Functional analysis of the human androgen receptor using synthetic and naturally occurring mutations

The human androgen receptor (hAR) is a ligand-activated transcription factor, and like other nuclear receptors, consists of a N-terminal modulatory domain, a central DNA-binding domain, and a C-terminal ligand-binding domain (LBD). Several missense mutations in the LBD cause androgen insensitivity syndrome (AI), a condition in XY individuals with absent or subnormal male primary and secondary sexual characteristics. On the other hand, abnormal expansion of a polyglutamine tract in the N-terminal domain of the hAR causes spinal and bulbar muscular atrophy (SBMA) which also affects males and causes milder forms of AI, in addition to adult-onset motor neuron degeneration and gradual wasting and weakening of the muscles of the limbs, face, throat, and tongue. However, it was not clear how and to what extent these mutations contribute to the clinical phenotype of the affected individuals. In order to investigate this matter, I used PCR site-directed mutagenesis to create plasmids expressing hARs with two pairs of missense mutations in the LBD (Val865Leu and Val865Met, and Arg839His and Arg839Cys), discovered in AI individuals with varying severity of the phenotype, and two abnormal expansions of the polyglutamine repeat discovered in SBMA patients (40 and 50 glutamines). I also synthesized plasmids expressing no glutamines (0 glutamines), 12 glutamines, or 20 glutamines in the same N-terminal region of the hAR. These plasmids were transiently expressed in heterologous cells (COS-1 and PC-3), and the mutant hARs were assayed for ligand binding, stability, and transactivational capacity.In contrast to the findings by others (McPhaul et al., 1992; Marcelli et al., 1994), in some instances involving identical mutations, I consistently observed a correlation between the biochemical phenotype of the mutant hARs and the clinical phenotype of AI individuals; that is, the more severe receptor phenotype was associated with the more severe AI. These results support the hypothesis that hAR phenotype is the dominant factor in the development of the secondary sexual characteristics in normal and affected individuals.I also observed a tight negative correlation between polyglutamine tract length and transactivational capacity. This suggests that polyglutamine modulates the activity of the hAR, and that hAR activity might be suppressed in various androgen-sensitive tissues (including motor neurons) in SBMA individuals, thereby contributing to the age of onset and/or progression of the disease, even if it cannot be the primary pathogenic agent of the disease