Displacement affinity chromatography of protein phosphatase one (PP1) complexes
Protein phosphatase one (PP1) is a ubiquitously expressed, highly conserved protein phosphatase that dephosphorylates target protein serine and threonine residues. PP1 is localized to its site of action by interacting with targeting or regulatory proteins, a majority of which contains a primary docking site referred to as the RVXF/W motif.
We demonstrate that a peptide based on the RVXF/W motif can effectively displace PP1 bound proteins from PP1 retained on the phosphatase affinity matrix microcystin-Sepharose. Subsequent co-immunoprecipitation experiments confirmed that each identified binding protein was either a direct PP1 interactor or was in a complex that contains PP1. Our results have linked PP1 to numerous new nuclear functions and proteins, including Ki-67, Rif-1, topoisomerase IIα, several nuclear helicases, NUP153 and the TRRAP complex.
This modification of the microcystin-Sepharose technique offers an effective means of purifying novel PP1 regulatory subunits and associated proteins and provides a simple method to uncover a link between PP1 and additional cellular processes.
KeywordsSepharose Bead NaSCN Random Peptide Library Peptide Elution HeLa Nuclear Extract
DNA-dependent protein kinase
transformation/transcription domain-associated protein
Rap1-interacting factor 1
nuclear mitotic apparatus protein 1
spliceosome-associated protein 155
myosin targeting subunit 110
p53 binding protein 2
PTB-associated splicing factor
ribonucleoprotein: PTB-binding 1
PP2A scaffolding or A subunit
E1A-binding protein p400
apoptosis stimulating proteins of p53
DEAD box family of RNA helicases
protein phosphatase one
protein phosphatase 2A
epidermal Langerhans cell protein
nuclear inhibitor of protein phosphatase-1: ZAP (ZAP3): YLPM-motif contain PP1 interactor
PP1 interactor (also PNUTS)
skeletal muscle glycogen binding PP1 interactor
THO complex 1
54 kDa nuclear RNA binding protein
TBP-interacting protein 49/RuvB-like 2
protein phosphatase four
recruits PP1 onto mitotic chromatin at anaphase
myosin-binding subunit of 85 kDa
Leucine-rich repeats (LRRs): ribonuclease inhibitor
Glucose-Regulated Protein of 78-kDa.
The phosphorylation of proteins is one of the most prevalent covalent modifications known, affecting essentially every aspect of cellular function [1, 2]. The protein kinases and phosphatases responsible are highly conserved across species and, with few exceptions, the kinases belong to one large gene family while the phosphatase complement is more complex and can be divided into three broad groups based on protein sequence, catalytic signature and substrate preference [3, 4, 5]. The action of protein phosphatases is tightly controlled with cellular targeting being an important means of regulation. Most phospho-serine and threonine dephosphorylation can be attributed to the PPM family and the more diverse PPP family, which includes PP1, PP2A, PP2B, and PP4 through to PP7. PP1 is thought to not exist as a free catalytic subunit in the cell, but to reside in complexes with a large array of targeting or regulatory subunits that define its function. Numerous PP1 docking proteins have been identified, but they most likely represent only a small fraction of the total number in the cell.
The microcystins are a group of cyclic peptides that bind with remarkable specificity and affinity to the type one, 2A and several recently identified protein phosphatases of the PPP family (e.g. PP4, PP6). Microcystin covalently couples to a conserved cysteine residue of PPP family members through its methyl-dehydroalanine residue [6, 7]. Nishiwaki et al  first used Microcystin-Sepharose to purify PP2A. We exploited a different synthetic approach whereby the carbon-carbon double bond of methyl-dehydroalanine in microcystin couples the latter to aminoethanethiol, which is then linked to a Sepharose bead. This generates a high affinity binding matrix for the microcystin-sensitive protein phosphatases that does not covalently couple the phosphatase . Microcystin-Sepharose has proved to be a powerful tool to purify these protein phosphatases and their associated regulatory subunits from a variety of tissues and cell types [9, 10, 11, 12].
