Signal Transduction Reaction Monitoring ... - ACS Publications

Article pubs.acs.org/jpr

Signal Transduction Reaction Monitoring Deciphers Site-Specific PI3K-mTOR/MAPK Pathway Dynamics in Oncogene-Induced Senescence Erik L. de Graaf,†,‡,∥,# Joanna Kaplon,§,# Shabaz Mohammed,†,‡,⊥ Lisette A. M. Vereijken,†,‡ Daniel P. Duarte,§ Laura Redondo Gallego,†,‡ Albert J. R. Heck,*,†,‡ Daniel S. Peeper,§ and A. F. Maarten Altelaar*,†,‡ †

Biomolecular Mass Spectrometry and Proteomics, Utrecht Institute for Pharmaceutical Sciences and Bijvoet Center for Biomolecular Research, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands ‡ Netherlands Proteomics Centre, Padualaan 8, 3584 CH Utrecht, The Netherlands § Division of Molecular Oncology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands S Supporting Information *

ABSTRACT: We report a straightforward strategy to comprehensively monitor signal transduction pathway dynamics in mammalian systems. Combining targeted quantitative proteomics with highly selective phosphopeptide enrichment, we monitor, with great sensitivity, phosphorylation dynamics of the PI3KmTOR and MAPK signaling networks. Our approach consists of a single enrichment step followed by a single targeted proteomics experiment, circumventing the need for labeling and immune purification while enabling analysis of selected phosphorylation nodes throughout signaling pathways. The need for such a comprehensive pathway analysis is illustrated by highlighting previously uncharacterized phosphorylation changes in oncogeneinduced senescence, associated with diverse biological phenotypes and pharmacological intervention of the PI3K-mTOR pathway. KEYWORDS: phosphoproteomics, ERK1/2, RPS6, MAPK, mTOR, PI3K, pathway, network biology, signal transduction, Ti4+-IMAC, SRM

■

INTRODUCTION Alterations in cellular signaling networks are the cause of many diseases and determine highly variable and microenvironmentdependent drug target potency.1 Therefore, understanding the response of cellular signaling networks to perturbations is of great clinical and biomedical interest. Signal transduction is still mostly studied by immunoblot analysis using phosphositespecific antibodies because of its great sensitivity and ease of use; however, the limited availability of these reagents and the low throughput of immunoblot analysis severely hinders the analysis of complete signaling networks. Other limiting factors of antibody-based protein phosphorylation assays are the semiquantitative and often ambiguous nature of obtained results. Phosphosites resident in highly conserved sequences are frequently undistinguishable between closely related protein isoforms, with ERK1 Y204 and ERK2 Y187 serving as prime examples. Moreover, close-proximity sites such as T202/Y204 on ERK1, T185/Y187 on ERK2, or S235/S236 and S240/S244 on RPS6, are indistinguishable by antibody binding. In an attempt to address these problems, dual site-specific antibodies have been developed that recognize multiple isoforms and © 2015 American Chemical Society

multiply phosphorylated domains; however, this approach does not allow for monitoring of possible mutually exclusive phosphorylation patterns and other cross-talk that may occur.2 The current lack of antibodies that are specific for close-proximity or protein isoform phosphosites has left the dynamics of these potentially important phosphosite differences largely unexplored, necessitating the development of methods with increased resolving power. Global shotgun mass spectrometry (MS)-based phosphoproteomics can address many of the issues previously described and allows for the analysis of thousands of phosphorylation events with high specificity for protein isoform and phosphosite localization.3 Currently, the main challenge of low phosphoprotein stoichiometry can be partially solved by separating phosphorylated peptides from the more abundant unphosphorylated peptides, using affinity-based techniques such as TiO24 or Fe3+/Ti4+-IMAC.5,6 Nonetheless, the large dynamic Received: March 18, 2015 Published: May 26, 2015 2906

DOI: 10.1021/acs.jproteome.5b00236 J. Proteome Res. 2015, 14, 2906−2914

Article

Journal of Proteome Research

Figure 1. Signal transduction reaction monitoring workflow. Biological triplicates of control (cycling), oncogene-induced (OIS), and OIS bypass (OISb) human fibroblast cells were grown. After lysis and digestion, two concentrations (high and low) of stable isotope standard (SIS) phosphopeptides for each phosphosite were spiked-in two separate aliquots to achieve similar levels of endogenous and SIS peptides. Subsequently, phosphopeptides were enriched from 200 μg of input material using single-stage Ti4+-IMAC, followed by a scheduled 2 h LC−SRM run for each sample.

biological relevant phosphorylation sites throughout the MAPK and PI3K-mTOR pathways to perform accurate targeted analysis of their dynamics.

range in the remaining phosphopeptide population greatly hampers global shotgun proteomics approaches. Detection reproducibility and sensitivity can be optimized using targeted MS, referred to as selected or multiple reaction monitoring (SRM/MRM). Moreover, when combined with stable isotope standard (SIS) phosphopeptides, technical variation can be reduced, resulting in increased accuracy in quantification.7 Although SRM has been established for monitoring protein abundance,8,9 its applicability to monitor selected protein phosphorylation events is still limited. The first in-depth analysis of phosphopeptides by SRM was pioneered by Wolf-Yadlin et al.,10 where they targeted global tyrosine phosphorylation dynamics upon EGFR stimulation, requiring iTRAQ peptide labeling and phosphotyrosine peptide immunepurification. Following this study, mainly low-throughput targeted phosphosite experiments have been described, analyzing a small number of phosphosites from high abundant proteins from total cell lysates7,11 or enriched for specific proteins11−13 or protein complexes.14 A recent approach by Soste et al.15 describes the targeted analysis of so-called sentinel phosphopeptides by SRM to monitor activation of cellular responses in yeast. Creating SRM assays for single LC−MS runs is still challenging, as they can be less sensitive than immunoblot analysis, hampering targeted analysis in highly complex backgrounds. Moreover, in the case of phosphopeptides, the phosphosite has to be accessible by proteolytic cleavage, producing MS responsive peptides. To increase the success rate of the phosphopeptide SRM assay development in this study, we focused on phosphopeptides that have been reported in our own and publicly available shotgun proteomics data sets. Moreover, the incorporation of a single enrichment step greatly increased sensitivity by removing background interference, thereby unlocking the targeted analysis of peptides that would otherwise be masked by the complex background. Building upon previous work, we developed an optimized method to achieve precise and reproducible monitoring of phosphorylation dynamics in specific signaling cascades, probing multiple signaling nodes throughout the pathway, while circumventing the need for large sample amounts, peptide labeling, specific antibodies or extensive sample preparation. The reported approach allowed us to select

