Purification and Identification of O-Glc N Ac-Modified Peptides Using

Dec 15, 2010 - Discipline of Pathology, School of Medical Sciences, University of Sydney, ... School of Molecular Bioscience, University of Sydney, Ne...
2 downloads 0 Views 6MB Size
ARTICLE pubs.acs.org/jpr

Purification and Identification of O-GlcNAc-Modified Peptides Using Phosphate-Based Alkyne CLICK Chemistry in Combination with Titanium Dioxide Chromatography and Mass Spectrometry Benjamin L. Parker,†,‡ Pankaj Gupta,§ Stuart J. Cordwell,‡,|| Martin R. Larsen,*,† and Giuseppe Palmisano*,† †

Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, 5230 Odense M, Denmark Discipline of Pathology, School of Medical Sciences, University of Sydney, New South Wales, 2006, Australia § Department of Chemistry, Binghamton University, The State University of New York, Binghampton, New York 13902, United States School of Molecular Bioscience, University of Sydney, New South Wales, 2006, Australia

)



bS Supporting Information ABSTRACT: A selective method for the enrichment of O-GlcNAcylated peptides using a novel CLICK chemistry reagent is described. Peptides modified by O-GlcNAc were enzymatically labeled with N-azidoacetylgalactosamine. The azide was then reacted with a phospho-alkyne using CLICK chemistry and O-GlcNAcGalNAzPO4-containing peptides were enriched using titanium dioxide chromatography. Modified peptides were analyzed using a combination of higher energy collision dissociation for identification and electron transfer dissociation to localize the site of O-GlcNAc attachment. The enrichment method was developed and optimized using an alphacrystallin standard protein and then applied to a soluble protein preparation of mouse brain tissue and a nuclear preparation generated from HeLa cells. A total of 42 unique O-GlcNAcylated peptides were identified, including 7 novel O-GlcNAc sites. KEYWORDS: O-GlcNAc, phosphorylation, HCD MS/MS, ETD MS/MS, chemoenzymatic labeling, titanium dioxide chromatography, post-translational modifications

’ INTRODUCTION The O-linked β-N-acetylglucosamine (O-GlcNAc) post-translational modification (PTM) is attached to serine and threonine residues on nucleocytoplasmic proteins.1 The O-GlcNAcylated state of a protein is dynamically regulated, in a process akin to kinase/phosphatase regulation of protein phosphorylation, by O-GlcNAc transferase and O-GlcNAcase2,3 and is highly dependent upon the nutrient status of the cell.4 Interestingly, augmentation of O-GlcNAc levels has been shown to protect cells from a variety of environmental stresses.5 There is also growing speculation that some cellular signaling networks may be regulated by an interplay between O-GlcNAcylation and phosphorylation at either the same or adjacent sites;6,7 however, it is unknown whether this includes only the targets of specific protein kinases or whether a relationship occurs at a wider level. Identification of O-GlcNAcylated sites in proteins would enable hierarchical clustering of peptide motifs to aid in answering this question. Furthermore, O-GlcNAcylation has been proposed to be involved in a number of disease states including diabetes8 and Alzheimer’s disease9 making site-specific assignment critical for a full understanding of this modification at the molecular level. r 2010 American Chemical Society

The analysis of O-GlcNAc presents a major analytical challenge, most importantly the low substoichiometric nature of the modified proteins/peptides resulting in a poor number of identifications in proteomics experiments. A number of strategies for enrichment of O-GlcNAc-modified peptides have been attempted including: lectins,10 chemical derivatization and affinity enrichment11 or covalent capture,12 and chemoenzymatic labeling.13,14 Such methods are generally hindered by their laborious nature, requirements for large starting amounts of protein, and poor specificity, resulting in the identification of relatively few O-GlcNAc-modified sites (10-150) from a complex protein mixture. This is compared to other modification-based proteomics experiments such as phosphoproteomics where typically more than 1000 sites are routinely identified. Here, we describe a novel enrichment procedure for O-GlcNAcylated peptides. The method relies on the enzymatic labeling of O-GlcNAc with an azide-containing galactose derivative using the Click-iT Enzymatic Labeling System. A phospho-alkyne group Received: June 7, 2010 Published: December 15, 2010 1449

dx.doi.org/10.1021/pr100565j | J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research is reacted with the azide using CLICK chemistry and the phosphoderivative can then be specifically enriched using an optimized TiO2 enrichment procedure generally used for phosphopeptide enrichment.15 The enriched O-GlcNAcGalNAzPO4containing peptides are then analyzed using a high accuracy LTQ Orbitrap mass spectrometer by HCD16 combined with ETD17 MS/MS. We optimized the method using an O-GlcNAc-containing peptide standard and bovine alpha-crystallin, and then applied it to a complex soluble protein preparation from mouse brain tissue and a nuclear preparation from HeLa cells.

’ MATERIALS AND METHODS Materials

Sequencing grade trypsin was obtained from Promega (Madison, WI). Click-iT O-GlcNAc peptide standard and Click-iT O-GlcNAc enzymatic labeling kit were obtained from Invitrogen (Eugene, OR). TiO2 beads were obtained from GL Sciences (Japan). POROS Oligo R3 reverse phase material was obtained from Applied Biosystems (Foster City, CA). Phosphatase inhibitor cocktail was obtained from Calbiochem (Darmstadt, Germany). All other chemicals were obtained from Sigma (Steinheim, Germany). Alpha-Crystallin Digest

Bovine alpha-crystallin was digested with trypsin (50:1 substrate/enzyme) in 20 mM ammonium bicarbonate, pH 7.9, for 12 h at 37 °C. Digested peptides were desalted using a POROS Oligo R3 microcolumn as previously described.18 The peptides were eluted with 70% acetonitrile (ACN) and 0.1% trifluoroacetic acid (TFA), and dried by vacuum centrifugation. Preparation of Mouse Brain Soluble Proteins

