Cancer Research CTRC-AACR San Antonio Breast Cancer Symposium AACR Conference on Molecular Diagnostics - 2008
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
Cancer Research 67, 1757-1768, February 15, 2007. doi: 10.1158/0008-5472.CAN-06-3700
© 2007 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
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 Lu, C.
Right arrow Articles by Sood, A. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lu, C.
Right arrow Articles by Sood, A. K.

Experimental Therapeutics, Molecular Targets, and Chemical Biology

Gene Alterations Identified by Expression Profiling in Tumor-Associated Endothelial Cells from Invasive Ovarian Carcinoma

Chunhua Lu1, Tomas Bonome3, Yang Li1, Aparna A. Kamat1, Liz Y. Han1, Rosemarie Schmandt1, Robert L. Coleman1, David M. Gershenson1, Robert B. Jaffe4, Michael J. Birrer3 and Anil K. Sood1,2

Departments of 1 Gynecologic Oncology and 2 Cancer Biology, University of Texas M. D. Anderson Cancer Center, Houston, Texas; 3 Cell and Cancer Biology Branch, National Cancer Institute, Bethesda, Maryland; and 4 Center for Reproductive Sciences, University of California, San Francisco, San Francisco, California

Requests for reprints: Anil K. Sood, Department of Gynecologic Oncology, University of Texas M. D. Anderson Cancer Center, Unit 1362, 1155 Herman Pressler, Houston, TX 77030. Phone: 713-745-5266; Fax: 713-792-7586; E-mail: asood{at}mdanderson.org.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Therapeutic strategies based on antiangiogenic approaches are beginning to show great promise in clinical studies. However, full realization of these approaches requires identification of key differences in gene expression between endothelial cells from tumors versus their normal counterparts. Here, we examined gene expression differences in purified endothelial cells from 10 invasive epithelial ovarian cancers and 5 normal ovaries using Affymetrix U133 Plus 2.0 microarrays. More than 400 differentially expressed genes were identified in tumor-associated endothelial cells. We selected and validated 23 genes that were overexpressed by 3.6- to 168-fold using real-time reverse transcription-PCR and/or immunohistochemistry. Among these, the polycomb group protein enhancer of Zeste homologue 2 (EZH2), the Notch ligand Jagged1, and PTK2 were elevated 3- to 4.3-fold in tumor-associated endothelial cells. Silencing these genes individually with small interfering RNA blocked endothelial cell migration and tube formation in vitro. The present study shows that tumor and normal endothelium differ at the molecular level, which may have significant implications for the development of antiangiogenic therapies. [Cancer Res 2007;67(4):1757–68]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Despite improvements in surgery and chemotherapy, mortality rates in women with advanced ovarian carcinoma have remained largely unchanged (1). Therefore, novel therapeutic strategies are needed. It is now well known that growth of tumors, both at the primary and metastatic sites, requires a blood supply for expansion beyond 1 to 2 mm (2). Targeting tumor angiogenesis by inhibiting endothelial cells that support tumor growth is particularly promising because of their presumed genetic stability. The recent success of a humanized monoclonal antibody (bevacizumab) against vascular endothelial growth factor in prolonging the lives of patients with advanced colon and breast carcinoma (3, 4) shows the promise of such approaches. However, the full spectrum of differences in the tumor vasculature compared with its normal counterpart is not known. Identification of additional targets on tumor endothelium may allow opportunities for developing new therapeutic approaches to inhibit angiogenesis in a tumor-specific manner.

Higher levels of proangiogenic cytokines and angiogenesis are associated with an increased risk of metastasis and poor prognosis in ovarian cancer (5, 6). To date, a small number of breast, colon, and brain cancers have been analyzed for gene expression changes in the tumor vasculature using serial analysis of gene expression (7–9). These studies showed the ability to define both tumor-specific endothelial genes and normal endothelial genes. Whereas selected genes in ovarian cancer vasculature have been characterized, there is little information about global gene expression alterations in ovarian cancer endothelium. This lack of data prompted us to carry out expression profiling on purified endothelial cells from invasive epithelial ovarian cancers and normal ovaries.

In recent years, whole genome expression profiling of cancer using methods such as microarray and serial analysis of gene expression has advanced our understanding of the molecular pathways involved in cancer onset and progression. However, global analysis of gene expression in specific cell populations within the tumor microenvironment is challenging and bulk tissue expression profiling may, in fact, mask gene changes in different cell types. We have recently used laser capture microdissection to isolate epithelial cells from ovarian cancers for microarray analyses (10), which elucidated changes in gene expression specific to the epithelial tumor cells. Profiling expression changes that occur in the tumor stroma, including the tumor endothelial cells, will likely provide insights into the mechanisms underlying tumor vascular growth, reveal additional targets for antiangiogenic therapies, and potentially offer new biomarkers for diagnosis and surveillance. However, the endothelium is enmeshed in a tissue complex consisting of vessel wall components, stromal cells, and epithelial cells. Only a small fraction of the cells within these tissues are endothelial. Moreover, gene analysis of specific cell types extracted from chemical reagent–fixed frozen tissue may not be accurate as the gene profile may be altered during the fixation process. In the present study, we immunopurified endothelial cells from human normal ovarian tissues and invasive epithelial cancers and investigated the gene expression profile using microarrays. Selected genes were validated to test the reliability of the microarray analysis. The gene expression profiles derived in the current study define unique alterations in vascular gene expression in epithelial ovarian carcinoma.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Sample preparation. Fresh tissue samples (5 normal ovaries and 10 epithelial high-grade, stage III or IV invasive serous ovarian cancers) were obtained from patients undergoing primary surgical exploration at the M. D. Anderson Cancer Center after approval from the Institutional Review Board. The minced tissue was digested with collagenase A, elastase, and DNase 1 at 37°C for 90 min to yield a single-cell suspension. A number of negative selections followed, including removal of platelets and RBCs by Percoll separation; removal of epithelial cells using M450 beads, which are prebound to BerEP4 antibody; and removal of leukocytes using anti-CD14, anti-CD45, and anti-CD64 beads (Dynal Biotech, Brown Deer, WI). Positive selection was done with P1H12 (CD146) immunobeads (P1H12 antibody was from Chemicon, Temecula, CA), and the beads linked to secondary antibody were from Dynal Biotech. Immunostaining was then done using von Willebrand factor and 4',6-diamidino-2-phenylindole nuclear staining to confirm the purification of endothelial cells.

