The ERK Signaling Cascade Inhibits Gonadotropin-Stimulated Steroidogenesis.

The response of granulosa cells to Luteinizing Hormone (LH) and Follicle- Stimulating Hormone (FSH) is mediated mainly by cAMP/protein kinase A (PKA) signaling. Notably, the activity of the extracellular signal-regulated kinase (ERK) signaling cascade is elevated in response to these stimuli as well. We studied the involvement of the ERK cascade in LH- and FSH-induced steroidogenesis in two granulosa-derived cell lines, rLHR-4 and rFSHR-17, respectively. We found that stimulation of these cells with the appropriate gonadotropin induced ERK activation as well as progesterone production downstream of PKA. Inhibition of ERK activity enhanced gonadotropin-stimulated progesterone production, which was correlated with increased expression of the Steroidogenic Acute Regulatory Protein (StAR), a key regulator of progesterone synthesis. Therefore, it is likely that gonadotropin-stimulated progesterone formation is regulated by a pathway that includes PKA and StAR, and this process is down-regulated by ERK, due to attenuation of StAR expression. Our results suggest that activation of PKA signaling by gonadotropins not only induces steroidogenesis but also activates down-regulation machinery involving the ERK cascade. The activation of ERK by gonadotropins as well as by other agents may be a key mechanism for the modulation of gonadotropin-induced steroidogenesis.

Modulation of the steroidogenic related activity according to the design of single-chain bovine FSH analogs.

Single-chain (SC) gonadotropins have been genetically engineered to increase the repertoire of analogs for potential use in humans and domestic animals. The major aim of the current study was to examine the steroidogenic related activity of SC FSH analogs carrying structural differences. To address this issue, we designed and expressed three SC bovine FSH analogs in CHO cells: (i) FSHßα in which the tethered subunit domains are linked in tandem; (ii) FSHßCTPα that contains the carboxy terminal peptide (CTP) of the human choriogonadotropin (hCG) ß subunit as a spacer, and (iii) FSHßboCTPα in which the linker is derived from a CTP-like sequence (boCTP) decoded from the bovine LHß DNA. The data suggested that the secretion efficiency of these variants from the transfected cells was unaffected by the presence or absence of the CTP linker, N-glycans were attached to the analogs and the hCGß-CTP domain in the FSHßCTPα variant was O-glycosylated. In a rat immortalized granulosa cell bioassay the potency of the three variants towards progesterone secretion varied. In immature mice, the analogs increased the ovary weight and induced StAR, Cyp11a (P450scc), Cyp17 (P450c17) and Cyp19 (P450aromatase) transcripts. However, the dose dependence and amplitude of these transcript levels differed in response to FSHßα, FSHßboCTPα and FSHßCTPα. Collectively, these data suggest that the design of the FSH analog can modulate the bioactivity in vitro and in vivo. A systematic analysis of receptor activation with ligands carrying structural differences may identify new regulatory factor/s involved in the pleiotropic FSH activity.

The expression of heparanase in normal and preeclamptic placentas.

OBJECTIVE: Heparanase plays a central role in processes of placentation. Abnormal placentation may result in inadequate uteroplacental blood flow, leading to unsuccessful pregnancy outcome and preeclampsia. We aimed to evaluate heparanase expression in placentas of preeclamptic patients. MATERIALS AND METHODS: Placental tissue samples were collected immediately after delivery from 9 preeclamptic patients and 3 healthy controls at term, and were analyzed by immunohistochemistry, western blot analysis and real-time PCR, with regard to the presence of heparanase. RESULTS: Immunohistochemistry staining for heparanase did not differ between normal and preeclamptic placental sections. On the other hand, western blot analysis revealed increased expression of heparanase in preeclpamptic placentas compared to controls, p = 0.001. Similarly, RT-PCR analysis showed also an increased expression of heparanase m-RNA compared to health controls, p = 0.005. CONCLUSION: Heparanase is over expressed in preeclamptic placentas compared to normal healthy controls, suggesting its role in the development of preeclampsia.

Dynamic distribution of ERK, p38 and JNK during the development of pancreatic ductal adenocarcinoma.