With only a few characterized exceptions, PP1 interacting proteins bind PP1 through their primary docking sequence called the RVXF/W motif . Their molecular interaction with PP1 has been visualized via PP1-peptide and PP1 regulatory subunit structures [14, 15]. It has also emerged that additional or secondary interaction sites often play a role in binding PP1 and likely contribute to PP1 isoform specificity recognition, substrate docking and modulation of PP1 activity [13, 15, 16, 17, 18, 19, 20]. Based on a compilation of demonstrated RVXF/W interaction motifs [21, 22, 23], the panning of a random peptide library , and mutagenesis and modeling studies  we noted preferences for particular amino acids within and adjacent to the RVXF/W motif. This led us to speculate that a PP1 interaction motif peptide, based on this comparison, could be a unique means to specifically disrupt PP1-targetting or regulatory subunit interactions [14, 25, 26]. Slight variation in the RVXF/W-motif combined with the now recognized additional, secondary PP1 interaction sites provide sufficient interaction specificity which may allow the development of drugs or peptide mimetics to abolish specific PP1 binding protein interactions in vivo. The idea of targeting protein-protein interaction domains with drugs has historically not been favored by the pharmaceutical industry yet, due to improved understanding of the underlying molecular mechanisms, it is now a concept that is growing in popularity [27, 28]. This idea has been explored with PKA anchoring proteins (AKAPs) where optimal RI and RII subunit binding peptides were derived from parent peptides and used to target PKA in vivo and displace it from its normal anchoring site . In addition to in vivo targeting, disrupting PP1-regulatory subunit interactions with a peptide would be an effective means to aid in identifying proteins in PP1 complexes and thus uncover new cellular processes regulated by this protein phosphatase.
Results and Discussion
RVRW peptide eluted samples were blotted for the known nuclear PP1 binding proteins ZAP, p99 (PNUTS) and NIPP1 (Figure 1c). This revealed that these proteins were readily displaced from the matrix by the peptide with essentially no further regulatory subunit being eluted with NaSCN. A direct comparison of elution conditions with the GKKRVRW ADLE peptide showed that displacement was much more effective in a higher ionic strength buffer (Figure 1d). Having established that the GKKRVRW ADLE peptide is an excellent tool to displace PP1 regulatory subunits, we tried several peptide concentrations to optimize displacement from the matrix (Figure 1e). This demonstrated that 0.1 mM peptide was sufficient to displace ZAP and p99, but 0.5 to 2 mM peptide was needed to remove all of the more abundant and perhaps higher affinity binding NIPP1 protein. At the 2 mM peptide concentration we noted that displacement was in fact not as effective for ZAP or p99. In subsequent experiments we eluted by first incubating with 0.5 mM, then 2 mM peptide to ensure all PP1 complex proteins were displaced. To ensure the specificity of the method, we included 2 independent negative controls. Nuclear phosphatases bound to the microcystin matrix were equally divided into 3 parts and eluted with either the GKKRVRW ADLE peptide, the GKKRVRW ADLE peptide where the key PP1 interacting amino acids (V and W) were replaced by alanine (GKKRARA ADLE), or a scrambled ZAP peptide (KLRGEVAKDWR). These fractions were blotted for ZAP, p99 and NIPP1 (Figure 1f) confirming the specificity of the displacement technique.
To extend this method to PP1 complexes outside the nucleus, we isolated glycogen particles from skeletal muscle where the well characterized PP1 regulatory subunit GM resides and functions to target PP1 to glycogen . As shown in Figure 1g, the peptide GKKRVRW ADLE effectively displaced GM from the microcystin-matrix. In addition, prior to the RVRW displacement, we performed an incubation with the GKKRARA ADLE peptide and, as predicted, due to the absence of the key V and W residues in the docking motif, this peptide did not displace the GM protein from the matrix which supports the idea that the peptide elution is specific for PP1 complexes. All subsequent experiments included the GKKRARA ADLE peptide elution and buffer wash step prior to GKKRVRW ADLE peptide displacement.
It is notable from the mass spectrometry data that we identified many nuclear processes or complexes not previously linked to PP1. For instance, we found numerous TIP60 (TRRAP) complex proteins . Immunoprecipitation of the TIP60 components TRRAP or p400 (Domino) were able to bring down PP1, confirming that PP1 is a previously unrecognized component of this complex. Putative interacting proteins or complexes identified here will need to be confirmed by co-immunoprecipitation experiments or other methods, and then ultimately all complexes, including the TRRAP complex, will need to be studied to define the role of PP1 binding to that protein or group of proteins.
