Titanium Dioxide Coated MALDI Plate for On Target Analysis of

A 4800 MALDI-TOF/TOF mass spectrometer (Applied Biosystems, Foster City, CA) was used in reflectron mode averaging 2500 laser shots in a random, unifo...
0 downloads 0 Views 2MB Size
Download PDF
Titanium Dioxide Coated MALDI Plate for On Target Analysis of Phosphopeptides Federico Torta,† Matteo Fusi,‡ Carlo S. Casari,‡ Carlo E. Bottani,‡ and Angela Bachi*,† Biological Mass Spectrometry Unit, Division of Genetics and Cell Biology, San Raffaele Scientific Institute, Via Olgettina 58, 20132 Milano, Italy, and Dipartimento di Chimica, Materiali e Ingegneria Chimica, NEMAS-Center for NanoEngineered Materials and Surfaces and IIT, Italian Institute of Technology, Politecnico di Milano, Via Ponzio 34/3, I-20133 Milan, Italy Received October 20, 2008

Protein phosphorylation controls many cellular processes and activities. One of the major challenges in the proteomic study of phosphorylation is the enrichment of substoichiometric phosphorylated peptides from complex mixtures. Titanium dioxide (TiO2)-based chromatography is now widely applied to isolate phosphopeptides because of its efficiency and flexibility. In this study, a novel TiO2 coated matrix assisted laser desorption ionization plate is presented and tested for the purification of phosphopeptides from complex mixtures. The novel feature of this approach is the deposition of a nanostructured TiO2 film on stainless steel plates by pulsed laser deposition (PLD). By using tryptic digests of R-casein, β-casein, and other nonphosphorylated proteins, the successful enrichment of phosphopeptides was possible with this novel device, called T-plate, even when working in the low fmol range, making the sample ready for mass spectrometric analysis in few minutes. Keywords: Phosphopeptides • titanium dioxide • MALDI • proteomics

Introduction Reversible phosphorylation of proteins is an important regulatory mechanism that occurs in all organisms. Early estimates are that one-third of cellular proteins are phosphorylated,1 and it is a reasonable conclusion that all cellular functions are directly or indirectly affected by protein phosphorylation. There are approximately 100000 potential phosphorylation sites in the human proteome, and just a small percentage of them are currently known, emphasizing the need for sensitive, high-throughput methods to identify, characterize, and monitor new sites of protein phosphorylation.2 Protein kinases and phosphatases can regulate and target protein phosphorylation at specific sites. Great interest in acquiring a complete knowledge of phosphorylation mechanisms is also due to the fact that more than a hundred different protein kinases have been implicated in human cancer.3 It is therefore not surprising that protein kinases have emerged as the most studied class of new drug target in oncology and other disease areas.4 Phospho-proteomics focuses on the analysis of protein phosphorylation and makes use of mass spectrometry (MS) and specific methods to purify phosphorylated proteins and peptides.5-7 Phospho-proteomics may facilitate the monitoring of protein kinase activities and pathways and aid drug discovery and development. Mass spectrometry can identify the amino acid sequence that defines a phosphorylation site, and, for example, the sequence may provide an indication of the kinase responsible * Corresponding author. Angela Bachi, Mass Spectrometry Unit DIBIT, San Raffaele Scientific Institute, Via Olgettina 58, I-20132 Milano, Italy. Fax: +390226432253. E-mail: [email protected]. † San Raffaele Scientific Institute. ‡ Politecnico di Milano.

