ALDH1-High Ovarian Cancer Stem-Like Cells Can Be Isolated from Serous and Clear Cell Adenocarcinoma Cells,and ALDH1 High Expression Is Associated with Poor Prognosis

Cancer stem-like cells (CSCs)/cancer-initiating cells (CICs) are defined as a small population of cancer cells that have high tumorigenicity. Furthermore, CSCs/CICs are resistant to several cancer therapies, and CSCs/CICs are therefore thought to be responsible for cancer recurrence after treatment and distant metastasis. In epithelial ovarian cancer (EOC) cases, disease recurrence after chemotherapy is frequently observed, suggesting ovarian CSCs/CICs are involved. There are four major histological subtypes in EOC, and serous adenocarcinoma and clear cell adenocarcinoma are high-grade malignancies. We therefore analyzed ovarian CSCs/CICs from ovarian carcinoma cell lines (serous adenocarcinoma and clear cell adenocarcinoma) and primary ovarian cancer cells in this study. We isolated ovarian CSCs/CICs as an aldehyde dehydrogenase 1 high (ALDH1^high) population from 6 EOC cell lines (3 serous adenocarcinomas and 3 clear cell adenocarcinomas) by the ALDEFLUOR assay. ALDH1^high cells showed greater sphere-forming ability, higher tumorigenicity and greater invasive capability, indicating that ovarian CSCs/CICs are enriched in ALDH1^high cells. ALDH1^high cells could also be isolated from 8 of 11 primary ovarian carcinoma samples. Immunohistochemical staining revealed that higher ALDH1 expression levels in ovary cancer cases are related to poorer prognosis in both serous adenocarcinoma cases and clear cell adenocarcinoma cases. Taken together, the results indicate that ALDH1 is a marker for ovarian CSCs/CICs and that the expression level of ALDH1 might be a novel biomarker for prediction of poor prognosis.     

Expression and methylation status of 14-3-3 sigma gene can characterize the different histological features of ovarian cancer

We hypothesize that 14-3-3 sigma gene expression and its regulation by methylation can characterize histological types of primary human epithelial ovarian cancer. To test this hypothesis, ovarian cancer cell lines and 54 ovarian cancer tissue samples were analyzed for expression and methylation of 14-3-3 sigma gene using methylation specific PCR. The results of our experiments demonstrate that 14-3-3 sigma gene was methylated and inactivated in ES-2 ovarian cell line, which was derived from clear cell adenocarcinoma. Treatment of this cell line with demethylating agent 5-aza-2'-deoxycytidine restored the expression of 14-3-3 sigma gene. In human ovarian cancer tissues, the expression of 14-3-3 sigma protein was inactivated in most of the ovarian clear cell carcinoma tissues. Interestingly, 14-3-3 sigma protein expression was positive in significantly higher percentages of serous (89.5%), endometrioid (90%), and mucinous (81.8%) ovarian adenocarcinoma tissues. The ovarian clear cell carcinoma samples with inactivated 14-3-3 sigma protein were highly methylated, suggesting that inactivation of 14-3-3 sigma gene is through DNA methylation. Using direct DNA sequencing, 14-3-3 sigma gene methylation on all the 17 CpG sites was significantly higher in ovarian clear cell carcinoma as compared to other histological types of ovarian cancer (serous, endometrioid, and mucinous). This is the first report suggesting that 14-3-3 sigma gene expression and methylation status can characterize histological features of different types of ovarian cancer.     

