DrFKBP52 contains all known critical domains and residues for function but does not potentiate receptor activity. (A) Amino acid sequence alignment for human FKBP52 (HsFKBP52, Genbank accession number NP_002005.1) and Danio rerio FKBP52 (DrFKBP52, Genbank accession number NP_958877) indicates that DrFKBP52 has 61% similarity to HsFKBP52. The FK1, FK2, and TPR domains are indicated in gray. Previously characterized domains critical for function including the proline-rich loop (a), FK linker and CKII phosphorylation site (b), and conserved C-terminal tail motif important for Hsp90 binding (c) are indicated. (B) Dihydrotestosterone-induced AR-mediated β-galactosidase reporter assays were performed in yeast in the presence or absence of the indicated expression vectors for DrFKBP52, HsFKBP52, and HsFKBP51. In all cases, androgen receptor signaling in cells expressing HsFKBP52 was significantly higher (p ≤ 0.0001) as compared with cells expressing Vector, HsFKBP51, and DrFKBP52.

Selection scheme for Danio rerio FKBP52 gain-of-function mutants. Libraries of random DrFKBP52 mutants were independently generated by error-prone PCR using primers (horizontal arrows) binding upstream in the GAPDH promoter (PGPD) or downstream in the transcriptional terminator (Term). Randomly generated mutant constructs were co-transformed with a linearized vector to facilitate homologous recombination between the common promoter and terminator regions on these fragments reconstituting TRP1-marked expression plasmids harboring the DrFKBP52 mutants. The parental strain contains a LEU2-marked androgen receptor (AR)-P723S expression plasmid and integrated HIS3 reporter gene driven by a hormone-responsive promoter element (HRE) such that growth in histidine-lacking medium is dependent on AR-P723S activity. Transformants were plated on selective growth medium supplemented with 10 mM 3-amino-1,2,4-triazole and 10 nM dihydrotestosterone (DHT), and colonies that grew, as well as those expressing human FKBP52, were selected for further analysis. Mutants exhibiting the gain-of-function phenotype were extracted from yeast and co-transformed with wild type AR into a secondary strain containing a hormone-responsive LacZ reporter plasmid, and assayed for the ability to potentiate wild type AR-mediated β-galactosidase activity. DrFKBP52 mutants that gained the ability to potentiate AR activity in these assays were sequenced to identify relevant mutations (inset). Yeast strains containing a hormone-inducible HIS3 gene and expressing AR-P723S plus either Vector, HsFKBP52, or DrFKBP52 were serially diluted and spotted on selective medium containing a growth-limiting concentration of 10 nM DHT (inset).

DrFKBP52 gain-of-function mutants isolated. The domain arrangement of wild type DrFKBP52 with amino acid numbering at the domain boundaries is illustrated at the top with identified mutants aligned below. Each mutant is coded based on the independent library from which it was isolated (I to R) and an isolate number. Repeatedly identified mutations of particular interest are shown in bold.

DrFKBP52 mutants potentiate androgen receptor activity. (A) Hormone-dependent β-galactosidase reporter gene activity was measured in yeast strains expressing wild type AR in the presence of 3 nM DHT. (B) FKBP activities were similarly determined in transfected 52KO MEF cells expressing wild-type AR and a hormone-dependent luciferase reporter. Protein expression levels were monitored by Western blot for the introduced FKBPs, AR, and endogenous GAPDH as a loading control. In all cases (A and B), androgen receptor activity in cells expressing DrFKBP52-A111V/T157R was significantly higher (p ≤ 0.0001) as compared with cells expressing Vector and DrFKBP52. There was no significant difference in androgen receptor activity in cells expressing human FKBP52 and DrFKBP52-A111V/T157R.

Predictive modeling and hydrophobicity scale of DrFKBP52 and DrFKBP52-A111V. The top structures represent the top five 3D models for each structural query of DrFKBP52 (A) and DrFKBP52-A111V (B) predicted by I-TASSER. The hydrophobicity on the surface of the loop was assessed (red = hydrophobic) in DrFKBP52 (A) and DrFKBP52-A111V (B) to allow us to further evaluate the structural changes that may be relevant in their functional divergence (bottom structures).

Predicted structural differences in the FK1 proline-rich loop. Homology modeling was used to generate predicted models of DrFKBP52 and DrFKBP52-A111V, in the comparison to human FKBP52 and FKBP51 to determine conformational changes induced by the DrFKBP52-A111V mutation. Crystal structures of HsFKBP51 (PDB ID: 1KT0), HsFKBP52 (PDB ID: 1Q1C), DrFKBP52 (predicted), and DrFKBP52-A111V (predicted) are aligned; the respective FK1 domains are shown in space-filled modeling and colored in hydrophobicity scale (blue = hydrophilic, red = hydrophobic). Note that the only differences between FKBP51 and HsFKBP52 within the loop region are at positions 119 and 124.