Discovery and Identification of Serine and Threonine Phosphorylated Proteins in Activated Mast Cells: Implications for Regulation of Protein Synthesis in the Rat Basophilic Leukemia Mast Cell Line RBL-2H3 Fredrik J. Olson,† Russell I. Ludowyke,‡ and Niclas G. Karlsson*,§ Proteome Systems Limited, Locked Bag 2073, North Ryde, Sydney, NSW 1670, Australia Received December 17, 2008
Mast cells are important in allergic inflammation and innate immunity. Antigen-induced activation via cell-surface receptors initiates a series of intracellular signaling events, leading to the secretion of inflammatory mediators. While many of the kinases involved in this process have been defined, their substrates are generally unknown. This study aimed to identify proteins phosphorylated by serine or threonine kinases in the early stages of mast cell activation, using the rat basophilic leukemia cell line RBL-2H3 as a model system. Cells were activated via FcεRI cross-linking, and lysed at different time points between 1-10 min. A novel, specific mixture of serine and threonine phospho-specific antibodies was utilized, and was shown to selectively detect proteins that were phosphorylated upon cell activation. The mixture of antibodies was used to immunoprecipitate such regulated phosphoproteins from cell lysates enriched in phosphoproteins by phospho-affinity chromatography. Immunoprecipitated proteins were analyzed by SDS-PAGE, Western blotting and liquid chromatography/mass spectrometry. With this approach, we highlighted a number of phosphoproteins, demonstrated differences in the phosphorylation/dephosphorylation rates among the regulated proteins, and identified eleven serineor threonine-phosphorylated proteins that are substrates of kinases involved in mast cell intracellular signaling. Among these were proteins with functions in protein metabolism, including elongation factor 2, calnexin and heat shock proteins; and cell structure, including moesin, tubulin and actin. The novel approach applied in this study proved useful for the identification of kinase substrates, and can readily be extended for use in similar phosphoproteomic studies. Keywords: Phosphorylation • Proteomics • Mast cell • Phospho-specific antibodies • Phosphoproteomics
Introduction Protein phosphorylation is a post-translational modification that is one of the major regulatory mechanisms in the cell. With the dynamic addition/removal of phosphate groups, a protein can be activated/deactivated or its function altered. Phosphoproteomics; studying the phosphoprotein content in a cell at any given moment, should address the problems in phosphorylation analysis, in addition to interpreting data to identify phosphorylation pathways and phosphorylation targets. This would not only provide a fundamental understanding of protein phosphorylation, but also has the potential for the discovery of novel biomarkers and drug targets for diseases.1 The analysis of protein phosphorylation is a challenging task, due to the low abundance of phosphoproteins, low stoichi* To whom correspondence should be addressed. E-mail, niclas.karlsson@ nuigalway.ie; phone, +353 (0)91 495606; fax, +353 (0)91 525700. † Current address: Center for Cardiovascular and Metabolic Research, Sahlgrenska Academy, University of Gothenburg, Sweden. ‡ Current address: Medical Affairs, Schering-Plough Pty Limited, North Ryde, Sydney, NSW 2113, Australia. § Current address: Centre for BioAnalytical Sciences, School of Chemistry, National University Ireland Galway, Ireland.
