A novel whole-cell lysate kinase assay identifies substrates of the p38 MAPK in differentiating myoblasts
- James DR Knight1, 2, 9,
- Ruijun Tian3, 4, 9,
- Robin EC Lee1, 2, 10,
- Fangjun Wang3, 5,
- Ariane Beauvais1,
- Hanfa Zou5,
- Lynn A Megeney1, 2, 6,
- Anne-Claude Gingras7, 8,
- Tony Pawson7, 8,
- Daniel Figeys3, 4 and
- Rashmi Kothary1, 2, 6Email author
© Knight et al; licensee BioMed Central Ltd. 2012
Received: 22 August 2011
Accepted: 6 March 2012
Published: 6 March 2012
The p38α mitogen-activated protein kinase (MAPK) is a critical mediator of myoblast differentiation, and does so in part through the phosphorylation and regulation of several transcription factors and chromatin remodelling proteins. However, whether p38α is involved in processes other than gene regulation during myogenesis is currently unknown, and why other p38 isoforms cannot compensate for its loss is unclear.
To further characterise the involvement of p38α during myoblast differentiation, we developed and applied a simple technique for identifying relevant in vivo kinase substrates and their phosphorylation sites. In addition to identifying substrates for one kinase, the technique can be used in vitro to compare multiple kinases in the same experiment, and we made use of this to study the substrate specificities of the p38α and β isoforms.
Applying the technique to p38α resulted in the identification of seven in vivo phosphorylation sites on six proteins, four of which are cytoplasmic, in lysate derived from differentiating myoblasts. An in vitro comparison with p38β revealed that substrate specificity does not discriminate these two isoforms, but rather that their distinguishing characteristic appears to be cellular localisation.
Our results suggest p38α has a novel cytoplasmic role during myogenesis and that its unique cellular localisation may be why p38β and other isoforms cannot compensate for its absence. The substrate-finding approach presented here also provides a necessary tool for studying the hundreds of protein kinases that exist and for uncovering the deeper mechanisms of phosphorylation-dependent cell signalling.
Keywordsdifferentiation FSBA kinase assay mitogen-activated protein kinase myoblast p38 phosphorylation quantitative MS
Protein kinases are well-known regulators of cell signalling and cellular behaviour that execute their function through the covalent attachment of an ATP-derived phosphate to protein substrates. To understand the function of any protein kinase on a large and cell-wide scale first requires the development of a substrate screening technique that allows for the proteins phosphorylated by a kinase of interest to be comprehensively identified, ideally in a single experiment. Although substrate-finding techniques exist, they are hindered by problems that prevent them from being easily or readily employed [1–4] and are generally limited to providing in vitro substrate identifications that may or may not be relevant in vivo. In vivo approaches currently available, such as that employed by Holt et al. , can associate a kinase with in vivo phosphorylation events, but direct phosphorylation cannot be inferred without additional experimentation. A simple technique that can identify direct in vivo substrates is an obvious need for the field.
The mitogen-activated protein kinase p38α is involved in several cellular processes, but its critical role during differentiation, and particularly the differentiation of myoblasts, has been a major focus. At the initiation of myoblast differentiation, p38α is known to phosphorylate several transcription factors and chromatin remodelling proteins, thereby inducing the expression of a myogenic gene program . Although much is known about p38α's role in this process, it is likely very partial, and whether p38α plays an important role in other processes during myoblast differentiation, such as cell fusion or sarcomere formation, is unknown. At the same time, there are questions regarding the other p38 isoforms and their role, or lack thereof, in myogenesis. p38β is also expressed in myoblasts and is activated in the same manner as p38α, but despite having a kinase domain 75% identical to that of p38α (72% sequence identity overall), p38β is unable to compensate for the loss of p38α, even when overexpressed [7–9]. The obvious and suspected explanation is that there are critical myogenic phosphorylations specific to the α isoform, but these have yet to be discovered and whether this assumption is correct is unknown.
Here we describe a simple approach for substrate finding that can be used to identify in vitro and in vivo substrates. The technique begins with treatment of cell lysate to inactivate endogenous kinases, followed by an in vitro assay using an exogenous kinase of interest, and concludes with quantitative mass spectrometry (MS) to identify phosphorylation sites specific to the added kinase. By using lysate derived from vehicle- or inhibitor-treated cells, this in vitro approach can be simultaneously coupled with biologically relevant information to identify direct substrates regulated by the kinase of interest in vivo. Applying this technique to p38α with lysate from differentiating myoblasts resulted in the identification of several new in vivo substrates that suggest novel functions for p38α during myogenesis. We did not identify a single phosphorylation specific to the p38α isoform compared with p38β, at least in terms of in vitro substrate specificity, but we did see a clear difference in cellular localisation during myoblast differentiation. This leads us to propose that although the kinase domains of p38α and β likely have the same capacity to phosphorylate substrates, there are major differences in actual substrate specificity in an in vivo context.