We recently discovered that oncogenic c-kit is highly expressed concomitantly with the development of pancreatic ductal adenocarcinoma (PDAC). Since oncogenic c-kit may activate major pathways of protein tyrosine phosphorylation, we decided to investigate this issue in the major protein phosphorylation cascades. In normal pancreas labeling with antiphosphorylated ERK1/2 (pERK1/2) antibody was mainly confined to islets of Langerhans in close overlapping with insulin containing cells. Phosphorylated p38 (pp38) showed a similar pattern of distribution, while only weak labeling was evident for pJNK and no labeling of pMEK was observed. As expected, general ERK1/2 (gERK1/2), general p38 (gp38), general JNK (gJNK) as well as general MEK (gMEK) were all evident in islets of Langerhans and in the exocrine tissue. In early development of PDAC, pERK1/2 and pp38 retained their localization in islets of Langerhans. Intensive staining of pERK1/2 was also evident in the cancerous ducts, while the labeling with antibodies to pp38 was more moderate. While pJNK staining in islets of Langerhans was weak, with no labeling in the cancerous ducts, antibodies to gJNK revealed intensive staining suggesting the weak staining of pJNK is not due to the lack of the enzyme. In a more advanced stage of PDAC the carcinomas were clearly stained with pERK1/2 and pp38, while moderate staining with pJNK was also evident. In liver metastases, the cancer cells were heavily labeled with all three phospho-MAPKs. It should be noted that the localization of all three kinases was mainly in the cell nuclei. In the more advanced stage of PDAC, heavy labeling was evident using antibodies to gERK1/2, gp38, gJNK and gMEK. However, no labeling to pMEK was evident in parallel sections. Our data suggest that both in normal and cancerous pancreas, most of the MAPK activities are located in islets of Langerhans and cancerous ducts. It is suggested that using inhibitors to protein kinases may attenuate the progression of the disease.

TADG-12 as an early marker in the development of pancreatic ductal adenocarcinoma (PDAC): involvement of insulin containing cells.

TADG-12 is a serine protease that was characterized as expressed in ovarian and gastric carcinomas. Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal cancers and its late detection results in poor prognosis. Therefore, we decided to examine whether TADG-12 appears early in PDAC development. In normal pancreas, pale to moderate immunostaining is present in islets of Langerhans, while exocrine tissue and ducts are free from labeling. In contrast, in cancer patients, who still preserve the integrity of the exocrine and the endocrine tissues, a pronounced immunolabelling of TADG-12 was evident mainly located in the insulin containing ß cells. In a more progressive stage of the disease TADG-12 was also evident in the deteriorated exocrine tissue. TADG-12 was also heavily labeled in islets of Langerhans, which were embedded in the stroma of the residual pancreatic tissue. Again, there was a considerable overlap between the labeling of insulin and TADG-12 in these islets. Close correlation between insulin and TADG-12 was also evident in islets of Langerhans surrounded by adipose cells. The TADG-12 labeled was confined to the cytoplasm and the membrane of the cells. In the progressive stage of PDAC, the cancerous ducts were clearly labeled with TADG-12 with no labeling of insulin. At high magnification the TADG-12 clearly labeled the cytoplasm and the cell wall membrane of duct cells, while the nuclei remained unstained upon incubation with antibodies to TADG-12. The present findings may assist in early detection of PDAC as well as targeting of TADG-12 in order to attenuate the rapid progression of the disease.

CD24 and Nanog identify stem cells signature of ovarian epithelium and cysts that may develop to ovarian cancer.

Ovarian cancer is the most lethal gynecological cancer. There is a general debate whether ovarian cancer is an intrinsic or an imported disease. We investigated whether in normal morphological appearance and in early stages of ovarian tumorgenesis typical cancer cell markers such as CD24 and Nanog are expressed. In 25% of normal appearing ovaries of post-menopausal women there was co-localization of CD24 and Nanog in the walls of the ovarian cysts, leaving the epithelial cells on the surface of these ovaries free of Nanog or CD24 expression. In benign ovarian tumors 37% of specimens were positive to CD24 and Nanog labeling while 26% of them were localized in the cyst walls. In contrast, in serous borderline tumors 79% specimens were labeled with CD24, 42% of them were localized in cysts and in 32% of them showed co-localization with CD24 and Nanog was evident: the rest were labeled in the ovarian epithelial cells. In serous ovarian carcinomas 81% specimens were labeled with CD24 antibodies. In 45% of them co-localization with Nanog was evident in the bulk of the cancerous tissue. In mucinous carcinomas no labeling with CD24 or Nanog was evident. In view of the synergistic effect of CD24 and Nanog expressed in malignant cancer development in other systems, it is suggested that such an analysis can be valuable for early detection of ovarian cancer. Moreover, the abundance of these markers in cysts in the development of ovarian cancer may suggest that they present an intrinsic source of the development of the highly malignant disease. Finally, since CD24 is exposed on the surface of the cancer cells, it may be highly beneficial to target these cells with antibodies to CD24 conjugated to cytotoxic drugs for more efficient treatment of this malignant disease.