3068 Journal of Proteome Research 2009, 8, 3068–3077 Published on Web 03/24/2009
ometry of phosphorylation, short times in which proteins are phosphorylated, and difficulties in ionizing phosphopeptides using mass spectrometry. Often, medium- to low-abundance proteins are not detected using two-dimensional (2D) gel electrophoresis of crude cell extracts, making this approach unsuitable for studying the dynamic phosphorylation of the notoriously low-abundance phosphoproteins in the presence of the high-abundance nonphosphorylated, nonregulatory proteins. Novel approaches to analyze the phosphoproteome were recently reviewed by Hjerrild et al.,2 such as approaches to identify phosphoproteins as well as approaches to identify specific phosphorylation sites. The latter often include enrichment of phosphopeptides before mass spectrometrical analysis. Phosphotyrosine containing peptides can be enriched by polyclonal antibodies,3 but due to the smaller size of phosphoserine and phosphothreonine groups, it has been more difficult to generate generic antibodies for these antigens. Instead, these phosphopeptides are usually enriched by methods like IMAC (immobilized metal affinity chromatography) or TiO2 affinity chromatography, prior to MS analysis.4-6 Other approaches include specific removal of the phospho groups 10.1021/pr8010809 CCC: $40.75
2009 American Chemical Society
Serine and Threonine Phosphorylation in Mast Cells by, for example, β-elimination followed by replacing the phospho group with a specific marker.2 We have previously studied the phenomenon of glycosylation regulation in the small intestine,7-9 where the regulation of glycosyltransferases during a helminth infection (Nippostrongylus brasilienis) has a profound affect on the glycosylation of the intestinal mucosa. Our working hypothesis is that this alteration of glycosyltransferase expression in the mucin producing goblet cells is controlled by molecular mediators secreted by intestinal immune cells. Indeed, intestinal mast cells increase their expression of cytokines during N. brasiliensis infection in rats,10 in conjunction with an increased IgE production in the mesenteric lymph node cells.11 Mast cells are important immune cells in innate immunity and allergic inflammation, and they are found at sites likely to encounter antigenic molecules, like the skin and mucosal surfaces.12,13 The mast cell membrane contains very high levels of IgE receptors,14 and the binding of antibodies to these receptors sensitizes the cell. Antigen induced activation of sensitized cells sets in motion a series of intracellular signaling events that result, within minutes, in the release of granule-bound inflammatory mediators such as histamine, tryptases, chymases and chemotactic factors.15,16 In addition, lipid mediators such as leukotrienes and prostaglandins are formed and released, and the synthesis of a large variety of cytokines and chemokines such as IL3, IL5 and TNF-R is initiated.13 Much of our understanding of the intracellular signaling events involved in the release of inflammatory mediators from mast cells has come from studies of the rat basophilic leukemia cell line, RBL-2H3.17 It is evident that one of the major processes that regulates the secretion is the transient phosphorylation of proteins at specific sites, catalyzed by protein kinases and protein phosphatases. Recently, Cao et al. identified substrates of tyrosine kinases in mast cells.3 However, although many of the serine and threonine kinases involved in mast cell degranulation have been identified, including protein kinase C (PKC)18,19 and Akt,20 the substrates of these serine and threonine kinases remain unexplored. The mapping of such substrates in mast cells is a challenging task, but would make a valuable contribution to the understanding of mast cells and allergic responses, as well as the regulation of intestinal defense against pathogens and the control of glycosylation and protein secretion in goblet cells. This study aimed to explore the use of generic antibodies against phosphoserine and phosphothreonine epitopes for phosphoproteomic studies of mast cell regulation. We mixed a variety of antibodies raised against phosphorylated substrate motives of different serine- and threonine kinases, and identified novel biological aspects of the mast cell degranulation process. We found that the antibody mixture was useful not only for Western blot applications, but also for immunoprecipitation to identify phosphoproteins and phosphorylation sites, and we gained novel insights into mast cell control of protein synthesis regulation during activation.