C2C12 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (vol/vol) foetal bovine serum with 100 U/ml penicillin, 100 μg/ml streptomycin and 250 ng/ml amphotericin B. To induce differentiation, cells were grown to 85-90% confluence, and the medium was changed to DMEM with 2% horse serum supplemented with penicillin, streptomycin and amphotericin B as described above. FLAG-tagged p38α and p38β were acquired from Addgene (Cambridge, MA, USA) . Constructs were transfected into C2C12 myoblasts with Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. To inhibit p38 activity, SB 202190 (Promega, Madison, WI, USA) solubilised in dimethyl sulphoxide (DMSO) or DMSO as control was added to differentiation media at 48 hours following the induction of differentiation, and medium with inhibitor was changed daily.
Cells were fixed with 4% formaldehyde and stained with the following antibodies: Flag M2 (1:1,000 dilution; Sigma-Aldrich, St Louis, MO, USA), myosin heavy chain (MyHC) (1:20 dilution; Developmental Studies Hybridoma Bank, Department of Biology, University of Iowa, Iowa City, IA, USA), Alexa Fluor 488 goat anti-mouse antibody (1:1,000 dilution; Molecular Probes/Invitrogen) and Alexa Fluor 555 goat anti-mouse antibody (1:1,000 dilution; Molecular Probes/Invitrogen). The differentiation index was calculated as the number of MyHC-positive nuclei divided by the total number of nuclei. The fusion index was quantified as the number of nuclei per MyHC-positive cell. Five fields of view at ×20 magnification were counted and averaged per replicate, with a total of three replicate experiments.
Statistical analyses were performed using StatPlus software (AnalystSoft Inc; http://www.analystsoft.com/en/products/statplus/). The data shown are means with SD, and Student's t-tests were performed to determine significance for the differentiation and fusion indices.
Western blot analysis
For Western blot analysis, cells were lysed in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris∙HCl, pH 7.5, 1% Nonidet P-40, 0.1% SDS, 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid (EDTA), 50 mM NaF, 200 μM Na3VO4, 1 mM phenylmethanesulphonylfluoride (PMSF), 10 μg/ml aprotinin, 10 μg/ml leupeptin and 10 μg/ml pepstatin), and 5× Laemmli buffer (300 mM Tris, pH 6.8, 0.01% bromophenol blue, 10% SDS, 50% glycerol and 5% β-mercaptoethanol) was added to 25 μg of protein per sample to a final concentration of 1× for SDS-PAGE. For fractionation, cells were lysed in hypotonic buffer (10 mM 4-(2-hydroxyethyl)-1-piperazineethanesulphonic acid (HEPES), pH 7.5, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM dithiothreitol (DTT), 50 mM NaF, 200 μM Na3VO4, 1 mM PMSF, 10 μg/ml aprotinin, 10 μg/ml leupeptin and 10 μg/ml pepstatin) and left on ice for 10 minutes. Cells were then passed through a 25-gauge needle three times and centrifuged at 500 g to pellet nuclei and unlysed cells. The supernatant was collected as whole cytoplasm. For further fractionation of the cytoplasm, the supernatant was centrifuged again at 5,000 g to pellet mitochondria and membrane fractions. The supernatant was then collected and centrifuged at 100,000 g to pellet any remaining cell particles, and the resulting supernatant was collected as cytosol. RIPA buffer was added to all fractions to a final concentration of 1× for complete lysis. Recombinant phosphorylated p38α (Millipore, Billerica, MA, USA) was dephosphorylated using λ-protein phosphatase (New England Biolabs, Inc, Ipswich, MA, USA) by adding 800 U of phosphatase to 200 ng of p38α diluted in λ-phosphatase buffer, and the sample was assayed at 30°C for 1 hour. Antibodies used for blotting were as follows: α-actinin (1:125 dilution; Abcam, Cambridge, MA, USA), COX IV (1:1,000 dilution; Abcam), GRP78 (1:500 dilution; Cell Signaling Technology, Danvers, MA, USA), lamin A/C (1:500 dilution; Abcam), MyHC (1:100 dilution; Developmental Studies Hybridoma Bank), MyoD (1:1,000 dilution; Santa Cruz Biotechnology, Santa Cruz, CA, USA), myogenin (1:100 dilution; Developmental Studies Hybridoma Bank), neural cell adhesion molecule (1:200 dilution; Abcam), p38α (1:500 dilution; Cell Signaling Technology), phospho-p38 (1:500 dilution; Abcam) and β-tubulin (1:1,000 dilution; Developmental Studies Hybridoma Bank). An Alpha Innotech HD2 imaging system (R&D Systems, Minneapolis, MN, USA) was used to quantify phospho-p38 and tubulin expression.