Expression and function of HOXA genes in normal and neoplastic ovarian epithelial cells

We studied the roles of three HOXA genes in cultured normal ovarian surface epithelial (OSE) cells and ovarian cancer cells. They included HOXA4 and HOXA7 because, by cDNA microarray analysis, these were more highly expressed in invasive ovarian carcinomas than in benign or borderline (noninvasive) ovarian tumors, and HOXA9 because it characterizes normal oviductal epithelium, which resembles ovarian serous adenocarcinomas. The three HOXA genes were more highly expressed when OSE cells were dividing and motile than when they were confluent and stationary, and also when they dispersed in response to EGF treatment or to reduced calcium concentrations in culture media. The expression of the HOXA genes varied among ovarian cancer cell lines, but was highest in lines with compact epithelial morphologies. We focused on HOXA4 as the most highly expressed in the ovarian carcinoma array. HOXA4 expression did not parallel proliferative activities of either OSE or ovarian cancer lines. Moreover, modifying HOXA4 expression in ovarian cancer cell lines did not alter either E-cadherin expression or CA125 secretion. However, HOXA4 downregulation enhanced EGFR phosphorylation and migration in serum-starved OSE and ovarian cancer cells in response to EGF, and enhanced migration of all ovarian cancer lines in 5% serum even without EGF treatment. Thus, HOXA4 expression does not correlate with proliferation or with epithelial differentiation, but it increases in response to OSE cell dispersion and negatively regulates EGFR activation and the motility of OSE and of ovarian cancer cells. HOXA4 expression was highest in cancer lines with compact epithelial growth patterns, suggesting, again, an anti-dispersion function. In summary, increased HOXA4 expression in ovarian cancer appears to constitute a tumor-suppressive, homeostatic response to aberrant cell behavior, and, in particular, to cell dispersion and migration.     

Identification of novel therapeutic targets in microdissected clear cell ovarian cancers

Clear cell ovarian cancer is an epithelial ovarian cancer histotype that is less responsive to chemotherapy and carries poorer prognosis than serous and endometrioid histotypes. Despite this, patients with these tumors are treated in a similar fashion as all other ovarian cancers. Previous genomic analysis has suggested that clear cell cancers represent a unique tumor subtype. Here we generated the first whole genomic expression profiling using epithelial component of clear cell ovarian cancers and normal ovarian surface specimens isolated by laser capture microdissection. All the arrays were analyzed using BRB ArrayTools and PathwayStudio software to identify the signaling pathways. Identified pathways validated using serous, clear cell cancer cell lines and RNAi technology. In vivo validations carried out using an orthotopic mouse model and liposomal encapsulated siRNA. Patient-derived clear cell and serous ovarian tumors were grafted under the renal capsule of NOD-SCID mice to evaluate the therapeutic potential of the identified pathway. We identified major activated pathways in clear cells involving in hypoxic cell growth, angiogenesis, and glucose metabolism not seen in other histotypes. Knockdown of key genes in these pathways sensitized clear cell ovarian cancer cell lines to hypoxia/glucose deprivation. In vivo experiments using patient derived tumors demonstrate that clear cell tumors are exquisitely sensitive to antiangiogenesis therapy (i.e. sunitinib) compared with serous tumors. We generated a histotype specific, gene signature associated with clear cell ovarian cancer which identifies important activated pathways critical for their clinicopathologic characteristics. These results provide a rational basis for a radically different treatment for ovarian clear cell patients.     

Comparative proteomics of ovarian epithelial tumors

We analyzed 12 ovarian epithelial tumors using 2D PAGE-based comparative proteomics to construct intra- and inter-tumoral distance map trees and to discover surrogate biomarkers indicative of an ovarian tumor. The analysis was performed after laser microdissection of 12 fresh-frozen tissue samples, including 4 serous, 5 mucinous, and 3 endometrioid tumors, with correlation with their histopathological characteristics. Ovarian epithelial tumors and normal tissues showed an apparent separation on the distance map tree. Mucinous carcinomas were closest to the normal group, whereas serous carcinomas were located furthest from the normal group. All mucinous tumors with aggressive histology were separated from the low malignant potential (LMP) group. The benign-looking cysts adjacent to the intraepithelial carcinoma (IEC) showed an expression pattern identical to that of the IEC area. The extent of change on the lineages leading to the mucinous and serous carcinoma was 1.98-fold different. The overall gene expression profiles of serous or endometrioid carcinomas appeared to be less affected by grade or stage than by histologic type. The potential candidate biomarkers screened in ovarian tumors and found to be significantly up-regulated in comparison to normal tissues were as follows: NM23, annexin-1, protein phosphatase-1, ferritin light chain, proteasome alpha-6, and NAGK (N-acetyl glucosamine kinase). In conclusion, ovarian mucinous tumors are distinct from other ovarian epithelial tumors. LMP mucinous tumors showing histologically aggressive features belong to mucinous carcinoma on the proteomic basis.     