Mouse brain (200 mg) was homogenized using a glass dounce homogenizer followed by tip-probe sonication in ice-cold lysis buffer (100 mM sodium carbonate, protease inhibitor cocktail, phosphatase inhibitor cocktail, 20 μM PUGNAc,19 pH 7.5) and centrifuged at 200 000g for 2 h. The supernatant containing soluble proteins was collected and digested on a 3000 molecular weight cutoff spin filter (Millipore) as previously described.20 Briefly, proteins were applied to the filter and centrifuged at 10 000g for 20 min at 25 °C. The protein sample on the filter was resuspended in 6 M urea, 2 M thiourea, and 100 mM ammonium bicarbonate (pH 7.9) containing 10 mM dithiothreitol (DTT) and incubated for 1 h at 30 °C followed by alkylation with 50 mM iodoacetamide (IAA) for 1 h at 30 °C in the dark. Samples were centrifuged at 10 000g for 20 min at 25 °C, resuspended in 100 mM ammonium bicarbonate, pH 7.9, and digested with trypsin (50:1 substrate/enzyme) for 12 h at 37 °C. Tryptic peptides were collected by centrifugation of the filter at 10 000g for 20 min at 25 °C. Preparation of HeLa Cell Nuclear Proteins

HeLa cells were grown in Dulbecco’s modified Eagle medium (Thermo), 10% fetal bovine serum containing 5% penicillinstreptomycin. In total, 1  107 cells were treated with 20 μM PUGNAc for 14 h and harvested by scraping. Cells were washed with phosphate-buffered saline and resuspended in hypotonic lysis buffer (10 mM HEPES, 1.5 mM MgCl2, 10 mM KCl, phosphatase inhibitor cocktail, protease inhibitor cocktail, and 20 μM PUGNAc, pH 7.9) and allowed to stand on ice for 15 min. Cells were lysed using a glass dounce homogenizer (40 strokes) and the crude nuclei collected by centrifugation at 500g for 15 min at 4 °C. Nuclei were resuspended in 250 mM sucrose and

ARTICLE

10 mM MgCl2 and layered onto 350 mM sucrose and 0.5 MgCl2 followed by centrifugation at 1430g for 5 min at 4 °C. The pellet containing enriched nuclei was resuspended in hypotonic lysis buffer containing 0.1% SDS and the nuclei were lysed using tip probe sonication. Nuclear proteins were precipitated using chloroform/methanol and resusupended in 6 M urea, 2 M thiourea, and 50 mM TEAB, pH 8.0, containing 10 mM DTT and incubated for 1 h at 30 °C followed by alkylation with 50 mM IAA for 1 h at 30 °C in the dark. The reaction mixture was diluted 5-fold with 50 mM TEAB and digested with trypsin (50:1 substrate/enzyme) for 16 h at 30 °C. Peptides were acidified to 2% formic acid and 0.1% TFA and purified by hydrophiliclipophilic-balance solid phase extraction (HLB-SPE) (Waters) according to the manufacturer’s instructions and dried by vacuum centrifugation. Synthesis of Cyclohexylammonium Pent-4-ynyl Phosphate

Phosphorous acid (1) (60 μL, 1.21 mmol) was dissolved in a mixture of 4-pentyn-1-ol (2) (3.90 mL, 41.95 mmol), and triethylamine (611 μL, 4.39 mmol) (Supporting Information). I2 was added over 5 min with stirring. The mixture was stirred for another 5 min, and then poured into acetone (35 mL) containing cyclohexylamine (1.2 mL). The mixture was kept for 2 h, and then the precipitate was recrystallized by ethanol containing a few drops of cyclohexylamine to produce dicyclohexylammonium pent-4-ynyl phosphate (3) (phospho-alkyne) as a white solid (220 mg, 51%). NMR data (not shown) were identical with the reported data.21 O-GlcNAc Enrichment by Chemoenzymatic Labeling and TiO2 Chromatography

Click-iT O-GlcNAc peptide standard, digested alpha-crystallin, digested soluble mouse brain proteins, and digested HeLa nuclear proteins were enzymatically labeled with GalNAz using galactosyltransferase-T1 (Gal-T1) (Y289L) from the Click-iT O-GlcNAc enzymatic labeling kit, essentially as described by the manufacturer. Briefly, peptides were resuspended in 40 mM HEPES, 50 mM sodium chloride, and 7 mM manganese chloride (pH 7.9) containing UDP-N-azidoacetylgalactosamine (UDPGalNAz) and Gal-T1 (Y289L) overnight at 4 °C. Digested mouse brain proteins were treated with 20 units of calf intestinal phosphatase (New England Biolabs) for 2 h at 37 °C. Digested HeLa nuclear proteins were treated with 50 units of calf intestinal phosphatase and 500 units lambda phosphatase (Sigma) for 2 h at 37 °C. The reaction mixtures were diluted 1:1 with 1% TFA and desalted using HLB-SPE. Peptides were resuspended in 50% methanol containing 2.5 mM dicyclohexanaminium pent-4-ynyl phosphate, 2 mM copper sulfate, 10 mM sodium ascorbate, and 1 mM Tris [(1-benzyl-1H-1,2,3-triazol-4-yl) methyl] amine (TBTA, in 4:1 t-butanol/DMSO) and incubated overnight at room temperature with gentle agitation. The reaction mixture was diluted 10-fold with TiO2 Loading Buffer (1 M glycolic acid in 80% ACN, 5% TFA) as previously described for phosphorylated peptides.15 This solution was incubated with 1 mg of TiO2 beads for 15 min at room temperature with gentle agitation. The suspension was then centrifuged at 1000g for 1 min. The supernatant was removed and the beads washed with TiO2 Loading Buffer. The suspension was centrifuged at 1000g for 1 min. The supernatant was removed and the beads washed with Washing Buffer 1 (80% ACN, 2% TFA) and centrifuged at 1000g for 1 min. The supernatant was removed and the beads were washed with Washing Buffer 2 (20% ACN, 0.2% TFA) and centrifuged at 1000g for 1 min. The supernatant was removed and the modified 1450

