Back

Canonical AREs are tumor suppressive regulatory elements in the prostate

Augello, M. A.; Chen, X.; Liu, D.; Lin, K.; Hakansson, A.; Sjostrom, M.; Khani, F.; Deonarine, L. D.; Liu, Y.; Travascio-Green, J.; Wu, J.; Loda, M.; Feng, F. Y.; Robinson, B. D.; Davicioni, E.; Sboner, A.; Barbieri, C.

2024-02-27 cancer biology
10.1101/2024.02.23.581466 bioRxiv
Show abstract

The androgen receptor (AR) is the central determinant of prostate tissue identity and differentiation, controlling normal, growth-suppressive prostate-specific gene expression1. It is also a key driver of prostate tumorigenesis, becoming "hijacked" to drive oncogenic transcription2-5. However, the regulatory elements determining the execution of the growth suppressive AR transcriptional program, and whether this can be reactivated in prostate cancer (PCa) cells remains unclear. Canonical androgen response element (ARE) motifs are the classic DNA binding element for AR6. Here, we used a genome-wide strategy to modulate regulatory elements containing AREs to define distinct AR transcriptional programs. We find that activation of these AREs is specifically associated with differentiation and growth suppressive transcription, and this can be reactivated to cause death in AR+ PCa cells. In contrast, repression of AREs is well tolerated by PCa cells, but deleterious to normal prostate cells. Finally, gene expression signatures driven by ARE activity are associated with improved prognosis and luminal phenotypes in human PCa patients. This study demonstrates that canonical AREs are responsible for a normal, growth-suppressive, lineage-specific transcriptional program, that this can be reengaged in PCa cells for potential therapeutic benefit, and genes controlled by this mechanism are clinically relevant in human PCa patients.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.