点击显示 收起
【关键词】 Domains
Many cellular signaling pathways share regulation by protein phosphatase-2A (PP2A), a widely expressed serine/threonine phosphatase, and the heterotrimeric G protein G12. PP2A activity is altered in carcinogenesis and in some neurodegenerative diseases. We have identified binding of G12 with the A subunit of PP2A, a trimeric enzyme composed of A (scaffolding), B (regulatory), and C (catalytic) subunits and demonstrated that G12 stimulated phosphatase activity (J Biol Chem 279: 54983-54986, 2004). We now show in substrate-velocity analysis using purified PP2A that Vmax was stimulated 3- to 4-fold by glutathione transferase (GST)-G12 with little effect on Km values. To identify the binding domains mediating the A-G12 interaction, an extensive mutational analysis was performed. Well-characterized mutations of A were expressed in vitro and tested for binding to GST-G12 in pull-down assays. G12 binds to A along repeats 7 to 10, and PP2A B subunits are not necessary for binding. To identify where A binds to G12, a series of 61 G12 mutants were engineered to contain the sequence Asn-Ala-Ala-Ile-Arg-Ser (NAAIRS) in place of 6 consecutive amino acids. Mutant G12 proteins were individually expressed in human embryonic kidney cells and analyzed for interaction with GST or GST-A in pull-down assays. The A binding sites were localized to regions near the N and C termini of G12. The expression of constitutively activated G12 (QL12) in Madin Darby canine kidney cells stimulated PP2A activity as determined by decreased phosphorylation of tyrosine 307 on the catalytic subunit. Based on crystal structures of G12 and PP2A A, a model describing the binding surfaces and potential mechanisms of G12-mediated PP2A activation is presented.
Heterotrimeric G proteins regulate fundamental processes in all eukaryotic cells. The canonical signaling pathway in which a membrane-bound G protein-coupled receptor activates G and G subunits to stimulate downstream effectors has been significantly extended in recent years. In particular, G protein signaling has been shown to regulate unique cellular functions in specialized membrane domains and interact with numerous protein-kinase signaling pathways (Luttrell and Luttrell, 2004; El-Shewy et al., 2006). Novel regulatory mechanisms of G protein signaling have also been identified through interactions of G subunits with proteins that modulate G function. These include regulator of G protein signaling (RGS) proteins that stimulate G GTPase activity and G protein regulatory (GPR) proteins that inhibit GDP release (Hollinger and Hepler, 2002; Lanier, 2004). G12 and G13 comprise one of the four heterotrimeric G protein families that also include Gs,Gi/o, and Gq.G12/13 regulate the actin cytoskeleton (Buhl et al., 1995) and epithelial cell junctions (Meigs et al., 2002; Meyer et al., 2003) in addition to other cellular processes, including transformation of fibroblasts (Jiang et al., 1993), stimulation of apoptosis (Berestetskaya et al., 1998), neurite retraction in PC12 cells (Katoh et al., 1998), directed cell movement within the developing embryo (Lin et al., 2005), and cell migration (Goulimari et al., 2005). G12 and G13 interact with the RGS domain of the Rho exchange factor p115-RhoGEF and G12 binds the RGS domain of axin (Kozasa et al., 1998; Stemmle et al., 2006). In addition, G12 and/or G13 bind to a diverse group of other signaling and structural proteins, including Bruton's tyrosine kinase, heat shock protein-90, the tight junction protein ZO-1, and E-cadherin (Jiang et al., 1998; Meigs et al., 2001; Vaiskunaite et al., 2001; Meyer et al., 2002; Waheed and Jones, 2002).
PP2A is composed of three subunits: A (scaffolding), B (regulatory), and C (catalytic). There are two isoforms for the A and C subunits ( and ). There are four families of B subunits (B/PR55, B'/B56/PR61, B'', and B''') encoding at least 14 separate gene products and several alternatively spliced variants that regulate localization of the core enzyme (A-C) and C subunit activity (Janssens and Goris, 2001). Catalytic subunits are not found as free monomers in cells but always in complex with the A subunit forming an A-C core enzyme. B subunits reversibly interact with the core enzyme to regulate PP2A function, and other proteins can mimic (replace) B subunits. We demonstrated previously G12 stimulation of PP2A phosphatase activity in vitro and in COS cells (Zhu et al., 2004). Many signaling pathways share involvement of G12 and PP2A, although until recently a direct link between these two proteins has been missing. Some cellular processes modulated by both G12/13 and PP2A include cellular transformation, growth, apoptosis, and stress responses. Furthermore, PP2A has been directly implicated in carcinogenesis and neurodegenerative diseases (including Alzheimer's) (Janssens and Goris, 2001). PP2A is also localized in epithelial cell tight junctions and regulates the phosphorylation state of numerous proteins during maintenance and assembly of the junction (Nunbhakdi-Craig et al., 2002). We and others have observed G12 in the tight junction (Dodane and Kachar, 1996; Meyer et al., 2002, 2003) suggesting that both G12 and PP2A are localized within the same subcellular microdomain. Taken together, we hypothesize that G12 regulation of PP2A is a critical signaling pathway important to diverse cellular functions and in the pathophysiology of certain disease states. To define the mechanism(s) of G12 regulation of PP2A and provide the basis for pharmacological modulation of these signaling pathways, we have performed extensive mutagenesis and in vitro binding analyses to identify the requisite binding domains of A and G12. In addition, we provide evidence that G12 activation in cells stimulates PP2A, and in vitro kinetic analysis reveals that G12 stimulates Vmax without significantly affecting Km values. Based on crystal structures of PP2A A and G12 subunits and the results of binding studies with mutant proteins, it is possible to propose a model of the mechanism by which G12 regulates PP2A function.
