Clinical Cancer Research Prevention Award Frontiers in Basic Cancer Research
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carson-Walter, E. B.
Right arrow Articles by Walter, K. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carson-Walter, E. B.
Right arrow Articles by Walter, K. A.
Clinical Cancer Research Vol. 11, 7643-7650, November 1, 2005
© 2005 American Association for Cancer Research

Human Cancer Biology

Plasmalemmal Vesicle Associated Protein-1 Is a Novel Marker Implicated in Brain Tumor Angiogenesis

Eleanor B. Carson-Walter1,3, Jessica Hampton1, Eveline Shue1, Daniel M. Geynisman1, Pramod Kumar Pillai1, Ramasri Sathanoori1, Stephen L. Madden4, Ronald L. Hamilton2 and Kevin A. Walter1,2,3

Authors' Affiliations: Departments of 1 Neurosurgery and 2 Pathology, University of Pittsburgh; 3 Brain Tumor Center, Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania; and 4 Genzyme Corporation, Framingham, Massachusetts

Requests for reprints: Kevin A. Walter, Department of Neurosurgery, University of Pittsburgh, Suite B400, 200 Lothrop Steet, Pittsburgh, PA 15213. Phone: 412-647-1025; Fax: 412-647-0989; E-mail: walterka{at}upmc.edu.


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 References
 
Purpose: Plasmalemmal vesicle associated protein-1 (PV-1) is up-regulated in the endothelium of human glioblastoma. We sought to further characterize the expression pattern of PV-1 in human brain tumors and interrogate its role in brain tumor angiogenesis.

Experimental Design: Quantitative reverse transcription-PCR and in situ hybridization were used to measure PV-1 expression in a panel of 46 human brain tumors and related pathologic states. Matrigel tubulogenesis assays and cell migration assays were used to show function of PV-1 in primary human endothelial cells (HMVEC) under gene knockdown conditions.

Results: PV-1 is selectively up-regulated in a variety of high-grade human brain tumors, including glioblastoma and metastatic carcinoma, as well as other cerebral disorders associated with blood-brain barrier disruption, such as acute ischemia. Expression levels were reduced in low-grade neoplasia; however, tumors associated with the ependyma and choroid plexus, known sites of PV-1 expression, also exhibited robust expression. Cerebral expression of PV-1 mRNA was confined to endothelial cells in all cases. PV-1 expression was induced in HMVEC cells in vitro by exposure to medium conditioned by U87MG and U251MG human brain tumor cell lines and by medium supplemented with exogenous vascular endothelial growth factor or scatter factor/hepatocyte growth factor. RNA interference–mediated inhibition of PV-1 induction in HMVEC cells blocked Matrigel-induced tubulogenesis and inhibited cell migration induced by conditioned medium or angiogenic growth factors.

Conclusions: Our results confirm that PV-1 is preferentially induced in the endothelium of high-grade human brain tumors. Inhibition of PV-1 expression is associated with failure of endothelial differentiation in vitro. PV-1 represents a novel marker of brain tumor angiogenesis and integrity of the blood-brain barrier and is a potential therapeutic target.

Malignant brain tumors (gliomas) represent a uniformly fatal form of cancer. Despite advances in neurosurgical techniques, chemotherapeutic regimens, and radiotherapy protocols, the median survival following surgical resection and adjuvant therapy is <1 year (1, 2). New treatment approaches that target specific molecular features of gliomas are warranted. One such feature is the ability of gliomas to induce microvascular neoangiogenesis capable of promoting tumor growth. Inhibiting the formation of this microvascular network may slow or stop tumor growth. Developing therapies based on this principle depends on the characterization of specific markers of brain tumor angiogenesis that could serve as targets.

We previously used serial analysis of gene expression (SAGE) to analyze gene expression patterns in microvascular endothelial cells derived from clinical brain tumor specimens compared with nonneoplastic brains to identify glioma endothelial markers (3). Of the 122 glioma endothelial markers that were up-regulated >4-fold in tumor endothelium, five were especially unique because their expression pattern seemed to be limited to the microvasculature (Table 1). One of these genes, plasmalemmal vesicle associated protein-1 (PV-1), was initially identified as a cell surface protein expressed in immature rodent brain endothelium (4). Further studies indicated that PV-1 localized to caveolae and transendothelial channels in fenestrated endothelium (5, 6). Interestingly, PV-1 was repressed in normal central nervous system (CNS) in association with formation of an intact blood-brain barrier (5). Because PV-1 expression seemed limited to tumor endothelium in the mature brain and it was a cell surface protein accessible to extracellular pharmaceutical compounds, we felt that further development of this molecule as a target for antiangiogenesis and possibly anticerebral edema strategies was warranted.


View this table:
[in this window]
[in a new window]

  		Table 1. Five genes up-regulated ≥4-fold in brain tumor-derived endothelial cells compared with nonneoplastic brain endothelium

 
We now present data to support the selective induction of PV-1 in high-grade brain tumors as well as related pathologic states of the cerebrovasculature characterized by hypoxic conditions and disruption of the blood-brain barrier. We also show that PV-1 is induced by human brain tumor cell lines as well as the well-characterized proangiogenic growth factors, vascular endothelial growth factor (VEGF) and scatter factor/hepatocyte growth factor, and that induction of PV-1 is required for endothelial cells to undergo tubulogenesis and migration in vitro.