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research

ARTICLE

Figure 1. Schematic diagram showing the enrichment of O-GlcNAcylated peptides by GalNAz enzymatic labeling and cycloaddition of a phosphoalkyne, followed by TiO2 chromatography. Inset shows the structure of the phospho-alkyne (Reagent 1).

peptides were eluted with 1% ammonium hydroxide, pH 11.3, by vortexing for 15 min followed by centrifugation at 1000g for 1 min. The supernatant was removed, acidified to 10% formic acid, 0.1% TFA and the modified peptides were purified using a POROS Oligo R3 reversed phase microcolumn as described previously.15 RPnLC-MS/MS Analysis

Enriched peptides were resuspended in 0.1% FA and separated by reversed phase chromatography on an in-house packed 17 cm  100 μm Reprosil-Pur C18-AQ column (3 μm; Dr. Maisch GmbH, Germany) using an Easy-LC nanoHPLC (Proxeon, Odense, Denmark). The HPLC gradient was 0-40% solvent B (A = 0.1% formic acid; B = 90% acetonitrile, 0.1% formic acid) for 120 min at a flow of 250 nL/min. Mass spectrometric detection was achieved using an LTQ Orbitrap XL or LTQ Orbitrap Velos (Thermo Scientific, MA). An MS scan (400-2000 m/z) was recorded in the Orbitrap set at a resolution of 60 000 at 400 m/z followed by data-dependent ETD MS/MS analysis of the two most intense ions with detection in the LTQ (O-GlcNAc standard, alpha-crystallin and mouse brain) or detection in the Orbitrap (HeLa nuclear preparation) followed by HCD MS/ MS analysis of the corresponding ions with detection in the Orbitrap. Parameters for acquiring ETD were as follows: activation time = 100 ms, Q-activation = 0.25, dynamic exclusion = enabled with repeat count 1, exclusion duration = 30 s and intensity threshold = 30 000 (XL) or 15 000 (Velos), supplemental activation enabled. Parameters for acquiring HCD were as follows: activation time = 5 ms, normalized energy = 35 or 45, dynamic exclusion = enabled with repeat count 1, exclusion duration = 30 s. Data Analysis

Raw files were analyzed using Proteome Discoverer v1.2 beta (Thermo Scientific). ETD MS/MS spectra were converted to . mgf files with spectral processing to remove neutral losses from charge reduced precursors and searched against the mammalian Swiss-Prot and reversed-concatenated IPI bovine (v3.58) databases for alpha-crystallin experiments, reversed-concatenated IPI mouse (v3.72) for mouse brain experiments, and reversedconcatenated IPI human (v3.72) for HeLa nuclei experiments using both Mascot and Sequest search algorithms. Database searches were performed with the following fixed parameters:

precursor mass tolerance 10 ppm; MS/MS mass tolerance 1.2 Da (ETD data) or 0.05 Da (HCD data); carbamidomethylation of cysteine; 2-missed cleavages and semitryptic cleavage (alphacrystallin) or trypsin cleavage (mouse brain and HeLa nuclei). Variable modifications included: methionine oxidation; serine, threonine, tyrosine phosphorylation and, serine, threonine O-GlcNAcGalNAzPO4 (þ611.18399 Da). HCD MS/MS spectra were manually interpreted for the presence of diagnostic fragmentation of O-GlcNAcGalNAzPO4 by performing XIC of m/z ranges 409.10-409.12 and 311.13-311.14 and, the observation of a peak at [M þ H - 611.18]1þ corresponding to the loss of the GlcNAcGalNAzPO4 modification. All spectra containing O-GlcNAcGalNAzPO4 and O-phosphate-modified peptides were confirmed manually and independently by the authors G.P. and B.L.P and only accepted when confirmed by both of the authors.

’ RESULTS AND DISCUSSION Enrichment of Phospho-Alkyne Derivatized O-GlcNAcylated Peptides Using TiO2 Chromatography

The enrichment scheme described here is summarized in Figure 1. Proteins were proteolytically digested and O-GlcNAcylated peptides enzymatically labeled using Gal-T1 (Y289L), which is capable of specifically adding a GalNAz residue to the 4 position of GlcNAc.13,22 Excess UDP-GalNAz was removed using reversed phase microcolumns and O-GlcNAcGalNAz-modified peptides reacted with phospho-alkyne (Reagent 1) through a copper-catalyzed 1,3-cycloaddition.23 The phospho-modified O-GlcNAcylated peptides were then enriched by TiO2 chromatography and analyzed by MS. Since TiO2 chromatography has been optimized for enrichment of phosphorylated peptides, these will be co-purified with the phospho-modified O-GlcNAcylated peptides unless a dephosphorylation step is introduced prior to the CLICK chemistry labeling. An additional complication of the analysis of O-GlcNAc-modified peptides by MS is the preferential fragmentation of O-GlcNAc to form oxonium ions and limited peptide backbone fragmentation by collision-induced dissociation. We have previously demonstrated the utility of parallel HCD/ETD fragmentation to simultaneously characterize N-glycans and the peptides to which they are attached.24 HCD was capable 1451

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research

ARTICLE

Figure 2. MS/MS spectra of the O-GlcNAc peptide standard TAPTSTIAPG ([M þ 2H]2þ = 763.84 m/z) tagged and enriched using the scheme described in Figure 1. (A) LTQ CID spectrum showing diagnostic ions at m/z 409.1 and 612.1, confirming the presence of the tagged peptide. The spectrum also shows limited peptide backbone fragmentation and a dominant [M þ H - GlcNAcGalNAzPO4]1þ ion. (B) HCD spectrum showing diagnostic ions at m/z 311.14, 373.09, and 409.11, confirming the presence of the tagged peptide. The spectrum also shows comprehensive peptide backbone fragmentation. (C) ETD spectrum localizing the site of modification to either T4 or S5.

of providing sequence information from peptides with very labile modifications and we therefore hypothesized that this approach could be used to characterize phosphate-derivatized O-linked glycopeptides. Thus, a combination of ETD and HCD was utilized in this study to identify O-GlcNAcylated peptides.