Chemicals, Antibodies, and cDNA Constructs. Purified PP2A core enzyme (A-C) was obtained from Upstate Biotechnology (Lake Placid, NY). Mouse -actin antibody was purchased from Sigma-Aldrich (St. Louis, MO), and p-Tyr307 antibody was from Epitomics (Burlingame, CA). The rabbit C-subunit antibody was kindly provided by Dr. David Virshup (University of Utah, Salt Lake City, UT). Recombinant G purified from baculovirus was kindly provided by Dr. Thomas Michel (Brigham and Women's Hospital, Boston, MA). G12 antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). G12 Tet-off MDCK cells and GST-G12 were described previously (Meyer et al., 2002). Glutathione-Sepharose was from GE Healthcare (Chalfont St. Giles, Buckinghamshire, UK). Full-length mouse A cDNA was subcloned into pGEX using standard techniques, and GST-A fusion protein was expressed and purified from Escherichia coli as described previously (Luo and Denker, 1999). Substitution mutants within myc-tagged QL12, in which regions of the cDNA encoding consecutive sextets of amino acids are replaced with DNA sequence encoding the sextet NAAIRS, were engineered using oligonucleotide-directed mutagenesis as described previously (Meigs et al., 2005). All constructs were verified by sequencing.
PP2A Phosphatase Activity. Phosphatase activity was determined as described previously (Zhu et al., 2004). In brief, phosphatase activity was determined by Malachite Green phosphatase assay (Upstate). The phospho-peptide (K-R-pT-I-R-R) was used as substrate to determine PP2A activity alone or PP2A in combination with GST or GST-G12 at 1:1 M ratios. A standard curve was generated in each assay, and PP2A activity was determined in duplicate for each concentration. The reaction was initiated by the addition of substrate at t = 0, and after 30 min, the reaction was terminated, absorbance was measured at 624 nm, and enzyme activity was determined (expressed in nanomoles of phosphate per minute per unit). Km and Vmax values were calculated from substrate-velocity curve analysis, and Lineweaver-Burk plots were generated using Prism software (GraphPad Software, San Diego, CA).
In Vitro Translation. A,B, and mutant A cDNAs in pBluescript or pcDNA3 were in vitro translated using 0.5 to 1 µg of plasmid, the appropriate RNA polymerase in a coupled rabbit reticulocyte translation system (TNT system; Promega, Madison, WI) plus [35S]methionine (PerkinElmer Life and Analytical Sciences, Boston, MA) as described previously (Denker et al., 1995). Protein expression was analyzed by SDS-PAGE and autoradiography.
GST Pull-Downs of A Subunits. 35S-labeled A subunits were precleared by incubation with glutathione agarose beads for 1 h at 4°C. Translates (5-20 µl, depending on translation efficiency) were incubated with GST or GST-G12 (0.5-1 µg) for 3 h in buffer A (50 mM HEPES, pH 7.5, 1 mM EDTA, 3 mM dithiothreitol, 10 mM MgSO4, and 0.05% polyoxyethylene-10-lauryl ether). Samples were centrifuged, washed three times with buffer A, and eluted with SDS-PAGE sample buffer. Samples were analyzed by SDS-PAGE and autoradiography. Band intensity was analyzed using NIH Image 1.63 (http://rsb.info.nih.gov/nih-image/) (Wayne Rasband), and relative intensity of precipitated A was expressed as a fraction of the total and normalized to the wild-type A control.
GST-G12 Pull-Downs of 35S-Labeled A and B Subunits. A and B subunits were in vitro translated and 35S-labeled in rabbit reticulocyte lysate as described above. Equivalent amounts (determined by densitometry) of the labeled proteins were mixed together and incubated with 1 µg of GST or GST-G12 for 3 h in buffer A. Samples were centrifuged, washed three times with buffer A, and eluted with SDS-PAGE sample buffer. Samples were analyzed by SDS-PAGE and autoradiography.
Preparation of G12 Mutants from Human Embryonic Kidney Cell Lysates. Human embryonic kidney (HEK293) cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Hyclone Laboratories, Logan, UT) and penicillin/streptomycin and were maintained at 37°C in a 5.0% CO2 atmosphere. For each G12 mutant, 7 µg of plasmid DNA were used to transfect a 10-cm dish of HEK293 cells at 60 to 80% confluence using Lipofectamine reagent (Invitrogen, Carlsbad, CA) in accordance with the manufacturer's instructions. At 48 h after transfection, cells were rinsed twice with phosphate-buffered saline, scraped from the dish, pelleted at 800g, and then 0.5 ml of ice-cold buffer A containing 1% polyoxyethylene-10-lauryl ether supplemented with protease inhibitors was used to resuspend each cell pellet. Samples were mixed by inversion at 4°C for 30 min and then centrifuged at 100,000g at 4°C for 1 h. Supernatants were snap-frozen in liquid N2 and stored at -80°C.
GST Pull-Down of G12 Mutants. GST-A and GST were expressed in E. coli and purified by immobilization onto glutathione-Sepharose as described previously (Luo and Denker, 1999; Meigs et al., 2001). For each mutant form of G12 tested, 45 µl of HEK293 supernatant (prepared as described above in buffer A + 1% detergent) was diluted to 450 µl using buffer A with no detergent (final detergent concentration, 0.1%). Five percent of each diluted lysate was set aside as "lysate load," and the remainder was equally divided and incubated with glutathione-Sepharose bound to either GST-A or GST. Samples were incubated for 2 h at 4°C, pelleted, and the supernatant was discarded. Beads were washed three times with buffer A + 0.05% polyoxyethylene-10-lauryl ether followed by resuspension of the Sepharose pellets in SDS sample buffer and heating to 72°C for 10 min. For each sample, 80% of the volume was subjected to SDS-PAGE and subsequent blot transfer. Immunoblots were blocked in TBST (50 mM Tris, pH 7.7, 150 mM NaCl, and 0.05% Tween 20) supplemented with 5% powdered milk for 2 h and then incubated with rabbit anti-G12 antibody (at 1:500 dilution) for 2 h. After three 10-min washes using TBST, alkaline phosphatase-conjugated anti-rabbit antibody was diluted 1:7500 in TBST + milk and applied for 1 h. Three additional washes were performed using TBST, and blots were developed using nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate reagent (Promega) according to the manufacturer's instructions. Results were documented using a Kodak Gel Logic 100 scanner, and band intensities were quantified using Kodak 1D image analysis software (Eastman Kodak, Rochester, NY). The remaining 20% of each sample volume was subjected to SDS-PAGE, and gels were stained overnight with Coomassie blue, destained, and photographed.
