In the United States, ovarian cancer is the fourth most common cause of cancer-related deaths among women. The most important prognostic factor for this cancer is tumor stage, or extent of disease at diagnosis. Although women with low-stage tumors have a relatively good prognosis, most women diagnosed with late-stage disease eventually succumb to their cancer. In an attempt to understand early events in ovarian carcinogenesis, and to explore steps in its progression, we have applied multiple molecular genetic techniques to the analysis of 21 early-stage (stage I/II) and 17 advanced-stage (stage III/IV) ovarian tumors. These techniques included expression profiling with cDNA microarrays containing approximately 18,000 expressed sequences, and comparative genomic hybridization to address the chromosomal locations of copy number gains as well as losses. Results from the analysis indicate that early-stage ovarian cancers exhibit profound alterations in gene expression, many of which are similar to those identified in late-stage tumors. However, differences observed at the genomic level suggest differences between the early- and late-stage tumors and provide support for a progression model for ovarian cancer development.

Of the cancers unique to women, ovarian cancer has the highest mortality rate. Over 26,000 women are diagnosed with this disease in the United States annually, and 60% of those diagnosed will die of the disease (1). There has been little change in ovarian cancer incidence and mortality over the past 5 decades and unfortunately there are a number of significant barriers to progress in its treatment. These include poor understanding of the underlying biology of this disease, inadequate screening tools as well as few early warning signs. The 5-year survival for patients with stage I disease can exceed 90%, but it is less than 25% for advanced-stage disease (2). These statistics underscore the need for better tools for the screening and staging of ovarian cancer.

Like other solid tumors, ovarian cancer is thought to result from an accumulation of genetic alterations. These events lead to changes in expression of many genes. Alterations in tumor suppressor genes such as p53 (3, 4), pRB (3, 5), NOEY2 (6), BRCA1 (7), and oncogenes such as K-ras (8), c-myc (9) and c-erbB-2 (10) have been shown to play an important role in ovarian carcinogenesis. There is very little information, however, on the sequence of genetic changes that are associated with progression of disease from early to advanced stage.

Examining tumors for alterations in gene expression is a potentially useful approach to identifying molecular differences between early- and late-stage ovarian carcinomas. Although there are several approaches to investigate differential gene expression in tumors, here we have used cDNA microarrays to develop expression profiles of early- and late-stage ovarian cancer (11, 12, 13). To extend our understanding of any stepwise genetic alterations that may underlie the assumed clinical progression from early- to late-stage ovarian cancer, CGH3 (14) was used to identify regions of genome loss and gain in the same series of tumors.

Tissue Processing and Tumor Selection.

Surgically removed ovarian tumors were snap-frozen in the surgical pathology unit of the Mayo Clinic. The tumor content of the specimens was assessed by H&E-stained sections. Only high-grade specimens containing more than 75% tumor were used for these experiments. Twenty normal OSEs from patients without cancer were used as normal controls. The epithelial nature of these brushings was verified by cytokeratin staining. The majority of these patients were between 45 and 65 years old, undergoing incidental oophorectomy at the time of pelvic surgery for other indications. All of the ovaries were examined pathologically and found to be benign or they were excluded. RNA from five such pooled brushings were profiled on cDNA microarrays along with five stage I, two stage II, and seven stage III ovarian tumors. The histology, grade, and stage of each tumor used in cDNA microarray, semiquantitative RT-PCR, and CGH studies are listed in Tables 1 and 5, respectively. Tumors were staged according to International Federation of Gynecology and Obstetrics criteria.

Ovarian Cell Lines.

Five of seven ovarian-carcinoma cell lines (OV 167, OV 177, OV 202, OV 207, and OV 266) were low-passage primary lines established at the Mayo Clinic (15). OVCAR-5 is a NIH cell line (16) and SKOV-3 was purchased from American Type Culture Collection (Manassas, VA). All of the cells were grown according to the provider’s recommendations.

cDNA Microarrays.

The cDNA microarray consisted of 25,000 elements from the Unigene set from Research Genetics Inc. (Huntsville, AL), which included 10,000 known genes; 13,000 ESTs; and 2,000 elements made up of control genes, dyes, bacterial genes, and water controls. After consolidating genes or ESTs belonging to the same Unigene clusters, we estimated that there were a total of 18,304 unique genes or ESTs. All of the genes and ESTs on the array were sequence verified before being arrayed on a high-precision robotics platform. Each 3- × 5-inch membrane contained up to 6,000 elements and each probe was hybridized to five sets of membranes in duplicate.

Total RNA Isolation and Labeling.

Total RNA was extracted from specific tumor samples using Trizol (Life Technologies, Inc., Rockville, MD) as recommended by the manufacturer after estimating the tumor content by standard histological methods. The integrity of the RNA was assessed by ethidium bromide staining after agarose gel electrophoresis. Total RNA was labeled with [α-33P]CTP after reverse transcription with oligo(dT)30. Fifteen μg of total RNA was annealed to 3 μg of oligo-dT30 primer in 13 μl of total volume and incubated at 70°C for 2 min in a PCR machine and held at 4°C. Reverse transcription of this annealed template was carried out in a final volume of 50 μl, 1× first strand buffer (DTT, 40 units of Rnase inhibitor, 400 units of Superscript II reverse transcriptase, 2 μl of 10 mm dNTP mix with dCTP at 0.1 mm concentration, 100 μCi of [α-33P] dCTP at 42°C for 60 min. The reaction was stopped by consecutive addition of the following reagents: 4 μl of 50 mm EDTA, 4 μl of 0.5 m NaOH, and 2 μl of diethylpyrocarbonate water. Heating at 65°C for 10 min degrades the RNA template in the reaction. Unincorporated nucleotides were removed from the labeled probe using a Clontech (Palo Alto, CA) Chroma Spin TE-30 spin column. The labeled probe with an estimated specific activity of 1 × 106 cpm/μl was preannealed to 100 mg of COT1 DNA to block repetitive sequences in a final volume of 100 μl. The blocked probe was heated for 5 min at 95°C followed by 65°C for 20 min and then added to the duplicate filters along with 5 μg/ml of denatured salmon sperm DNA. The filters were hybridized overnight at 65°C in a Hybaid hybridization oven. Each probe was hybridized to duplicate filters. The next day, the filters were washed for 15 min at 65°C with 4% SDS solution in 20 mm sodium phosphate (pH 7.2) with 1 mm EDTA followed by 1% SDS solution in the same buffer with three changes at 65°C. After the wash, the filters were dried between whatman papers and baked at 80°C for 1 h. The filters were imaged on a storm 840 phosphoimager after 24–48 h exposure.

Image Analysis and Data Recovery.

The ArrayVision program (Imaging Research Inc., St. Catharines, ON, Canada) was used to quantify the spot intensities of the scanned images for each membrane. For each membrane, the spot intensities were normalized by the median intensity for the entire array, resulting in a new median intensity of 1.0. Spots that were <0.1 after normalization were thresholded to 0.1 to account for the level of background noise. After normalization and thresholding, spot intensities were averaged across the duplicate filters for each tissue sample. Fold-regulation for each tumor, at each gene, was determined by calculating the ratio of tumor expression:average expression for the five (pooled) control normal OSEs.

Semiquantitative RT-PCR.

Reverse transcribed cDNAs (50–100 ng) were used in a multiplex reaction with gene-specific primers and GAPDH, forward primer (5′-ACCACAGTCCATGCCATCAC-3′) and reverse primer (5′-TCCACCACCCTGTTGCTTGTA-3′) yielding a 450-bp product. The PCR reaction mixtures consist of 50 mm Tris-HCl (pH 8.3), 1.5 mm MgCl2, 400 μm concentration of each primer for the specific gene to be analyzed (50 μm for the GAPDH primers), and 0.5 units of Taq polymerase (Promega, Madison, WI), in a 12.5-μl reaction volume. The conditions for amplification were as follows: 94°C for 3 min, then 29 cycles of 94°C for 30 s, 50–62°C for 30 s depending on the gene-specific primers being tested, and 72°C for 30 s in a Perkin-Elmer-Cetus (Norwalk, CT) 9600 Gene-Amp PCR system. The products of the reactions were resolved on a 1.6% agarose gel. Band intensities were quantified using the Gel Doc 1000 photo-documentation system (Bio-Rad, Hercules, CA) and its associated software. Gene specific primers were as follows: for Gas 1, forward, CGC GCC TCG TCT CCT TTC CC, and reverse, GGC GCG TGG GCT AAA AGA GC; for PAI1, forward AAT CGC AAG GCA CCT CTG AG, and reverse, GAT CTG GTT TAC CAT CTT TT; for StAR, forward, GAC CCC ACC ACT GCC ACA TT, and reverse, GAT CTT AGA CTT GCA GGC TT; for AREG, forward, CCG CTG CGA AGG ACC AAT GA, reverse, CTA TGA CTT GGC AGT CAG TC; for FGF7, forward, TAA TGC ACA AAT GGA TAC, and reverse, ATT GCC ATA GGA AGA AAG; for HPR6, forward, TGC CTA GCG CGG CCC AAC, and reverse, CAG ACT GGA CTG TTA CAA ATG; for ITM2A, forward, CGC AGC CCG AAG ATT CAC TAT G, and reverse, R-CTT ATT ACC AAG GAC ACT CTA TCT; and for decorin, forward, CCT GGT TGT GAA AAT ACA TGA, and reverse, TGA CAT TAA CAA GAT TTT GCC.

CGH.

Metaphase spreads from normal human lymphocytes were prepared using standard protocols (17). The slides were aged for 2–3 days prior to denaturation at 70°C by 70% formamide in 2× SSC, followed by dehydration in a series of ethanol-water mixtures. The slides were treated with proteinase K, at a concentration of 0.1 μg/ml in 20 mm Tris (pH 7.5)-2 mm CaCl2, prior to hybridization. The CGH procedure was similar to published standard protocols (14, 17). Twenty images were captured using a Nikon Labophot-2 microscope equipped with an automatic filter wheel and an 83,000 filter set (Chroma, Brattleboro, VT) with single band pass exciter filters for UV/FITC (490 nm), 4′,6-diamidino-2-phenylindole (360 nm), and rhodamine (570 nm), and were analyzed using the QUIPS CGH software version 3.12 (Vysis Inc., IL). Using this software, the ratio of rhodamine:FITC signal is expressed as a red:green ratio, with deviations from a 1:1 ratio indicative of gain or loss of chromosome material. The lower and upper limits for gain and loss were established by performing control CGH experiments with a well-characterized tumor cell line, IMR32 (18), and DNA derived from male or female normal tissue. On the basis of these findings, a 95% confidence interval for gain and loss was set at 1.20 and 0.80, respectively, with gene amplification defined as gain >1.5.

