Vignette 1 (Kevin T. Nash, Pushkar A. Phadke, Andra R. Frost, Douglas R. Hurst, Natalya Frolova, John C. Kappes, Yujiang Jia, Stephen Barnes): The primary obstacle to cancer cure is development of distant metastases. At the time of diagnosis, tumors have shed many cells into the circulation rendering anti-metastatic therapies moot since the cells have already seeded other tissues. Recent findings with metastasis suppressors indicate that tumor cells expressing KISS1 are able to complete every step of the metastatic cascade except proliferation at the secondary site. Tumor cells inoculated at orthotopic sites (i.e., intradermal) form progressively growing tumors that shed cells which remain dormant in the lungs. Tumor cells inoculated intravenously seed the lungs, but fail to establish macroscopic lesions. The nascent KISS1 protein (~17 kDa) has never been detected while a secreted internal 54 amino acid polypeptide (termed metastin or kisspeptin-54 (KP54)) had previously been isolated from placenta. To assess whether KISS1 or KP54 secretion is necessary for metastasis suppression, an internally FLAG-tagged KISS1 was constructed with (designated KFM) or without (designated ÄSS) the putative signal peptide, transfected into metastatic human C8161.9 melanoma cells and evaluated for tumor growth and metastasis following orthotopic or intravascular injection. ÄSS expressing cells no longer produced detectable KISS1 or KP54 in culture media. Moreover, ÄSS-transfected cells were as metastatic as parental cells, while KFM-transfected cells were suppressed for metastasis. Both still allowed local tumor growth. Media was isolated from cells expressing KFM, immunoprecipitated using anti-KISS1 and anti-FLAG antibodies, analyzed by mass spectrometry and internal sequencing and verified to be polypeptides derived from KISS1. Therefore, secretion of KISS1 is necessary for its anti-metastatic effects.Efforts are underway to determine whether processing is also required. As a secreted molecule that impacts metastasis, KISS1 (products) are likely drug-able and may be useful for inhibiting colonization of tumors at secondary sites.

Vignette 2 (Kedar S. Vaidya, Pushkar A. Phadke, Sitaram Harihar, Daryll B. DeWald, Graham Casey): The BRMS1 metastasis suppressor associates with histone deacetylase complexes which, in turn, control gene expression. Recently, data suggest that BRMS1 selectively regulates signaling through phosphatidylinositol-4,5-bisphosphate and NFêB. We hypothesized that BRMS1 expression determines which microenvironments are permissive/restrictive for cellular survival and growth, perhaps explaining why BRMS1 expressing cells grow at orthotopic, but not at ectopic sites. MDA-MB-231/435 human breast cancer cells and BRMS1 transfectants (231BRMS1 /435BRMS1) were exposed to epidermal growth factor (EGF, 0-50 ng/ml) and downstream effectors were studied by immunoblot, calcium mobilization and F-actin organization. EGF stimulated AKT-Ser473 phosphorylation in 231 and 435 but phosphorylation was reduced > 50% in 231BRMS1 and was completely abolished in 435BRMS1. EGF failed to modulate p42/p44 MAPK phosphorylation in 231 and 231BRMS1; whereas, 435BRMS1 showed marked reduction in MAPK phosphorylation. 231BRMS1 had decreased EGFR mRNA and protein while 435BRMS1 EGFR levels were undetectable. 231BRMS1/435BRMS1 did not mobilize calcium or reorganize F-actin following EGF or PDGF treatment. Interestingly, 231BRMS1/435BRMS1 cell expression of PDGFR was unchanged, indicating a deficit in downstream signaling. 231BRMS1/435BRMS1 cells were attenuated for Akt, but not ERK, response to insulin treatment, while TPA treatment showed attenuated ERK signaling without alteration of Akt signaling. Thus, BRMS1 differentially attenuates responses to mitogenic signals by alternation of cell surface receptors and downstream pathways.

The Women’s Cancers Section performs basic and translational research on the molecular biology of breast cancer. Three major projects are under investigation: (1) What is the role of the nm23 metastasis suppressor gene in breast cancer progression? (2) What molecular events are involved in breast cancer metastasis to the brain? (3) What molecular events are responsible for early neoplastic progression in the breast?

The nm23 family of genes was discovered by Dr. Steeg on the basis of its reduced expression in highly metastatic murine melanoma cell lines, as compared to related, poorly metastatic cell lines. Eleven transfection studies have documented that overexpression of nm23 cDNA in metastastic cell lines results in a significant decrease in metastatic potential in vivo, without an effect on tumorigenicity, establishing nm23 as a metastasis suppressor gene. Increased Nm23 expression induced morphological and functional differentiation in transfection studies involving breast and neuroendocrine cells. Reduced Nm23 expression has been correlated with poor patient survival and the presence of metastases or other histopathological indicators of aggressive clinical course in breast and other cohort studies, although it does not represent an independent prognostic factor.