1932 Journal of Proteome Research 2009, 8, 1932–1942 Published on Web 02/11/2009

for the modification.8 The main goals of a phospho-proteomic experiment are the enrichment of phosphate-containing proteins or peptides and their characterization by MS. Thus, the major challenge for the application of phosphoproteomics lies within successfully subtracting phosphoproteins from the whole cell lysate (phosphorylation is usually substoichiometric) with a focus on identification of lowabundant phosphoproteins. To date, several strategies have been developed to enrich samples for phosphopeptides and to remove nonphosphorylated acidic peptides. One commonly used strategy is immobilized metal ion affinity chromatography (IMAC),9,10 a standard method for specifically enriching phosphorylated peptides prior to MS, where Fe3+ or Ga3+ ions are commonly used. The IMAC approach is based on ionic interactions, and it has been reported that peptides rich in glutamic, aspartic, cysteine, and histidine residues are often copurified. Alternatively, metal oxides such as ZrO2,11 TiO2, 12,13 and Al2O314 have been used to selectively concentrate phosphopeptides, where the phosphate functional groups can bind to the surface of metal oxide particles. Because of the stability of the metal oxides over a wide pH range, acidic buffers can be employed to avoid the binding of nonphosphorylated peptides to the active surface.15 Recently, new improved procedures were reported for the purification of phosphorylated peptides using TiO2, which substantially enhances the enrichment efficiency for phosphopeptides without any prior chemical modification.15-17 For example, titanium oxide nanoparticles, produced by chemical methods, are characterized by a high effective surface area and showenhancedperformancesinphosphopeptideenrichment.18,19 In this framework, it is demanding the exploitation of deposi10.1021/pr8008836 CCC: $40.75

 2009 American Chemical Society

research articles

On Target Analysis of Phosphopeptides tion techniques allowing a fine control and design of material properties down to the nanoscale. Among all available deposition techniques, physical vapor deposition takes advantage of a high degree of cleanliness and control on the produced material. In particular, pulsed laser deposition (PLD), based on the production of a vapor/plasma from a material target by intense laser pulses, allows the production of nanostructured films, surfaces, and nanoparticles with a peculiar control on the structure, stoichiometry, and morphology even for complex materials.20 Mass spectrometry instrumentations, data processing, and analysis strategies to characterize phosphopeptides after various enrichment procedures have been extensively described and reviewed.21-23 Common for these MS strategies is that proteins are enzymatically digested into peptides, which are then analyzed by MS either directly or after enrichment. The analysis of the molecular mass of a phosphopeptide will indicate the presence or absence of a phosphate group. The verification of a phosphate group and location of the exact phosphorylation site on a phosphopeptide can be performed using MS/MS, where the peptide is activated in the gas phase and fragmented to produce informative fragment ions. Extensive work has to be done to detect phosphopeptides in biological samples with adequate sensitivity, and new technological tools are focusing on the enrichment step. Our goal consisted in simplifying this procedure, providing a unique tool designed to permit phosphopeptide enrichment and detection with the same platform, eliminating tedious washing and centrifugation steps, speeding up analysis, and reaching a high sensitivity due to the benefits given by the use of a nanoscaled technology. Recently, new methods based on a MALDI plate that can be used for both enrichment and detection of phosphopeptides have been proposed. For example, a recent report described the use of gold-coated silicon wafers derivatized with a nitrilotriacetic acid moiety for phosphopeptide enrichment by Fe(III)IMAC.24 These IMAC-MALDI wafers allowed the enrichment and detection of low-picomole quantities of ovalbumin and R-casein tryptic phosphopeptides. Also Qiao et al.16 developed functionalized MALDI plates by sintering of commercial TiO2 nanopowders and demonstrated the selective capture of phosphoproteins from tryptic digest mixtures of BSA and β-casein. Additionally, a separate study based on a film deposition approach illustrated the enrichment and direct detection of casein tryptic phosphopeptides on a MALDI plate coated with ZrO2 by means of reactive ion landing.25 For such systems, enhanced performances may be achieved by a proper control on material properties down to the nanoscale. In this framework, PLD is particularly attractive since it allows assembling ultrafine particles in a nanostructured film with tailored morphology and structure. In this article, the use of a MALDI plate coated by PLD with a nanostructured titanium dioxide layer (T-plate), for the enrichment of tryptic phosphopeptides from R- and β-casein in complex mixtures, is reported. After optimization of the enrichment and analysis conditions, high selectivity and good sensitivity toward phosphopeptides were demonstrated using the newly developed T-plate, which also offers the possibility of being reused after cleaning the active surface. Most importantly, this platform will work both as a device for the purification and for the analysis of phosphorylated and nonphosphorylated peptides.