Cluster Analyses.

Hierarchical cluster analyses were performed using the Cluster software package (19). Genes were clustered using the Pearson correlation coefficient as the distance metric, and clusters were agglomerated using the average linkage criterion. Graphical displays of our cluster analyses were obtained using the TreeView program (19).

Gene Expression Patterns of Early-Stage (I/II) and Late-Stage (III/IV) Tumors.

We examined gene expression in seven early- and seven late-stage ovarian tumors (Table 1) using cDNA microarray filters containing approximately 25,000 members of the Unigene set and control probes. Each tumor was examined in duplicate, with results indicating excellent reproducibility (Pearson correlation, >0.97 in all cases). The overall degree of similarity in gene expression between duplicate microarray experiments and among different tissue specimens was assessed using scatter plots. A representative scatter plot for two duplicate filters is shown in Fig. 1,A (Pearson correlation, 0.97), and Fig. 1 B displays the scatter plot of an early-stage tumor (OV106) versus one of the pooled normal epithelial brushings (Pearson correlation coefficient, 0.76).

Analysis of the Differential Expression in Early- and Late-Stage Ovarian Tumors.

Table 2 displays the number of genes for which we found varying levels of differential expression relative to normal ovarian epithelial cell brushings. Generally speaking, more genes were determined to be down-regulated than up-regulated. For example, 46 genes are at least 10-fold down-regulated in at least seven of the tumors that were expression-profiled. The number of genes that are 10-fold up-regulated in at least seven tumors is only 12. A similar trend was evident when the tumors were grouped by stage (Table 3). Heirarchical clustering for a set of genes selected for maximal variance (without prejudice toward differential expression) are shown in Fig. 2. The dendrogram in Fig. 2 shows the tight clustering of normal samples, whereas separation between early- and late-stage samples is far less compelling. Figs. 3 and 4 display heirarchical clustering for a set of genes that are ≥5-fold up- and down-regulated in a significant number of early- and late-stage tumors, respectively, relative to the pooled epithelial cell brushings. It is evident from these analyses that several of the genes are aberrantly regulated to the same extent in both early- and late-stage tumors.

To assess the extent to which genes in specific functional categories are differentially expressed in both early- and late-stage tumors, expressed sequences showing five times the expression changes in a majority of early-and late-stage tumors were grouped in one of four categories: cell-cell interactions; intermediate filament markers; cell cycle and growth regulators; and genes involved in invasion and metastasis (Fig. 5). This analysis revealed that genes involved in cell-cell interactions such as cadherin 11 (20), cadherin 2 (21), and nidogen (enactin; Ref. 22) were all down-regulated in a majority of tumors, whereas genes involved in invasion and metastasis, including matrilysin (MMP 7; Refs. 23, 24), gelatinase (MMP 9; Ref. 25), matrix metalloproteinase 10 and 12 (Ref. 23) were up-regulated in a majority of tumors. Genes belonging to the intermediate filament category, such as vimentin (26) and keratin 10 (27) were down-regulated in both early- and late-stage tumors, whereas keratins 8 (28), 13 (29), and 18 (28) were up-regulated in both stages. For the cell cycle/growth regulators, cyclin D2 (30, 31), cyclin dependent kinase 5, growth arrest-specific 1 (32), and the E2F-related transcription factor DP1 (33) were all down-regulated. Other genes in this category, such as cyclin E2 (34), proliferating cell nuclear antigen (35), and transcription factor E2F5 (36), were up-regulated in both early and late-stage tumors.

The genes with an average-fold change of at least 10-fold change across all tumors are listed in Table 4. Several of these genes, such as mucin 1 (37), ceruloplasmin ferroxidase (38), claudin 4, and HE4 (39, 40) have been previously identified as up-regulated in ovarian tumors.

Validation of Microarray Results by Semiquantitative RT-PCR.

To validate the expression levels of genes from the profiling analysis, we performed semiquantitative RT-PCR, with GAPDH as a control in seven ovarian tumor cell lines, on 20 early (I/II)- and 16 late (III/IV)-stage tumors (Table 1) with a set of genes that showed 5-fold differential regulation in at least 50% of the tumors. Fig. 6,A shows the expression levels of a subset of the genes in ovarian tumor cell lines compared with short-term cultures of ovarian epithelial cells. Fig. 6 B shows the expression levels of four of these genes in primary ovarian tumors compared with normal epithelial cell brushings. Among the genes with the most striking difference in expression in tumors relative to normal epithelial cell brushings, FGF7 showed complete loss of expression in all of both early- and late-stage serous tumors and in six of the seven ovarian tumor cell lines. Decorin, a small proteoglycan involved in the activation of EGFR (41) pathway showed lower levels of expression in 22 of 31 tumors and in five of seven cell lines. The expression of plasminogen activator inhibitor 1 (PAI1; Ref. 42), a well-characterized serine protease inhibitor, was also lost in most (25 of 36) of the tumors tested. Among the genes showing frequent up-regulation by microarray analysis, PUMP-1 (24, 43) and HE4, an epidydimis-specific whey acidic protein (40), showed over-expression in almost all of the tumors tested (data not shown).

CGH Analysis.

To investigate whether any of the results from the gene expression analysis could be associated with genomic alterations, we performed CGH on a set of 29 ovarian tumors. The overall trend suggested similar changes in both early- and late-stage tumors (Table 5; Fig. 7). There were, however, conspicuous differences between the two tumor groups. For instance, the CGH results indicated that losses are more common than gains in the early-stage tumors. Consistent losses were observed at chromosomal regions of 2q (5 of 12), 4q (8 of 12), 5q (8 of 12), 6q (6 of 12), 13q (6 of 12), 18q (3 of 14), and Xq (5 of 12). Additional losses were also observed on chromosomes 7, 9, and 15 (2 cases); and 3, 12, 17, and 21 (1 case each). Gains involving chromosomes 20 (7 of 12); 1 (6 of 12); 3 and 8 (5 of 14); 19 (4 of 12); 11, 12, 16, 17, and 22 (3 of 14); 2, 5, 6, and 18 (1 of 12) were also observed. The chromosomal arms that showed frequent loss in the late-stage tumors were nearly identical to those observed in the early-stage tumors, and included 2q (5 of 17), 4q (9 of 17), 5q (9 of 17), 6q and 13q (5 of 17), 18q (6 of 13), and Xq (10 of 17). Less frequent losses were seen on chromosomes 9 (3 cases); 1p (2 cases); and 3, 7, 8, 11, 14, and 15 (1 case each). Gains on chromosomes 8 (76%), 20 (70%), 1 and 3 (58%), 12 (52%), 11 (41%), and 2 (35%) were evident in at least six of the late-stage tumors. Less frequent gains were observed on chromosomes 19 and 17 (4 of 17); 5 and 7 (3 of 17); 9, 16, and 22 (2 of 13); and 6 and 10 (1 case each).

Amplifications were mainly confined to the late-stage ovarian tumors. These involved regions at 8q23–24 (8 cases); 20q12–13 (6 cases); 3q24–26, 11q12–13, and 12p11–13 (5 cases each); 2p24–25 (3cases); 1p36, 9p21–23, 11q24–25, and 17q24–25 (2 cases each); and 1 case each on 1q32–44, 2q23–24, 3q13, 5p14–15, 7p21, 7q35, 8q11–21, and 22q12. In contrast, amplifications in the early-stage cancers were observed in just six instances: three times at chromosome 3q24–26 and one case each on 3q23, 8q23–24, and 11q24–25.

This report represents the first communication of an investigation involving the genome-wide examination of changes in gene expression and chromosomal regions for early- and late-stage ovarian cancer. In our cDNA microarray analysis, we identified several differentially expressed genes the potential role of which in carcinogenesis have been described previously. However, this analysis also identified several genes, both of known and unknown function, the role of which in tumor development has yet to be elucidated.

We validated the differential expression of several of the genes identified through microarray analysis by semiquantitative RT-PCR of RNA from both ovarian tumor cell lines and primary tumors. The down-regulation (at least 5-fold) of genes such as hydroxy-δ-5 steroid dehydrogenase, 3β (HSD3B1; Ref. 44), steroidogenic acute regulatory protein (StAR; Ref. 45, 46), amphiregulin (47), glutathione S-transferase A3 (48), paternally expressed gene 3 (49), and integral membrane protein 2A (50) have not previously been associated with ovarian cancer. Other genes such as decorin (41, 51), platelet-derived growth factor receptor α (52), cadherin 11 (21), cyclin D2 (53), E2F-related transcription factor DP1 (33), NOEY2 (6), and secreted frizzled related protein (SFRP; Ref. 54), have functions that can be or have been linked with carcinogenesis. Our expression analysis further revealed the consistent up-regulation of several genes, including HE4 (40), matrilysin (MMP 7), ceruloplasmin ferroxidase (38), claudin 4 (39), cyclin D1 (55), mucin 1 (56) protease serine 8 (57), keratins 7 (58) and 8 (28), as well as several ESTs. We have also validated the expression levels of HE4, MMP7, ITM2A, HSD3B1, and PEG3 by real time RT-PCR in a smaller panel of primary tumors, and the results were similar to the semiquantitative RT-PCR analysis (data not shown).

To address the issue of whether the changes that we observed at the expression level were present at the genomic level, CGH analysis was performed on 12 early- and 17 late-stage tumors. Chromosomes 2, 4, 5, 6, 13, and 18 showed regions of loss in both stages at a similar frequency (59, 60, 61). Previous genomic sequence copy number screenings of ovarian tumors by conventional and molecular cytogenetics have reported similar findings (62). High copy number gains or amplifications were mainly observed in the late-stage tumors (62). Chromosomal regions involved in amplification were 3q24–26 (63), 8q23–24 (64), 11q12–13 (55), 12p11–12 (65), 17q24–25 (66), and 20q12–13 (67).

Inspection of the data generated by CGH and expression profiling allows one to speculate on a mechanistic basis for the differential expression of at least some of the genes. For example, CGH results indicated high-level gains of 8q24 and 11q13 in multiple tumors (Fig. 8), and the microarray analysis of the same tumors indicated an increase in expression of genes/ESTs from these same regions: i.e., cyclin D1 (PRAD1), uncoupling protein 2, folate receptor, and matrix metalloproteinase 7, from 11q13; and MYCC (68), PTK2 protein tyrosine kinase 2 (69), RAD21 homologue (70), and several ESTs from 8q24 (data not shown), respectively. From the combined results, it seems reasonable to consider that the overexpression of these genes may be the result of gene amplification. Similarly, loss of genomic sequences from chromosomal region 4q could be responsible for the low expression of LIM (71), Hevin (72), MAP kinase 10 and several ESTs detected by the expression profiling from the same region.