FSBA treatment and substrate labelling
Cells were lysed in Nonidet P-40 buffer (50 mM Tris∙HCl, pH 7.8, 150 mM NaCl, 1% (vol/vol) Nonidet P-40, 1 mM PMSF, 10 μg/ml aprotinin, 10 μg/ml leupeptin and 10 μg/ml pepstatin). Lysate was treated at a concentration of 2 mg/ml with 20 mM 5'-4-fluorosulphonylbenzoyladenosine (FSBA) solubilised in DMSO and placed at 30°C for 1 hour. The sample was then diluted down to 1:5 with Nonidet P-40 buffer minus protease inhibitors and desalted using Millipore Amicon ultrafiltration columns with a 10 kDa molecular weight cutoff. Following concentration, the sample was diluted to 4 mg/ml with Nonidet P-40 buffer and diluted 1:2 with 2× kinase assay buffer (40 mM 3-morpholinopropane-1-sulphonic acid (MOPS), pH 7.2, 50 mM β-glycerophosphate, 10 mM ethylene glycol tetraacetic acid (EGTA), 2 mM Na3VO4, 2 mM DTT, 50 mM MgCl2, 400 μM cold ATP and 5 μCi [γ-32P]ATP). Recombinant p38α or p38β (Millipore) was added to a final concentration of 0.5% (wt/wt) total protein. Control and kinase-added samples were assayed at 30°C for 1.5 hours. For one-dimensional SDS-PAGE, 5× Laemmli buffer was added following the assay to 1×, and the sample was electrophoresed. For two-dimensional electrophoresis, 17-cm ReadyStrip immobilised pH gradient (IPG) strips (Bio-Rad Laboratories, Hercules, CA, USA) were directly rehydrated with labelled lysate diluted in rehydration buffer (7 M urea, 2 M thiourea, 4% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulphonate (CHAPS) and 1% DTT) following the manufacturer's directions. Isoelectric focusing was performed on a PROTEAN IEF Cell (Bio-Rad Laboratories) under the following conditions: 200 V for 1 hour, 500 V for 1 hour, 5,000 V ramp for 5 hours and 5,000 V for 80,000 VH. IPG strips were then equilibrated following the manufacturer's instructions and overlaid onto a 12% SDS-PAGE gel. Following electrophoresis, gels were dried and imaged. For one-dimensional electrophoresis, 100 μg of lysate was used per reaction. For two-dimensional electrophoresis, 300 μg of lysate was used.
For in vitro substrate identification, assays were performed as described above with the following modifications. A quantity of 1.5 mg of lysate was treated with FSBA, the sample was desalted and 2× kinase assay buffer was added (40 mM MOPS, pH 7.2, 50 mM β-glycerophosphate, 10 mM EGTA, 2 mM Na3VO4, 2 mM DTT, 50 mM MgCl2 and 2 mM cold ATP). The sample was then split into three 500-μg aliquots, and 5 μg of heat-inactivated p38α was added to the control, 5 μg of active p38α was added to the second aliquot and 5 μg of active p38β was added to the third aliquot. The samples were then assayed for 3 hours at 30°C.
For in vivo substrate identification, assays were performed as above with the following modifications. At 48 hours of differentiation, cells were treated with 5 μM of SB 202190 or an equivalent amount of DMSO as vehicle. Twenty-four hours later the cells were lysed in Nonidet P-40 buffer. Lysate (1 mg) from DMSO-treated cells or 2× 1 mg of lysate from SB 202190-treated cells was inhibited with FSBA, the samples were desalted and 2× kinase assay buffer was added to each. A quantity of 5 μg of active p38α was added to one of the lysate aliquots from SB-treated cells. The samples were then assayed for 3 hours at 30°C.
