The first goal of the project was the study of the structural properties of the activation domain of the receptor (AR), that is intrinsically disordered. This involved detailed studies using nuclear magnetic resonance spectroscopy, that allowed uncovering that the regions of sequence known as activation are partially folded, that was expected, but also that the polyglutamine tract of the androgen receptor was highly helical and that its helical content correlated with its degree of expansion, that was not. The work was published in ACS Chemical Biology (2016) and Biophysical Journal (2016) and led to a secondary but highly interesting research line on the study of the structural properties of polyglutamine sequences (see below).
Next came the study of how these regions of sequence interact with other AR domains, with different proteins and with small molecules discovered by phenotypic screening. We first studied the interaction between this domain of the C-terminal domain of sub-unit II of TFIIF, a general transcription factor, which allowed us to identify that it is mediated by a key motif in the AD, 433WHTLF437, putting forward a new therapeutic target for castration resistant prostate cancer. The work was published in Structure in 2018 and was highlighted by the editors of the journal with a 'highlight' piece by AR expert Marianne Sadar. In a related work we studied the interaction with Bag-1, that interacts with the region known as Tau-5, and reported the finding in eLife (2017).
The project reached a turning point when we discovered that the activation domain undergoes liquid liquid phase separation and other scientists working on activation domains put forward the hypothesis that this is a general mechanism for transactivation. We published our findings in a collaborative paper in Molecular Cell (2018) and have since then carried out work to understand the precise steps by which this takes place in AR, that we have now fully characterized and will reveal in an upcoming publication, still in preparation, that we consider will result in a milestone paper, that main one coming out of the CONCERT project, that will have impact in different fields.
We discovered that prior to activation the activation domain forms a complex with molecular chaperones such as Hsp40 and Hsp70, that bind to the motif 23FQNLF27, and that hormone binding to the ligand binding domain leads to its interaction with the same motif, displacing the chaperones and leading to liquid liquid phase separation. This phase transition is driven by interactions between aromatic residues in the domain and is indispensable for the function of the receptor: mutations of the residues to Ser, that is not aromatic, leads to a loss of ability to translocate to the nucleus and activate transcription - some of these findings were presented in a paper published in Nature Communications (2019) but the majority of the findings are not yet in the public domain.
Our findings on the structural properties of the polyglutamine tract of AR led to discovery that the helices that they form are stabilized by glutamine side chain to main chain interactions. This finding could explain why the transcriptional activity of the receptor depends on the length of the polyglutamine tract and, most importantly, why expansion of the tracts beyond 37 residues leads to spinobulbar muscular atrophy (SBMA), also known as Kennedy disease, a rare neuromuscular disease that remains not cured. This finding was reported in a second Nature Communications paper (2019), that has had substantial impact in the field (60 citations in 2 years) and was highlighted in Faculty of 1000.
The translational potential of the project is reflected in us filing a number of patents for the exploitation (three: 2016, 2021, 2022) of the results obtained and the creation of the company Nuage Therapeutics in 2021, that will carry out drug discovery for castration resistant prostate cancer. In addition the secondary project on polyglutamine helices has also clear translational potential as we have used it to design peptides that are fully helical and that we envisage to use for the inhibition of protein protein interactions in a paper that is under review in Nature Communications and that is already available as a preprint in Biorxiv.