The method was developed using the TAPTSTIAPG standard O-GlcNAcylated peptide (Invitrogen). Figure 2A shows the LTQ CID MS/MS spectrum of the O-GlcNAcGalNAzPO4 derivatized peptide. This spectrum shows dominant fragments originating from the loss of the sugars and very little peptide 1452

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research

ARTICLE

Figure 3. ETD MS/MS spectra of the alpha-A-crystallin peptide, AIPVSREEKPSSAPSS detected in (A) its tagged O-GlcNAcylated form ([M þ 3H]3þ = 752.02 m/z) and, (B) its reciprocal O-phosphorylated form ([M þ 3H]3þ = 574.61 m/z).

backbone fragmentation. Figure 2B shows the corresponding HCD MS/MS spectrum with an almost complete series of peptide backbone fragment ions confirming the sequence. In addition, the intact peptide with and without the N-acetylglucosamine is observed, as well as a number of intense ions at m/z 409.11, 373.09, and 311.14 corresponding to the loss of the GalNAzPO4, GalNAzPO4 - 2H2O, and GalNAzPO4 - H3PO4, respectively. The advantage of analyzing O-GlcNAcGalNAzPO4containing peptides using HCD versus LTQ CID is twofold. First, a greater peptide backbone fragmentation is observed due to multiple collisions in the octopole allowing for a greater confidence in peptide identification. Second, the high mass accuracy measurement of the fragments also increases the confidence for both the identification of the peptide and the identification of diagnostic O-GlcNAcGalNAzPO4 fragments. Spectra of modified peptides can be found regardless of the intensity of the fragment ions by performing XIC ( 25 ppm of the diagnostic ions. Figure 2C shows the corresponding ETD MS/MS spectrum

localizing the modification site to the underlined threonine or serine in the standard TAPTSTIAPG peptide. Interestingly, the spectrum also contains an intense ion at m/z 1320.08 corresponding to the loss of m/z 207 that may arise from fragmentation of the bond next to the azido ring. It should be noted that O-GlcNAcGalNAzPO4 fragments were also detected in the ETD MS/MS spectrum and presumably arise from supplemental activation of the precursor ion. Enrichment of O-GlcNAcylated Peptides from Alpha-Crystallin

Since we were readily able to detect the modified O-GlcNAcylated standard peptide using our experimental approach, we next decided to attempt purification of O-GlcNAcylated peptides from among a simple peptide (single protein) mixture. Alphacrystallin is a heavily modified eye lens protein that contains a number of previously published phosphorylated and O-GlcNAcylated amino acid residues.25 A tryptic digest of commercially available ‘purified’ bovine alpha-crystallin was subjected to the 1453

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research

ARTICLE

Table 1. Identified GlcNAcylated Peptides Identified in the Mouse Braina

a

“x” denotes modification of previous amino acid. Asterisk (*) indicates site not annotated in dbOGAP but previously described.

enrichment scheme described in Figure 1, but without phosphatase treatment. A total of 8 unique O-GlcNAcylated peptides containing 2 nonredundant sites, as well as 45 unique phosphopeptides containing 21 nonredundant sites were identified from both alpha-A-crystallin and alpha-B-crystallin (Supporting Table 1, Supporting Information). In the set of 8 unique O-GlcNAcylated peptides, we observed several peptide variants that may originate from either degradation during the purification procedure or represent physiologically relevant isoforms. Panels A and B in Figure 3 illustrate the ETD MS/MS spectra of the peptide AIPVSREEKPSSAPSS3þ from alpha-A-crystallin detected in both its O-GlcNAcylated and phosphorylated forms, respectively. In addition to the identification of O-GlcNAc and O-phosphate sites on alpha-crystallin, a total of 108 phosphorylation sites were identified from bovine proteins co-purified from the commercial alpha-crystallin preparation (FDR < 1%) (Supporting Table 2, Supporting Information). Additional copurifying O-GlcNAc-containing peptides were also detected by the diagnostic ions in the HCD spectra but their sequences could not be identified due to poor quality spectra (data not shown), most likely a consequence of the very low abundance of these peptides. Identification of O-GlcNAcylated Peptides from Mouse Brain Soluble Proteins

To test the enrichment strategy on a complex biological sample, we digested proteins from isolated mouse brain tissue and dephosphorylated the sample with calf intestinal alkaline phosphatase. O-GlcNAcylated peptides were tagged and enriched with TiO2 chromatography and analyzed by nRPLCMS/MS employing both HCD and ETD parallel fragmentation. A total of 14 unique GlcNAcylated peptides were identified and all sites could be unambiguously assigned (Table 1 and Supporting Information). Of the 14 sites, 2 were attached to an asparagine within the consensus motif for N-linked glycosylation (NxS/T/C) on sulfated glycoprotein 1. The identification of a single N-GlcNAc has been previously described;26 however, its biological role has not been studied. The 12 identified O-GlcNAc sites represented 9 proteins, which were searched against the database for O-GlcNAcylated proteins, dbOGAP (https://cbsb. lombardi.georgetown.edu/OGAP.html) to assess their novelty. Three O-GlcNAc-modified sites were novel, including Ser-539 on Traf2 and NCK-interacting protein kinase (P83510) and Ser260 on tankyrase-1-binding protein (P58871). An additional