After assaying the samples, they were precipitated by methanol chloroform, then redissolved in 200 μl of 8 M urea and 50 mM Tris∙HCl, pH 8.1, with sonication. The samples were then reduced with 20 mM DTT for 1 hour at 60°C and alkylated by 100 mM iodoacetamide for 30 minutes at room temperature in the dark. Subsequently, the samples were diluted to 2 M urea with 50 mM Tris∙HCl, pH 8.1, and digested with trypsin at a protein-to-trypsin ratio of 50:1 (wt/wt) for 16 hours at 37°C. Next, the digested samples were acidified to pH 2 using 10% (vol/vol) formic acid. Dimethyl labelling of the samples was performed as reported previously  and is described briefly as follows. The acidified peptides were loaded onto C18 solid phase extraction (SPE) columns (50 mg of packing material). After brief washing with 50 mM sodium phosphate buffer, pH 7.5, 3 ml of light, intermediate and heavy labelling reagents were loaded onto C18 SPE columns trapped with control, p38α- and p38β-labelled samples, respectively. After being washed with 0.1% (vol/vol) formic acid, the labelled samples were eluted with 80% acetonitrile (ACN) (vol/vol) and 0.1% (vol/vol) formic acid, then dried by vacuum centrifugation.
Phosphopeptide enrichment by TiO2 was carried out as reported previously  with modifications. The dried samples were redissolved with 65% ACN/2% trifluoroacetic acid (TFA)/saturated glutamic acid and combined. TiO2 beads suspended in 65% ACN/2% TFA/saturated glutamic acid were added into the above samples with a peptide to TiO2 bead ratio of 1:4 (wt/wt). After being nutated for 40 minutes, the TiO2 beads were recovered by centrifugation and washed thoroughly with 65% ACN/2% TFA. Finally, the enriched phosphopeptides were eluted with 10% (vol/vol) NH3·H2O and dried by vacuum centrifugation.
Online liquid chromatography tandem mass spectrometry analysis
Online liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis was performed as reported previously [13, 14] with modifications. The dried sample was redissolved with 0.1% formic acid and loaded onto a biphasic trap column (200 μm ID × 10 cm; 5-cm reversed phase column packed with ReproSil-Pur C18 resin (5 μm at 200 Å; Dr.Maisch GmbH, Ammerbuch-Entringen, Germany) and a 5-cm monolith strong cation exchange (SCX) column). The trapped phosphopeptides were eluted from the trap column onto a C18 tip column (75 μm ID × 20 cm, 3 μm at 200 Å; Dr.Maisch GmbH) by a series of salt washes at increasing concentrations (0, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 70, 80, 90, 100, 150, 200 and 1,000 mM). Each fraction was then separated by reversed phase-based gradient elution and detected using an LTQ Orbitrap XL Fourier transform mass spectrometer (Thermo Scientific, Waltham, MA, USA). The reversed phase gradient was set as follows: 0% to 5% ACN for 2 minutes, 5% to 30% ACN for 90 minutes and 30% to 80% ACN for 5 minutes. After flushing with 80% ACN for 10 minutes, the column was equilibrated with 0.1% formic acid aqueous solution for 13 minutes. The LTQ Orbitrap XL Fourier transform mass spectrometer was operated in positive ionization mode. A voltage of 1.8 kV was applied. MS and MS/MS spectra were acquired in a data-dependent mode, and one full MS scan was followed by ten MS/MS scans. The resolution was set at 60,000 at m/z 400 after accumulation to a target value of 500,000.
Protein identification and quantification
All MS/MS spectra in one acquired raw file were converted to a single *.mgf file using DTASuperCharge version 2.0a7 (Matrix Science, Boston, MA, USA). The *.mgf file was queried against the International Protein Index mouse database version 3.52 (EMBL-EBI; http://www.ebi.ac.uk/IPI/IPIhelp.html) using Mascot version 2.1 (Matrix Science). To evaluate the false discovery rate (FDR), reversed sequences were appended to the database. Cysteine residues were searched as a static modification of +57.0215 Da; methionine residues were searched with a variable modification of +15.9949 Da; and serine, threonine and tyrosine residues were searched with a variable modification of +79.9663 Da. Light, intermediate and heavy dimethylation of peptide amino termini and lysine residues were set as variable modifications of +28.0313 Da, +32.0564 Da and +36.0757 Da, respectively. Peptides were queried using full tryptic cleavage constraints with up to two missed cleavage sites. The mass tolerances were 7 ppm for parent masses and 0.5 Da for fragment masses. Phosphopeptides with a Mascot score ≥ 30 (rank 1, P ≤ 0.05, bold red required) were selected and quantified (FDR < 0.01). Phosphorylation site localisation and phosphopeptide quantification were performed using a dimethyl-adapted version of MSQuant version 2.0a81. For each peptide, the putative site of phosphorylation yielding the highest posttranslational modification (PTM) score was accepted (PTM score > 13 required, as described previously ). Peptide ratios were obtained by calculating the extracted ion chromatograms of the light, medium and heavy forms of the peptide using the monoisotopic peaks only, and protein ratios were calculated from the average of all quantified peptides. All MSQuant outputs of the same online multidimensional separation were then imported into StatQuant version 1.2.2, and the quantified phosphopeptides were sorted together and exported.