We have investigated the biochemical mechanism of action of Nm23-H1 in suppressing metastasis. Site directed mutation experiments showed a correlation between the histidine protein kinase activity of Nm23-H1 and its ability to suppress motility in vitro. We have recently identified a physiologic substrate for Nm23-H1 as a histidine protein kinase, the kinase suppressor of ras (Ksr). Ksr is a putative scaffold protein for the Erk arm of the Map kinase pathway. Using MDA-MB-435 breast carcinoma cells, we showed that Nm23-H1 overexpression altered Ksr binding of Hsp90, Ksr degradation and tumor cell sensitivity to geldanamycin inhibition of colonization. The data indicate that common signaling pathways may be affected by the expression and activity of metastasis suppressor genes.

Translational projects are under way in the section based on the hypothesis that elevation of Nm23 expression in micrometastatic breast and possibly other tumor cells may impact their colonization, motility, and differentiation with a clinical benefit. Analysis of the nm23-H1 promoter revealed its regulation by a set of mammary specific transcription factors, controlled in the MMTV-LTR by glucocorticoids. While traditional glucocorticoids were ineffective at elevating the Nm23-H1 expression of metastatic breast carcinoma cell lines, medroxyprogesterone acetate (MPA), through an unusual post-translational, glucocorticoid receptor based interaction, elevated cell line Nm23-H1 expression and inhibited colonization in vitro. In vivo analysis of MPA is underway.

Brain metastases occur in at least 15% of breast cancer metastatic patients and confer a dismal prognosis. We have undertaken a microarray analysis of surgically resected brain metastases of breast cancer, and compared these data to a cohort of independent primary breast tumors matched for hormone receptor status, patient age and TNM stage. Our preliminary data indicate that brain metastases are distinct from primary tumors and that translationally important targets are apparent.

The development of preventive strategies for women at high risk for breast cancer will require a “molecular map” of the cancer progress. A new initiative is to identify proteins differentially expressed among human premalignant breast lesions using proteomics and mass spectroscopy sequencing. Fifty-seven proteins were identified in a proteomic comparison of normal ductal/lobular units and DCIS, many of which were new to the breast cancer literature. These proteins suggest that breast cancer exhibits massive alterations in subcellular trafficking of lipids, proteins, putative prevention compounds, ions, etc. Functional analysis of Rab 11 expression is underway and indicates a functional impact on the EGF receptor pathway.

Information to be posted soon.

Metastasis – the spread of cancer from its site of origin to distant, vital organs – is responsible for most cancer deaths, and can occur years after apparently successful treatment of the primary tumor, due to metastatic dormancy. Metastasis is not just an end-point; it also is a process, with potential treatment opportunities at various points along the process. However, the biology of metastasis and dormancy remain poorly understood. A better understanding of these processes will be important for improved ways to prevent or treat metastatic disease. We are using various imaging modalities including in vivo video-microscopy, coupled with quantitative analyses of the fate of cancer cells during the metastasis process, to study metastasis and dormancy in experimental animal models. The insights gained from these basic biology studies are providing new ways to look at metastasis, its prevention and its treatment.

The cause of death for the vast majority of cancer patients is the development of metastases at sites distant from that of the primary tumor. For pediatric osteosarcoma patients this pattern of progression is unfortunately predictable. Despite successful management of the primary tumor through multimodality approaches the development of metastases, most commonly to the lungs, is cause of death for the majority of ostesoarcoma patients. We believe that opportunities to improve outcomes for patients who present with metastases and those at risk for metastatic progression require an improved understanding of metastasis biology.

Our approach to the problem of metastasis is founded in the development, characterization and use of a diversity of relevant in vivo (animal) models of metastasis. These models include transplantable syngeneic murine tumor models, genetically engineered mouse models, human-mouse xenograft models and naturally occurring cancers that develop in pet animals. The use of several model systems to study problems in metastasis provides an opportunity to emphasize strengths and minimize weaknesses of any single model system.

To extract greatest value from our modeling efforts in metastasis we have integrated existing and novel imaging strategies to the use our approach. Accordingly, we have become “consumers” of several imaging strategies and have begun to define the types of questions that are best answered within each model and imaging strategy. This work includes the modification of single cell fluorescent and nano-particle imaging to understand early events occurring during metastasis, the use of biolumincent and MRI studies to follow cancer progression and response to therapy, and the development of targeting imaging approaches that may be translated from the laboratory to the clinic. Examples of our use of this multi-model approach to study the biology of osteosarcoma metastasis will be discussed.