Total RNA amplification for Affymetrix GeneChip hybridization and image acquisition. To successfully generate sufficient labeled cRNA for microarray analysis from 25 ng of total RNA, two rounds of amplification were necessary. For the first-round synthesis of double-stranded cDNA, 25 ng of total RNA were reverse transcribed using the Two-Cycle cDNA Synthesis Kit (Affymetrix, Santa Clara, CA) and oligo-dT24-T7 (5'-GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG-3') primer according to the manufacturer's instructions followed by amplification with the MEGA script T7 Kit (Ambion, Inc., Austin, TX). After cleanup of the cRNA with a GeneChip Sample Cleanup Module IVT column (Affymetrix), second-round double-stranded cDNA was amplified using the IVT Labeling Kit (Affymetrix). A 15.0-µg aliquot of labeled product was fragmented by heat and ion-mediated hydrolysis at 94°C for 35 min in 24 µL of H2O and 6 µL of 5x Fragmentation Buffer (Affymetrix). The fragmented cRNA was hybridized for 16 h at 45°C in a Hybridization Oven 640 to a U133 plus 2.0 oligonucleotide array (Affymetrix). Washing and staining of the arrays with phycoerythrin-conjugated streptavidin (Molecular Probes, Eugene, OR) was completed in a Fluidics Station 450 (Affymetrix). The arrays were then scanned using a confocal laser GeneChip Scanner 3000 and GeneChip Operating Software (Affymetrix).

Data normalization and filtering. Global normalization at a target value of 500 was applied to all 15 arrays under consideration using GeneChip Operating Software (Affymetrix). Normalized data were uploaded into the National Cancer Institute Microarray Analysis Database for quality control screening and collation before downstream analyses.5 Biometric Research Branch (BRB) ArrayTools version 3.2.2 software, developed by Drs. Richard Simon and Amy Peng Lam of the Biometrics Research Branch of the National Cancer Institute, was used to filter and complete the statistical analysis of the array data. BRB-ArrayTools is a multifunctional Excel add-in that contains utilities for processing and analyzing microarray data using the R version 2.0.1 environment (R Development Core Team, 2004). Of the 47,000 transcripts represented on the array, hybridization control probe sets and probe sets scored as absent at {alpha}1 = 0.05 or marginal at {alpha}2 = 0.065 were excluded. In addition, only those transcripts present in >50% of the arrays and displaying a variance in the top 50th percentile were evaluated.

Class comparison analysis. Differentially expressed genes were identified for tumor and normal endothelial cell specimens using a multivariate permutation test in BRB-ArrayTools (11). A total of 2,000 permutations were completed to identify the list of probe sets with a false discovery rate of <10% at a confidence of 95%. Differential expression was considered significant at P < 0.001. A random-variance t test was selected to permit the sharing of information among probe sets within class variation without assuming that all of the probe sets possess the same variance (12). A global assessment of whether expression profiles were different between classes was also done. During each permutation, the class labels were reassigned randomly and the P value for each probe set was recalculated. The proportion of permutations yielding at least as many significant genes as the actual data set at P < 0.001 was reported as the significance level of the global test.

Pathway analysis. Differentially regulated genes identified in a series of 48 late-stage (III and IV), high-grade (3) microdissected papillary serous ovarian carcinomas, as compared with 10 normal ovarian surface epithelial brushings (10), were categorized by cellular component according to the Gene Ontology ontological hierarchy. Epithelial genes associated with the cell membrane, extracellular matrix, and extracellular region were used as central nodes to identify signaling pathways modulated in tumor-associated endothelial cell isolates. This was accomplished using PathwayAssist version 3.0 software (Iobion Informatics LLC, La Jolla, CA). This software package contains more than 500,000 documented protein interactions acquired from MedLine using the natural language processing algorithm MEDSCAN. The proprietary database can be used to develop a biological association network to identify putative coregulated signaling pathways using expression data.

Quantitative real-time PCR validation. Quantitative real-time RT-PCR was done on 100 ng of double-amplified product from the 15 specimens using primer sets specific for 23 select genes and the housekeeping genes GAPDH, GUSB, and cyclophilin. An iCycler iQ Real-time PCR Detection System (Bio-Rad Laboratories, Hercules, CA) was used in conjunction with the QuantiTect SYBR Green RT-PCR Kit (Qiagen, Inc., Valencia, CA) according to previously described cycling conditions (13). To calculate the relative expression for each gene, the 2–{Delta}{Delta}CT method was used, averaging the CT values for the three housekeeping genes for a single reference gene value (14).

