A gene signature identified using a mouse model of androgen receptor-dependent prostate cancer predicts biochemical relapse in human disease

Vanessa C Thompson, Tanya K Day, Tina Bianco-Miotto, Luke A Selth, Guangzhou Han, Mervyn Thomas, Grant Buchanan, Howard I Scher, Colleen C Nelson, Australian Prostate Cancer BioResource, Norman M Greenberg, Lisa M Butler, Wayne D Tilley

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

Mutations in the androgen receptor (AR) have been detected in experimental and clinical prostate tumors. Mice with enforced prostate-specific expression of one such receptor variant, AR-E231G, invariably develop prostatic intraepithelial neoplasia by 12 weeks and metastatic prostate cancer by 52 weeks. The aim of this study was to identify genes with altered expression in the prostates of AR-E231G mice at an early stage of disease that may act as drivers of AR-mediated tumorigenesis. The gene expression profile of AR-E231G prostate tissue from 12-week-old mice was compared to an equivalent profile from mice expressing the AR-T857A receptor variant (analogous to the AR-T877A variant in LNCaP cells), which do not develop prostate tumors. One hundred and thirty-two genes were differentially expressed in AR-E231G prostates. Classification of these genes revealed enrichment for cellular pathways known to be involved in prostate cancer, including cell cycle and lipid metabolism. Suppression of two genes upregulated in the AR-E231G model, ADM and CITED1, increased cell death and reduced proliferation of human prostate cancer cells. Many genes differentially expressed in AR-E231G prostates are also deregulated in human tumors. Three of these genes, ID4, NR2F1 and PTGDS, which were expressed at consistently lower levels in clinical prostate cancer compared to nonmalignant tissues, formed a signature that predicted biochemical relapse (hazard ratio 2.2, p = 0.038). We believe that our findings support the value of this novel mouse model of prostate cancer to identify candidate therapeutic targets and/or biomarkers of human disease.

Original languageEnglish
Pages (from-to)662-672
Number of pages11
JournalInternational Journal of Cancer
Volume131
Issue number3
DOIs
Publication statusPublished - 1 Aug 2012
Externally publishedYes

Keywords

  • androgen receptor
  • gene signature
  • mouse model
  • prostate cancer

Fingerprint Dive into the research topics of 'A gene signature identified using a mouse model of androgen receptor-dependent prostate cancer predicts biochemical relapse in human disease'. Together they form a unique fingerprint.

Cite this