Motor neuron degeneration in Spinal and Bulbar Muscular Atrophy: molecular approaches to counteract mutant androgen receptor neurotoxicity

Spinal bulbar muscular atrophy (SBMA) is a motoneuron disease (MN) associated with the expression of an androgen receptor (AR) containing an elongated polyglutamine (polyQ) tract. The AR ligand testosterone triggers ARpolyQ toxic properties by inducing peculiar structures and aberrant (misfolded) conformations in the mutant protein. In cells, misfolded ARpolyQ toxicity is counteracted by the protein quality control system. This system is composed of chaperones and degradative pathways (proteasome and autophagy), and is partial activation or blockage may result in ARpolyQ accumulation. An improved autophagy-mediated degradation of aggregating ARpolyQ is thought to be beneficial. Indeed, the pharmacological induction of autophagic pathway regulators increases ARpolyQ clearance attenuating its toxicity. Autophagy relies on dynein-mediated retrograde transport of specific chaperones (e.g.: the HSPB8-BAG3-HSC70 complex) which bind the misfolded proteins cargoes. Notably, colchicine, which induces HSPB8, the key component of the complex, increases misfolded protein clearance in MNs. By using motoneuronal and MNs diferentiated from iPSCs expressing ARpolyQ, we found that inhibition of dynein-mediated retrograde transport results in a reduced accumulation and enhanced clearance of ARpolyQ. We found that, not autophagy, but the potentiation of a proteasome-mediated clearance is responsible for the reduction of aggregating insoluble species of ARpolyQ. This process is mediated by the upregulation of the HSC70 co- chaperone BAG1. In fact, overexpression of exogenous BAG1 increased the proteasome-mediated clearance of these misfolded proteins, while BAG1 silencing blocked the enhanced ARpolyQ clearance associated to dynein inhibition. We postulated that when the misfolded ARpolyQ is poorly transported to degradion by autophagy or stored in aggresome, the cells activate a compensatory mechanism based on BAG1 to target the HSC70-bound cargo to the proteasome in a dynein-independent manner. In an analysis performed in a SBMA mice model (AR113Q mice) we found that in muscle of symptomatic SBMA mice, both BAG1 and BAG3 were robustly upregulated. Notably, the BAG3:BAG1 ratio was also increased suggesting a preferential misfolded ARpolyQ routing to autophagy, rather than to proteasome. In addition, in the same muscles most of the autophagic markers (Beclin-1, ATG10, p62/SQSTM1, LC3) analysed were upregulated together with the major players of the HSPB8-mediated PQC response (HSPB8 and its co-chaperone BAG3, and the HSPB8 interactors HSPB2 and HSPB3). These changes were accompanied by the upregulation of Pax-7, myogenin, E2- ubiquitin ligase UBE2Q1, TGF?1, PGC-1? and acetylcholine receptor (AchR), but not of MyoD, and of two E3-ligases (MuRF-1 and Cullin3). Thus an alteration in HSPB8-based PQC machinery may represent muscle-specific biomarkers useful to assess SBMA progression in mice and patients in response to pharmacological treatments