Interaction of p130 with, and Consequent Inhibition of, the Catalytic Subunit of Protein Phosphatase 1α*

The protein p130 was originally isolated from rat brain as an inositol 1,4,5-trisphosphate-binding protein with a domain organization similar to that of phospholipase C-δ1 but which lacks phospholipase C activity. Yeast two-hybrid screening of a human brain cDNA library for clones that encode proteins that interact with p130 has now led to the identification of the catalytic subunit of protein phosphatase 1α (PP1cα) as a p130-binding protein. The association between p130 and PP1cα was also confirmedin vitro by an overlay assay, a “pull-down” assay, and surface plasmon resonance analysis. The interaction of p130 with PP1cα resulted in inhibition of the catalytic activity of the latter in a p130 concentration-dependent manner. Immunoprecipitation and immunoblot analysis of COS-1 cells that stably express p130 and of mouse brain extract with antibodies to p130 and to PP1cα also detected the presence of a complex of p130 and PP1cα. The activity of glycogen phosphorylase, which is negatively regulated by dephosphorylation by PP1cα, was higher in COS-1 cells that stably express p130 than in control COS-1 cells. These results suggest that, in addition to its role in inositol 1,4,5-trisphosphate and Ca2+ signaling, p130 might also contribute to regulation of protein dephosphorylation through its interaction with PP1cα.

The Ins(1,4,5)P 3 -binding protein with a molecular mass of 130 kDa, termed p130, was a previously unidentified molecule (2,3). The predicted amino acid sequence of rat p130 shares 38.2% identity with that of rat PLC-␦1; the five identified domains of PLC-␦1 (PH, EF-hand, putative catalytic (X and Y), and C2 domains) are all present in p130. The domain organization of p130 suggests that the protein is likely to possess a fold similar to that of PLC-␦1, a notion that is supported by the results of limited proteolysis with trypsin (8). However, p130 exhibits some distinct characteristics. It is larger than the PLC-␦ isozymes, and it possesses unique regions both at the NH 2 terminus, preceding the PH domain, and at the COOH terminus. Moreover, the residues within the catalytic domain of PLC-␦ that are critical for enzyme activity (His 356 and Glu 390 ) are not conserved in p130 (9). The PH domain of p130, like that of PLC-␦1, is important for the binding of Ins(1,4,5)P 3 (10). Other molecules that show sequence similarity to p130, including human PLC-L (11) and the K10F12.3 gene product of Caenorhabditis elegans (12), have also been described. Otsuki et al. (13) recently isolated a cDNA from mouse brain that encodes a protein with 66% sequence identity to PLC-L; they therefore termed this protein PLC-L 2 and renamed the original PLC-L as PLC-L 1 . Furthermore, the gene for human type2 p130 (PLC-L 2 ) has also been cloned (14). All of these proteins exhibit characteristic NH 2 -and COOH-terminal extensions and replacement of critical catalytic residues. The identification of a p130-related molecule in such a simple organism as C. elegans suggests that this family of proteins diverged early from other PLC isozymes. We propose that this distinct family of PLC-related proteins be designated the PLC-related catalytically inactive protein (PRIP) family (comprising PRIP-1 and -2 subfamilies).
To investigate the physiological functions of PRIP family proteins, we previously examined the possible role of the binding of inositol compounds to the PH domain of p130 (10,(15)(16)(17). Our results suggested that p130, which is localized predominantly in the cytoplasm, contributes to Ins(1,4,5)P 3 -mediated Ca 2ϩ signaling. The high affinity binding of Ins(1,4,5)P 3 to the PH domain of p130 might also serve to sequester Ins(1,4,5)P 3 and therefore prevent its interaction with Ins(1,4,5)P 3 receptors and metabolizing enzymes (18).
We have now applied the yeast two-hybrid system to identify proteins that interact with p130. With the unique NH 2 -terminal region of p130 as the bait for screening a human brain cDNA library, we isolated two positive clones, one of which was shown to encode the catalytic subunit of protein phosphatase 1␣ (PP1c␣). To characterize the interaction between p130 and PP1c␣, we studied the association of these two proteins both in vitro and in living cells, we delineated further the region of p130 that is responsible for binding to PP1c␣, and we examined the effect of such binding on the enzymatic activity of PP1c␣.