   Materials and Methods
Top
 Abstract
 Materials and Methods
Results
 Discussion
 References
 
Clinical specimens. All clinical specimens were provided by the Brain Tumor Bank of the University of Pittsburgh as approved under University of Pittsburgh Institutional Review Board protocol 021164. All samples were de-identified according to Health Insurance Portability and Accountability regulations by a university-approved honest broker before being provided for this research. This research falls under NIH guidelines for exempt research, not requiring informed consent, and was approved by the University of Pittsburgh Institutional Review Board for this exemption.

Chemicals and reagents. All chemicals and reagents were from Fisher Scientific (Pittsburgh, PA) unless otherwise specified. PCR primers were generated by Integrated DNA Technologies (Coralville, IA). TaqMan probes were generated by Synthegen (Houston, TX).

Cell lines and culture conditions. Human dermal microvascular endothelial cells (HMVEC; Cambrex BioScience, Walkersville, MD) were maintained in EBM/EGM-2MV growth medium with 1x growth supplements (human epidermal growth factor, hydrocortisone, GA-1000, fetal bovine serum, VEGF, human fibroblast growth factor-B, R3-insulin-like growth factor-I, and ascorbic acid; Cambrex BioScience). U87MG tumor cells (American Type Culture Collection, Manassas, VA) were maintained in MEM Eagle with Earle's salts and L-glutamine (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum, 0.1 mmol/L MEM nonessential amino acids, 0.1 mmol/L MEM sodium pyruvate, sodium bicarbonate, 1 unit/mL penicillin, and 1 µg/mL streptomycin (Invitrogen). U251MG tumor cells (a generous gift of Hideho Okada, University of Pittsburgh, Pittsburgh, PA) were maintained in DMEM with L-glutamine and sodium pyruvate (Invitrogen), supplemented with 10% fetal bovine serum, 1 unit/mL penicillin, and 1 µg/mL streptomycin. Cells were maintained in a 5% CO2 atmosphere at 37°C. HMVEC cells were used at passages 3 to 6 in all experiments. To condition the medium, 6 x 105 U87MG or U251MG tumor cells were grown in 5 mL medium in a T25 flask for 24 hours. Endothelial cells were exposed to the tumor-conditioned medium for a total of 24 hours before immunoblotting. For growth factor inductions, 70% confluent HMVEC cells were maintained in VEGF-free medium for 24 hours, then stimulated with either fresh VEGF-free medium or 50 nmol/L phorbol 12-myristate 13-acetate (Sigma, St. Louis, MO), 50 ng/mL VEGF (BD Biosciences, Bedford, MA), or 48 ng/mL scatter factor/hepatocyte growth factor (BD Biosciences) in VEGF-free medium for 48 hours before analysis.

RNA isolation and reverse transcription. Total RNA was isolated from 10 to 30 mg of snap frozen, homogenized tissue or 106 cultured cells using the Qiagen RNeasy Mini kit (Qiagen, Valencia, CA) according to the protocol of the manufacturer. cDNA was generated using the SuperScript First-Strand Synthesis System for reverse transcription-PCR (RT-PCR; Invitrogen) according to the protocol of the manufacturer. Briefly, 5 µg of total RNA was combined with oligo(dT)12-18 primers, heat denatured, and reverse transcribed with SuperScript II RT at 42°C for 50 minutes. A RT-minus control was done for each reverse transcription reaction. Products were phenol-chloroform extracted, diluted 1:10, and stored at –20°C. Equivalent volumes of RT-plus and RT-minus products were used as templates for amplification.

PCR amplification. PCR reactions were done using Platinum Taq Polymerase (Invitrogen). Standard reaction conditions were 1x PCR buffer, 200 mmol/L deoxynucleotide triphosphates (Invitrogen), 4% DMSO (Sigma), 1.5 mmol/L MgCl2, 1 µmol/L sense primer, 1 µmol/L antisense primer, 1.5 units Platinum Taq, and 2 µL cDNA in a total reaction volume of 25 µL. Reactions were overlaid with mineral oil and cycled in a PTC-200 DNA Engine thermalcycler (MJ Research, Waltham, MA). Primers for human EF1{alpha} or KDR were used for normalization of RT-PCR experiments.

TaqMan. Real-time quantitative PCR data and technical expertise were provided by personnel of the TaqMan core facility of the Genomics and Proteomics Core Laboratories of the University of Pittsburgh. In brief, total RNA isolated from tumor and normal specimens was reverse transcribed using the ABI High-Capacity cDNA Archive kit (Applied Biosystems, Foster City, CA) according to the protocol of the manufacturer. Quantitative PCR was done using the TaqMan Universal PCR Master Mix and ABI Prism 7900HT with sequence detection software v2.1 (Applied Biosystems). Each reaction was done in triplicate. PV-1 results for each sample were averaged and normalized to von Willebrand factor (vWF). The fold increase in PV-1 expression (tumor versus normal) was calculated. Statistical significance was determined using a nonparametric Wilcoxon rank sum test and JMP IN software. Primer and probe sequences are available upon request.

