Expression of ras oncogene p21 protein in normal and neoplastic ovarian tissues: correlation with histopathologic features and receptors for estrogen, progesterone, and epidermal growth factor

Diagnostic approach using the expression profiling of the P53 tumor suppressor gene and its related proteins in ovarian epithelial tumors

The initial aim of this study was to examine the expression profiles of P53 and its upstream genes, downstream genes, and cell cycle regulators to determine whether these markers are useful for making a differential diagnosis among the benign, borderline, and malignant ovarian epithelial tumors. Between borderline and malignant tumors, the increased expression levels of P53, Bax, Cyclin E, and cyclin-dependent kinase-2 as well as the decreased expression levels of growth arrest and DNA damage (GADD45) and murine double minute-2 (MDM2) were significantly associated with malignancy (P < 0.01, each). Using the receiver operating curve (ROC), the most reliable cutoff value of the added-up staining scores of those markers was 4.5 with 79% sensitivity and 89% specificity for malignancy. Between benign and borderline tumors, the P21 and Bax expression levels were significantly higher in borderline tumors, whereas the Bcl-2 expression level was much higher in benign tumors (P < 0.01, each). Using the ROC, the cutoff value of the added-up staining scores used to discriminate between the two groups was 2.5 with 70% sensitivity and 74% specificity for borderline tumors. Thus, for the differential diagnosis between borderline and malignant tumors, the cutoff value 4.5 of the cumulative staining scores can be used. However, the cutoff value 2.5 for discrimination between benign and borderline tumors may not be useful because of its relatively low sensitivity and specificity. In addition, the P53, GADD45, Cyclin E, and MDM2 expression levels in malignant ovarian tumors might be useful for determining the histologic grade and type.

Retinoids and ovarian cancer

Each year, an estimated 26,000 women in the United States are diagnosed with ovarian cancer. During any given year, approximately 14,500 women die from this disease. Ovarian cancer is the seventh most common cancer in women worldwide, after breast, cervix, colon/rectum, stomach, corpus uteri, and lung cancers. In the U.S., ovarian cancer is the second most common gynecologic cancer, and is the fourth leading cause of solid tumor cancer deaths among women. Currently, postoperative chemotherapy of ovarian cancer is still suboptimal. Drug resistance is a common problem resulting in only 20 approximately 30% overall 5-year survival rates. Clearly, continued development of alternative therapeutic strategies is essential for the management of this fatal disease. A number of recent studies have suggested that retinoids may play a potential role as an ovarian cancer chemotherapeutic agent. Retinoids, the natural and synthetic derivatives of vitamin A, have been shown to inhibit the growth of human ovarian cancer cells both in vivo and in culture. This review will initially summarize what is known about the pathological and molecular characteristics of ovarian carcinoma. It will then describe retinoid metabolism and the role of the cellular and nuclear retinoid binding proteins in mediating retinoid action. Following this general review of retinoids and their function, data supporting the role of retinoic acid as a suppresser of ovarian carcinoma cell growth will be presented. Particular attention will be paid to studies suggesting that members of the RB family of proteins and RB2/p130, in particular, are the molecular targets responsible for retinoid mediated inhibition of ovarian carcinoma cell growth. This review will then conclude with a brief discussion of two synthetic retinoids, 4 HPR R(fenretinide) and AHPN/CD437, which have been shown to induce apoptosis in ovarian tumor cells. It will be clear from the studies summarized in this review that retinoids represent a potentially powerful alternative to present chemotherapeutic approaches to the treatment of late stage ovarian cancer.

ras gene activation and infrequent mutation in papillary serous carcinoma of the peritoneum

OBJECTIVE

The ras genes are a well-studied family of proto-oncogenes whose involvement in many cancers has been delineated. However, K-ras mutations have not previously been examined in papillary serous carcinoma of the peritoneum (PSCP), a tumor which resembles serous epithelial ovarian carcinoma (SEOC) both in histology and epidemiology. In this study we examine the role of the K-ras oncogene in PSCP compared to SEOC.