Analysis of Tyr307 Phosphorylation in G12-Expressing MDCK Cells. Wild-type G12 and QL12-MDCK cells were cultured ± doxycycline (dox) for 72 h to induce G12 expression using conditions described previously (Meyer et al., 2002). In brief, MDCK cell lysates were prepared by scraping cells in buffer B (1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris, pH 7.5, 150 mM NaCl, 4 mM EDTA, 1 mM Na3VO4, 25 mM NaF, and protease inhibitor cocktail), briefly sonicated, centrifuged, and the resulting supernatant saved for analysis. Protein concentration was determined by BCA protein assay kit (Pierce, Rockford, IL), and equivalent protein amounts were analyzed by SDS-PAGE and Western blots as described above.
Modeling of Crystal Structures. Cnd4 version 4.1 was obtained from http://www.ncbi.nlm.nih.gov/Structure/CN3D/cn3d.shtml, and G12 crystal structure (1ZCA; Kreutz et al., 2006) and A crystal structure (1B3U_A) were retrieved from the Protein Data Base and independently studied using Cnd4. Mutations were highlighted in yellow and various aspects of the crystal structure were hidden. Images were exported as PNG files, and resolution/size was adjusted in Adobe Photoshop and then labeled and assembled in Adobe Illustrator (Adobe Systems, Mountain View, CA).
We demonstrated previously increased PP2A phosphatase activity when core enzyme (A-C) was incubated with GST-G12 or recombinant G12 but not with several other purified G subunits (Zhu et al., 2004). In addition, there were no detectable differences in affinity of GDP-, GTPS-, or AlF -4 liganded G12 for A in immunoprecipitation studies, and recombinant G12 stimulated phosphatase activity similarly in the inactive and active conformations. Based on the similar binding and stimulation of PP2A by GST-G12 and baculovirus-purified G12, GST-G12 was used to determine Km and Vmax values for PP2A core enzyme from substrate-velocity and Lineweaver-Burk plots in the presence of equimolar GST or GST-G12. Figure 1 shows the initial velocity versus substrate plot of mean PP2A activity ± preincubation with GST (control) or GST-G12 in three separate experiments each performed in duplicate over a range of substrate concentrations. Figure 1 (inset) shows the Lineweaver-Burk plot of the velocity versus substrate data for PP2A alone, PP2A + GST, and PP2A + GST-G12. The Km and Vmax values for each condition were determined in GraphPad Prism as described under Materials and Methods and are summarized in Table 1. A 3- to 4-fold stimulation of Vmax by GST-G12 compared with GST was observed, with no significant effect on Km values. Preincubation of GST-G12 with GTPSor G subunits had no demonstrable effect on the kinetics (results not shown). However, it is not known whether GST-G12 exchanges guanine nucleotides or interacts with G. Nevertheless, these findings are consistent with our previous observations with purified G12. Preincubation of G12 with GTPS (Zhu et al., 2004), or G (data not shown) had no effect on steady-state phosphatase activities. Together, these findings suggest that the stimulation of PP2A phosphatase activity in vitro is not dependent on the activated G12 conformation or affected by G12 complexing with G.
Fig. 1. G12 stimulates the Vmax value of PP2A core enzyme. PP2A core enzyme (A-C) was preincubated with GST or GST-G12 at a 1:1 M ratio, and substrate velocity was determined over a range of substrate concentrations as described under Materials and Methods. Results are the mean of three experiments, each done in duplicate. The Lineweaver-Burk plot was obtained using GraphPad Prism and is shown as an inset. The summary of Vmax and Km values is shown in Table 1.
TABLE 1 Summary of Km and Vmax values determined from substrate velocity analysis as described under Materials and Methods
Values are the mean ± S.D. of three experiments performed in duplicate.
To gain additional insights into how G12 stimulates PP2A phosphatase activity, we sought to identify the amino acid regions of A and G12 necessary for the interaction. A series of well-characterized mutants of A (Ruediger et al., 1992, 1994) were [35S]methionine-labeled, incubated with GST or GST-G12, and analyzed in GST pull-down assays (Fig. 2). GST-G12 and GST proteins are shown in Fig. 2A, and the [35S]methionine-labeled A mutants are shown in Fig. 2B. The results of the pull-downs are shown in Fig. 2C next to a schematic of each A mutant protein. Wild-type A, consisting of 589 amino acids, was analyzed in parallel with each mutant, and the binding of each mutant is expressed relative to the wild-type control in that experiment. The A subunit is composed of 15 tandem HEAT repeats (huntingtin, elongation, A subunit, target of rapamycin) each composed of approximately 39 amino acids. The repeats in the A subunit were proposed to be composed of two superimposed -helices yielding an elongated molecule often described as a hook (Fig. 7) (Ruediger et al., 1992). This structure was subsequently confirmed, and molecular details were extended by crystal structure analysis (Groves et al., 1999). Mutational analysis of A has identified the domains required for B and C subunit binding (Ruediger et al., 1992, 1994). C subunits interact with repeats 11 to 15 whereas the heterogeneous regulatory B subunits require repeats 1 to 10 for binding. The recent crystal structure of trimeric A-B561-C confirms C subunit interaction with A repeats 11 to 15 but reveals B' subunit binding to repeats 2 to 7 (Cho and Xu, 2007). In mutational studies with various B subunits, A repeat 5 was required for all known B subunit interactions (Ruediger et al., 1999). The analysis shown in Fig. 2 indicates that G12 binds to HEAT repeats 7 to 10 within A. There was a partial reduction in binding (25-50% of control) seen with deletion of repeat 7 (7) or repeat 8 (8), and there was complete loss of binding with deletion of repeat 9 (9) or repeat 10 (10). Deletion of N-terminal repeats had no effect on G12 binding (1, 1-2, and 1-3), and mutants with single repeats 3, 4, or 5 deleted also bound normally. The normal interaction of G12 with 5 indicates that B subunit binding is not required for the G12 interaction. Deletion of any single repeat between 11 and 15 prevents binding of the C subunit (Ruediger et al., 1992), yet G12 interacted normally with 15 and 14-15 and was only slightly reduced with the larger 11-15 deletion. Consistent with the single repeat deletions, deletion of repeats 9 to 15 (9-15) resulted in complete loss of binding to G12.