Gene expression patterns in ovarian carcinomas

We used DNA microarrays to characterize the global gene expression patterns in surface epithelial cancers of the ovary. We identified groups of genes that distinguished the clear cell subtype from other ovarian carcinomas, grade I and II from grade III serous papillary carcinomas, and ovarian from breast carcinomas. Six clear cell carcinomas were distinguished from 36 other ovarian carcinomas (predominantly serous papillary) based on their gene expression patterns. The differences may yield insights into the worse prognosis and therapeutic resistance associated with clear cell carcinomas. A comparison of the gene expression patterns in the ovarian cancers to published data of gene expression in breast cancers revealed a large number of differentially expressed genes. We identified a group of 62 genes that correctly classified all 125 breast and ovarian cancer specimens. Among the best discriminators more highly expressed in the ovarian carcinomas were PAX8 (paired box gene 8), mesothelin, and ephrin-B1 (EFNB1). Although estrogen receptor was expressed in both the ovarian and breast cancers, genes that are coregulated with the estrogen receptor in breast cancers, including GATA-3, LIV-1, and X-box binding protein 1, did not show a similar pattern of coexpression in the ovarian cancers.     

Epigenetic analysis of sporadic and Lynch-associated ovarian cancers reveals histology-specific patterns of DNA methylation

Diagnosis and treatment of epithelial ovarian cancer is challenging due to the poor understanding of the pathogenesis of the disease. Our aim was to investigate epigenetic mechanisms in ovarian tumorigenesis and, especially, whether tumors with different histological subtypes or hereditary background (Lynch syndrome) exhibit differential susceptibility to epigenetic inactivation of growth regulatory genes. Gene candidates for epigenetic regulation were identified from the literature and by expression profiling of ovarian and endometrial cancer cell lines treated with demethylating agents. Thirteen genes were chosen for methylation-specific multiplex ligation-dependent probe amplification assays on 104 (85 sporadic and 19 Lynch syndrome-associated) ovarian carcinomas. Increased methylation (i.e., hypermethylation) of variable degree was characteristic of ovarian carcinomas relative to the corresponding normal tissues, and hypermethylation was consistently more prominent in non-serous than serous tumors for individual genes and gene sets investigated. Lynch syndrome-associated clear cell carcinomas showed the highest frequencies of hypermethylation. Among endometrioid ovarian carcinomas, lower levels of promoter methylation of RSK4, SPARC, and HOXA9 were significantly associated with higher tumor grade; thus, the methylation patterns showed a shift to the direction of high-grade serous tumors. In conclusion, we provide evidence of a frequent epigenetic inactivation of RSK4, SPARC, PROM1, HOXA10, HOXA9, WT1-AS, SFRP2, SFRP5, OPCML, and MIR34B in the development of non-serous ovarian carcinomas of Lynch and sporadic origin, as compared to serous tumors. Our findings shed light on the role of epigenetic mechanisms in ovarian tumorigenesis and identify potential targets for translational applications.     