■

RESULTS AND DISCUSSION Using our method we comprehensively quantify phosphorylation events in both the MAPK and PI3K-mTOR signal transduction pathways. The success of this approach is determined by the combination of the sensitive and quantitatively reproducible Ti4+-IMAC enrichment strategy6,16 with targeted SRM mass spectrometry and using low-cost crude synthetic SIS phosphopeptides at two different concentrations (Figure 1). The Ti4+-IMAC pre-enrichment step allows for a fast removal of interfering nonphosphorylated peptides as well as removal of impurities from the crude SIS phosphopeptide mixture. The removal of interfering background by affinity or immune enrichment greatly increases the overall sensitivity of SRM, allowing the targeted analysis of peptides and proteins from highly complex samples, routinely not accessible.17 Samples were mixed with two different amounts of SIS peptide allowing for intensities of SIS and NAT (natural endogenous) versions to differ less than 10-fold, ensuring accurate and reproducible quantification, to prevent quantification problems regarding upper and lower limits of detection and signal response linearity. Finally, to measure the selected phosphosites, we constructed robust interference-free SRM assays for the samples studied here, using reproducible retention times and coeluting SIS reference peptides for relative fragment ion intensity comparison together with SRM-triggered MS/MS for identification and phosphosite localization (Supplementary Tables 1 and 2 in the Supporting Information (SI)). The SRM-based method was applied to characterize the dynamics of the MAPK and PI3K-mTOR signal transduction pathways in oncogene-induced senescence (OIS).18,19 OIS is a largely irreversible state of cell cycle arrest, which can be triggered by the unscheduled activation of oncogenes.20 It has been shown to act alongside death programs to suppress tumorigenesis.19 For this purpose, “control” cycling cells (Cycl), OIS cells (upon introduction of BRAFV600E, a common oncogene and strong inducer of OIS), and OIS-bypassing cells (OISb),21 were analyzed in biological triplicates. Additionally, 2907


Article


Figure 2. Phosphosite dynamics of several nodes in the MAPK and PI3K-mTOR analyzed by immunoblot and SRM analysis. (A) Immunoblot analysis on three biological replicates probed with antibodies for ERK1/2 and phosphosites T202/Y204 and Y185/Y187, p70S6K and phosphosite T412, 4EBP1 and phosphosite S65, and RPS6 and phosphosites S235/236 and S240/244. (B) SRM analysis was performed on three biological replicates, with one or two SIS spike-in concentrations for each replicate. (See Supplementary Table 1 in the SI.) Significant regulations were highlighted with an asterisk comparing average signals (two-sided t test, p < 0.05, s.e.m. error bars). Using SRM, protein isoform phosphorylation domains as well as close-proximity phosphosites could be dissected, untraceable by immunoblot analysis.

allowed for the analysis of the single tyrosine sites (Y204/ Y187), of which phosphorylation is known to precede that of the threonine residues (T202/T185), showing lower levels of single tyrosine site phosphorylation on both ERK1 Y204 and ERK2 Y187 in OIS cells compared with cycling cells (Supplementary Figure 3 in the SI). This lower phosphorylation level might be a direct consequence of the increase observed for the dual phosphorylation in OIS. We also observed a similar trend of phosphorylation on RRAS2 and MEK2 phosphorylation that are both MAPK pathway nodes upstream of ERK1/2. Interestingly, downregulation of MEK2 T394 phosphorylation in OIS versus cycling cells is observed, which has been described to be phosphorylated by ERK1/2 in a negative feedback loop, reducing MEK and MAPK pathway activity.24 Another major proliferation-controlling pathway is mainly controlled by PI3K and mTOR. Activation of PI3K-mTOR triggers protein synthesis via the phosphorylation of 4E-binding protein 1 (4EBP1) and the ribosomal protein S6 kinase (p70S6K). Indeed, in our results, both immunoblot and SRM assays confirmed a strong mTOR-dependent regulation of 4EBP1 S65, as evidenced by the total loss of S65 phosphorylation upon treatment with the dual PI3K/mTOR inhibitor, BEZ235 (Figure 2B, Supplementary Figure 4 in the SI). Interestingly, a strong reduction of growth-stimulating 4EBP1 phosphorylation was observed in OIS and OISb, albeit less pronounced in the latter. A similar trend in signaling perturbation was observed for p70S6K phosphorylation on T412 (often referred to as T389) and S452, however, not for S441 and S447 (Figures 2 and 3A). Phosphorylation of p70S6K T412 is often used as a read-out for mTOR activity in immunoassays. This site is not accessible by SRM after tryptic

Cycl, OIS, and OISb cells treated with BEZ235 (a dual PI3K/ mTOR inhibitor), were included to screen for mTOR-sensitive perturbations. On the basis of our previous inquiries on phosphorylation dynamics in OIS22 we selected the MAPK and mTOR pathways for targeted analysis and identified the most important nodes functionally described in literature and UniProt. We used both our own shotgun proteomics data as well as MS data from a public repository (Phosphositeplus23) to identify previously detected phosphopeptides belonging to the MAPK or mTOR pathway. The chosen sites were restricted to phosphopeptides observed after tryptic digestion, as both our own shotgun data and the majority of reported phosphosites in the database are obtained from trypsin-based proteomics experiments. This results in several functional sites in these pathways to be absent, and these should be targeted with alternative proteolytic enzymes. In total, 306 phosphorylation events in 51 phospho-patterns from 27 different signaling proteins were reproducibly quantified in 18 samples (Supplementary Table 2 in the SI). MAPK and PI3K-mTOR Phosphorylation Dynamics

