Saraswati Sukumar, Ph.D.

The research goal of the Sukumar laboratory is to obtain a molecular profile of breast cancer and to apply this knowledge to the early detection, diagnosis and therapy of breast cancer. Toward this end, SAGE, array, and proteomic analysis of breast carcinomas is being performed on noninvasive (ductal carcinoma in situ), invasive (invasive ductal carcinoma), and metastatic breast cancer (bone, ovarian, brain metastasis). The goal is to obtain mRNA and protein profiles of each cell type composing the lesion. This will help derive gene expression signatures typical for each stage of breast cancer for epithelial, stromal, and endothelial cells. By such careful analysis, candidate genes are identified, characterized and utilized for early detection and/or therapy of breast cancer.

Previous gene searches in breast cancer performed here identified a number of potential candidate tumor suppressors and oncogenes. These belonged to families of genes involved in a variety of functions in the cell. Transcription factors, such as HOXA5, WT1, RAR-beta and Twist; cell cycle regulators such as Cyclin D2 and 14.3.3 sigma; and tight junction proteins, such as Claudin-3, -4 and -7, were selected for further analysis. Biological analysis of these genes revealed molecular alterations that could lead to their aberrant function in breast cancer cells, results of which were published in the last three years. Hypermethylated promoter regions of a number of genes serve as tumor- specific markers for early detection of breast cancer. As described in a paper in Lancet in 2001, a marker panel consisting of Cyclin D2, RAR-beta and Twist was capable of detecting almost all breast cancers (96 percent) with a high level of specificity and sensitivity. This marker panel detected cancer in two high-risk women who were mammographically normal. This panel has now been improved by adding more markers, such as RASSF1A, HIN-1, ER-a and ER-b, and by developing a multiplex, quantitative methylation-specific PCR (Q-MSP). Q-MSP will allow accurate determination of the relative amount of methylation present in each gene, using less than nanogram amounts of DNA, in an objective manner. To test this premise, these markers are currently being applied to ductal cells obtained in two current trials aimed at validating the markers for early detection in high-risk women.

Soon, this panel will be tested as intermediate- response markers in a chemoprevention trial in collaboration with three other institutions in the United States. A combination of an objective test, such as MSP, with a subjective test, such as cytology, has the potential to provide an adjunct to mammography for early detection of breast cancer. This test is also being modified to detect tumor-specific DNA in serum from breast cancer patients.

The long-term goal is to provide a molecular blood test for early detection of breast cancer. Interrogating the hypothesis that breast cancer can be treated and prevented by delivery of agents via the intraductal route, both conventional and novel agents were tested. First, treatment of rats bearing carcinogen-induced tumors with Doxil results in complete regression of the tumors. Doxil administration via this route in a prevention setting, soon after carcinogen administration, also protected 90 percent of the treated glands. This mode of delivery can achieve therapeutic effects with less than one-third the systemic dose of the drug. This concept has also been tested in the HER2/neu mouse model for breast cancer. Here again, potent antitumor effects on established tumors were achieved by intraductal instillation of Doxil. Doxil also prevented the development of mammary tumors in this system. With the goal of eliminating the primary tumor and its metastasis, chemotherapy is being combined with immunotherapy (in collaboration with the Jaffee laboratory). Targeting genes that were found by our previous SAGE analysis searches to be overexpressed in breast cancer (the Claudins 3 and 4), we are testing therapies using Clostridium perfringens enterotoxin (CPE). CPE binds to the overexpressed tight junction proteins Claudin 3 and 4 on the tumor cells, resulting in tumor cell kill. No apparent cytotoxic effects are observed on normal epithelial cells. Additionally, the systemic toxic effects of CPE on the gastrointestinal tract are bypassed by utilizing the intraductal route of delivery. A combination of basic and preclinical research efforts in the laboratory aims at translating discoveries made at the bench to the clinic as rapidly as possible.