In situ hybridization. The protocol is adapted from St. Croix et al. (6). Digoxygenin-labeled antisense RNA probes covering the body of the target cDNA were generated by PCR amplification with incorporation of a T7 promoter (5'-CTAATACGACTCACTATAGGGAGA) into the antisense primer. In vitro transcription was done using the digoxygenin RNA Labeling kit (SP6/T7) and T7 RNA polymerase (Roche, Indianapolis, IN) according to the protocol of the manufacturer. Sections (8-10 µm) were cut onto CSS-100 silylated slides (CEL Associates, Pearland, TX). Briefly, sections were deparaffinized in xylenes and rehydrated in 100% ethanol, 70% ethanol, and 50% ethanol for 5 minutes each. Slides were rinsed in diethylpyrocarbonate-distilled water followed by incubation in 0.1 N HCl for 20 minutes and second rinse in diethylpyrocarbonate-distilled water. Slides were immersed in 10 µg/mL proteinase K (Invitrogen) diluted in 1x TBS (Quality Biologicals, Gaithersburg, MD) for 40 minutes at 37°C then rinsed in 1x TBST (1x TBS/0.05% Tween 20). Hybridizations were done by diluting digoxygenin-labeled RNA probe cocktail to 200 ng/mL in RNA hybridization solution (DakoCytomation, Carpinteria, CA) and heat denaturing for 3 minutes at 85°C. Sections were covered in HybriWell hybridization chambers (Molecular Probes, Eugene, OR) and the denatured hybridization solution was added to fill the chamber. Slides were hybridized overnight at 55°C. Slides were rinsed twice in 2x SSC for 5 minutes at 45°C and treated with RNase A/T1 (Ambion, Austin, TX) diluted 1:35 in 2x SSC at 37°C for 30 minutes. Slides were washed in 2x SSC/50% formamide DI (American Bioanalytical, Natick, MA) for 30 minutes at 55°C, followed by 0.08x SSC (DakoCytomation) for 30 minutes at 55°C. Slides were rinsed with 1x TBST and peroxidase blocking reagent (DakoCytomation) was added to each section for 30 minutes. Slides were rinsed in 1x TBST for 5 minutes. Normal rabbit IgG (DakoCytomation) was diluted 1:20 in buffer [100 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl, and 1% casein] to generate blocking buffer. Blocking buffer was added to sections for 30 minutes at room temperature. For detection, sections were incubated with rabbit horseradish peroxidase-antidigoxygenin primary (DakoCytomation) diluted 1:150 in blocking buffer for 45 minutes. Sections were washed in 1x TBST and one drop of biotinyl-tyramide (DakoCytomation) was added to each slide for 10 minutes. Slides were again washed in 1x TBST. Sections were incubated in rabbit horseradish peroxidase-antibiotin secondary (DakoCytomation) diluted 1:150 in blocking buffer for 20 minutes in the dark then rinsed in 1x TBST. One drop of biotinyl-tyramide was added to sections and slides were incubated for 5 minutes in the dark then rinsed in 1x TBST. Sections were incubated with rabbit alkaline phosphatase–antibiotin tertiary (DakoCytomation) diluted 1:75 in blocking buffer for 20 minutes in the dark and rinsed in 1x TBST. Signal was detected by staining with Fast Red (Sigma) followed by counterstaining with Probe/Hematoxylin counterstain (Biomeda, Foster City, CA). Slides were mounted with Crystal/Mount (Biomeda).

Antibodies. A rabbit polyclonal antibody was raised against the peptide "CPIDPASLEEFKRKILESQR" from the carboxy terminus of the human PV-1 protein (Quality Controlled Biochemicals, Hopkinton, MA). Anti-ß-actin goat polyclonal was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Goat anti-rabbit-horseradish peroxidase and rabbit anti-goat-horseradish peroxidase secondary antibodies were from Jackson ImmunoResearch Laboratories (West Grove, PA).

Western blotting. Cells were washed twice with Dulbecco's PBS, lysed in ice-cold radioimmunoprecipitation assay buffer (167 mmol/L NaCl, 0.5% NP40, 0.05% NaDOC, 0.1% SDS, and 50 mmol/L Tris-HCl) supplemented with protease inhibitors (Calbiochem, San Diego, CA). The protein lysate was homogenized through a 20 g needle and centrifuged for 5 minutes at 10,000 x g. Protein lysates were quantified using the Pierce Coomassie Protein Assay kit (Pierce, Rockford, IL) according to the protocol of the manufacturer. Proteins were electrophoresed through 4% to 20% Tris-glycine polyacrylamide Novex gels (Invitrogen) under denaturing conditions and were transferred onto Immobilon polyvinylidene difluoride membranes (Invitrogen). Membranes were blocked overnight at 4°C in 10% bovine serum albumin/TBS with 0.05% Tween 20 (TBST). Proteins were then exposed to an anti-human-PV1 rabbit polyclonal diluted 1:100 in 5% BSA/TBST at 4°C overnight. Membranes were washed with 5% BSA/TBST for 3 x 10 minutes. The proteins were exposed to a goat anti-rabbit-horseradish peroxidase secondary diluted 1:1,000 in 5% BSA/TBST for a total of 1 hour. Membranes were washed 3 x 10 minutes with 5% BSA/TBST and then washed an additional 6 x 10 minutes with TBST. Antibody binding was visualized using ECL Western Blotting Detection Reagents according to directions of the manufacturer (Amersham Biosciences, Little Chalfont Buckinghamshire, United Kingdom). Membranes were stripped with 60 mmol/L Tris-HCl, 2% SDS, and 100 mmol/L ß-mercaptoethanol and reprobed as above using 1:1,000 anti-ß-actin goat polyclonal antibody and 1:5,000 rabbit anti-goat secondary antibody.

