Nuclear export of Smad2 and Smad3 by RanBP3 facilitates termination of TGF-beta signaling

Abstract

Smad2 and Smad3 (Smad2/3) are key intracellular signal transducers for TGF-beta signaling, and their transcriptional activities are controlled through reversible phosphorylation and nucleocytoplasmic shuttling. However, the precise mechanism underlying nuclear export of Smad2/3 remains elusive. Here we report the essential function of RanBP3 in selective nuclear export of Smad2/3 in the TGF-beta pathway. RanBP3 directly recognizes dephosphorylated Smad2/3, which results from the activity of nuclear Smad phosphatases, and mediates nuclear export of Smad2/3 in a Ran-dependent manner. As a result, increased expression of RanBP3 inhibits TGF-beta signaling in mammalian cells and Xenopus embryos. Conversely, depletion of RanBP3 expression or dominant-negative inhibition of RanBP3 enhances TGFbeta-induced antiproliferative and transcriptional responses. In conclusion, our study supports a definitive role for RanBP3 in mediating Smad2/3 nuclear export and terminating TGF-beta signaling.

Figure 1. RanBP3 inhibits TGF-β-induced transcriptional responses in mammalian cells

(A) The SBE-luc reporter activity was measured in HaCaT cells that were co-transfected with Myc-RanBP3 and treated with 2 ng/ml of TGF-β (20 h). Values and error bars represent mean and standard deviation of three experiments. (B) Effect of RanBP3 on the natural p21 promoter in HaCaT cells. (C) RanBP3-wv, defective in Ran-binding, fails to affect SBE-luc response. SBE-Luc reporter was co-transfected into HaCaT cells with Myc-RanBP3 (WT) or Myc-RanBP3-wv (wv), and treated with SB431542 (SB) or TGF-β. The expression level of RanBP3 (TGF-β-treated) was examined by using Western blotting (bottom). (D) Effect of RanBP3 on BMP responses. C2C12 cells were transfected with Myc-RanBP3 and Id1-Luc reporter. Luciferase activity was measured 20 h after treatment with 25 ng/ml of BMP2. (E) Quantitative real-time RT-PCR (qRT-PCR) analysis of p15 mRNA. RanBP3-OE (stably expressing Flag-RanBP3) cells and parental HaCaT cells were stimulated with TGF-β up to 8 h before total RNAs were harvested. (F) qRT-PCR analysis of p21 mRNA. (G) qRT-PCR analysis of PAI-1 mRNA. (H) Western analysis of p21 protein. RanBP3-OE or parental HaCaT cells were treated with TGF-β for 4 h and harvested at indicated times. The expression levels of RanBP3, p21, Smad2/3 and Smad2/3 phosphorylation were examined by Western blotting with appropriate antibodies as indicated. β-actin blot serves as a loading control. (I) Mesodermal and endodermal marker induction in Xenopus ectodermal explants. The gene expression in uninjected control animal caps (lane 1), and in whole embryos processed in the absence or presence of reverse transcriptase in RT-PCRs (lanes 8 and 9) are shown. The expression level of EF-1α is used as the loading control.

Figure 2. Knockdown of RanBP3 enhances TGF-β growth inhibitory and transcriptional responses

(A) RanBP3 expression in HaCaT cell lines stably expressing RanBP3 shRNAs was examined by anti-RanBP3 western blotting. Whole cell lysates were prepared from HaCaT cell lines stably expressing shRNA-494 (designated RanBP3-KD1), shRNA-676 (RanBP3-KD2) or empty pSRG vector (CTRL). β-actin serves as a loading control. (B) Luciferase activity of SBE-Luc was examined in RanBP3 knockdown stable cells and control HaCaT cells. (C) Western blotting analysis of PAI-1, p15 and p21. Whole cell lysates from RanBP3-KD1 and control HaCaT cells were immunoblotted with antibodies shown on the right. An anti-β-actin immunoblot serves as a loading control. (D) Proliferation of HaCaT stable clone RanBP3-KD1 and control cells was examined using MTT assays (Promega) according to manufacture's instructions. (E) Proliferation of HaCaT stable clone RanBP3-KD3 and control cells. RanBP3 expression was examined by anti-RanBP3 western blotting (left).