EXPERIMENTAL PROCEDURES
Materials-Cloning vectors pGBT9 and pACT2, a human brain cDNA library, and yeast strains HF7c and SFY526 were obtained from CLONTECH (Palo Alto, CA). All restriction endonucleases and DNAmodifying enzymes were from Toyobo (Tokyo, Japan). Dropout yeast selection medium and dropout base medium were from BIO101 (Vista, CA). YPD medium for yeast and bacterial medium were obtained from Becton Dickinson (Sparks, MD). Polyvinylidene difluoride (PVDF) membranes were from Millipore (Bedford, MA). [␥-32 P]ATP (222 terabecquere/mmol) was obtained from DuPont-New England Nuclear. A large scale plasmid preparation kit, QIAfilter Plasmid Giga kit, and nitrilotriacetic acid-agarose beads for purification of His 6 -tagged proteins were from Qiagen (Chatsworth, CA). Protein G-Sepharose, glutathione-Sepharose 4B beads, and pGEX vectors were from Amersham Pharmacia Biotech. Ins(1,4,5)P 3 was synthesized as described (19). The catalytic subunit of cAMP-dependent protein kinase (PKA) was obtained from Promega (Madison, WI), and wild-type rabbit PP1c␣ was from Calbiochem-Novabiochem (La Jolla, CA). A soluble form of PtdIns(4,5)P 2 , diC 8 -PtdIns(4,5)P 2 , was obtained from Echelon Research Laboratories (Salt Lake City, UT). G M peptide (GRRVSFADNFGFN) and its random sequence (GNFRGFRSADFVN) were synthesized using the Fmoc (N-(9-fluorenyl)methoxycarbonyl) cleavage method on an Advanced ChemTech 348MPS peptide synthesizer, and the purity was checked by applying the sample to a Bondashere 5-C18 column mounted on a high-performance liquid chromatography column (more than 90%). Other reagents used were of the highest grade available.
For construction of pGBT9-p130D (amino acids 848 to 1096), a SpeI site (nt 3011) was introduced into pcMT3 by site-directed mutagenesis, and the resulting plasmid was digested with SpeI (nt 3011) and XhoI (nt 5233). The released 2.2-kilobase pair fragment was then ligated into the SpeI-SalI sites of pGBT9. For construction of pACT2-PP1c␣(⌬29 -163) a positive clone obtained from the yeast two-hybrid screening, pACT2-PP1c␣, was digested with PvuII (nt 116 and 815) and self-ligated. For expression of recombinant PP1c␣ in Escherichia coli, the BamHI fragment of pACT2-PP1c␣ was ligated into the BamHI site of pGEX-3X; the resulting construct encodes a fusion protein of glutathione S-transferase (GST) and PP1c␣.
Yeast Two-hybrid Screening and ␤-Galactosidase Assay-Yeast twohybrid screening of a human brain cDNA library cloned in the pACT2 vector was performed in yeast strain HF7c with the bait plasmids pGBT9-p130PH or pGBT9-p130D. Transformants (total of 2 ϫ 10 6 ) were plated and selected with a combination of tryptophan, leucine, and histidine. The positive clones identified by two-hybrid screening were sequenced with an ABI 373A automated DNA sequencer. The domains required for the interaction between p130 and PP1c␣ were investigated by expression of various combinations of bait and target plasmids in yeast SFY526 cells and measurement of ␤-galactosidase activity.
GST Fusion Protein Precipitation and Protein Overlay Analyses-The recombinant GST-PP1c␣ fusion protein was purified from E. coli by affinity chromatography, and recombinant full-length p130 (amino acids 24 to 1096) and the PH domain of p130 (p130PH; amino acids 95 to 232) were prepared as described previously (8,10). For "pull-down" assays, GST-PP1c␣ was incubated for 1 h at 4°C with glutathione-Sepharose 4B beads in binding buffer (50 mM Tris-HCl (pH 7.5), 5 mM MgCl 2 , 100 mM NaCl, 10% glycerol, 0.5 mg/ml bovine serum albumin (BSA), 5 mM 2-mercaptoethanol). The beads were washed with 50 volumes of binding buffer and then incubated (6 g of GST-PP1c␣) for 1 h at 4°C, with gentle rotation, in a total volume of 150 l with recombinant full-length p130 or p130PH. After washing of the beads five times with 500 l of binding buffer, bound proteins were eluted with 50 l of a solution containing 50 mM Tris-HCl (pH 8.0) and 10 mM reduced glutathione and were then subjected to SDS polyacrylamide gel electrophoresis (PAGE) and immunoblot analysis with antibodies to p130 (2F9) or to p130PH (3,8).