Small interfering RNA. Small interfering RNA (siRNA) duplexes were generated against PV-1 (Integrated DNA Technologies). The targeted PV1-KD sequences were GAAGCUCAACUUCACCACCTT (sense) and GGUGGUGAAGUUGAGCUCCTT (antisense). The control PV1-scr sequences were CAGCACAUGCAUGUACACGTT (sense) and CGUGUACAUGCAUGUGCUGTT (antisense). Duplexes were transfected using the Targefect siRNA Transfection kit, according to the protocol of the manufacturer (Targeting Systems, Santee, CA). Briefly, cells were seeded at 12 x 104 cells/T25 flask and brought to 60% to 70% confluence for the day of transfection. Transfection complexes were formed in Opti-MEM I medium (Invitrogen) containing transfection reagents and 75 pmol of siRNA duplex at 37°C for 25 minutes. Cells were washed twice with Opti-MEM I medium (Invitrogen) and exposed to transfection complex medium for 2 hours at 37°C in a 5% CO2 atmosphere. The cells were replenished with the appropriate growth medium and incubated overnight. Fresh growth medium was supplied the next morning and cells were maintained for 96 hours before Matrigel induction.

Matrigel assays. Matrigel (BD Biosciences) was thawed on ice overnight at 4°C and 300 µL was spread per well of a 24-well plate. The plates were polymerized for 30 minutes at 37°C. Cells were counted using trypan blue exclusion to ensure viability and plated in appropriate growth medium at a density of 6 x 104 viable cells per well. Cells were grown for 24 hours on Matrigel with tubulogenesis observed at 6 hours postseeding. In each condition, five randomly selected fields of view were photographed in each well for assessment of average tubule number and average tubule length. Images were captured using a Zeiss AxioCAM digital camera attached to a microscope at x10 magnification. Image analysis of tubule length and number was carried out using Zeiss KS 300 3.0 software. Statistical significance was determined using the Wilcoxon/Kruskal-Wallis rank sum test and JMP IN software.

Cell migration assays. HMVEC cells were grown in T25 flasks to 70% confluence and transfected with 70 pmol of PV1-KD or PV1-scr RNA interference (RNAi) duplex in DMEM medium for 2 hours. Twenty-four hours following transfection, cells were harvested with trypsin-EDTA (Invitrogen), suspended in growth medium supplemented with 10% fetal bovine serum, and allowed to recover for 2 hours under normal growth conditions. Twenty-four-well, 8 µm transwell inserts (Corning, Inc., Corning, NY) were coated with 0.01% porcine gelatin (Sigma) overnight at 4°C. Conditioned medium (400 µL) from U87MG/VEGF overexpressing cells (generous gift of Dr. Shih Yuan Cheng, Department of Pathology, University of Pittsburgh, Pittsburgh, PA), conditioned medium from U87MG cells, control U87MG medium supplemented with 50 ng/mL VEGF, or control U87MG medium alone were added to the lower chambers and 1 x 105 HMVEC cells in 70 µL were added to the top of the inserts. Cells were incubated for 6 hours under normal growth conditions and stained with CellTracker Green (Invitrogen). Unmigrated cells were removed from the top of the filter with a Q-tip. Transmigration of cells were determined by counting total cells in five high power fields per well. Experiments were done in duplicate. Statistical significance was determined using the Wilcoxon/Kruskal-Wallis rank sum test and JMP IN software.


   Results
Top
 Abstract
 Materials and Methods
 Results
Discussion
 References
 
Expression of PV-1. Previous SAGE results had shown selective up-regulation of PV-1 in the endothelial cells of human glioblastoma compared with nonneoplastic endothelium (Table 1; ref. 3). To verify that PV-1 was induced in human brain tumors, we performed RT-PCR using PV-1–specific primers on mRNA isolated from frozen clinical specimens of two nonneoplastic brains and five high-grade gliomas. Additionally, we wondered whether PV-1 induction was glioma specific, or possibly a more general marker of CNS neoplasia, so five specimens of metastatic adenocarcinoma to the brain were analyzed as well. The RT-PCR yielded PV-1 mRNA in the glioma and metastatic mRNA samples but no PV-1 mRNA expression in the normal brain samples. Representative data are shown in Fig. 1A.



View larger version (40K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 1. A, semiquantitative RT-PCR for PV-1 and EF-1{alpha} on RNA isolated from clinical specimens of epileptic cortex (N), metastatic adenocarcinoma (Met), and glioblastoma multiforme (GBM). B, TaqMan real-time PCR data quantifying fold increase of PV-1 in metastatic adenocarcinoma and glioblastoma multiforme. Results are normalized to vWF as a control endothelial marker gene. Bars, SE. *P = 0.01, Wilcoxon rank sum test.

 
We next did quantitative real-time PCR on our samples to better quantify the up-regulation of PV-1 in the tumors. Again, the results confirmed an ~4-fold induction of PV-1 mRNA in all tumors when normalized to vWF, a blood vessel marker (P = 0.01, Wilcoxon rank sum test; Fig. 1B). These results showed that the increase in PV-1 signal was not simply due to an increased number of blood vessels. Although we did not detect PV-1 expression in our normal samples with standard RT-PCR, the more sensitive TaqMan protocol was able to measure low levels of PV-1 expression.