Materials and Methods Cell Culture, Activation, and Lysis. The RBL-2H3 mast cells were maintained as an adherent monolayer in RPMI 1640 media supplemented with 10% (v/v) fetal bovine serum, 2 mM glutamine, 25 mM HEPES, pH 7.2-7.5, 100 U/mL penicillin and 100 µg/mL streptomycin at 37 °C in a humidified 5% CO2 incubator. The cells were maintained in 75 cm2 vented tissue culture flasks and passaged every 2-3 days, using trypsin
research articles digestion, to a maximum of 25 passages. For antigen activation, cells were primed overnight with 75 ng/mL DNP specific IgE (Sigma-Aldrich, St Louis, MO) in culture media. The cells were then washed three times with activation buffer (119 mM NaCl, 5 mM KCl, 25 mM PIPES (free acid), 5.6 mM dextrose, 0.4 mM MgCl2, pH 7.25) supplemented with 1 mM CaCl2 immediately prior to use. Cells were equilibrated at 37 °C in a water bath, and 100 ng/mL DNP-BSA (Calbiochem, Darmstadt, Germany) was added to the activation buffer to cross-link the FcεRI high affinity IgE receptors. Control cells were incubated with supplemented activation buffer alone for 10 min. After 1-10 min of activation, reactions were stopped by placing the cells on ice. An aliquot of the supernatant was collected for a secretion assay. The remaining supernatant was removed and PSS lysis buffer (100 mM NaCl, 50 mM Na-pyrophosphate, 50 mM NaF, 2 mM Na-orthovanadate, 20 mM HEPES, 5 mM EDTA, 10 mM EGTA, 1% CHAPS, 1% NP40, 2 µg/mL PUGNAc, and 0.1% protease inhibitor cocktail) was added. While kept on ice, the cells were scraped off and collected. After 5-10 min incubation on ice, the samples were sonicated four times to ensure complete lysis. Cell lysates were stored at -20 °C. The amount of β-hexosaminidase released into the media was assayed using an absorbance assay with p-nitrophenyl-Nacetyl-b-D-glucosamide (Sigma-Aldrich) as the substrate.18 Phosphoprotein Enrichment. Cleared cell lysate (2.5 mg) was subjected to buffer exchange on a PD10 column (GE Healthcare Life Sciences, Uppsala, Sweden) equilibrated in TBS with 0.25% CHAPS, and eluted in 3.5 mL of the lysis buffer supplied in the Phosphoprotein Purification Kit from QIAGEN (Venlo, Netherlands), with 0.25% CHAPS. The detergent added to the equilibration buffer was necessary to prevent protein precipitation on the column. Phosphoproteins were then purified by affinity chromatography using the Phosphoprotein Purification Kit according to the manufacturer’s instructions. Eluted fractions were pooled and concentrated on Amicon Ultra-4 or Ultrafree-MC Centrifugal Filter Units (MWCO 5 kDa; Millipore, Billerica, MA). SDS-PAGE and Western Blotting. Proteins were separated on SDS-polyacrylamide gels (Proteome Systems, Boston, MA), and then either silver or Coomassie Blue stained according to standard protocols, or transferred to Immobilon-P membranes (Millipore) at 300 mA and +10 °C for 30 min on an ElectrophoretIQ3 (Proteome Systems Ltd., Sydney, Australia) instrument. The membranes were stained with Direct Blue 71 for total protein staining, then wetted in methanol and blocked with 1% bovine serum albumin in TBS-Tween. Membranes were incubated at +4 °C overnight with primary antibody (Table 1) in blocking solution, washed, incubated at room temperature for 1 h with horseradish peroxidase (HRP)conjugated secondary antibody in blocking solution, washed repeatedly in TBS-Tween, and developed with the Western Lightning (Perkin-Elmer, Waltham, MA) chemiluminescence reagent. Western blot signal intensities in the time course study were quantified using the Multi Gauge v3.0 image analysis software (Fujifilm Life Science, Tokyo, Japan). The signal intensity of each protein band was normalized, first against the p31 band in the corresponding lane, and subsequently against the 1 min data point of the same protein. Protein Identification. Gel plugs were excised from 1D-SDSpolyacrylamide gels, washed twice in 50 mM NH4HCO3/50% acetonitrile, and proteolytically digested with trypsin (Promega Corp, Sydney, Australia) in 50 mM NH4HCO3 at +34 °C Journal of Proteome Research • Vol. 8, No. 6, 2009 3069
research articles
Olson et al.
