Fig. 10. Model of TAZ-mediated transcriptional regulation. TAZ functions as a co-activator through binding PPXY motifs within the activation domains of transcription factors via its WW domain. 14-3-3 retains phosphorylated TAZ in the cytosol, negatively regulating this activity. TAZ also binds to the N-terminal PDZ domain of NHERF-2 at the plasma membrane. NHERF-2 interacts with ion channels, receptors and cytosolic signaling proteins, tethering them to the actin cytoskeleton through the ezrin–radixin–moesin (ERM) proteins. Thus, TAZ may link events at the plasma membrane and cytoskeleton to nuclear transcription.

Fig. 8. Co-expression of 14-3-3 localizes TAZ to the cytoplasm. (A) HeLa cells expressing GFP–TAZ wild-type (upper) or GFP–TAZ S89A (lower) together with Xpress-tagged 14-3-3 were stained with an anti-Xpress antibody. TAZ and 14-3-3 were visualized by green (left) or red (middle) fluorescence, respectively. Merged images are shown in the right panels. Bar = 10 µm. (B) Quantitative localization of GFP–TAZ WT or S89A in cells expressing 14-3-3 was performed as described in Figure 7. Results are the mean ± SD from three separate experiments.

Fig. 2. Phosphorylation of TAZ is required for the binding to 14-3-3. (A) [³⁵S]methionine-labeled TAZ was synthesized by in vitro translation in the presence of various concentrations of the protein kinase inhibitor K252a (0, 0.1, 1 and 10 µM). The amount of TAZ present in the lysate (middle) and bound by GST–14-3-3 (top) was quantitated (bottom). (B) ³⁵S-labeled TAZ was incubated with GST–14-3-3ε wild-type, the E180K mutant, the K49E mutant or GST control beads and analyzed by autoradiography. (C) Ser89 of TAZ is the binding site of 14-3-3 in vitro. ³⁵S-labeled TAZ (wild-type or S89A mutant) was pulled-down with GST–14-3-3 beads or GST control beads.

Fig. 4. TAZ contains a strong intrinsic transactivation domain in its C-terminus. (A) NIH-3T3 cells were transfected with 100 ng of tk-GALpx3-LUC (tkGAL) or tk-LUC (tk), 1 ng of pRL-EF, 500 ng of pEF-BOS backbone vector and 100 ng of the plasmid expressing the indicated GAL4(1–93) fusion protein. The luciferase activities relative to those obtained with GAL4(1–93) are shown. Results are the mean ± SD from three separate experiments. (B) Schematic illustration of mouse TAZ and its deletion constructs, and the transcriptional activity of each fragment fused to GAL4(1–93). NIH-3T3 cells were transfected with 100 ng of tk-GALpx3-LUC, 1 ng of pRL-EF, 500 ng of pEF-BOS and 100 ng of the plasmid expressing the indicated GAL4–TAZ fusion protein. The luciferase activities relative to those of GAL4(1–93) are indicated. Results are the mean ± SD from three separate experiments.

Fig. 5. The TAZ WW domain binds to a PPXY motif. (A) Blot overlay analysis verifies PPXY motif selection for the TAZ WW domain. GST or GST fusion proteins containing the WW domains of YAP and TAZ were analyzed by SDS–PAGE, transferred to membranes and probed using biotin-labeled peptides corresponding to the WW domain-binding motifs PPXY, PPLP, PPR and pS/T-P. In control experiments, the PPLP peptide bound strongly to the FBP11 WW domain, and the PPR peptide bound strongly to the FBP30 WW domains (data not shown). (B) The activation domains of many transcription factors contain a PPXY motif. DBD denotes the DNA-binding domain, AD the activation domain, and ID the autoinhibitory domain of PEBP2α.

Fig. 9. TAZ interacts specifically with the N-terminal PDZ domain of NHERF-2. (A) 293T cells were transfected with Flag-tagged YAP or TAZ constructs and His-tagged NHERF or NHERF-2, as indicated. Cell lysates were immunoprecipitated with anti-Flag followed by blotting with anti-His (upper) to detect NHERF and NHERF-2 (arrows). A cross-reacting immunoglobulin band is indicated by an arrowhead. The same blot was reprobed with anti-Flag (middle) to show TAZ or YAP expression. The amount of His-tagged NHERF or NHERF-2 within the total cell lysates is shown (lower). (B) TAZ interacts with PDZ1 of NHERF-2. [³⁵S]methionine-labeled TAZ was pulled-down with beads containing GST (lane 1), GST–NHERF-2 (lane 2), GST–NHERF-2N (containing the PDZ1) (lane 3) or GST–NHERF-2C (containing the PDZ2) (lane 4). Ten percent of the input is also shown (lane 5). (C) Z-mode scanning of HeLa cells expressing both GFP–TAZ (left panel) and RFP–NHERF-2 (middle panel). A merged image is shown in the right panel. Co-localization at the plasma membrane is indicated by arrowheads. Bar = 10 µm.