The data from the microarray analysis revealed that the majority of genes that were differentially expressed in ovarian cancer showed aberrant transcript levels in both early- and late-stage tumors. At initial consideration, such results seem contrary to the genetic dogma associated with solid-tumor evolution that holds that gene alterations, and presumably, therefore, gene expression changes, become more prevalent as tumors become increasingly malignant. Interestingly, the CGH data indicate some differences between early- and late-stage tumors, especially the more common finding of regional gain and/or amplification in the late-stage tumors. Consequently, the CGH data provide some support for the genetic evolution of this solid tumor.

The inconsistency observed between the expression profiling and the CGH data could potentially be attributable to epigenetic changes such as methylation. Inactivation of genes attributable to CpG island methylation is very well documented (73). Evidence seems to indicate that these changes are probably very early events in carcinogenesis (74, 75). This may partially explain why the majority of genes are aberrantly regulated in early-stage tumors. It is also well documented that amplification events in carcinogenesis are late events (76, 77, 78) and is consistent with what is observed from the CGH data. Abnormal expression of genes is not attributable only to gene amplification or gains. Other genomic changes like rearrangements (translocations) are also responsible for alteration in gene expression. CGH can only aid in detecting net chromosomal losses or gains in tumor cells and does not account for rearrangements, thus overlooking changes that could be responsible for abnormal expression detected by microarray analysis.

An aspect of our finding that may be especially important to the clinical treatment of ovarian cancer is the determination of extensive, aberrant gene expression in early-stage tumors. One of the assumptions of most screening strategies is that a significant fraction of stage III/IV disease metastasizes from stage I lesions. These data are consistent with that possibility. In addition, poorly differentiated stage I cancers have been treated with cytotoxic chemotherapy. These data suggest that early-stage lesions have most of the genetic changes required for metastatic spread, consistent with a need for aggressive therapy. Further study of the genes that show consistent, high-level expression changes may result in the identification of markers useful for early cancer detection and may further point to potential therapeutic targets.

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

      
1

Supported by NIH Grant CA48031 (to D. I. S.) and Department of Defense Grant DAMD 17-99-1-9504 (to D. I. S. and V. S.), and by the Mayo Foundation.

            
3

The abbreviations used are: CGH, comparative genomic hybridization; OSE, ovarian epithelial cell brushing; EST, expressed sequence tag; RT-PCR, reverse transcription-PCR; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Fig. 1.
Fig. 1. Representative scatter plots of cDNA microarray analysis. A, duplicate filters of OV 106 hybridization (Pearson correlation, 0.97). B, scatter plot of an early-stage tumor (OV106) versus one of the pooled normal epithelial brushings (Pearson correlation coefficient, 0.76). The lines, 2-fold differences in expression levels between samples.
View largeDownload slide

Representative scatter plots of cDNA microarray analysis. A, duplicate filters of OV 106 hybridization (Pearson correlation, 0.97). B, scatter plot of an early-stage tumor (OV106) versus one of the pooled normal epithelial brushings (Pearson correlation coefficient, 0.76). The lines, 2-fold differences in expression levels between samples.

Fig. 1.
Fig. 1. Representative scatter plots of cDNA microarray analysis. A, duplicate filters of OV 106 hybridization (Pearson correlation, 0.97). B, scatter plot of an early-stage tumor (OV106) versus one of the pooled normal epithelial brushings (Pearson correlation coefficient, 0.76). The lines, 2-fold differences in expression levels between samples.
View largeDownload slide

Representative scatter plots of cDNA microarray analysis. A, duplicate filters of OV 106 hybridization (Pearson correlation, 0.97). B, scatter plot of an early-stage tumor (OV106) versus one of the pooled normal epithelial brushings (Pearson correlation coefficient, 0.76). The lines, 2-fold differences in expression levels between samples.

Close modal
Fig. 2.
Fig. 2. Hierarchical clustering of 19 ovarian epithelial samples, using Eisen’s Stanford clustering package. Eighteen thousand expressed sequences were profiled against five normal ovarian epithelial (NOE), two early-stage clear cell, one early-stage endometrioid, four early-stage serous, and seven late-stage (stage-III) serous samples. The 50 annotated clones with maximal normalized expression variance (SD divided by mean of expression) are shown. Data were log-transformed after addition of a baseline noise compensation term and were clustered using centered Pearson correlation. Whereas the normal samples neatly cluster together, separation between early- and late-stage samples is less compelling.
View largeDownload slide

Hierarchical clustering of 19 ovarian epithelial samples, using Eisen’s Stanford clustering package. Eighteen thousand expressed sequences were profiled against five normal ovarian epithelial (NOE), two early-stage clear cell, one early-stage endometrioid, four early-stage serous, and seven late-stage (stage-III) serous samples. The 50 annotated clones with maximal normalized expression variance (SD divided by mean of expression) are shown. Data were log-transformed after addition of a baseline noise compensation term and were clustered using centered Pearson correlation. Whereas the normal samples neatly cluster together, separation between early- and late-stage samples is less compelling.

Fig. 2.
Fig. 2. Hierarchical clustering of 19 ovarian epithelial samples, using Eisen’s Stanford clustering package. Eighteen thousand expressed sequences were profiled against five normal ovarian epithelial (NOE), two early-stage clear cell, one early-stage endometrioid, four early-stage serous, and seven late-stage (stage-III) serous samples. The 50 annotated clones with maximal normalized expression variance (SD divided by mean of expression) are shown. Data were log-transformed after addition of a baseline noise compensation term and were clustered using centered Pearson correlation. Whereas the normal samples neatly cluster together, separation between early- and late-stage samples is less compelling.
View largeDownload slide

Hierarchical clustering of 19 ovarian epithelial samples, using Eisen’s Stanford clustering package. Eighteen thousand expressed sequences were profiled against five normal ovarian epithelial (NOE), two early-stage clear cell, one early-stage endometrioid, four early-stage serous, and seven late-stage (stage-III) serous samples. The 50 annotated clones with maximal normalized expression variance (SD divided by mean of expression) are shown. Data were log-transformed after addition of a baseline noise compensation term and were clustered using centered Pearson correlation. Whereas the normal samples neatly cluster together, separation between early- and late-stage samples is less compelling.

Close modal
Fig. 3.
Fig. 3. Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more up-regulated genes in early- and late-stage tumors. Red, fold up-regulation of five times or more. Green boxes, fold regulation between 1 and 2. Different shades of green, values above 2 but less than 3.
View largeDownload slide

Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more up-regulated genes in early- and late-stage tumors. Red, fold up-regulation of five times or more. Green boxes, fold regulation between 1 and 2. Different shades of green, values above 2 but less than 3.

Fig. 3.
Fig. 3. Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more up-regulated genes in early- and late-stage tumors. Red, fold up-regulation of five times or more. Green boxes, fold regulation between 1 and 2. Different shades of green, values above 2 but less than 3.
View largeDownload slide

Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more up-regulated genes in early- and late-stage tumors. Red, fold up-regulation of five times or more. Green boxes, fold regulation between 1 and 2. Different shades of green, values above 2 but less than 3.

Close modal
Fig. 4.
Fig. 4. Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more down-regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold regulation between 1 and 2. Different shades of red, values above 2 but less than 3.
View largeDownload slide

Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more down-regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold regulation between 1 and 2. Different shades of red, values above 2 but less than 3.

Fig. 4.
Fig. 4. Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more down-regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold regulation between 1 and 2. Different shades of red, values above 2 but less than 3.
View largeDownload slide

Hierarchical clustering of 14 ovarian tumor samples, using Eisen’s Stanford clustering package. Data were log-transformed and clustered by centered Pearson correlation. Cluster tree analysis of five times or more down-regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold regulation between 1 and 2. Different shades of red, values above 2 but less than 3.

Close modal
Fig. 5.
Fig. 5. Cluster tree analysis of gene categories. five times or more differentially regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold up-regulation. A, intermediate filament genes; B, cell-cell interactions; C, cell invasion and metastasis; and D, cell cycle and growth regulators.
View largeDownload slide

Cluster tree analysis of gene categories. five times or more differentially regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold up-regulation. A, intermediate filament genes; B, cell-cell interactions; C, cell invasion and metastasis; and D, cell cycle and growth regulators.

Fig. 5.
Fig. 5. Cluster tree analysis of gene categories. five times or more differentially regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold up-regulation. A, intermediate filament genes; B, cell-cell interactions; C, cell invasion and metastasis; and D, cell cycle and growth regulators.
View largeDownload slide

Cluster tree analysis of gene categories. five times or more differentially regulated genes in early- and late-stage tumors. Green boxes, fold down-regulation of five times or more. Red boxes, fold up-regulation. A, intermediate filament genes; B, cell-cell interactions; C, cell invasion and metastasis; and D, cell cycle and growth regulators.

Close modal
Fig. 6.
Fig. 6. A, agarose gel showing the products of semiquantitative RT-PCR in the ovarian cell lines. M, 100-bp ladder; Lane 1, short-term cultures of normal ovarian epithelial cells (OSE54); Lane 2, OV 167; Lane 3, OV 177; Lane 4, OV 202; Lane 5, OV 207; Lane 6, OV 266; Lane 7, OVCAR 5; Lane 8, SKOV; Lane 9, water control (H2O). Probes: panel 1, growth arrest-specific gene 1 (GAS1); panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, steroidogenic acute regulatory protein (STAR); panel 4, amphiregulin (AREG); panel 5, fibroblast growth factor 7 (FGF 7); panel 6, human progesterone binding protein (HPR6); panel 7, integral membrane protein 2A (ITM2A), panel 8, decorin; panel 9, GAPDH. B, agarose gel showing the products of the result of semiquantitative RT-PCR resolved on a 1.6% agarose gel. On top of the figure, sample numbers; on top of the tumor numbers, the staging information for these tumors. M, 100-bp ladder; Lane 1, normal epithelial cell brushings (B). Panel 1, decorin; panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, integral membrane protein 2A (ITM2A); panel 4, fibroblast growth factor 7 (FGF 7) and GAPDH.
View largeDownload slide

A, agarose gel showing the products of semiquantitative RT-PCR in the ovarian cell lines. M, 100-bp ladder; Lane 1, short-term cultures of normal ovarian epithelial cells (OSE54); Lane 2, OV 167; Lane 3, OV 177; Lane 4, OV 202; Lane 5, OV 207; Lane 6, OV 266; Lane 7, OVCAR 5; Lane 8, SKOV; Lane 9, water control (H2O). Probes: panel 1, growth arrest-specific gene 1 (GAS1); panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, steroidogenic acute regulatory protein (STAR); panel 4, amphiregulin (AREG); panel 5, fibroblast growth factor 7 (FGF 7); panel 6, human progesterone binding protein (HPR6); panel 7, integral membrane protein 2A (ITM2A), panel 8, decorin; panel 9, GAPDH. B, agarose gel showing the products of the result of semiquantitative RT-PCR resolved on a 1.6% agarose gel. On top of the figure, sample numbers; on top of the tumor numbers, the staging information for these tumors. M, 100-bp ladder; Lane 1, normal epithelial cell brushings (B). Panel 1, decorin; panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, integral membrane protein 2A (ITM2A); panel 4, fibroblast growth factor 7 (FGF 7) and GAPDH.