Experimental Procedures Materials. Acetonitrile (MeCN), trifluoroacetic acid (TFA), phosphoric acid, formic acid, ammonium hydroxide (NH4OH), 2,5-dihydroxybenzoic acid (DHB), R-cyano-hydroxy-cinnamicacid (CHCA), angiotensin I human, sequence-grade modified trypsin, R-casein, β-casein, RNase, bovine serum albumin (BSA), and myoglobin were purchased from Sigma. A Ti target (purity 99.995%) was purchased from Kurt J. Lesker Company. A Pure Oxygen (5.0) cylinder was purchased from Rivoira s.p.a. Preparation of Standard Protein Digests. R- and β-casein, RNase, BSA, and myoglobin were dissolved in 50 mM ammonium bicarbonate, pH 7.8, including 10 mM dithiothreitol (DTT) and incubated at 56 °C for 30 min. The proteins were subsequently alkylated using 40 mM iodoacetamide for 1 h at room temperature. The reaction was quenched with the addition of 10 mM DTT, and digested using trypsin (2% w/w) at 37 °C for 12 h. MALDI-TOF Mass Spectrometry. Matrix-assisted laser desorption ionization (MALDI) mass spectra were acquired on a Voyager DE-STR time-of-flight mass spectrometer (Applied Biosystems, Framingham, MA) equipped with a nitrogen laser operated at 337 nm. Ions were accelerated with a 20 kV pulse, with a delayed extraction period of 170 ns. Spectra were generated by averaging between 200 and 400 laser pulses. Laser intensity was set to optimize the signal-to-noise ratio and the resolution of mass peaks of the analyte. All spectra were internally calibrated and processed via the Data Explorer (version 4.9) software. MALDI-TOF/TOF Mass Spectrometry. A 4800 MALDI-TOF/ TOF mass spectrometer (Applied Biosystems, Foster City, CA) was used in reflectron mode averaging 2500 laser shots in a random, uniform pattern. MS/MS experiments were performed in reflectron mode with CID gas on and the precursor mass window set to relative with a value of 200 (FWHM). MALDI T-Plate Fabrication. Standard stainless steel MALDI TOF and MALDI TOF/TOF plates were coated with titanium oxide thin films (200 nm thickness) by pulsed laser deposition. The stainless steel deposition chamber is equipped with a turbomolecular pump connected to a primary pump and with a gas inlet system with pressure gauges for fine pressure control from high vacuum (10-5 Pa) to the atmosphere. Here a Ti target was ablated in 20 Pa O2 pressure exploiting UV laser pulses (duration ≈ 7 ns) from a quadrupled Nd:YAG laser (λ ) 266 nm, 10 Hz repetition rate). Then 2600 laser pulses were focused on the target with an energy density (fluence) of about 3 J/cm2. Titanium oxide was deposited at room temperature and subsequently annealed in air at 400 °C for 1 h. A copper grid was applied to the MALDI plates for titanium oxide deposition in specific wells. Phosphopeptide Enrichment. Peptide mixtures, diluted 1:5 to the desired concentration in the loading solution (DHB 50 mg/mL in 80% MeCN/5% TFA when not otherwise specified), were deposited onto the plate surface and incubated for 1 min at room temperature. After application, each peptide spot was washed with a micropipette by drawing in and expelling the loading solution first and then the washing solution (80% MeCN/2% TFA) onto the deposited sample. After the final Journal of Proteome Research • Vol. 8, No. 4, 2009 1933

research articles

Torta et al.