Several nodes in the MAPK pathway were expected to show changes in phosphorylation, as both OIS and OISb cells overexpress the constitutively active BRAFV600E mutant. Indeed, under each condition we observed a specific upregulation of the downstream ERK1/2 TEY dual phosphorylation motif, as seen in both immunoblot and SRM analyses. In addition, we detected a strong activating BRAF S729 phosphorylation under these conditions in the SRM measurements (Figure 2B). Comparing the two ERK isoforms, similar trends in phosphorylation were observed in both immunoblot and SRM analyses, indicating a level of redundancy between ERK1/2 (Figure 2, Supplementary Figure 4 in the SI). SRM 2908


Article


Figure 3. Regulations in the PI3K-mTOR and MAPK pathway measured by SRM. (A) Phosphosites specifically regulated in OIS. (B) Phosphosites sensitive to PI3K-mTOR inhibition by BEZ235 treatment. Significantly up- or down-regulated phosphosites (three biological replicates, two-sided t test, p < 0.05) are color-coded in green and red circles, respectively. Protein interaction and kinase substrate relationships were extracted from the UniProtKB and phosphositeplus databases and manually curated using literature.

detecting only dual phosphorylation on S235/S236 or S240/ S244. As this highly conserved and confined domain is functionally important and phosphorylated by two different kinases from two pathways, phosphorylation patterns are most likely more complex than merely two separate dual phosphorylations. Therefore, this confined multiply phosphorylated RPS6 domain poses an interesting subject to investigate further with SRM. By using antibodies to detect RPS6 S235/S236 and S240/ S244 phosphorylation, no substantial differences in phosphorylation were observed between cycling and OIS and OISb cells (Figure 2A, Supplementary Figure 4 in the SI). Treatment with BEZ235 revealed a strong dependence of these sites on mTOR activity in cycling cells and an altered response in OIS and OISb, showing only partial loss of phosphorylation in OIS and to a lesser extent in OISb. Exploiting the specificity of the SRM approach, we were able to dissect different patterns for S235, S236, and S240 phosphorylation, associated with different biological conditions (Figure 2B). For example, a strong mTOR-independent induction of single S236 phosphorylation was observed in OIS. When interpreting the results obtained by the S235/S236 antibody, one might falsely conclude that the phosphorylation of both sites is mainly dependent on mTOR; however, SRM revealed that S236 is strictly phosphorylated by

digestion, however, could be targeted using for example Glu-C. Conversely, for p70S6K S452, no site-specific antibody is available and no responsible kinase or biological role has been reported. By SRM analysis, we show an identical regulation of S452 to T412, including a total loss of phosphorylation upon BEZ235 treatment, indicating S452 as a novel putative PI3KmTOR phosphorylation site. These results highlight the superior resolving power of our method, revealing quantitative differences in TEY-motif phosphorylation on different ERK1/2 isoforms as well as monitoring biologically interesting sites that would remain undetected using classical approaches. SRM Reveals Differential Phosphorylation Patterns on RPS6

Ribosomal protein S6 (RPS6), a well-studied downstream target of both the PI3K-mTOR and MAPK pathway, functions to integrate proliferation promoting signals from both pathways on a single protein. Phosphorylated RPS6 is increased in many types of cancer and is proposed to be a marker for drug resistance.25,26 RPS6 is primarily known as a downstream mTOR target and is phosphorylated on S235, S236, S240, and S244 by the mTOR conveyor p70S6K; however, RSKs (RSK1/ 2) from the MAPK pathway can phosphorylate S235 and S236 as well.27 In all studies on RPS6 signaling, antibodies were used 2909


Article


The mutant form of BRAF shows a 4-fold upregulation in OIS over cycling cells, while the phosphosite-specific regulation is over 20-fold. An example of observed phosphorylation dynamics instigated by altered protein expression is the protein RRAS2, where protein expression and S186 phosphorylation dynamics are on par. In the case of RPS6, protein expression levels do not change between the different conditions, while its phosphorylation status is highly dynamic, revealing an active status of its upstream kinases. It has been suggested before to normalize phosphorylation quantification using protein expression changes,32 which in the case of TSC1, by SRM showing only a slight decrease in phosphorylation in OIS and even less in OISb, would lead to a more dramatic downregulation of phosphorylation, as protein expression is up almost 2-fold. Since, the biological significance of truly differential phosphorylation versus altered phosphorylation status caused by altered protein expression is not clear, we decided to report this information separately. Future investigations into pathway dynamics in biological systems, including samples from cell lines, primary cells or tissues, can directly use the PI3K-mTOR pathway assays and method developed herein. An automated protocol using robotic systems could be explored in the future, to further aid the implementation of SRM analysis as a standard tool for molecular biologists to analyze signal transduction pathways. A restriction of the current method is the focus on phosphosites observed on tryptic peptides, which excludes several sites in the PI3K and mTOR pathways for analysis; however, this restriction can be tackled by using alternative proteolytic enzymes to allow targeting of a broader spectrum of phosphosites.33 In conclusion, we report here a strategy to comprehensively monitor pathway signal transduction dynamics in a site-specific manner. Precise and phosphosite-specific quantification with increased throughput is demonstrated, allowing the characterization of complex protein phosphorylation patterns associated with a particular biological phenotype. The need for such a comprehensive full pathway analysis method is illustrated by highlighting previously unknown and undetectable changes in protein phosphorylation in OIS both before and after pharmacological inhibition of mTOR.

RSK1/2 in OIS. This is also in line with an increase in RSK1/2 phosphorylation observed in OIS (Supplementary Figure 3 in the SI). SRM analysis on single S240 phosphorylation showed comparable results to dual S240/S244 immunoblot analysis, that is, no difference between control, OIS, and OISb cells before BEZ235 treatment and significant differences between the three cell types when treated with BEZ235. Interestingly, the phosphorylation of the dual (S236/S240) and the triply phosphorylated domain (S235/S236/S240) were strongly (albeit not exclusively) dependent on PI3K-mTOR. In addition, in samples without BEZ235 treatment, triple phosphorylation showed a significant downregulation in OIS compared to cycling and OISb, identical to S235/S240 dual phosphorylation (Supplementary Figure 3 in the SI). The complex differential phosphorylation patterns on RPS6, which was revealed by SRM analysis only, indicate an important role of multiple signaling pathways regulating a confined domain on a single protein. In summary, SRM analysis revealed a strong and specific regulation of different mTOR substrates in cells that undergo OIS, which is associated with remarkably high phosphosite pattern differentiation, which has not been previously reported. Pathway Deep SRM Analysis

