Promotion of importin alpha-mediated nuclear import by the phosphorylation-dependent binding of cargo protein to 14-3-3

Abstract

14-3-3 proteins are phosphoserine/threonine-binding proteins that play important roles in many regulatory processes, including intracellular protein targeting. 14-3-3 proteins can anchor target proteins in the cytoplasm and in the nucleus or can mediate their nuclear export. So far, no role for 14-3-3 in mediating nuclear import has been described. There is also mounting evidence that nuclear import is regulated by the phosphorylation of cargo proteins, but the underlying mechanism remains elusive. Myopodin is a dual-compartment, actin-bundling protein that functions as a tumor suppressor in human bladder cancer. In muscle cells, myopodin redistributes between the nucleus and the cytoplasm in a differentiation-dependent and stress-induced fashion. We show that importin alpha binding and the subsequent nuclear import of myopodin are regulated by the serine/threonine phosphorylation-dependent binding of myopodin to 14-3-3. These results establish a novel paradigm for the promotion of nuclear import by 14-3-3 binding. They provide a molecular explanation for the phosphorylation-dependent nuclear import of nuclear localization signal-containing cargo proteins.

Figure 1.

Myopodin interacts with 14-3-3 via two consensus 14-3-3–binding motifs. (A) Myopodin contains two NLSs, a PPXY motif, two 14-3-3–binding motifs, and an actin-binding site. Arrows indicate potential interactions with other proteins or with protein domains (WW). The Myo-2 fragment (aa 187–420) was used as bait in a yeast two-hybrid screen. (B) 14-3-3β colocalizes with the Z-disc marker α-actinin in mouse skeletal muscle. (C) In undifferentiated C2C12 myoblasts, myopodin is predominantly found in the nucleus (top left), whereas 14-3-3β is preferentially found in the cytoplasm (top right). Blocking nuclear export with LMB does not affect the nuclear localization of myopodin (bottom left), but leads to the nuclear accumulation of 14-3-3β (bottom right). (D) Heat shock causes nuclear accumulation of 14-3-3β in differentiated myotubes, as visualized by double labeling with DAPI. (E) Myopodin from mouse skeletal muscle (left) and C2C12 myoblast extracts (right) specifically binds to GST–14-3-3β, but not to GST alone. (F) Coimmunoprecipitation experiments show that endogenous myopodin interacts with 14-3-3β in adult mouse heart. (left) IP with anti–14-3-3β; (right) IP with antimyopodin. No binding was found with a control IgG. (G) Myopodin binds to all 14-3-3 isoforms except 14-3-3τ. 14-3-3β lacking the NH₂-terminal dimerization domain (βΔN) can bind to myopodin, albeit to a lesser extent. In contrast, 14-3-3ɛ carrying a point mutation (ɛK49E) that is known to abrogate target binding does not interact with myopodin. (H) The two consensus 14-3-3–binding motifs, 14-3-3#1 (sequence RSLASVP) and 14-3-3#2 (sequence RSVTSP), were deleted separately (Δ14-3-3#1 and Δ14-3-3#2) and together (Δ14-3-3#1 + 2). Both FLAG–Δ14-3-3#1 and FLAG–Δ14-3-3#2 display dramatically reduced binding to 14-3-3β as compared with full-length myopodin (full-length FLAG). Deletion of both 14-3-3–binding motifs (FLAG–Δ14-3-3#1 + 2) abrogates binding to 14-3-3β.

Figure 2.

14-3-3β is required for nuclear import of myopodin. (A) Overexpression of full-length GFP–14-3-3β does not change the nuclear localization of endogenous myopodin (top left), whereas the dominant negative form GFP–14-3-3βΔN causes the nuclear exclusion of myopodin (bottom left). LMB does not reverse the cytoplasmic localization of myopodin that is caused by GFP–14-3-3βΔN (bottom right). (B) Confocal microscopy reveals a predominantly nuclear localization of GFP-tagged, full-length myopodin (top). Deletion of both NLSs (ΔNLS#1 + 2) or both 14-3-3–binding sites (Δ14-3-3#1 + 2) dramatically decreases the nuclear localization of GFP–myopodin. The combined deletion of all four binding motifs (Δ14-3-3#1 + 2 + ΔNLS#1 + 2; bottom) virtually abrogates the nuclear localization of myopodin (bottom). (C) Quantitative analysis is presented as a percentage of nuclear GFP–myopodin. 73.4% of GFP full-length myopodin is found in the nucleus. Deletion of both NLSs separately (ΔNLS#1 and ΔNLS#2) or together (ΔNLS#1 + 2) decreases the amount of nuclear GFP–myopodin to 14.8%, 13.4%, and 14.2%, respectively. Removal of 14-3-3–binding sites (Δ14-3-3#1, Δ14-3-3#2, and Δ14-3-3#1 + 2) reduces nuclear myopodin to 20.0%, 21.8%, and 19.1%, respectively. Combined deletion of all four motifs (Δ14-3-3#1 + 2 + ΔNLS#1 + 2) causes a further reduction of nuclear myopodin to 6.4%. 60.5% of GFP (control) is found in the nucleus. Statistical significance was confirmed by analysis of variance between groups (ANOVA; P < 0.001). Error bars indicate standard deviation.

Figure 3.