Modulation of c-kit expression in pancreatic adenocarcinoma: a novel stem cell marker responsible for the progression of the disease.

Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal cancers because of late symptoms and resistance to chemotherapy and radiation therapy. We have investigated the appearance of c-kit, a stem cell marker, in both normal adult pancreatic tissue and in cancerous tissue. Apart from some very pale staining of islets of Langerhans, normal pancreas was devoid of staining with antibodies to c-kit. In contrast, in cancerous tissue that still preserves the overall integrity of the pancreatic tissue, there was a clear labeling in islets of Langerhans, which seemed to be co-localized with insulin containing ß cells. In other cases, where the pancreatic tissue was completely deteriorated, intensive labeling was clearly evident in remnants of both the exocrine and the endocrine tissues. The duct cells of the adenocarcinoma were moderately but clearly labeled with antibodies to c-kit. In contrast, in metastasis of PDAC, very intensive labeling of c-kit was evident. The location of KRAS, which is strongly associated with PDAC, was also analyzed at the initial stages of the disease, when islets of Langerhans still preserve their integrity to a large extent. KRAS was found exclusively in islets of Langerhans and overlapped in its location with insulin and c-kit expressing cells. It is suggested that the modulation of the expression of c-kit, visualized by antibodies to the oncogene molecule, may play an important role in the formation and progression of PDAC. The absence of c-kit in normal pancreas and its appearance in PDAC is probably due to a mutational event, which probably allows conversion of the ß cells into cancer stem cells (CSC). Co-expression of both c-kit and KRAS, typical markers for CSC with overlapping with insulin in islets of Langerhans, strongly support the notion that ß-cells play a central role in the development of PDAC. The use of specific drugs that can attenuate the kinase activity of c-kit or target KRAS expressing cancer cells should be tested in order to attenuate the progression of this lethal disease.

The role of the 3' region of mammalian gonadotropin ß subunit gene in the luteinizing hormone to chorionic gonadotropin evolution.

CGß subunits comprise a unique carboxyl-terminal peptide (CTP) that has multiple O-linked glycans and extends serum half-life of the protein. It has evolved by incorporating a previously untranslated region of the LHß gene into the reading frame. Although CTP-like sequences are encrypted in the LHß genes of several mammals, the CGß subunit developed only in primates and equids. To study this restriction in evolution, we examined whether the cryptic CTP decoded from the bovine LHß gene (boCTP) possesses key characteristics of the human (h) CGß-CTP. The boCTP does not impede several crucial aspects of hormone biosynthesis, but compared to the hCGß-CTP, the stretch lacks O-glycans and determinants for circulatory survival. O-glycan deficiency and the associated incapacity to extend serum half-life is a major drawback of the boCTP. This may explain why LH did not evolve into CG in ruminants and consequently alternative mechanisms evolved to delay luteolysis early in gestation.

LGR5 and Nanog identify stem cell signature of pancreas beta cells which initiate pancreatic cancer.

Pancreas cancer, is the fourth leading cause of cancer death but its cell of origin is controversial. We compared the localization of stem cells in normal and cancerous pancreas using antibodies to the stem cell markers Nanog and LGR5. Here we show, for the first time, that LGR5 is expressed in normal pancreas, exclusively in the islets of Langerhans and it is co-localized, surprisingly, with Nanog and insulin in clusters of beta cells. In cancerous pancreas Nanog and LGR5 are expressed in the remaining islets and in all ductal cancer cells. We observed insulin staining among the ductal cancer cells, but not in metastases. This indicates that the islet's beta cells, expressing LGR5 and Nanog markers are the initiating cells of pancreas cancer, which migrated from the islets to form the ductal cancerous tissue, probably after mutation and de-differentiation. This discovery may facilitate treatment of this devastating cancer.