Epigenetic analysis of sporadic and Lynch-associated ovarian cancers reveals histology-specific patterns of DNA methylation

Diagnosis and treatment of epithelial ovarian cancer is challenging due to the poor understanding of the pathogenesis of the disease. Our aim was to investigate epigenetic mechanisms in ovarian tumorigenesis and, especially, whether tumors with different histological subtypes or hereditary background (Lynch syndrome) exhibit differential susceptibility to epigenetic inactivation of growth regulatory genes. Gene candidates for epigenetic regulation were identified from the literature and by expression profiling of ovarian and endometrial cancer cell lines treated with demethylating agents. Thirteen genes were chosen for methylation-specific multiplex ligation-dependent probe amplification assays on 104 (85 sporadic and 19 Lynch syndrome-associated) ovarian carcinomas. Increased methylation (i.e., hypermethylation) of variable degree was characteristic of ovarian carcinomas relative to the corresponding normal tissues, and hypermethylation was consistently more prominent in non-serous than serous tumors for individual genes and gene sets investigated. Lynch syndrome-associated clear cell carcinomas showed the highest frequencies of hypermethylation. Among endometrioid ovarian carcinomas, lower levels of promoter methylation of RSK4, SPARC, and HOXA9 were significantly associated with higher tumor grade; thus, the methylation patterns showed a shift to the direction of high-grade serous tumors. In conclusion, we provide evidence of a frequent epigenetic inactivation of RSK4, SPARC, PROM1, HOXA10, HOXA9, WT1-AS, SFRP2, SFRP5, OPCML, and MIR34B in the development of non-serous ovarian carcinomas of Lynch and sporadic origin, as compared to serous tumors. Our findings shed light on the role of epigenetic mechanisms in ovarian tumorigenesis and identify potential targets for translational applications.     

Argyrophilia in ovarian serous tumors. A comparative study in 127 epithelial ovarian tumors.

The distribution of argyrophil cells in epithelial ovarian tumors was studied in 127 cases. The results showed that not only mucinous tumors and endometrioid tumors contained argyrophil cells, but also some serous tumors expressed argyrophilia. 31% of serous tumors including 40% of serous adenocarcinomas contained variable numbers of argyrophil cells. Argyrophilia has been demonstrated in mucinous tumors, endometrioid tumors and Brenner tumors before. However, this is the first time the presence of argyrophilia in serous tumors has been noticed. Moreover, the argyrophil cells in 5 serous carcinomas showed reactivity with Neuroendocrine (chromogranin A) antibody but not with serotonin. The expression pattern of argyrophilia in the serous tumors was different from that of the mucinous tumors; in the former, argyrophil granules appeared in apical portions or throughout the cytoplasm of single or clustered cells. In addition, the argyrophilia in some serous tumors and endometrioid tumors decreased after diastase digestion. Ultrastructurally, no typical neurosecretory granule was found in the argyrophilic serous tumors. The findings in this study suggest that argyrophilia could be quite frequently found in ovarian epithelial tumors and in itself is not a very specific differential characteristic of carcinoid tumors. The argyrophilia found in a variety of epithelial ovarian tumors might lend additional support to the histogenesis and close relationship between the common epithelial tumors of the ovary.     

Classification of cancer cell lines using an automated two-dimensional liquid mapping method with hierarchical clustering techniques

A two-dimensional liquid mapping method was used to map the protein expression of eight ovarian serous carcinoma cell lines and three immortalized ovarian surface epithelial cell lines. Maps were produced using pI as the separation parameter in the first dimension and hydrophobicity based upon reversed-phase HPLC separation in the second dimension. The method can be reproducibly used to produce protein expression maps over a pH range from 4.0 to 8.5. A dynamic programming method was used to correct for minor shifts in peaks during the HPLC gradient between sample runs. The resulting corrected maps can then be compared using hierarchical clustering to produce dendrograms indicating the relationship between different cell lines. It was found that several of the ovarian surface epithelial cell lines clustered together, whereas specific groups of serous carcinoma cell lines clustered with each other. Although there is limited information on the current biology of these cell lines, it was shown that the protein expression of certain cell lines is closely related to each other. Other cell lines, including one ovarian clear cell carcinoma cell line, two endometrioid carcinoma cell lines, and three breast epithelial cell lines, were also mapped for comparison to show that their protein profiles cluster differently than the serous samples and to study how they cluster relative to each other. In addition, comparisons can be made between proteins differentially expressed between cell lines that may serve as markers of ovarian serous carcinomas. The automation of the method allows reproducible comparison of many samples, and the use of differential analysis limits the number of proteins that might require further analysis by mass spectrometry techniques.     