Fig. 6.
Fig. 6. A, agarose gel showing the products of semiquantitative RT-PCR in the ovarian cell lines. M, 100-bp ladder; Lane 1, short-term cultures of normal ovarian epithelial cells (OSE54); Lane 2, OV 167; Lane 3, OV 177; Lane 4, OV 202; Lane 5, OV 207; Lane 6, OV 266; Lane 7, OVCAR 5; Lane 8, SKOV; Lane 9, water control (H2O). Probes: panel 1, growth arrest-specific gene 1 (GAS1); panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, steroidogenic acute regulatory protein (STAR); panel 4, amphiregulin (AREG); panel 5, fibroblast growth factor 7 (FGF 7); panel 6, human progesterone binding protein (HPR6); panel 7, integral membrane protein 2A (ITM2A), panel 8, decorin; panel 9, GAPDH. B, agarose gel showing the products of the result of semiquantitative RT-PCR resolved on a 1.6% agarose gel. On top of the figure, sample numbers; on top of the tumor numbers, the staging information for these tumors. M, 100-bp ladder; Lane 1, normal epithelial cell brushings (B). Panel 1, decorin; panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, integral membrane protein 2A (ITM2A); panel 4, fibroblast growth factor 7 (FGF 7) and GAPDH.
View largeDownload slide

A, agarose gel showing the products of semiquantitative RT-PCR in the ovarian cell lines. M, 100-bp ladder; Lane 1, short-term cultures of normal ovarian epithelial cells (OSE54); Lane 2, OV 167; Lane 3, OV 177; Lane 4, OV 202; Lane 5, OV 207; Lane 6, OV 266; Lane 7, OVCAR 5; Lane 8, SKOV; Lane 9, water control (H2O). Probes: panel 1, growth arrest-specific gene 1 (GAS1); panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, steroidogenic acute regulatory protein (STAR); panel 4, amphiregulin (AREG); panel 5, fibroblast growth factor 7 (FGF 7); panel 6, human progesterone binding protein (HPR6); panel 7, integral membrane protein 2A (ITM2A), panel 8, decorin; panel 9, GAPDH. B, agarose gel showing the products of the result of semiquantitative RT-PCR resolved on a 1.6% agarose gel. On top of the figure, sample numbers; on top of the tumor numbers, the staging information for these tumors. M, 100-bp ladder; Lane 1, normal epithelial cell brushings (B). Panel 1, decorin; panel 2, plasminogen activator inhibitor 1 (PAI1); panel 3, integral membrane protein 2A (ITM2A); panel 4, fibroblast growth factor 7 (FGF 7) and GAPDH.

Close modal
Fig. 7.
Fig. 7. Ideograms of the results of CGH analysis of 29 ovarian tumors. Vertical lines to the right of schematic chromosomes, gains; vertical lines to the left, losses. Thicker line to the right, the presence of high copy gain/amplification. A, CGH analysis of early-stage (I/II) tumors. Copy number changes of 12 tumors indicate a pattern of losses. In particular, loss of chromosomes 2, 4, 5, 6, 13, and 18 are frequent. Gains were less frequent, mainly involving chromosomes 1, 3, 6, 8 11, 19, and 20. B, CGH analysis of late-stage (III/IV) tumors. Copy number changes of 17 tumors indicate a pattern of losses and gene amplification; thicker lines, high level gains/amplifications (∼>5 copy number). In contrast to the early-stage ovarian carcinomas, there are more chromosomal gains, and more tumors with localized gene amplification, predominantly involving the 3q24–26, 8q23–24, 11q12–13, 12p12–13, and 20q12–13 region. Other amplifications were detected on chromosomal arms 1p, 1q, 2p, 2q, 5p, 7q, 9p, 11q, 16p, 17q, and 18p. Losses were common on chromosomes 2, 4, 5,6, 13, and 18.
View largeDownload slide

Ideograms of the results of CGH analysis of 29 ovarian tumors. Vertical lines to the right of schematic chromosomes, gains; vertical lines to the left, losses. Thicker line to the right, the presence of high copy gain/amplification. A, CGH analysis of early-stage (I/II) tumors. Copy number changes of 12 tumors indicate a pattern of losses. In particular, loss of chromosomes 2, 4, 5, 6, 13, and 18 are frequent. Gains were less frequent, mainly involving chromosomes 1, 3, 6, 8 11, 19, and 20. B, CGH analysis of late-stage (III/IV) tumors. Copy number changes of 17 tumors indicate a pattern of losses and gene amplification; thicker lines, high level gains/amplifications (∼>5 copy number). In contrast to the early-stage ovarian carcinomas, there are more chromosomal gains, and more tumors with localized gene amplification, predominantly involving the 3q24–26, 8q23–24, 11q12–13, 12p12–13, and 20q12–13 region. Other amplifications were detected on chromosomal arms 1p, 1q, 2p, 2q, 5p, 7q, 9p, 11q, 16p, 17q, and 18p. Losses were common on chromosomes 2, 4, 5,6, 13, and 18.

Fig. 7.
Fig. 7. Ideograms of the results of CGH analysis of 29 ovarian tumors. Vertical lines to the right of schematic chromosomes, gains; vertical lines to the left, losses. Thicker line to the right, the presence of high copy gain/amplification. A, CGH analysis of early-stage (I/II) tumors. Copy number changes of 12 tumors indicate a pattern of losses. In particular, loss of chromosomes 2, 4, 5, 6, 13, and 18 are frequent. Gains were less frequent, mainly involving chromosomes 1, 3, 6, 8 11, 19, and 20. B, CGH analysis of late-stage (III/IV) tumors. Copy number changes of 17 tumors indicate a pattern of losses and gene amplification; thicker lines, high level gains/amplifications (∼>5 copy number). In contrast to the early-stage ovarian carcinomas, there are more chromosomal gains, and more tumors with localized gene amplification, predominantly involving the 3q24–26, 8q23–24, 11q12–13, 12p12–13, and 20q12–13 region. Other amplifications were detected on chromosomal arms 1p, 1q, 2p, 2q, 5p, 7q, 9p, 11q, 16p, 17q, and 18p. Losses were common on chromosomes 2, 4, 5,6, 13, and 18.
View largeDownload slide

Ideograms of the results of CGH analysis of 29 ovarian tumors. Vertical lines to the right of schematic chromosomes, gains; vertical lines to the left, losses. Thicker line to the right, the presence of high copy gain/amplification. A, CGH analysis of early-stage (I/II) tumors. Copy number changes of 12 tumors indicate a pattern of losses. In particular, loss of chromosomes 2, 4, 5, 6, 13, and 18 are frequent. Gains were less frequent, mainly involving chromosomes 1, 3, 6, 8 11, 19, and 20. B, CGH analysis of late-stage (III/IV) tumors. Copy number changes of 17 tumors indicate a pattern of losses and gene amplification; thicker lines, high level gains/amplifications (∼>5 copy number). In contrast to the early-stage ovarian carcinomas, there are more chromosomal gains, and more tumors with localized gene amplification, predominantly involving the 3q24–26, 8q23–24, 11q12–13, 12p12–13, and 20q12–13 region. Other amplifications were detected on chromosomal arms 1p, 1q, 2p, 2q, 5p, 7q, 9p, 11q, 16p, 17q, and 18p. Losses were common on chromosomes 2, 4, 5,6, 13, and 18.

Close modal
Fig. 8.
Fig. 8. CGH profile of an ovarian tumor illustrating a high copy gain/amplification in the 11q12–13 region of chromosome 11. The transcription profile of the same tumor by cDNA microarray analysis exhibited overexpression of cyclin D1, uncoupling protein 2, folate receptor, metalloproteinase 7, and some ESTs from this region.
View largeDownload slide

CGH profile of an ovarian tumor illustrating a high copy gain/amplification in the 11q12–13 region of chromosome 11. The transcription profile of the same tumor by cDNA microarray analysis exhibited overexpression of cyclin D1, uncoupling protein 2, folate receptor, metalloproteinase 7, and some ESTs from this region.

Fig. 8.
Fig. 8. CGH profile of an ovarian tumor illustrating a high copy gain/amplification in the 11q12–13 region of chromosome 11. The transcription profile of the same tumor by cDNA microarray analysis exhibited overexpression of cyclin D1, uncoupling protein 2, folate receptor, metalloproteinase 7, and some ESTs from this region.
View largeDownload slide

CGH profile of an ovarian tumor illustrating a high copy gain/amplification in the 11q12–13 region of chromosome 11. The transcription profile of the same tumor by cDNA microarray analysis exhibited overexpression of cyclin D1, uncoupling protein 2, folate receptor, metalloproteinase 7, and some ESTs from this region.