Besides obtaining a few specific phosphosite pattern quantifications, SRM also enables comprehensive monitoring of many nodes in a pathway using a single run per sample, illustrated here for the PI3K-mTOR and MAPK pathways (Figure 3A). For example, it has been reported previously that phosphorylation of PRAS40 at S183 and T246 by mTOR and AKT, respectively, represses its inhibitory function on mTORC1 signaling.28 Interestingly, our pathway analysis revealed that both S183 and T246 phosphorylation sites of PRAS40 are reduced in OIS, hinting at an important role for PRAS40 in the reduced mTORC1 activity associated with OIS. In line with this, PDK1, an upstream AKT-PRAS40 kinase, showed lower S241 phosphorylation in OIS, which has previously been associated with decreased PDK1 activity.29 Next, we systematically searched for pathway components specifically regulated by the PI3K-mTOR pathway (Figure 3B). Analysis of BEZ235-treated cells confirmed that phosphorylation of specific sites on p70S6K and 4EBP1 is PI3K-mTORdependent, whereas all measured sites on RSK1/2 are not. Other potential direct or indirect PI3K-mTOR substrates showing regulation upon BEZ235 treatment include TSC2 S1388 and S1411 and eIF4B S445. The increased sensitivity of phosphopeptide SRM through the single enrichment step precludes a direct comparison to protein expression changes, as unmodified counterparts are not measured. That said, the analysis of low abundant proteins in complex samples such as human cells, has proven to be challenging in single LC−SRM runs and require more elaborate approaches including enrichment, depletion, and fractionation.17,30,31 In a previous report, we performed extensive fractionation before high-resolution LC−MS/MS to quantify the unmodified proteins in the same model system,22 and as a reference we have included these results when available for the proteins studied in this report (Supplementary Table 2 in the SI). From these data, we can extract additional information on the regulation of several of the phosphosites. Most of the observed proteins show little to no expression regulation between the different conditions, indicating kinase activity being responsible for the observed dynamics in our SRM data.

■

MATERIALS AND METHODS

Cell Culture

The human diploid fibroblast (HDF) cell line Tig3 expressing the ecotropic receptor, hTERT, and sh-p16INK4A (Tig3 (et)16i) was derived from Tig3 cell line, a kind gift of Anno Kumiko, an assistant professor at Hiroshima University (Japan). Tig3 cell line was STR-profiled and tested negative for mycoplasma. Tig3 (et)-16i cell line was maintained in DMEM with 4.5 mg/mL glucose and 0.11 mg/mL sodium pyruvate, supplemented with 9% fetal bovine serum (PAA), 2 mM glutamine, 100 units/mL penicillin, and 0.1 mg/mL streptomycin (GIBCO). The Phoenix packaging cell line was used for the generation of ecotropic retroviruses. For infections, filtered (pore size 0.45 mm) viral supernatant, supplemented with 4−8 μg/mL polybrene, was used. In general, a single infection round of 6 h was sufficient to infect at least 90% of the population. For senescence experiments, cells were infected with shCEBP/β-encoding or control retrovirus, selected pharmacologically (puromycin) and subsequently infected with 2910


Article

Journal of Proteome Research BRAFV600E-encoding or control virus. After 9 days of selection (blastomycin), cells were collected for mass spectrometry (MS) and immunoblotting assays. For screening mTOR-sensitive perturbations, cells were pretreated with 500 nM mTOR inhibitor BEZ235 for 24 h. Results represent data from three independent experiments.

trations of 5 mM dithiothreitol and 10 mM iodoacetamide, respectively. The first enzymatic digestion step was performed with Lys-C at 37 °C for 4 h in lysis buffer (enzyme/substrate 1:75). For the second digestion, samples were diluted to 2 M urea using 50 mM ammonium bicarbonate and incubated overnight with trypsin (Promega, USA) at 37 °C (enzyme/ substrate 1:100). The resulting endogenous peptides were acidified with 10% formic acid (FA) and split into two aliquots of 500 μg peptides for each sample. Next, aliquots were spiked with either 10 or 0.1 pmol of heavy stable isotope standard (SIS) peptides. Crude synthetic SIS peptides containing one Cterminal heavy Lysine (6C13,2N15: +8 Da) or heavy Arginine (6C13, 4N15: +10 Da) amino acid, were purchased from JPT technologies (Berlin, Germany). The SIS stock was created by mixing equimolar quantities of all peptides in 50% acetonitrile (ACN), 1% FA and divided into 100 pmol aliquots stored at −20 °C. Prior to phosphopeptide enrichment, peptide mixtures were desalted on Sep-Pak C18 columns (Waters, Milford, MA), dried to completion in vacuum, and stored at −80 °C. Conventionally, highly purified synthetic peptides are used for high precision absolute quantification. In our novel approach, we used more inexpensive crude synthetic phosphopeptides for each site and in two different concentrations per sample to allow for high-throughput, sensitive and reproducible relative quantification.

Plasmids

The plasmids pMSCV-blast and pMSCV-blast-BRAFV600E as well as pRS-puro and pRS-puro-C/EBPβ#1 were previously described.19,21 Antibodies

Antibodies for immunoblotting were purchased from Cell Signaling Technology (Beverly, MA). Antibodies used include: phospho-p44/42 MAPK (ERK1/2) (T202,Y204/T185,Y187; clone E10; #9106), total p44/42 MAPK (ERK1/2) (#9102), phospho-p70S6 (T389/T412; clone 108D2; #9234), total p70S6 (#9202), phospho-4EBP1 (S65; clone 174A9; #9456), total 4EBP1 (#9452), phospho-RPS6 (S235/S236; #2211), phospho-RPS6 (S240/S244; #2215), and total RPS6 (clone 5G10; #2217). All antibodies were profiled for use in the immunoblot analysis of human cell line extracts. Validation profiles and multiple references are available on the Web site of the provider (http://www.cellsignal.com). Protein Extraction and Immunoblot Analysis