Fig. 2. G12 binds to repeats 7 to 10 on A. A, Coomassie blue stained gel of GST-G12 and GST fusion proteins. B, autoradiogram of [35S]methionine labeled and in vitro translated wild-type A and A mutants. Translates (0.5-2 µl) were analyzed by SDS-PAGE and autoradiography. Exposure time was 4 to 12 h. White vertical lines indicate separate gels. C, wild-type (WT) A schematic with 15 boxed repeats is shown on top. The deletions of N- and C-terminal repeats are shown as GAPs, and single repeat deletions are highlighted by a black box. A representative autoradiogram of GST pull-down with GST-G12 and GST is shown for each mutant in the adjacent panel. The input of labeled A protein is 1% of the total used in the binding assay. Wild-type binding was set at 100%, and in each experiment, a wild-type control was included for comparison of mutants. Binding for each mutant was scored from quantification of band intensity determined by densitometry and expressed relative to wild-type control. ++++, >75% of wild type; +++, 51 to 75%; ++, 25 to 50%; +, 10 to 24%; and -, <10%. Each mutant was tested two to four times, and scoring is the average value obtained from the individual experiments. Exposure times ranged from 1 to 7 days.
Fig. 7. Crystal structures of A and G12 binding domains. A and B, the PP2A-A crystal structure was analyzed using Cn3d as described under Materials and Methods. The N and C termini are noted. In A, the G12 binding region (repeats 7-10) is colored yellow within this hook-shaped structure. In B, the structure has been rotated to highlight an intrarepeat surface (the inside region where B and C subunits bind) and the interrepeat loops on the outside. Repeats 9 (R9) and 10 (R10) are shown in yellow. C, crystal structure of G12 (amino acids 47-379), with the N terminus denoted by a yellow dash. The 4 and 5 helices and the 6 sheet are noted with arrows, and the 5 helix ends at the C terminus. The NAAIRS mutants with reduced or absent binding to A are highlighted yellow. D, the identical structure in C revealing only the amino acids identified in the interaction sites. With the exception of t-u, the contact regions are predominantly within the C-terminal 60 amino acids and near the N terminus. Note that although the N-terminal structure is missing in the crystal, it is predicted to lie in close proximity to the C terminus. E, the structure in D has been rotated 90° and up so that now the C terminus (5 helix) is projecting directly out of the image.
These results indicate that B subunit interactions are not required for G12 binding to the core enzyme. However, this does not exclude the possibility that G12 can interact with the PP2A holoenzyme (A-B-C). To test whether G12 may interact with the holoenzyme, A and B subunits were in vitro translated in rabbit reticulocyte lysate, labeled with [35S]methionine and allowed to equilibrate with the endogenous subunits to form holoenzymes (as shown previously; Ruediger et al., 1992). Pull-down experiments of labeled A and/or B with GST or GST-G12 were analyzed for detection of labeled subunits with the precipitated protein complex. As shown in Fig. 3, there is readily detectable A and B subunits in the GST-G12 pull-down that was not seen with GST. This finding suggests that G12 may interact with a holoenzyme in this assay, and we have not excluded the possibility of a direct G12-B interaction. The B signal was stronger than for A in this experiment, but because the amounts of endogenous A and B in the rabbit reticulocyte lysate are unknown, it is not possible to compare the relative binding of these subunits.
Fig. 3. G12 binds to labeled A and B subunits in vitro. Autoradiogram of pull-down with GST-G12 or GST of [35S]methionine-labeled A and B that were in vitro-translated (IVT) separately and then mixed together (A + B) and equilibrated with endogenous proteins in the rabbit reticulocyte lysate for 30 min at 23°C before analysis as described under Materials and Methods. Exposure time is 48 h.
To map the determinants within G12 that are important for its interaction with PP2A A, we first engineered a series of amino acid substitution mutants that span the full length of G12. To produce each mutant, a sextet of consecutive amino acids in the primary sequence of myc-tagged QL12 (constitutively active G12) was replaced by the sequence NAAIRS. This series of mutations was created in the QL12 protein to facilitate the study of G12 binding proteins that would be expected to be downstream effectors in canonical G protein signaling. The NAAIRS sequence is believed to be a well-tolerated substitution in proteins because of its appearance in both -helical and -sheet secondary structures (Wilson et al., 1985) and was recently used to successfully engineer a mutant of G12 that was selectively uncoupled from a known binding partner, p115RhoGEF (Meigs et al., 2005). It was shown previously that wild type and QL12 interact with PP2A A (Zhu et al., 2004). The myc-tagged QL12 was used as starting material so that the subsequently engineered NAAIRS mutants could be distinguished from endogenous G12 when expressed in cultured mammalian cells (Meigs et al., 2005). The NAAIRS mutants were given consecutive alphabetical designations: a to z followed by aa to zz, ending with aaa to kkk (Fig. 4). We were able to express 61 of these 63 mutants in HEK293 cells, as determined by immunoblot analysis (data not shown). Mutant w was not engineered because of the positioning of the myc epitope tag within it, and mutant zz was undetectable in cell lysates. Mutants were detergent-extracted from HEK293 lysates and then screened for binding to GST-A as described under Materials and Methods.