Figure 3. RanBP3 enhances nuclear export of Smad2/3

(A) Smad3 nuclear accumulation. HaCaT cells were transfected with Myc-RanBP3, and 24 h after transfection, treated with TGF-β for 2 h and fixed. The localization of endogenous Smad3 was examined by indirect immunostaining with anti-Smad3 antibody. Cells that received Myc-RanBP3 were demonstrated by anti-Myc immunostaining, indicated by white arrows. Quantification of nuclear Smad3 (immunostaining intensity) in RanBP3-transfected cells (RanBP3⁺, n=15) and RanBP3 non-transfected cells (RanBP3^-, n=45) from multiple experiments were shown on the right. Error bars represent standard deviation. (B) Distribution of Smad2. HaCaT stable cells expressing Flag-RanBP3 (RanBP3-OE) or shRanBP3-433 (RanBP3-KD1) as well as parental cells were treated with TGF-β for 2 h and fixed. RanBP3 and Smad2 were visualized by indirect immunostaining with anti-RanBP3 or anti-Smad2 antibodies, respectively. (C) Effect of RanBP3 depletion on the level of nuclear Smad2. RanBP3-KD1 and parental HaCaT cells were treated with TGF-β for 2 h to induce nuclear accumulation of Smad2. Cells were washed 3 times to remove TGF-β and then treated with SB431542 (5 nM) and/or LMB (20 ng/ml) for another 2 h before fixation. Endogenous Smad2 and Smad4 were visualized by indirect immunostaining with anti-Smad2 and anti-Smad4 antibodies, respectively. Fluorescence images were acquired with the same exposure parameters for intensity comparison. (D) Cytoplasmic and nuclear fractionation. RanBP3-KD1 and parental HaCaT cells were treated with TGF-β for 1 h, cells were washed 3 times to remove TGF-β and treated with SB431542 up to 4 h. After cells were then harvested at indicated time, both the nuclear and cytoplasmic fractions were collected. The total Smad2/3 levels as well as the P-Smad2 level were examined. The nuclear-localized protein Lamin A/C and cytoplasmic-localized protein GADPH demonstrate separation of the fractions and proper sample loading. Relative level of cytoplasmic Smad3 (Smad3/GAPDH) was quantified using NIH image. The arbitrary unit for cytoplasmic Smad3 in control cells at 4 h of SB431542 treatment was set to 1. Values and error bars represent mean and standard deviation of three experiments.

Figure 4. RanBP3 mediates Smad2 export dependent of its Ran-binding ability

(A) In vitro export of Smad2. Stable HaCaT cells expressing GFP-Smad2 were permeabilized with digitonin (30 ng/ml) for 5 min and then incubated in a reaction buffer containing either BSA or Ran at 30°C or on ice. GFP-Smad2 was examined by anti-Smad2 antibody Western blotting (∼80 kDa). Lamin A/C and GADPH indicate levels of nuclear and cytoplasmic proteins. Relative level of nuclear GFP-Smad2 (Smad2:Lamin A) was quantified from multiple experiments. The arbitrary unit for nuclear GFP-Smad2 in lane 4 was set to 1. Values and error bars represent mean and standard deviation of three experiments. (B) The effect of RanBP3 or RanBP3-wv protein on Smad2 export. (C) A quantitative Smad2 export assay (see Experimental Procedures) was used to examine the effect of RanBP3 or RanBP3-wv mutant on Smad2 export. Values and error bars represent mean and standard deviation of three experiments.

Figure 5. RanBP3 interacts with TGF-β specific Smad2 and Smad3

(A) Co-immunoprecipitation (Co-IP) of Smads and RanBP3. HEK293T cells were transfected with indicated expression plasmids for RanBP3, Smad1 or 2, and a constitutively active receptor ALK5(T202D) (lane 3) or ALK3(Q233D) (lane 5). Levels of Smads and RanBP3 in the IP products and whole cell lysates (WCL) were analyzed by Western blotting. (B) Co-IP of RanBP3 and Smad3. (C) Endogenous RanBP3-Smad2/3 interaction. HaCaT cells at 90% confluency were treated with or without TGF-β for 2 h before harvest. RanBP3-bound Smad2/3 were immunoprecipitated with an anti-RanBP3 antibody and detected by anti-Smad2/3 Western blotting. The anti-mouse IgG immunoprecipitates in lane 3 serve as negative control. (D) Mapping of RanBP3-binding domains of Smad3. [S³⁵]-labeled Smad3 MH1 domain, MH2 domain as well as the linker region plus MH2 domain was incubated with GST-RanBP3 protein on glutathione beads. Retrieved proteins were separated by SDS-PAGE and visualized by autoradiography. GFP serves as a negative control. (E) Mapping of Smad-binding domains of RanBP3. (F) Ran-binding domain of RanBP3 (RanBP3-R) interacts with Smad3 in GST-pulldown assay. (G) The MH1 and MH2 domains confer the maximum binding of Smad3 to RanBP3. HEK293T cells were transfected with Myc-RanBP3, Flag-Smad1, Flag-Smad3 as well as Flag-Smad1/3 domain chimera as indicated. Co-IP experiments were conducted similarly as in A.