Phosphopeptide Enrichment

Cycling, OIS, or OISb cells were harvested 9 days after BRAFV600E infection. Cells were scraped with 1× PBS, centrifuged @ 4000 rpm for 4 min at 4 °C, and the pellets were frozen or used immediately. Fresh or frozen pellets were lysed on ice in RIPA buffer supplemented with protease inhibitor cocktail (Roche) and phosphatase inhibitors (10 mM βglycerophosphate, 2 mM sodium fluoride, 0.2 mM sodium orthovanadate, 1 mM sodium pyrophosphate). Lysates were sonicated for 1 min (5 s on/off interval) and centrifuged at 4 °C and 1200 rpm for 10 min, and supernatants were transferred to fresh Eppendorf tubes. Protein concentrations were determined using Bradford assay (Bio-Rad). Protein samples were prepared in 4× sample buffer (Invitrogen) supplemented with 2.5% β-mercaptoethanol. Proteins were separated on 4− 12% polyacrylamide gels (Invitrogen), transferred onto a nitrocellulose membrane (Whatman), and blocked in blocking buffer (4% milk in 1× TBS-Tween) for 1 h at room temperature. The membrane was probed with the indicated primary antibodies (overnight at 4 °C in 4% milk in 1× TBSTween), followed by 1 h of incubation with the corresponding secondary antibodies conjugated with horseradish peroxidase (HRP) enzyme. For detection of the signal, the membrane was incubated 1 min with ECL reagent (Amersham Biosciences) and visualized on films (GE Healthcare). The immunoblot analysis was performed on three biological replicates. Quantification of the depicted immunoblot TIFF images was performed by Quantify One software version 4.6.7 (BIO-RAD).

Two Ti4+-IMAC columns for each sample were prepared and processed in parallel using a microcentrifuge, as described previously.6 In brief, microcolumns were created by loading 500 μg Ti4+-IMAC beads onto GELoader tips (Eppendorf) with a C8 plug. Ti4+-IMAC columns were pre-equilibrated two times with 30 μL of loading buffer (80% ACN, 6% trifluoroacetic acid (TFA)). Next, all samples were reconstituted in 250 μL of loading buffer, and 100 μL was loaded onto two equilibrated Ti4+-IMAC columns, corresponding to 200 μg natural peptides (NAT) and 4 or 0.04 pmol SIS per column. The second column was used as a backup and was only analyzed when needed. Next, Ti4+-IMAC columns were washed with 60 μL of washing buffer A (50% ACN, 0.5% TFA, 200 mM NaCl) and subsequently with 40 μL of washing buffer B (50% ACN, 0.1% TFA). Bound peptides were eluted by 30 μL of 10% ammonia into 30 μL of 10% FA. Finally, the remainder of the peptides was eluted with 5 μL of (80% ACN, 2% FA). The collected eluate was further acidified by adding 20 μL of 10% TFA and subsequently desalted using homemade reversed phase tips loaded with 10 μL of Aqua C18 5 μm beads. Tips were washed twice with 30 μL of 0.1% TFA, followed by elution of peptides with 20 μL 80% ACN, 1% FA. Prior to liquid chromatography (LC)−MS analysis, peptide mixtures were dried to completion in vacuum and stored at −80 °C. Signal Transduction Reaction Monitoring Assay Development

Sample Preparation

To confirm SIS peptide sequence and phosphosite localization and consistent LC−MS observability, all ordered crude SIS peptides were run separately on different LC−MS setups. Crude peptides were characterized by data-dependent acquisition (DDA) LC−MS runs on an Orbitrap Q-Exactive (Thermo Scientific) because we have demonstrated previously that this type of MS/MS spectra is more beneficial for selected reaction monitoring (SRM) assay development than conventional ion trap CID spectra.34 Peptides that failed to be identified or for which phosphosites were not localized by DDA

Frozen HDF cell pellets were lysed by sonication in lysis buffer (8 M urea in 50 mM ammonium bicarbonate, supplemented with 1 tablet Complete mini EDTA-free Cocktail (Roche) and 1 tablet PhosSTOP phosphatase inhibitor Cocktail (Roche) per 50 mL of lysis buffer). After centrifugation (20 000g, 30 min at 4 °C), the supernatant was collected and assayed for protein content using the bicinchoninic acid assay (BCA) kit following manufacturer instructions (Pierce, Rockford, IL). Protein reduction and alkylation were performed using final concen2911


Article

Journal of Proteome Research LC−MS runs were reanalyzed by targeted ETD fragmentation on an Orbitrap Elite (Thermo Scientific) for better phosphopeptide identification and localization or by SRMtriggered MS/MS on a TSQ Vantage (Thermo Scientific) for increased sensitivity. Subsequently, all beam-type MS/MS spectra (i.e., HCD and TSQ spectra) were used for SRM assay development. Assay development was performed using Skyline35 to pick the eight most abundant fragment ions for SRM-mode validation and collision energy optimization. Collision energy optimization was achieved by ramping the collision energy (CE) from 7.5 eV lower to 7.5 eV higher than the calculated CE (standard skyline equation) in steps of 1.5 eV. For the final assay, the five most abundant peptide fragment ions were manually picked that showed no (auto-) interference in standard (SIS) or endogenous (NAT) peptides measured in HeLa or HDF cell lysates enriched by Ti4+-IMAC. Peptide ID was confirmed by matching retention time (RT), relative fragment ion intensities, and SRM-triggered MS/MS runs that were scheduled in-between sample batches. Representative phosphopeptide RT reproducibility and XICs are shown in Supplementary Figure 1 in the SI. Out of 49 phosphosites (on 51 phosphopeptides), 48 could be pinpointed with a localization certainty higher than 95% resulting in a total average localization probability of 98.6%. In some cases, tryptic phosphopeptides inevitably contained miscleavages or methionine residues. Miscleavages (MC) in phosphopeptides are frequently observed when negatively charged phosphomoieties are in close proximity to positively charged arginine or lysine residues, thereby hampering cleavage by trypsin. Our data and other studies36,37 show that quantification standard deviations are similar between MC peptides and non-MC peptides, indicating MC peptides are reproducibly formed under our sample preparation conditions and therefore can be used for reproducible phosphosite quantification. In the case of methionine containing peptides, SRM assays for both the normal and oxidized forms were generated and measured in the final sample; however, when one of the two oxidation states was not present in a natural or SIS peptide, the peptide isoform was not taken into account for further analysis. In only two cases were both oxidation states present on both the SIS and NAT peptide. SRM analysis of these peptides showed identical regulation of both isomers with small deviations between biological replicate samples (Supplementary Figure 2 in the SI), indicating that oxidation was similar on both SIS and NAT peptides in all samples. SRM assay parameters for the phosphopeptides analyzed in this work can be found in Supplementary Table 1 in the SI.

