Recent advances in microarray technologies promise more precise molecular classification of tumors based on gene expression profiles, which will allow a more well-defined basis for prognosis and therapeutic strategies. In order to take full advantage of this emerging technology, characteristic gene expression profiles or """"""""molecular signatures"""""""" must be identified in tumors and associated with specific disease states as well as responses to particular therapies. However, the known heterogeneity of many cancers, combined with the thousands of genes available for interrogation by microarray technologies, make evident the need for more rational approaches to identifying new molecular signatures of clinical relevance. We and others have demonstrated that certain members of the Signal Transducer and Activator of Transcription (STAT) family of proteins, particularly Stat3, have important roles in human cancer. Constitutive activation of Stat3 in human tumors is predicted to induce a permanent alteration in global gene expression patterns associated with distinct molecular signatures. Our central hypothesis is that Stat3-regulated genes define characteristic Stat3 molecular signatures of clinical relevance in human cancers. To test this hypothesis, we will use gene expression profiling by oligonucleotide microarray techniques applied across species (mouse and human) and across tumors (breast and prostate cancer), in combination with a variety of conventional and novel data analysis methods. By integrating with ongoing clinical trials at Moffitt Cancer Center, our specific aims address the following critical questions relating to our central hypothesis. (1) Can we identify global gene expression patterns associated with Stat3 activation in transformed mouse cells that predict universal elements of Stat3 molecular signatures in human tumors? (2) Are there Stat3 molecular signatures associated with specific molecular subtypes of human breast cancer that predict response to chemotherapy? (3) Can Stat3 molecular signatures be used to predict recurrence in prostate cancer patients following prostatectomy? These studies will not only reveal molecular subtypes of tumors with well-defined clinicopathological characteristics but also provide new insights into the mechanistic basis of cancer.
Showing the most recent 10 out of 69 publications