Localization of PV-1. To confirm that PV-1 mRNA was not only up-regulated in human brain tumors but that its expression localized to the microvasculature, we did in situ hybridization on paraffin-embedded normal brain and brain tumor samples using riboprobes directed against PV-1 and the endothelial-specific markers KDR (VEGFR2) and vWF. The microvasculature of both the glial tumors and the metastatic adenocarcinomas stained brightly with both PV-1 and KDR. There was no expression of PV-1 detected in any location other than endothelium (Fig. 2). In contrast, in situ hybridization to normal brain showed comparatively nominal expression of PV-1 within the cortex (Fig. 2A). During examination of the nonneoplastic brain specimens, we determined that vWF showed more complete staining of the cerebrovasculature than KDR (data not shown), so we elected to use vWF as the endothelial control gene for the remainder of the in situ studies. Control staining for vWF and KDR was identical for the high-grade tumors.



View larger version (121K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 2. In situ hybridization for PV-1 in clinical specimens of CNS pathology. In each pair, superior image represents staining for PV-1 and inferior image represents control staining for endothelial-specific antigen. KDR was used as the control for images (A-C) and vWF was used as the control for (D-H). Pathologic diagnoses: epileptic cortex (A), glioblastoma multiforme (B), metastatic adenocarcinoma (C), anaplastic astrocytoma (D), oligodendroglioma (E), cerebral primitive neuroectodermal tumor (F), medulloblastoma (G), and acute ischemia (H).

 
We next expanded our examination of the PV-1 expression pattern to a panel of high- and low-grade brain tumors as well as several additional pathologic brain specimens with an abnormal vasculature. PV-1 staining was uniformly detected in the glioblastomas and metastatic tumors, but was also demonstrable in a lesser percentage of the lower-grade tumors (Fig. 2). In all cases, PV-1 expression was localized to endothelial cells. Interestingly, PV-1 expression was evident in tumors arising from the choroid plexus and ependyma, two sites where PV-1 expression remains present in the adult CNS. PV-1 was also up-regulated in brain tissue surgically resected from patients undergoing decompressive operations for acute ischemia/stroke, indicating a possible role for hypoxia in PV-1 gene induction. Table 2 summarizes these results.


View this table:
[in this window]
[in a new window]

  		Table 2. Expression of PV-1 in clinical samples

 
Induction of PV-1 expression by human brain tumor cell lines. PV-1 is expressed at very low levels in primary human endothelial cells but can be induced by phorbol 12-myristate 13-acetate treatment (7). To further elucidate the relationship between brain tumors and PV-1, we sought to determine whether PV-1 expression could be up-regulated by exposure of endothelial cells to medium conditioned by human glioblastoma cell lines. HMVEC cells were cultured in the presence of U87MG or U251MG conditioned medium for 24 hours and then harvested for protein analysis. Western blotting showed a clear induction of PV-1 in HMVEC cells exposed to both the U87MG and U251MG medium (Fig. 3A). Furthermore, PV-1 was induced by the individual addition of proangiogenic growth factors VEGF or scatter factor/hepatocyte growth factor, which are secreted by gliomas (Fig. 3B). Phorbol 12-myristate 13-acetate induction of PV-1 served as a positive control.



View larger version (66K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 3. Western blot for PV-1. A, HMVEC cells grown on plastic express low levels of PV-1 protein (control) but are induced to express increased levels when cultured in medium conditioned by the malignant glioma cell lines U87MG or U251MG. B, HMVEC cells grown in the absence of VEGF express little to no PV-1. PV-1 expression is induced by the addition of scatter factor/hepatocyte growth factor (SF/HGF) or VEGF to the medium. Phorbol 12-myristate 13-acetate (PMA) induction of PV-1 serves as a positive control. C, PV-1 induction by U87MG conditioned medium is blocked in HMVEC cells transfected with RNAi complexes targeted against PV-1 (PV1-KD). Induction in cells transfected with a scrambled RNAi complex (PV1-Scr) is unchanged compared with control HMVEC cells.

 
Small interfering RNA knockdown of PV-1 expression. We next generated a previously validated siRNA duplex targeted against the PV-1 mRNA molecule (PV1-KD) and a scrambled control duplex (PV1-Scr; ref. 7). Fluorescently tagged versions of the siRNA duplexes were used to optimize transfection conditions into endothelial cells (data not shown). Transfection of the targeted PV1-KD knockdown duplex into HMVEC cells prevented the up-regulation of PV-1 protein upon treatment with U87MG conditioned medium (Fig. 3C). In contrast, the scrambled, untargeted duplex, PV1-Scr, had nominal effect upon PV-1 induction by U87MG conditioned medium. PV1-KD also blocked induction by U251MG (data not shown).

Knockdown of PV-1 inhibits tubulogenesis and migration in vitro. To address the role of PV-1 in the angiogenic response, we examined whether PV-1 expression was required for endothelial cells to undergo Matrigel-induced tubulogenesis. HMVEC cells were transfected with PV1-KD or PV1-Scr duplexes or no siRNA. The cells were allowed to recover from the transfection, checked for viability, and then plated onto Matrigel. Within 6 hours, the untreated cells showed a robust tubulogenic response (Fig. 4A). The cells treated with the nonspecific PV1-Scr duplex also underwent tubulogenesis (Fig. 4B). However, the cells treated with the PV1-KD duplex failed to undergo tubulogenesis in Matrigel, suggesting that PV-1 expression is critical for the tubulogenic response in vitro. PV-1 knockdown in HMVEC cells significantly inhibited both tubule number (P = 0.002, Wilcoxon rank sum) and length (P = 0.009, Wilcoxon rank sum) compared with control HMVEC cells and HMVEC cells transfected with a scrambled RNAi duplex (Fig. 4C and D).