Fig. 4. G12 amino acid sequence, NAAIRS mutants, and summary of binding assays. Amino acid sequence of G12 indicating sites of NAAIRS substitutions. The starting point for the mutants was myc-tagged QL12 cDNA and is described under Materials and Methods. The myc tag was inserted between proline 139 and valine 140 (mutant, w). The w sequence, therefore, was not converted to a NAAIRS mutant. Sextets replaced by the sequence NAAIRS are flanked by vertical bars. The activating Q-to-L substitution is indicated by a dashed square. Regions in which replacement by NAAIRS resulted in a 75 to 90% decrease in binding to A are indicated by open boxes, and those that resulted in a 90 to 100% decrease are indicated by shaded boxes. Mutants with >50% binding were tested twice, and those with <50% binding (open and shaded boxes) were tested three times.
Next, an A GST fusion protein (GST-A) was constructed and purified as described under Materials and Methods. GST-A migrates at its predicted molecular mass of 80 kDa during SDS-PAGE (Fig. 5A). GST-A was tested for its ability to precipitate myc-tagged QL12 and the full panel of NAAIRS substitution mutants. Seven or eight mutants were screened in each experiment, and myc-tagged QL12 was included as an internal control and was precipitated by A in all experiments (Fig. 5B). Whereas most of the NAAIRS mutants were precipitated by GST-A, several mutants displayed severely impaired or abolished binding to A (Fig. 5B, and summarized in Fig. 4). To provide a quantitative analysis of the relative affinity of these mutants for A and to account for differing levels of expression of these mutants in HEK293 cells, we calculated a normalized ratio using the densities of the bands on Western blots for the starting material and the pull-down. This was compared with the same ratio determined for myc-QL12 within the same experiment. These results are summarized in Fig. 4. The majority of the NAAIRS mutants showed a normalized ratio that was greater than 50% of the control value (no marking in Fig. 4), but several mutants displayed a dramatic loss of interaction with A. Mutants showing moderate-to-severe impairment of interaction (t, rr, bbb, eee, fff, and jjj; 10-25% of control) are shown in open boxes in Fig. 4, and those with abolished interactions (<10% of control) are highlighted in the gray boxes in Fig. 4. These latter mutants (a, f, i, aaa, ccc, ddd, and hhh), with the exception of mutant u, were localized near the N and C termini of G12 (Fig. 4). As an additional control, myc-QL12 and all NAAIRS mutants were tested for precipitation by glutathione-Sepharose bound GST; this complex did not precipitate myc-QL12 or any of its mutant variants (Fig. 5B).
Fig. 5. Representative G12 Western blots of NAAIRS mutants with intact and disrupted A binding. A, purification and immobilization of GST-fused A. GST-A and GST, previously expressed in bacteria and immobilized on glutathione-Sepharose (see Materials and Methods), were subjected to SDS-PAGE and then stained using Coomassie blue. Molecular mass standards (in kilodaltons) are indicated on the right. B, immunoblot analysis of G12 precipitation. Lysates from HEK293 cells expressing either myc-tagged QL12 or each indicated NAAIRS substitution mutant of this protein (e.g., I-myc, DDD-myc) were pulled down with GST alone or GST-A as described under Materials and Methods. For each panel, the left lane (GST-A) indicates G12 that was precipitated by GST-A, the middle lane (GST) indicates G12 that was precipitated by GST alone, and the right lane (load) indicates a fraction (5%) of the HEK293 cell lysate that was set aside before precipitation.
Our next goal was to determine whether G12 regulates PP2A activity in cells. To address this question, we used well-characterized MDCK cells with inducible expression of G12 and QL12 (Meyer et al., 2002, 2003). PP2A activity can be assessed by the relative tyrosine phosphorylation on the catalytic subunit (Tyr307), and decreased phosphorylation is associated with increased PP2A activity (Chen et al., 1992). Figure 6 shows that endogenous G12 in MDCK cells is not readily detectable by Western blot with available anti-G12 antibodies. However, after 72 h in the absence of doxycycline, QL12 was readily detectable. The Western blots were stripped and reprobed for actin, total C subunit, and phosphotyrosine 307 levels. There was a small but consistent reduction in normalized phosphotyrosine 307 levels in QL12-expressing MDCK cells (-dox) that was not seen in +dox QL12 cells (no QL12 expression) or in wild-type G12-MDCK cells ± dox (Fig. 6). This finding is consistent with G12 activation of PP2A phosphatase activity in MDCK cells and our prior finding of G12/PP2A-stimulated dephosphorylation of the microtubule-associated protein tau in transfected COS cells (Zhu et al., 2004).
Fig. 6. QL12 expression in MDCK cells reduces Tyr307 phosphorylation. Previously characterized, stably transfected MDCK II cells with inducible expression of G12 (WT) or a constitutively active mutant of G12 (QL) were cultured for 72 h ± doxycycline and analyzed by Western blotting as described under Materials and Methods. Equivalent amounts of total protein (50 µg) were analyzed by SDS-PAGE and Western blotting with indicated antibodies. The nitrocellulose was sequentially stripped and reprobed. Band intensities were determined using NIH Image software and phosphotyrosine 307 density in -dox is compared with +dox control for WT and QL12-MDCK cells. Results are the mean of four separate experiments analyzed with GraphPad Prism software. Two-tailed t test indicated significant reduction; *, p = 0.03.
We reported previously that G12 binds to the A subunit of PP2A and stimulates its activity (Zhu et al., 2004). We now identify the specific regions of G12 and A necessary for this interaction and show that G12 binding to PP2A core enzyme stimulates catalytic activity with little effect on substrate affinity. Analysis of well-characterized deletions of individual HEAT repeats in A identifies repeats 7 to 10 as the critical region for G12 interaction. G12 binding does not require the B subunit to be bound, but in vitro, G12 may interact with the PP2A holoenzyme. Analysis of 61 distinct G12 mutants identifies regions near the N and C termini as important for G12 binding to A. Furthermore, using a cell culture model we provide evidence for G12-dependent activation of PP2A. Based on these results and crystal structures of G subunits and PP2A, their interacting surfaces are modeled and possible mechanism(s) of regulation are discussed (Fig. 7).