Two initiation sites of early detection of colon cancer revealed by localization of pERK1/2 in the nuclei or in aggregates at the perinuclear region of the tumor cells.

We have used human specimens and antibodies to pERK1/2 to detect early development of colon cancer using indirect immunocytochemistry. Two distinct sites were stained; one at the tip of the colon crypts and the other in the stromal tissue associated with the colonic tissue. These foci represent early stages of colon cancer initiation sites as established by enhanced Kirsten Rat Sarcoma Virus (KRAS) and the lack of p53 staining. The enhanced KRAS coincides with the initiation of tumor growth revealed by pERK1/2, both in the tip of the colon crypts, as well as in the stromal initiation site of the colon tumors. Foci of pERK1/2 staining were also detected in 50% of stromal tissue and tips of colon crypts, which were classified as normal tissues, adjacent to the malignant tissue according to general morphology. However, in colon specimens, where no malignancy was observed, no accumulation of pERK1/2 was observed. The staining of pERK1/2 at the stromal foci of the apparently non-malignant tissue appeared as aggregates in the perinuclear region, while in the colon epithelium it appeared in the cell nuclei. In low-grade colon cancer that was still free of induced mutated p53, staining of pERK1/2 was prominent in the cell nuclei, both in the stroma tissue and the tip of the colon crypts. In the intermediate stage, that exhibited significant p53 staining, only a fraction of p53-free tumor cells was labeled with pERK1/2 antibody, while in high-grade tumors, all cells of tumors were labeled with antibodies to p53, but not with antibodies to pERK1/2. We suggest that the down regulation in pERK1/2 labeling is due to the mitogenic capacity of the tumor cells, which are shifted from being driven by nuclear pERK1/2 to mutated p53 expression. We also found that the cytoplasm of low grade tumors was positive for epiregulin, while this labeling decreased in high-grade tumors. We found that the tumors arising from the stroma demonstrated poor structural differentiation, while the tumors initiating from the epithelial cells of the colon demonstrated high structural differentiation. We conclude that pERK1/2 is a sensitive marker of early colon cancer, which disappears at later stages of cancer development. Moreover, pERK1/2 staining can distinguish between tumor cells originating from the tip of the colon crypts and those developing in the stroma, which is present in the close vicinity to colon epithelial tissue, and thus can assist in selecting the appropriate therapy.

Localization of the stem cell markers LGR5 and Nanog in the normal and the cancerous human ovary and their inter-relationship.

LGR5 and Nanog were recently characterized as stem cell markers in various embryonic, adult and cancer stem cells. However, there are no data on their precise localization in the normal adult ovary, which may be important for the initial steps of development of ovarian cancer, the most lethal gynecological cancer. We analyzed by immunocytochemistry the precise localization of these markers in normal ovary (11 specimens, age range 43-76), in borderline specimens (12 specimens), and in serous ovarian cancer (12 specimens of stage II) which comprises the vast majority (80%) of all ovarian cancer. Surprisingly, we revealed that both Nanog and LGR5 are clearly localized in the epithelial cells of the normal ovary. However, in 5 of 12 ovaries there was no labeling at all, while in 3 ovaries staining of Nanog was more prominent with only weak labeling of LGR5. In addition, we found in 3 of 11 ovaries clear labeling in foci of both LGR5 and Nanog antibodies, with partial overlapping. Occasionally, we also found in the stroma foci labeled by either Nanog or LGR5 antibodies. In general, the stroma area of tissue sections labeled with LGR5 was much greater than that labeled with Nanog. In borderline tumors a significant portion of the specimens (7 of 12) was labeled exclusively with Nanog and not with LGR5. In ovarian carcinomas almost 100% of the cells were exclusively labeled only with Nanog (6 of 12 of the specimens) with no labeling of LGR5. These data may suggest the potential of ovaries from postmenopausal women, which express Nanog, to undergo transformation, since Nanog was shown to be oncogenic. We conclude that Nanog, which probably plays an important role in ovarian embryonic development, may be partially silenced in fertile and post-menopausal women, but is re-expressed in ovarian cancer, probably by epigenetic activation of Nanog gene expression. Expression of Nanog and LGR5 in normal ovaries and in borderline tumors may assist in the early detection and improved prognosis of ovarian cancer. Moreover, targeting of Nanog by inhibitory miRNA or other means may assist in treating this disease.