For overlay analysis, samples were fractionated by SDS-PAGE, and the separated proteins were transferred electrophoretically to a PVDF membrane. After the blocking of nonspecific sites with 5% dried skim milk, the membrane was incubated for 1 h at room temperature with a protein probe (10 g/ml). The membrane was then washed and incubated with antibodies to the probe protein followed by alkaline phosphatase-conjugated secondary antibodies, after which immune complexes were detected by enzymatic reaction.
Immunoprecipitation-COS-1 cells, COS-1 p130 cells (COS-1 cells stably expressing p130) (18), and mouse brain extract were subjected to immunoprecipitation with a specific monoclonal antibody to p130 (2F9) or polyclonal antibodies to PP1c␣ (Santa Cruz Biotechnology, Santa Cruz, CA). Cells (5 ϫ 10 6 ) or mouse brain (wet weight, 0.2 g) were homogenized in 0.5 ml of a solution containing 20 mM HEPES-NaOH (pH 7.4), 130 mM NaCl, 5 mM EDTA, and a mixture of protease inhibitors. The homogenate was centrifuged (14,000 ϫ g, 20 min, 4°C), and the resulting supernatant was incubated, with gentle rotation, for 1 h at 4°C with 30 g of antibodies to p130 or to PP1c␣ that had been premixed with 10 l of a 50% slurry of protein G-Sepharose in phosphate-buffered saline containing 0.1% BSA. The beads were then washed twice with a homogenizing solution (described above) containing 0.2% Triton X-100, boiled in SDS sample buffer, and subjected to SDS-PAGE and immunoblot analysis with antibodies to PP1c␣ or to p130. Assay of Glycogen Phosphorylase Activity-COS-1 or COS-1 p130 cells (2 ϫ 10 6 ) were lysed by three freeze-thaw cycles in a solution containing 50 mM NaCl, 10 mM MES-NaOH (pH 6.0), 1 mM EDTA, and 10 mM 2-mercaptoethanol. The lysate was subjected to centrifugation at 15,000 ϫ g for 30 min, and the resulting supernatant was assayed for glycogen phosphorylase activity as described (20).

Analysis of Protein-Protein Interaction in Real Time-Protein-pro-
Assay of Phosphatase Activity-Phosphatase activity was determined in a reaction mixture (40 l) containing 139.2 mM KCl, 20 mM 4-morpholinepropanesulfonic acid-KOH (pH 7.0), 0.1 mM MnCl 2 , 0.5 mM dithiothreitol, BSA (0.5 mg/ml), 2 M phosphorylated myosin light chain (from bovine stomach), 3.4 nM recombinant rabbit skeletal muscle PP1c␣, and various concentrations of recombinant full-length p130 or p130PH, in the absence or presence of 10 M Ins(1,4,5)P 3 . The mixture minus the substrate was incubated for 10 min at 25°C, and the reaction was started by the addition of phosphorylated myosin light chain and stopped after 20 min by the addition of 0.2 ml of ice-cold 10% trichloroacetic acid. The unphosphorylated and phosphorylated myosin light chains were separated by two-dimensional electrophoresis, and the density of each spot was determined as described (21).