1 μg recombinant HSP27 (Enzo Life Sciences, Farmingdale, NY, USA), recombinant Rps27 (Abnova, Taipei City, Taiwan) or recombinant Smad9 (Abnova) were diluted in 30 μl of 1× kinase assay buffer (20 mM MOPS, pH 7.2, 25 mM β-glycerophosphate, 5 mM EGTA, 1 mM Na3VO4, 1 mM DTT, 25 mM MgCl2, 200 μM cold ATP and 2.5 μCi [γ-32P]ATP). A quantity of 500 ng of p38α or p38β was then added, and the samples were assayed at 30°C for 1 hour. Laemmli buffer at 5× concentration was added to 1× to terminate the assays, then the samples were electrophoresed and the gels were dried and imaged. Peptide assays were performed similarly with 5 μg of the following peptides synthesised by Biomatik (Cambridge, ON, Canada): HSP27 S180-APLPKAVTQSAEITIPVTF, HSP27 A180-APLPKAVTQAAEITIPVTF, Rps27 S27-KHKKKRLVQSPNSYFMDVK, Rps27 A27-KHKKKRLVQAPNSYFMDVK, Smad9 T136-NPYHYQRVETPVLPPVLVP and Smad9 A136-NPYHYQRVEAPVLPPVLVP. The frequency logo was generated using WebLogo .
FSBA inhibits endogenous protein kinases
Kinase-specific substrate labelling
Cell lysate treated with FSBA can be desalted to remove any unbound inhibitor and a pool of protein with no inherent kinase activity is generated. A kinase of interest can then be added with a kinase assay buffer, and any labelling that subsequently occurs is due to the added kinase as opposed to an endogenous one. To specifically label substrates of p38α, a kinase assay buffer containing [γ-32P]ATP was added to FSBA-treated C2C12 lysate, along with recombinant p38α, and the sample was assayed at 30°C. Substrates labelled by p38α appeared as bands following one-dimensional gel electrophoresis or as spots in two-dimensional gel electrophoresis with no contaminating signal from endogenous kinases (Figure 1B and Additional file 1 Figure S1). Although this type of approach is excellent for visualising phosphorylation, it is very difficult to identify phosphorylated proteins through spot-picking and MS. We therefore sought an alternative gel- and radioactive-free approach for identifying phosphorylated proteins.
Quantitative MS coupled with a phosphopeptide enrichment to identify substrates
Substrate specificity does not distinguish p38α and p38β
If substrate specificity does not distinguish the two p38 isoforms, and yet p38β is unable to compensate for the loss of p38α, there must be an alternative characteristic that discriminates them. An obvious possibility is cellular localisation. If p38α and p38β localise differently within the cell, then p38β would simply be unable to fulfil p38α's role, even though it might have the catalytic potential to do so. To study this, we overexpressed FLAG-tagged p38α and p38β in C2C12 cells and assessed their localisation during differentiation (Figure 4E). Whereas p38α has a ubiquitous localisation pattern, p38β is found only at the periphery of the cell. p38α therefore has access to a substrate pool that p38β does not, highlighting a major reason why p38β cannot compensate for the loss of p38α.
Identification of novel in vivo p38α substrates
Although the approach we have outlined can identify in vitro kinase substrates, such in vitro approaches will also turn up irrelevant substrates that are phosphorylated only because of the absence of the appropriate cellular context. For example, we identified just over 400 phosphorylation sites in the screen described above, 158 of which are in vitro p38 targets. If these numbers are representative, it suggests that p38 could be responsible for approximately 40% of all cellular phosphorylations in differentiating myoblasts, a number that seems unlikely, given that there are about 500 mouse and human kinases .
Newly discovered in vivo p38α substrates
Neuroblast differentiation-associated protein AHNAK
Protein IWS1 homolog
78 kDa glucose-regulated protein
Membrane-associated progesterone receptor component 1
E3 SUMOa protein ligase RanBP2
We have described here a simple technique for kinase substrate finding that uses whole-cell lysate, can identify sites of phosphorylation in addition to the protein phosphorylated, has the ability to identify both in vivo and in vitro substrates, and can be used to compare the in vitro substrate specificity of two kinases in the same experiment. Performing the assay with p38α and p38β revealed that, although differences in preference are apparent, there are no obvious phosphorylation sites specific to p38α. However, a major characteristic that distinguishes p38α from p38β during myoblast differentiation is localisation, with p38α located ubiquitously throughout the cell while p38β is found only at the cell periphery. We also have shown that p38α can phosphorylate several cytosolic phosphorylation sites in vivo and that this may be a critical but previously unknown function for it. These results demonstrate the utility of the substrate screening technique we have developed and how its use not only can find proteins phosphorylated by a kinase of interest but also can answer particular questions and propose novel avenues of research.