In total, 89 phosphopeptides were measured in light and heavy, resulting in 884 transitions. All transitions were measured using polytyrosine tuned S-lens values, transition-optimized CE values, scheduled peptide RT window of 1.75 to 3 min, and a cycle time of 1.8 s, resulting in a maximum of ∼60 concurrent transitions. DDA LC−MS and SRM Data Analysis

Peptide MS/MS spectra from DDA LC−MS and SRM−MS/ MS were identified using Proteome Discoverer 1.4 (Thermo Scientific, Bremen) and Mascot 2.4 (Matrix Science, London), and phosphosite localization was performed using the phosphoRS38 PD node. SRM data were analyzed using Skyline.35 For the final SRM data set, transitions with remaining interference, signal intensities above 13E6, or low signal-tonoise ratios were discarded for further analysis. To find samples or peptides with interfering irreproducible transitions, we performed a t test comparing the average relative fragment intensity ratio between light and heavy peptides from biological triplicate runs. All transitions with a p value below 0.1 indicated inequality between NAT and SIS relative fragment intensities and were therefore discarded. Additionally, all samples were inspected manually for transition interference and consistency in transition peak shape, peak integration, and RT, using Skyline. In the end, only peptides with a minimum of three transitions for each SIS and NAT were kept for further analysis, resulting in an average of 4,2 transitions per peptide isotopologue. To achieve relative quantification, we used crude synthetic peptides for determining relative ratios. To prevent quantification problems regarding upper and lower limits of detection and signal response linearity, we used preferably samples with equal intensity for SIS and NAT versions (SIS/NAT ratio close to 1). Therefore, each sample was split into two aliquots and each was spiked with a different SIS concentration (with a 100fold difference), allowing for each peptide analyte to select one of the two (or both) spike-in concentrations, ideally obtaining a SIS-NAT fold-difference less than 10; however, because of strong regulations in NAT levels between different phenotypes the ratio NAT/SIS could inherently not always be kept less than 10-fold. To obtain phosphosite abundances relative to control, we divided each SIS normalized phosphopeptide abundance by the average intensity of the control biological triplicate. Finally, two sided student t tests with equal variances were used to calculate significant (p < 0.05) changes between biological triplicate samples.

■

SRM LC−MS Data Acquisition

ASSOCIATED CONTENT

S Supporting Information *

All samples were reconstituted in 10 μL of 10% FA, and 30% was analyzed on a TSQ Vantage coupled to an Easy nLC-1000 LC system configured with a single easy spray analytical column (ES803; 50 cm × 75 μm ID, 2 μm particles, 100 Å pore size) (Thermo Scientific) using 2 h runs. Briefly, samples were loaded with 6 μL of buffer A (0.1% FA) at 800 bar limited to a maximum of 300 nL/min, column equilibration was performed with 3 μL of buffer A at 800 bar, and peptides were separated on a gradient of 0−65 min 3−25% buffer B (0.1% FA, 99,9% ACN) and 65−75 min 25−40% buffer B at 200 nL/min, followed by a column wash for 10 min at 200 nL/min 100% buffer B. The TSQ Vantage spray voltage was set to 2.3 kV and was further configured to select peptides in Q1 at 0.7 fwhm and fragment them at 1.5 mTorr Argon in the second quadrupole.

Supplementary Table 1: All phosphosites and phosphopeptides SRM assay coordinates. Supplementary Table 2: All phosphosite and phosphopeptide identification and quantification results. Supplementary Figure 1: RT reproducibility. Supplementary Figure 2: Peptide methionine oxidation. Supplementary Figure 3: All phosphosite measurements in the PI3KmTOR and MAPK pathway. Supplementary Figure 4: Semiquantitative plots of immunoblots depicted in Figure 2A. Supplementary Figure 5: Annotated spectra of the doubly and triply phosphorylated RPS6 peptides. The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jproteome.5b00236. All raw data, annotated spectra, and SRM quantification results are available 2912