Two-hydrid Screening for Proteins That Interact with p130 -
Screening of a human brain cDNA library with a bait plasmid (2) encoding the unique NH 2 -terminal region of rat p130 (amino acids 24 to 298), including the PH domain and a portion of the EF-hand motif (Fig. 1A), yielded 51 positive clones of a total of 2 million clones examined; no positive clones were obtained with a bait plasmid encoding the COOH-terminal region (amino acids 848 to 1096) of p130. 10 of the 51 clones identified proved to be false positives, and the remaining 41 clones were divided into two groups on the basis of analysis of their inserts by polymerase chain reaction amplification and restriction enzyme digestion. Sequencing revealed that one of these 41 clones encoded full-length PP1c␣, a 37,510-Da protein composed of 330 amino acids. To delineate more precisely the region of p130 required for binding to PP1c␣, we used plasmids encoding smaller portions of p130 as baits in two-hybrid analysis with the plasmid encoding full-length PP1c␣ (Fig. 1A). Positive signals were obtained with a bait plasmid (3) encoding amino acids 24 to 222, as well as with that (1) encoding full-length p130. Neither a bait plasmid (4) encoding amino acids 24 to 82 nor one (5) encoding residues 222 to 298 yielded a positive signal (Fig. 1B). A plasmid encoding a PP1c␣ mutant lacking amino acids 30 to 162 did not yield a positive signal with any of the p130 bait plasmids examined. These results thus suggested that the region of p130 composed of residues 83 to 222 interacts with that of PP1c␣ comprising residues 30 to 162.
Association of p130 with PP1c␣ in Vitro-We next examined the interaction of p130 and PP1c␣ in vitro by several methods. The association was first analyzed with an overlay assay ( Fig.  2A). Extracts of nontransformed E. coli and of bacteria expressing a GST-PP1c␣ fusion protein, as well as recombinant GST-PP1c␣ purified from such a latter extract, were fractionated by SDS-PAGE, and the separated proteins were transferred to a PVDF membrane and probed with antibodies to PP1c␣ to confirm that the prominent band that migrated at a position corresponding to a molecular size of 37 kDa was indeed PP1c␣. Duplicate membranes were incubated in the presence of a recombinant p130 fragment containing the PH domain (p130PH; amino acids 95 to 232), recombinant full-length p130 (residues 24 to 1096), or BSA (negative control). After washing, the membranes were exposed to the corresponding antibodies to p130PH or to p130. Both the recombinant GST-PP1c␣ present in the bacterial extract and the purified protein interacted with both full-length p130 and p130PH. Together with the results from the yeast two-hybrid analysis, these data indicate that residues 95 to 222 of p130 (which include the entire PH domain and the 20 residues preceding it) mediate the interaction of this protein with PP1c␣.
The GST-PP1c␣ fusion protein was also subjected to a pulldown assay with recombinant p130 or p130PH (Fig. 2B). Incu-bation of a GST-PP1c␣ resin with p130PH and subsequent immunoblot analysis of bead-bound proteins with antibodies to p130PH revealed that p130PH was precipitated by GST-PP1c␣ and that this interaction was sensitive to the presence of low concentrations of full-length p130 but not to PLC-␦1 (Fig. 2B,  panel a). In contrast, although full-length p130 also bound to GST-PP1c␣ (but not to GST alone), this interaction was not sensitive to the presence of p130PH (Fig. 2B, panel b). These results indicate that, although the PH domain of p130 is primarily responsible for the binding of this protein to PP1c␣, other regions of p130 also contribute to the interaction between these two proteins.
Analysis of the interaction of various regulatory subunits with PP1c␣ has led to the identification of a consensus sequence for binding, (K/R)(K/R)(V/I)XF (22). The sequence VSF (residues 95 to 97) is present in the region of p130 shown to bind to PP1c␣. To determine whether this sequence participates in the interaction of p130 with PP1c␣, we examined the effect of a peptide (G M peptide, GRRVSFADNFGFN) that has been shown to inhibit the association between PP1c␣ and several regulatory subunits (22). This peptide inhibited the interaction of PP1c␣ with either full-length p130 or p130PH, whereas a random peptide with the same amino acid composition had no such effect (Fig. 2B, panel c). To confirm the role of the VSF sequence of p130 in the interaction of this protein with PP1c␣, we expressed in and purified from E. coli p130 fragments comprising amino acids 82 to 232. Whereas the wildtype fragment bound to PP1c␣, fragments containing either V95L or F97A mutations bound to the lesser extent (Fig. 2B,  panel d).
