Molecular insights into the functional interplay between androgen receptor and its cofactors

Expansion over 38 repeats of the polymorphic CAG tandem repeat in the first exon in androgen receptor gene (AR) causes Spinal and Bulbar Muscular Atrophy (SBMA), a progressive neurodegenerative disease that fully manifests only in males. An interesting therapeutic approach is to enhance the production of other active and non-pathogenic AR isoforms. An example is given by AR-A, a shorter isoform of the full-length protein produced by translation initiation at the Met-188, lacking the polyQ tract. The hypothesis of this work is that by selectively triggering the second AUG-mediated translation we will produce the active and non-toxic AR-A. The preliminary results presented suggest a possible expression of this isoform in WT mice during physiological aging, in particular in the spinal cord and brainstem, with an increase AR-A/AR-full length ratio and we partially confirm this results in knock-in AR113Q mice at 3 months of age. For a long time, SBMA has been thought to primarily be a motor neuron disorder. However, in the last decade a key emerging aspect of this disease is the primary involvement of peripheral tissues, such as skeletal muscle. The second aim of this project is to investigate the role of androgen receptor coregulators in SBMA skeletal muscle and in particular understanding if targeting transcriptional coregulators aberrantly overexpressed could be a potential therapeutic approach to modulate the toxic gain-of-function of the polyQ-expanded AR without exacerbating the loss-of-function. Leveraging cell lines, mice, and patient-derived specimens we found that two AR coregulators, lysine-specific demethylase 1 (LSD1) and protein arginine methyltransferase 6 (PRMT6), are aberrantly overexpressed in the skeletal muscle of male SBMA mice, while expression is much lower in female mice, and overexpression is attenuated by surgical castration. PolyQ-AR binds AREs present in the promoters of Lsd1 and Prmt6 in SBMA myotubes. Notably, binding is specific for mutant AR, occurs in the absence of ligand and is enhanced by androgens. LSD1 and PRMT6 cooperatively and synergistically transactivate AR, and their effect is enhanced by polyQ-AR. Both LSD1 and PRMT6 have a well-studied role as histone writers, but they could also work on non-histone proteins. Post translational modifications are crucial for AR function and activity. PRMT6’s arginine methylation enhances the polyQ-expanded AR toxicity preventing Akt phosphorylation and so degradation. LSD1-mediated lysine demethylation has not yet investigated in SBMA. We demonstrate in this work that LSD1 overexpression modifies AR lysine methylation status, reducing the mono- and di-methylation status of both normal and polyQ-expanded AR suggesting that possible LSD1 is demethylating the receptor in one or more lysine residues. We selected some candidate lysines (K318, K639, K659 and K823) starting from Mass Spectrometry data and we characterized the correspondent lysine-methylation defective mutants in terms of stabilization and activation following ligand binding. The preliminary results obtained suggest that K639 and K659 could be two interesting lysines involved in LSD1’s control of AR transactivation, because LSD1 failed in their transactivation. Pharmacological and genetic silencing of these two coregulators attenuates polyQ-AR transactivation in SBMA cells and suppresses toxicity in SBMA flies. In this project, we provide proof-of-principle that selectively targeting AR coregulators that are aberrantly expressed in SBMA vulnerable tissues is a valuable therapeutic strategy. We perform a preclinical approach based on miRNA-mediated silencing of LSD1 and PRMT6 attenuates disease manifestations in SBMA mice. A crucial finding here is that LSD1 and PRMT6 overexpression is an early pathological process specifically in skeletal muscle fashion and in particular we show that polyQ-AR aberrantly enhances the expression of its own coregulators in SBMA skeletal muscle.