Immunohistochemical staining. Paraffin sections were stained for the following antibodies: rabbit anti-Fyn at 1:400 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), rabbit anti–focal adhesion kinase (FAK) at 1:50, mouse anti–matrix metalloproteinase (MMP)-9 at 1:40 (Oncogene Research Products, Boston, MA), anti–ß2-arrestin at 1:200 (Santa Cruz Biotechnology), anti-PLXDC1 at 1:200 (Abcam, Inc., Cambridge, MA), or anti-Jagged1 at 1:200 (Santa Cruz Biotechnology) diluted in PBS at 4°C. After three washes in PBS, sections were incubated with secondary antibody for 1 h at room temperature. Positive reactions were rendered visible by incubating the slides with stable 3,3-diaminobenzidine for 5 to 10 min. Sections were rinsed with distilled water, counterstained with Gill's hematoxylin for 30 s, and mounted with Universal Mount (Research Genetics, Huntsville, AL). The intensity of protein expression in the endothelial cells was evaluated using OPTIMAS 6.5 software and the mean absorbance was calculated from five normal ovarian tissue and five ovarian cancer samples. Ten vessels were randomly selected from each sample for the measurements.

Small interfering RNA. The small interfering RNA (siRNA) constructs were purchased from Qiagen (Germantown, MD): a control sequence with no homology to any human mRNA (as determined by BLAST search) and separate sequences designed to target EZH2, Jagged1, or PTK2 mRNA. The EZH2 siRNA was targeted to the region corresponding to residues 85 to 106 of human EZH2 (NM004456). The Jagged1 siRNA target sequence is 5'-CTGCATTTAGGGAGTATTCTA-3'. For in vitro delivery, siRNA (5 µg) was incubated with 30 µL of RNAiFect transfection reagent (Qiagen) for 10 min at room temperature and added to cells in culture at 80% confluence in 35-mm culture plates.

Cell migration assay. Unstimulated motility was determined in membrane invasion culture system chambers containing polycarbonate filter (with 10-µm pores) that had been soaked in 0.1% gelatin, as previously described (15). Human umbilical vein endothelial cells (HUVEC; 1 x 105) were seeded in each upper well, allowed to incubate at 37°C for 6 h in DMEM containing 15% serum, and subsequently processed as described for the invasion assay.

Tube formation assay. Matrigel (12.5 mg/mL) was thawed at 4°C and 50 µL were quickly added to each well of a 96-well plate and allowed to solidify for 10 min at 37°C. The wells were then incubated for 6 h at 37°C with HUVECs (20,000 per well), which had previously been treated for 18 h with the indicated siRNA. The formation of capillary-like structures was examined microscopically and photographs (x50) were taken using a Retiga 1300 camera and a Zeiss Axiovert S100 microscope. The extent to which capillary-like structures formed in the gel was quantified by analysis of digitized images to determine the thread length of the capillary-like network using a commercially available image analysis program (Northern Eclipse, North Tonawanda, NY).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Endothelial isolation and confirmation of cell purity. Five normal ovaries and 10 invasive epithelial ovarian cancers were obtained and subjected to negative and positive immunoselection. Before carrying out microarray analysis, we tested the purity of all samples for endothelial cells with the endothelial cell markers P1H12 and von Willebrand factor. Immunostaining revealed that the immunopurification technique yielded endothelial cell purity of >95% in all samples (Fig. 1A ). Thus, our isolation technique resulted in a highly pure population of endothelial cells for subsequent analyses.


Figure 1
View larger version (45K):
[in this window]
[in a new window]

  		Figure 1. A, immunofluorescence (von Willebrand factor and Hoechst) staining of endothelial cells isolated from human normal ovary and ovarian cancer tissues. Endothelial cells were isolated from human normal ovarian and invasive epithelial ovarian cancer tissues by negative and positive immunoselection procedures as described in Materials and Methods. All photomicrographs were taken at original magnification (x100). B, quantitative real-time PCR validation of endothelial cell microarray data. Gene expression in 10 tumor isolates was calculated as mean fold change relative to five normal endothelial specimens (normal = 1) using the 2–{Delta}{Delta}CT method. Real-time validation confirmed significant differential expression of 23 up-regulated genes, including HES4, FYN, VAV2, PGF, ECGF1, PTK2, TNFAIP6, EZH2, EGFL6, STC1, CSPG2, ADAM12, COL5A3, COL18A1, PCOLCE, PMAIP1, CENTA2, MMP9, NPTX2, TMEPAI, ARRB2, JAG1, and PLXDC1, in tumor endothelial cells versus normal isolates. Of these genes, many (e.g., VAV2, TNFAIP6, and EZH2) have not been previously described in endothelial cells. C, immunohistochemical staining for PTK2, Fyn, MMP9, ß2-arrestin, Jagged1, and PLXDC1 in human normal ovarian and invasive epithelial ovarian cancers. All photomicrographs were taken at original magnification (x100 or x200). D, the intensity of protein expression in the endothelial cells from five normal ovarian and five ovarian cancer samples was evaluated using the OPTIMAS 6.5 software. For each protein, the intensity of staining in the normal ovarian samples was set at 1 and the fold change was calculated for the tumor endothelium. *, P < 0.01; **, P < 0.001.

 
Class comparison analysis of endothelial cell isolates. Total RNA from purified endothelial cells was subjected to microarray analysis using the Affymetrix Human U133 Plus 2.0 GeneChip platform. A multivariate permutation t test (P < 0.001), providing 95% confidence that the number of false discoveries did not exceed 10% of the complete gene list, identified 1,149 genes that were differentially regulated ≥2-fold in endothelium derived from epithelial ovarian cancers compared with normal ovarian tissue (Supplementary Table S1). In addition, global analysis of the gene list returned a P value of <5 x 10–4, indicating that tumor and endothelial isolates possess distinct expression profiles.

Genes up-regulated in ovarian cancer endothelium. Of the 652 genes that were up-regulated ≥2-fold, 35 genes were elevated at least 6-fold in tumor endothelium (Table 1 ), with 9 being elevated >10-fold. Multiple genes encoding proteins, such as collagens, involved in extracellular matrix function, TNFAIP6, ADAMTS4, MMP9, and MMP11, had increased expression in tumor vasculature compared with normal ovarian vasculature. As expected, the {alpha}v integrin (vitronectin receptor) was elevated 2.5-fold in tumor endothelium. Most of these genes have previously been shown to have increased expression in colon and breast cancer vasculature (7, 8) and may reflect gene alterations regardless of tumor type.