Differential localization of LGR5 and Nanog in clusters of colon cancer stem cells.

One paradigm of cancer development claims that cancer emerges at the niche of tissue stem cells and these cells continue to proliferate in the tumor as cancer stem cells. LGR5, a membrane receptor, was recently found to be a marker of normal colon stem cells in colon polyps and is also expressed in colon cancer stem cells. Nanog, an embryonic stem cell nuclear factor, is expressed in several embryonic tissues, but Nanog expression is not well documented in cancerous stem cells. Our aim was to examine whether both LGR5 and Nanog are expressed in the same clusters of colon stem cells or cancer stem cells, using immunocytochemistry with specific antibodies to each antigen. We analyzed this aspect using paraffin embedded tumor tissue sections obtained from 18 polyps and 36 colon cancer specimens at stages I-IV. Antibodies to LGR5 revealed membrane and cytoplasm immunostaining of scattered labeled cells in normal crypts, with no labeling of Nanog. However, in close proximity to the tumors, staining to LGR5 was much more intensive in the crypts, including that of the epithelial cells. In cancer tissue, positive LGR5 clusters of stem cells were observed mainly in poorly differentiated tumors and in only a few scattered cells in the highly differentiated tumors. In contrast, antibodies to Nanog mainly stained the growing edges of carcinoma cells, leaving the poorly differentiated tumor cells unlabeled, including the clustered stem cells that could be detected even by direct morphological examination. In polyp tissues, scattered labeled cells were immunostained with antibodies to Nanog and to a much lesser extent with antibodies to LGR5. We conclude that expression of LGR5 is probably specific to stem cells of poorly differentiated tumors, whereas Nanog is mainly expressed at the edges of highly differentiated tumors. However, some of the cell layers adjacent to the carcinoma cell layers that still remained undifferentiated, expressed mainly Nanog with only a few cells labeled with antibodies to LGR5. Considering the different sites and pattern of expression in the tumor, our data imply that targeting the clustered stem cells expressing LGR5 in poorly differentiated colon cancer may require different strategies than targeting the stem cells expressing Nanog in the highly differentiated tumors. Alternatively, combined application of specific inhibitory miRNAs to Nanog and to LGR5 expression may assist therapeutically.

Use of multiple biomarkers for the localization and characterization of colon cancer stem cells by indirect immunocytochemistry.

In this study, we used LGR5, γ-synuclein, p53, KRAS and epiregulin antibodies to localize stem cells by indirect immunocytochemistry in paraffin sections of normal and cancerous colon tissues. In the normal colon tissue, no staining of cells with LGR5, γ-synuclein, p53 and KRAS antibodies was observed, apart from a few scattered cells in between the colon villi that were faintly stained with antibodies to LGR5. Staining of highly differentiated cancer tissue with LGR5 antibodies revealed single cells or clusters of up to 4 cells in the interior space of the carcinoma cell layers. Staining of poorly differentiated cancer tissues (stage I-IV) revealed 9-81 clustered stem cells. The number of clustered stem cells increased significantly with the tumor stage, when comparing stage II to stage IV (p<00048). Occasionally, the clustered stem cells appeared in the interphase between the colon stroma and the tumor tissue. Surprisingly, antibodies to p53 clearly stained the clusters of stem cells both in the nuclei and the cytoplasm. The staining of the nuclei of other cells in the undifferentiated tumors was in general weaker, and no staining was found in the cytoplasm. Antibodies to γ-synuclein heavily stained the endothelial cells of the blood vessels and some other scattered cells in the highly differentiated tumors. Antibodies to γ-synuclein heavily stained the stem cells in both the cytoplasm and the nuclei of poorly differentiated tumors. Antibodies to KRAS stained the cytoplasm and the nuclei of stem cells in poorly differentiated tumors and also stained the cytoplasm of some scattered cells. Antibodies to epiregulin stained the cytoplasm of normal colon tissue cells in the crypt-villus axis. The antibodies weakly stained the highly differentiated tumor cells and moderately stained the moderately differentiated tumor cells. Of note, the antibodies intensively stained the clustered stem cells of the poorly differentiated tumor cells. These antibodies also clearly stained the clustered stem cells of poorly differentiated tumors but were not specific as they clearly stained cells in the crypt-villus axis of the normal colon wall. Our results show that LGR5 antibodies can serve as a reliable marker for colon cancer stem cells. Once the colon stem cells are identified, the targeting of specific drugs to kill these cells should be attempted in the future in order to cure this disease. Moreover, the fact that we did not find any stained cells with antibodies to LGR5 in normal tissues apart from a few scattered cells, suggests that the normal colon stem cells differ from the tumor stem cells at least as regards the expression of this protein. In addition, antibodies to γ-synuclein, p53 and KRAS only stained the tumor stem cells and not the normal tissue. Thus, they can serve as multiple biomarkers for the localization of colon cancer stem cells by indirect immunofluorescence.

