzed. In normal Mrp1 brain1 expression has been shown both in the endothelium and in the brain parenchyma (Régina et al. 1998). Mrp1 expression has been related to the intrinsic resistance of human brain tumors (Mohri et al. 2000; Haga et al. 2001). In these studies no distinction was made between expression in the tumor vasculature or the tumor cells themselves. We show here that both cell types express the Mrp1 gene. These findings demonstrate that the peritumoral environment influences EC differentiation within glioma tumors. Glioma cells may also maintain the barrier properties in vessels from surrounding tissue that infiltrate the tumor. Thus, in brain tumors the multidrug resistance phenomenon may be due both to the ECs and to the tumor cells. As P-gp expression in brain tumor vasculature might be involved in the high resistance of gliomas to chemotherapy, studies using intracerebral models may be more appropriate as P-gp disappears from the vasculature of subcutaneous CNS-1 model. MMPs are crucial to glioma invasion and angiogenesis (Van Meter et al. 2001). Our results show that brain tumor cells primarily express a MMP-2 activity whereas ECs generally express a MMP-9 activity. The most striking change between normal brain ECs and intracerebral brain tumor ECs is the high increase in MMP-9 activity. Studies using in situ hybridization and immunohistochemistry has also shown that, in human gliomas, MMP-2 expression is primarily detected in glioma cells whereas MMP-9 expression is predominantly in vascular structures (Vince et al. 1999; Raithatha et al. 2000). No comparison between normal and tumoral brain was made in those studies. Interestingly, it has been reported that the cerebrospinal fluid (CSF) of patients with malignant gliomas contains MMP-2 and MMP-9 whereas only MMP-2 is found in the CSF of healthy patients (Friedberg et al. 1998). Furthermore, a study with transgenic mice showed recently that during pancreatic islet carcinogenesis, the absence of MMP-2 does not impair angiogenesis whereas expression of MMP-9 in the vasculature was an essential component of the angiogenic switch (Bergers et al. 2000). Our study supports the idea that MMP-9, and not MMP-2, is the major matrix-degrading enzyme expressed by angiogenic ECs. The difference between normal and tumoral brain endothelium extracellular matrix visualized by reticulin staining may result from the increased expression of matrix-degrading enzyme MMP-9 that we observed in ECs isolated from intracerebral gliomas. The combination of these results shows that the tumor cells surrounding ECs in gliomas are able to influence the invasive phenotype of the ECs. In contrast to intracerebral glioma ECs, a near-absence of gelatinase activity is seen with subcutaneous glioma ECs, providing additional evidence for a paracrine regulation which occurs only in the intracerebral glioma model, which may be representative of regulation taking place in human brain tumors.

The analysis of in vivo-derived ECs is a step above in vitro cell culture analysis. Our results show that there are strong molecular differences in the phenotypes of normal and tumoral brain endothelium as shown by differences in the expression of important targets for brain cancer therapy, such as P-gp and MMPs. The establishment of specific tumor cell properties has been shown before to depend on tumor cells implantation at their histological origin (for review, see Killion et al. 1999). We further demonstrate that the same is true for ECs within the tumors as we demonstrate ECs phenotypical differences whether the tumor cells are inoculated orthotopically or ectotopically. Such observations may have significant implications for the development of antiangiogenic therapies.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References

We thank Dr William F Hickey (Darmouth Medical School, Lebanon, NH) for providing us with the CNS-1 cell line. This work was supported by grants to RB from the Natural Sciences and Engineering Research Council of Canada. AR received a postdoctoral fellowship from the Fonds de la Recherche en Science du Québec.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Chemicals
  5. Glioma cell lines
  6. Migration assay
  7. Tumor implantation
  8. Isolation of EC fractions
  9. Western blot analysis
  10. RT-PCR analysis
  11. Zymographic analysis
  12. Densitometric analysis
  13. Results
  14. Growth characteristics of CNS-1 model
  15. EC markers expression in isolated cell fractions
  16. Expression of P-glycoprotein
  17. Matrix metalloproteinases activity
  18. Discussion
  19. Acknowledgements
  20. References
  • Aroca F., Renaud W., Bartoli C., Bouvier-Labit C. and Figarella-Branger D. (1999) Expression of PECAM-1/CD31 isoforms in human brain gliomas. J. Neurooncol. 43, 1925.
