Development and application of proteomic tools for the a

Investigators

Abstract

Laser-capture microdissection (LCM), invented by Dr. Lance Liotta at the NCI, has enabled researchers for the first time to be able to analyze purified populations of cells directly from patient tissue samples. These populations include, but are not limited to, cancer epithelial cells- cells which comprise only 1-2% of the total population of cells found in any non-hematopoetic derived tumor (the rest being stromal cells, fibroblasts, infiltrating lymphocytes, etc.). In a close collaborative effort, I have become the co-director of the FDA-NCI Clinical Proteomics Program. This program has invented and developed several key enabling technologies for the proteomic analysis of cancer cells from a variety of tissue types (prostate, breast, lung, esophageal, lung, brain, ovary, colon) derived from laser-capture microdissected cells. These cells have been analyzed via 2D-PAGE protein fingerprinting, signal pathway profiling, phosphoprotein analysis, SELDI protein chip retentate mapping analysis, multiplexed scanning immunoblot analysis. Additionally, serum and body fluid tumor marker identification is being performed via SELDI and 2D-PAGE analysis. We have found that we can successfully recover and analyze proteins from the LCM and have identified over one hundred proteins from ovarian, breast, lung, colon, prostate and esophageal cancer that are differentially regulated (turned off or turned on). Over 450 proteins have been identified so far that track with the malignant phenotype. We are validating these now using a novel tool we have invented using high-throughput microarray-based lysates, antibody arrays, and phosphoprotein pattern profiling. We have now taken these tools and are implementing proteomic tools in actual human clinical trials to analyze the signaling pathways of cells form biopsy before during and after targeted therapeutics. Moreover, we have successfully employed high throughput mass spectrometry for LCM derived cell, and body fluids analysis/biomarker discovery. Through a CRADA with Correlogic Systems, Inc. we have developed a novel artificial intelligence-based pattern for diagnostic analysis of cancer from high dimensional high-throughput mass spectrometry. Highly accurate patterns have been uncovered that can detect early stage lung, breast, pancreatic, colon and prostate cancer. We are currently using this approach to analyze serum from ovarian cancer patients and found we can detect early stage cancer nearly 100% of the time. Moreover, we have been able to use this approach for detection of early stage drug induced toxicity.