novel O-GlcNAc site was identified at Thr-2356 on the previously annotated O-GlcNAcylated protein, Piccolo (Q9QYX7). The remaining 9 identified sites have been previously annotated including Thr-1354, Thr-1517, Ser-1772, Thr-2945 on bassoon (O88737), Thr-27 on beta-synuclein (Q91ZZ3), Thr-205 on YTH domain family protein 3 (Q8BYK6), Thr-579 on host cell factor C1 (Q61191), Ser-367 on Arf-GAP domain and FG repeats-containing protein 1 (Q8K2K6) and Ser-151 on tubulin polymerization-promoting protein (Q7TQD2). Despite treating the sample with calf intestinal phosphatase, we also identified 485 nonredundant phosphopeptides using the Mascot search algorithm (FDR < 1%) (Supporting Table 3). A total of 141 nonphosphorylated or non-GlcNAcylated peptides were also identified, indicating high enrichment efficiency for phosphate-modified peptides (>75%). Of the 9 O-GlcNAcylated proteins identified, phosphorylation sites were identified on 4: Ser-595, Thr-1090, Ser-1304, Ser-1409 on piccolo; Ser-610, Ser740 on Traf2 and NCK-interacting protein kinase; Ser-1375 on tankyrase-1-binding protein; and Thr-15, Ser-31 (or Ser-34) on tubulin polymerization-promoting protein. No sites shared with O-GlcNAcylation were identified. These data confirm that an advantage of our method is the ability to perform simultaneous enrichment of phosphorylated and O-GlcNAcylated peptides within a single experiment. While this is undoubtedly useful, it is not clear how the relative abundance of the two modifications will affect the ultimate enrichment efficiency. It is certainly possible that phosphorylation is a far more common and thus abundant modification than O-GlcNAc, which could result in poor recovery of O-GlcNAcylated peptides relative to phosphopeptides. For example, in the mouse brain, we observed only 12 O-GlcNAc-modified peptides compared to 450 phosphopeptides, even despite phosphatase treatment. This suggests that simultaneous purification would best be performed on relatively simple mixtures, such as purified protein complexes of 2-50 proteins. Alternatively, each sample could be analyzed in parallel by ‘splitting’ into ‘phosphatase treated’ (for O-GlcNAc-containing peptides) and ‘nonphosphatase treated’ (for phosphopeptides) fractions, which would maintain the biological relevance and allow direct comparisons of occupied sites, particularly in light of the ‘yin-yang’ theory of phosphate versus O-GlcNAc modification. The unambiguous identification of the novel O-GlcNAc site at Thr-2356 in Piccolo was possible through the observation of the 1454

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research

ARTICLE

Table 2. Identified O-GlcNAcylated Peptides Identified from HeLa Nuclear Extracta

a “x” denotes modification of previous amino acid. Highlighted ion scores indicate ambiguous site assignments. Asterisk (*) indicates site not annotated in dbOGAP but previously described.

c13 and c14 ions (Supporting Information p. 144). Piccolo is a high molecular mass protein associated with the active zones of presynaptic neurons. It has been hypothesized that Piccolo may participate in active zone formation and may act as a scaffolding protein of key molecules involved in synaptic vesicle recycling. The effects of post-translational modifications on Piccolo protein function have not been studied; however, knock-down studies of Piccolo in cultured hippocampal neurons influences presynaptic function by negatively regulating synaptic vesicle recycling through the modulation of synapsin-1a dynamics.27 The identification of a novel O-GlcNAc site at Ser-539 on Traf2 and NCK-interacting protein kinase (TNIK) was unambiguously achieved through the identification of the c1 and z12 ions (Supporting Information

p. 156). TNIK is a germinal center kinase that has been shown to activate the c-Jun N-terminal kinase pathway and affect F-actin structure and cell-spreading.28 Furthermore, TNIK is an activator of the Wnt signaling pathway and activates the transcription factor T-cell factor-4, which is essential for colorectal carcinogenesis.29 Identification of O-GlcNAcylated Peptides from HeLa Cell Nuclear Proteins

Since O-GlcNAc is predominantly described in the literature as a modification found on nucleocytoplasmic proteins, we further tested our enrichment strategy using digested proteins from a crude nuclei preparation from HeLa cells. We also extensively treated the sample with a phosphatase cocktail (alkaline 1455

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research

ARTICLE

Figure 4. ETD MS/MS spectrum of the peptide, VATTSVITIVK from Transcription factor MafK enriched from HeLa cell nuclear protein extracts, in its tagged O-GlcNAcylated form ([M þ 2H]2þ = 871.95 m/z) localizing the site of modification to Thr4 (Thr134 in MafK).

phosphatase and lambda phosphatase) in an attempt to reduce the number of co-purifying phosphopeptides and thus increase the number of O-GlcNAcylated peptide identifications. O-GlcNAcylated-peptides were again labeled and enriched by TiO2 chromatography and analyzed by nRPLC-MS/MS employing both HCD and ETD parallel fragmentation. A total of 30 unique O-GlcNAcylated peptides were identified with 19 O-GlcNAc sites unambiguously identified (Table 2 and Supporting Information). Of the 19 sites, 4 have not previously been reported including Thr-134 on Transcription factor MafK (O60675), Ser-518 on Host cell factor 1 (P51610), Ser-198 on YTH family domain protein 1 (Q9BYJ9) and Thr-650 on JmjC (Jumonji) domain-containing histone demethylation protein 2C (Q15652). In addition to this, 3 novel O-GlcNAcylated peptides were identified where the site could not be localized including Ser-475, Ser-477, or Ser-478 on cytokeratin 8; Ser-689, Thr-690, or Ser-691 on RNA-binding protein 26 (Q5T8P6); and Thr456 or Thr-460 on TGF-beta-activated kinase 1 and MAP3K7binding protein 2 (Q9NYJ8). The increased number of O-GlcNAc-modified peptide identifications (30 from HeLa nuclei versus 12 from mouse brain) likely reflects both the use of a phosphatase cocktail treatment and the reduction in the complexity of the sample by using a nuclear protein fraction. The phosphatase cocktail resulted in a significantly lower number of identified phosphopeptides compared to the soluble mouse brain preparation with only 23 nonredundant phosphopeptides identified (Supporting Table 4). The novel O-GlcNAc site identified at Thr-134 on Transcription factor MafK was unambiguously localized by the observation of z7 and z8 ions (Figure 4.). MafK is a member of the small Maf family of basic leucine zipper (bZip) proteins (MafK, MafF, and MafG) that function as both transcriptional activators and repressors. Maf bZips transcription factors have been implicated in erythroid differentiation30 and the regulation of the antioxidant

