Title: Characterization of androgen receptor variant 7 dimerization in prostate cancer


Speaker: Fatma Özgün


Time: August 28, 2018, 13:00


Place: SNA A52

Koç University

Rumeli Feneri Yolu

Sariyer, Istanbul


Thesis Committee

Assoc. Prof. Dr. Nathan A. Lack (Advisor, Koç University)

Assoc. Prof. Dr. Tuğba Bağcı Önder (Koç University)

Prof. Dr. Batu Erman (Sabancı University)

Assoc. Prof. Dr. Nurhan Özlü (Koç University)

Assist. Prof. Dr. Martin E. van Royen (Erasmus MC)



Prostate cancer is an extremely common disease that affects an estimated 1 out of seven men in their lifetime. While the cancer can often be treated through surgery or radiotherapy, those patients with late-stage, recurrent or metastatic forms of prostate cancer are often given androgen deprivation therapy. While this treatment is initially effective the cancer almost always develops resistance, and progresses to castrate resistant prostate cancer (CRPC). Recent studies have proposed that androgen receptor variants, including ARv7 and ARv567es, can cause AR signaling in the absence of androgen and may drive CRPC. As these variants do not need androgen to initiate transcription, they therefore are intrinsically resistant to all clinical approved therapeutics.


While AR variants correlate with CRPC, it is unclear how they initiate transcription as they are missing a key domain involved in activation and dimerization. Therefore, to better understand the mechanism of variant dimerization we developed an acceptor photobleaching FRET based methodology to test if the most commonly observed variant, ARv7, can form dimers. FRET measurements were analyzed with a custom methodology designed to track single nuclei and record the FRET intensities. We demonstrated that ARv7 preferentially formed heterodimers with flAR, with no ARv7/ARv7 homodimerization being detected. To quantify the relative affinity of the ARv7 with flAR, we conducted FRET with a “tunable” system that can vary the flAR and ARv7 expression. With this model, we monitored the loss of flAR homodimerization due to competition with increasing concentrations of unlabeled flAR or ARv7. We demonstrated that ARv7 interacts with flAR with a same affinity as flAR. To test if the observed variant dimerization was specific to only ARv7, we also conducted FRET with ARv567es. Similar to ARv7, we could readily observe flAR/ARv567es heterodimers but no ARv567es homodimers.


To better understand the bound DNA occupancy of the ARv7/flAR heterodimer, we performed fluorescence recovery after photobleaching (FRAP) experiments. We found that the ARv7 residence time on DNA was markedly shorter than flAR regardless of heterodimerization. These results suggest that ARv7 does not form long-term interactions with DNA. In fact, the occupancy time of the variant was similar to a mutant flAR that cannot bind to DNA. This was not specific to only ARv7, as the recovery time of ARv567es was similar to DNA binding mutant flAR. Using the previously described “tunable” system we found that co-expression of unlabeled protein did not dramatically affect DNA occupancy.


In conclusion, ARv7 and ARv567es interact with flAR to form heterodimers. These variants do not form stably bound fractions on DNA and have much shorter occupancy. Further studies are needed to better understand the function of AR variant heterodimers in CRPC as a potential therapeutic target.