Given that p130 contains four consensus motifs for phosphorylation by PKA ( 74 Arg Arg Thr Ser 77 , 90 Arg Lys Lys Thr 93 , 104 Lys Ile Ser 107 , and 567 Arg Arg Val Ser 570 [underlining refers to phosphorylatable residues] one of which ( 104 Lys Lys Ile Ser 107 ) is present in p130PH, it was possible that p130 associates with PP1c␣ because it is a substrate for phosphatase activity of this enzyme. Indeed, p130 was phosphorylated by PKA (Fig. 2C, lane 2), although the precise site (or sites) phos-

FIG. 1. Yeast two-hybrid analysis of the interaction between p130 and PP1c␣.
A, schematic representation of plasmids encoding various regions of p130 (left) and PP1c␣ (right) that were used for analysis of the domains required for binding. The initial screening of the human brain cDNA library for p130-binding proteins was performed with p130 bait plasmid 2 (pGBT9-p130PH(24 -298)). Bait plasmids 3 (XhoI-KpnI fragment of p130 cDNA), 4 (XhoI-BamHI), and 5 (KpnI-SmaI) encode amino acids 24 to 222, 24 to 82, and 222 to 298 of p130, respectively. These constructs were introduced into yeast strain SFY526, together with pACT2-PP1c␣ or pACT2-PP1c␣(⌬30 -162), the latter of which was prepared from the former by digestion with PvuI and selfligation. B, ␤-Galactosidase assay of protein-protein interaction. Two-hybrid analysis was performed with SFY256 cells transformed with the indicated p130 (1 to 6) and PP1c␣ plasmids. The activity of ␤-galactosidase was determined with a filter assay. phorylated remains to be determined. However, this explanation for the interaction between p130 and PP1c␣ is unlikely, because phosphorylated p130 did not associate with PP1c␣, whereas p130 treated with ATP alone (without PKA) bound to PP1c␣ (Fig. 2C, lane 1).
The association between p130 and PP1c␣ was further confirmed by surface plasmon resonance analysis. Full-length p130 was introduced into the analysis chamber after immobilization of GST-PP1c␣ onto the sensor chip. Positive signals indicative of protein-protein interaction were generated in a p130 concentration-dependent manner and were abolished by washing away of the applied p130 (Fig. 2D). The dissociation constant was calculated to be 1.2 Ϯ 0.1 nM (mean Ϯ S.E. of values from five independent determinations). Replacement of the full-length p130 molecule with p130PH yielded a dissociation constant in the micromolar range, consistent with the results obtained with pull-down assays (Fig. 2B, panels a and b).
We next investigated whether the activity of PP1c␣ is affected by the association with p130. The dephosphorylation of phosphorylated smooth muscle myosin light chain (21) by recombinant rabbit skeletal muscle PP1c␣ was inhibited by fulllength p130 in a concentration-dependent manner (Fig. 3). Recombinant p130PH also inhibited the activity of PP1c␣, al-though higher concentrations of p130PH than of full-length p130 were required for this effect.
The effects of Ins(1,4,5)P 3 and water-soluble (short-chain) PtdIns(4,5)P 2 on the association of p130 with PP1c␣, as well as on the inhibition of PP1c␣ activity by p130, were also examined, given that the site of p130 responsible for the association with PP1c␣ was shown to be located immediately upstream of the PH domain and that PH domains mediate binding to Ins(1,4,5)P 3 or PtdIns(4,5)P 2 . The presence of Ins(1,4,5)P 3 or short-chain PtdIns(4,5)P 2 at a concentration of 10 M in the reaction mixture for the pull-down assay had no effect on the interaction of p130 with PP1c␣ (data not shown), and 10 M Ins(1,4,5)P 3 had no effect on p130-induced inhibition of PP1c␣ activity (Fig. 3).