The recent convergence of technologies for expression profiling and intravital imaging has revealed the identities of the genes involved in the survival, adjuvant-resistance and chemotaxis of invasive cancer cells inside living tumors. These genes fall into well defined pathways and are coordinately regulated in invasive tumor cells. This pattern is called the invasion signature. The invasion signature indicates that invasive cancer cells are a population that is neither proliferating nor apoptotic but highly chemotactic to macrophage-secreted EGF. Of particular relevance to the migratory behavior of invasive cancer cells is the finding that the genes coding for pathways leading to the minimum motility machine, i.e., the cofilin, capping protein and Arp2/3 pathways, that regulate -actin polymerization at the leading edge, and the directionality of cell protrusion during chemotaxis to EGF, are coordinately up-regulated. This latter result is particularly relevant to the contribution of the tumor microenvironment to metastasis because chemotaxis to blood vessels is involved in the escape of cancer cells from primary mammary tumors. Key genes in the invasion signature have been studied for their ability to alter metastatic outcome and these results confirm the importance of the invasion signature in predicting metastasis. The invasion signature predicts that invasive tumor cells, upon escaping the primary tumor, are non-dividing and resistant to chemotherapy, predictions that have been confirmed in vitro. Finally, the invasion signature provides several new target opportunities for preventing metastasis and testing of these will be discussed.

Information will be posted soon.

The overall goal of our research is to define the cellular and molecular determinants of the step-wise pathways mediating epithelial carcinogenesis. Our research is based upon the premise that, in addition to intrinsic changes occurring within neoplastic cells, e.g., activation of oncogenes and inactivation of tumor suppressor genes, extrinsic factors, e.g., inflammation, extracellular matrix (ECM) remodeling and angiogenesis also regulate critical properties of tumor evolution.

The model system we study is a transgenic mouse model of squamous cell carcinoma (SCC) development where human papillomavirus type 16 (HPV16) oncogenes are expressed in mitotically active basal keratinocytes in skin epidermis, e.g., K14-HPV16 transgenic mice (Coussens et al., 1996). While HPV16 oncogenes act as ‘initiators’ of neoplastic progression in this model, we have identified ‘host’ inflammatory cells, e.g., mast cells and neutrophils (Coussens et al 1999; deVisser et al 2004, 2005), and specific stromal-derived extracellular proteinases, e.g., matrix metalloproteinases (MMPs) (Coussens et al., 2000; vanKempen et al., 2002), chymase/mMCP-4 (Coussens et al., 1999), tryptase/mMCP-6 (Coussens et al., 1999) and their endogenous tissue inhibitors (Rhee et al., 2004) as critical extrinsic regulators of epithelial neoplastic evolution. By genetically modulating the host immune response and/or presence of various extracellular proteinases with specific gene knockouts and double transgenic mice, we have been able to manipulate the temporal dynamics, overall incidence and malignant potential of carcinomas that develop in HPV16 mice. Our findings that inflammatory leukocytes contribute in a dominant manner to the activation of neoplasia-associated angiogenesis functionally link inflammation as a potentiator of the neoplastic process (Coussens et al., 2000; van Kempen et al., 2002; de Visser et al, 2005). More recently, we demonstrated that B lymphocytes are required for establishing chronic inflammatory states and promotion of de novo epithelial carcinogenesis (de Visser et al., 2005). Adaptive immune-deficiency in HPV16 mice results in failure to recruit innate immune cells into premalignant tissue, and as a consequence, responding pathways downstream of inflammatory cell activation, e.g. tissue remodeling, angiogenesis, keratinocyte hyperproliferation and cancer development, are significantly attenuated. Importantly, these necessary characteristics of premalignant progression are restored by transfer of B lymphocytes or serum from HPV16 mice into T and B lymphocyte-deficient/HPV16 mice (deVisser et al, 2005), indicating that B lymphocytes play a crucial role as initiators of chronic inflammation associated with premalignant progression, and thus potentiate the neoplastic cascade downstream of oncogene expression. While there have been many studies on the role of B lymphocytes and/or immunoglobulins as mediators of acute inflammation, this is the first to provide clear and convincing functional proof, via the combined use of gene knock out mice, adoptive B cell transfer and serum transfer, to demonstrate that B cells play such important roles in epithelial carcinogenesis, and particularly at the earliest stages of carcinogenesis, during premalignant progression. Altogether, these studies indicate that molecules downstream of inflammation and/or protease action play fundamental roles in discrete stages of tumor development.

Based on our data implicating inflammatory leukocytes and their products, e.g., proteases, as important regulators of epithelial carcinogenesis, we hypothesized that altering the susceptibility of critical MMP substrates to proteolytic cleavage might similarly alter neoplastic development. Thus, we tested whether the equilibrium between synthesis, accumulation and degradation of a key MMP target, e.g., type I collagen, was central for the cascade of cellular and genetic changes necessary to manifest carcinoma development and/or metastasis in HPV16 mice (vanKempen et al., Manuscript in preparation). Our studies reveal that both neoplastic progression to SCC and metastatic capability of SCCs is significantly reduced in an environment where matrix remodeling is restricted. In combination, HPV16/immune altered, HPV16/protease altered and HPV16/protease-substrate altered mice have allowed us to test the hypothesis that the ‘host response’ to neoplastic progression represents a fundamental aspect of carcinogenesis and reveals important new targets for therapeutic intervention.