Table 1. Antibodies Used in This Study antibody
short name a
Phospho-(Ser) PKC Substrate Antibody Phospho-(Ser/Thr) PKA Substrate Antibodya Phospho-(Ser/Thr) Akt Substrate Antibodya Phosphothreonine Antibody Phospho-Threonine Antibody (P-Thr-Polyclonal) Anti-myosin heavy chain Phospho-eEF2 (Thr57) Antibody HRP/Goat anti-rabbit Ig
PKC PKA Akt Zymed CSPT Anti-myosin Anti-eEF2 -
dilution used
1:1000 1:1500 1:1500 1:1000 1:3000 1:2000 1:1000 1:10000
antibody type
Rabbit Ig Rabbit Ig Rabbit Ig Rabbit Ig Rabbit Ig Rabbit Ig Rabbit Ig Goat Ig
manufacturer
cat. no. b
Cell Signaling Cell Signaling Cell Signaling Zymedc Cell Signaling BTId Cell Signaling DAKOe
2261 9621 9611 71-8200 9381 BT-564 2331 P0448
a The antibody was used in the phosphoantibody mixture, using the same dilutions. b Cell Signaling Technology, Inc. (Danvers, MA). Laboratories, Inc. (South San Francisco, CA). d Biomedical Technologies, Inc. (Stoughton, MA). e DAKO (Glostrup, Denmark).
overnight. Peptides were extracted from the gel plug by sonication in 50 mM NH4HCO3 and analyzed by LC-ESI-MS/ MS using an LCQ DECA system (Thermo Electron Corp., Waltham, MA) as described previously,21 with protein identities assigned by SEQUEST (Thermo Electron) and MASCOT (Matrix Science, Boston, MA) searching in an in-house database. The database was created from Swiss-Prot/TrEMBL and NCBI protein entries, and comprised all protein sequence information from Rattus sp. The criteria for valid peptide assignments were an xCorr score over 2.50 and a ∆CN value over 0.10 in SEQUEST, and a Mowse score over 40 in MASCOT. Immunoprecipitation. Phosphoprotein purified and concentrated samples were reduced with 2 mM dithiothreitol and alkylated with 50 mM iodoacetamide. The accessibility of the phosphorylation site for the antibodies was previously found to increase after reduction and alkylation prior to immunoprecipitation.22 The resulting sample (15 µL; corresponding to the phosphoprotein purified fraction from 250 µg of total cell lysate) was diluted with 8 vol of loading buffer (1.7% thesit, 150 mM NaCl, and 50 mM HEPES, pH 7.5). Magnetic Protein G Dynabeads (10 µL of 50% suspension; Invitrogen, Carlsbad, CA) were preincubated with a mixture of the Akt (6 µL), PKA (6 µL) and PKC (9 µL) antibodies at +4 °C overnight. After washing the beads in 100 µL of loading buffer, the samples were added and incubated on a spinning wheel at room temperature for 3 h. The nonbound fraction was collected, the beads were washed three times with loading buffer, and bound proteins were released from the beads with 2× 15 µL of 50 mM glycine, pH 2.8.
c
Zymed
Figure 1. Effect of phosphatase inhibitors on cell lysates. Total cell lysates prepared from activated RBL-2H3 cells in the presence or absence of phosphatase inhibitors (PI) were separated by SDSPAGE (4-20%) and Western blotted to PVDF membrane. Phosphoproteins were detected by immunostaining with the phosphoantibody mixture (Akt, PKC and PKA) presented in this paper.