Association between p130 and PP1c␣ in Intact Cells-To determine whether p130 and PP1c␣ interact in living cells, we first examined COS-1 cells that stably express recombinant p130 (COS-1 p130 cells) (18). Immunoblot analysis of extracts of both control COS-1 cells (which lack endogenous p130) and COS-1 p130 cells with antibodies to PP1c␣ revealed that both cell lines express similar amounts of PP1c␣ (Fig. 4A, a). Cell extracts were then subjected to immunoprecipitation with antibodies to either p130 (Fig. 4A, b) or PP1c␣ (Fig. 4A, c), and the with ATP alone (lane 1) was incubated with recombinant GST-PP1c␣ immobilized on glutathione-Sepharose 4B beads, after which bead-bound protein was eluted with reduced glutathione and subjected to SDS-PAGE and immunoblot analysis with antibodies to p130 (b) or to PP1c␣ (c). D, surface plasmon resonance analysis. GST-PP1c␣ was immobilized on a sensor chip and exposed to various concentrations of full-length p130 (0.23, 2.3, 23, 230, and 2300 nM) for 360 s before the application of buffer alone; data are expressed in relative units (RU). resulting precipitates were subjected to immunoblot analysis with the same two types of antibodies. Stable association of p130 with PP1c␣ was apparent in COS-1 p130 cells but not in control COS-1 cells (Fig. 4A). We also examined whether these two proteins interact in mouse brain, which contains both molecules (Fig. 4B). PP1c␣ was detected in p130 immunoprecipitates (Fig. 4B, b), and p130 was detected in PP1c␣ immunoprecipitates (Fig. 4B, c) prepared from mouse brain.
PP1c␣ is thought to catalyze protein dephosphorylation reactions that underlie many aspects of cell function (23,24). Glycogen phosphorylase, which catalyzes the conversion of glycogen to glucose 1-phosphate, is a substrate for PP1c␣ in a wide variety of cell types (25); its dephosphorylation by this phosphatase results in inhibition of phosphorylase activity. Measurement of glycogen phosphorylase activity in extracts of COS-1 and COS-1 p130 cells yielded values of 68 Ϯ 6 and 130 Ϯ 9 nmol per milligram of protein per 5 min (means Ϯ S.E. of six independent determinations), respectively. These results thus indicate that glycogen phosphorylase is phosphorylated to a greater extent in COS-1 p130 cells than in COS-1 cells. DISCUSSION With the use of its specific NH 2 -terminal region as a bait, we applied the yeast two-hybrid screen to identify human brain proteins that bind to p130, designated henceforth as PRIP-1. This approach identified PP1c␣ as one such protein. PP1 is a widely expressed serine-threonine protein phosphatase that exists in several isoforms, including ␣, ␣2, ␥1, ␥2, and ␦ (23,24). Various regulatory subunits have been shown to associate with PP1c␣ and thereby to influence its catalytic activity (26). For example, G M and G L subunits function to target PP1c␣ to glycogen granules; phosphorylation of these subunits by PKA induces their dissociation from PP1c␣, whereas that triggered by insulin promotes their association with and activation of PP1c␣, resulting in inhibition of glycogen breakdown. The association of I-1 (inhibitor 1) or DARPP-32 (dopamine-and cAMP-regulated phosphoprotein of 32 kDa) with PP1c␣ appears not to affect phosphatase activity, whereas phosphorylation of I-1 or DARPP-32 by PKA induces marked inhibition of such activity. Our results now suggest that PRIP-1 also functions as a regulatory subunit of PP1c␣ that inhibits phosphatase activity. The binding of Ins(1,4,5)P 3 or PtdIns(4,5)P 2 to PRIP-1 had no effect on its association with or inhibition of PP1c␣. Our previous observations suggested that Ins(1,4,5)P 3 may be a physiological ligand for PRIP-1 and that this protein is localized predominantly to the cytosol (18). PRIP-1 may therefore serve not only to inhibit the activity of PP1c␣ but also to target this enzyme to the cytosol.
