Greg Bloom | PubMed Health

PubMed2005

Iterative microarray and RNA interference-based interrogation of the SRC-induced invasive phenotype.

Src kinase has long been recognized as a factor in the progression of colorectal cancer and seems to play a specific role in the development of the metastatic phenotype. In spite of numerous studies conducted to elucidate the exact role of Src in cancer progression, downstream targets of Src remain poorly understood. Gene expression profiling has permitted the identification of large sets of genes that may be functionally interrelated but it is often unclear as to which molecular pathways they belong. Here we have developed an iterative approach to experimentally reconstruct a network of gene activity regulated by Src and contributing to the invasive phenotype. Our strategy uses a combination of phenotypic anchoring of gene expression profiles and loss-of-function screening by way of RNA-mediated interference. Using a panel of human colon cancer cell lines exhibiting differential Src-specific activity and invasivity, we identify the first two levels of gene transcription responsible for the invasive phenotype, where first-tier genes are controlled by Src activity and the second-tier genes are under the influence of the first tier. Specifically, perturbation of first-tier gene activity by either pharmacologic inhibition of Src or RNA-mediated interference-directed knockdown leads to a loss of invasivity and decline of second-tier gene activity. The targeting of first-tier genes may be bypassed altogether because knockdown of second-tier genes led to a similar loss of invasive potential. In this manner, numerous members of a "transcriptional cascade" pathway for metastatic activity have been identified and functionally validated.

Biomarkers, Tumor↗

PubMed2004

Multi-platform, multi-site, microarray-based human tumor classification.

The introduction of gene expression profiling has resulted in the production of rich human data sets with potential for deciphering tumor diagnosis, prognosis, and therapy. Here we demonstrate how artificial neural networks (ANNs) can be applied to two completely different microarray platforms (cDNA and oligonucleotide), or a combination of both, to build tumor classifiers capable of deciphering the identity of most human cancers. First, 78 tumors representing eight different types of histologically similar adenocarcinoma, were evaluated with a 32k cDNA microarray and correctly classified by a cDNA-based ANN, using independent training and test sets, with a mean accuracy of 83%. To expand our approach, oligonucleotide data derived from six independent performance sites, representing 463 tumors and 21 tumor types, were assembled, normalized, and scaled. An oligonucleotide-based ANN, trained on a random fraction of the tumors (n = 343), was 88% accurate in predicting known pathological origin of the remaining fraction of tumors (n = 120) not exposed to the training algorithm. Finally, a mixed-platform classifier using a combination of both cDNA and oligonucleotide microarray data from seven performance sites, normalized and scaled from a large and diverse tumor set (n = 539), produced similar results (85% accuracy) on independent test sets. Further validation of our classifiers was achieved by accurately (84%) predicting the known primary site of origin for an independent set of metastatic lesions (n = 50), resected from brain, lung, and liver, potentially addressing the vexing classification problems imposed by unknown primary cancers. These cDNA- and oligonucleotide-based classifiers provide a first proof of principle that data derived from multiple platforms and performance sites can be exploited to build multi-tissue tumor classifiers.

Algorithms↗