METHODS

Using single-strand conformational polymorphism analysis and cycle sequencing protocols, we evaluated our collection of 51 cases of PSCP for K-ras mutations and compared these findings with the experience in SEOC. We then examined 5 cases of PSCP for activation of ras proteins and MAP kinase to evaluate the potential involvement of the ras pathway in PSCP tumorigenesis.

RESULTS

We found only one K-ras mutation in our 51 cases (2%) of PSCP compared to three mutations in 46 cases (6.5%) of high-grade, late-stage SEOC. This was not significantly different (P > 0.10). In the single PSCP case with a K-ras mutation, the mutation was found in only one of five tumor sites tested. All four mutations involved a single nucleotide alteration in codon 12 (GGT to GTT, Gly to Val). To evaluate the ras pathway in PSCP, we used the known activated ras binding domain on Raf-1 to perform an assay to test for activated ras. We identified ras activation in 4 of 5 PSCP cases tested and, to confirm that the activation was functional, we tested and found similar activation of MAP kinase, a downstream mediator for K-ras expression.

CONCLUSIONS

K-ras mutations occur at low rates in both PSCP and high-grade, late-stage SEOC, and therefore K-ras mutations are not involved in the development of these two diseases. Finding the mutation in only one of multiple tumor sites in the PSCP case supports growing evidence for a multifocal origin of PSCP. Our findings of ras activation in four of five cases of PSCP suggest that ras activation by mechanisms other than genetic mutation is important for PSCP tumorigenesis.

Expression of brx proto-oncogene in normal ovary and in epithelial ovarian neoplasms

OBJECTIVE

We previously identified a protein, Brx, that interacted with estrogen receptor alpha. Sequence analysis determined that Brx is a novel member of the Dbl family of oncoproteins involved in signaling pathways that regulate cell growth. Because the Brx protein was found to be highly expressed in hormoneresponsive breast epithelium, the objective of this study was to determine whether Brx was expressed in both normal and neoplastic ovarian tissues.

STUDY DESIGN

A polyclonal antiserum directed against the Brx protein was used to perform immunolocalization on sections from 5 normal ovaries and 20 ovarian neoplasms. Chromosomal localization of the brx gene was accomplished by means of fluorescence in situ hybridization. Northern and Western blot analyses were performed on extracts prepared from human ovaries.

RESULTS

Brx protein was localized to the cytoplasm of granulosa cells from mature graafian follicles, the corpus luteum, and islands of hilar cells in normal ovaries. In tumors with low malignant potential and ovarian carcinomas the neoplastic epithelium stained strongly for Brx protein. Northern and Western blot analyses, respectively, confirmed expression of Brx messenger ribonucleic acid and protein in normal ovary. Finally, the brx gene was localized to 15q25.

CONCLUSION

The proto-oncogene brx is expressed in specific normal human ovarian tissues and is also present in ovarian epithelial neoplasms.

Adenovirus-mediated p53 growth inhibition of ovarian cancer cells is independent of endogenous p53 status

OBJECTIVE

The aim of this study was to determine the effect of transfection of adenovirus-mediated wild-type p53 into ovarian cancer cells with both wild-type and mutant endogenous p53.

STUDY DESIGN

Eight human ovarian cancer cell lines were used: three with p53 mutations, one that is p53 null, and four with wild-type p53. The recombinant p53 adenovirus (Adp53) contains the cytomegalovirus promoter, wild-type p53 cDNA, and SV40 polyadenylation signal in a minigene cassette inserted into the E1-deleted region of modified Ad5. The transduction efficiency of cells was assessed using a beta-gal-containing adenovirus. Cell-counting assays were used to evaluate the effect of transfection with Adp53 on the growth of cells. P53 expression was evaluated using Western blot. Cell cycle analysis and apoptosis studies were done using a tunnel-based assay and fluorescent activated cell sorting.

RESULTS

Transduction efficiencies varied between cell lines. More than 90% growth inhibition occurred in seven of eight cell lines after infection with adenovirus-mediated p53 if a viral dose leading to at least 50% of cells infected was used. Regardless of endogenous p53 status, apoptosis occurred in cells infected with p53.

CONCLUSIONS

Ovarian cancer cells are growth inhibited by transfection with adenovirus-mediated p53 regardless of their endogenous p53 status. Growth inhibition is related to transduction efficiency.

