PTEN binds MAGI-2 through a PDZ domain-mediated interaction. (A) Structure of MAGI-2. The two clones isolated in the two-hybrid screen (clones 4.1 and 20.1) are indicated by arrows. PDZ0 indicates a probable PDZ domain that does not have the consensus GLGF sequence. The numbering of PDZ domains 1–5 is based on the previously published nomenclature for MAGI-1 (30). GuK, guanylate kinase domain. (B) ³⁵S-labeled in vitro transcribed/translated MAGI-2 protein from clone 20.1 and clones of individual PDZ domains 2 or 4 was pulled down with GST alone or full-length PTEN-GST beads. Input represents 20% of the protein used. The bottom panel shows Coomassie staining of protein bound to beads. (C) 293T cells were transfected with wild-type (wt) or mutant FLAG-PTEN constructs and HA-MAGI-2. Lysates were immunoprecipitated (IP) with anti-FLAG antibody and immunoblotted with anti-HA antibody 12CA5 to detect MAGI-2 or anti-FLAG antibody to detect PTEN. (D) Transfection and immunoprecipitation (IP) were performed in 293T cells as indicated in C. (E) MAGI-2 was immunoprecipitated (IP) from the homogenate of fresh mouse brain (lane 4) by using antisera raised against the WW domains of S-SCAM (synaptic scaffolding molecule), the rat homologue of MAGI-2 (27). PTEN was immunoprecipitated (lane 5) by using polyclonal antisera (28) or an independently derived PTEN antibody (19) (not shown). Anti-Akt polyclonal antisera (New England Biolabs) (lane 3), pre-IP lysate (lane 6), and no lysate (lane 7) were used as controls. Immunoblots were analyzed by using rabbit polyclonal antisera against GST-MAGI-2.

Effect of MAGI-2 on PTEN activity. (A) 293T cells were transfected with MAGI-2, HA-Akt, and different concentrations of wild-type or mutant PTEN plasmid as indicated. The Akt kinase activity was measured by immunoprecipitation using an anti-HA antibody, followed by a kinase assay using histone H2B as substrate. MAGI-2 and HA-Akt expression was verified by immunoblotting. (B) Quantitative results from two independent experiments are shown.

The PDZ binding motif of PTEN affects protein stability. (A) Schematic diagram of PTEN mutants. WT, wild type. (B) Coimmunoprecipitation experiments were performed in 293T cells transfected with the indicated plasmids as described in Fig. 1. IP, immunoprecipitation. (C) Rat-1 (Upper) and LNCaP (Lower) cells were infected with retrovirus expressing the indicated PTEN mutant, then selected in G418. Lysates were analyzed for expression of FLAG-PTEN at multiple time points. Results from the 25-day time point are shown.

Subcellular localization of PTEN and MAGI-2. MDCK cells were plated onto fibronectin-treated cover slips, then transfected with FLAG-PTEN or HA-MAGI-2. After 24 h the cells were washed, fixed in paraformaldehyde, permeabilized in 0.2% Triton X-100, and blocked in 3% BSA in PBS containing 1 mM CaCl₂ and 1 mM MgCl₂. Cells were incubated in primary antibody in blocking buffer for 1 h at 37°C. Cells were washed in blocking solution plus 0.2% Triton. Cells were incubated with secondary antibody in blocking buffer for 45 min, washed, and then visualized by using a Zeiss LSM310 laser scanning confocal microscope.

The PTEN PDZ-binding motif is required for complete Akt suppression. (A) Rat-1 fibroblasts stably expressing wild-type (WT) FLAG-PTEN, FLAG-PTEN 1–377, or Neo control were serum-starved overnight, then challenged with serum for 30 min. Phospho-Akt, total Akt, and PTEN levels were measured by immunoblot (16). (B) Pulse–chase experiments were performed in 293T cells by transfection with indicated plasmids. PTEN was immunoprecipitated by using anti-FLAG antibody and visualized by autoradiography. Signals quantitated by a PhosphorImager are shown relative to time 0 of the chase. (C) NIH 3T3 cells stably expressing indicated PTEN proteins were harvested after serum stimulation at indicated times and processed as described in A, except anti-PTEN antibody (Santa Cruz Biotechnology) was used to measure PTEN expression.