MicroRNA Profiling of BRCA1/2 Mutation-Carrying and Non-Mutation-Carrying High-Grade Serous Carcinomas of Ovary

Background

MicroRNAs (miRNA) are 20∼25 nucleotide non-coding RNAs that inhibit the translation of targeted mRNA, and they have been implicated in the development of human malignancies. High grade serous ovarian carcinomas, the most common and lethal subtype of ovarian cancer, can occur sporadically or in the setting of BRCA1/2 syndromes. Little is known regarding the miRNA expression profiles of high grade serous carcinoma in relation to BRCA1/2 status, and compared to normal tubal epithelium, the putative tissue of origin for high grade serous carcinomas.

Methodology/Principal Findings

Global miRNA expression profiling was performed on a series of 33 high grade serous carcinomas, characterized with respect to BRCA1/2 status (mutation, epigenetic silencing with loss of expression or normal), and with clinical follow-up, together with 2 low grade serous carcinomas, 2 serous borderline tumors, and 3 normal fallopian tube samples, using miRNA microarrays (328 human miRNA). Unsupervised hierarchical clustering based on miRNA expression profiles showed no clear separation between the groups of carcinomas with different BRCA1/2 status. There were relatively few miRNAs that were differentially expressed between the genotypic subgroups. Comparison of 33 high grade serous carcinomas to 3 normal fallopian tube samples identified several dysregulated miRNAs (false discovery rate <5%), including miR-422b and miR-34c. Quantitative RT-PCR analysis performed on selected miRNAs confirmed the pattern of differential expression shown by microarray analysis. Prognostically, lower level miR-422b and miR-34c in high grade serous carcinomas were both associated with decreased disease-specific survival by Kaplan-Meier analysis (p<0.05).

Conclusions/Significance

High grade serous ovarian carcinomas with and without BRCA1/2 abnormalities demonstrate very similar miRNA expression profiles. High grade serous carcinomas as a group exhibit significant miRNA dysregulation in comparison to tubal epithelium and the levels of miR-34c and miR-422b appear to be prognostically important.     

Normalizing genes for quantitative RT-PCR in differentiating human intestinal epithelial cells and adenocarcinomas of the colon

As for other mRNA measurement methods, quantitative RT-PCR results need to be normalized relative to stably expressed genes. Widely used normalizing genes include beta-actin and glyceraldehyde-3-phosphate dehydrogenase. It has, however, become clear that these and other normalizing genes can display modulated patterns of expression across tissue types and during complex cellular processes such as cell differentiation and cancer progression. Our objective was to set the basis for identifying normalizing genes that displayed stable expression during enterocytic differentiation and between healthy tissue and adenocarcinomas of the human colon. We thus identified novel potential normalizing genes using previously generated cDNA microarray data and examined the alterations of expression of two of these genes as well as seven commonly used normalizing genes during the enterocytic differentiation process and between matched pairs of resection margins and primary carcinomas of the human colon using real-time RT-PCR. We found that ribosomal phosphoprotein P0 was particularly stable in all intestinal epithelial cell extracts, thereby representing a particularly robust housekeeping reference gene for the assessment of gene expression during the human enterocytic differentiation process. On the other hand, beta-2-microglobulin generated the best score as a normalizing gene for comparing human colon primary carcinomas with their corresponding normal mucosa of the resection margin, although others were found to represent acceptable alternatives. In conclusion, we identified and characterized specific normalizing genes that should significantly improve quantitative mRNA studies related to both the differentiation process of the human intestinal epithelium and adenocarcinomas of the human colon. This approach should also be useful to validate normalizing genes in other intestinal contexts.     