View larger version (78K):
[in this window]
[in a new window]
[Download PPT slide]
  		Fig. 4. Knockdown of PV-1 expression by RNAi blocks angiogenic responses of HMVEC cells. A, HMVEC cells grown on Matrigel form a microtubule network. This network is unaffected by transfection with a scrambled RNAi complex (PV1-Scr; B) but is effectively blocked in HMVEC cells transfected with RNAi targeted against PV-1 (PV1-KD; C). D, tubule number and length are both significantly inhibited by PV-1 knockdown compared with control HMVEC cells or cells transfected with scrambled RNAi duplexes, n = 5 low power fields per condition, *P = 0.002, tubule number, and P = 0.009, tubule length, Wilcoxon rank sum test. E, migration of HMVEC cells through gelatin-coated transwell inserts in response to U87MG/VEGF conditioned medium (U87/VEGF), U87MG conditioned medium (U87), growth medium supplemented with VEGF, or untreated medium (NO TX) is specifically inhibited by transfection with the PV1-KD RNAi duplex compared with the PV1-Scr–transfected cells or control cells (P = 0.002 versus each, Wilcoxon rank sum). The PV1-Scr control duplex does not inhibit migration compared with the control cells in response to any chemoattractant.

 
To further substantiate the role of PV-1 in angiogenesis, we used a modified transwell assay to assess migration of HMVEC cells in response to proangiogenic chemoattractants. HMVEC cells were added to top wells of gelatin-coated transwell inserts and exposed to medium conditioned by U87MG/VEGF-overexpressing cells, parental U87MG cells, VEGF-supplemented growth medium, or growth medium alone. Untransfected, control HMVEC cells or those transfected with the nonspecific PV1-Scr RNAi duplex showed increased migration in response to chemoattractants (Fig. 4E). However, transfection with the PV1-KD RNAi duplex significantly inhibited cell migration in response to every condition compared with control HMVEC cells and HMVEC cells transfected with scrambled RNAi duplexes (P = 0.002 versus control and scrambled RNAi, Wilcoxon rank sum test). These data provide further evidence that PV-1 is required during the angiogenic response of endothelial cells in vitro.


   Discussion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
References
 
Caveolae are specialized invaginations of the plasma membrane (50-100 nm in diameter) also known as "lipid rafts" formed as a result of localized accumulation of cholesterol, glycosphingolipids, and structural proteins, reviewed in (ref. 8). In addition to a classic role in transcellular transport, they are also recognized to play roles in receptor-mediated uptake, receptor tyrosine kinase signal transduction, apoptosis, and lipid metabolism. The principal structural component of caveolae, caveolin-1, is widely expressed in brain endothelial microvascular cells (8–11). Unlike in fenestrated endothelia where caveolae are involved in transport, microvesicular transport is absent in the mature blood-brain barrier and caveolin-1 seems to play a role in maintaining blood-brain barrier integrity through associations with P-glycoprotein and occludin (12–15).

Under conditions promoting angiogenesis, the role of caveolae seems to change. VEGF stimulation of endothelial cells in vitro leads to increased caveolae and fenestrae formation (16). These changes occur without an increase in cellular levels of caveolin-1. Caveolae also become critical for cytokine signal transduction, such as those from the chemokine monocyte chemoattractant protein-1 (17), which are responsible for inflammatory events at the blood-brain barrier and for proper cellular trafficking of proteins associated with extracellular matrix turnover such as matrix metalloproteinase-9 and fibronectin (18, 19). In total, blocking caveolin-1 expression with antisense RNA effectively blocks angiogenesis using in vitro models (18, 19) and indicates that targeting caveolae may be an attractive target to block tumor-induced angiogenesis in patients.

Unfortunately, the widespread expression of caveolin-1 in the mature blood-brain barrier and its apparent role in both maintaining the intact blood-brain barrier as well as promoting angiogenesis makes it a poor target for possible therapies. In this paper, we describe a valuable alternative therapeutic target, PV-1.

PV-1 was initially described in association with microvesicular transport in immature rodent brain endothelium as a cell surface protein whose expression is suppressed in the mature CNS (4). Further microstructural studies have indicated that multiple homodimers of PV-1 localize to the stomatal diaphragms of caveolae in fenestrated endothelium (20). Induction of PV-1 in primary endothelial cells triggered the de novo formation of stomatal diaphragms of caveolae and transendothelial channels, whereas silencing of PV-1 prevented the formation of the diaphragms of caveolae as well as fenestrae and transendothelial channels (7). Whereas PV-1 is normally expressed in adult lung, kidney, and endocrine tissues, its expression is lost in the CNS in association with formation of an intact blood-brain barrier (5). Recent genomic data suggest that the cDNA encoding the murine endothelial marker, MECA-32, is identical to PV-1, and, indeed, MECA-32 expression is suppressed during differentiation of the blood-brain barrier in mice (21).