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

  		Table 1. Genes up-regulated by ≥6-fold in the tumor-associated endothelium

 
Several transcription factors were up-regulated in the ovarian cancer vasculature. For example, HEYL was increased 3-fold, which is also elevated in breast and colon cancer endothelium (7, 8). In addition, several novel transcription factors were identified, including E2F transcription factor 3 (E2F3; plays a role in cell proliferation; ref. 16), runt-related transcription factor 1 (RUNX1; has a direct role in angiogenesis; ref. 17), signal transducer and activator of transcription 2 (STAT2; involved in cellular proliferation; ref. 18), the SNAIL-related zinc-finger transcription factor SLUG (SNAI2; ref. 19), and Twist1 (20). These genes were elevated 2- to 18-fold in the ovarian cancer vasculature relative to normal ovarian endothelial cells. Whereas some of these genes have previously been shown to have direct or indirect effects on angiogenesis, the functional significance of others remains to be shown.

Whereas vascular endothelial growth factor (VEGF)–based targeting has improved response rates to therapy and overall survival in many cancers, most patients still eventually die from cancer. Therefore, additional targets are likely to be required to achieve curative therapy. Genes that are overexpressed on both tumor cells and tumor-associated endothelial cells may be particularly appealing as targets for antivascular therapy due to their ability to target both the epithelial and stromal compartments. For example, epidermal growth factor receptor (EGFR) expression was increased by 3.5-fold in the tumor endothelium. EGFR is known to be overexpressed in ovarian carcinomas and is predictive of poor outcome (21). We have previously shown that EGFR is overexpressed and phosphorylated in tumor endothelial cells, and dual targeting of VEGF receptor and EGFR in combination with paclitaxel is highly efficacious (6). Similarly, nonreceptor kinases such as FAK (or PTK2; 3.1-fold increase) and Fyn (4.7-fold increase), which are known to play functionally significant roles for tumor cells and endothelial cells, may represent novel targets for antivascular approaches (15).

Genes down-regulated in ovarian cancer endothelium. The reduction of gene expression in tumor versus normal vasculature may reveal genes that function to suppress tumor and/or vascular growth. Therefore, we next identified genes that were down-regulated in endothelial cells derived from ovarian cancer tissue. There were 497 genes with ≥2-fold decrease in expression in tumor endothelium, with 17 decreased at least 6-fold (Table 2 ). Interestingly, monoamine oxidase B (MAOB), a gene responsible for detoxification and degradation of monoamines, was decreased by 6.4-fold in the tumor endothelial cells (22). Decorin, a small multifunctional proteoglycan with antiangiogenic properties, was decreased by 4.8-fold (23). Several other genes with potential antiangiogenic or antiproliferative roles, such as fibulin-5 (FBLN-5) and checkpoint suppressor 1 (CHES1), were down-regulated by 4.5- and 4.3-fold, respectively (24, 25). The functional role of these and other down-regulated genes in the context of tumor angiogenesis remains to be determined.


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

  		Table 2. Down-regulated vascular genes in invasive epithelial ovarian cancer

 
Genes determined to be potential tumor endothelial markers. We identified differentially regulated genes uniquely expressed in tumor-associated endothelium versus normal endothelium, and compared these to tumor-associated epithelial cells versus ovarian surface epithelium as potential endothelial-specific markers. We have previously reported expression profiling of microdissected papillary serous ovarian cancers using the same microarray methods (10). The current list of differentially expressed genes in tumor-associated endothelial cells was compared with the gene list identified for laser microdissected tumor epithelial cells. A total of 534 differentially regulated genes were uniquely altered (up- or down-regulated) in the endothelial cells. The 28 genes with the greatest (≥6-fold) level of increase in the tumor related endothelial cells are listed in Table 3 .


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

  		Table 3. Genes specifically regulated in tumor endothelium

 
Validation of gene expression alterations. The 23 validated genes were selected in two rounds. To show the reproducibility of the microarray analysis, a series of 17 genes were selected at random, spanning a range of fold-changes (3.6–155.3; Fig. 1B). Of 17 primer sets, 15 yielded specific quantitative real-time PCR products when analyzed using Universal Human Reference RNA (Stratagene, La Jolla, CA), with 13 reaching statistical significance in tumor (n = 10) and normal (n = 5) isolates (P < 0.05) including PLXDC1, ARRB2, HES4, PGF, EGFL6, ADAM12, COL5A3, COL18A1, PCOLCE, PMAIP1, CENTA2, TMEPAI, and NPTX2. To substantiate the pathway analysis (presented below), a second set of genes implicated in endothelial tumor cell signaling was assessed. From a series of 12 genes, suitable primer sets were obtained for 10 genes. All 10 pathway members were successfully validated (P < 0.05) including FYN, VAV2, ECGF1, PTK2, TNFAIP6, EZH2, STC1, MMP9, JAG1, and CSPG2. These data provide important confirmation of the gene expression alterations identified by the microarray analysis.

To further examine whether the gene expression alterations identified by the microarray analysis also occur at the protein level, we did immunohistochemical staining for selected proteins on five normal ovaries and five invasive epithelial ovarian cancers. The microarray analysis identified FAK (PTK2; 3.1-fold), Fyn (4.7-fold), MMP9 (9.4-fold), ß2-arrestin (4.8-fold), Jagged1 (4.3-fold), and PLXDC1 (10.2-fold) as being significantly increased in tumor-associated endothelial cells, and these changes were validated by real-time RT-PCR (Fig. 1). Immunohistochemical peroxidase staining confirmed that both FAK and Fyn were indeed overexpressed in the tumor-associated endothelial cells (Fig. 1C and D) in all samples. There were no obvious differences in protein expression between arterioles and venules. Similarly, increased expression of MMP9, ß2-arrestin, Jagged1, and PLXDC1 was also confirmed at the protein level (Fig. 1C and D).