Villin: a cytoskeletal protein and a differentiation marker expressed in some human adenocarcinomas

We studied the expression of villin, a microfilament-associated, actin-binding protein typical of brush-border microvilli, in a variety of human carcinomas by applying immunofluorescence microscopy to frozen sections and immunoblotting methods to tissue extracts using a rabbit antiserum and a monoclonal antibody specific for villin. All of the 24 primary and metastatic colorectal adenocarcinomas tested were uniformly and strongly positive for villin, with the immunocytochemical labeling concentrated at the luminal cell margin. In poorly differentiated tumor areas, rudimentary tubules were stained. All of the six tubular adenocarcinomas of the stomach studied as well as two adenocarcinomas of the gall bladder and a hepatocellular carcinoma were also villin-positive. Villin was detectable in 12 of 14 adenocarcinomas of the pancreas; in some of these cases, its distribution was heterogeneous. Among 21 renal cell carcinomas investigated, positivity for villin was seen in nine of 13 clear cell tumors (especially those of grade II), and in all four chromophilic cell tumors; however, all four chromophobe cell tumors studied were negative. Four of 11 endometrial but none of nine ovarian carcinomas were (uniformly or focally) villin positive. Of 18 adenocarcinomas of the lung studied, one was uniformly and four focally positive for villin, while the remainder were negative. All of the other epithelial tumors studied, including 12 adenocarcinomas of the breast and seven epithelial or biphasic pleural mesotheliomas, were villin negative. Our results show that the expression of villin in intestinal epithelial cells is consistently maintained in their corresponding carcinomas, even when the organized brush-border structure has been lost. The presence of villin in some endometrial and pulmonary adenocarcinomas--in contrast to its absence in the respective normal epithelia--suggests that this protein is newly expressed during hyperplasia, dysplasia, or carcinogenesis. Determining the presence or absence of villin and its immunocytochemical staining pattern in metastatic adenocarcinomas may be of some help in determining the type and site of the primary tumor.     

Primary Human Ovarian Epithelial Cancer Cells Broadly Express HER2 at Immunologically-Detectable Levels

The breadth of HER2 expression by primary human ovarian cancers remains controversial, which questions its suitability as a universal antigen in this malignancy. To address these issues, we performed extensive HER2 expression analysis on a wide panel of primary tumors as well as established and short-term human ovarian cancer cell lines. Conventional immunohistochemical (IHC) analysis of multiple tumor sites in 50 cases of high-grade ovarian serous carcinomas revealed HER2 overexpression in 29% of evaluated sites. However, more sensitive detection methods including flow cytometry, western blot analysis and q-PCR revealed HER2 expression in all fresh tumor cells derived from primary ascites or solid tumors as well as all established and short-term cultured cancer cell lines. Cancer cells generally expressed HER2 at higher levels than that found in normal ovarian surface epithelial (OSE) cells. Accordingly, genetically-engineered human T cells expressing an HER2-specific chimeric antigen receptor (CAR) recognized and reacted against all established or primary ovarian cancer cells tested with minimal or no reactivity against normal OSE cells. In conclusion, all human ovarian cancers express immunologically-detectable levels of HER2, indicating that IHC measurement underestimates the true frequency of HER2-expressing ovarian cancers and may limit patient access to otherwise clinically meaningful HER2-targeted therapies.     

Distinct Expression Levels and Patterns of Stem Cell Marker,Aldehyde Dehydrogenase Isoform 1 (ALDH1), in Human Epithelial Cancers

