Mitochondria-nucleus shuttling FK506-binding protein 51 interacts with TRAF proteins and facilitates the RIG-I-like receptor-mediated expression of type I IFN

Abstract

Virus-derived double-stranded RNAs (dsRNAs) are sensed in the cytosol by retinoic acid-inducible gene (RIG)-I-like receptors (RLRs). These induce the expression of type I IFN and proinflammatory cytokines through signaling pathways mediated by the mitochondrial antiviral signaling (MAVS) protein. TNF receptor-associated factor (TRAF) family proteins are reported to facilitate the RLR-dependent expression of type I IFN by interacting with MAVS. However, the precise regulatory mechanisms remain unclear. Here, we show the role of FK506-binding protein 51 (FKBP51) in regulating the dsRNA-dependent expression of type I IFN. The binding of FKBP51 to TRAF6 was first identified by "in vitro virus" selection and was subsequently confirmed with a coimmunoprecipitation assay in HEK293T cells. The TRAF-C domain of TRAF6 is required for its interaction, although FKBP51 does not contain the consensus motif for interaction with the TRAF-C domain. Besides TRAF6, we found that FKBP51 also interacts with TRAF3. The depletion of FKBP51 reduced the expression of type I IFN induced by dsRNA transfection or Newcastle disease virus infection in murine fibroblasts. Consistent with this, the FKBP51 depletion attenuated dsRNA-mediated phosphorylations of IRF3 and JNK and nuclear translocation of RelA. Interestingly, dsRNA stimulation promoted the accumulation of FKBP51 in the mitochondria. Moreover, the overexpression of FKBP51 inhibited RLR-dependent transcriptional activation, suggesting a scaffolding function for FKBP51 in the MAVS-mediated signaling pathway. Overall, we have demonstrated that FKBP51 interacts with TRAF proteins and facilitates the expression of type I IFN induced by cytosolic dsRNA. These findings suggest a novel role for FKBP51 in the innate immune response to viral infection.

Figure 1. FKBP51 is a novel TRAF6-binding protein.

(A) The schematic structure of mouse FKBP51. FK1 and FK2 indicate FKBP12-like domains. TPR1, TPR2, and TPR3 indicate tetratricopeptide motifs. An amino acid sequence showing a peptide determined with IVV selection to be a TRAF6-binding peptide. (B) Protein–protein interaction network in MAVS-mediated type I IFN induction obtained from the PPI database (http://genomenetwork.nig.ac.jp/index_e.html). The three dotted lines show the interactions newly identified in this study. The red dotted line indicates the interaction identified with IVV selection. (C) Coimmunoprecipitation of FKBP51 with TRAF6 and its mutants (ΔR, ΔRZ, ΔNC, and ΔC). The TRAF6 mutants expressed in the cells are indicated on the top of the panels. WT and C indicate full-length TRAF6 and Flag-tagged expression vector, respectively. The upper panel shows the western blotting of immunoprecipitates using anti-Myc antibody to detect Myc-tagged FKBP51. Middle panels show the western blotting of immunoprecipitates using anti-Myc antibody to detect Flag-tagged TRAF6 and mutants. The lower panel shows the western blotting of total cell lysates using anti-Myc antibody. One representative experiment of three is shown. (D) Schematic structures of TRAF6 and its deletion mutants used in this study. “RING” indicates the RING-finger domain. “ZINC” indicates the region of six zinc-finger domains. “C-C” indicates the coiled-coil domain. “TRAF-C” indicates the TRAF domain conserved in all TRAF family members. The Flag tag (abbreviated in this figure) was connected to the N-terminal of the WT protein and mutants. The binding ability of each protein to FKBP51, as determined in Figure 1C, is indicated on the right of each structure. “+” indicates positive for binding, and “–” indicates negative for binding. (E) FKBP51 preferentially binds TRAF3 and TRAF6. Flag-tagged TRAF proteins and Myc-tagged FKBP51 were expressed in HEK293T cells and immunoprecipitated with an anti-Flag antibody. The TRAF proteins expressed in the cells are indicated on the top of the panels. The upper panel shows the western blotting of samples immunoprecipitated with anti-Myc to detect FKBP51. The middle panel shows the western blotting of samples immunoprecipitated with anti-Flag antibody. The lower panel shows the western blotting of the cell lysate with anti-Myc antibody. One representative experiment of three is shown.

Figure 2. FKBP51 facilitates the cytosolic poly I:C-dependent expression of type I IFN.

(A) Knockdown of FKBP51 in MEF cells and L929 cells. Total cell lysates from FKBP51-knockdown cells (siFKBP51) and control cells (siControl) stimulated by lipofection with poly I:C or control lipofection were analyzed with western blotting using anti-FKBP51 antibody (upper panel) or anti-tubulin antibody (lower panel) as the loading control. (B) Reduction in the expression of IFNβ and ISG15 in FKBP51-knockdown MEF cells and L929 cells stimulated by lipofection with poly I:C. FKBP51-knockdown or control MEF and L929 cells were unstimulated or stimulated by lipofection with poly I:C for 1, 2 and 4 h. The expression levels of IFNβ (Ifnb) and ISG15 (Isg15) were evaluated with a qPCR analysis and normalized to the level of GAPDH mRNA. Data are the means ± SD of triplicate determinations. *P<0.05 and **P<0.01; Student’s t test with a two-tailed distribution and two-sample equivalent variance parameters. (C) Reduction in the expression of IFNβ in FKBP51-deficient (Fkbp51 ^−/−) MEF cells stimulated with dsRNA. FKBP51-deficient and control MEF cells were transfected with poly I:C by lipofection. The expression of IFNβ and ISG15 was evaluated with a qPCR analysis. Data are the means ± SD of triplicate determinations and are representative of three independent experiments. ***P<0.001 and **P<0.01; Student’s t test with a two-tailed distribution and two-sample equivalent variance parameters.