We have previously described PV-1 as a glioma-specific endothelial marker based on its differential expression in SAGE studies (3). PV-1 was dramatically up-regulated in endothelial cells derived from patients with glioblastoma multiforme in our SAGE studies but minimally expressed in patients with temporal lobe epilepsy (nonneoplastic controls). We have now confirmed that PV-1 is dramatically and specifically up-regulated in the endothelial cells of gliomas and have expanded our analysis to a panel of brain tumors and other pathologic disorders. There is a clear association between PV-1 expression and abnormal cerebrovascular integrity. PV-1 expression is robust in malignant brain tumors, including both primary glioblastoma multiforme and secondary metastatic carcinoma. Interestingly, PV-1 expression was also robust in CNS tissue resected for decompression following an acute ischemic infarct. Surgery for this indication is done immediately following the event, indicating PV-1 up-regulation occurs rapidly in response to external stimuli, rather than occurring as a late effect of vascular remodeling. Generally, PV-1 expression was highest in lesions with a disrupted blood-brain barrier that would be expected to enhance on imaging studies, such as magnetic resonance imaging or computed tomography. There were exceptions to this, however, as a large medulloblastoma with bright control endothelial staining for vWF did not show significant staining for PV-1 by in situ hybridization. It is always possible that although up-regulation did occur in this setting, it fell outside the ability of our in situ probes to detect.

We confirmed a direct link between CNS malignancy and endothelial PV-1 expression by demonstrating that cultured medium from human glioma cell lines U87MG and U251MG was capable of boosting PV-1 expression in HMVEC cells grown in vitro. Previous experiments by our laboratory and others indicate that HMVEC cells grown on plastic express basal levels of PV-1 that can also be increased by growth on Matrigel (3) or exposure to phorbol 12-myristate 13-acetate (7). Additionally, we showed that VEGF and scatter factor/hepatocyte growth factor induced PV-1 expression in vitro, suggesting that secretion of these growth factors by gliomas might increase protein expression in vivo. Interestingly, we have observed that fetal astrocytes can also induce PV-1 expression in HMVEC cells (data not shown), potentially reflecting the immature microvascular phenotype as astrocytes are known to influence cerebrovascular integrity and can release agents, such as ET-1, glutamate, inetrleukin-6, tumor necrosis factor-{alpha}, and macrophage inflammatory protein-2, that increase blood-brain barrier permeability (reviewed in ref. 22). Together, these data indicate that PV-1 expression is tightly regulated in response to external stimuli, including those provided by replicating glioma cell lines.

We validated PV-1 as a therapeutic target for blocking angiogenesis. Matrigel-induced tubulogenesis was inhibited by RNAi knockdown of PV-1 expression. Although results are presented using HMVEC cells, similar phenomena were also seen with human umbilical vascular endothelial and primary brain endothelial cell lines (data not shown). PV-1 knockdown did not affect cellular viability as measured by trypan blue exclusion, indicating the effects of PV-1 knockdown are selective for tubulogenesis rather than merely representing a fatal event for the endothelial cell population.

In summary, PV-1 expression is selective in the CNS and limited to disease states associated with blood-brain barrier disruption, such as malignancy and stroke. Blocking PV-1 through RNAi also blocks angiogenesis in an in vitro model system. Selective targeting of the CNS to allow sparing of the systemic vasculature through polymer-mediated, site-specific delivery or convection-enhanced therapy is a rapidly expanding field (23). PV-1 may, therefore, be a suitable antiangiogenic target for brain tumor therapy and cerebral edema and may represent a selective molecule through which caveolae-mediated biological phenomena can be controlled.


   Footnotes
 
Grant support: National Institute of Neurologic Disease and Stroke grant K08 NS046461, Elsa U. Pardee Foundation, Neurosurgery Research and Education Foundation, and Copeland Fund of the Pittsburgh Foundation (K.A. Walter).

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.

Received 5/18/05; revised 7/15/05; accepted 8/ 9/05.