A comparison of the proteins eluted and identified from rat and human nuclei reveals that most of the proteins are the same (although the number of rat proteins is greater), but several differences are apparent. First, some differences are not surprising given that the HeLa cell line expresses numerous proteins that differ from the rat liver cell proteome. In addition, there are quite a number of proteins in the rat samples that appear in multiple bands excised from the gel, but this only occurs a few times in the HeLa bands. This is likely due to the fact that preparation of the rat nuclei from liver tissue is more time consuming and thus we expect increased proteolysis of proteins and therefore many more hits of the same protein in bands of many masses. Comparing rat liver versus cells grown in culture also means that we started with much more protein from the rat preparation and this certainly contributed to the identification of more lower abundance proteins.
To date several approaches have been used to define the PP1-interactome of the cell, including total (NaSCN) elution from microcystin-Sepharose followed by overlay analysis, SILAC mass spectrometry based PP1 binding partner investigation and most recently an antibody array approach for selected putative targets [11, 31, 36, 37]. Any methodology has its limitations; our work certainly represents a subgroup of the total PP1 targeted protein/complexes of the nucleus for reasons discussed above and because our nuclear extraction procedure likely only released some fraction of the total PP1 complexes. It is clear that many more PP1 binding partners exist than originally thought and it will be the combination of multiple approaches that will ultimately define the plethora of roles for PP1 in the nucleus and other cellular compartments.
We have developed a method to specifically elute proteins that reside in PP1 complexes or bind directly to PP1 after retention on the affinity matrix microcystin-Sepharose. This approach should be valuable to isolate new PP1 binding proteins and to link PP1 function to as yet unrecognized PP1 regulated cellular processes across a broad range of tissues, cell types and organisms. This technique also lends further credibility to the notion that it could be possible to develop peptides as tools for the specific disruption of protein protein interactions with potential usage for therapeutic purposes.
HeLa nuclear extracts
For HeLa cells, nuclei were isolated and extracted from ~1 × 109 cells grown in spinner flasks using the method outlined at http://www.lamondlab.com. Sonicated, extracted nuclei were made to 0.42 M NaCl, mixed end-over-end for 10 min. at 4°C and clarified by centrifugation at 10,000 g for 10 min. This clarified extract was then mixed with the microcystin matrix.
Affinity displacement chromatography
Microcystin-Sepharose was prepared as described [7, 9]. Control matrix was prepared by coupling to Tris instead of aminoethanethiol-microcystin. In a typical nuclear preparation, the livers of 10 rats were homogenized, nuclei isolated and proteins extracted . Nuclear proteins were incubated end over end with 1 mL of microcystin-Sepharose or control matrix for 1 h at 4°C and then washed in a column with 250 mL of buffer A (25 mM Tris-HCl pH 7.5, 0.1 mM EGTA, 1 mM benzamidine, 0.1 mM PMSF, 0.1% β-ME) plus 300 mM NaCl. All peptides were dissolved in buffer A plus 300 mM NaCl and the pH adjusted to 7.5 with 1 N NaOH by spotting on pH paper. The matrix (1 mL) was first incubated with 2 mL 0.5 mM peptide for 30 min. and the eluted protein collected with the addition of 2 mL buffer A plus 300 mM NaCl. The matrix was then incubated with 2 mL 2 mM peptide for 30 min. and the eluted protein collected with the addition of 5 mL buffer A plus 300 mM NaCl. After washing with an additional 30 mL of buffer A plus 300 mM NaCl, the column was eluted with 3 M sodium isothiocyanate in buffer A . The peptides used were RPKRKRKNSRVTF SEDDEII (from human NIPP1), GKKRVRW ADLE (from human ZAP), GKKRARA ADLE (from human ZAP with alanine substitutions for V and W) and KLRGEVAKDWR (scrambled GKKRVRW ADLE). The 0.5 and 2 mM peptide elutions were pooled and immediately concentrated in a centriprep 10, then centricon 10 to 40 μL and boiled in 5× SDS-cocktail. The NaSCN elution was dialyzed extensively and concentrated to 40 μL as described for the peptide elutions and boiled in 5× SDS-cocktail. Samples were run on 10% SDS-PAGE or in some cases on 4–12% gradient gels (Invitrogen).