Amino acid residues 95 to 97 of PRIP-1, located upstream of the PH domain, appear to contribute to the binding site for PP1c␣. The fragment of PRIP-1 comprising residues 24 to 222 interacted with PP1c␣ in the yeast two-hybrid assay, and p130PH (PRIP-1PH) (residues 95 to 232) as well as the fulllength molecule, associated with PP1c␣ in vitro, as demonstrated with a variety of binding assays. A G M peptide that disrupts the interaction of PP1c␣ with several regulatory subunits and that contains the VSF (residues 95 to 97) sequence of PRIP-1 also inhibited the association of PRIP-1 with PP1c␣. Furthermore, mutation of residues 95 or 97 of PRIP-1 prevented the association of this protein with PP1c␣. Other regions of the PRIP-1 molecule may also interact with PP1c␣, as suggested by the observations that the full-length molecule bound to PP1c␣ was not displaced by an excess amount of PRIP-1PH and that the dissociation constant obtained by surface plasmon resonance analysis for the interaction with PP1c␣ was smaller for the full-length molecule than for PRIP-1PH. However, the observation that the G M peptide was similarly Association between p130 and PP1c␣ in intact cells and mouse brain. A, panel a, extracts prepared from COS-1 or COS-1 p130 cells were subjected to immunoblot analysis (IB) with antibodies to p130 or to PP1c␣, as indicated. Panels b and c, extracts of COS-1 (lanes 1) and COS-1 p130 (lanes 3) cells were subjected to immunoprecipitation (IP) with antibodies to p130 (b) or to PP1c␣ (c), and the resulting precipitates were subjected to immunoblot analysis with the same two types of antibodies. Lanes 2 correspond to COS-1 p130 cell extract subjected to immunoprecipitation with protein G-Sepharose in the absence of antibodies. Upper bands in lanes 1 and 3 in the blot analyzed with anti-PP1c␣ in panel b are immunoglobulin heavy chains. B, mouse brain extract was analyzed as described in A; lanes 2 correspond to brain extract subjected to immunoprecipitation with the indicated antibodies, whereas lanes 1 correspond to extract subjected to precipitation with protein G-Sepharose in the absence of antibodies. effective in inhibiting the association of PP1c␣ with full-length PRIP-1 and with PRIP-1PH suggests rather that other regions of PRIP-1 promote the interaction of the region containing residues 95 to 97 with PP1c␣.
Phosphorylation of PRIP-1 by PKA resulted in inhibition of the association between PRIP-1 and PP1c␣. Although the phosphorylated residues of PRIP-1 that underlie this effect remain to be identified, T93, which is located immediately upstream of the putative binding site for PP1c␣, is a likely candidate.
PP1c␣ contributes to the regulation of many aspects of cellular metabolism, including glycogen metabolism (through dephosphorylation of phosphorylase kinase, glycogen phosphorylase, and glycogen synthase) and lipid metabolism (through dephosphorylation of acetyl-CoA carboxylase, hormonedependent lipase, and hydroxymethylglutaryl-CoA reductase). Furthermore, it participates in the regulation of Ca 2ϩ transport (through dephosphorylation of phospholamban and Ca 2ϩ channel proteins), smooth muscle contraction (through dephosphorylation of myosin light chain), DNA replication (through dephosphorylation of histones H2B and H1), and protein synthesis (through dephosphorylation of initiation factor eIF-2, RNA-dependent protein kinase, heat shock protein, S6 protein, and S6 kinase) (24,25). It remains to be determined which of these cellular activities are physiologically regulated by PRIP-1 through its interaction with PP1c␣. Our data do suggest, however, that the association between PRIP-1 and PP1c␣ occurs in living cells, and we have shown that the activity of glycogen phosphorylase, which is regulated exclusively by phosphorylation, was increased in COS-1 cells by the expression of PRIP-1, probably as a result of the interaction of PRIP-1 with, and the consequent inhibition of, PP1c␣. Glycogen phosphorylase may therefore be a physiological target for regulation by the interaction of PRIP-1 with PP1c␣.
In summary, we have shown that (i) p130, which belongs to the PRIP family of proteins and is here renamed PRIP-1, associates with PP1c␣ through a G M peptide-like region located upstream of the PH domain; (ii) association with PRIP-1 results in inhibition of the catalytic activity of PP1c␣ as measured in vitro with phosphorylated myosin light chain as substrate; and (iii) glycogen phosphorylase activity was increased by expression of PRIP-1 in intact cells, likely as a result of inhibition of PP1c␣ and accumulation of the phosphorylated, active form of glycogen phosphorylase. In addition to its role in Ins(1,4,5)P 3 and Ca 2ϩ signaling (18), PRIP-1 might therefore also contribute to the regulation of protein dephosphorylation. Given that the binding of Ins(1,4,5)P 3 to the PH domain of PRIP-1 had no effect on the association of PRIP-1 with PP1c␣ or on its inhibition of PP1c␣ activity, PRIP-1 may contribute to both Ca 2ϩ signaling and regulation of protein dephosphorylation simultaneously and, in some instances, cooperatively.