Cytogenetic, morphologic and oncogene analysis of a cell line derived from a heterologous mixed mullerian tumor of the ovary

A cell line was established from a mixed mullerian tumor of the ovary and designated LN1. Histopathologic analysis of the fresh tumor specimen demonstrated a highly aneuploid heterologous tumor comprised of undifferentiated mesodermal components with carcinomatous cells present as a smaller population. Long-term in vitro culture resulted in the establishment of a cell line that exhibits an epithelial-like morphology and expresses epithelial antigens cytokeratin, epithelial membrane antigen, and carcinoma antigen TAG-72. These cells also express mesenchymal intermediate filaments, vimentin, and desmin. Karyotypic analysis revealed a basic triploid pattern with multiple chromosomal abnormalities, most notably an isochromosome of the short arm of five present in three copies. Analysis of oncogene expression revealed that LN1 cells constitutively express mRNA for c-ras, c-erbB2, and p53. The expression of mRNA for cellular oncogenes correlated with the presence of corresponding oncoproteins, p21H-ras, p21K-ras, and p185erB2 and mutant p53 protein. In summary, coexpression of epithelial and mesenchymal antigens by LN1 cells lends support to the hypothesis that epithelial and mesenchymal elements comprising mixed mullerian tumors of the ovary are derived from a common stem cell precursor. Furthermore, this cell line represents a functional in vitro model to evaluate the biologic activities of these unusual and highly aggressive ovarian malignancies.

Oncogenes in gynecologic oncology

Advances in molecular biology have facilitated the recent investigation of gynecological malignancies. The presence of certain oncogenes within gynecological tumors indicates that transformation may be associated with genetic alteration of normal regulatory processes. This paper reviews several oncogenes that have been implicated in the transformation of gynecological tissues.

Expression of cyclin D1 and c-Ki-ras gene product in human epithelial ovarian tumors

The expression of cyclin D1 gene product in human ovarian tumors was studied. We found that cyclin D1 is expressed at high levels in several ovarian cancer cell lines. Immunohistochemical study also showed that a significant proportion of primary ovarian tumor tissues overexpressed cyclin D1 gene product. Clear nuclear staining of cyclin D1 protein was detected in 28% of the cases. We also characterized the expression of c-Ki-ras gene product in ovarian cancer cell lines and tumor tissues. Amplification or overexpression of this proto-oncogene has been reported in ovarian tumors from Taiwan. These results show that c-Ki-ras is strongly expressed in PA-1 and NIH:OVCAR-3 cells in which cyclin D1 also expressed at high levels. Specific cytoplasmic staining of c-Ki-ras protein was detected in 11 tumors (52%). Statistical analyses show a strong positive correlation between cyclin D1 and c-Ki-ras immunoexpression. Thus, these data support the ideas that cyclin D1 may be involved in the pathogenesis of ovarian cancer, and coactivation of cyclin D1 and c-Ki-ras gene expression may represent one of the major pathways that lead to the development of ovarian cancer in Taiwan.

Influence of cell type on the steroidogenic potential and basal cyclic AMP levels of ras-oncogene-transformed rat cells

Transformation with ras oncogenes causes loss, maintenance or modulation of differentiation, depending on the developmental history of the target cells. In the present study, we examined steps in signal transduction that may underlie some of this variation, using steroidogenic cells of adult rats as the model system. Steroidogenesis in normal cells is regulated by cyclic AMP and protein kinase A (the cAMP/PKA pathway). We showed previously that transformation with v-Ki-ras induces constitutive progesterone secretion in ovarian and adrenocortical cells that are normally steroidogenic (ovarian granulosa and adrenal glomerulosa cells) and also in developmentally related cells that are normally nonsteroidogenic (ovarian surface epithelium and adrenal capsular fibroblasts), but not in unrelated nonsteroidogenic cells, such as muscle fascia fibroblasts and peritoneal mesothelium. In the present study, basal cAMP levels in all transformed ovarian and adrenal cell-lines were increased over basal levels in normal cells, and of transformed muscle fascia and mesothelial cell-lines. As in normal cells, transformation-induced steroidogenesis was stimulated by cAMP and was PKA dependent. A comparison of malignancy-related characteristics showed that transformed cells from nonsteroidogenic organs were more tumorigenic in vivo and less sensitive to growth inhibition by cAMP in vitro than transformed ovarian and adrenocortical cells. The results show that the abnormal, constitutive steroidogenesis induced by the viral form of the Kirsten ras oncogene (v-Ki-ras) in certain cell types is associated with tissue-specific increases in basal cAMP levels. Thus, although the ras oncogenes function primarily through other signal transduction pathways, transformation with ras oncogenes alters PKA-mediated signal transduction in a manner that is developmentally determined.(ABSTRACT TRUNCATED AT 250 WORDS)