The technique we have developed has several advantages over existing substrate finding approaches. It can provide direct substrate identifications and information on in vivo relevance in a single experiment. Whole-cell lysate can be used as a source of candidate substrates, allowing screens to be performed on samples of specific relevance to the area of interest instead of on protein or peptide arrays. Our approach for eliminating endogenous kinase activity in lysate is nondenaturing, decreasing the likelihood of false-positive and false-negative substrate phosphorylations that would result from the heat-inactivating approach that others have employed [31, 32]. At the same time, the recombinant kinase used in the assay does not require mutation or unnatural ATP analogues as other approaches do , making the technique applicable to any kinase that can be made and purified in an active form. Although we used dimethyl labelling for quantitative MS, the technique is fully compatible with iTRAQ, which allows up to eight samples to be compared in a single experiment, meaning that simultaneous in vivo substrate identification and kinase comparisons can be performed. Radioactivity and gel work are not required, but can be used to visualise phosphorylations and make qualitative comparisons between kinases on one-dimensional gels. The lysate requirements are also relatively low; 1.5 to 3 mg were used in these assays, which may make the technique difficult to employ with some primary cells lines, but it is easily applicable to most secondary cell lines and to tissues as well. We have treated several different lysate types with FSBA, and complete inhibition of endogenous kinase activity occurred in all, meaning that our approach is likely universally applicable. In this study our approach for identifying in vivo substrates utilised an inhibitor, but it is more versatile than that, and can be applied to study phosphorylations triggered by a specific stimulus (using stimulated versus unstimulated cells for lysate pools) without requiring a specific inhibitor be available. Finally, although we have performed the in vivo approach on unfractionated lysate, when combined with fractionation prior to LC-MS/MS, the number of substrates identified could be much greater than what we have demonstrated in principle here.
As mentioned, a major advantage of the technique presented here is that it can be used to compare the substrate profiles of two or more kinases, and we used this to study the substrate specificity of the p38α and β isoforms. However, we were unable to identify any phosphorylations that were specific to p38α. There are apparent differences in preference for substrates between p38α and p38β, which could be functionally relevant. Assuming that p38β had the same ubiquitous localisation pattern as p38α, a low preference for certain critical myogenic phosphorylations could result in a difference in cell behaviour if p38β were the only isoform present. However, it seems likely that differentiation would still occur, but possibly at a reduced rate or in a compromised way. This is not the case, as p38β-containing myoblasts that lack p38α fail to differentiate at all [7–9, 34]. Rather, we believe what truly distinguishes the two isoforms is their localisation. p38α has a known critical role in the nucleus, and our results suggest that the same may be true for the cytoplasm; therefore, p38β would be unable to compensate because it is found solely at the periphery of the cell. A similar distinction exists for two other closely related kinases, focal adhesion kinase (FAK) and proline-rich tyrosine kinase 2 (PYK2) . In fibroblasts, FAK localises to focal adhesions, whereas PYK2 has a perinuclear distribution. PYK2, which has a kinase domain 60% identical to that of FAK, can compensate for the loss of FAK provided it has a focal adhesion targeting sequence. Together these results suggest that rather than substrate specificity, it is other characteristics, in these cases localisation, that distinguish closely related kinases. Whether the same holds true for the entire kinase family is an intriguing question.
In addition to p38α's known role in regulating gene expression at the onset of myoblast differentiation, our results suggest that it is likely to have a cytosolic role as well, possibly in regulating its own pathway. We have found that p38α is present in the cytosol, active p38 increases in the cytoplasm with differentiation, p38α can phosphorylate several cytosolic proteins, three of the cytosolic proteins contain predicted D domains, and connections between these proteins and the p38α activation pathway have been previously reported. Together these results suggest a previously unrecognised role for p38α in the cytosol.
While the role of the p38α MAPK during myoblast differentiation has been studied extensively with regard to gene regulation, our evidence suggests that this kinase is likely involved in other processes as well. The differentiation of myoblasts into a mature myotube involves extensive morphological and functional changes that p38α might regulate, both through the initiation of a myogenic gene program and through direct phosphorylation and functional modification of cytosolic proteins. No one has previously been able to identify a p38α-specific substrate, and our large-scale screen suggests these may exist only within the context of cellular localisation. In addition, the substrate screening technique we have developed should serve as a useful tool for studying the other kinases known to be involved in myogenesis, as well as the hundreds of other protein kinases that exist.
