Title: New insights in androgen receptor functioning and screening for androgen receptor antagonists
Other Titles: Nieuwe inzichten in de werking van de androgeenreceptor en screening naar androgeenreceptor antagonisten
Authors: Helsen, Christine
Issue Date: 7-Dec-2012
Abstract: Androgens, such as testosterone and dihydrotestosterone, are the male sex hormones that are responsible for the early development of the male reproductive organs but also for the secondary sexual characteristics that come up at puberty and for maintaining the reproductive function in adult life. Androgens exert their effects by binding to an intracellular receptor, named the androgen receptor (AR). The AR is in fact a ligand-dependent transcription factor that can switch genes on or off by binding as a homodimer to certain elements in the DNA. It is a member of the nuclear receptor superfamily that share structural and functional characteristics. All nuclear receptors consist of the same building blocks or domains. From N- to C-terminus, there is the aminoterminal domain, the DNA-binding domain, the hinge region and the ligand-binding domain. The DBD and the LBD have the highest sequence conservation between nuclear receptors and are structurally similar. Despite the existence of independent domains, each with their own, intrinsic function, the domains also interact or communicate with each other. The N/C-interaction between NTD and LBD and the DNA-dependent dimerization via the DBD in the homodimer are well documented. Via in silico models and mutational analysis we have provided evidence for communication between the DBD and the LBD of the AR. Communication between the LBD of PPAR and the DBD of RXR was found earlier to be important for DNA binding by the PPAR-RXR heterodimer (Chandra et al. 2008). The first in silico model for the AR was therefore based on the crystal structure of the PPAR-RXR heterodimer bound to DNA. Due to the high sequence and structural conservation of DBD and LBD between nuclear receptors we were able to align the protein structure of the DBD and the LBD of the AR onto the PPAR-LBD and RXR-DBD. The second in silico model of DBD-LBD communication in the AR was obtained after a docking analysis that calculates the most optimal interaction modes between AR-DBD and AR-LBD based on the energy-lowering contacts made at the surface of DBD and LBD. We selected residues that were important for interaction based on the in silico models and on mutations that were reported to be involved in the androgen insensitivity syndrome. We thus created mutant ARs with a mutation on the surface of the DBD or on the surface of the LBD and compared their ligand-binding and DNA-binding ability together with their transactivation capacity to the properties of the wild-type AR. We found three mutations in the DBD (K590A, K592A and E621A) that do not affect DNA binding, but reduce ligand-binding and transactivation of the AR. In the LBD, we found four mutations (D695N, R710A, F754S and P766A) that did not affect ligand binding but reduce DNA binding and transactivation. We suggest that these residues are involved in DBD-LBD communications as they can affect the function of a domain they are not situated in.The AR-signaling pathway is an important therapeutic target in advanced and metastatic prostate cancer since prostate tissue is very hormone-responsive. For small and localized prostate tumors, surgery and radiotherapy remain the preferred treatments. Initially prostate cancer cells respond very well to androgen deprivation or androgen blockage. Androgen deprivation is achieved by preventing testosterone production in the testis with GnRH analogues or by inhibiting the androgen biosynthesis. Androgen blockage, on the other hand, is achieved by directly inhibiting the activity of the AR with AR-antagonists or antiandrogens. Current, clinically used antiandrogens such as Bic and HOFl, are relatively weak and many patients have developed therapy-resistance. During androgen deprivation and blockage, the tumor can adapt to the low levels of androgens and the presence of antiandrogens by several, often AR-related mechanisms and become resistant to the applied treatment. This stage of the disease can still be treated with AR-targeted therapeutics, but these need to be more effective or have a distinct mechanism of action compared to the previous treatments. A promising candidate which is now investigated in clinical trials is the second-generation AR-antagonist MDV3100. In order to expand the options for androgen blockade we have set up a screening to detect AR-antagonists in a compound library. To detect AR-antagonists we created stable cell lines that express the human AR and contain an ARE-driven luciferase reporter. The hit-rate of the screening was 1.3% with a false positive rate of only 0.1%. Three compounds (MEL-3, MEL-4 and MEL-6) were selected based on their binding affinity for the AR, their drug-likeliness and patentability of their structure. These compounds were thoroughly characterized for their antagonistic effects on expression of androgen-regulated genes and on proliferation of AR-positive and AR-negative prostate cancer cell lines. MEL-3, MEL-4 and MEL-6 all performed better than Bic in our in vitro tests, but not always better than MDV3100. With RNA-seq we confirmed that MEL-3, MEL-4 and MEL-6 indeed target the AR. The MEL-compounds have a distinct mechanism of action compared to Bic and MDV3100 since each compound demonstrated a distinct pattern of repressive effects on the proliferation of a set of AR-positive cell lines: LNCaP, LAPC4 and VCaP. MEL-6 was selected as the overall best-performing and most specific compound and will now be further developed. To investigate whether the MEL-compounds are suitable as second- or third-line antiandrogens we determined their antagonistic effects in models of resistance to Bic, HOFl and MDV3100. As model for Bic- and HOFl-resistance we used mutant ARs reported to convert the antagonists into agonists: AR W741C for Bic and AR T877A for HOFl. While MEL-6 is a full antagonist on AR wt, AR T877A and AR W741C; MEL-3 and MEL-4 are partial agonists on AR T877A. MEL-6 can thus be used as second-line antiandrogen to treat prostate tumors expressing AR T877A and AR W741C which escaped therapy with HOFl and Bic, respectively. Prostate tumors expressing AR T877A should not be treated with MEL-3 andMEL-4 since they act as agonists in absence of androgens. Since MDV3100 was not yet FDA-approved at the time of writing, little information on therapy-resistance was available. We therefore created a MDV3100-resistant LNCaP-subline by long-term growth in the presence of RD162/MDV3100 as a model of resistance for MDV3100. After 12 months of therapy with RD162, the LNCaP-Rr cell line demonstrated a 2- to 3-fold increase of AR protein level with no detected mutations. The transcriptional activity of the AR in LNCaP-Rr has become hypersensitive to the presence of serum, but not to androgens. Proliferation of LNCaP-Rr is still androgen-responsive but surprisingly both androgens and serum demonstrate growth-suppressive effects. The resistant cells thus proliferate best in absence of androgens and serum. MEL-3, MEL-4 and MEL-6 can, however, still suppress proliferation of this MDV3100-resistant cell line which underlines the crucial role of the AR in the mechanism of resistance. Additional, resistant, AR-positive cell lines are now being developed, and studies are planned to verify the effects of MEL-6 in xenotransplant and explant models. However, our current data indicate already that third-generation antiandrogens like MEL-6 could be viable candidates for the treatment of MDV3100-resistant PrCa.
Publication status: published
KU Leuven publication type: TH
Appears in Collections:Faculty of Medicine - miscellaneous
Biochemistry Section (Medicine) (-)
Laboratory of Molecular Endocrinology

Files in This Item:
File Status SizeFormat
Thesis-CH.pdf Published 3954KbAdobe PDFView/Open

These files are only available to some KU Leuven staff members


All items in Lirias are protected by copyright, with all rights reserved.