Figure 6. RanBP3 binds to Smads in its unphosphorylated form in the nucleus

(A) Effect of increasing dosages of PPM1A on the Smad3-RanBP3 interaction. HEK293T cells were transfected with expression plasmids as indicated. (B) Binding of RanBP3 to Smad3 mutants in co-IP experiments. Myc-RanBP3 was co-transfected into HEK293T cells with Smad2 wild type (WT), Smad2 mutant lacking the C-terminal serine phosphorylation (2SA), or active Smad2 mutant harboring C-terminal serine phosphorylation-mimetic residues (2SD). (C) The interaction between Myc-RanBP3 and Flag-Smad2(2SD) in the presence of HA-Smad4. (D) Effect of RanBP3 on the interaction between HA-Smad4 and Flag-Smad2(2SD). (E) Binding of RanBP3 to the un-phosphorylated MH2 domain of Smad2 (S2MH2). Recombinant S2MH2 or phosphorylated S2MH2 (P-S2MH2) was incubated with GST-RanBP3 proteins (or GST only as negative control) on glutathione beads. S2MH2 or GST proteins were demonstrated by Western blots with anti-Smad2 or anti-GST antibody, respectively. (F) Visualization of the RanBP3-Smad2/3 interactions in BiFC assays. HaCaT cells were co-transfected with expression plasmids as indicated. Images of reconstituted YFP fluorescence (cells in cyan boxes) were acquired 24 h after transfection. Cells that received the YC-RanBP3/YN pair or the YN-Smad2/YC pair were also immunostained with anti-RanBP3 (image c) or anti-Smad2 antibody (image d), respectively. DAPI staining indicates the nucleus. (G) Effect of TGF-β on the RanBP3-Smad3 interaction in BiFC assays. YC-RanBP3/YN-Smad3 co-transfected HaCaT cells were treated for 2 h with TGF-β (a), overnight with SB431542 (b) or 0.2% FBS (c). (H) DNA-binding assay of Smad3. The DNA-Smad complex was affinity-purified using biotinylated SBE from RanBP3-KD1 cells and parent HaCaT cells, and then examined by using Western blotting with the indicated antibodies.

Figure 7. RanBP3 controls Smad2/3 export and TGF-β signaling via its contact with Smad2/3

(A) Effect of Flag-NLS-RanBP3-R (Flag-NLS-R) on the RanBP3-Smad2 interaction. HEK293T cells were transfected with indicated expression plasmids and co-IP experiments were carried as described in Figure 4A. (B) Effect of Flag-NLS-R on the RanBP3-Smad3 interaction. (C) Effect of Flag-NLS-R on RanBP3-mediated Smad2 export was examined using a quantitative Smad2 export assay. Values and error bars represent mean and standard deviation of two experiments. (D) Effect of Flag-NLS-R on the SBE-luc reporter activity. Flag-NLS-R was co-transfected into HaCaT cells, and the SBE-Luc reporter activity was measured 20 h after TGF-β stimulation. (E) A working model for Smad2/3 transport. During TGF-β signal transduction (indicated by green arrows), receptor-phosphorylated Smad2/3 form a complex with Smad4. The heteromeric Smad complex is then recruited to transcriptional machinery via its binding to the promoter, Smad-cooperating transcription factors (TFX) and co-activators (Co-A) such as p300/CBP, TAZ and the Mediator complex. During signal termination (red arrows), Smad2/3 become dissociated and dephosphorylated by PPM1A. Dephosphorylated Smad2/3 is then exported out of the nucleus by RanBP3 in a Ran-dependent manner. Signal transduction may be re-initiated, dependent on receptor activity. TβR, TGF-β receptor complex; GTF, general transcription factors.