14-3-3 binding to myopodin is required for the interaction between myopodin and importin α. (A) Endogenous 14-3-3β from C2C12 myoblasts coimmmunoprecipitates with myopodin (FLAG-full-length). FLAG–Δ14-3-3#1 and FLAG–Δ14-3-3#2 show impaired binding to 14-3-3β. No binding to 14-3-3β is found for FLAG–Δ14-3-3#1 + 2. In contrast, deletion of one or both NLSs (FLAG-ΔNLS#1, FLAG-ΔNLS#2, and FLAG-ΔNLS#1 + 2) does not interfere with the binding of myopodin to 14-3-3β. Combined deletion of all four motifs (FLAG–Δ14-3-3#1 + 2 + ΔNLS#1 + 2) abrogates the binding of myopodin to 14-3-3β. (B) Binding of 14-3-3β (top) and importin α (middle) to FLAG-tagged myopodin (bottom). Single NLS deletions (FLAG-ΔNLS#1 and FLAG-ΔNLS#2) do not impair the expression of myopodin or the interaction of myopodin with 14-3-3β or importin α. In contrast, the absence of both NLSs (FLAG-ΔNLS#1 + 2) causes loss of importin α binding, whereas 14-3-3β binding is preserved. Deletion of 14-3-3–binding motifs in myopodin (FLAG–Δ14-3-3#1, FLAG–Δ14-3-3#2, and FLAG–Δ14-3-3#1 + 2) abrogates the interaction of myopodin with 14-3-3β and with importin α. (C) The 14-3-3 inhibitory peptide R18 abrogates the binding of myopodin to 14-3-3β (top) and to importin α (bottom). (D) Immobilized GST–importin α, but not GST alone, binds purified FLAG-myopodin in the presence of purified FLAG–14-3-3β. In contrast, no interaction is found between GST–importin α and FLAG–14-3-3β or FLAG-myopodin alone.

Figure 4.

Two phosphorylated residues in myopodin mediate 14-3-3 binding. (A) SDS-PAGE analysis showing reduced molecular weight of purified FLAG-myopodin after dephosphorylation with λ-PPase. (B) Dephosphorylation of myopodin abrogates binding to GST–14-3-3β. The interaction is also prevented by the 14-3-3–blocking peptide R18. (C) Binding of purified FLAG-myopodin to endogenous 14-3-3β from C2C12 myoblasts is abrogated by dephosphorylation of myopodin (top). In contrast, binding of endogenous myopodin to purified FLAG–14-3-3β is not affected by the dephosphorylation of 14-3-3β (bottom). (D) Dephosphorylation with λ-PPase abrogates the interaction of FLAG-myopodin (bottom) with endogenous 14-3-3β (top) and importin α (middle) from X. laevis extracts. (E) Putative phosphoacceptor sites within the 14-3-3–binding motifs of myopodin. S225 in motif#1, as well as T272 and S273 in motif#2, were substituted with alanine to remove putative phosphorylation sites. Replacement with aspartic acid or glutamic acid was done to mimic phosphorylation. (F) Purified FLAG-tagged, wild-type myopodin interacts with GST–14-3-3β. Substitution of S225 or T272 with alanine (S225A, T272A, and S225AT272A) abrogates 14-3-3β binding. Replacement of S273 with alanine (S273A) does not interfere with the binding of myopodin to 14-3-3β. (G) Substitutions of S225 or T272 with aspartic or glutamic acid, respectively, does not alter the binding of myopodin to 14-3-3 (top). However, after dephosphorylation, only S225DT272E retains strong binding to GST–14-3-3β. Single mutations (S225D and T272E) bind significantly less, and wild-type binding is abrogated.

Figure 5.

Two phosphorylated residues regulate nuclear localization of myopodin. (A) Confocal imaging of GFP-tagged myopodin in C2C12 myoblasts. Wild-type myopodin shows a predominantly nuclear localization (top), and S225AT272A displays a dramatically decreased nuclear localization (middle). In contrast, S225DT272E shows a primarily nuclear localization (bottom). Rhodamine-labeled phalloidin identifies myopodin-induced actin bundles, and DAPI visualizes nuclei. (B) Quantitative analysis is presented as a percentage of nuclear GFP–myopodin. 77.4% of wild-type myopodin is detected in the nucleus. S225A and T272A show 20.1% and 20.5% nuclear myopodin, respectively. S225AT272A shows a further decrease to 10.5%. In contrast, S273A does not alter nuclear localization (78.8%). S225D, S273D, and T272E show 75.6%, 75.7%, and 80.3% nuclear localization, respectively. S225DT272E displays 83.5% nuclear localization. Statistical significance was confirmed by ANOVA (P < 0.001). Error bars indicate standard deviation.

Figure 6.

A model for phosphorylation- and 14-3-3-dependent nuclear import of myopodin. (1) When phosphoacceptor sites in 14-3-3–binding motifs#1 (S225) and #2 (T272) are not phosphorylated, myopodin cannot interact with 14-3-3. Therefore, the NLSs in myopodin are not accessible for importin α binding, and myopodin cannot enter the nucleus. (2) After phosphorylation by serine/threonine protein kinases, myopodin binds to 14-3-3, rendering the NLSs accessible for importin α binding. (3) Importin α binds to the NLSs and mediates the nuclear import of myopodin (4).