JDRK was supported by a Vanier Canada Graduate Scholarship, a Canadian Institutes of Health Research (CIHR) Canada Graduate Scholarship and a Multiple Sclerosis Society of Canada (MSSC) Studentship. LAM holds the Mach Gaensslen Chair in Cardiac Research and was funded by CIHR. ACG is a Canada Research Chair in Functional Proteomics and the Lea Reichmann Chair and is funded by CIHR grant MOP-84314. DF acknowledges a Canada Research Chair in Proteomics and Systems Biology as well as grants from the Natural Sciences and Engineering Research Council of Canada and the J-Louis Lévesque Foundation. RK holds a University Health Research Chair at the University of Ottawa and was supported by grants from CIHR and the MSSC.
- Berwick DC, Tavaré JM: Identifying protein kinase substrates: hunting for the organ-grinder's monkeys. Trends Biochem Sci 2004, 29:227–232.PubMedView Article
- Johnson SA, Hunter T: Kinomics: methods for deciphering the kinome. Nat Methods 2005, 2:17–25.PubMedView Article
- Manning BD, Cantley LC: Hitting the target: emerging technologies in the search for kinase substrates. Sci STKE 2002, 2002:pe49.PubMedView Article
- Sopko R, Andrews BJ: Linking the kinome and phosphorylome: a comprehensive review of approaches to find kinase targets. Mol Biosyst 2008, 4:920–933.PubMedView Article
- Holt LJ, Tuch BB, Villén J, Johnson AD, Gygi SP, Morgan DO: Global analysis of Cdk1 substrate phosphorylation sites provides insights into evolution. Science 2009, 325:1682–1686.PubMedView Article
- Lluís F, Perdiguero E, Nebreda AR, Muñoz-Cánoves P: Regulation of skeletal muscle gene expression by p38 MAP kinases. Trends Cell Biol 2006, 16:36–44.PubMedView Article
- Li Y, Jiang B, Ensign WY, Vogt PK, Han J: Myogenic differentiation requires signalling through both phosphatidylinositol 3-kinase and p38 MAP kinase. Cell Signal 2000, 12:751–757.PubMedView Article
- Perdiguero E, Ruiz-Bonilla V, Gresh L, Hui L, Ballestar E, Sousa-Victor P, Baeza-Raja B, Jardí M, Bosch-Comas A, Esteller M, Caelles C, Serrano AL, Wagner EF, Muñoz-Cánoves P: Genetic analysis of p38 MAP kinases in myogenesis: fundamental role of p38α in abrogating myoblast proliferation. EMBO J 2007, 26:1245–1256.PubMedView Article
- Wang H, Xu Q, Xiao F, Jiang Y, Wu Z: Involvement of the p38 mitogen-activated protein kinase α, β, and γ isoforms in myogenic differentiation. Mol Biol Cell 2008, 19:1519–1528.PubMedView Article
- Enslen H, Raingeaud J, Davis RJ: Selective activation of p38 mitogen-activated protein (MAP) kinase isoforms by the MAP kinase kinases MKK3 and MKK6. J Biol Chem 1998, 273:1741–1748.PubMedView Article
- Boersema PJ, Raijmakers R, Lemeer S, Mohammed S, Heck AJ: Multiplex peptide stable isotope dimethyl labeling for quantitative proteomics. Nat Protoc 2009, 4:484–494.PubMedView Article
- Li QR, Ning ZB, Tang JS, Nie S, Zeng R: Effect of peptide-to-TiO 2 beads ratio on phosphopeptide enrichment selectivity. J Proteome Res 2009, 8:5375–5381.PubMedView Article
- Wang F, Dong J, Jiang X, Ye M, Zou H: Capillary trap column with strong cation-exchange monolith for automated shotgun proteome analysis. Anal Chem 2007, 79:6599–6606.PubMedView Article
- Wang F, Chen R, Zhu J, Sun D, Song C, Wu Y, Ye M, Wang L, Zou H: A fully automated system with online sample loading, isotope dimethyl labeling and multidimensional separation for high-throughput quantitative proteome analysis. Anal Chem 2010, 82:3007–3015.PubMedView Article
- Olsen JV, Blagoev B, Gnad F, Macek B, Kumar C, Mortensen P, Mann M: Global, in vivo , and site-specific phosphorylation dynamics in signaling networks. Cell 2006, 127:635–648.PubMedView Article
- Crooks GE, Hon G, Chandonia JM, Brenner SE: WebLogo: a sequence logo generator. Genome Res 2004, 14:1188–1190.PubMedView Article
- Hashimoto E, Takio K, Krebs EG: Amino acid sequence at the ATP-binding site of cGMP-dependent protein kinase. J Biol Chem 1982, 257:727–733.PubMed
