| Cancer arises via selective accumulation of rate limiting mutations that subvert growth arrest and apoptotic programs. In the case of prostate cancer, this process occurs over the course of decades, but it ultimately results in an extremely common and potentially lethal disease. Progress in diagnosing and treating prostate cancer will extend from a better understanding of the gene regulatory networks that are corrupted during tumor initiation and progression. In this work, we have utilized genomic technologies, mouse models and human tissue resources to better characterize the molecular genetics of prostate tumorigenesis. A survey of global gene expression changes in benign and cancerous prostate specimens identified overexpression of the hepsin gene as a common feature of prostate cancer. Hepsin encodes a transmembrane serine protease, and it represents a potentially useful target for rationale drug design. Additional work focused on the androgen signaling axis. We used microarrays to characterize the androgen response program of mouse and human prostate models. A comparison of mouse and human androgen targets identified conserved patterns of expression among select target genes, and this comparison informed a bioinformatics screen for conserved cis-regulatory elements in one AR target, the FKBP5 gene. Another conserved androgen target gene encodes the homeodomain transcription factor Nkx3.1. The Nkx3.1 gene is located at 8p21 in the human genome, a site of frequent loss-of-heterozygosity in prostate cancer. Conditional deletion of even a single Nkx3.1 allele in the mouse prostate resulted in a hyperplasia/PIN phenotype. This phenotype arises from an extension of the proliferative phase during luminal cell maturation following Nkx3.1 allelic loss. We identified Nkx3.1 target genes, which show varying sensitivities to haploid Nkx3.1 deletion. Subsequent analysis revealed that Nkx3.1 gene dosage determines the relative probabilities of stochastic activation or inactivation of a given target gene.{09}Thus, haploinsufficiency at the Nkx3.1 locus can result in complete inactivation of a given target gene in a subpopulation of luminal cells. These findings establish a model for early tumorgenesis wherein quantitative changes in tumor suppressor dosage can result in complete inactivation of crucial effector pathways. |