Figure 1. Pictures of the modified plate. (A) Digital picture of a MALDI plate after deposition of titanium dioxide in selected wells. Cross-sectional (B) and top view (C) SEM images of the deposited TiO2 thin film. Table 1. List of the Identified Phosphorylated Peptides (M + H)+

peptide sequencea

(RS1) 1660.79 (RS1) 1832.70 (RS1) 1927.69 (RS1) 1943.89 (RS1) 1951.95 (RS1) 2935.87 (RS2) 3008.22 (β) 2061.83 (β) 2556.20 (β) 3122.27

VPQLEIVPNpSAEER YLGEYLIVPNpSAEER DIGpSEpSTEDQAMEDIK DIGpSEpSTEDQAMoxEDIK YKVPQLEIVPNpSAEER KEKVNELpSKDIGpSEpSTEDQAMEDIKQ NANEEEYSIGpSpSpSEEpSAEVATEEVK FQpSEEQQQTEDELQDK FQpSEEQQQTEDELQDKIHPF RELEELNVPGEIVEpSLpSpSpSEESITR

a

Sequences are cited from ref 13.

wash, the surfaces were allowed to dry, and the matrix was spotted directly. Elution of Peptides from the T-Plate and Identification by NanoLC-MS/MS. For an initial estimation of phosphopeptide recovery from the new surface, 1 pmol of the phosphopeptide AKAVDGpYVKPQIKQ (m/z 1624.84) was loaded on the T-plate and washed with 80% MeCN/2% TFA. The phosphopeptide was then eluted with 1 µL of 0.5% NH4OH, mixed with 1 µL of angiotensin I (0.5 pmol, m/z 1296.48), that is used as internal standard to correct variations due to inherent inhomogeneity of matrix/peptides crystals. Then 0.5 µL of this mixture was spotted onto a normal MALDI plate and analyzed with the standard matrix (10 mg/mL CHCA in 70% MeCN/0.1% TFA). As control, 1 µL of the phosphopeptide stock solution (1 pmol/µL) was directly mixed with 0.5 pmol of angiotensin and spotted on a normal plate, followed by the addition of 0.5 µL 1934

Journal of Proteome Research • Vol. 8, No. 4, 2009

of matrix solution. Samples were analyzed using a Voyager MALDI-TOF mass spectrometer in reflectron mode. Recoveries were calculated using the formula: (relative signal after elution/ relative control signal) × 100. When using the T-plate as a device for phosphopeptide enrichment from a complex mixture before LC-MS/MS analysis, we omitted the placement of the matrix on the sample after the optimized enrichment procedure. In this case, DHB was substituted by 1 M glycolic acid in 80% MeCN/5% TFA (or just with 80% MeCN/2% TFA for simple mixtures) as loading and washing solution. After elution with 0.5% NH4OH, the sample was acidified with 5% formic acid and dried under vacuum. The sample was then resuspended in 10% formic acid and analyzed with nanoLC-MS/MS. Five microliters of eluate were injected in a capillary chromatographic system (EasyLC, Proxeon Biosystem, Denmark). Peptide separations occurred on a RP homemade 10-cm reverse phase spraying fused silica capillary column (75 µm i.d. × 10 cm), packed with 3-µm ReproSil 100C18 (Dr. Maisch GmbH, Germany) by using eluents A (H2O with 2% v/v MeCN, 0.1% v/v formic acid) and B (MeCN with 2% v/v H2O with 0.1% v/v formic acid). A 65 min gradient (0.2 µL/min flow rate) from 8% to 50% B was used to achieve separation. The LC system was connected to an LTQ-Orbitrap mass spectrometer (ThermoScientific, Bremen, Germany) equipped with a nanoelectrospray ion source (Proxeon Biosystems, Odense, Denmark). Full scan mass spectra were acquired in the LTQ Orbitrap mass spectrometer with the resolution set to 60000. The acquisition mass range for each sample was from m/z 350 to 1500 Da. The five most intense doubly and triply charged ions were automatically selected and fragmented in the ion trap. Target ions already selected for the MS/MS were dynamically excluded for 60 s. Data Analysis. Tandem mass spectra obtained with the LTQOrbitrap were extracted by Raw2MSM, version 1.5_2007.02.22. All MS/MS data were searched against the bovine International Protein Index (IPI) protein sequence database (IPI bovine 20081001) using an in-house Mascot server (Matrix Science, London, UK; version 2.1.04), assuming the digestion enzyme trypsin with one miss cleavage allowed. Mascot was searched with a fragment ion mass tolerance of 0.60 Da and a parent ion tolerance of 10.0 ppm. The iodoacetamide derivative of cysteine was specified as a fixed modification and oxidation of methionine and phosphorylation (ST) as variable modifications.