Differential staining of γ synuclein in poorly differentiated compared to highly differentiated colon cancer cells.

Synuclein α, ß and γ are proteins usually found in neurodegenerative diseases. However, interestingly synucleins are expressed in cancer cells of several organs including ovary, mammary gland and colon. By immunocytochemistry using specific antibodies to γ synuclein (SNCG), we examined the distribution of this protein in poorly differentiated, compared to highly differentiated colon cancer cells. In poorly differentiated cancer cells tumors were very frequently stained intensely with antibodies to SNGG, suggesting high expression of this protein. In contrast, in highly differentiated cells, there was no labeling. Labeled cells could be found only at the edges or in between the lobules of the differentiated tumor cells. However, in moderately differentiated tumors, a weak cytoplasmic staining of SNCG was evident. Interestingly in cancer patients (stage II-IV) both poorly and highly differentiated tumor cells were often present in the same patient. Labeled cancer cells with SNCG were evident also in lymph nodes, around the wall of blood vessels and in fat tissue, where only poorly differentiated cancer cells were exclusively present. Since cancer cells with poor differentiation are believed to be aggressive with metastases formation it is suggested that SNCG can serve as a marker for the potential of the tumor cell for the rapid spreading and metastazing of the non-differentiated tumors.

Two initiation sites of early detection of colon cancer, revealed by localization of pERK1/2 in the nuclei or in aggregates at the perinuclear region of tumor cells.

We have used human specimens and antibodies to pERK1/2 to detect early development of colon cancer, using indirect immunocytochemistry. Two distinct sites were stained; one at the tip of the colon villi and the other in the stromal tissue, associated with the colon tissue. These foci represent early stages of colon cancer initiation, as established by enhanced KRAS, and lack of p53 staining. It should be noted, however, that the enhanced KRAS coincides with the initiation of tumor growth revealed by pERK1/2 only in the tip of the colon villi but not in the stromal initiation site of the colon tumors. Interestingly, foci of pERK1/2 staining were also detected within 50% of stromal tissue and tips of colon villi, that were classified as normal tissues, distal from the malignant one according to general morphology. The staining of pERK1/2 at the stromal foci of this apparently non-malignant tissue appeared as aggregates at the perinuclear region, while at the colon epithelium, it appeared at the cell nuclei. At low-grade of colon cancer, that was still free of induced mutated p53, staining of pERK1/2 was prominent at the cell nuclei both at the stroma tissue and the tip of the colon villi. In intermediate stage, that exhibited a significant p53 staining, only a fraction of p53-free tumor cells was labeled with pERK1/2 antibody, while in high-grade tumors, all cells of tumors were labeled with antibodies to p53, but not with pERK1/2. We also found that the cytoplasm of low-grade tumors was positive for epiregulin, while this labeling decreased in high-grade tumors. Interestingly, we found that the tumors initiating from the stroma demonstrated poor structural differentiation, while the tumors initiating from the epithelial cells of the colon demonstrated high structural differentiation. It is concluded that pERK1/2 is a sensitive early marker of colon cancer, which disappears at later stages of cancer development. Moreover, pERK1/2 staining can distinguish between tumor cells originated from the tip of the colon villi and those originated in the stroma, associated with the colon tissue, and thus can assist in selecting the appropriate therapy.