  • Arosarena O., Guerin C., Brem H. and Laterra J. (1994) Endothelial differentiation in intracerebral and subcutaneous experimental gliomas. Brain. Res. 640, 98104.
  • Barth R. F. (1998) Rat brain tumor models in experimental neuro-oncology: the 9L, C6, T9, F98, RG2(D74), RT-2 and CNS-1 gliomas. J. Neurooncol. 36, 91102.
  • Beaulieu E., Demeule M., Ghitescu L. and Beliveau R. (1997) P-glycoprotein is strongly expressed in the luminal membranes of the endothelium of blood vessels in the brain. Biochem. J. 326, 539544.
  • Beaulieu E., Kachra Z., Mousseau N., Delbecchi L., Hardy J. and Beliveau R. (1999) Matrix metalloproteinases and their inhibitors in human pituitary tumors. Neurosurgery. 45, 14321440.
  • Becker I., Becker K. F., Meyermann R. and Hollt V. (1991) The multidrug-resistance gene MDR1 is expressed in human glial tumors. Acta Neuropathol. (Berl.) 82, 516519.
  • Beranek J. T. (2002) Endothelial hyperplasia: an important indicator of actual angiogenesis. Br. J. Cancer. 86, 658.
  • Bergers G., Brekken R., McMahon G., Vu T. H., Itoh T., Tamaki K., Tanzawa K., Thorpe P., Itohara S., Werb Z. and Hanahan D. (2000) Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis. Nat. Cell. Biol. 2, 737744.
  • Boado R. J., Black K. L. and Pardridge W. M. (1994) Gene expression of GLUT3 and GLUT1 glucose transporters in human brain tumors. Brain. Res. Mol. Brain. Res. 27, 5157.
  • Boado R. J., Li J. Y. and Pardridge W. M. (2000) Selective Lutheran glycoprotein gene expression at the blood–brain barrier in normal brain and in human brain tumors. J. Cereb. Blood. Flow. Metab. 20, 10961102.
  • Demeule M., Labelle M., Regina A., Berthelet F. and Beliveau R. (2001) Isolation of endothelial cells from brain, lung, and kidney: expression of the multidrug resistance P-glycoprotein isoforms. Biochem. Biophys. Res. Commun. 281, 827834.
  • Friedberg M. H., Glantz M. J., Klempner M. S., Cole B. F. and Perides G. (1998) Specific matrix metalloproteinase profiles in the cerebrospinal fluid correlated with the presence of malignant astrocytomas, brain metastases, and carcinomatous meningitis. Cancer 82, 923930.
  • Haga S., Hinoshita E., Ikezaki K., Fukui M., Scheffer G. L., Uchiumi T. and Kuwano M. (2001) Involvement of the multidrug resistance protein 3 in drug sensitivity and its expression in human glioma. Jpn. J. Cancer. Res. 92, 211219.
  • Iwata S., Nakagawa K., Harada H., Oka Y., Kumon Y. and Sakaki S. (1999) Endothelial nitric oxide synthase expression in tumor vasculature is correlated with malignancy in human supratentorial astrocytic tumors. Neurosurgery 45, 2428.
  • Killion J. J., Radinsky R. and Fidler I. J. (1999) Orthotopic models are necessary to predict therapy of transplantable tumors in mice. Cancer Metastasis Rev. 17, 279284.
  • Kruse C. A., Molleston M. C., Parks E. P., Schiltz P. M., Kleinschmidt-DeMasters B. K. and Hickey W. F. (1994) A rat glioma model, CNS-1, with invasive characteristics similar to those of human gliomas: a comparison to 9L gliosarcoma. J. Neurooncol. 22, 191200.
  • Leon S. P., Folkerth R. D. and Black P. M. (1996) Microvessel density is a prognostic indicator for patients with astroglial brain tumors. Cancer 77, 362372.