- Kamps MP, Taylor SS, Sefton BM: Direct evidence that oncogenic tyrosine kinases and cyclic AMP-dependent protein kinase have homologous ATP-binding sites. Nature 1984, 310:589–592.PubMedView Article
- Russo MW, Lukas TJ, Cohen S, Staros JV: Identification of residues in the nucleotide binding site of the epidermal growth factor receptor/kinase. J Biol Chem 1985, 260:5205–5208.PubMed
- Zoller MJ, Nelson NC, Taylor SS: Affinity labeling of cAMP-dependent protein kinase with p -fluorosulfonylbenzoyl adenosine: covalent modification of lysine 71. J Biol Chem 1981, 256:10837–10842.PubMed
- Knight JD, Qian B, Baker D, Kothary R: Conservation, variability and the modeling of active protein kinases. PLoS One 2007, 2:e982.PubMedView Article
- Schwartz D, Gygi SP: An iterative statistical approach to the identification of protein phosphorylation motifs from large-scale data sets. Nat Biotechnol 2005, 23:1391–1398.PubMedView Article
- Manning G, Whyte DB, Martinez R, Hunter T, Sudarsanam S: The protein kinase complement of the human genome. Science 2002, 298:1912–1934.PubMedView Article
- Andrés V, Walsh K: Myogenin expression, cell cycle withdrawal, and phenotypic differentiation are temporally separable events that precede cell fusion upon myogenesis. J Cell Biol 1996, 132:657–666.PubMedView Article
- Diebold ML, Koch M, Loeliger E, Cura V, Winston F, Cavarelli J, Romier C: The structure of an Iws1/Spt6 complex reveals an interaction domain conserved in TFIIS, Elongin A and Med26. EMBO J 2010, 29:3979–3991.PubMedView Article
- Yoh SM, Cho H, Pickle L, Evans RM, Jones KA: The Spt6 SH2 domain binds Ser2-P RNAPII to direct Iws1-dependent mRNA splicing and export. Genes Dev 2007, 21:160–174.PubMedView Article
- Asally M, Yasuda Y, Oka M, Otsuka S, Yoshimura SH, Takeyasu K, Yoneda Y: Nup358, a nucleoporin, functions as a key determinant of the nuclear pore complex structure remodeling during skeletal myogenesis. FEBS J 2011, 278:610–621.PubMedView Article
- Obenauer JC, Cantley LC, Yaffe MB: Scansite 2.0: Proteome-wide prediction of cell signaling interactions using short sequence motifs. Nucleic Acids Res 2003, 31:3635–3641.PubMedView Article
- Warde-Farley D, Donaldson SL, Comes O, Zuberi K, Badrawi R, Chao P, Franz M, Grouios C, Kazi F, Lopes CT, Maitland A, Mostafavi S, Montojo J, Shao Q, Wright G, Bader GD, Morris Q: The GeneMANIA prediction server: biological network integration for gene prioritization and predicting gene function. Nucleic Acids Res 2010, (38 Web server):W214-W220.
- Knight JD, Kothary R: The myogenic kinome: protein kinases critical to mammalian skeletal myogenesis. Skelet Muscle 2011, 1:29.PubMedView Article
- Huang SY, Tsai ML, Chen GY, Wu CJ, Chen SH: A systematic MS-based approach for identifying in vitro substrates of PKA and PKG in rat uteri. J Proteome Res 2007, 6:2674–2684.PubMedView Article
- Troiani S, Uggeri M, Moll J, Isacchi A, Kalisz HM, Rusconi L, Valsasina B: Searching for biomarkers of Aurora-A kinase activity: identification of in vitro substrates through a modified KESTREL approach. J Proteome Res 2005, 4:1296–1303.PubMedView Article
- Blethrow JD, Glavy JS, Morgan DO, Shokat KM: Covalent capture of kinase-specific phosphopeptides reveals Cdk1-cyclin B substrates. Proc Natl Acad Sci USA 2008, 105:1442–1447.PubMedView Article
- Ruiz-Bonilla V, Perdiguero E, Gresh L, Serrano AL, Zamora M, Sousa-Victor P, Jardí M, Wagner EF, Muñoz-Cánoves P: Efficient adult skeletal muscle regeneration in mice deficient in p38β, p38γ and p38δ MAP kinases. Cell Cycle 2008, 7:2208–2214.PubMedView Article
- Klingbeil CK, Hauck CR, Hsia DA, Jones KC, Reider SR, Schlaepfer DD: Targeting Pyk2 to β1-integrin-containing focal contacts rescues fibronectin-stimulated signaling and haptotactic motility defects of focal adhesion kinase-null cells. J Cell Biol 2001, 152:97–110.PubMedView Article