Results and Discussion Pulsed Laser Deposition (PLD) of Titanium Dioxide. Pulsed laser deposition has been employed to functionalize MALDITOF and MALDI-TOF/TOF plates. PLD is a versatile technique for the deposition of nanostructured films with a fine control on the material properties.26,27 In PLD, an intense pulsed laser beam is focused on a solid target placed in a vacuum chamber. Energy absorption from the laser leads to target ablation followed by ejection of ablated species (atoms, ions, electrons, etc.) in the form of an expanding plasma plume. Ablated species are deposited on a substrate where film growth takes place. The plasma expansion is strongly forward directed and allows for the easy production of patterned deposits using shadow masking techniques, while the film thickness can be carefully controlled tuning the number of laser shots. Of particular interest is the possibility to grow nanostructured and cluster-assembled films with controlled morphology by ablating material in the presence of a moderate gas pressure, and by exploiting plasma confinement effects due to the increased

On Target Analysis of Phosphopeptides

research articles

Figure 2. Optimization of phosphopeptide enrichment. MALDI mass spectra of 500 fmol of tryptic β-casein peptides before (A) and after enrichment when loading in 65% MeCN/0.1% TFA (B), 80% MeCN/2% TFA (C), and DHB 50 mg/mL in 80% MeCN/5% TFA (D). An asterisk indicates phosphorylated peptides.

collision rate of ablated species.28,29 The morphology and structure of the deposited material can be finely tuned by adjusting the gas pressure and the target-to-substrate distance.30 Titanium oxide dots (Ø ≈ 3 mm, thickness ≈ 200 nm) (Figure 1A) were deposited by ablating a titanium target in 20 Pa of pure oxygen and employing a Cu mask applied to the MALDI plate. As-deposited films have a porous structure organized at the nanoscale and characterized by columnar growth as shown in the SEM images reported in Figure 1B and C. As-deposited films are mainly constituted by amorphous titanium oxide; details on the structural and morphological characterization are reported elsewhere.31 The deposition conditions were chosen as a compromise between a high porosity (i.e., high surface area) and a sufficiently high degree of mechanical stability of the film as well as a good adhesion to the substrate. This assures a robust structure against loading and washing procedures with the possibility of repeated use of the same plate. In order to increase film cristallinity and to enhance film stability without substantially modifying the morphology, as-deposited films have been subsequently annealed in air for 1 h at 400 °C. Raman and X-ray diffraction investigations (not shown) revealed that annealed films were mainly constituted by nanocrystalline anatase phase. Direct Phosphopeptide Enrichment on a Titanium Dioxide Coated MALDI Plate. The compatibility of the modified MALDI plate with routine analysis, as protein mass fingerprinting

(PMF) from tryptic protein digests, was first tested. In standard experimental conditions (instrument settings and matrix as in conventional PMF), the results obtained with the titanium coated plate were comparable to those coming from a stainless steel target (data not shown). This finding confirmed that the new plate could be used in our instrumental system and that titanium dioxide nanofilm can support routine PMF analysis. However, our new device was created with the goal of being dedicated to specific characterization of phosphorylated peptides. Thus, the titanium dioxide coated surface of the MALDI plate was examined for its capability of selectively trapping the phosphopeptides from a tryptic digest of β-casein (500 fmol), chosen as a model compound. The sample, diluted 1:5 in the loading solution (65% MeCN/0.1% TFA), was deposited on a titanium dioxide coated well and allowed to stand at room temperature for 1 min; the spot was then rinsed with loading and washing solutions (see Experimental Procedures) to remove unbound and bound nonphosphorylated peptides. The remaining peptides mixture was then covered with DHB (20 mg/mL) + CHCA (5 mg/mL) in 50% MeCN matrix solution (including 1% phosphoric acid, 0.1% TFA). A list of the tryptic phosphorylated peptides derived from caseins is shown in Table 1. Figure 2A shows the mass spectrum of the tryptic digest of β-casein before rinsing; the phosphorylated peptides are detected only at abundance less than 10% compared to the nonphosphorylated peptides. However, when washing proceJournal of Proteome Research • Vol. 8, No. 4, 2009 1935

research articles

Torta et al.

