Figure 1.
USF2 is phosphorylated by GSK3β.
(A) USF2 was immunoprecipitated from GSK3β+/+ and GSK3β−/− cells and phospho-USF2 protein levels were detected with phospho-threonine or phospho-serine antibodies. (B) USF2 was immunoprecipitated from HeLa cells transfected with pcDNA3-GSK3β-WT-HA. Following SDS-PAGE phosphoproteins were visualized with the Pro-Q Diamond Phosphoprotein Gel Stain. The total protein amount was detected by silver staining and GSK3β expression was verified by Western blotting. (C) Where indicated, HeLa cells transfected as above were treated with the GSK3 selective inhibitors BIO (1 µM), LiCl (10 mM), 1-Azakenpaullone (Aza, 7.5 µM) for 1 h. Proteins were isolated 24 h after transfection and detected by Western blotting. (D) Cells were transfected with expression vectors for USF2 and GSK3β and the cell extract was incubated with calf intestinal phosphatase (CIP) or only with buffer (Mock). Proteins were detected by Western blotting.
Figure 2.
GSK3β-mediated phosphorylation occurs in two USF2 domains.
(A) Schematic representation of the USF2 deletion mutants used to identify the domains that are phosphorylated by GSK3β. (B–E) Purified GST-tagged USF2 proteins were incubated with recombinant human GSK3β in the presence of [γ−32P] ATP. Proteins were separated by SDS-PAGE and incorporated radioactivity was detected by autoradiography. The total amount of proteins was detected by Coomassie staining. (F) HeLa cells were transfected with expression vectors for USF2 or the empty vector. USF2 was immunoprecipitated from the total cell extract and then incubated with recombinant human GSK3β in the presence of [γ−32P] ATP. Proteins were separated by SDS-PAGE and incorporated radioactivity was detected by autoradiography. The total amount of proteins was detected by silver staining.
Figure 3.
GSK3β-mediated phosphorylation of USF2 affects its transactivity, DNA binding and target gene expression.
(A) HeLa cells were cotransfected with pFR-5Gal4-RE-Luc, an expression vector for constitutively active GSK3β-S9A and WT or mutant pcDNA6-Gal4-USF2 (1-231) or the appropriate empty Gal4 vector. The luciferase activity was calculated as fold induction compared to the Gal4-USF2 (1-231)-WT luciferase activity after subtracting the values from the empty Gal4 expression vector. *, significant differences control vs. GSK3β. (B) Representative Western blot of the transfected constructs. 50 µg of protein from transfected cells were probed with an antibody against Gal4, HA-tag and α-tubulin. (C) Quantitative RT-PCR analyses of FAS, HO-1, PAI-1 and USF2 mRNA levels in GSK3β+/+ and GSK3β−/− cells. *, significant differences WT vs. GSK3β−/−. (D) Western Blot analyses of FAS, HO-1 and PAI-1 expression in GSK3β+/+ and GSK3β−/− cells. 50 µg of protein were subjected to Western analysis with antibodies against FAS, HO-1, PAI-1 or β-catenin, c-Myc or α-tubulin; the latter served as a loading control. (E) ChIP was performed in GSK3β+/+ and GSK3β−/− cells with either USF2 antibody, control IgG or RNA Pol II antibody. The quantitative PCR was performed with primers amplifying the FAS, HO-1, and PAI-1 promoter containing the USF2 binding sites, and with primers amplifying the β-actin promoter binding RNA Pol II as outlined in Materials and Methods. *, significant differences WT vs. GSK3β−/−.
Figure 4.
GSK3β-mediated phosphorylation of USF2 affects activation of target gene promoters.
(A–C) HeLa cells were cotransfected with the indicated luciferase gene construct and with WT or mutant p3xFLAG-USF2-myc-CMV24 or the appropriate empty vector. The measured luciferase activity is plotted as fold induction compared to the luciferase activity measured in the control transfected with the empty expression vector. *, significant difference WT vs. mutant. (D) Representative Western blot of the transfected constructs. 50 µg of protein from transfected cells were probed with an antibody against myc-tag and α-tubulin.
Figure 5.
Phosphorylation of USF2 by GSK3β increases domain (residue) distance.
(A–C) Simulated structures of wild type (A), S155 phosphorylated (B), and T230 phosphorylated (C) USF2 from the MD simulation trajectories. The side chains of Ser155 and Thr230 are shown in stick representation. Phosphorylated amino acids are labeled with an asterisk (*). The distance between side chain oxygen of S155 and T230 were analyzed (D). Phosphorylation of S155 (red) or T230 (green) increases the distance compared to non-phosphorylated USF-2 (black).
Figure 6.
The half-life of USF2 is induced upon GSK3β-dependent phosphorylation.
(A) GSK3β+/+ and GSK3β−/− cells were treated with 10 µg/ml CHX for the indicated time periods. Proteins were isolated, separated by SDS-PAGE and detected by Western blotting. Protein levels were quantified and the relative protein level of USF2 was blotted against the duration of CHX treatment for estimation of the half-life. The dashed line indicates the USF2 half-life where 50% of the USF2 protein level was reached. (B) Representative Western Blot. 50 µg of protein from were probed with an antibody against USF2 and α-tubulin.
Figure 7.
The phosphorylation of USF2 by GSK3β affects cell migration.
(A, B, C, D) GSK3β−/− cells were transfected with vectors allowing expression of USF2-S155A/T230A or USF2-S155D/T230D or an empty vector. Cellular viability, proliferation and cell migration were monitored by MTT, BrdU (A) and cell migration (B, C) assays. (B) Photographs from a representative Transwell chamber experiment. (C) Data represent the absorbance of crystal violet at 595 nm relative to the control. *, significant difference between GSK3β−/− cells and GSK3β−/− cells + USF2-S155D/T230D. (D) The expression of USF2 was controlled by Western blotting. 50 µg of protein from transfected cells were probed with an antibody against USF2 and α-tubulin.