Epiregulin as a marker for the initial steps of ovarian cancer development.

Epiregulin (Ep) was found to be produced in non-cancer ovarian cells in response to gonadotropin stimulation as well in ovarian cancer cells in an autonomous manner. However, there were no systematic follow-up studies of Ep expression in the development of different stages of ovarian cancer. Using specific antibodies to Ep and the indirect immunocytochemistry methods, we found that in normal ovary the staining for Ep was mainly confined to the epithelial cells, while the stromal cells were only occasionally and moderately stained. In contrast in benign serous and mucinous tumors most of the tumor cells showed a clear staining in the cytoplasm. In borderline serous and mucinous tumors the staining was much more intensive, and appear occasionally in aggregated form. In serous, mucinous and endometrioid carcinomas labeling remain high, with more frequent aggregated form. It is suggested that follow-up of the expression of Ep can serve as a reliable early indication of the development of ovarian cancer. Moreover, the cytoplasmic aggregation of Ep may suggest a specific mechanism of the release of this growth factor to the extracellular space in order to exert its autocrine and paracrine effect on the family of the EGF receptors.

Nuclear localization of phosphorylated ERK1 and ERK2 as markers for the progression of ovarian cancer.

We examined the possibility that the localization of phosphorylated ERK1 and ERK2 (pERK1/2) can serve as a marker for the development of benign and borderline tumors as well as carcinoma of the ovary by an immunohistochemical method on ovarian paraffin sections, obtained from women aged 41-83 years. In normal tissue, 28.3% of nuclei were labeled, mainly confined to the epithelial cells at the surface of the ovary. In benign serous tumors, the label rose to 55.0%, while the intensity of the staining was weak. In contrast, in borderline serous tumors and in ovarian serous carcinoma (stage II) 52.1% and 70.3% of nuclei, respectively, were labeled with a high intensity. In mucinous benign tumors, the number of labeled nuclei was as in the control, but in addition, 49.4% of the cells demonstrated high concentration of pERK1/2 in aggregated form that was evident in the cytoplasm of the cells. In the mucinous and endometrioid ovarian carcinomas (stage II) very intensive labeling was found in 60% and 77.3% of cells, respectively. It is, therefore, suggested that since nuclear pERK1/2 can be mitogenic, it can serve as a reliable marker for the progression of ovarian cancer. Interestingly, the intense labeling of pERK1/2 was mainly confined to the peripheral areas of ovarian endometrioid carcinoma (stage II). In addition, all tumor cells in this class of cancer were positively stained with mutated p53. It seems, therefore, that immunohistochemical staining of normal and ovarian tumor cells with anti-pERK1/2 is a reliable marker for early detection of the cancer, which may assist in the early diagnosis and prognosis of this lethal disease.

[Novel role of growth factors in ovary function].

The development of the DNA microarray technique facilitated systematic studies of the modulation of gene function. Considerable attention has been focused on members of the growth factor family to elucidate the main regulators of oocyte maturation and ovarian follicle rupture. Among these growth factors, it was found, both in rodents and in humans, that amphiregulin (Ar) and epiregulin (Ep) of the epidermal growth factor (EGF) family were dramatically up-regulated by gonadotrophins in the intact ovary and in primary granulosa cells, respectively. Their role in cumulus expansion and oocyte maturation was established in rodents, and their synthesis under LH stimulation in granulosa cells was demonstrated in humans. To be activated, Ar and Ep must be cleaved by a disintegrin and metalloproteinases (ADAMs) family. However, the precise processing of Ar and Ep by the cumulus cells is still obscure. Future investigations using DNA microarray technique may reveal the repertoire of genes activated in Ar- and Ep-stimulated cumulus cells and may help elucidate the molecular basis of ovulation. EFG-like factors are also involved in triggering ovarian cancer The author hypothesized that the normal ovary maintains cyclicity in the formation of these growth factors preventing the ovary from developing ovarian cancer In ovarian cancer these growth factors are continuously formed in an autocrine manner, leading to transformation and subsequently to ovarian cancer. These growth factors are essential for both normal and neoplastic transformation of the ovary. Taking into consideration these growth factors in the treatment of ovarian malfunction may be one way of curing ovarian cancer.