Figure 3. Matrix optimization. MALDI mass spectra of 500 fmol β-casein digest onto the coated plate by using (A) DHB/CHCA 1:1 (15 mg/mL), (B) DHB/CHCA 4:1 (20 mg/mL:5 mg/mL). All the matrices were dissolved in 50% MeCN/0.1% TFA/1% phosphoric acid. Phosphopeptides are indicated with asterisks.

dures were carried out on the modified plate after the deposition of the sample, phosphopeptides started to appear in the mass spectrum, as displayed in Figure 2B where all three phosphopeptides (m/z 2061.83, 2556.20, and 3122.27) derived from β-casein are detected. Compared to the tryptic digest, the signal-to-noise ratio (S/N) of the phosphorylated peptides enriched by the TiO2 surface is significantly improved, and at the same time, the signals of contaminating nonphosphorylated peptides are clearly reduced. These results indicate that the coated plate can be used for efficient enrichment of phosphopeptides. Peaks at m/z 1968, 2462, 2935, 3028, correspond to metastable ions due to the loss of HPO3- from the phosphorylated fragments. To investigate the effect of the loading and washing buffers on the selective binding of phosphorylated peptides to the titanium dioxide coated plate, the percentages of acetonitrile, TFA, and DHB were varied from 50 to 80%, 0.1 to 5%, and 50 to 150 mg/mL, respectively. A major factor in nonspecific binding on TiO2-based affinity columns is hydrophobic interaction. Kokubu et al.32 have proved that acetonitrile could break up interactions between hydrophobic peptides and the metal oxides. As shown in previous publications, TFA-acetonitrile buffers show good selectivity in the enrichment of phosphopeptides.15 Titania is known to have amphoteric properties.33 In acidic solutions, TiO2 behaves as a Lewis acid with positively 1936

Journal of Proteome Research • Vol. 8, No. 4, 2009

charged titanium atoms, thereby displaying anion-exchange properties. In our experimental system, the use of 65% MeCN with 0.1% TFA as loading solution enhanced the specific enrichment of phosphopeptides but only partially removed nonphosphorylated peptides (Figure 2B). Even when increasing MeCN (80%) and TFA (2%) content, several nonphosphorylated peptides were still present on the surface of the plate (Figure 2C). As suggested by Larsen et al.,13 the presence of DHB in the loading solution can effectively improve the selectivity for phosphopeptides when using titania microcolumns, by competing for the binding to TiO2 with other nonphosphorylated groups and was thus added to our loading/ washing buffers. Solutions containing DHB (50 to 150 mg/mL depending on the complexity of the deposited sample) in 80% MeCN/5% TFA were employed for loading and for the first washing step. Another rinsing was then performed with 80% MeCN/2% TFA before deposition of the matrix. As shown in Figure 2D, DHB enhanced the selective enrichment of β-casein phosphorylated peptides, dramatically reducing nonspecific binding to TiO2. Phosphopeptides from S1-R-casein were also present in the original sample as contaminants and became evident after this procedure at m/z 1660.79, 1927.69, and 1951.95. In previous publications, TiO2 microcolumns have been used to enrich samples before LC-MS/MS analysis.34 In this context,

On Target Analysis of Phosphopeptides

research articles

Figure 4. Phosphopeptide enrichment from a complex sample. MALDI mass spectra of a mixture containing 500 fmol tryptic peptides of R- and β-casein, RNase, myoglobin, and BSA before (A) and after enrichment (B) onto the coated plate. Phosphopeptides are indicated with asterisks. Table 2. Quantification of Phosphopeptides Enrichment in a Complex Mixturea m/z

