Comparing gene expression profiles of cell culture models with profiles of tumor xenografts and patient tumors: Implications for lung and breast cancers.

摘要

Tumor cell lines are relied upon extensively for cancer investigations, yet cultured cells in an in vitro environment differ considerably in behavior as compared to those of the same cancer cells that proliferate in vivo and form tumors in vivo. At the same time, despite the many differences between the in vitro and in vivo systems, cell cultures are a powerful model for understanding the molecular biology of cancers. In the first part of this thesis, I present an analysis of mRNA profile data obtained from tumors grown from lung cancer cells implanted in immunodeficient mice, in order to characterize the influence of the in vivo microenvironment on cancer gene expression, as well as the transcriptomic programs associated with tumor formation. In the second part of this thesis, I present a series of examples where I integrate mRNA profile data from in vitro cell line experiments with public datasets of human tumor gene expression, in order both to show the pathological relevance of the in vitro models and to aid in the selection of genes from the in vitro data for follow-up studies. These case studies center primarily on the transcriptional programs of estrogen-mediated growth and of estrogen-independent growth in breast cancer.