Signaling pathways modulated in tumor endothelium and their functional significance. Ovarian epithelial carcinomas arise from molecular events occurring in the epithelial layer, which affect changes in gene expression within surrounding nonepithelial cell populations. For endothelial cells, this altered signaling environment stimulates proliferation, migration, and tumor vascularization. To identify epithelial genes that may be responsible for these changes and the endothelial signaling pathways that are affected, a series of laser microdissected papillary serous epithelial cell isolates and ovarian surface epithelial brushings were compared, as previously described (10). Pathway diagrams were generated using PathwayAssist version 3.0 software (Fig. 2A and B ). The genes comprising the pathway suggest involvement in endothelial cell proliferation, tube formation, and cell motility. The genes in this pathway are summarized in Fig. 2C, along with the inferred biological function. To test the biological significance of some of these genes, we selected three genes: the novel polycomb group protein enhancer of Zeste homologue 2 (EZH2), the notch receptor ligand Jagged1, and PTK2. EZH2 plays an important role in many biological processes and is downstream of Akt activation (26), making it a potential antiangiogenic target. SiRNA was used to silence EZH2 expression (Fig. 3A ) in HUVECs and its effects on migration and tube formation were examined. In comparison with control nonsilencing siRNA, EZH2 silencing resulted in a 85% decrease in endothelial tube formation on Matrigel (Fig. 3B). EZH2-targeted siRNA completely blocked VEGF-stimulated migration of HUVECs (Fig. 3C). Similarly, to determine the functional relevance of Jagged1 (27) for endothelial cell function, we tested the effects of Jagged1 silencing with siRNA (Fig. 3A) on migration and tube formation. Jagged1-targeted siRNA reduced tube formation by 80% (Fig. 3B) and blocked VEGF-stimulated HUVEC migration (Fig. 3C). Similar results were noted with PTK2 silencing with siRNA (Fig. 3A–C). These data indicate that the novel differentially expressed genes in the tumor-associated endothelial cells play functionally significant roles in angiogenesis.


Figure 2
View larger version (28K):
[in this window]
[in a new window]

  		Figure 2. Analysis of putative pathways stimulated in tumor endothelial cells by papillary serous ovarian epithelial tumor cells. Pathway diagrams were generated with the assistance of PathwayAssist version 3.0 software using gene expression data. Genes included in the pathway were required to have a fold change value of ≥2.0. For some genes, multiple probe sets were averaged. A, differentially expressed genes modulated in tumor endothelial cells and their associated interactions are described. Green oval, gene is up-regulated in endothelial tumor samples; red oval, down-regulation; gray oval, genes displaying no change in expression. Green ovals with a red border, secreted genes up-regulated in ovarian epithelial tumor cells. B, pathways implicated in cell proliferation, tube formation (JAG1, JAG2, NOTCH1, HESR1, EZH2, and STC1), and cell motility. C, table describing pathway genes and their inferred biological function.

 

Figure 3
View larger version (37K):
[in this window]
[in a new window]

  		Figure 3. A, siRNA-mediated silencing of EZH2, Jagged1, and PTK2 was assessed using immunofluorescence and Western blot. B, effect of EZH2, Jagged1, or PTK2 silencing on HUVEC tube formation. C, effect of EZH2, Jagged1, or PTK2 silencing on HUVEC migration. *, P < 0.01.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The major finding of the present study is that ovarian cancer–associated endothelial cells contain a large number of gene alterations in comparison with normal ovarian endothelial cells. We identified genes that are unique to the tumor vasculature and those that overlap in expression with tumor cells. To the best of our knowledge, this is the largest study of genomic profiling of ovarian cancer–associated endothelial cells.

Cancer is a heterogeneous disease that requires multimodality therapy. To date, most of the therapeutic approaches for ovarian cancer have focused on chemotherapy, which primarily targets proliferating tumor cells. However, despite initial responses, most tumors eventually develop resistance. Over the last few years, the critical role of the microenvironment in ovarian cancer growth and progression has been established (6, 28, 29). The presumed genetic stability of the microenvironment components such as endothelial cells makes them an attractive therapeutic target. Indeed, biological therapies aimed at the microenvironment, such as bevacizumab and VEGF Trap, are starting to show promise in clinical trials (4) and preclinical models (30, 31) in ovarian and other cancers. However, it is likely that additional targets will be required to achieve further gains in therapeutic benefit. Genomic profiling of tumor-associated endothelial cells, as described in the current study, is a powerful method for identifying novel genes that may be potential targets or biomarkers.

It has become evident that endothelial cells vary phenotypically depending on the organ of origin (32). To date, there are limited data about genomic differences in tumor-associated endothelial cells from specific tumors. A small number of either colon, breast, or brain tumors have been evaluated previously using serial analysis of gene expression (7–9). Whereas there was overlap in the gene profile of endothelial cells from ovarian versus these other cancers, there were also clear differences. For example, genes such as {alpha}v-integrin, ADAMTS4, HEYL, and MMP9 were increased in ovarian cancer vasculature and have been noted to have increased expression in the vasculature from other cancers as well. However, several unique genes such as TNFAIP6, E2F3, EZH2, and RUNX1 were also discovered. Whether these are indeed unique to ovarian cancer vasculature or tumor vasculature more broadly will require further investigation.