1952 2062 1661 1267

enrichment factor

≈ 3.2 ≈ 10.2 ≈4 ≈ 0.08

a The enrichment factor for 3 representative phosphopeptides (m/z 1952, 2062, 1661) and 1 non-phosphopeptide (m/z 1267) was calculated as MALDI intensity after washing/MALDI intensity before washing.

the use of DHB as a nonphosphopeptide excluder can contaminate the LC system and the inlet of the mass spectrometer.15 For this reason, different compounds were tested for the compatibility with LC-MS/MS analyses and, at the same time, for the efficiency in selecting for phosphorylated peptides. As shown by Jensen and Larsen,15 glycolic acid can be an acceptable substitute for DHB in the enrichment phase and does not contaminate the LC system. Thus, having in mind the possibility of using our titanium dioxide surface as a preparative device before LC-MS/MS analyses, 1 M glycolic acid in 80% MeCN/5% TFA was tested as loading solution. Also in this experimental condition, the sample laid for 1 min on the target after deposition and before washing. Glycolic acid behaves like DHB, also when using our approach, selectively

enriching for phosphopeptides and, after recovering from the surface, making the sample compatible with LC-MS analyses (as shown below). Selection of MALDI Matrix. The most used MALDI matrix in the study of phosphopeptides is DHB in 50% MeCN 1% phosphoric acid. It was found that, when using DHB, the detection limit for phosphopeptides is generally improved in comparison to other matrices, and the presence of phosphoric acid can additionally increase their signal intensity.35 However, inhomogeneous crystal formation in the DHB preparation requires a search for sweet spots in the sample. Spectrum acquisition is therefore time-consuming and difficult to reproduce, especially if the MALDI analysis is performed in automatic mode. As suggested by Laugesen and Roepstorff,36 a mixture of CHCA and DHB (1:1) as MALDI matrix can improve reproducibility and reduce the time of the measurements in common proteomic experiments. More recently, Chen and Chen37 suggested that this mixture is also extremely useful when analyzing low concentrations of phosphopeptides. According to their observations 2,5-DHB and CHCA are immiscible in solution, which helps the crystallization of 2,5DHB. As phosphopeptides cocrystallize with 2,5-DHB, but not with CHCA in the two-matrix mixture, the detection limit is then improved. As it was tested in our measurements, this matrix is really useful in this type of study. However, we noted Journal of Proteome Research • Vol. 8, No. 4, 2009 1937

research articles

Torta et al.

Figure 5. Versatility of the T-plate. (A) MALDI MS spectrum of 500 fmol β-casein digest after enrichment onto a 4800 MALDI TOF-TOF titanium dioxide coated plate. Phosphopeptides are indicated with asterisks. (B) Example of MS/MS spectrum of phosphopeptide FQpSEEQQQTEDELQDK enriched on the titanium dioxide coated plate.

that after deposition of the matrix on the TiO2 surface, most of the DHB crystals tend to grow as short needles at the rim of the spot while the rest of the area is filled with CHCA small crystals (data not shown). Thus, when shooting in the central area of the sample spot, the ionization of nonphosphorylated ions together with phosphorylated ones (Figure 3A) was observed. We then decided to increase the ratio of DHB/CHCA in favor of DHB, and when using DHB/CHCA 4:1, better results were obtained, deriving from an increased surface homogeneity and an intense response from all of the sample surface as shown in Figure 3B. Purification of Phosphopeptides from the Digest of Protein Mixtures. To evaluate the specificity of the T-plate and its ability to selectively capture phosphopeptides from a more complex sample, equal amounts (500 fmol) of a tryptic digest of nonphosphorylated proteins (RNase, myoglobin and albumin) and phosphoproteins (R-casein and β-casein) were mixed. Figure 4A displays the direct MALDI mass spectrum of the tryptic digest product of the mixed sample. Phosphopeptide ions are not observed in the MALDI mass spectrum prior to enrichment; the most intense peaks are all nonphosphorylated peptides. Compared to the spectrum in Figure 2A, where only the β-casein digest was deposited, signals from phosphorylated peptides became weaker or even disappeared, due to ion suppression from other abundant nonphosphopeptides. However, after enrichment, phosphopeptide ion peaks appeared in 1938