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 References
 

  1. Shapiro WR, Green SB, Burger PC. Randomized trial of three chemotherapy regimens and two radiotherapy regimens in postoperative treatment of malignant gliomas. Brain Tumor Cooperative Group Trial 8001. J Neurosurg 1989;71:1–9.[Medline]
  2. Walker MD, Green SB, Byar DP. Randomized comparisons of radiotherapy and nitrosoureas for treatment of malignant glioma after surgery. N Engl J Med 1980;303:1323–9.[Abstract]
  3. Madden SL, Cook BP, Nacht M, et al. Vascular gene expression in nonneoplastic and malignant brain. Am J Pathol 2004;165:601–8.[Abstract/Free Full Text]
  4. Hallman R, Mayer DN, Berg EL, Broermann R, Butcher EC. Novel mouse endothelial cell surface marker is suppressed during differentiation of the blood brain barrier. Dev Dyn 1995;202:325–32.[Medline]
  5. Hnasko R, McFarland M, Ben-Jonathan N. Distribution and characterization of plasmalemma vesicle protein-1 in rat endocrine glands. J Endocrinol 2002;175:649–61.[Abstract]
  6. St. Croix B, Rago C, Velculescu V, et al. Genes expressed in human tumor endothelium. Science 2000;289:1197–202.[Abstract/Free Full Text]
  7. Stan RV, Tkachenko E, Niesman IR. PV1 is a key structural component for the formation of the stomatal and fenestral diaphragms. Mol Biol Cell 2004;15:3615–30.[Abstract/Free Full Text]
  8. Galbiati F, Razani B, Lisanti MP. Emerging themes in lipid rafts and caveolae. Cell 2001;106:403–11.[CrossRef][Medline]
  9. Virgintino D, Robertson D, Errede M, et al. Expression of caveolin-1 in human brain microvessels. Neuroscience 2002;115:145–52.[CrossRef][Medline]
  10. Ikezu T, Ueda H, Trapp BD, et al. Affinity-purification and characterization from the brain: differential expression of caveolin-1, -2, and -3 in brain endothelial and astroglial cell types. Brain Res 1998;804:177–92.[CrossRef][Medline]
  11. Cameron PL, Ruffin JW, Bollag R, Rasmussen H, Cameron RS. Identification of caveolin and caveolin-related proteins in the brain. J Neurosci 1997;17:9520–35.[Abstract/Free Full Text]
  12. Liebner S, Kniesel U, Kalbacher H, Wolburg H. Correlation of tight junction morphology with the expression of tight junction proteins in blood-brain barrier endothelial cells. Eur J Cell Biol 2000;10:707–17.
  13. Kniesel U, Wolburg H. Tight junctions of the blood-brain barrier. Cell Mol Neurobiol 2000;20:57–76.[CrossRef][Medline]
  14. Jodoin J, Demeule M, Fenart L, et al. P-glycoprotein in blood-brain barrier endothelial cells: interaction and oligomerization with caveolins. J Neurochem 2003;87:1010–23.[CrossRef][Medline]
  15. Virgintino D, Robertson D, Errede M, et al. Expression of P-glycoprotein in human cerebral cortex microvessels. J Histochem Cytochem 2002;50:1671–6.[Abstract/Free Full Text]
  16. Esser S, Wolburg K, Wolburg H, et al. Vascular endothelial growth factor induces endothelial fenestrations in vitro. J Cell Biol 1998;140:947–59.[Abstract/Free Full Text]
  17. Ge S, Pachter JS. Caveolin-1 knockdown by small interfering RNA suppresses responses to the chemokine monocyte chemoattractant protein-1 by human astrocytes. J Biol Chem 2004;279:6688–95.[Abstract/Free Full Text]
  18. Sottile J, Chandler J. Fibronectin matrix turnover occurs through a caveolin-1 dependent process. Mol Biol Cell 2005;16:757–68.[Abstract/Free Full Text]
  19. Phillips PG, Birnby LM. Nitric oxide modulates caveolin-1 and matrix metalloproteinase-9 expression and distribution at the endothelial cell/tumor cell interface. Am J Physiol Lung Cell Mol Physiol 2004;286:L1055–65.[Abstract/Free Full Text]
  20. Stan RV. Multiple PV1 dimers reside in the same stomatal or fenestral diaphragm. Am J Physiol Heart Circ Physiol 2004;286:H1347–53.[Abstract/Free Full Text]
  21. Hallmann R, Mayer DN, Berg EL, Broermann R, Butcher EC. Novel mouse endothelial cell surface marker is suppressed during differentiation of the blood brain barrier. Dev Dyn 1995;202:325–32.
  22. Abbott NJ. Astrocyte-endothelial interactions and blood brain barrier permeability. J Anat 2002;200:629–38.[CrossRef][Medline]
  23. Walter KA, Tamargo RJ, Olivi A, Burger PC, Brem H. Intratumoral chemotherapy. Neurosurgery 1995;37:1128–45.[Medline]



This article has been cited by other articles:


Home page
 IOVSHome page
N. F. Schafer, U. F. O. Luhmann, S. Feil, and W. Berger
Differential Gene Expression in Ndph-Knockout Mice in Retinal Development
Invest. Ophthalmol. Vis. Sci., February 1, 2009; 50(2): 906 - 916.
[Abstract] [Full Text] [PDF]

Home page
 JCBHome page
S. Liebner, M. Corada, T. Bangsow, J. Babbage, A. Taddei, C. J. Czupalla, M. Reis, A. Felici, H. Wolburg, M. Fruttiger, et al.
Wnt/{beta}-catenin signaling controls development of the blood-brain barrier
J. Cell Biol., November 3, 2008; 183(3): 409 - 417.
[Abstract] [Full Text] [PDF]

Home page
 J. Am. Soc. Nephrol.Home page
K. Ichimura, R. V. Stan, H. Kurihara, and T. Sakai
Glomerular Endothelial Cells Form Diaphragms during Development and Pathologic Conditions
J. Am. Soc. Nephrol., August 1, 2008; 19(8): 1463 - 1471.
[Abstract] [Full Text] [PDF]

Home page
 Cardiovasc ResHome page
N. M.S. van den Akker, V. Caolo, L. J. Wisse, P. P.W.M. Peters, R. E. Poelmann, P. Carmeliet, D. G.M. Molin, and A. C. Gittenberger-de Groot
Developmental coronary maturation is disturbed by aberrant cardiac vascular endothelial growth factor expression and Notch signalling
Cardiovasc Res, May 1, 2008; 78(2): 366 - 375.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
L. A. Sanders (Strickland) and H. Koeppen
Comments on Plasmalemmal Vesicle Associated Protein-1 as a Novel Marker Implicated in Brain Tumor Angiogenesis
Clin. Cancer Res., April 15, 2006; 12(8): 2649 - 2649.
[Full Text] [PDF]

Home page
 Clin. Cancer Res.Home page
E. B. Carson-Walter and K. A. Walter
Comments on plasmalemmal vesicle associated protein-1 as a novel marker implicated in brain tumor angiogenesis.
Clin. Cancer Res., April 15, 2006; 12(8): 2649 - 2649.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carson-Walter, E. B.
Right arrow Articles by Walter, K. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carson-Walter, E. B.
Right arrow Articles by Walter, K. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online