Article


(9) Kennedy, J. J.; Abbatiello, S. E.; Kim, K.; Yan, P.; Whiteaker, J. R.; Lin, C.; Kim, J. S.; Zhang, Y.; Wang, X.; Ivey, R. G.; Zhao, L.; Min, H.; Lee, Y.; Yu, M. H.; Yang, E. G.; Lee, C.; Wang, P.; Rodriguez, H.; Kim, Y.; Carr, S. A.; Paulovich, A. G. Demonstrating the feasibility of largescale development of standardized assays to quantify human proteins. Nat. Methods 2014, 11 (2), 149−155. (10) Wolf-Yadlin, A.; Hautaniemi, S.; Lauffenburger, D. A.; White, F. M. Multiple reaction monitoring for robust quantitative proteomic analysis of cellular signaling networks. Proc. Natl. Acad. Sci. U. S. A. 2007, 104 (14), 5860−5865. (11) Held, J. M.; Britton, D. J.; Scott, G. K.; Lee, E. L.; Schilling, B.; Baldwin, M. A.; Gibson, B. W.; Benz, C. C. Ligand binding promotes CDK-dependent phosphorylation of ER-alpha on hinge serine 294 but inhibits ligand-independent phosphorylation of serine 305. Mol. Cancer Res. 2012, 10 (8), 1120−1132. (12) Zhang, G.; Fang, B.; Liu, R. Z.; Lin, H.; Kinose, F.; Bai, Y.; Oguz, U.; Remily-Wood, E. R.; Li, J.; Altiok, S.; Eschrich, S.; Koomen, J.; Haura, E. B. Mass spectrometry mapping of epidermal growth factor receptor phosphorylation related to oncogenic mutations and tyrosine kinase inhibitor sensitivity. J. Proteome Res. 2011, 10 (1), 305−319. (13) Bauer, M.; Ahrne, E.; Baron, A. P.; Glatter, T.; Fava, L. L.; Santamaria, A.; Nigg, E. A.; Schmidt, A. Evaluation of data-dependent and -independent mass spectrometric workflows for sensitive quantification of proteins and phosphorylation sites. J. Proteome Res. 2014, 13 (12), 5973−5988. (14) Bisson, N.; James, D. A.; Ivosev, G.; Tate, S. A.; Bonner, R.; Taylor, L.; Pawson, T. Selected reaction monitoring mass spectrometry reveals the dynamics of signaling through the GRB2 adaptor. Nat. Biotechnol. 2011, 29 (7), 653−658. (15) Soste, M.; Hrabakova, R.; Wanka, S.; Melnik, A.; Boersema, P. J.; Maiolica, A.; Wernas, T.; Tognetti, M.; von Mering, C.; Picotti, P. A sentinel protein assay for simultaneously quantifying cellular processes. Nat. Methods 2014, 11 (10), 1045−1048. (16) de Graaf, E. L.; Giansanti, P.; Altelaar, A. F.; Heck, A. J. Single step enrichment by Ti4+-IMAC and label free quantitation enables indepth monitoring of phosphorylation dynamics with high reproducibility and temporal resolution. Mol. Cell. Proteomics 2014, 13 (9), 2426−2434. (17) Bisson, N.; James, D. A.; Ivosev, G.; Tate, S. A.; Bonner, R.; Taylor, L.; Pawson, T. Selected reaction monitoring mass spectrometry reveals the dynamics of signaling through the GRB2 adaptor. Nat. Biotechnol. 2011, 29 (7), 653−658. (18) Collado, M.; Serrano, M. Senescence in tumours: evidence from mice and humans. Nat. Rev. Cancer 2010, 10 (1), 51−57. (19) Michaloglou, C.; Vredeveld, L. C.; Soengas, M. S.; Denoyelle, C.; Kuilman, T.; van der Horst, C. M.; Majoor, D. M.; Shay, J. W.; Mooi, W. J.; Peeper, D. S. BRAFE600-associated senescence-like cell cycle arrest of human naevi. Nature 2005, 436 (7051), 720−724. (20) Kuilman, T.; Michaloglou, C.; Mooi, W. J.; Peeper, D. S. The essence of senescence. Genes Dev. 2010, 24 (22), 2463−2479. (21) Kuilman, T.; Michaloglou, C.; Vredeveld, L. C.; Douma, S.; van Doorn, R.; Desmet, C. J.; Aarden, L. A.; Mooi, W. J.; Peeper, D. S. Oncogene-induced senescence relayed by an interleukin-dependent inflammatory network. Cell 2008, 133 (6), 1019−1031. (22) de Graaf, E. L.; Kaplon, J.; Zhou, H.; Heck, A. J.; Peeper, D. S.; Altelaar, A. F. Phosphoproteome dynamics in onset and maintenance of Oncogene-Induced Senescence. Mol. Cell. Proteomics 2014, 13 (8), 2089−2100. (23) Hornbeck, P. V.; Kornhauser, J. M.; Tkachev, S.; Zhang, B.; Skrzypek, E.; Murray, B.; Latham, V.; Sullivan, M. PhosphoSitePlus: a comprehensive resource for investigating the structure and function of experimentally determined post-translational modifications in man and mouse. Nucleic Acids Res. 2012, 40 (Database issue), D261−D270. (24) Brunet, A.; Pages, G.; Pouyssegur, J. Growth factor-stimulated MAP kinase induces rapid retrophosphorylation and inhibition of MAP kinase kinase (MEK1). FEBS Lett. 1994, 346 (2−3), 299−303. (25) Qian, Z. R.; Ter-Minassian, M.; Chan, J. A.; Imamura, Y.; Hooshmand, S. M.; Kuchiba, A.; Morikawa, T.; Brais, L. K.; Daskalova, A.; Heafield, R.; Lin, X.; Christiani, D. C.; Fuchs, C. S.; Ogino, S.;

for download at the PASSEL data repository (acc. no.: PASS00552; http://www.peptideatlas.org/PASS/PASS00552).

■

AUTHOR INFORMATION

Corresponding Authors

*A.J.R.H.: E-mail: [email protected]. Tel: +31 (0)30 2536797. *A.F.M.A.: E-mail: [email protected]. Tel: +31 (0)30 2539554. Present Addresses ∥ E.L.d.G.: Fondazione Pisana per la Scienza ONLUS, Via Panfilo Castaldi 2, 56121 Pisa, Italy. ⊥ S.M.: New Biochemistry building, Department of Chemistry and Department of Biochemistry, University of Oxford, South Parks Road, OX1 3QU, Oxford, United Kingdom.

Author Contributions #

E.L.d.G. and J.K. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by The Netherlands Proteomics Center, and the PRIME-XS project (grant agreement number 262067) funded by the European Union seventh Framework Programme. A.F.M.A. was supported by The Netherlands Organization for Scientific Research (NWO) with a VIDI grant (723.012.102) and D.S.P. with a NWO VICI grant and a Queen Wilhelmina Award grant from the Dutch Cancer Society (KWF Kankerbestrijding). This work is part of the project Proteins At Work, financed by The Netherlands Organisation for Scientific Research (NWO) as part of the National Roadmap Large-scale Research Facilities of The Netherlands (project number 184.032.201).