PP2A is composed of three subunits: A (scaffolding), B (regulatory), and C (catalytic). Catalytic subunits are always found in complex with the A subunit forming an A-C core enzyme. B subunits reversibly interact with A-C to regulate PP2A function, and other proteins can mimic B subunits. For instance, tumor (T) antigens can displace B subunits to affect PP2A activity, and this has been implicated in carcinogenesis (Janssens and Goris, 2001). SV40 small-T antigen requires A repeats 3 to 6, and Polyoma small T and middle T antigens require repeats 2 to 8 (Ruediger et al., 1992, 1994). The results of the binding experiments with A deletion mutants shown in Fig. 2 reveal a unique pattern of G12 interaction. G12 binds to repeats 7 to 10, and a complete loss of binding occurs when either repeat 9 or repeat 10 is deleted. This pattern is distinct from B and C subunit binding and from binding of T antigens, allowing for potentially complex regulation of the enzyme. B subunits interact along the first 10 repeats, and all tested B subunits require repeat 5 to bind. G12 binds 5 and wild-type A with similar affinity, indicating that B subunit binding to the A-C core enzyme complex is not required for G12 interaction. G12 binding does not overlap on A with C subunit binding (repeats 11-15), a finding consistent with simultaneous binding of C subunit and G12 to A. This proximity is likely to be important for inducing conformational changes resulting in enhanced phosphatase activity. The B561 subunit makes direct contact with the C subunit and is postulated to affect substrate specificity (Cho and Xu, 2007). In an analogous manner, the binding of G12 to the site immediately adjacent to the C subunit may permit direct interaction(s) of G12 with the C subunit.
The results shown in Fig. 3 suggest that a tetrameric complex of G12 with the holoenzyme may form in vitro. Consistent with this model, the crystal structure of A-B561-C reveals binding of B' to A repeats 2 to 7 (Cho and Xu, 2007), indicating that for this specific holoenzyme, the G12 binding sites at repeats 8 to 10 should be accessible to G12 for binding. However, we have not excluded the possibility that G12 binds to B in the absence of A or C subunits. Furthermore, our analysis is unable to distinguish G12 binding to the intrarepeat loops (inner surface of hook) from binding to the inter-repeat loops (outer surface, Fig. 7B). The diversity in structure of B subunits makes it likely that there are differences in how B subunits interact with A. Thus, the regulation of PP2A by B subunits and G12 may depend on the specific B subunit. A more refined mutational analysis using purified proteins will be necessary to determine whether G12 binds to B in the absence of A and C subunits and whether G12 interacts with PP2A holoenzymes.
The mutational analysis of G12 indicates that regions near the N and C termini are the domains critical for binding A. This analysis was performed with a comprehensive set of mutations created in the QL (activated) G12 protein. The crystal structure of the G12 N terminus is not known, and the unique N-terminal structure of G12 is believed to inhibit G12 protein expression and crystallization (Kreutz et al., 2006). To crystallize G12, it was necessary to generate a chimeric G12 missing the first 46 amino acids and containing the first 28 amino acids of Gi1 (Kreutz et al., 2006). The N-terminal G12 sequence missing from the crystal structure contains 2 of 8 regions in which NAAIRS mutants (a and f; Fig. 4) resulted in nearly complete loss of binding to A. The N terminus of G subunits contributes to protein binding and contains lipid modifications important for membrane attachment (Busconi and Denker, 1997; Busconi et al., 1997); G12/13 are palmitoylated within this region (Jones and Gutkind, 1998). The N terminus of G subunits also binds to G (Denker et al., 1992). However, QL12 has a low affinity for G, and the addition of G did not affect stimulation of PP2A phosphatase activity. This indicates that G and the N-terminal sites used for binding G are not critical for the interaction of G12 with A. NAAIRS mutant b (Fig. 4) mutates the palmitoylation site (cysteine 11), and the normal pull-down of mutant b suggests that palmitoylation is not required for the interaction. Although the structure of the G12 N terminus is unknown, it is predicted to lie in close proximity to the C terminus (Fig. 7). The C terminus of G12 contained the largest region important for binding to A. Most of the region necessary for A binding comprises the G12 4 helix, 6 sheet, and 5 helix that ends at the C terminus. This region forms an exposed surface (Fig. 7, C-E) available for potential interaction with repeats 7 to 10 of A. NAAIRS mutants t and u are located apart from the N- and C-terminal binding regions (Fig. 7, C-E) and are not likely to directly contribute to interactions with A but may induce secondary conformational changes in the 4-6-5 region. These in vitro observations do not exclude the possibility that G12 activation is necessary for PP2A activation in cells. The high protein concentrations in vitro may permit stimulation of PP2A by virtue of the binding that occurs in both conformations. In the G12-expressing MDCK cells (Fig. 6), there was no change in tyrosine 307 phosphorylation with wild-type G12, but with expression of QL12, there was loss of phosphorylation at tyrosine 307 on the C subunit, a finding consistent with increased PP2A catalytic activity. Nevertheless, it remains to be established whether both active and inactive G12 stimulate PP2A in cells. The lack of an effect for wild-type G12 on PP2A activity could result from the protein being denatured, although in other studies, G12 reversibly interacts with the basolateral membrane (Meyer et al., 2002) and modulates paracellular flux (Meyer et al., 2003).