Potential criteria for cohort selection in chemoprevention trials of epithelial ovarian cancer

Epithelial ovarian cancer is a heterogeneous disease. Epidemiologic studies have identified risk factors for this disease including advanced age, nulliparity, history of infertility, early age at menarche, late age at menopause, and perhaps ovulation induction. Cohort selection that includes women who have potential precursor lesions and alterations of select biomarkers may prove useful in the design of chemoprevention trials of epithelial ovarian cancer. Nuclear morphometry, specific genetic alterations, and markers of proliferation and differentiation may be useful biomarkers to monitor the efficacy of specific interventions.

Expression of ras oncogene p21 protein in normal and neoplastic laryngeal tissues: correlation with histopathological features and epidermal growth factor receptors

Western blotting analysis of the p21 ras oncoprotein was performed in seven normal laryngeal mucosa specimens and 43 primary laryngeal cancers. Varying p21 levels, expressed as optical density (OD), were found in normal mucosa (median 1.94 OD, range 0.90-2.17 OD) and in primary laryngeal tumours (median 1.74 OD, range 0.30-6.37 OD). When p21 expression in laryngeal cancer was compared with the normal counterpart, higher levels were found in neoplastic than in normal laryngeal tissue (median 2.54 OD, range 1.76-6.37 OD, vs median 1.94 OD, range 0.90-2.17 OD) (P = 0.023). Immunohistochemical analysis demonstrated that most of the tumour cells (more than 70%) were immunostained while the stromal component was unreactive. No correlation between p21 expression and tumour location, stage and histopathological grade was observed. The correlation between ras p21 protein expression and epidermal growth factor receptor (EGFR) levels was also investigated. EGFR-positive cases did not show any difference in p21 expression with respect to EGFR-negative cases (median 1.52 OD, range 0.30-6.37 OD, vs median 1.84 OD, range 0.93-3.71 OD). Our findings suggest that overexpression of p21 protein is associated with a malignant phenotype in laryngeal cancer. Further studies should be undertaken to evaluate whether the assessment of p21 protein expression may have clinical significance in laryngeal cancer.

Immunodetection of the p21-ras products in human normal and preneoplastic tissues and solid tumors: a review

Immunohistochemical detection of p21-ras to identify and characterize preneoplastic or neoplastic lesions in human tissues is reviewed. Information concerning the commercially available antibodies is presented. Antibodies DWP, Ras-10, Y13-259, YA6-172, NCC-001, and NCC-004 are fully documented with respect to their behavior in appropriate specificity tests and appear to be reliable reagents. After reviewing the data we have identified three groups of tissues or organs with respect to positive immunostaining for p21-ras as the significant criterion of malignancy. These three groups comprise (1) tissues for which no definite conclusion could be drawn (colon, lung, bladder, ovary, and neural and odontogenic tissues) despite occasional claims to the contrary, (2) tissues for which conclusions were negative (pancreas and stomach), and (3) tissues for which p21-ras staining positively discriminated malignant from normal tissues (liver, uterus, and salivary gland). Immunohistochemically detectable levels of products from a mutated ras gene could be demonstrated in a fraction of the samples from colon, lung, and bladder carcinomas, as well as in some histologically normal tissues adjacent to a colon carcinoma. The possibility that a higher relative intensity of the immunostaining reaction for p21-ras might discriminate malignant tissues from normal tissues or benign lesions in breast, pancreas, stomach, lung, uterus, or thyroid samples is suggested. Further studies now appear warranted and a strategy is proposed to validate the conclusions reached thus far.