Localization of activated STAT3 in mouse ovarian cancer cells

The signal transducer and activator of transcription-3 (STAT3) is frequently constitutively activated in EOC cell lines and primary human tumors. STATs are cytoplasmic transcription factors that mediate signals from cytokines, growth factor receptors and non-receptor tyrosine kinases. Activation of STAT3 results in induction of genes involved in proliferation, apoptosis and angiogenesis. While these pathways are clearly important in cell transformation and tumorigenesis, numerous studies have also demonstrated a role for STAT3 in motility, migration and invasion in both normal cells and tissues and in cancer cells. With specific regard to ovarian cancer, recent studies have shown that expression of constitutively activated (phosphorylated) STAT3 (pSTAT3) is correlated with advanced stage invasive primary EOCs and that inhibition of STAT3 results in decreased motility of EOC cell lines in vitro. Unexpectedly, these studies showed that pSTAT3 localized not only to nuclei in accordance with its role as a transcription factor, but also to focal adhesions in EOC cells, co-localizing with focal adhesion kinase (FAK), Src and paxillin. Our own studies have confirmed these results, and further suggest that pSTAT3 plays a critical role in focal complex signaling in EOC cells. The functional significance of focal complex localization of pSTAT3 is unclear, but these structures are strongly associated with cellular movement, further supporting a role for pSTAT3 in pro-metastatic functions such as motility and invasion. Notably, STAT3 deficient keratinocytes have migration defects associated with aberrant phosphorylation of the focal adhesion adapter protein p130Cas. Moreover, STAT3-mediated transcriptional activation results in expression of genes involved in motility and invasion. For example, during zebrafish gastrulation, STAT3 induces LIV-1, which is required for nuclear localization and activity of SNAIL, a factor involved in migration and invasion of epithelial cells. In cancer cells, STAT3 has been shown to induce expression of MMPs. In addition to its canonical role as a transcription factor, STAT3, like other transcription factors (e.g., p53 and β-catenin), may have additional, transcription-independent functions. For example, STAT3 binds to stathmin, a protein involved in microtubule dynamics and cell motility, and inhibits its ability to destabilize microtubules; these activities are independent of both the transcriptional activity and phosphorylation state of STAT3.

Src is a non-receptor tyrosine kinase that has been shown to be activated in numerous human cancers, including EOC. Inhibition of 

Localization of activated STAT3 in human ovarian cancer cells

Src in ovarian carcinoma cells has been shown to inhibit angiogenesis and tumorigenesis in vivo in human xenograft models and in enhanced paclitaxel-mediated cytotoxicity in ovarian carcinoma cells in culture. The effects of Src activation, much like activation of STAT3, include proliferation, increased cell survival, angiogenesis, cell migration and invasion. Moreover, Src has been shown to directly activate of STAT3, in some cases acting downstream of growth factor receptor tyrosine kinases to mediate STAT activation, and STAT3 is absolutely required for v-Src mediated transformation. Thus the finding that activated STAT3 co-localizes to focal adhesions and that this co-localization depends on Src, suggests that Src-mediated activation of STAT3 in focal adhesions may play an important role in EOC. The established roles of both Src and STAT3 in motility, migration and invasion and the importance of focal adhesions in cell movement suggests that the interaction between Src and STAT3 may be an essential component of ovarian cancer cell motility and invasion, perhaps mediated by activation of MMPs.

Based on its high frequency of activation in EOCs, association with advanced stage tumors and pleiotropic activities, STAT3 may be an excellent target for therapeutic intervention in EOC. We are currently using human and murine ovarian carcinoma cell lines in both in vitro and in vivo analyses to determine if activated STAT3 directly influences ovarian cancer cell motility and invasion and if this function is Src-dependent. In addition, our studies are aimed at identifying activated STAT3-induced mediators of the migratory and invasive phenotype of EOC cells. Finally, we will determine of the effects of inhibition of STAT3 and/or its downstream mediators on migration and invasion of EOC cells in vitro and in vivo.

Co-localization of HEF1 and FAK in ovarian cancer cells

Recent studies indicate that HEF1/NEDD9/Cas-L (hereafter called HEF1) is an important mediator of glioblastoma invasion and melanoma and lung tumor metastasis. HEF1 is a scaffolding protein that has dual functions at focal adhesions during interphase and at the centrosome during cell division. At focal adhesions, HEF1 mediates the activation of proteins that promote migration and invasion, including FAK and Src. Constitutive activation of FAK and Src are common features in primary EOCs and ovarian carcinoma cell lines. Other studies have shown loss or depletion of HEF1 in lymphoid cells results in defects in chemokine signaling, migration and homing, based in part on impaired CXCL12(SDF-1)/CXCR4 signaling. This signaling pathway is also important in ovarian cancer metastasis. In addtion, HEF1 binding to Aurora A at centrosomes is necessary for the activation and stabilization of Aurora A: overexpression and activation of Aurora A have also been noted as contributing to EOC. Finally, recent preliminary data from our laboratory suggest that HEF1 mediates the localization of activated STAT3 at focal adhesions. Based on these observations we hypothesize that HEF1 plays a central role in EOC pathogenesis. The goals of this project are to: 1) Determine the role of HEF1 in focal complex signaling in normal ovarian epithelium and EOC cells, 2) Determine the requirement of HEF1 expression in ovarian cancer development and progression in vivo, 3) Determine the relevance of HEF1 expression in human ovarian tumors and 4) Validate HEF1 as a therapeutic target for treatment of EOC.