Fig. 6. TAZ functions as a transcriptional co-activator through its WW domain in a manner that is regulated by 14-3-3 and PDZ domain binding. (A) Jurkat cells were transfected with 100 ng of tk-GALpx3-LUC, 1 ng of pRL-EF, 200 ng of pEF-TAZ-N-Flag and 100 ng of plasmids expressing either GAL(1–93) (G) or GAL4(1–93) fused to a 33 amino acid region from PEBP2αB/AML1 (G-PY33) having wild-type or mutated versions of the PPXY motif. The luciferase activities relative to those obtained with GAL(1–93) are shown. Results are the mean ± SD from three separate experiments. (B) P19 cells were transfected with 200 ng of pFL56–3, 1 ng of pRL-EF, 200 ng of a pEF-based plasmid expressing full-length PEBP2αB1, PEBP2αB1(P1A) or PEBP2αB1(1–371), and the indicated amount of pEF-TAZ-N-Flag. Luciferase activities relative to those obtained without effector plasmids are shown. Results are the mean ± SD from three separate experiments. (C) Jurkat cells were transfected with 100 ng of tk-GALpx3-LUC, 1 ng of pRL-EF and 100 ng of GAL4(1–93) (open bars) or GAL4(1–93)-PY33 (filled bars) together with the indicated amounts of TAZ constructs. WT, pEF-TAZ-N-Flag; S89A, pEF-TAZ-N-Flag (S89A); WT/–4, pEF-TAZ-N-Flag/–4; S89A/–4, pEF-TAZ-N-Flag (S89A)/–4. Results are the mean ± SD from three separate experiments.

Fig. 3. TAZ associates with 14-3-3 through the phosphoserine-binding motif RSHpSSP in vivo. (A) 293T cell lysates (30 µg) were incubated with or without calf intestine alkaline phosphatase (CIP) [lanes 1 and 4, no phosphatase; lane 2, 200 U of CIP; lane 3, 200 U of CIP and a mixture of phosphatase inhibitors (Inh.)] and pulled-down with GST–14-3-3ε beads (lanes 1, 2 and 3) or GST–14-3-3ε (K49E) beads (lane 4). Proteins bound to beads were eluted, fractionated by SDS–PAGE and immunoblotted with an anti-TAZ antibody. No TAZ was detected in pull-downs using GST alone (data not shown). (B) 293T cells were transfected with Flag-TAZ wild-type (WT) or mutant (S89A). Lysates were immunoprecipitated with anti-Flag and immunoblotted for 14-3-3 (top). Blots were reprobed using anti-Flag (bottom). Cells transfected with pEF-BOS were used as a control. The band in the control lane and the upper band in the sample lanes is the cross-reacting immunoglobulin heavy chain.

Fig. 7. Subcellular localization of TAZ. (A) COS7 cells transiently expressing GFP–TAZ (a) and corresponding F-actin staining with rhodamine–phalloidin (b). Bar = 10 µm. In addition to cytoplasmic staining, TAZ is concentrated along specific portions of the plasma membrane (arrowheads) and within the nucleus. The nuclear staining of the upper cell in (a) is diffuse, while that of the lower cell is punctate, though these details are obscured in this exposure. (B) HeLa cells expressing GFP–TAZ (a), GFP–TAZ lacking the C-terminal four amino acids (d) or GFP–YAP (g). (b), (e) and (h) are the corresponding DAPI staining. (c) and (f) are Z-mode scanning of 293T cells expressing GFP–TAZ (c) and GFP–TAZ/-4 (f). Removal of the PDZ-binding motif eliminated punctate nuclear staining and plasma membrane staining (arrowheads). Bar = 10 µm. (C) Quantitative analysis of TAZ localization. The subcellular distribution of the various GFP fusion proteins was scored according to whether it was higher in the nucleus (N>C), evenly distributed between the nucleus and cytoplasm (N=C) or higher in the cytoplasm (N<C). The percentage of cells showing punctate nuclear staining (N.P.) was also quantitated. Results are the mean ± SD from three separate experiments.

Fig. 1. Characterization of TAZ. (A) Isolation of TAZ as a 14-3-3-binding protein. ³⁵S-labeled proteins from cDNA pool #636 were incubated with GST or GST–14-3-3-beads and bound proteins analyzed by SDS–PAGE/autoradiography. The arrow indicates a partial fragment of human TAZ. The mobility of molecular mass standards in kDa is indicated. (B) Sequence alignment of mouse and human TAZ with mouse, human and chicken YAP. The region surrounding the WW domain (indicated by a line) is boxed in yellow, the transcriptional activation domain in green, and the N-terminal region of homology in orange. The 14-3-3-binding site is indicated in red, and the conserved PDZ-binding motif is boxed in blue. The coiled-coil domain is shown as a cylinder. (C) The domain structure of mouse TAZ and mouse YAP65. CC, coiled-coil domain; SH3-Binding, SH3-binding motif; Pro-rich, proline-rich sequence; S89 of TAZ and S112 of YAP are the 14-3-3-binding site. Transactivation domains are dotted. (D) Immunoblot of endogenous and overexpressed TAZ: MDCK (lane 1); NIH-3T3 (lane 2); 293T (lane 3); and 293T cells transfected with pEF-TAZ (lane 4). The arrow indicates TAZ, and the arrowhead indicates YAP. The mobility of molecular mass standards is indicated. No immunostaining was observed using the pre-immune serum (data not shown). (E) Expression of TAZ determined by northern blot. A mouse (left panels) and human tissue blot (right panels) are probed as indicated. The positions of RNA size markers in kilobases are shown.