Close modal
Table 1

Tumor cohorta

HistologyStageGradecDNA microarrayRT-PCR
Cl cell OV 106 I 3 +b + 
Cl cell OV 267 I 3  + 
Cl cell OV 496 I 3 + + 
Endo OV 51 I 3  + 
Endo OV 78 I 3  + 
Endo OV 88 I 3  + 
Endo OV 105 I 3  + 
Endo OV 338 I 3  + 
Endo OV 647 I 3 + + 
Serous OV 6 I 3  + 
Serous OV 17 I 3 + + 
Serous OV 20 I 3  + 
Serous OV 90 I 3  + 
Serous OV 234 I 3  + 
Serous OV 363 I 3  + 
Serous OV 526 I 3 + + 
Cl cell OV 102 II 3  + 
Endo OV 296 II 3  + 
Serous OV 149 II 3  + 
Serous OV 354 II 3  + 
Serous OV 401 II 3 + + 
Serous OV 402 II 3 + + 
Serous OV 414 II 3  + 
Cl cell OV 176 III 3  + 
Endo OV 93 III 3  + 
Endo OV 110 III 3  + 
Endo OV 259 III 3  + 
Serous OV 4 III 3 + + 
Serous OV 11 III 3 + + 
Serous OV 13 III 3 + + 
Serous OV 16 III 3 + + 
Serous OV 29 III 3 + + 
Serous OV 150 III 3 + + 
Serous OV 167 III 3  + 
Serous OV 206 III 3  + 
Serous OV 208 III 3  + 
Serous OV 461 III 3  + 
Serous OV 472 III 3 + + 
Serous OV 97 IV 3  + 
HistologyStageGradecDNA microarrayRT-PCR
Cl cell OV 106 I 3 +b + 
Cl cell OV 267 I 3  + 
Cl cell OV 496 I 3 + + 
Endo OV 51 I 3  + 
Endo OV 78 I 3  + 
Endo OV 88 I 3  + 
Endo OV 105 I 3  + 
Endo OV 338 I 3  + 
Endo OV 647 I 3 + + 
Serous OV 6 I 3  + 
Serous OV 17 I 3 + + 
Serous OV 20 I 3  + 
Serous OV 90 I 3  + 
Serous OV 234 I 3  + 
Serous OV 363 I 3  + 
Serous OV 526 I 3 + + 
Cl cell OV 102 II 3  + 
Endo OV 296 II 3  + 
Serous OV 149 II 3  + 
Serous OV 354 II 3  + 
Serous OV 401 II 3 + + 
Serous OV 402 II 3 + + 
Serous OV 414 II 3  + 
Cl cell OV 176 III 3  + 
Endo OV 93 III 3  + 
Endo OV 110 III 3  + 
Endo OV 259 III 3  + 
Serous OV 4 III 3 + + 
Serous OV 11 III 3 + + 
Serous OV 13 III 3 + + 
Serous OV 16 III 3 + + 
Serous OV 29 III 3 + + 
Serous OV 150 III 3 + + 
Serous OV 167 III 3  + 
Serous OV 206 III 3  + 
Serous OV 208 III 3  + 
Serous OV 461 III 3  + 
Serous OV 472 III 3 + + 
Serous OV 97 IV 3  + 
a

Cl, clear; OV, ovarian; Endo, endometrioid.

b

+, the tumors used in each analysis.

View Large
Table 2

Differential gene expression in ovarian tumors: number of sequences with expression changes >2-fold, 5-fold, 10-fold, and 20-fold

No. of tumorsDown-regulatedUp-regulated
2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold
14 148 24 9 2 11 1 0 0 
13 262 48 16 6 30 6 1 0 
12 409 56 19 6 43 9 2 1 
11 565 68 21 7 67 11 5 1 
10 719 87 24 8 108 12 5 1 
9 945 108 27 11 168 15 6 1 
8 1261 135 34 12 265 25 8 1 
7 1750 179 46 12 435 36 12 3 
6 2346 248 56 18 794 51 18 5 
5 3115 409 115 42 1417 72 27 7 
4 4159 565 144 52 2510 116 32 12 
3 5718 781 187 65 4303 198 51 25 
2 8245 1159 279 81 7051 388 105 42 
1 12106 2940 1044 414 11178 1552 480 183 
No. of tumorsDown-regulatedUp-regulated
2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold
14 148 24 9 2 11 1 0 0 
13 262 48 16 6 30 6 1 0 
12 409 56 19 6 43 9 2 1 
11 565 68 21 7 67 11 5 1 
10 719 87 24 8 108 12 5 1 
9 945 108 27 11 168 15 6 1 
8 1261 135 34 12 265 25 8 1 
7 1750 179 46 12 435 36 12 3 
6 2346 248 56 18 794 51 18 5 
5 3115 409 115 42 1417 72 27 7 
4 4159 565 144 52 2510 116 32 12 
3 5718 781 187 65 4303 198 51 25 
2 8245 1159 279 81 7051 388 105 42 
1 12106 2940 1044 414 11178 1552 480 183 
View Large
Table 3

Differential gene expression as a function of tumor stage

No. of tumorsStage I/II downStage III/IV downStage I/II upStage III/IV up
2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold
7 200 35 11 2 406 56 18 6 19 2 0 0 48 10 3 0 
6 413 62 19 6 739 91 22 9 64 8 2 1 99 12 4 1 
5 759 86 26 8 1393 154 36 11 185 19 7 1 239 16 5 1 
4 1322 152 38 13 2360 321 99 42 498 39 15 3 714 27 8 2 
3 2500 275 65 16 3365 508 127 47 1311 92 36 9 1615 65 17 8 
No. of tumorsStage I/II downStage III/IV downStage I/II upStage III/IV up
2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold2-fold5-fold10-fold20-fold
7 200 35 11 2 406 56 18 6 19 2 0 0 48 10 3 0 
6 413 62 19 6 739 91 22 9 64 8 2 1 99 12 4 1 
5 759 86 26 8 1393 154 36 11 185 19 7 1 239 16 5 1 
4 1322 152 38 13 2360 321 99 42 498 39 15 3 714 27 8 2 
3 2500 275 65 16 3365 508 127 47 1311 92 36 9 1615 65 17 8 
View Large
Table 4

Genes with expression level changes >10-fold in both early- and late-stage tumors

Cluster identificationCluster title
10-fold down-regulated genes in early- and late-stage tumors  
 Hs.a38586 Hydroxy-δ-5-steroid dehydrogenase, 3 β- and steroid δ-isomerase 1 
 Hs.213499 ESTs 
 Hs.49075 Human DNA sequence from clone 22D12 on chromosome Xq21.1-21.33 
 Hs.188543 ESTs 
 Hs.129879 ESTs 
 Hs.79946 Cytochrome P-450, subfamily XIX (aromatization of androgens) 
 Hs.180616 CD36 antigen (collagen type I receptor, thrombospondin receptor)-like 1 
 Hs.159867 ESTs 
 Hs.1275 Amphiregulin (schwannoma-derived growth factor) 
 Hs.41585 ESTs 
 Hs.94107 ESTs 
 Hs.3132 Steroidogenic acute regulatory protein 
 Hs.105033 ESTs 
 Hs.109333 ESTs 
 Hs.117848 ESTs 
 Hs.102488 Glutathione S-transferase A3 
 Hs.54751 ESTs 
 Hs.17109 Integral membrane protein 2A 
 Hs.21858 CAG repeat containing (glia-derived nexin I α) 
 Hs.54751 ESTs 
 Hs.9914 Follistatin 
 Hs.139033 Paternally expressed gene 3 
 Hs.105806 Granulysin 
 Hs.169228 δ (Drosophila)-like 1 
 Hs.37012 Homo sapiens DNA sequence from PAC 434O14 on chromosome 1q32.3-41 
 Hs.148493 Cathepsin B 
10-fold up-regulated genes in early and late stage tumors  
 Hs.5372 Claudin 4 
 Hs.2719 Epididymis-specific, whey-acidic protein type, HE 4 
 Hs.16696 E74-like factor 3 (ets domain transcription factor) 
 Hs.19222 Ecotropic viral integration site 1 
 Hs.896603 Mucin 1, transmembrane 
 Hs.24743 ESTs 
 Hs.111461 Ceruloplasmin (ferroxidase) 
 Hs.206297 ESTs 
 Hs.104472 ESTs 
 Hs.23291 ESTs 
 Hs.73149 Paired box gene 8 
 Hs.2256 Matrix metalloproteinase 7 (matrilysin, uterine) 
 Hs.204238 Lipocalin 2 (oncogene 24p3) 
 Hs.76550 ESTs 
 Hs.36451 Glutamate receptor, ionotropic, N-methyl d-aspartate 2C 
 Hs.155097 Carbonic anhydrase II 
Cluster identificationCluster title
10-fold down-regulated genes in early- and late-stage tumors  
 Hs.a38586 Hydroxy-δ-5-steroid dehydrogenase, 3 β- and steroid δ-isomerase 1 
 Hs.213499 ESTs 
 Hs.49075 Human DNA sequence from clone 22D12 on chromosome Xq21.1-21.33 
 Hs.188543 ESTs 
 Hs.129879 ESTs 
 Hs.79946 Cytochrome P-450, subfamily XIX (aromatization of androgens) 
 Hs.180616 CD36 antigen (collagen type I receptor, thrombospondin receptor)-like 1 
 Hs.159867 ESTs 
 Hs.1275 Amphiregulin (schwannoma-derived growth factor) 
 Hs.41585 ESTs 
 Hs.94107 ESTs 
 Hs.3132 Steroidogenic acute regulatory protein 
 Hs.105033 ESTs 
 Hs.109333 ESTs 
 Hs.117848 ESTs 
 Hs.102488 Glutathione S-transferase A3 
 Hs.54751 ESTs 
 Hs.17109 Integral membrane protein 2A 
 Hs.21858 CAG repeat containing (glia-derived nexin I α) 
 Hs.54751 ESTs 
 Hs.9914 Follistatin 
 Hs.139033 Paternally expressed gene 3 
 Hs.105806 Granulysin 
 Hs.169228 δ (Drosophila)-like 1 
 Hs.37012 Homo sapiens DNA sequence from PAC 434O14 on chromosome 1q32.3-41 
 Hs.148493 Cathepsin B 
10-fold up-regulated genes in early and late stage tumors  
 Hs.5372 Claudin 4 
 Hs.2719 Epididymis-specific, whey-acidic protein type, HE 4 
 Hs.16696 E74-like factor 3 (ets domain transcription factor) 
 Hs.19222 Ecotropic viral integration site 1 
 Hs.896603 Mucin 1, transmembrane 
 Hs.24743 ESTs 
 Hs.111461 Ceruloplasmin (ferroxidase) 
 Hs.206297 ESTs 
 Hs.104472 ESTs 
 Hs.23291 ESTs 
 Hs.73149 Paired box gene 8 
 Hs.2256 Matrix metalloproteinase 7 (matrilysin, uterine) 
 Hs.204238 Lipocalin 2 (oncogene 24p3) 
 Hs.76550 ESTs 
 Hs.36451 Glutamate receptor, ionotropic, N-methyl d-aspartate 2C 
 Hs.155097 Carbonic anhydrase II 
a

Hs., unigene cluster IDs.