A number of genes that were identified are known to play functionally significant roles in both tumor cells and tumor-associated endothelial cells. For example, PTK2 (FAK) has been shown to play a role in ovarian cancer cell migration and invasion (15) and survival (33). With regard to endothelial cells, PTK2 plays a pivotal role in angiogenesis related to late embryonic development (34) and modulates endothelial cell migration (35). Data from our study show that PTK2 expression was up-regulated in tumor-related endothelium. We have recently shown that PTK2 silencing with siRNA sensitized tumor cells to docetaxel chemotherapy in vitro (33). Moreover, in vivo FAK silencing with siRNA incorporated in neutral liposomes was highly efficacious in both chemotherapy-sensitive and chemotherapy-resistant ovarian cancer models through both direct and indirect antitumor effects by decreasing VEGF levels (36). Based on the results of the current study, we anticipate that such effects may be even greater when FAK is suppressed directly in endothelial cells.

It is well known that Src family tyrosine kinases play an essential role in the signaling of integrin-mediated biological processes such as cell proliferation, differentiation, actin organization, and cell migration (37). The Src family member Fyn plays a key role in endothelial cell signaling pathways resulting in stimulation of endothelial migration and tube formation (38). In the present study, gene profiling revealed that Fyn expression was increased by 4.7-fold in tumor-related endothelium, and this increase was validated at both mRNA and protein levels. Based on its known role in endothelial cell function and significant up-regulation in tumor-associated endothelial cells, Fyn may be an antiangiogenic target.

One of the novel genes identified from the current study is EZH2, a member of the polycomb group of genes (PcG), which are important for transcriptional regulation through nucleosome modification, chromatin remodeling, and interaction with other transcription factors (39). EZH2 is controlled by E2F and is also involved in p53-regulated cell cycle control (40). EZH2 was previously shown to be overexpressed in prostate, breast, and other cancers (41). To the best of our knowledge, this is the first report that EZH2 is overexpressed in tumor-associated endothelial cells. In the present study, the microarray data yielded a 2.9-fold higher EZH2 expression in ovarian cancer–related endothelium compared with normal ovary. We also showed the functional relevance of EZH2 for endothelial cell migration and tube formation using RNA interference. HUVECs were used for these assays and it is possible that the role of EZH2 may be different in ovarian or mesenteric endothelial cells. Whether EZH2 is a valid antiangiogenic target will require further investigation, but its role in endothelial function combined with its significance for tumor cell function makes it an appealing candidate.

Pathway analysis identified coordinated signaling events stimulated by transformed ovarian epithelial cells, which may be modulating tumor endothelial cell behavior. The proangiogenic effect of EZH2 on endothelial cells and increased CCNE1 levels may be associated, in part, with epithelial cell–induced VEGF signaling. In cortical neuron precursors, VEGF induces E2F3 expression (42). Whereas neuron development and migration is seemingly unrelated to angiogenesis, there is increasing evidence that the two processes may use analogous pathways (43). Enhanced E2F3 expression may result in elevated levels of EZH2 and CCN1, both of which are direct targets of the transcription factor (44, 45). Notch signaling has also been linked to the regulation of tube formation and the ability of endothelial cells to establish a mature phenotype (46). Both JAG2, secreted by epithelial tumor cells, and endothelial cell–derived JAG1 can activate the Notch pathway. HESR1 is a Notch responsive transcription factor that has been implicated in the regulation of endothelial cell tube formation (47). Consequently, the ability of JAG1 to stimulate tube formation may be mediated through the downstream targets of this gene. In addition to promoting endothelial cell tube formation and proliferation, secreted epithelial factors may also drive endothelial cell motility. For example, SPP1 can engage the {alpha}v integrin receptor and support directional cell migration (48). Furthermore, it can interact with the CD44 receptor stimulating VAV2 activity (49). This interaction may be stabilized by TNFAIP6, resulting in CD44 receptor clustering and enhanced signaling (50). The net effect of these interactions is increased FAK stimulation and endothelial cell migration. Signaling events originating from epithelial tumor cells and their downstream endothelial effectors represent a reservoir of putative targets for therapeutic intervention. Of particular interest are the novel genes E2F3, EZH2, and TNFAIP6, which may directly participate in critical proangiogenic pathways.

In summary, the expression profile of ovarian cancer–associated endothelial cells is distinct and unique. However, there are similarities with tumor vasculature in other organs. Moreover, there are multiple genes that have increased expression in both tumor-associated endothelial cells and tumor cells. Additional work is needed to define the role of the novel genes identified here in processes related to tumor vascularization, invasion, and metastatic growth. Some of these may offer opportunities for therapeutic intervention.


   Acknowledgments
 
Grant support: NIH grants CA10929801 and CA11079301, Program Project Development Grant from the Ovarian Cancer Research Fund, The Marcus Foundation, The Zarrow Foundation, and the University of Texas M. D. Anderson Cancer Center Specialized Program of Research Excellence in Ovarian Cancer grant P50 CA 083639.

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.

We thank Drs. Isaiah J. Fidler and Robert Langley at the University of Texas M. D. Anderson Cancer Center (Houston, TX) for helpful input and discussions about this work; Joseph Celestino for assistance with specimen collection; Donna Reynolds for assistance with immunohistochemistry; and Susan Davis and Catherine Rodgers for assistance with manuscript preparation.


