Large FK506-binding proteins shape the pharmacology of rapamycin

Abstract

The immunosuppressant and anticancer drug rapamycin works by inducing inhibitory protein complexes with the kinase mTOR, an important regulator of growth and proliferation. The obligatory accessory partner of rapamycin is believed to be FK506-binding protein 12 (FKBP12). Here we show that rapamycin complexes of larger FKBP family members can tightly bind to mTOR and potently inhibit its kinase activity. Cocrystal structures with FKBP51 and FKBP52 reveal the modified molecular binding mode of these alternative ternary complexes in detail. In cellular model systems, FKBP12 can be functionally replaced by larger FKBPs. When the rapamycin dosage is limiting, mTOR inhibition of S6K phosphorylation can be enhanced by FKBP51 overexpression in mammalian cells, whereas FKBP12 is dispensable. FKBP51 could also enable the rapamycin-induced hyperphosphorylation of Akt, which depended on higher FKBP levels than rapamycin-induced inhibition of S6K phosphorylation. These insights provide a mechanistic rationale for preferential mTOR inhibition in specific cell or tissue types by engaging specific FKBP homologs.

Fig 1

Large FKBPs bind to the FRB domain of mTOR and inhibit its kinase activity in vitro. (A) Rapamycin dependence of the interaction of FKBP51 and FKBP52 with the FRB domain of mTOR. The binding of the isolated FK1 domains of FKBP51 and FKBP52 to the FRB domain was tested in a GST pulldown assay. FKBP51 FK1 and FKBP52 FK1 at 0.5, 1.0, and 1.5 μM were incubated with a 1 μM concentration of the GST-FRB domain on GST beads in the absence (lanes 3 and 9) or presence of 1 μM rapamycin (lanes 4 to 6 and 10 to 12). After washing, matrix-bound proteins were analyzed by SDS-PAGE. Lane 1 is the GST-FRB domain and FKBP52 FK1 at 1.5 μM each. Lanes 2 and 8 are controls for unspecific binding of 1.5 μM FKBP51 FK1 or FKBP52 FK1 in the absence of the GST-FRB domain. Lane 7 is a molecular size standard. (B) mTORC1 kinase activity is inhibited by large FKBPs in vitro. Kinase assays were performed by incubating immunoprecipitated mTORC1 with EGFP–4E-BP1 or HA-S6 kinase at 37°C in the presence or absence of 500 nM FKBP, 500 nM rapamycin (R), or 100 nM torin-1 (T). After 30 min, the samples were boiled and 4E-BP1 phospho-T37/46 and S6 kinase phospho-T389 signal intensities were assessed by Western blotting (top), quantified, and normalized to those of a sample treated with mTORC1 only (bottom). Results for FKBPs are displayed as standard deviations from the mean.

Fig 2

Crystal structures of the complexes of FKBP51 and FKBP52 with rapamycin and mTOR. (A) Ribbon representation of the ternary minimal complexes of FKBP12 (2FAP) and of the FK1 domains (aa 1 to 143) of FKBP51 and FKBP52 with rapamycin and the FRB domain of mTOR. The FKBP domains were superimposed and are in red (FKBP12), pale blue (FKBP51 FK1), and yellow (FKBP52 FK1), respectively. The 40s and 80s loops, which directly contact the FRB domain, are indicated. For clarity, only a single rapamycin molecule (dark blue sticks) and a single FRB domain (green ribbon) are shown. (B) Direct protein-protein contacts in the ternary complexes of FKBP12 (left) and FKBP51 (right). The orientation is the same as in panel A. Interacting residues are shown as sticks. van der Waals interactions are gray dashed lines, and hydrogen bonds are black lines. FRB domains and FKBPs are shown as backbone traces. (C) Network of ordered water molecules at the FKBP-FRB domain interface. The proteins are shown as surface representations, and the water molecules are shown as red spheres. Hydrogen bonds are indicated by dashed lines. Water molecules that directly mediate interactions between the FRB domain and FKBPs are circled.