The identification and definition of cancer initiating cells (CICs) within solid tumors has become a topic of intense interest. By definition, CICs: 1) exist as a minority population within a tumor, 2) are characterized by a distinctive profile of cell surface markers that allow differential isolation from the bulk tumor population and 3) give rise to tumors consisting of mixed populations of cell types, similar in heterogeneity to the tumor from which they were initially isolated. Based on these criteria, many laboratories have reported the isolation of CICs that express various cell surface markers (e.g., ATP transporters, cell surface glycoproteins, receptor tyrosine kinases, etc.) from hematopoietic and solid tumors. Because the expression of these putative CIC markers can be somewhat inconsistent between different tumor types or even among tumors of the same organ (e.g., EOCs), the functional significance of the expression of individual and/or multiple markers to the CIC phenotype is unknown. Whether the expression of such markers is cause or consequence of the CIC phenotype also remains largely unknown. Finally, whether the CICs represent a static sub-population of cells that stably express a defined set of cell surface markers and exhibit the CIC phenotype as opposed to a population of cells that are relatively plastic, i.e., a sub-population of cells in which cell surface marker expression and phenotype may be largely context-specific remains to be determined.

Several recent studies and our own preliminary data suggest the existence of CICs in human EOC. One study identified spontaneously transformed clones of cells from a patient ascites specimen based on anchorage-independent growth as tumor spheroids that express several cell surface markers characteristic of CICs (e.g., Nestin, Nanog, Oct-4, CD44 and c-Kit (CD117)). These cells could be serially transplanted into immunocompromised mice resulting in tumors similar in histology to the original human tumor. In our own work, we have characterized an ovarian cancer ‘side population’ (SP) from human and murine ovarian carcinoma cells defined as cells that exclude the DNA binding dye Hoecsht 33342, express ABCG2 (BCRP), form measurable tumors with shorter latency than non-SP cells and are resistant to chemotherapeutic agents including doxorubicin and paclitaxel. In addition, among murine carcinoma cell lines evaluated, we found expression of CD44 and CD117 (c-Kit) in the SP. More recently, ovarian CICs were isolated from primary human EOCs based on growth as self-renewing non-adherent spheroids, resistance to cisplatin or paclitaxel, induction of stem cell markers (e.g., Bmi-1, stem cell factor, Notch, Nanog, Nestin, ABCG2 (BCRP), and OCT-4) and expression of the cell surface markers CD44 and CD117. Taken together, these studies suggest several common features of ovarian CICs including enhanced expression of drug efflux transporters, relative chemoresistance, transcriptional activation of stem cell markers and the expression of the common CIC cell surface markers CD44 and CD117. The goals of this project are to determine 1) whether CD44 and/or CD117 are critical mediators of the ovarian cancer initiating cell phenotype 2) whether inhibition of CD44 and/or CD117 in ovarian cancer initiating cells influences drug sensitivity and 3) whether anti-CD44 therapy using hyaluronan oligomers (o-HA) affects the tumorigenic properties of unselected ovarian carcinoma cells or their CD44+/CD117+ subpopulations.

Activation of Aurora A (Aur A) kinase by genomic amplification and/or increased expression is a common feature of human epithelial ovarian carcinoma (EOC) cell lines and primary tumors. A high level of Aur A expression in tumors is associated with supernumerary centrosomes and overall decreased survival. Aur A has become a target of interest for therapeutic intervention, because of its frequent activation in primary tumors and its interaction with p53, a gene commonly mutated in EOC. Aur A can exert either positive or negative effects on tumor cell growth, centrosome amplification and transformation potential depending on the functional status of p53. Although it is well-established that Aur A plays important roles in cell cycle regulation and chromosomal stability, recent work suggests novel functions for Aur A in mediating cancer cell migration, invasion and epithelial-to-mesenchymal transition (EMT). Interestingly, studies in xenograft models revealed that Aur kinase inhibitors induce mitotic arrest and apoptosis in a small percentage of tumor cells that does not fully account for the inhibition of tumor growth observed. This raises the possibility that Aur A regulation of metastatic functions and differentiation status may contribute to the regression of tumors observed in these studies. Elucidation of the mechanisms by which Aur A regulates the metastatic phenotype and epithelial cell plasticity will fill a gap in our knowledge of how this kinase contributes to ovarian tumorigenesis, and aid in the intelligent design of future therapies. The goals of this proposal are to determine the requirement for Aur A in regulating migration, invasion and differentiation status of EOC cells in vitro and to assess the efficacy of Aur A kinase-specific inhibition in suppressing ovarian tumor growth and dissemination in vivo.

Our laboratory is interested in the development and utilization of magnetic resonance, bioluminescent and fluorescent imaging techniques in vivo imaging and quantitation of tumor burden in mice.

MRI of mouse with bilateral ovarian tumors

Detection of ovarian tumor with a molecular imaging probe

Orthotopic ovarian tumor detected by bioluminescent imaging