View Large
Table 5

CGH results

Patient no. (histology; stage)CGH changes
OV106 (Clear Cell; I) −1p32-pter;+1q;+3q22-26a;−4q;−6q24-27;+8q;+12p;−13q21-qter;−18q12-21;−Xq21-23 
OV496 (Clear Cell; I) +1p32-pter;−2q21-qter;−3;−4q12-28;+5p;−5q;−6;−12q;−13q21-33;+16;+18;+19;−Xq 
OV51 (Endo; I) NAD 
OV105 (Endo; I) +1q31-42;−2q22-24;−4q13-24;−5q12-23;−6q21-23;+11q22-24;+12q23-qter;−13q21-31;−15q21-24;+16p12-13;+19;+20q;−21 
OV338 (Endo; I) +1q21-qter;+6;+8;−9p;−21q 
OV647 (Endo; I) −4q;−5q;+6q24-27;−7q21-31;+16p12-13.2;+17;−18q;+20 
OV6 (Serous; I) +1p35-qter;−1p21-31;−2q23-32;−4q21-31.3;−5q14-32;+8q;−13q13-33;+19;+20q;+22q;−Xq21-qter 
OV90 (Serous; I) −1q24-31;+3q24-qter;−4q13-25;−5q21-23;−6;+8q23-qter;−9p13-23;+12p11.2-qter;−13q21-qter;−15q21;+17q24-25;−18q12-21;+20q12-qter;+22q 
OV234 (Serous; I) −1p33-35;+2q22;−4q22-31;−5q14-21;−6q14-16;−10p;+12p12-13;+17q;+20q 
OV102 (Clear Cell; II) −1p12-31;+1q31-42;−2q34-36;+3q24-26;−17p 
OV401 (Serous; II) +3q;+3q24;+3q26;−4q;+5p;−5q;−6;−7p;+8q24;+11q23-24;−18q21-22;+20;−Xq21-qters 
OV402 (Serous; II) +1q41-44;−2q24-32;+3q24-27;−5q21-23;−6q21-23;+11q13;−13q21-31;−18q12-22;+19;+20q;−Xq 
OV176 (Clear Cell; III) +3q25-26;+8q;+17q;+20q12-13 
OV453 (Clear Cell; III) +3q24-26.3;+17q24-25 
OV522 (Clear Cell; III) +2p24-25 
OV623 (Clear Cell; III) +8q 
OV78 (Endo; III) +1q21-41;−2p14-16;+3q13.1-13.3;−4q;+5p;+8q21.1-24;+9p21-23;+11q12-13;+17q23-25;−18q21-22;+20q;−X 
OV93 (Endo; III) +1q32-44;+2p24-25;+2q23-24;+3q13.3-qter;+3q24-26;−4q;−5;−6q22-qter;−8p;+8q;+9p22-23;+12p12-13;−13q31-34;−18q12-23;+20q13 
OV110 (Endo; III) +1p36;−4q13-32;−5q13-22;+8q24;+10;+11q24-25;+12p12-13;−14q23-24;+16q21-23;+19q;+20q 
OV259 (Endo; III) −1p13-22;+1q;+2q35-37;+3q13;+3q25-27;−4q23-qter;+5p;−5q14-qter;−6q;−7p;−11;+12p12-13;−13q;+19q;+20q12-13;+22q12.2-12.3;−Xq 
OV4 (Serous III) −2q;+3q26.1-26.3;−4q22-24;−5q13-23;−6q16.3-25;+7q33-35;+8q23-24;+11q13-14;+12p11-13;−18q12.3-22;+20q12-13.2 
OV11 (Serous III) −4q;−5q13.3-23.2;+8q23-24;+11q23-24;+12p12-13;−Xq21-qter 
OV16 (Serous III) +1q;−18q22;+22;−Xq23-25 
OV29 (Serous III) +1q21-22;−2q22-24;−5q;+8q;+8q23-24;+11q12-13;−18q12-qter;+18p11.2-11.3;+20q;−Xq 
OV72 (Serous III) +1p32-36.3;+1q31-32;−2q22-34;+5p13-15;−5q;+6p21.1-pter;+7p21;+8;+8q23-24;−9p;+11q12-13;+12p12-13;−18q12-qter;+19;+20;−Xq 
OV150 (Serous III) +1q21-24;+2p24;−3p;+3q26-1-26.3;−4p;+5p14-15;−6q13-21;+8q24;−9p24;+11q13;+12p12-13;−13q21-22;16p12-13.2;−18q;+20q12-13;−Xq 
OV392 (Serous III) +3q26;−4q221;−5q12-21;−6q24-27;+7q32-qter;+8q24.1-24.3;+12p13;−13q14-22;+20q;−X 
OV448 (Serous III) +1p36;−2q31-34;−5q14-23.3;+17q24-25;+19;+20q12-13.2;−X 
OV319 (Endo; IV) −1p31;+1q;+2p24;+2q24;+3q21-26;−4q32-qter;+8q11.2-21.2;+8q23-24;−9p23;−13q31;−15q21-qter;−18q;+20q13;−X 
Patient no. (histology; stage)CGH changes
OV106 (Clear Cell; I) −1p32-pter;+1q;+3q22-26a;−4q;−6q24-27;+8q;+12p;−13q21-qter;−18q12-21;−Xq21-23 
OV496 (Clear Cell; I) +1p32-pter;−2q21-qter;−3;−4q12-28;+5p;−5q;−6;−12q;−13q21-33;+16;+18;+19;−Xq 
OV51 (Endo; I) NAD 
OV105 (Endo; I) +1q31-42;−2q22-24;−4q13-24;−5q12-23;−6q21-23;+11q22-24;+12q23-qter;−13q21-31;−15q21-24;+16p12-13;+19;+20q;−21 
OV338 (Endo; I) +1q21-qter;+6;+8;−9p;−21q 
OV647 (Endo; I) −4q;−5q;+6q24-27;−7q21-31;+16p12-13.2;+17;−18q;+20 
OV6 (Serous; I) +1p35-qter;−1p21-31;−2q23-32;−4q21-31.3;−5q14-32;+8q;−13q13-33;+19;+20q;+22q;−Xq21-qter 
OV90 (Serous; I) −1q24-31;+3q24-qter;−4q13-25;−5q21-23;−6;+8q23-qter;−9p13-23;+12p11.2-qter;−13q21-qter;−15q21;+17q24-25;−18q12-21;+20q12-qter;+22q 
OV234 (Serous; I) −1p33-35;+2q22;−4q22-31;−5q14-21;−6q14-16;−10p;+12p12-13;+17q;+20q 
OV102 (Clear Cell; II) −1p12-31;+1q31-42;−2q34-36;+3q24-26;−17p 
OV401 (Serous; II) +3q;+3q24;+3q26;−4q;+5p;−5q;−6;−7p;+8q24;+11q23-24;−18q21-22;+20;−Xq21-qters 
OV402 (Serous; II) +1q41-44;−2q24-32;+3q24-27;−5q21-23;−6q21-23;+11q13;−13q21-31;−18q12-22;+19;+20q;−Xq 
OV176 (Clear Cell; III) +3q25-26;+8q;+17q;+20q12-13 
OV453 (Clear Cell; III) +3q24-26.3;+17q24-25 
OV522 (Clear Cell; III) +2p24-25 
OV623 (Clear Cell; III) +8q 
OV78 (Endo; III) +1q21-41;−2p14-16;+3q13.1-13.3;−4q;+5p;+8q21.1-24;+9p21-23;+11q12-13;+17q23-25;−18q21-22;+20q;−X 
OV93 (Endo; III) +1q32-44;+2p24-25;+2q23-24;+3q13.3-qter;+3q24-26;−4q;−5;−6q22-qter;−8p;+8q;+9p22-23;+12p12-13;−13q31-34;−18q12-23;+20q13 
OV110 (Endo; III) +1p36;−4q13-32;−5q13-22;+8q24;+10;+11q24-25;+12p12-13;−14q23-24;+16q21-23;+19q;+20q 
OV259 (Endo; III) −1p13-22;+1q;+2q35-37;+3q13;+3q25-27;−4q23-qter;+5p;−5q14-qter;−6q;−7p;−11;+12p12-13;−13q;+19q;+20q12-13;+22q12.2-12.3;−Xq 
OV4 (Serous III) −2q;+3q26.1-26.3;−4q22-24;−5q13-23;−6q16.3-25;+7q33-35;+8q23-24;+11q13-14;+12p11-13;−18q12.3-22;+20q12-13.2 
OV11 (Serous III) −4q;−5q13.3-23.2;+8q23-24;+11q23-24;+12p12-13;−Xq21-qter 
OV16 (Serous III) +1q;−18q22;+22;−Xq23-25 
OV29 (Serous III) +1q21-22;−2q22-24;−5q;+8q;+8q23-24;+11q12-13;−18q12-qter;+18p11.2-11.3;+20q;−Xq 
OV72 (Serous III) +1p32-36.3;+1q31-32;−2q22-34;+5p13-15;−5q;+6p21.1-pter;+7p21;+8;+8q23-24;−9p;+11q12-13;+12p12-13;−18q12-qter;+19;+20;−Xq 
OV150 (Serous III) +1q21-24;+2p24;−3p;+3q26-1-26.3;−4p;+5p14-15;−6q13-21;+8q24;−9p24;+11q13;+12p12-13;−13q21-22;16p12-13.2;−18q;+20q12-13;−Xq 
OV392 (Serous III) +3q26;−4q221;−5q12-21;−6q24-27;+7q32-qter;+8q24.1-24.3;+12p13;−13q14-22;+20q;−X 
OV448 (Serous III) +1p36;−2q31-34;−5q14-23.3;+17q24-25;+19;+20q12-13.2;−X 
OV319 (Endo; IV) −1p31;+1q;+2p24;+2q24;+3q21-26;−4q32-qter;+8q11.2-21.2;+8q23-24;−9p23;−13q31;−15q21-qter;−18q;+20q13;−X 
a

Boldface denotes amplification.

