Trafficking of androgen receptor mutants fused to green fluorescent protein: a new investigation of partial androgen insensitivity syndrome.

The naturally occurring mutations of the androgen receptor (AR), detected in patients with androgen insensitivity syndrome (AIS), are currently analyzed by in vitro assays. Unfortunately, these assays do not always permit the demonstration of a direct relationship between the in vitro activity of the receptor and the severity of the phenotype (in particular, for mutations detected in patients with partial AIS). We recently studied the trafficking of wild-type AR, fused to the green fluorescent protein (GFP) in living cells. In the present study, we applied this method for the analysis of AR mutants to find out whether it could be a complementary method of investigation of AIS. After construction of the GFP-AR mutant fusion proteins, the androgen-binding characteristics, nuclear transfer capacities, and transcriptional activities were evaluated. The nuclear transfer was quantified in the presence of various concentrations of dihydrotestosterone (DHT). We studied two mutants associated with partial AIS: G743V and R840C. The androgen-binding characteristics of both mutants were affected, in comparison with normal AR. Although the affinities were similar, the dissociation rate of GFP-AR-G743V was twice that of GFP-AR-R840C. In transcriptional assay, both mutants were active only at high concentrations of androgen. The nuclear trafficking of the mutants was evaluated by two parameters: 1) the rate of nuclear transfer; and 2) the maximal amount of receptors imported into the nucleus. At 10(-6) mol/L DHT, the GFP-AR mutants entered into the nucleus in a fashion similar to that of GFP-AR-wt. At 10(-7) mol/L DHT, the rate and maximal degree of nuclear import were both reduced, even more, for GFP-AR-G743V. The difference between mutants was more pronounced at 10(-9) mol/L DHT, because GFP-AR-G743V entered into the nucleus with even slower kinetics. Though the androgen-binding affinity and transcriptional activity assays did not reveal major differences between mutants, the dissociation rate and the trafficking capacity measurements permitted the activity of the mutants to be differentiated. We observed that the nuclear transfer capacities of these mutants are in correlation with the severity of the phenotype. The GFP-AR model provides an opportunity both to observe the dynamics of the hormone/receptor complex in living cells and to study the impact of the ligand-binding domain mutation, as opposed to certain in vitro techniques. Because the nuclear import capacity correlates well with the degree of androgen insensitivity, the GFP-AR is a useful complementary tool to understanding the phenotype/genotype relationship of AR function in patients with AIS.