Search

In mice in which human androgen receptor (AR) replaces the endogenous murine gene, variation in the length of a polymorphic N-terminal polyglutamine tract affects initiation, progression and therapy response of prostate tumors in the TRAMP mouse model. This provides a genetic paradigm in which to dissect AR functions that determine response to treatment. We developed a mouse model that more accurately reflects human disease, utilizing ETV1 overexpression and heterozygosity for PTEN loss, coupled with variation of the androgen axis via AR alleles differing in Q tract length. In vitro, Q tract length was shown to influence ligand-independent AR activation shorter Q tract lengths led to greater phosphorylation of a site in the AR hinge region, enhancing both transactivation and nuclear shuttling of the receptor. The new tumor model proved milder than expected, with a lethal phenotype occurring only at late age in a subset of mice, perhaps similar to PCa occurrence in man. Gene expression profiles in this model showed effects of ETV1 prior to any notable pathophysiology, supporting the notion of ETV1 overexpression as an early oncogenic event. The expression patterns revealed a strong antagonism between AR and ETV1 for a subset of AR target genes, and this antagonism was abrogated in a background of PTEN loss. Antagonism at the molecular level may lead to synergy at the disease level and may be subject to variation in AR allelic strength and androgen levels. Further knowledge of these variables may lead to better predictors for response to androgen ablation therapy.

In mice in which human androgen receptor (AR) replaces the endogenous murine gene, variation in the length of a polymorphic N-terminal polyglutamine tract affects initiation, progression and therapy response of prostate tumors. This provides a genetic paradigm in which to dissect AR functions that determine response to treatment. We are studying the role of the AR Q tract in ligand-independent AR activation in vitro and in a mouse model with prostate cancer ontogeny similar to human. In the mouse model, molecular correlates of differential response to castration will be determined using bioinformatic analysis of microdissected tumor samples. In the third year of this award, in in vitro studies, we showed that ARs with different Q tract lengths are differentially responsive to growth factor activation likely to be influential under castrate conditions. At least some of this differential activity appears to affected by phosphorylation at S-650. At this time, nearly all the experimental mice aged for tumorigenesis (containing a short or long Q-tract AR allele [12Q vs. 48Q], an ETV1 transgene and heterozygous loss of PTEN) have reached their time point and prostate samples have been collected for histology and biochemical analysis. Since tumorigenesis was much slower than anticipated, we performed some interim analysis on ETV1 transgenics wild type for PTEN to test AR efficacy. This will be an important control for the mice deficient in PTEN since both ETV1 and PTEN prove to directly influence AR action, which has relevance to human as well as mouse prostate cancer progression.

In mice in which human androgen receptor (AR) replaces the endogenous murine gene, variation in the length of a polymorphic N-terminal polyglutamine tract affects initiation, progression and therapy response of prostate tumors. This provides a genetic paradigm in which to dissect AR functions that determine response to treatment. We are studying the role of the AR Q tract in ligand-independent AR activation in vitro and in a mouse model with prostate cancer ontogeny similar to human. In the mouse model, molecular correlates of differential response to castration will be determined using bioinformatic analysis of microdissected tumor samples. In the second year of this award, we have accrued the experimental mice from the strains constructed last year, using a modified genetic strategy thought to optimize similarity to human tumorigenesis. In the in vitro studies, we showed that ARs with different Q tract lengths are differentially responsive to signal transduction cascades likely to be influential under castrate conditions. We have tested several different pathways of such ligandindependent activation for their sensitivity to Q tract length. A phosphorylation site in the AR hinge region was examined by mutagenesis for a role in mechanisms underlying Q tract effects. Finally, we have begun to characterize differential prostate pathology in the mouse model as some animals have reached their designated end points.

In mice in which human androgen receptor (AR) replaces the endogenous murine gene, variation in the length of a polymorphic N-terminal polyglutamine tract affects initiation, progression and therapy response prostate tumors. This provides a genetic paradigm in which to dissect AR functions that determine response to therapy. We are studying the role of the AR Q tract in ligand-independent AR activation in vitro and in a mouse model with prostate cancer ontogeny similar to human. In the mouse model, molecular correlates of differential response to castration will be determined using bioinformatic analysis of microdissected tumor samples. In the first year of this award, we have constructed the necessary mouse strains to generate experimental animals, using a modified approach for optimal tumorigenesis. In the in vitro studies, we have shown that AR Q tract length confers differential activation in a promoter- and cell type-dependent manner. In addition, ARs with different Q tract lengths are differentially responsive to signal transduction cascades likely to be influential under castrate conditions. Furthermore, phosphorylation of AR by growth factor activation correlates inversely with Q tract length, providing a potential mechanism by which this polymorphism impacts AR function in castration recurrent prostate cancer., The original document contains color images.