   Footnotes
 
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

C. Lu and T. Bonome contributed equally to this work. M.J. Birrer and A.K. Sood share senior authorship.

5 http://nciarray.nci.nih.gov/index.shtml Back

Received 10/10/06. Revised 11/26/06. Accepted 12/ 4/06.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Cannistra SA. Cancer of the ovary. N Engl J Med 1993;329:1550–9.[Free Full Text]
  2. Folkman J. What is the evidence that tumors are angiogenesis dependent? J Natl Cancer Inst 1990;82:4–6.[Free Full Text]
  3. Hurwitz H, Fehrenbacher L, Novotny W, et al. Bevacizumab plus irinotecan, fluorouracil, and leucovorin for metastatic colorectal cancer. N Engl J Med 2004;350:2335–42.[Abstract/Free Full Text]
  4. Jain RK, Duda DG, Clark JW, Loeffler JS. Lessons from phase III clinical trials on anti-VEGF therapy for cancer. Nat Clin Pract Oncol 2006;3:24–40.[CrossRef][Medline]
  5. Cooper BC, Ritchie JM, Broghammer CL, et al. Preoperative serum vascular endothelial growth factor levels: significance in ovarian cancer. Clin Cancer Res 2002;8:3193–7.[Abstract/Free Full Text]
  6. Thaker PH, Yazici S, Nilsson MB, et al. Antivascular therapy for orthotopic human ovarian carcinoma through blockade of the vascular endothelial growth factor and epidermal growth factor receptors. Clin Cancer Res 2005;11:4923–33.[Abstract/Free Full Text]
  7. Parker BS, Argani P, Cook BP, et al. Alterations in vascular gene expression in invasive breast carcinoma. Cancer Res 2004;64:7857–66.[Abstract/Free Full Text]
  8. St Croix B, Rago C, Velculescu V, et al. Genes expressed in human tumor endothelium. Science 2000;289:1197–202.[Abstract/Free Full Text]
  9. 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]
  10. Bonome T, Lee JY, Park DC, et al. Expression profiling of serous low malignant potential, low-grade, and high-grade tumors of the ovary. Cancer Res 2005;65:10602–12.[Abstract/Free Full Text]
  11. Simon R, McShane L, Radmacher M, Wright G, Zhao Y. Design and analysis of DNA microarray investigations. New York: Springer-Verlag; 2004.
  12. Wright GW, Simon RM. A random variance model for detection of differential gene expression in small microarray experiments. Bioinformatics 2003;19:2448–55.[Abstract/Free Full Text]
  13. Donninger H, Bonome T, Radonovich M, et al. Whole genome expression profiling of advance stage papillary serous ovarian cancer reveals activated pathways. Oncogene 2004;23:8065–77.[CrossRef][Medline]
  14. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(–{Delta}{Delta}C(T)) method. Methods 2001;25:402–8.[CrossRef][Medline]
  15. Sood AK, Coffin JE, Schneider GB, et al. Biological significance of focal adhesion kinase in ovarian cancer: role in migration and invasion. Am J Pathol 2004;165:1087–95.[Abstract/Free Full Text]
  16. Black EP, Hallstrom T, Dressman HK, West M, Nevins JR. Distinctions in the specificity of E2F function revealed by gene expression signatures. Proc Natl Acad Sci U S A 2005;102:15948–53.[Abstract/Free Full Text]
  17. Iwatsuki K, Tanaka K, Kaneko T, et al. Runx1 promotes angiogenesis by down-regulation of insulin-like growth factor-binding protein-3. Oncogene 2005;24:1129–37.[CrossRef][Medline]
  18. Gomez D, Reich NC. Stimulation of primary human endothelial cell proliferation by IFN. J Immunol 2003;170:5373–81.[Abstract/Free Full Text]
  19. Perez-Mancera PA, Gonzalez-Herrero I, Perez-Caro M, et al. SLUG in cancer development. Oncogene 2005;24:3073–82.[CrossRef][Medline]
  20. Mironchik Y, Winnard PT, Jr., Vesuna F, et al. Twist overexpression induces in vivo angiogenesis and correlates with chromosomal instability in breast cancer. Cancer Res 2005;65:10801–9.[Abstract/Free Full Text]
  21. Berchuck A, Rodriguez GC, Kamel A, et al. Epidermal growth factor receptor expression in normal ovarian epithelium and ovarian cancer. I. Correlation of receptor expression with prognostic factors in patients with ovarian cancer. Am J Obstet Gynecol 1991;164:669–74.[Medline]
  22. Grimsby J, Toth M, Chen K, et al. Increased stress response and ß-phenylethylamine in MAOB-deficient mice. Nat Genet 1997;17:206–10.[CrossRef][Medline]
  23. Sulochana KN, Fan H, Jois S, et al. Peptides derived from human decorin leucine-rich repeat 5 inhibit angiogenesis. J Biol Chem 2005;280:27935–48.[Abstract/Free Full Text]
  24. Albig AR, Schiemann WP. Fibulin-5 antagonizes vascular endothelial growth factor (VEGF) signaling and angiogenic sprouting by endothelial cells. DNA Cell Biol 2004;23:367–9.[CrossRef][Medline]
  25. Scott KL, Plon SE. CHES1/FOXN3 interacts with Ski-interacting protein and acts as a transcriptional repressor. Gene 2005;359:119–26.[CrossRef][Medline]
  26. Cha TL, Zhou BP, Xia W, et al. Akt-mediated phosphorylation of EZH2 suppresses methylation of lysine 27 in histone H3. Science 2005;310:306–10.[Abstract/Free Full Text]
  27. Zeng Q, Li S, Chepeha DB, et al. Crosstalk between tumor and endothelial cells promotes tumor angiogenesis by MAPK activation of Notch signaling. Cancer Cell 2005;8:13–23.[CrossRef][Medline]
  28. Sood AK, Seftor EA, Fletcher MS, et al. Molecular determinants of ovarian cancer plasticity. Am J Pathol 2001;158:1279–88.[Abstract/Free Full Text]