Figure 3. FKBP51 enhances NDV-dependent expression of type I IFN.

(A) Knockdown of FKBP51 in MEF cells. Total cell lysates from FKBP51-knockdown cells (siFKBP51) and control knockdown cells (siControl) infected with NDV or without infection were analyzed by western blotting using anti-FKBP51 antibody (upper panel) and anti-tubulin antibody (lower panel) as the loading control. One representative experiment of three is shown. (B) FKBP51-knockdown and control MEF cells were infected with NDV (MOI = 10). At 8 h after infection, the expression of IFNβ was evaluated by qPCR and normalized to the expression of β-actin mRNA. Data are the means ± SD of triplicate determinations and are representative of three independent experiments. *P<0.05; Student’s t test with a two-tailed distribution and two-sample equivalent variance parameters.

Figure 4. FKBP51 enhances the cytosolic poly I:C-dependent activations of IRF3, RelA, and JNK.

(A) Immunoblotting analysis of total cell lysates from FKBP51-knockdown or control MEF cells stimulated by lipofection with or without poly I:C. Total cell lysates were recovered at various time points after stimulation and analyzed by western blotting using anti-phospho IRF3, anti-IRF3, anti-FKBP51 and anti-tubulin antibodies. (B) Immunoblotting analysis of total cell lysates (whole) and nuclear extracts (Nuclear) from FKBP51-knockdown or control MEF cells stimulated by lipofectamin with poly I:C. Total cell lysates and nuclear extracts were recovered at various time points after poly I:C stimulation and analyzed by anti-RelA. Anti-Parp1 was used for loading control of nuclear fraction. Anti-tubulin was used to monitor contamination of cytosolic fraction. (C) Immunoblotting analysis of total cell lysates from FKBP51-knockdown or control MEF cells stimulated by Lipofection with or without poly I:C. Total cell lysates were recovered at various time points after poly I:C stimulation and analyzed by western blotting using anti-phospho p38, anti-p38, anti-phospho JNK, anti-JNK1 and 2, anti-FKBP51, and anti-tubulin antibodies.

Figure 5. FKBP51 preferentially accumulates in the mitochondria after cytosolic dsRNA stimulation.

(A) Cytoplasmic accumulation of FKBP51 in MEF cells after stimulation with cytosolic dsRNA. MEF cells were stimulated by Lipofectamine with or without poly I:C. Endogenous FKBP51 was detected with an anti-FKBP51 antibody. The nuclei were visualized by staining with propidium iodide. One representative experiment of three is shown. (B) Cytoplasmic accumulation of FKBP51 in MEF cells after NDV infection. Endogenous FKBP51 was detected with an anti-FKBP51 antibody. The nuclei were visualized by staining with propidium iodide. One representative experiment of three is shown. (C) Subcellular localization of FKBP51 in mitochondria. MEF cells were stimulated by lipofectamine with poly I:C or NDV infection. Endogenous FKBP51 was detected with an anti-FKBP51 antibody. The mitochondria were visualized by immunostaining with anti-TOM20 antibody. One representative experiment of three is shown.

Figure 6. FKBP51 is a possible scaffolding protein in RIG-I-dependent signaling.

(A) Binding of FKBP51 to IRF7 in HEK293T cells. Combinations of Myc-tagged FKBP51 and Flag-tagged TRAF6, MAVS, TBK1, NEMO, IRF7, IRF3, or control were transiently expressed in HEK293T cells. The transfected genes are indicated on the top of the panels. IP and INPUT panels indicate the western blotting analysis of the immunoprecipitated samples and the cell lysates used for immunoprecipitation, respectively. The antibodies used for western blotting are shown on the right of the panels. One representative experiment of two independent experiments is shown. (B) Luciferase activity of HEK293T cells transfected with a combination of plasmids encoding Myc-tagged FKBP51 or control Myc, the Flag-tagged CARD domain of RIG-I (RIG-I-CARD) or control Flag, ISRE-driven luciferase, and β-actin-promoter-driven β-gal. The amount of transfected plasmid encoding FKBP51 or RIG-I-CARD is indicated below the graph. Relative luminescence units (RLU) were normalized to the activity of β-gal. The fold activation was determined as the normalized RLU for each sample relative to those of the sample transfected with the control vector, which indicated the basal activation. Data are the means ± SD of triplicate determinations and are representative of three independent experiments. **P<0.01; Student’s t test with a two-tailed distribution and two-sample equivalent variance parameters. Western blotting analysis of transfected samples using anti-Myc antibody or anti-Flag antibody. Subcellular localization of overexpressed FKBP51 is shown in the bottom panels. (C) Luciferase activity of HEK293T cells transfected with combinations of two plasmids encoding FKBP51 and the CARD domain of MDA-5 (MDA-5-CARD), TBK1, MAVS, IRF7, IRF3 or its control vector. The combinations of the transfected plasmids are indicated. Luciferase activity is shown as the fold induction relative to the basal activation, as in Fig. 6B. Data are the means ± SD of triplicate determinations and are representative of three independent experiments. **P<0.01, *P<0.05; Student’s t test with a two-tailed distribution and two-sample equivalent variance parameters.