Our research is focused on understanding of cancer pathogenesis with a particular attention to the integrative assessment of biological consequences of genetic alterations in the context of cell lineage development and cellular interactions with microenvironment.

Embryonic cells from 129/Ola mice used in our lab. Click image to enlarge

Success of such studies greatly depends on the availability of comprehensively characterized immunocompetent mouse models that accurately mimic human cancers. Therefore, we have established a series of genetically defined mouse models of cancers associated with alterations of p53 and Rb tumor suppressors and their pathways. The rationale for choosing these genes is two-fold: (1) alterations in p53 and Rb and/or their pathways occur in over 80% of all human cancers and (2) multiple functions of p53 and Rb have been extensively described and characterized at the molecular and cellular levels providing an appropriate starting point for our studies. Conventional knock out of the Rb gene leads to embryonic lethality, while mice with a single copy of wild-type Rb (Rb+/-) and p53-deficient mice succumb to rapidly progressing neoplasms with a limited relevance to major human cancers associated with deficiency of these genes. Thus, mouse models allowing cell type-restricted conditional inactivation of p53 and Rb tumor suppressor genes have been either generated by our lab or received from other investigators.

Our main areas of interest include understanding the role of stem cell compartment in carcinogenesis, studies of epithelial ovarian cancer pathogenesis and modeling metastasis. We are also interested in pursuing technology-oriented research based on cross-disciplinary collaborations are described below.

Understanding the role of stem cell compartment in carcinogenesis. A significant body of information exists about contribution of stem cells to hematopoietic malignancies. However, solid neoplasms have been difficult to study. Using prostate epithelium-specific inactivation of p53 and Rb, we have developed a new autochthonous mouse model of metastatic prostate cancer. In this model neoplasms exhibit features of both luminal and neuroendocrine differentiation and are marked with multiple signature gene expressions commonly found in human prostate carcinomas. Intriguingly, all malignant neoplasms arise only from the proximal region of prostatic ducts, the compartment highly enriched for prostatic stem/progenitor cells. Our preliminary observations indicate that synergistic effects of p53 and Rb alterations on prostate carcinogenesis are particularly effective in the context of the stem cell compartment. Further studies should address specific roles of p53 and Rb in development, maintenance and malignant transformation of the prostate stem cell compartment. We have also recently established new models of breast and soft tissue neoplasia which should be very useful to studying the relationship between stem cell biology and malignant transformation.

Studies of epithelial ovarian cancer pathogenesis. Epithelial ovarian cancer (EOC) is the 5th leading cancer type among cancer-related deaths in women in the United States. Until recently, available models of EOC were based on either transplantation systems or non-mammalian models such as egg-laying aged hen. Thus, comprehensive evaluation of epithelial ovarian carcinogenesis in the context of immunocompetent mammals was impossible. Due to its symptomless development and the lack of accurate animal models, EOC pathogenesis remains among the least understood of all major cancers.

Mouse epithelial ovarian carcinoma induced by conditional inactivation of p53 and Rb Click image to enlarge

Our laboratory developed an approach for introduction of defined genetic alterations exclusively into the ovarian surface epithelium (OSE) in situ in time-controlled manner. We have taken advantage of the enclosed anatomical location of the mouse ovary within the bursa, which allows for selective exposure of the OSE to inducing agents and established a technique of trans-infundibular intrabursal administration of replication-deficient recombinant adenovirus expressing Cre-recombinase. Using this approach in combination with Cre-loxP mediated gene inactivation in mice with floxed p53 and Rb, we demonstrated that OSE-restricted p53 and Rb inactivation leads to epithelial ovarian carcinogenesis in 97% of mice. Importantly, mouse EOC closely resembles human serous adenocarcinoma of the ovary. Similar to progression in human counterparts, ovarian neoplasms spread intraperitoneally, form ascites, and metastasize to the contralateral ovary, the lung and the liver. Thus, the first genetically defined model of sporadic EOC developing in immunocompetent animal has been established. This model directly proves the capacity of alterations in p53 and Rb-mediated pathways to cause EOC and is particularly useful for modeling of postnatally induced carcinogenesis.

Using this model we have recently determined that p53 transcriptionally regulates expression of microRNAs family miR-34. MicroRNAs (miRNAs) are a recently discovered class of non-coding RNAs which control gene expression either by degradation of target mRNAs or, more commonly for animal miRNAs, by posttranscriptional repression in a mechanism similar to siRNA-mediated gene silencing. Notably, among computationally predicted effectors of miR-34 are Ezh2, Met, cyclin D1, Cdk4, Cdk6, Cdk7, and E2F3, indicating existence of novel mechanisms of p53-mediated regulation. We are currently pursing evaluation of miR34 functions in ovarian cancer and identification of critical downstream targets.

Modeling metastasis. Metastatic progression is the most lethal feature of advanced cancer. Currently available models of metastasis are mainly based on either xenograft systems or mice with generalized transgene overexpression.

Chimeric mouse prepared by Andrea Flesken-Nikitin

As discussed above, we have developed a new genetically defined mouse model of metastatic prostate cancer associated with deficiency for p53 and Rb pathways. In that model cancer progresses from prostatic intraepithelial neoplasia to invasive adenocarcinoma and gives rise to extensive metastasis in the liver, the lung and the regional lymph node, sites commonly targeted in the human disease. This model is at clear advantage to such surrogate substitution approaches because (1) metastatic process occurs as advanced stage of carcinogenesis associated with deficiency of prototypical tumor suppressor genes; (2) it develops spontaneously with high frequency and synchronicity; and (3) metastasis takes place in immunocompetent mice. Importantly, this model recapitulates most aspects of human prostate cancer at the molecular level, including expression of both epithelial and neuroendocrine markers and should be particularly useful in gaining further understanding of the processes that govern metastasis in human prostate cancer.

Technology-oriented research based on cross-disciplinary collaborations. It is our strong conviction that advances in modern life sciences greatly depend on the development of new technologies resulting from successful interactions among investigators with diverse disciplinary backgrounds. At Cornell University we have been fortunate to set-up a number of collaborations allowing us to apply the cutting edge photonics and nanofabrication to such challenging areas of cancer research as molecular imaging and targeting. Together with Drs. Warren Zipfel and Rebecca Williams (Biomedical Engineering), and Watt Webb (Applied Physics) we are developing and assessing nonlinear optical techniques for the early detection of cancer. Together with Dr. Ulrich Wiesner (Materials Science and Engineering) we are evaluating biodistribution and diagnostic value of CU dots as a safer alternative to quantum dots.