View Large
1
Landis S. H., Murray T., Bolden S., Wingo P. A. Cancer statistics, 1999.
CA Cancer J. Clin.
,
49
: 
8
-31,  
1999
.
2
Friedlander M. L. Prognostic factors in ovarian cancer.
Semin. Oncol.
,
25
: 
305
-314,  
1998
.
3
Dong Y., Walsh M. D., McGuckin M. A., Cummings M. C., Gabrielli B. G., Wright G. R., Hurst T., Khoo S. K., Parsons P. G. Reduced expression of retinoblastoma gene product (pRB) and high expression of p53 are associated with poor prognosis in ovarian cancer.
Int. J. Cancer
,
74
: 
407
-415,  
1997
.
4
Gershenson D. M., Deavers M., Diaz S., Tortolero-Luna G., Miller B. E., Bast R. C., Jr., Mills G. B., Silva E. G. Prognostic significance of p53 expression in advanced-stage ovarian serous borderline tumors.
Clin. Cancer Res.
,
5
: 
4053
-4058,  
1999
.
5
Aprelikova O. N., Fang B. S., Meissner E. G., Cotter S., Campbell M., Kuthiala A., Bessho M., Jensen R. A., Liu E. T. BRCA1-associated growth arrest is RB-dependent.
Proc. Natl. Acad. Sci. USA
,
96
: 
11866
-11871,  
1999
.
6
Yu Y., Xu F., Peng H., Fang X., Zhao S., Li Y., Cuevas B., Kuo W. L., Gray J. W., Siciliano M., Mills G. B., Bast R. C., Jr. NOEY2 (ARHI), an imprinted putative tumor suppressor gene in ovarian and breast carcinomas.
Proc. Natl. Acad. Sci. USA
,
96
: 
214
-219,  
1999
.
7
Catteau A., Harris W. H., Xu C. F., Solomon E. Methylation of the BRCA1 promoter region in sporadic breast and ovarian cancer: correlation with disease characteristics.
Oncogene
,
18
: 
1957
-1965,  
1999
.
8
Scambia G., Masciullo V., Benedetti Panici P., Marone M., Ferrandina G., Todaro N., Bellacosa A., Jain S. K., Neri G., Piffanelli A., Mancuso S. Prognostic significance of ras/p21 alterations in human ovarian cancer.
Br. J. Cancer
,
75
: 
1547
-1553,  
1997
.
9
Tanner B., Hengstler J. G., Luch A., Meinert R., Kreutz E., Arand M., Wilkens C., Hofmann M., Oesch F., Knapstein P. G., Becker R. C-myc mRNA expression in epithelial ovarian carcinomas in relation to estrogen receptor status, metastatic spread, survival time. FIGO stage, and histologic grade and type.
Int. J. Gynecol. Pathol.
,
17
: 
66
-74,  
1998
.
10
Bourguignon L. Y., Zhu H., Chu A., Iida N., Zhang L., Hung M. C. Interaction between the adhesion receptor, CD44, and the oncogene product, p185HER2, promotes human ovarian tumor cell activation.
J. Biol. Chem.
,
272
: 
27913
-27918,  
1997
.
11
Wang K., Gan L., Jeffery E., Gayle M., Gown A. M., Skelly M., Nelson P. S., Ng W. V., Schummer M., Hood L., Mulligan J. Monitoring gene expression profile changes in ovarian carcinomas using cDNA microarray.
Gene (Amst.)
,
229
: 
101
-108,  
1999
.
12
Alaiya A. A., Franzen B., Hagman A., Silfversward C., Moberger B., Linder S., Auer G. Classification of human ovarian tumors using multivariate data analysis of polypeptide expression patterns.
Int. J. Cancer
,
86
: 
731
-736,  
2000
.
13
Schena M., Shalon D., Davis R. W., Brown P. O. Quantitative monitoring of gene expression patterns with a complementary DNA microarray.
Science (Wash. DC)
,
270
: 
467
-470,  
1995
.
14
Kallioniemi A., Kallioniemi O. P., Sudar D., Rutovitz D., Gray J. W., Waldman F., Pinkel D. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors.
Science (Wash. DC)
,
258
: 
818
-821,  
1992
.
15
Conover C. A., Hartmann L. C., Bradley S., Stalboerger P., Klee G. G., Kalli K. R., Jenkins R. B. Biological characterization of human epithelial ovarian carcinoma cells in primary culture: the insulin-like growth factor system.
Exp. Cell Res.
,
238
: 
439
-449,  
1998
.
16
Hamilton T. C., Young R. C., Ozols R. F. Experimental model systems of ovarian cancer: applications to the design and evaluation of new treatment approaches.
Semin. Oncol.
,
11
: 
285
-298,  
1984
.
17
Dracopoli N. C. .
Current Protocols in Human Genetics
, John Wiley & Sons, Inc. New York  
1999
.
18
Tumilowicz J. J., Nichols W. W., Cholon J. J., Greene A. E. Definition of a continuous human cell line derived from neuroblastoma.
Cancer Res.
,
30
: 
2110
-2118,  
1970
.
19
Eisen M. B., Spellman P. T., Brown P. O., Botstein D. Cluster analysis and display of genome-wide expression patterns.
Proc. Natl. Acad. Sci. USA
,
95
: 
14863
-14868,  
1998
.
20
Kremmidiotis G., Baker E., Crawford J., Eyre H. J., Nahmias J., Callen D. F. Localization of human cadherin genes to chromosome regions exhibiting cancer-related loss of heterozygosity.
Genomics
,
49
: 
467
-471,  
1998
.
21
Peralta Soler A., Knudsen K. A., Tecson-Miguel A., McBrearty F. X., Han A. C., Salazar H. Expression of E-cadherin and N-cadherin in surface epithelial-stromal tumors of the ovary distinguishes mucinous from serous and endometrioid tumors.
Hum. Pathol.
,
28
: 
734
-739,  
1997
.
22
Zedlacher M., Schmoll M., Zimmermann K., Horstkorte O., Nischt R. Differential regulation of the human nidogen gene promoter region by a novel cell-type-specific silencer element.
Biochem. J.
,
338
: 
343
-350,  
1999
.
23
Moser T. L., Young T. N., Rodriguez G. C., Pizzo S. V., Bast R. C., Jr., Stack M. S. Secretion of extracellular matrix-degrading proteinases is increased in epithelial ovarian carcinoma.
Int. J. Cancer
,
56
: 
552
-559,  
1994
.
24
Tanimoto H., Underwood L. J., Shigemasa K., Parmley T. H., Wang Y., Yan Y., Clarke J., O’Brien T. J. The matrix metalloprotease pump-1 (MMP-7. Matrilysin): a candidate marker/target for ovarian cancer detection and treatment.
Tumor Biol.
,
20
: 
88
-98,  
1999
.
25
Shibata K., Kikkawa F., Nawa A., Tamakoshi K., Suganuma N., Tomoda Y. Increased matrix metalloproteinase-9 activity in human ovarian cancer cells cultured with conditioned medium from human peritoneal tissue.
Clin. Exp. Metastasis
,
15
: 
612
-619,  
1997
.
26
Ferrari S., Battini R., Kaczmarek L., Rittling S., Calabretta B., de Riel J. K., Philiponis V., Wei J. F., Baserga R. Coding sequence and growth regulation of the human vimentin gene.
Mol. Cell. Biol.
,
6
: 
3614
-3620,  
1986
.
27
Rogaev E. I., Rogaeva E. A., Ginter E. K., Korovaitseva G. I., Farrer L., Shlenskii A. B., Prytkov A. N., St. George-Hyslop P., Mordovtsev V. N. Mapping the gene for palmoplantar hyperkeratosis (thylosis) to chromosome 17 in the 17q12–q24 region.
Genetica (Dordr.)
,
30
: 
326
-329,  
1994
.
28
Waseem A., Gough A. C., Spurr N. K., Lane E. B. Localization of the gene for human simple epithelial keratin 18 to chromosome 12 using polymerase chain reaction.
Genomics
,
7
: 
188
-194,  
1990
.
29
Waseem A., Alam Y., Dogan B., White K. N., Leigh I. M., Waseem N. H. Isolation, sequence and expression of the gene encoding human keratin 13[published erratum appears in Gene, 221: 287, 1998].
Gene
,
215
: 
269
-279,  
1998
.
30
Palmero I., Holder A., Sinclair A. J., Dickson C., Peters G. Cyclins D1 and D2 are differentially expressed in human B-lymphoid cell lines.
Oncogene
,
8
: 
1049
-1054,  
1993
.
31
Palmero I., Peters G. Perturbation of cell cycle regulators in human cancer.
Cancer Surv.
,
27
: 
351
-367,  
1996
.
32
Ruaro M. E., Stebel M., Vatta P., Marzinotto S., Schneider C. Analysis of the domain requirement in Gas1 growth suppressing activity.
FEBS Lett.
,
481
: 
159
-163,  
2000
.
33
Johnson D. G., Cress W. D., Jakoi L., Nevins J. R. Oncogenic capacity of the E2F1 gene.
Proc. Natl. Acad. Sci. USA
,
91
: 
12823
-12827,  
1994
.
34
Gudas J. M., Payton M., Thukral S., Chen E., Bass M., Robinson M. O., Coats S. Cyclin E2, a novel G1 cyclin that binds Cdk2 and is aberrantly expressed in human cancers.
Mol. Cell. Biol.
,
19
: 
612
-622,  
1999
.
35
Liu F. S., Chen J. T., Liu S. C., Shih A., Shih R. T., Ho E. S. Expression and prognostic significance of proliferating cell nuclear antigen and Ki-67 in malignant ovarian germ cell tumors.
Chung Hua I Hsueh Tsa Chih (Taipei)
,
62
: 
695
-702,  
1999
.
36
Polanowska J., Le Cam L., Orsetti B., Valles H., Fabbrizio E., Fajas L., Taviaux S., Theillet C., Sardet C. Human E2F5 gene is oncogenic in primary rodent cells and is amplified in human breast tumors.
Zhonghua Yi Xue Za Zhi
,
28
: 
126
-130,  
2000
.
37
Giuntoli R. L., II, Rodriguez G. C., Whitaker R. S., Dodge R., Voynow J. A. Mucin gene expression in ovarian cancers.
Cancer Res.
,
58
: 
5546
-5550,  
1998
.
38
Senra Varela A., Lopez Saez J. J., Quintela Senra D. Serum ceruloplasmin as a diagnostic marker of cancer.
Cancer Lett.
,
121
: 
139
-145,  
1997
.
39
Hough C. D., Sherman-Baust C. A., Pizer E. S., Montz F. J., Im D. D., Rosenshein N. B., Cho K. R., Riggins G. J., Morin P. J. Large-scale serial analysis of gene expression reveals genes differentially expressed in ovarian cancer.
Cancer Res.
,
60
: 
6281
-6287,  
2000
.
40
Schummer M., Ng W. V., Bumgarner R. E., Nelson P. S., Schummer B., Bednarski D. W., Hassell L., Baldwin R. L., Karlan B. Y., Hood L. Comparative hybridization of an array of 21,500 ovarian cDNAs for the discovery of genes overexpressed in ovarian carcinomas.
Gene (Amst.)
,
238
: 
375
-385,  
1999
.
41
Nash M. A., Loercher A. E., Freedman R. S. In vitro growth inhibition of ovarian cancer cells by decorin: synergism of action between decorin and carboplatin.
Cancer Res.
,
59
: 
6192
-6196,  
1999
.
42