Activated G12 and G13 have been shown to bind and stimulate the PP5 family of serine/threonine phosphatases (Yamaguchi et al., 2002, 2003) through an interaction with the tetratricopeptide repeat domain. This observation, in addition to the studies presented here, suggest that G12/13 may be important regulators of cellular phosphatases. Many studies have linked G protein signaling to protein kinase pathways, and the interactions of G12 and G13 with protein phosphatases suggest that G protein signaling through G12/13 may counter-regulate protein phosphorylation occurring in response to activation of other G protein signaling pathways. Although G12/13 stimulation of protein phosphatases may be a broad regulatory mechanism, the modes of activation of PP5 versus PP2A by G12 seem to be distinct. PP5 is a single subunit phosphatase stimulated by both G12 and G13. In contrast, we did not detect G13 stimulation of PP2A (Zhu et al., 2004), although preliminary results indicate that G13 can bind to the GST-A fusion protein (data not shown). In addition, there is no sequence homology between the tetratricopeptide repeat domain of PP5 and the HEAT repeats 7 to 10 of A.
The numerous functions of PP2A require sophisticated mechanisms to allow for specific regulation. The diverse family of B subunits can modulate PP2A catalytic activity and direct the core enzyme to specific substrates. Our finding that G12 regulates PP2A provides a novel mechanism for regulation. Future studies are needed to determine how this occurs in cells, and it remains to be determined whether G12 functions as a novel "B" subunit or instead acts in concert with B subunits to regulate PP2A activity. However, these two mechanisms are not mutually exclusive, and the mechanism(s) may depend on the specific B subunit in the PP2A trimer. The question of whether there are differences in PP2A stimulation in cells by active and inactive G12 requires additional study. The findings in this report provide the basis for future studies to dissect aberrant signaling seen in some pathological conditions. For example, the G12-PP2A interaction can be disrupted by deletion of a single HEAT repeat that would not be expected to affect other subunit interactions.
Acknowledgements
We thank C. Todd DeMarco for extensive technical assistance and David Virshup, Gernot Walter, Pat Casey, Daniel Kaplan, and Patrick Kelly for helpful discussions.
ABBREVIATIONS: RGS, regulator of G protein signaling; PP2A, protein phosphatase-2A; MDCK, Madin Darby canine kidney; dox, doxycycline; HEAT repeat, 39 amino acid motif named for huntingtin, elongation, A subunit, target of rapamycin; HEK, human embryonic kidney; GST, glutathione transferase; PAGE, polyacrylamide gel electrophoresis; TBST, Tris-buffered saline/Tween 20; GTPS, guanosine 5-3-O-(thio)triphosphate; WT, wild type.
【参考文献】
Berestetskaya YV, Faure MP, Ichijo H, and Voyno-Yasenetskaya TA (1998) Regulation of apoptosis by -subunits of G12 and G13 proteins via apoptosis signal-regulating kinase-1. J Biol Chem 273: 27816-27823.[Abstract/Free Full Text]
Buhl AM, Johnson NL, Dhanasekaran N, and Johnson GL (1995) G12 and G13 stimulate Rho-dependent stress fiber formation and focal adhesion assembly. J Biol Chem 270: 24631-24634.[Abstract/Free Full Text]
Busconi L, Boutin PM, and Denker BM (1997) N-terminal binding domain of G subunits: involvement of amino acids 11-14 of Go in membrane attachment. Biochem J 323: 239-244.
Busconi L and Denker BM (1997) Analysis of the N-terminal binding domain of Go . Biochem J 328: 23-31.
Chen J, Martin BL, and Brautigan DL (1992) Regulation of protein serine-threonine phosphatase type-2A by tyrosine phosphorylation. Science (Wash DC) 257: 1261-1264.[Abstract/Free Full Text]
Cho U and Xu W (2007) Crystal structure of a protein phosphatase 2a heterotrimer holoenzyme. Nature (Lond) 445: 53-57.
Denker BM, Boutin PM, and Neer EJ (1995) Interactions between the amino- and carboxyl-terminal regions of G subunits: analysis of mutated Go/Gi2 chimeras. Biochemistry 34: 5544-5553.
Denker BM, Neer EJ, and Schmidt CJ (1992) Mutagenesis of the amino terminus of the subunit of the G protein Go. In vitro characterization of o interactions. J Biol Chem 267: 6272-6277.[Abstract/Free Full Text]
Dodane V and Kachar B (1996) Identification of isoforms of G proteins and PKC that colocalize with tight junctions. J Membr Biol 149: 199-209.
El-Shewy HM, Johnson KR, Lee MH, Jaffa AA, Obeid LM, and Luttrell LM (2006) Insulin-like growth factors mediate heterotrimeric G protein-dependent ERK1/2 activation by transactivating sphingosine 1-phosphate receptors. J Biol Chem 281: 31399-31407.[Abstract/Free Full Text]
Goulimari P, Kitzing TM, Knieling H, Brandt DT, Offermanns S, and Grosse R (2005) G12/13 is essential for directed cell migration and localized Rho-Dia1 function. J Biol Chem 280: 42242-42251.[Abstract/Free Full Text]
Groves MR, Hanlon N, Turowski P, Hemmings BA, and Barford D (1999) The structure of the protein phosphatase 2A PR65/A subunit reveals the conformation of its 15 tandemly repeated HEAT motifs. Cell 96: 99-110.
Hollinger S and Hepler JR (2002) Cellular regulation of RGS proteins: modulators and integrators of G protein signaling. Pharmacol Rev 54: 527-559.[Abstract/Free Full Text]
Janssens V and Goris J (2001) Protein phosphatase 2A; a highly regulated family of serine/threonine phosphatases implicated in cell growth and signaling. Biochem J 353: 417-439.
Jiang H, Wu D, and Simon MI (1993) The transforming activity of activated G12. FEBS Lett 330: 319-322.
Jiang Y, Ma W, Wan Y, Kozasa T, Hattori S, and Huang XY (1998) The G protein G12 stimulates Bruton's tyrosine kinase and a rasGAP through a conserved PH/BM domain. Nature (Lond) 395: 808-813.
Jones TL and Gutkind JS (1998) G12 requires acylation for its transforming activity. Biochemistry 37: 3196-3202.