Results Mast Cell Activation/Phosphorylation. RBL-2H3 cells (model mast cells) were activated via FcεRI cross-linking in the presence of 1 mM CaCl2. The maximal secretion rate was observed after 2-2.5 min, in accordance with previous findings for this cell type.23,24 Accordingly, 2.5 min of activation was chosen to represent a highly activated state. Phosphorylation/ dephosphorylation is a dynamic event, and information about natural kinase substrates could be lost due to the action of phosphatases. While it has been suggested that the use of phosphatase inhibitors in Phosphoproteomics could provide a distorted picture of the phosphorylation,5 we found that addition of phosphatase inhibitors to the lysis buffer was necessary in order to preserve intact phosphorylated proteins (Figure 1). To minimize the effect of remaining active kinases on non-natural targets after lysis, we ensured that the cells were efficiently lysed in large excess of cold lysis buffer with inclusion of detergents, and the lysates subsequently stored in -20 °C. Phosphoproteins in Activated Mast Cells. Generic phosphoaffinity chromatography was used to enrich phosphoproteins 3070
Journal of Proteome Research • Vol. 8, No. 6, 2009
from the cell lysates. The phosphoprotein fraction was compared to the total lysate fraction and the flow-through fraction by SDS-PAGE (Figure 2A). The silver stained gel showed a different protein composition and, as expected, a considerably lower protein amount in the phosphoprotein fraction compared with the total lysate, indicating that the enrichment method was selective. Western blotting with phosphorylationspecific antibodies confirmed that phosphoproteins were retained on the column and eluted in the phosphoprotein fraction (Figure 2B). While enrichment of phosphoproteins was achieved, it was also evident that a considerable amount of phosphoprotein was lost in the flow through fraction. The antibody reactivity was only slightly stronger in the enriched sample, despite it being 20 times more concentrated compared with the total lysate. Preparative gel electrophoresis was performed in order to identify proteins in the phosphoprotein enriched fraction. Thirty-six high-abundance proteins were identified after excision of protein bands detected by Coomassie Blue staining from
Serine and Threonine Phosphorylation in Mast Cells
Figure 2. Phosphoprotein enrichment by phospho-affinity chromatography. SDS-PAGE (4-20%) of total lysate from activated RBL-2H3 cells (TL), and flow-through (FT) and phosphoprotein enriched (Ph) fractions after phospho-affinity chromatography on activated mast cell lysates. The gels were (A) silver stained for total protein visualization, or (B) Western blotted to PVDF membranes and immunostained with the phosphoantibody mixture (Akt, PKC and PKA) presented in this paper. Concentration factors of flow-through and phosphoprotein enriched fractions are given above the panels.
1D SDS-PAGE gels, while an additional 40 low-abundance proteins were identified after cutting the entire lane into 0.5 cm long gel pieces and analyzing each piece (Supporting Information Table 1). The UniProtKB database (http://www.uniprot.org) reports phosphorylation sites in 61 of these 76 proteins, indicating that enrichment of phosphoproteins had been accomplished. Identification of phosphoproteins alone gives little or no information about proteins that are phospho-regulated upon cell activation, since constantly phosphorylated and housekeeping phosphoproteins will be among the identified proteins. Applying the enrichment step on not only the activated mast cells, but also on nonactivated controls, and searching for quantitative differences in the resulting samples in order to identify regulated phosphoproteins would be tempting. However, with the loss of quantitative information implicated by the enrichment step, this approach turned out to be unsuccessful. Further investigations were needed to focus on phosphothreonine and phosphoserine proteins that were specifically regulated during mast cell activation. We therefore explored the use of phosphothreonine and phosphoserine specific antibodies in order to further investigate the mast cell activation. Western Blotting with Phosphoserine and Phosphothreonine Antibodies. With the aim to detect proteins that become phosphorylated specifically during mast cell activation, a number of phosphoserine and phosphothreonine reactive antibodies (Table 1) were screened for their ability to highlight phosphorylation changes in the mast cell activation model. Cell lysates from control and activated cells were subjected to SDSPAGE and Western blotting, and incubated with the different antibodies (Figure 3). All of the antibodies reacted with a range of phosphoproteins. From this data, it is obvious that the more generic antibodies against phosphothreoninesCSPT and
research articles
Figure 3. Screening of phosphoantibodies. Total cell lysates prepared from activated (a) or control (c) RBL-2H3 cells were separated by SDS-PAGE (3-8%) and Western blotted to PVDF membrane. (A) Immunostaining with five different phosphoprotein antibodies. The PKC, PKA and Akt antibodies detected several proteins that were phosphorylated upon cell activation (indicated with crosses), while the CSPT and Zymed antibodies detected phosphoproteins not regulated in the early stage of cell activation. (B) The same blot stained with Direct Blue 71 for total protein visualization, prior to antibody incubation.