Pujade-Lauraine E., Lu H., Mirshahi S., Soria J., Soria C., Bernadou A., Kruithof E. K., Lijnen H. R., Burtin P. The plasminogen-activation system in ovarian tumors.
Int. J. Cancer
,
55
: 
27
-31,  
1993
.
43
Shigemasa K., Tanimoto H., Sakata K., Nagai N., Parmley T. H., Ohama K., O’Brien T. J. Induction of matrix metalloprotease-7 is common in mucinous ovarian tumors including early stage disease.
Med. Oncol.
,
17
: 
52
-58,  
2000
.
44
Milewich L., Shaw C. E., Mason J. I., Carr B. R., Blomquist C. H., Thomas J. L. 3β-hydroxysteroid dehydrogenase activity in tissues of the human fetus determined with 5α-androstane-3β,17 β-diol and dehydroepiandrosterone as substrates.
J. Steroid Biochem. Mol. Biol.
,
45
: 
525
-537,  
1993
.
45
Lee H. K., Yoo M. S., Choi H. S., Kwon H. B., Soh J. Retinoic acids up-regulate steroidogenic acute regulatory protein gene.
Mol. Cell. Endocrinol.
,
148
: 
1
-10,  
1999
.
46
Brand C., Souchelnytskiy S., Chambaz E. M., Feige J. J., Bailly S. Smad3 is involved in the intracellular signaling pathways that mediate the inhibitory effects of transforming growth factor-β on StAR expression.
Biochem. Biophys. Res. Commun.
,
253
: 
780
-785,  
1998
.
47
Funatomi H., Itakura J., Ishiwata T., Pastan I., Thompson S. A., Johnson G. R., Korc M. Amphiregulin antisense oligonucleotide inhibits the growth of T3M4 human pancreatic cancer cells and sensitizes the cells to EGF receptor-targeted therapy.
Int. J. Cancer
,
72
: 
512
-517,  
1997
.
48
Letourneau S., Palerme J. S., Delisle J. S., Beausejour C. M., Momparler R. L., Cournoyer D. Coexpression of rat glutathione S-transferase A3 and human cytidine deaminase by a bicistronic retroviral vector confers in vitro resistance to nitrogen mustards and 1-β-d-arabinofuranosylcytosine in murine fibroblasts.
Cancer Gene Ther.
,
7
: 
757
-765,  
2000
.
49
Li L. L., Szeto I. Y., Cattanach B. M., Ishino F., Surani M. A. Organization and parent-of-origin-specific methylation of imprinted Peg3 gene on mouse proximal chromosome 7.
Genomics
,
63
: 
333
-340,  
2000
.
50
Tuckermann J. P., Pittois K., Partridge N. C., Merregaert J., Angel P. Collagenase-3 (MMP-13) and integral membrane protein 2a (Itm2a) are marker genes of chondrogenic/osteoblastic cells in bone formation: sequential temporal, and spatial expression of Itm2a, alkaline phosphatase, MMP-13, and osteocalcin in the mouse.
J. Bone Miner. Res.
,
15
: 
1257
-1265,  
2000
.
51
Iozzo R. V., Moscatello D. K., McQuillan D. J., Eichstetter I. Decorin is a biological ligand for the epidermal growth factor receptor.
J. Biol. Chem.
,
274
: 
4489
-4492,  
1999
.
52
Shawver L. K., Schwartz D. P., Mann E., Chen H., Tsai J., Chu L., Taylorson L., Longhi M., Meredith S., Germain L., Jacobs J. S., Tang C., Ullrich A., Berens M. E., Hersh E., McMahon G., Hirth K. P., Powell T. J. Inhibition of platelet-derived growth factor-mediated signal transduction and tumor growth by N-[4-(trifluoromethyl)-phenyl]5- methylisoxazole-4-carboxamide.
Clin. Cancer Res.
,
3
: 
1167
-1177,  
1997
.
53
Sicinski P., Donaher J. L., Geng Y., Parker S. B., Gardner H., Park M. Y., Robker R. L., Richards J. S., McGinnis L. K., Biggers J. D., Eppig J. J., Bronson R. T., Elledge S. J., Weinberg R. A. Cyclin D2 is an FSH-responsive gene involved in gonadal cell proliferation and oncogenesis.
Nature (Lond.)
,
384
: 
470
-474,  
1996
.
54
Roth W., Wild-Bode C., Platten M., Grimmel C., Melkonyan H. S., Dichgans J., Weller M. Secreted frizzled-related proteins inhibit motility and promote growth of human malignant glioma cells.
Oncogene
,
19
: 
4210
-4220,  
2000
.
55
Shigemasa K., Tanimoto H., Parham G. P., Parmley T. H., Ohama K., O’Brien T. J. Cyclin D1 overexpression and p53 mutation status in epithelial ovarian cancer.
J. Soc. Gynecol. Investig.
,
6
: 
102
-108,  
1999
.
56
Dong Y., Walsh M. D., Cummings M. C., Wright R. G., Khoo S. K., Parsons P. G., McGuckin M. A. Expression of MUC1 and MUC2 mucins in epithelial ovarian tumours.
J. Pathol.
,
183
: 
311
-317,  
1997
.
57
Yu J. X., Chao L., Ward D. C., Chao J. Structure and chromosomal localization of the human prostasin (PRSS8) gene.
Genomics
,
32
: 
334
-340,  
1996
.
58
Sack M. J., Roberts S. A. Cytokeratins 20 and 7 in the differential diagnosis of metastatic carcinoma in cytologic specimens.
Diagn. Cytopathol.
,
16
: 
132
-136,  
1997
.
59
Takakura S., Okamoto A., Saito M., Yasuhara T., Shinozaki H., Isonishi S., Yoshimura T., Ohtake Y., Ochiai K., Tanaka T. Allelic imbalance in chromosome band 18q21 and SMAD4 mutations in ovarian cancers.
Genes Chromosomes Cancer
,
24
: 
264
-271,  
1999
.
60
Allan G. J., Cottrell S., Trowsdale J., Foulkes W. D. Loss of heterozygosity on chromosome 5 in sporadic ovarian carcinoma is a late event and is not associated with mutations in APC at 5q21–22.
Hum. Mutat.
,
3
: 
283
-291,  
1994
.
61
Watson R. H., Roy W. J., Jr., Davis M., Hitchcock A., Campbell I. G. Loss of heterozygosity at the α-inhibin locus on chromosome 2q is not a feature of human granulosa cell tumors.
Gynecol. Oncol.
,
65
: 
387
-390,  
1997
.
62
Wasenius V. M., Jekunen A., Monni O., Joensuu H., Aebi S., Howell S. B., Knuutila S. Comparative genomic hybridization analysis of chromosomal changes occurring during development of acquired resistance to cisplatin in human ovarian carcinoma cells.
Genes Chromosomes Cancer
,
18
: 
286
-291,  
1997
.
63
Hu L., Zaloudek C., Mills G. B., Gray J., Jaffe R. B. In vivo and in vitro ovarian carcinoma growth inhibition by a phosphatidylinositol 3-kinase inhibitor (LY294002).
Clin. Cancer Res.
,
6
: 
880
-886,  
2000
.
64
Baker V. V., Borst M. P., Dixon D., Hatch K. D., Shingleton H. M., Miller D. c-myc amplification in ovarian cancer.
Gynecol. Oncol.
,
38
: 
340
-342,  
1990
.
65
Fukumoto M., Estensen R. D., Sha L., Oakley G. J., Twiggs L. B., Adcock L. L., Carson L. F., Roninson I. B. Association of Ki-ras with amplified DNA sequences, detected in human ovarian carcinomas by a modified in-gel renaturation assay.
Cancer Res.
,
49
: 
1693
-1697,  
1989
.
66
Gogusev J., de Joliniere J. B., Telvi L., Doussau M., du Manoir S., Stojkoski A., Levardon M. Genetic abnormalities detected by comparative genomic hybridization in a human endometriosis-derived cell line.
Mol. Hum. Reprod.
,
6
: 
821
-827,  
2000
.
67
Courjal F., Cuny M., Rodriguez C., Louason G., Speiser P., Katsaros D., Tanner M. M., Zeillinger R., Theillet C. DNA amplifications at 20q13 and MDM2 define distinct subsets of evolved breast and ovarian tumours.
Br. J. Cancer
,
74
: 
1984
-1989,  
1996
.
68
Schraml P., Kononen J., Bubendorf L., Moch H., Bissig H., Nocito A., Mihatsch M. J., Kallioniemi O. P., Sauter G. Tissue microarrays for gene amplification surveys in many different tumor types.
Clin. Cancer Res.
,
5
: 
1966
-1975,  
1999
.
69
Agochiya M., Brunton V. G., Owens D. W., Parkinson E. K., Paraskeva C., Keith W. N., Frame M. C. Increased dosage and amplification of the focal adhesion kinase gene in human cancer cells.
Oncogene
,
18
: 
5646
-5653,  
1999
.
70
Hoque M. T., Ishikawa F. Human chromatid cohesin component hRad21 Is phosphorylated in M phase and associated with metaphase centromeres.
J. Biol. Chem.
,
276
: 
5059
-5067,  
2001
.
71
Bouju S., Pietu G., Le Cunff M., Cros N., Malzac P., Pellissier J. F., Pons F., Leger J. J., Auffray C., Dechesne C. A. Exclusion of muscle specific actinin-associated LIM protein (ALP) gene from 4q35 facioscapulohumeral muscular dystrophy (FSHD) candidate genes.
Neuromuscul. Disord.
,
9
: 
3
-10,  
1999
.
72
Nelson P. S., Plymate S. R., Wang K., True L. D., Ware J. L., Gan L., Liu A. Y., Hood L. Hevin, an antiadhesive extracellular matrix protein, is down-regulated in metastatic prostate adenocarcinoma.
Cancer Res.
,
58
: 
232
-236,  
1998
.
73
Robertson K. D., Jones P. A. DNA methylation: past, present and future directions.
Carcinogenesis (Lond.)
,
21
: 
461
-467,  
2000
.
74
Cheng P., Schmutte C., Cofer K. F., Felix J. C., Yu M. C., Dubeau L. Alterations in DNA methylation are early, but not initial, events in ovarian tumorigenesis.
Br. J. Cancer
,
75
: 
396
-402,  
1997
.
75
Rideout W. M., III, Eversole-Cire P., Spruck C. H., III, Hustad C. M., Coetzee G. A., Gonzales F. A., Jones P. A. Progressive increases in the methylation status and heterochromatinization of the myoD CpG island during oncogenic transformation.
Mol. Cell. Biol.
,
14
: 
6143
-6152,  
1994
.
76
Calugi A., Eleuteri P., Cavallo D., Naso G., Albonici L., Lombardi M. P., Manzari V., Romanini C., DeVita R. Detection of cellular heterogeneity by DNA ploidy, 17 chromosome, and p53 gene in primary carcinoma and metastasis in a case of ovarian cancer.
Int. J. Gynecol Pathol.
,
15
: 
77
-81,  
1996
.
77
Anderson M. W., Reynolds S. H., You M., Maronpot R. M. Role of proto-oncogene activation in carcinogenesis.
Environ. Health Perspect.
,
98
: 
13
-24,  
1992
.
78
Ranzani G. N., Pellegata N. S., Previdere C., Saragoni A., Vio A., Maltoni M., Amadori D. Heterogeneous protooncogene amplification correlates with tumor progression and presence of metastases in gastric cancer patients.
Cancer Res.
,
50
: 
7811
-7814,  
1990
.
©2001 American Association for Cancer Research.
2001