  • Matter A. (2001) Tumor angiogenesis as a therapeutic target. Drug Discov. Today 6, 10051024.
  • Miyawaki T., Sohma O., Mizuguchi M. and Takashima S. (1995) Development of endothelial nitric oxide synthase in endothelial cells in the human cerebrum. Brain Res. Dev. Brain. Res. 89, 161166.
  • Mohri M., Nitta H. and Yamashita J. (2000) Expression of multidrug resiatance-associated protein (MRP) in gliomas. J. Neurooncol. 49, 105115.
  • Papadopoulos M. C., Saadoun S., Davies D. C. and Bell B. A. (2001) Emerging molecular mechanisms of brain tumour oedema. Br. J. Neurosurg. 15, 101108.
  • Pardridge W. M. (1999) Blood–brain barrier biology and methodology. J. Neurovirol. 5, 556569.
  • Peoc'h M., Le Duc G., Trayaud A., Farion R., Le Bas J. F., Pasquier B. and Remy C. (1999) Quantification and distribution of neovascularization following microinjection of C6 glioma cells in rat brain. Anticancer Res. 19, 30253030.
  • Raithatha S. A., Muzik H., Rewcastle N. B., Johnston R. N., Edwards D. R. and Forsyth P. A. (2000) Localization of gelatinase-A and gelatinase-B mRNA and protein in human gliomas. Neuro-oncology 2, 145150.
  • Régina A., Koman A., Piciotti M., El Hafny B., Center M. S., Bergmann R., Couraud P. O. and Roux F. (1998) Mrp1 multidrug resistance-associated protein and P-glycoprotein expression in rat brain microvessel endothelial cells. J. Neurochem. 71, 705715.
  • Risau W. (1998) Development and differentiation of endothelium. Kidney Int. Suppl. 67, S3S6.
  • Sawada T., Kato Y., Sakayori N., Takekawa Y. and Kobayashi M. (1999) Expression of the multidrug-resistance P-glycoprotein (Pgp, MDR-1) by endothelial cells of the neovasculature in central nervous system tumors. Brain Tumor Pathol. 16, 2327.
  • Seetharaman S., Barrand M. A., Maskell L. and Scheper R. J. (1998) Multidrug resistance-related transport proteins in isolated human brain microvessels and in cells cultured from these isolates. J. Neurochem. 70, 11511159.
  • St Croix B., Rago C., Velculescu V., Traverso G., Romans K. E., Montgomery E., Lal A., Riggins G. J., Lengauer C., Vogelstein B. and Kinzler K. W. (2000) Genes expressed in human tumor endothelium. Science 289, 11971202.
  • Toth K., Vaughan M. M., Peress N. S., Slocum H. K. and Rustum Y. M. (1996) MDR1 P-glycoprotein is expressed by endothelial cells of newly formed capillaries in human gliomas but is not expressed in the neovasculature of other primary tumors. Am. J. Pathol. 149, 853858.
  • Van Meter T. E., Rooprai H. K., Kibble M. M., Fillmore H. L., Broaddus W. C. and Pilkington G. J. (2001) The role of matrix metalloproteinase genes in glioma invasion: co-dependent and interactive proteolysis. J. Neurooncol. 53, 213235.
  • Vince G. H., Wagner S., Pietsch T., Klein R., Goldbrunner R. H., Roosen K. and Tonn J. C. (1999) Heterogeneous regional expression patterns of matrix metalloproteinases in human malignant gliomas. Int. J. Dev. Neurosci. 17, 437445.
  • Wang J. L., Liu Y. H., Lee M. C., Nguyen T. M., Lee C., Kim A. and Nguyen M. (2000) Identification of tumor angiogenesis-related genes by subtractive hybridization. Microvasc. Res. 59, 394397.
  • Williams K. C., Zhao R. W., Ueno K. and Hickey W. F. (1996) PECAM-1 (CD31) expression in central nervous system and its role in experimental allergic encephalomyelitis in the rat. J. Neurosci. Res. 45, 747757.