Isolation of rabbit skeletal muscle glycogen particles and chromatography on MC-Sepharose
Skeletal muscle from the back and hind legs of one rabbit was removed, placed on ice, minced and homogenized with 2.5 L/kg homogenization buffer (2 mM EDTA, 2 mM EGTA, 0.1% B-ME, 1 mM benzamidine, 0.1 mM PMSF) in a blender and centrifuged in a Beckman J6 at 4200 rpm for 30 min. at 4°C. The supernatant was decanted through glass wool in a funnel and the pH of the solution lowered to 6.1 with 1 M acetic acid. After 15 min on ice, glycogen was pelleted by centrifugation at 4200 rpm for 30 min. at 4°C. The glycogen pellet was resuspend in 100 mL of the following buffer (50 mM Tris-HCl pH 7, 2 mM EGTA, 5% V/V glycerol, 4 μg/mL leupeptin, 1 mM benzamidine, 0.1 mM PMSF, and 0.1% β-ME.) and loaded on a 1 mL microcystin-Sepharose column equilibrated with microcystin-Sepharose buffer A. After incubation for 1.5 hour with an end-over-end machine, then washing with microcystin-Sepharose buffer A plus 0.3 M NaCl (200 column volumes or until no protein comes off in a Bradford assay) the matrix was eluted with peptides and NaSCN as described above.
For protein identification by mass spectrometry bands were excised from colloidal blue stained gels, digested with trypsin and LC-MS/MS was performed as in Ulke-Lemée et al . Peak lists were searched with Mudpit scoring using Mascot version (v2.2). We used the criteria of 2 matching peptides for a positive identification.
Antibodies to proteins or an equivalent amount of pre-immune serum purified IgG were coupled to Protein A-Sepharose (GE Healthcare) and incubated with HeLa cell extracts (from 4 × 10 cm plates) end over end for 2 h at 4°C. After sedimentation, IP pellets were washed 2× with PBS, 2× with PBS plus 0.05% NP-40 and 150 mM NaCl and 2 more times in PBS, before boiling in SDS-cocktail. In all cases, antibody and pre-immune serum IPs were incubated with identical amounts of protein, washed in parallel and eluted with an equal volume of cocktail to allow direct comparison on western blots.
Antibodies were kindly provided by individuals or purchased as indicated in brackets. PP1, NIPP1, p99 (PNUTS), SAP155, ZAP and their use are detailed in Tran et al . Other antibodies were PR65 (B. Hemmings), GM and M110 (P. Cohen), DPK-1 (SP Lees-Miller), NUP153 (M. Lohka), DDX5 (F. Fuller-Pace), RAVER-1 (B. Jockusch), TOPOIIα (E. Kurz), TRRAP, Envoplakin, HDAC6, PSF (Santa Cruz), Ki-67, p53bp2 [ASPP2], p54nrb/nono (BD Biosciences), Rif-1, p84 [Thoc1], P400 (Abcam), Filamin-A (Chemicon), NUMA-1 (Nova Biologicals), DDX18, DDX17 [DNA helicaseA] (Bethyl Lab), HDAC1 (Cell Signaling), PP2A-C (Transduction Labs).
This work was supported by the Natural Sciences and Engineering Research Council of Canada and the Alberta Cancer Board.
- 5.Kerk D, Templeton G, Moorhead GB: Evolutionary radiation pattern of novel protein phosphatases revealed by analysis of protein data from the completely sequenced genomes of humans, green algae, and higher plants. Plant Physiol. 2008, 146 (2): 351-367. 10.1104/pp.107.111393.PubMedCentralCrossRefPubMedGoogle Scholar
- 28.Guergnon J, Dessauge F, Dominguez V, Viallet J, Bonnefoy S, Yuste VJ, Mercereau-Puijalon O, Cayla X, Rebollo A, Susin SA, et al.: Use of penetrating peptides interacting with PP1/PP2A proteins as a general approach for a drug phosphatase technology. Mol Pharmacol. 2006, 69 (4): 1115-1124. 10.1124/mol.105.019364.CrossRefPubMedGoogle Scholar
- 29.Alto NM, Soderling SH, Hoshi N, Langeberg LK, Fayos R, Jennings PA, Scott JD: Bioinformatic design of A-kinase anchoring protein-in silico: a potent and selective peptide antagonist of type II protein kinase A anchoring. Proc Natl Acad Sci USA. 2003, 100 (8): 4445-4450. 10.1073/pnas.0330734100.PubMedCentralCrossRefPubMedGoogle Scholar
- 32.Ulke-Lemee A, Trinkle-Mulcahy L, Chaulk S, Bernstein NK, Morrice N, Glover M, Lamond AI, Moorhead GB: The nuclear PP1 interacting protein ZAP3 (ZAP) is a putative nucleoside kinase that complexes with SAM68, CIA, NF110/45, and HNRNP-G. Biochim Biophys Acta. 2007, 1774 (10): 1339-1350.CrossRefPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.