Reasearch Highlights
Stanford University is at the forefront of developing experimental and analytical methods for the molecular profiling of human cells and tissues and in using these methods to improve cancer diagnosis and treatment. Here is a summary of recent discoveries by program researchers.
- Pioneering work by Dr. Brown and co-workers produced practical DNA microarray technology and a suite of software for genome-wide analysis of gene expression [1-6]. Dr. Brown and his colleagues originated experimental designs and analysis methods for many applications of DNA microarray technology, including the first genome-wide profiles of mRNA transcript levels [1, 4], systematic characterization of physiological and developmental gene expression programs [1, 7-41] and the effects of specific mutations on global gene expression programs [1, 10, 11, 13, 16, 17, 19, 22, 23, 29, 30, 32, 36, 42, 43], DNA copy-number alterations [44-47], binding sites of transcription factors and other DNA-binding proteins [48-52], ribosome density on mRNAs [7, 21, 53, 54], mRNA decay [38, 55], and sub-cellular localization of mRNAs [55, 56].
- Dr. Pollack and co-workers developed the first method for genome-wide mapping of DNA copy number alterations at single-gene resolution [47]. His laboratory has used array Comparative Genomic Hybridization (CGH) to systematically examine diverse cancers [44, 46, 47, 57-61] and has identified new, prognostically significant copy number alterations (Proc. Natl. Acad. Sci, 2002), a new candidate tumor-suppressor gene in prostate cancer (Cancer Res, 2005), and has directly demonstrated that gene expression profiling predicts improves outcome prediction in acute myelogenous leukemia (New Engl. J. Med, 2004).
- Using DNA microarrays, Drs. Brown, Jeffrey, Pollack, Brooks, So, van de Rijn, Tibshirani, Hastie, Lowe and colleagues have carried out large-scale systematic studies of global gene expression patterns in human breast [44, 47, 62-77], prostate [14, 57, 78-81], lung [64, 82, 83], kidney [18, 84, 85], stomach [64, 86-88], liver [89-91], pancreatic [92], testicular [37], and ovarian cancers [10, 93], as well as in glioblastoma [94], diverse soft-tissue sarcomas [46, 58, 62, 76, 88, 95-99], leukemias [100-102] and lymphomas [76, 103-110]. To enable further systematic investigations by other researchers, free and unrestricted access to the complete data from all these studies is provided by the Stanford Microarray Database [111-120].
- Dr. Dahl and colleagues are attempting to identify high-risk characteristics of children with AML to define a group of patients who do not require aggressive life threatening therapy [100, 121]. Using PAM analysis in collaboration with Dr. Tibshirani, he was able to define subsets of childhood AML samples positive for FLT-3 mutations that had defined prognostic significance. Blood, 104:2646-54. 2004.
- Drs. Chang, Brown, Pollack, Brooks, van de Rijn, Jeffrey, Dahl, Tibshirani and Hastie, using a variety of approaches, discovered and validated novel prognostic patterns of gene expression that contributed significant new information for predicting progression or metastasis of several common human cancers, including lymphoma [103, 105, 106, 110], leukemia [100], glioblastoma [94], cancers of the breast [62-66, 75], prostate [79], lung [64, 83], stomach [64, 86, 87] and ovary. These signatures have led to the classification of tumors into novel subclasses with significantly different prognoses.
- Using DNA micorarrays, Drs. Chang, Brown, Tibshirani, van de Rijn, Hastie and co-workers have discovered gene expression programs that link wound healing and hypoxia to cancer progression. They have demonstrated that the appearance of a “wound healing” signature and a “hypoxia-response” signature can independently predict both progression and metastasis in breast cancer [63, 64]. This research provides new insight to the role of the wound-response pathway and hypoxia and oxygen recruitment in the fundamental biology of solid tumors.
- DNA display technology developed by Dr. Harbury and co-workers is the first and still the only practical method for in vitro evolution of a synthetic chemical library [122-124]. This technology provides a rapid combinatorial synthesis approach combined with the ability to engage in rounds of enrichment for compounds with the correct binding activity and has tremendous potential for the discovery of new compounds with applications to detection, diagnosis and treatment of cancer.
- Dr. Zare has pioneered many experimental methods using state-of-the-art capillary electrophoresis, mass spectrometry and microfluidics to quantitatively analyze the molecules present in a single cell. Most relevant to this program has been his continuing work on microscale resolution and analysis of complex chemical mixtures and its application to basic and applied biological problems [125-144].
- Drs. Herschlag and Brown developed microarray methods for genome-wide profiling of translational regulation and RNA decay rates and a systematic approach to investigating the specificity and regulatory logic of RNA binding proteins on a genome-wide scale [7, 21, 38, 53-56].
- Dr. Bogyo pioneered an approach to profiling the in vivo activity of proteases using chemical probes, and has applied this approach to profiling proteases human cancer [145-153] (BioTechniques, 2005).
- Dr. Chu, in collaboration with Dr. Tibshirani, invented and implemented several of the most powerful and widely used methods for analyzing microarray data [101, 154-162], including Significance Analysis of Microarrays (SAM) [162], which identifies genes with reproducible changes in expression and supplies a false detection rate from permutation of the data. The predictive analysis of microarrays (PAM) method, using Nearest Shrunken Centroids finds and cross-validates genes that classify a sample by proximity to the nearest centroid for gene expression [161]. Supervised principle component analysis (superPC) uses a step-wise approach to identify principle components in expression patterns correlated with a clinical parameter and evaluate their predictive power [101, 157]. These methods have been widely and successfully used in hundreds of published studies.
- Dr. Sherlock has led the development and continual improvement of the Stanford Microarray Database, one of the largest, and probably the most widely used systematic database for global gene expression data, as well as a web-accessible tool-kit for microarray data analysis [111-116, 118, 163, 164].
- Drs. Brown, Mark Davis and Daniel Chen (Immunology and Immunotherapy; Program 7) and colleagues developed methods using peptide-MHC microarrays to manipulate living cells, investigate their responses to specific molecular signals, and profile the binding specificity of their surface receptors [165, 166].
- Dr. Brown and colleagues developed a protein microarray technology that allows sensitive, quantitative detection and measurement of hundreds of different proteins in complex mixtures (like human serum), or for profiling patterns of antibody (or autoantibody) specificity [167, 168]. This technology has been widely adopted, and is currently being employed by Dr. Robinson for his studies in Program 7.
Trial Administration