response element.31 Interestingly, the identified O-GlcNAc site at Thr-134 on MafK is substituted for an alanine in MafF. Host cell factor 1 (HCFC) is a component of the MLL5 multiprotein complex that functions to methylate Lys-4 of histone H3. The MLL5 complex contains O-GlcNAc transferase (OGT) and O-GlcNAcylation of the MLL5 histone lysine methyltransferase is necessary to induce acitivity.32 It is therefore likely that O-GlcNAcylation of HCFC may be associated with its role in the MLL5 complex. The JmjC domain-containing protein is a histone demethylase responsible for the demethylation of Lys-9 in histone H3. This provides further evidence of a role for O-GlcNAc protein modification in this process. The identification of 42 O-GlcNAcylated peptides (30 from a HeLa cell nuclear subcellular fraction and 12 from complex mouse brain tissue) is comparable with other studies examining O-GlcNAc enrichment on a global scale. For example, 34 O-GlcNAcylated peptides were identified from nuclear and S100 cytoplasmic fractions of mouse brain using chemoenzymatic labeling and amoxy biotin/avidin enrichment combined with SCX and off-line fractionation.33 Lectin weak affinity chromatography coupled to ETD allowed identification of 58 novel O-GlcNAc sites from the murine postsynaptic density pseudoorganelle; however, 28 of these were identified on a single protein, Bassoon.27 In arguably the most extensive ‘O-GlcNAcome’ characterized to-date, 141 O-GlcNAcylated peptides were identified from a midbody and spindle preparation of HeLa cells using chemoenzymatic labeling and photocleavable biotin enrichment combined with off-line fractionation and FETD on an LTQ-FT MS.7 Taken together, these results confirm the difficulty in identifying large numbers of O-GlcNAc modification sites in semicomplex and complex biological samples. Our method has the advantage of being rapid, highly specific, and relatively simple to perform and contains few steps that will lead to significant protein/peptide 1456

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research losses. Furthermore, our study did not utilize off-line peptide prefractionation; thus, it is likely that with extensive peptide separation, either prior to or after labeling, and TiO2 enrichment, for example, by strong cation exchange (SCX) or hydrophilic interaction liquid chromatography (HILIC), the method will lead to the purification and identification of large numbers of O-GlcNAcylated peptides and facilitate quantitative comparative analyses.

’ CONCLUSIONS An enrichment strategy that is able to purify and identify O-GlcNAcylated peptides has been described, and could be a valuable tool to study the complexity and function of this modification. As the method takes advantage of the strong interaction of phosphate with TiO2, it may be an important tool to address the relationship between O-GlcNAcylation and phosphorylation as simultaneous enrichment of these modifications is possible. It is likely, however, that the abundance of phosphorylation in complex biological systems could overwhelm the purification procedure, resulting in poor final yield of O-GlcNAcmodified peptides from complex mixtures. A parallel approach using split phosphatase treated and untreated samples will potentially be the most fruitful technique to yield high numbers of peptide identifications containing these modifications. The combination of the enrichment protocol with parallel ETD/ HCD fragmentation techniques enabled peptide identification and O-GlcNAc site assignments from a single experiment. HCD also provided diagnostic O-GlcNAcGalNAzPO4 fragment ions, as well as high mass accuracy sequence information. Despite the complementary nature of ETD/HCD fragmentation techniques, potential limitations of the strategy described here include: (i) the relatively low fragmentation efficiency of ETD that may result in ambiguous site assignment; (ii) the low sensitivity of HCD acquired in an Orbitrap XL versus an Orbitrap Velos; and (iii) the combination of ETD/HCD is a relatively slow fragmentation technique resulting in an extended duty cycle and a reduction in the number of acquired MS/MS spectra compared to other fragmentation techniques such as CID or PQD. The first point is further complicated by a reduction in the charge density induced by the addition of the GalNAzPO4 group, which may further reduce the fragmentation efficiency. The method developed by Wang et al. may have some advantage with respect to charge density, as their enriched GlcNAcylated peptides contain an additional charge induced by the photochemical cleavage of the biotin group.14 A recent improvement to increase the fragmentation efficiency in ETD is the use of a front-end ETD source (FETD) on an LTQ-FT.7 The implementation of FETD with a faster ETD reagent such as azulene (Prof. Donald Hunt HUPO Congress Oral Presentation, September 2010) combined with the parallel fragmentation of ETD/HCD on an LTQ-Orbitrap Velos is likely to address the three limitations of the strategy described above. We utilized this procedure to identify the known O-GlcNAc sites from alpha-crystallin through a number of peptide isoforms. We then identified 12 unique O-GlcNAcylated peptides from intact mouse brain tissue including 3 novel sites, and 485 nonredundant phosphopeptides from a single RPLC-MS/MS analysis. Subsequently, we employed a phosphatase cocktail and identified 30 unique O-GlcNAcylated peptides from a nuclear preparation of HeLa cells, including 4 novel O-GlcNAc sites, again from a single RPLC-MS/MS analysis. The identification

ARTICLE

of 30 O-GlcNAcylated peptides from a prefractionated subcellular preparation likely to be enriched in nucleocytoplasmic proteins and in the presence of extensive phosphatase treatment indicates the analytical challenge presented by O-GlcNAc modification. It is probable, therefore, that the combination of this enrichment strategy with extensive multidimensional liquid chromatography separation will be needed to provide a dramatic increase in the number O-GlcNAcylated-peptides identified from a complex protein mixture. The requirement for TiO2 purification will also enable the analysis of unbound flow-through to identify nonmodified peptides and thus gain insights into site occupancy, an issue that is very important to investigate the hypothesis that O-GlcNAcylation and phosphorylation compete for identical sites to regulate signal pathways. The enrichment procedure described here, coupled to existing O-GlcNAc and O-phosphate enrichment methods, and supported by high resolution and mass accuracy mass spectrometry, will lead to improved coverage of the O-GlcNAcome and facilitate quantitative analysis to determine the role of O-GlcNAc in biological systems.