Fig 3

TOR-FKBP interactions in eukaryotic cells. (A) Overexpression of FKBP homologs restores rapamycin sensitivity in yeast cells. Yeast strain ΔFPR1p, lacking FKBP12, was transformed with plasmids encoding human FKBP12, FKBP51 FK1, FKBP52 FK1, or full-length FKBP52 under the control of the Gal promoter. A 1:10 dilution series of the resulting strains was spotted onto yeast minimal medium plates lacking histidine with 0.2% galactose as the carbon source with or without supplementation with the indicated concentrations of rapamycin and grown for 3 to 4 days at 30°C. (B) mTOR interacts with FKBP12, FKBP51, and FKBP52 in a rapamycin-dependent manner. HEK293 cells were transfected with a FLAG-mTOR or FLAG-mTOR-S2035T expression plasmid. At 30 min prior to lysis, rapamycin (Rap), FK506 (FK) (final concentration, 25 nM each), or the dimethyl sulfoxide vehicle (veh.) was added. The cell lysates were subjected to immunoprecipitation with anti-FLAG antibody, and the eluates were analyzed by immunoblotting. (C) FKBP12 and FKBP51 interact with the mTORC1 component Raptor. HEK293 cells were transfected with myc-Raptor and FLAG-FKBPs, starved, and stimulated with FBS in the absence or presence of 20 nM rapamycin for 60 min. Cell lysates were subjected to immunoprecipitation with anti-FLAG antibody. Beads were washed four times with CHAPS-containing lysis buffer and eluted by boiling with SDS sample buffer for 5 min. Supernatants were analyzed by SDS-PAGE and immunoblotting. Expression of FKBPs was induced by SG medium.

Fig 4

Contributions of FKBPs to the cellular action of rapamycin. (A) Effect of rapamycin on S6K phosphorylation in FKBP12 knockdown HeLa cells. HeLa cells were transfected with siRNA for FKBP12 or nontargeting controls, starved, and stimulated with 10% FBS–100 nM insulin for 60 min in the presence of 0 to 3 nM rapamycin. Cells were lysed and analyzed by immunoblotting. FKBP25 probing served as an additional loading control. (B) Overexpression of FKBP12 or FKBP51 has different effects on S6 kinase phosphorylation. HeLa cells were transfected with an FKBP12 or FKBP51 overexpression vector, starved for 16 h, and stimulated with 10% FBS–100 nM insulin for 60 min in the absence or presence of 0.25 nM rapamycin. Thereafter, cells were lysed and subjected to SDS-PAGE and Western blotting. The membrane was probed with antibodies against the indicated proteins or antigens. Unrelated lanes between mock-transfected controls and FKBP12 and FKBP51 overexpression samples have been removed for clarity. (C) In FKBP12 knockdown cells, rapamycin-induced Akt hyperphosphorylation is blunted. Wild-type (wt) and FKBP12 knockdown neuroblastoma cells were starved for 24 h and stimulated with 10% FCS–100 nM insulin for 60 min in the presence of 0 to 10 nM rapamycin. Cells were lysed and subjected to SDS-PAGE and immunoblotting. S6 kinase and Akt phosphorylation was assessed with the appropriate phospho-specific antibodies. (D) FKBP12 is necessary for rapamycin-induced Akt hyperphosphorylation. Wild-type and FKBP12 knockdown SH-SY5Y neuroblastoma cells were starved for 24 h and stimulated with 10% FCS–100 nM insulin for 60 min in the presence of 0 to 500 nM rapamycin or 100 nM torin-1. Cells were lysed, and Akt S473 phosphorylation was assessed with a phospho-antibody-based FRET assay. Mean values of two independent duplicates are shown. Values were normalized to 1 for 0 nM rapamycin for each cell type. (E) FKBP51 can replace FKBP12 in enabling rapamycin (Rap)-induced Akt hyperphosphorylation. FKBP12 knockdown SH-SY5Y cells were transfected with FKBP12, FKBP51, or a control plasmid and treated 24 h later with 10 nM rapamycin or 100 nM torin-1 for 60 min. Cellular Akt S473 phosphorylation was determined with a homogeneous time-resolved FRET assay. Mean values of three independent data points are shown. Results were normalized to those of dimethyl sulfoxide (DMSO)-treated controls (100%) for each transfection condition. ***, P < 0.001; **, P < 0.01.

Fig 5

FKBP12-selective ligands do not block the cellular effects of rapamycin. SHSY-5Y cells (A and B) and HeLa cells (C and D) were treated with dimethyl sulfoxide, rapamycin (Rap), or torin-1 (100 nM) in the absence or presence of the FKBP12-selective inhibitor Biricodar (Biri) (40) or compound 44 (cmpd44) (58) or the nonselective FKBP inhibitor FK1706 or FK506 (45) at the indicated concentration. After 60 min, cellular Akt phosphorylation (A and C) and cellular mTOR phosphorylation (B and D) were determined with a homogeneous time-resolved assay. The panselective FKBP inhibitors, but not the FKBP12-selective inhibitors, blocked the effect of rapamycin. For all experiments, mean values from at least three data points are shown. The dimethyl sulfoxide control was set to 100% phosphorylation, while torin-1 was defined as 0% phosphorylation. The values are presented as means ± standard deviations. Two-way analysis of variance and a priori testing were used for statistical analysis. Comparison with dimethyl sulfoxide treatment: ***, P < 0.001; **, P < 0.01; *, P < 0.05. Comparison with treatment with rapamycin alone: +++, P < 0.001; ++, P < 0.01; +, P < 0.05.