Aldehyde dehydrogenase isoform 1 (ALDH1) has been proved useful for the identification of cancer stem cells. However, our knowledge of the expression and activity of ALDH1 in common epithelial cancers and their corresponding normal tissues is still largely absent. Therefore, we characterized ALDH1 expression in 24 types of normal tissues and a large collection of epithelial tumor specimens (six cancer types, n = 792) by immunohistochemical staining. Using the ALDEFUOR assay, ALDH1 activity was also examined in 16 primary tumor specimens and 43 established epithelial cancer cell lines. In addition, an ovarian cancer transgenic mouse model and 7 murine ovarian cancer cell lines were analyzed. We found that the expression levels and patterns of ALDH1 in epithelial cancers are remarkably distinct, and they correlate with their corresponding normal tissues. ALDH1 protein expression levels are positively correlated with ALDH1 enzymatic activity measured by ALDEFLUOR assay. Long-term in vitro culture doesn''t significantly affect ALDH1 activity in epithelial tumor cells. Consistent with research on other cancers, we found that high ALDH1 expression is significantly associated with poor clinical outcomes in serous ovarian cancer patients (n = 439, p = 0.0036). Finally, ALDH^br tumor cells exhibit cancer stem cell properties and are resistant to chemotherapy. As a novel cancer stem cell marker, ALDH1 can be used for tumors whose corresponding normal tissues express ALDH1 in relatively restricted or limited levels such as breast, lung, ovarian or colon cancer.     

Expression of Serum Amyloid A in Human Ovarian Epithelial Tumors: Implication for a Role in Ovarian Tumorigenesis

Serum amyloid A (SAA) is an acute phase protein which is expressed primarily in the liver as a part of the systemic response to various injuries and inflammatory stimuli; its expression in ovarian tumors has not been described. Here, we investigated the expression of SAA in human benign and malignant ovarian epithelial tumors. Non-radioactive in situ hybridization applied on ovarian paraffin tissue sections revealed mostly negative SAA mRNA expression in normal surface epithelium. Expression was increased gradually as epithelial cells progressed through benign and borderline adenomas to primary and metastatic adenocarcinomas. Similar expression pattern of the SAA protein was observed by immunohistochemical staining. RT-PCR analysis confirmed the overexpression of the SAA1 and SAA4 genes in ovarian carcinomas compared with normal ovarian tissues. In addition, strong expression of SAA mRNA and protein was found in the ovarian carcinoma cell line OVCAR-3. Finally, patients with ovarian carcinoma had high SAA serum levels, which strongly correlated with high levels of CA-125 and C-reactive protein. Enhanced expression of SAA in ovarian carcinomas may play a role in ovarian tumorigenesis and may have therapeutic application. (J Histochem Cytochem 58:1015–1023, 2010)     

Differential analysis of ovarian and endometrial cancers identifies a methylator phenotype

Despite improved outcomes in the past 30 years, less than half of all women diagnosed with epithelial ovarian cancer live five years beyond their diagnosis. Although typically treated as a single disease, epithelial ovarian cancer includes several distinct histological subtypes, such as papillary serous and endometrioid carcinomas. To address whether the morphological differences seen in these carcinomas represent distinct characteristics at the molecular level we analyzed DNA methylation patterns in 11 papillary serous tumors, 9 endometrioid ovarian tumors, 4 normal fallopian tube samples and 6 normal endometrial tissues, plus 8 normal fallopian tube and 4 serous samples from TCGA. For comparison within the endometrioid subtype we added 6 primary uterine endometrioid tumors and 5 endometrioid metastases from uterus to ovary. Data was obtained from 27,578 CpG dinucleotides occurring in or near promoter regions of 14,495 genes. We identified 36 locations with significant increases or decreases in methylation in comparisons of serous tumors and normal fallopian tube samples. Moreover, unsupervised clustering techniques applied to all samples showed three major profiles comprising mostly normal samples, serous tumors, and endometrioid tumors including ovarian, uterine and metastatic origins. The clustering analysis identified 60 differentially methylated sites between the serous group and the normal group. An unrelated set of 25 serous tumors validated the reproducibility of the methylation patterns. In contrast, >1,000 genes were differentially methylated between endometrioid tumors and normal samples. This finding is consistent with a generalized regulatory disruption caused by a methylator phenotype. Through DNA methylation analyses we have identified genes with known roles in ovarian carcinoma etiology, whereas pathway analyses provided biological insight to the role of novel genes. Our finding of differences between serous and endometrioid ovarian tumors indicates that intervention strategies could be developed to specifically address subtypes of epithelial ovarian cancer.