’ ASSOCIATED CONTENT

bS

Supporting Information Synthesis of cyclohexylammonium pent-4-ynyl phosphate; tables and annotated MS/MS spectra of O-GlcNAcylated and phosphorylated peptides identified from alpha-crystallin, phosphorylated peptides identified from bovine proteins co-purified with alpha-crystallin, GlcNAcylated peptides identified from mouse brain and from HeLa nuclei. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*G.P.: Department of Biochemistry and Molecular Biology, The University of Southern Denmark, Denmark; (e-mail) [email protected]. M.R.L.: Department of Biochemistry and Molecular Biology, The University of Southern Denmark, Denmark; (e-mail) [email protected].

’ ACKNOWLEDGMENT The authors wish to thank Lene Jakobsen for instrument maintenance and Frank Kjeldsen for helpful discussions and instrument optimization. This study was supported by the Lundbeck Foundation (MRL-Junior Group Leader Fellowship) and the Danish Natural Science Research Council (MRL 09-065989), and the National Health and Medical Research Council (NHMRC) of Australia (SJC 571002). B.L.P. is the recipient of a University of Sydney Postgraduate Award and a Faculty of Medicine Alumni Award. ’ REFERENCES (1) Hart, G. W.; Housley, M. P.; Slawson, C. Cycling of O-linked B-N-acetylglucosamine on nucleocytoplasmic proteins. Nature 1999, 446, 1017–1022. (2) Kreppel, L. K.; Blomberg, M. A.; Hart, G. W. Dynamic glycosylation of nuclear and cytosolic proteins. Cloning and characterization of a unique O-GlcNAc transferase with multiple tetratricopeptide repeats. J. Biol. Chem. 1997, 272 (14), 9308–9315. (3) Hart, G. W. Dynamic O-linked glycosylation of nuclear and cytoskeletal proteins. Annu. Rev. Biochem. 1997, 66, 315–335. 1457

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458

Journal of Proteome Research (4) Kreppel, L. K.; Hart, G. W. Regulation of a cytosolic and nuclear O-GlcNAc transferase. Role of the tetratricopeptide repeats. J. Biol. Chem. 1999, 274 (45), 32015–3222. (5) Zachara, N. E.; O’Donnell, N.; Cheung, W. D.; Mercer, J. J.; Marth, J. D.; Hart, G. W. Dynamic O-GlcNAc modification of nucleocytoplasmic proteins in response to stress. A survival response of mammalian cells. J. Biol. Chem. 2004, 279 (29), 30133–30142. (6) Wang, Z.; Gucek, M.; Hart, G. W. Cross-talk between GlcNAcylation and phosphorylation: site-specific phosphorylation dynamics in response to globally elevated O-GlcNAc. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (37), 13793–13798. (7) Wang, Z.; Udeshi, N. D.; Slawson, C.; Compton, P. D.; Sakabe, K.; Cheung, W. D.; Shabanowitz, J.; Hunt, D. F.; Hart, G. W. Extensive crosstalk between O-GlcNAcylation and phosphorylation regulates cytokinesis. Sci. Signal. 2010, 3 (104), No. ra2. (8) Copeland, R. J.; Bullen, J. W.; Hart, G. W. Cross-talk between GlcNAcylation and phosphorylation: roles in insulin resistance and glucose toxicity. Am. J. Physiol.: Endocrinol. Metab. 2008, 295 (1), E17–28. (9) Liu, F.; Iqbal, K.; Grundke-Iqbal, I.; Hart, G. W.; Gong, C. X. O-GlcNAcylation regulates phosphorylation of tau: a mechanism involved in Alzheimer’s disease. Proc. Natl. Acad. Sci. U.S.A. 2004, 101 (29), 10804–10809. (10) Vosseller, K.; Trinidad, J. C.; Chalkley, R. J.; Specht, C. G.; Thalhammer, A.; Lynn, A. J.; Snedecor, J. O.; Guan, S.; Medzihradszky, K. F.; Maltby, D. A.; Schoepfer, R.; Burlingame, A. L. O-linked N-acetylglucosamine proteomics of postsynaptic density preparations using lectin weak affinity chromatography and mass spectrometry. Mol. Cell. Proteomics 2006, 5 (5), 923–934. (11) Wells, L.; Vosseller, K.; Cole, R. N.; Cronshaw, J. M.; Matunis, M. J.; Hart, G. W. Mapping sites of O-GlcNAc modification using affinity tags for serine and threonine post-translational modifications. Mol. Cell. Proteomics 2002, 1 (10), 791–804. (12) Klement, E.; Lipinszki, Z.; Kupihar, Z.; Udvardy, A.; Medzihradszky, K. F. Enrichment of O-GlcNAc modified proteins by the periodate oxidationhydrazide resin capture approach. J. Proteome Res. 2010, 9 (5), 2200–2206. (13) Khidekel, N.; Arndt, S.; Lamarre-Vincent, N.; Lippert, A.; Poulin-Kerstien, K. G.; Ramakrishnan, B.; Qasba, P. K.; Hsieh-Wilson, L. C. A chemoenzymatic approach toward the rapid and sensitive detection of O-GlcNAc posttranslational modifications. J. Am. Chem. Soc. 2003, 125 (52), 16162–16163. (14) Wang, Z.; Udeshi, N. D.; O’Malley, M.; Shabanowitz, J.; Hunt, D. F.; Hart, G. W. Enrichment and site mapping of O-linked N-acetylglucosamine by a combination of chemical/enzymatic tagging, photochemical cleavage, and electron transfer dissociation mass spectrometry. Mol. Cell. Proteomics 2010, 9 (1), 153–160. (15) Larsen, M. R.; Thingholm, T. E.; Jensen, O. N.; Roepstorff, P.; Jorgensen, J. D. Highly selective enrichment of phosphorylated peptides from peptide mixtures using titanium dioxide microcolumns. Mol. Cell. Proteomics 2005, 4, 873–886. (16) Olsen, J. V.; Macek, B.; Lange, O.; Makarov, A.; Horning, S.; Mann, M. Higher-energy C-trap dissociation for peptide modification analysis. Nat. Methods 2007, 4 (9), 709–712. (17) McAlister, G. C.; Phanstiel, D.; Good, D. M.; Berggren, W. T.; Coon, J. J. Implementation of electron-transfer dissociation on a hybrid linear ion trap-orbitrap mass spectrometer. Anal. Chem. 2007, 79 (10), 3525–3534. (18) Thingholm, T. E.; Jorgensen, T. J.; Jensen, O. N.; Larsen, M. R. Highly selective enrichment of phosphorylated peptides using titanium dioxide. Nat. Protoc. 2006, 1 (4), 1929–1935. (19) Haltiwanger, R. S.; Grove, K.; Philipsberg, G. A. Modulation of O-linked N-acetylglucosamine levels on nuclear and cytoplasmic proteins in vivo using the peptide O-GlcNAc-beta-N-acetylglucosaminidase inhibitor O-(2-acetamido-2-deoxy-D-glucopyranosylidene)aminoN-phenylcarbamate. J. Biol. Chem. 1998, 273 (6), 3611–3617. (20) Manza, L. L.; Stamer, S. L.; Ham, A. J.; Codreanu, S. G.; Liebler, D. C. Sample preparation and digestion for proteomic analyses using spin filters. Proteomics 2005, 5 (7), 1742–1745.