Zymedsreacted mainly with proteins that did not change in their phosphorylation upon cell activation. This was in agreement with the common hypothesis that only a small fraction of the proteins in the cells are expected to be phosphoregulated during cell activation. However, the remaining three antibodiessproduced from consensus sequences of phosphorylated peptides generated by Akt-, PKA-, and PKC-kinasessall reacted with a number of proteins that were phosphorylated upon activation. Furthermore, all three antibodies detected unique phosphoproteins. Thus, these antibodies would all contribute to identifying proteins undergoing serine or threonine phosphorylation upon mast cell activation. Interestingly, both PKC and Akt are known phosphoserine and phosphothreonine kinases activated by the FcεRI mediated IgE stimulation of mast cells,20 and seven of the detected phosphoproteins were a direct consequence of the activation of these kinases. In addition, three phosphoproteins were detected with the antibody directed toward PKA substrates. PKA is involved in cyclic AMP signaling in mast cells.25,26 The fact that many kinases show identical or overlapping sequence specificity (http://www.hprd.org) indicates that any prediction of activation of specific kinases using these antibodies would have to be validated by other methods. To generate a simple method to identify proteins phosphorylated during mast cell activation, the PKC, PKA and Akt Journal of Proteome Research • Vol. 8, No. 6, 2009 3071
research articles
Olson et al.
Figure 4. Phosphoantibody mixture for detection of regulated phosphoproteins. Total cell lysates prepared from activated (a) or control (c) RBL-2H3 cells were separated by SDS-PAGE (4-20%) and Western blotted to PVDF membrane. PKC, PKA and Akt antibodies were mixed and used for immunostaining of the blot. This mixture of antibodies selectively detected proteins that were phosphorylated upon cell activation, with a 31 kDa protein as an exception.
antibodies were mixed and used simultaneously in a Western blotting experiment. The dilution of each antibody was optimized to give a good signal relative to the others. The mixture was used to compare activated and control mast cells (Figure 4). A few phosphorylated proteins were detected in the control sample, but the number detected in the activated sample was remarkably higher. Thus, mixing these three antibodies was a good approach to address phosphorylation changes in the mast cell model, and a useful tool for the further identification of phosphoproteins. Mast Cell Phosphorylation/Dephosphorylation. To follow the phosphorylation events over time in the activated mast cells, cell lysates were prepared at different time points, from 1 to 10 min after induction. Lysates were subjected to SDSPAGE, blotted and incubated with the phosphoantibody mixture. Figure 5A shows the results from two independent activation experiments. Already 1 min after activation, phosphorylation of many proteins was observed. During the timecourse, the strongest phospho-reactivity was observed in the samples collected after 1 and 2.5 min, validating that the 2.5 min time-point of activation provided a good reflection of an activated mast cell. Beyond 2.5 min, the degree of phosphorylation of most proteins declined. However, the rate of decline differed between different phosphoproteins, and semiquantification of the staining intensities revealed three general patterns of regulation (Figure 5B-D). Some proteins were still heavily phosphorylated (more than 60% of the 1 min signal intensity) after 10 min (Figure 5B). Another group of proteins maintained their degree of phosphorylation between 1 and 5 min, before becoming dephosphorylated (Figure 5C). Finally, one protein at around 60 kDa was highly phosphorylated at 1 min, and then im3072
Journal of Proteome Research • Vol. 8, No. 6, 2009
Figure 5. Phosphorylation during the first 10 min of mast cell activation. RBL-2H3 cells were activated by DNP-BSA and incubated in activation buffer for between 1 and 10 min (1, 2.5, 5, and 10 min) before being lysed. Control RBL-2H3 cells (c) were incubated in activation buffer in the absence of DNP-BSA for 10 min before being lysed. (A) Total cell lysates were separated by SDS-PAGE (4-20%), Western blotted to PVDF membranes, and immunostained with the phosphoantibody mixture (PKC, PKA and Akt) for detection of regulated phosphoproteins. Two independent experiments are presented. Annotation of detected proteins with the estimated molecular weight is presented between the blots. Quantification of the protein staining demonstrated different patterns of regulation, in protein groups with >60% (B), 30-60% (C), or
