Human expanded polyglutamine androgen receptor mutants in neurodegeneration as a novel ligand target

Androgen receptor (AR) plays key roles in various biological events, including pathological processes such as prostate cancer, androgen-insensitive syndrome, and spinal and bulbar muscular atrophy (SBMA). SBMA is caused by mutation of the expanded polyglutamine (polyQ) stretches in the AR gene. Recently, we established a Drosophila SBMA model that expresses the expanded polyQ hAR mutant in eyes, which monitors neurodegeneration as a rough eye phenotype. In addition, we showed that androgen binding to the mutant hAR causes structural alterations, leading to the onset of neurodegeneration in the fly eyes. In the present study, we examined whether the ligand-induced neurodegeneration via the hAR mutant is coupled with the known ligand-induced transactivation function of hAR. By testing several known AR antagonists and several of their structure-related compounds, we unexpectedly found that none of the AR ligands antagonized the hAR mutant neurodegeneration function, and surprisingly, compound 4-(4,4-dimethyl-2,5-dioxo-1-imidazolidinyl)-2-trifluoromethylbenzonitrile (RU56279) was more potent in inducing neurodegeneration. However, in vitro and in vivo mammalian assays showed that RU56279 exhibited the expected antagonistic activity with the same potency as those of the other compounds. Thus, these findings suggest the presence of a novel ligand-induced function of the polyQ hAR mutant in neurodegeneration that could not be prevented by known antagonists for the hAR transactivation function.