Katoh H, Aoki J, Yamaguchi Y, Kitano Y, Ichikawa A, and Negishi M (1998) Constitutively active G12, G13 and Gq induce Rho-dependent neurite retraction through different signaling pathways. J Biol Chem 273: 28700-28707.[Abstract/Free Full Text]
Kozasa T, Jiang X, Hart MJ, Sternweis PM, Singer WD, Gilman AG, Bollag G, and Sternweis PC (1998) p115 RhoGEF, a GTPase activating protein for G12 and G13. Science (Wash DC) 280: 2109-2111.[Abstract/Free Full Text]
Kreutz B, Yau DM, Nance MR, Tanabe S, Tesmer JJ, and Kozasa T (2006) A new approach to producing functional G subunits yields the activated and deactivated structures of G (12/13) proteins. Biochemistry 45: 167-174.
Lanier SM (2004) AGS proteins, GPR motifs and the signals processed by heterotrimeric G proteins. Biol Cell 96: 369-372.
Lin F, Sepich DS, Chen S, Topczewski J, Yin C, Solnica-Krezel L, and Hamm H (2005) Essential roles of G12/13 signaling in distinct cell behaviors driving zebrafish convergence and extension gastrulation movements. J Cell Biol 169: 777-787.[Abstract/Free Full Text]
Luo Y and Denker BM (1999) Interaction of heterotrimeric G protein Go with Purkinje cell protein-2: evidence for a novel nucleotide exchange factor. J Biol Chem 274: 10685-10688.[Abstract/Free Full Text]
Luttrell DK and Luttrell LM (2004) Not so strange bedfellows: G-protein-coupled receptors and Src family kinases. Oncogene 23: 7969-7978.
Meigs TE, Fedor-Chaiken M, Kaplan DD, Brackenbury R, and Casey PJ (2002) G12 and G13 negatively regulate the adhesive functions of cadherin. J Biol Chem 25: 24594-24600.
Meigs TE, Fields TA, McKee DD, and Casey PJ (2001) Interaction of G12 and G13 with the cytoplasmic domain of cadherin provides a mechanism for beta -catenin release. Proc Natl Acad Sci USA 2: 519-524.
Meigs TE, Juneja J, DeMarco CT, Stemmle LN, Kaplan DD, and Casey PJ (2005) Selective uncoupling of G12 from Rho-mediated signaling. J Biol Chem 280: 18049-18055.[Abstract/Free Full Text]
Meyer TN, Hunt J, Schwesinger C, and Denker BM (2003) G12 regulates epithelial cell junctions through Src tyrosine kinases. Am J Physiol 285: C1281-C1293.
Meyer TN, Schwesinger C, and Denker BM (2002) Zonula occludens-1 is a scaffolding protein for signaling molecules. G12 directly binds to the src homology 3 domain and regulates paracellular permeability in epithelial cells. J Biol Chem 277: 24855-24858.[Abstract/Free Full Text]
Nunbhakdi-Craig V, Machleidt T, Ogris E, Bellotto D, White CL 3rd, and Sontag E (2002) Protein phosphatase 2A associates with and regulates atypical PKC and the epithelial tight junction complex. J Cell Biol 158: 967-978.[Abstract/Free Full Text]
Ruediger R, Fields K, and Walter G (1999) Binding specificity of protein phosphatase 2A core enzyme for regulatory B subunits and T antigens. J Virol 73: 839-842.[Abstract/Free Full Text]
Ruediger R, Hentz M, Fait J, Mumby M, and Walter G (1994) Molecular model of the A subunit of protein phosphatase 2A: interaction with other subunits and tumor antigens. J Virol 68: 123-129.[Abstract/Free Full Text]
Ruediger R, Roeckel D, Fait J, Bergqvist A, Magnusson G, and Walter G (1992) Identification of binding sites on the regulatory A subunit of protein phosphatase 2A for the catalytic C subunit and for tumor antigens of simian virus 40 and polyomavirus. Mol Cell Biol 12: 4872-4882.[Abstract/Free Full Text]
Stemmle LN, Fields TA, and Casey PJ (2006) The regulator of G protein signaling domain of axin selectively interacts with G12 but not G13. Mol Pharmacol 70: 1461-1468.[Abstract/Free Full Text]
Vaiskunaite R, Kozasa T, and Voyno-Yasenetskaya TA (2001) Interaction between the G subunit of heterotrimeric G12 protein and Hsp90 is required for G12 signaling. J Biol Chem 276: 46088-46093.[Abstract/Free Full Text]
Waheed AA and Jones TL (2002) Hsp90 interactions and acylation target the G protein G12 but not G13 to lipid rafts. J Biol Chem 277: 32409-32412.[Abstract/Free Full Text]
Wilson IA, Haft DH, Getzoff ED, Tainer JA, Lerner RA, and Brenner S (1985) Identical short peptide sequences in unrelated proteins can have different conformations: a testing ground for theories of immune recognition. Proc Natl Acad Sci USA 82: 5255-5259.[Abstract/Free Full Text]
Yamaguchi Y, Katoh H, Mori K, and Negishi M (2002) G12 and G13 interact with Ser/Thr protein phosphatase type 5 and stimulate its phosphatase activity. Curr Biol 12: 1353-1358.
Yamaguchi Y, Katoh H, and Negishi M (2003) N-terminal short sequences of subunits of the G12 family determine selective coupling to receptors. J Biol Chem 278: 14936-14939.[Abstract/Free Full Text]
Zhu D, Kosik KS, Meigs TE, Yanamadala V, and Denker BM (2004) G12 directly interacts with PP2A: evidence for G12-stimulated PP2A phosphatase activity and dephosphorylation of microtubule-associated protein, . J Biol Chem 279: 54983-54986.[Abstract/Free Full Text]
作者单位：Renal Division, Brigham and Women's Hospital, and Harvard Medical School, Boston, Massachusetts (B.M.D., D.Z.); Department of Biology, University of North Carolina at Asheville, Asheville, North Carolina (R.I.T., T.E.M.); and Department of Pathology, University of California at San Diego, La Jolla,