  29. Liotta LA, Kohn EC. The microenvironment of the tumour-host interface. Nature 2001;411:375–9.[CrossRef][Medline]
  30. Hu L, Hofmann J, Holash J, Yancopoulos GD, Sood AK, Jaffe RB. Vascular endothelial growth factor trap combined with paclitaxel strikingly inhibits tumor and ascites, prolonging survival in a human ovarian cancer model. Clin Cancer Res 2005;11:6966–71.[Abstract/Free Full Text]
  31. Byrne AT, Ross L, Holash J, et al. Vascular endothelial growth factor-trap decreases tumor burden, inhibits ascites, and causes dramatic vascular remodeling in an ovarian cancer model. Clin Cancer Res 2003;9:5721–8.[Abstract/Free Full Text]
  32. Langley RR, Ramirez KM, Tsan RZ, Van Arsdall M, Nilsson MB, Fidler IJ. Tissue-specific microvascular endothelial cell lines from H-2K(b)-tsA58 mice for studies of angiogenesis and metastasis. Cancer Res 2003;63:2971–6.[Abstract/Free Full Text]
  33. Halder J, Landen CN, Jr., Lutgendorf SK, et al. Focal adhesion kinase silencing augments docetaxel-mediated apoptosis in ovarian cancer cells. Clin Cancer Res 2005;11:8829–36.[Abstract/Free Full Text]
  34. Shen TL, Park AY, Alcaraz A, et al. Conditional knockout of focal adhesion kinase in endothelial cells reveals its role in angiogenesis and vascular development in late embryogenesis. J Cell Biol 2005;169:941–52.[Abstract/Free Full Text]
  35. Kaczmarek E, Erb L, Koziak K, et al. Modulation of endothelial cell migration by extracellular nucleotides: involvement of focal adhesion kinase and phosphatidylinositol 3-kinase-mediated pathways. Thromb Haemost 2005;93:735–42.[Medline]
  36. Halder J, Kamat AA, Landen CN, et al. Focal adhesion kinase targeting using in vivo short interfering RNA delivery in neutral liposomes for ovarian carcinoma therapy. Clin Cancer Res 2006;12:4916–24.[Abstract/Free Full Text]
  37. Playford MP, Schaller MD. The interplay between Src and integrins in normal and tumor biology. Oncogene 2004;23:7928–46.[CrossRef][Medline]
  38. Mochizuki Y, Nakamura T, Kanetake H, Kanda S. Angiopoietin 2 stimulates migration and tube-like structure formation of murine brain capillary endothelial cells through c-Fes and c-Fyn. J Cell Sci 2002;115:175–83.[Abstract/Free Full Text]
  39. Varambally S, Dhanasekaran SM, Zhou M, et al. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 2002;419:624–9.[CrossRef][Medline]
  40. Tang X, Milyavsky M, Shats I, Erez N, Goldfinger N, Rotter V. Activated p53 suppresses the histone methyltransferase EZH2 gene. Oncogene 2004;23:5759–69.[CrossRef][Medline]
  41. Kleer CG, Cao Q, Varambally S, et al. EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc Natl Acad Sci U S A 2003;100:11606–11.[Abstract/Free Full Text]
  42. Zhu Y, Jin K, Mao XO, Greenberg DA. Vascular endothelial growth factor promotes proliferation of cortical neuron precursors by regulating E2F expression. FASEB J 2003;17:186–93.[Abstract/Free Full Text]
  43. Carmeliet P. Blood vessels and nerves: common signals, pathways and diseases. Nat Rev Genet 2003;4:710–20.[CrossRef][Medline]
  44. Bracken AP, Pasini D, Capra M, Prosperini E, Colli E, Helin K. EZH2 is downstream of the pRB-E2F pathway, essential for proliferation and amplified in cancer. EMBO J 2003;22:5323–35.[CrossRef][Medline]
  45. Humbert PO, Verona R, Trimarchi JM, Rogers C, Dandapani S, Lees JA. E2f3 is critical for normal cellular proliferation. Genes Dev 2000;14:690–703.[Abstract/Free Full Text]
  46. Sainson RC, Harris AL. Hypoxia-regulated differentiation: let's step it up a Notch. Trends Mol Med 2006;12:141–3.[CrossRef][Medline]
  47. Henderson AM, Wang SJ, Taylor AC, Aitkenhead M, Hughes CC. The basic helix-loop-helix transcription factor HESR1 regulates endothelial cell tube formation. J Biol Chem 2001;276:6169–76.[Abstract/Free Full Text]
  48. Liaw L, Skinner MP, Raines EW, et al. The adhesive and migratory effects of osteopontin are mediated via distinct cell surface integrins. Role of {alpha}vß3 in smooth muscle cell migration to osteopontin in vitro. J Clin Invest 1995;95:713–24.[Medline]
  49. Lin YH, Yang-Yen HF. The osteopontin-CD44 survival signal involves activation of the phosphatidylinositol 3-kinase/Akt signaling pathway. J Biol Chem 2001;276:46024–30.[Abstract/Free Full Text]
  50. Lesley J, Gal I, Mahoney DJ, et al. TSG-6 modulates the interaction between hyaluronan and cell surface CD44. J Biol Chem 2004;279:25745–54.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 JCOHome page
C. N. Landen Jr, M. J. Birrer, and A. K. Sood
Early Events in the Pathogenesis of Epithelial Ovarian Cancer
J. Clin. Oncol., February 20, 2008; 26(6): 995 - 1005.
[Abstract] [Full Text] [PDF]

Home page
 Cancer Res.Home page
J. Halder, Y. G. Lin, W. M. Merritt, W. A. Spannuth, A. M. Nick, T. Honda, A. A. Kamat, L. Y. Han, T. J. Kim, C. Lu, et al.
Therapeutic Efficacy of a Novel Focal Adhesion Kinase Inhibitor TAE226 in Ovarian Carcinoma
Cancer Res., November 15, 2007; 67(22): 10976 - 10983.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplementary Data
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 Lu, C.
Right arrow Articles by Sood, A. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Lu, C.
Right arrow Articles by Sood, A. K.

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