Figure 1.
Pak1 phosphorylates BAD (aa 104–141) primarily on S111 instead of S112/136.
(A) Pak1 phosphorylates S112/136A double mutant of BAD. 4 µg of each GST fusion protein of fragment BAD wild type (WT aa 104–141) or mutant BAD (aa 104–141) with the mutation of serine-to-alanine at 112, 136 or both sites was used as substrate for in vitro kinase assay with Pak1 in the present or absence of Cdc42L61, which was used to activate Pak1. Kinase assays were carried out using γ- [32P] ATP for 30 min at 30°C as described in Materials and Methods. One quarter of each reaction was separated by 10% SDS-PAGE, and phosphorylated proteins were detected by autoradiography. The position of phosphorylated Pak1, Cdc42 L61, and BAD WT and mutants were determined according to their sizes as shown in the Figure S1C. (B) Phosphoamino acid analysis of phosphorylated GST-BAD (aa 104–141). GST-WT BAD (aa 104-141) was phosphorylated by Pak1, and the proteins were run through SDS-PAGE. The phosphorylated BAD was isolated and subjected to acid hydrolysis and analyzed on a TLC plate as described in Materials and Methods. P-S, P-T and P-Y represent phosphor-serine, phosphor-threonine and phosphor-tyrosine, respectively. Pi indicates orthophosphate. (C–E) Mutation at S111 abolished BAD phosphorylation by Pak1. Kinase assay was carried out as in Figure 1A, except that substrates were the indicated GST-BAD mutants (aa 104-141). Panel E shows phosphorimager quantification of the bands in panel D (n = 3).
Figure 2.
Phosphorylation of BAD at S111 by Pak1 in 293T cells.
(A) Characterization of S111 and S112 antibodies used in this study. Purified GST-WT BAD (aa 104-141) and BAD (aa 104-141) mutants as indicated in the figure were incubated with or without Pak1 and phosphorylated with γ- [32P] ATP for 30 min at 30°C (upper panel) and subjected to autoradiography, or phosphorylated with unlabeled ATP (lower two panels) and probed with the S111 and S112 antibodies. (B and C) 293T cells were transfected with the indicated plasmids along with a BAD WT or S111A mutant plasmid. Pak1T423E, CDC42L61 and RacL61 are activated mutants of Pak, Cdc42 and Rac respectively. Pak1 KD is a kinase dead Pak1 mutant. Equal amounts of extracts were incubated with glutathione beads in a GST pulldown experiment. Samples were separated by 10% SDS-PAGE and then blotted with anti-phospho-S111 BAD antibody to assess BAD phosphorylation. Total GST-BAD and Pak1 are shown in the lower panels. (D) Endogenous BAD was immunoprecipitated with an anti-BAD antibody and blotted with S111 phospho-specific antibody. Total BAD is shown below the blot.
Figure 3.
Pak1 stimulates phosphorylation of Ser112 on BAD via Raf-1.
(A) Effects of kinase inhibitors on BAD S112 phosphorylation. 293T cells were co-transfected with expression vectors encoding full length (FL) WT BAD and Pak1 (KD or T423E). 24 hr after transfection, cells were starved for 16 hr and treated with GW5074 (5 µM), PD098059 (20 µM), Rapamycin (5 µM) or H89 (5 µM) for 2.5 hr as indicated. Equal amounts of proteins were used for Western blot to assess BAD phosphorylation at S112 or S111. The cell lysates were also subjected to immunoblotting with anti-BAD, anti-Myc, anti-phospho-ERK and anti-ERK antibodies. (B) Effects of inhibitors on forskolin stimulated activation of BAD phosphorylation. 293T cells were transfected with full length (FL) WT BAD for 24 hr, starved for 16 hr and treated with vehicle (DMSO), GW5074 (5 µM), or H89 (5 µM) for 2.5 hr prior to forskolin (50 µM) treatment. Equal amounts of proteins were used for Western blot to assess BAD phosphorylation at S112. The cell lysates were also subjected to immunoblotting with anti-BAD.
Figure 4.
Mutations on BAD S111 reduces Pak1 phosphorylation of BAD at S112.
(A) 293T cells were co-transfected with expression vectors encoding full length (FL) BAD (WT, S111A, or S112A) and Pak1 (KD or T423E). Equal amounts of proteins were used for Western blotting to assess BAD phosphorylation at S112. The cell lysates were also subjected to immunoblotting with anti-BAD and anti-Pak1 antibodies. (B) 293T cells were transfected with expression vectors encoding full length (FL) BAD (WT, S111A, or S112A). Equal amounts of proteins were used for Western blot to assess BAD phosphorylation at S112. The cell lysates were also subjected to immunoblotting with anti-Pak1 antibodies.
Figure 5.
Bcl2 and 14-3-3 binding to BAD and model.
(A) Effects of S111, S112 and S136 mutation(s) on BAD/Bcl-2 and BAD/14-3-3 complex formation. 293T cells were transfected with the full length (FL) BAD WT or mutant plasmids as indicated in the figure. Equal amounts of extracts were incubated with glutathione beads in a GST pulldown experiment. Samples were separated by 10% SDS-PAGE and then blotted with anti-Bcl-2 or anti 14-3-3 antibodies. Total Bcl-2 and 14-3-3 were also determined by Western blotting. (B and C) Protein levels were evaluated through densitometry and two different blots for were quantified. The ratio of immunoprecipitated and blotted Bcl-2 to total Bcl-2, and immunoprecipitated and blotted 14-3-3 to total 14-3-3 were plotted as B and C.
Figure 6.
Pak1 inhibition reduced BAD S111 and S112 phosphorylation.
(A) BAD S111 phosphorylation was observed in a lung cancer cell line H358 and the MPNST cell lines 90-8 and STS26. (B) Effects of protein kinase inhibitors on BAD S111 and S112 phosphorylation. The MPNST cells, ST88-14, were starved and treated with vehicle (DMSO), IPA3 (20 µM), GW5074 (5 µM), or H89 (5 µM) for 15 hr. Equal amounts of proteins were used for Western blot to assess BAD phosphorylation at S111and S112. The cell lysates were also subjected to immunoblotting with anti-BAD. As a control, Raf-1 phosphorylation at S338 was examined. (C) Model showing that Pak1 phosphorylates BAD at S111 directly, but most of its protective effects are mediated through Raf-1 and phosphorylation of S112. S111 also appears to facilitate phosphorylation at S112 because mutating it reduces phosphorylation by Pak1, but not phosphorylation by PKA. S136 is the most important phosphorylation site for the BAD/14-3-3 complex formation as previously observed. The enhanced Bcl-2 binding by the triple mutant is likely the result of the release of BAD from 14-3-3 sequestering in addition to the enhanced Bcl-2 association.