ARTICLE

(21) Lee, S. E.; Elphick, L. M.; Anderson, A. A.; Bonnac, L.; Child, E. S.; Mann, D. J.; Gouverneur, V. Synthesis and reactivity of novel gamma-phosphate modified ATP analogues. Bioorg. Med. Chem. Lett. 2009, 19 (14), 3804–3807. (22) Ramakrishnan, B.; Qasba, P. K. Structure-based design of beta 1,4-galactosyltransferase I (beta 4Gal-T1) with equally efficient N-acetylgalactosaminyltransferase activity: point mutation broadens beta 4Gal-T1 donor specificity. J. Biol. Chem. 2002, 277 (23), 20833–20839. (23) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. A stepwise huisgen cycloaddition process: copper(I)-catalyzed regioselective “ligation” of azides and terminal alkynes. Angew. Chem., Int. Ed. 2002, 41 (14), 2596–2599. (24) Scott, N. E.; Parker, B. L.; Connolly, A. M.; Paulech, J.; Edwards, A. V.; Crossett, B.; Falconer, L.; Kolarich, D.; Djordjevic, S. P.; Hojrup, P.; Packer, N. H.; Larsen, M. R.; Cordwell, S. J. Simultaneous glycan-peptide characterization using hydrophilic interaction chromatography and parallel fragmentation by CID, HCD and ETD-MS applied to the N-linked glycoproteome of Campylobacter jejuni. Mol. Cell. Proteomics 2010in press. (25) Viner, R. I.; Zhang, T.; Second, T.; Zabrouskov, V. Quantification of post-translationally modified peptides of bovine alpha-Crystallin using tandem mass tags and electron transfer dissociation. J. Proteomics 2009, 72 (5), 874–885. (26) Chalkley, R. J.; Thalhammer, A.; Schoepfer, R.; Burlingame, A. L. Identification of protein O-GlcNAcylation sites using electron transfer dissociation mass spectrometry on native peptides. Proc. Natl. Acad. Sci. U.S.A. 2009, 106 (22), 8894–8899. (27) Leal-Ortiz, S.; Waites, C. L.; Terry-Lorenzo, R.; Zamorano, P.; Gundelfinger, E. D.; Garner, C. C. Piccolo modulation of Synapsin1a dynamics regulates synaptic vesicle exocytosis. J. Cell Biol. 2008, 181 (5), 831–846. (28) Fu, C. A.; Shen, M.; Huang, B. C.; Lasaga, J.; Payan, D. G.; Luo, Y. TNIK, a novel member of the germinal center kinase family that activates the c-Jun N-terminal kinase pathway and regulates the cytoskeleton. J. Biol. Chem. 1999, 274 (43), 30729–30737. (29) Shitashige, M.; Satow, R.; Jigami, T.; Aoki, K.; Honda, K.; Shibata, T.; Ono, M.; Hirohashi, S.; Yamada, T. Traf2- and Nckinteracting kinase is essential for Wnt signaling and colorectal cancer growth. Cancer Res. 2010, 70 (12), 5024–5033. (30) Igarashi, K.; Itoh, K.; Hayashi, N.; Nishizawa, M.; Yamamoto, M. Conditional expression of the ubiquitous transcription factor MafK induces erythroleukemia cell differentiation. Proc. Natl. Acad. Sci. U.S.A. 1995, 92 (16), 7445–7449. (31) Dhakshinamoorthy, S.; Jaiswal, A. K. Small maf (MafG and MafK) proteins negatively regulate antioxidant response elementmediated expression and antioxidant induction of the NAD(P)H: Quinone oxidoreductase1 gene. J. Biol. Chem. 2000, 275 (51), 40134– 40141. (32) Fujiki, R.; Chikanishi, T.; Hashiba, W.; Ito, H.; Takada, I.; Roeder, R. G.; Kitagawa, H.; Kato, S. GlcNAcylation of a histone methyltransferase in retinoic-acid-induced granulopoiesis. Nature 2009, 459 (7245), 455–459. (33) Khidekel, N.; Ficarro, S. B.; Peters, E. C.; Hsieh-Wilson, L. C. Exploring the O-GlcNAc proteome: direct identification of O-GlcNAcmodified proteins from the brain. Proc. Natl. Acad. Sci. U.S.A. 2004, 101 (36), 13132–13137.

1458

dx.doi.org/10.1021/pr100565j |J. Proteome Res. 2011, 10, 1449–1458