■

REFERENCES

(1) Chappell, W. H.; Steelman, L. S.; Long, J. M.; Kempf, R. C.; Abrams, S. L.; Franklin, R. A.; Basecke, J.; Stivala, F.; Donia, M.; Fagone, P.; Malaponte, G.; Mazzarino, M. C.; Nicoletti, F.; Libra, M.; Maksimovic-Ivanic, D.; Mijatovic, S.; Montalto, G.; Cervello, M.; Laidler, P.; Milella, M.; Tafuri, A.; Bonati, A.; Evangelisti, C.; Cocco, L.; Martelli, A. M.; McCubrey, J. A. Ras/Raf/MEK/ERK and PI3K/ PTEN/Akt/mTOR inhibitors: rationale and importance to inhibiting these pathways in human health. Oncotarget 2011, 2 (3), 135−164. (2) Hunter, T. The age of crosstalk: phosphorylation, ubiquitination, and beyond. Mol. Cell 2007, 28 (5), 730−738. (3) Altelaar, A. F.; Munoz, J.; Heck, A. J. Next-generation proteomics: towards an integrative view of proteome dynamics. Nat. Rev. Genet. 2013, 14 (1), 35−48. (4) Pinkse, M. W.; Uitto, P. M.; Hilhorst, M. J.; Ooms, B.; Heck, A. J. Selective isolation at the femtomole level of phosphopeptides from proteolytic digests using 2D-NanoLC-ESI-MS/MS and titanium oxide precolumns. Anal. Chem. 2004, 76 (14), 3935−3943. (5) Villen, J.; Gygi, S. P. The SCX/IMAC enrichment approach for global phosphorylation analysis by mass spectrometry. Nat. Protoc. 2008, 3 (10), 1630−1638. (6) Zhou, H.; Ye, M.; Dong, J.; Corradini, E.; Cristobal, A.; Heck, A. J.; Zou, H. Mohammed, S., Robust phosphoproteome enrichment using monodisperse microsphere-based immobilized titanium (IV) ion affinity chromatography. Nat. Protoc. 2013, 8 (3), 461−480. (7) Gerber, S. A.; Rush, J.; Stemman, O.; Kirschner, M. W.; Gygi, S. P. Absolute quantification of proteins and phosphoproteins from cell lysates by tandem MS. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (12), 6940−6945. (8) Picotti, P.; Bodenmiller, B.; Mueller, L. N.; Domon, B.; Aebersold, R. Full dynamic range proteome analysis of S. cerevisiae by targeted proteomics. Cell 2009, 138 (4), 795−806. 2913


Article

Journal of Proteome Research Kulke, M. H. Prognostic significance of MTOR pathway component expression in neuroendocrine tumors. J. Clin. Oncol. 2013, 31 (27), 3418−3425. (26) Corcoran, R. B.; Rothenberg, S. M.; Hata, A. N.; Faber, A. C.; Piris, A.; Nazarian, R. M.; Brown, R. D.; Godfrey, J. T.; Winokur, D.; Walsh, J.; Mino-Kenudson, M.; Maheswaran, S.; Settleman, J.; Wargo, J. A.; Flaherty, K. T.; Haber, D. A.; Engelman, J. A. TORC1 suppression predicts responsiveness to RAF and MEK inhibition in BRAF-mutant melanoma. Sci. Transl. Med. 2013, 5 (196), 196ra198. (27) Roux, P. P.; Shahbazian, D.; Vu, H.; Holz, M. K.; Cohen, M. S.; Taunton, J.; Sonenberg, N.; Blenis, J. RAS/ERK signaling promotes site-specific ribosomal protein S6 phosphorylation via RSK and stimulates cap-dependent translation. J. Biol. Chem. 2007, 282 (19), 14056−14064. (28) Vander Haar, E.; Lee, S. I.; Bandhakavi, S.; Griffin, T. J.; Kim, D. H. Insulin signalling to mTOR mediated by the Akt/PKB substrate PRAS40. Nat. Cell Biol. 2007, 9 (3), 316−323. (29) Casamayor, A.; Morrice, N. A.; Alessi, D. R. Phosphorylation of Ser-241 is essential for the activity of 3-phosphoinositide-dependent protein kinase-1: identification of five sites of phosphorylation in vivo. Biochem. J. 1999, 342 (Pt 2), 287−292. (30) Keshishian, H.; Addona, T.; Burgess, M.; Kuhn, E.; Carr, S. A. Quantitative, multiplexed assays for low abundance proteins in plasma by targeted mass spectrometry and stable isotope dilution. Mol. Cell. Proteomics 2007, 6 (12), 2212−2229. (31) Shi, T.; Fillmore, T. L.; Sun, X.; Zhao, R.; Schepmoes, A. A.; Hossain, M.; Xie, F.; Wu, S.; Kim, J. S.; Jones, N.; Moore, R. J.; PasaTolic, L.; Kagan, J.; Rodland, K. D.; Liu, T.; Tang, K.; Camp, D. G., 2nd; Smith, R. D.; Qian, W. J. Antibody-free, targeted massspectrometric approach for quantification of proteins at low picogram per milliliter levels in human plasma/serum. Proc. Natl. Acad. Sci. U. S. A. 2012, 109 (38), 15395−15400. (32) Wu, R.; Dephoure, N.; Haas, W.; Huttlin, E. L.; Zhai, B.; Sowa, M. E.; Gygi, S. P. Correct interpretation of comprehensive phosphorylation dynamics requires normalization by protein expression changes. Mol. Cell. Proteomics 2011, 10 (8), M111 009654. (33) Tsiatsiani, L.; Heck, A. J. R. Proteomics beyond trypsin. FEBS J. 2015, DOI: 10.1111/febs.13287. (34) de Graaf, E. L.; Altelaar, A. F.; van Breukelen, B.; Mohammed, S.; Heck, A. J. Improving SRM assay development: a global comparison between triple quadrupole, ion trap, and higher energy CID peptide fragmentation spectra. J. Proteome Res. 2011, 10 (9), 4334−4341. (35) MacLean, B.; Tomazela, D. M.; Shulman, N.; Chambers, M.; Finney, G. L.; Frewen, B.; Kern, R.; Tabb, D. L.; Liebler, D. C.; MacCoss, M. J. Skyline: an open source document editor for creating and analyzing targeted proteomics experiments. Bioinformatics 2010, 26 (7), 966−968. (36) Dickhut, C.; Feldmann, I.; Lambert, J.; Zahedi, R. P. Impact of Digestion Conditions on Phosphoproteomics. J. Proteome Res. 2014, 13 (6), 2761−2770. (37) Glatter, T.; Ludwig, C.; Ahrne, E.; Aebersold, R.; Heck, A. J.; Schmidt, A. Large-scale quantitative assessment of different in-solution protein digestion protocols reveals superior cleavage efficiency of tandem Lys-C/trypsin proteolysis over trypsin digestion. J. Proteome Res. 2012, 11 (11), 5145−5156. (38) Taus, T.; Kocher, T.; Pichler, P.; Paschke, C.; Schmidt, A.; Henrich, C.; Mechtler, K. Universal and confident phosphorylation site localization using phosphoRS. J. Proteome Res. 2011, 10 (12), 5354−5362.

2914