Journal of Proteome Research • Vol. 8, No. 4, 2009

the mass spectrum as the most abundant species (as shown in Figure 4B). The results demonstrate that the plate coated with TiO2 is very effective in selectively trapping phosphopeptides even from a very complex sample. To quantify the performance of the T-plate extraction procedure, the ratios between the measured intensity for 3 phosphorylated peptides and for a nonphosphorylated peptide, before and after enrichment, have been reported in Table 2 (average after 3 measurements, instrumental parameters as reported in Experimental Procedures). As shown in the table, after washing the plate surface, the signals from the phosphorylated species are largely enhanced, while signals from nonphosphopeptides are nearly abolished, eliminating any previous ionization suppression effect. MS/MS Measurements Using the T-Plate. The possibility to perform MALDI MS/MS experiments after enrichment of phosphopeptides can greatly improve the significance of this newly developed protocol, giving the chance to obtain the peptide sequence of unknown phosphorylated samples. Thus, a stainless steel 4800 MALDI-TOF/TOF plate was partially covered with a titanium dioxide film. After deposition and selective enrichment of 500 fmol of β-casein digest by using the previously optimized protocol, we confirmed the possibility to adapt this technique to different instruments (Figure 5A). Even when increasing the complexity of the deposited samples, results comparable with those already shown with the Voyager

On Target Analysis of Phosphopeptides

research articles

Figure 6. (A) MALDI mass spectrum of 25 fmol tryptic digest of β-casein after enrichment onto the 4800 MALDI TOF-TOF titanium dioxide coated plate and (B) MS/MS spectrum of phosphopeptide FQpSEEQQQTEDELQDK selected from the same sample. Phosphopeptides are indicated with asterisks.

MALDI-TOF were obtained (data not shown). The possibility to acquire MS/MS spectra was then tested on selected phosphorylated peptides. As an example, shown in Figure 5B, the singly phosphorylated peptide corresponding to m/z 2061 was easily fragmented, as on a conventional plate, allowing the unambiguous identification and localization of the phosphorylation site. We think that the use of a matrix mixture of DHB and CHCA facilitated the fragmentation of ionized peptides. It has been already reported that while DHB leads to the formations of peptide ions with low internal energy, CHCA is recognized as a hot matrix that triggers extensive fragmentation of phosphopeptides during the analysis.32,38,39 To further demonstrate the utility of this approach, the MS/MS spectra of the m/z 2061 ion before and after enrichment (Figure S1 panel A and B, respectively, in Supporting Information) were registered and searched with Mascot. As shown, the quality of

the spectrum largely improved after washing the plate and the corresponding results identified the monophosphorylated peptide of β-casein only in the case of the enriched sample. Sensitivity and Durability of the Device. To test the detection limit of phosphopeptides when using the T-plate, the amount of the β-casein tryptic digest was lowered to 25 fmol, and as shown in Figure 6A, only two phosphopeptide-related peaks were still observed in the mass spectrum. This result shows that this on-target enrichment technique is sensitive enough to detect phosphopeptide samples at low concentrations. In experiments characterized by low amounts of starting protein, the sample was enriched for phosphorylated peptides by using weaker loading conditions, when compared to complex and more concentrated samples, meaning that the loading and washing solutions were only made by 80% MeCN/2% TFA, in order to have less competition for the titanium binding sites. Journal of Proteome Research • Vol. 8, No. 4, 2009 1939

research articles

Torta et al.

Figure 7. MALDI mass spectrum after elution with NH4OH and deposition onto a stainless steel plate of the sample represented in Figure 2B.

The MS/MS analysis of phosphorylated peptides in this low concentration range is also feasible as shown in Figure 6B by the good fragmentation spectrum of the m/z 2061 ion. After assessing the specificity and sensitivity of the T-plate, the durability of the plate performances was also tested. After each enrichment procedure, the active surfaces were regenerated by first soaking in 0.5% NH4OH for 10 min, followed by soaking in excess water and ethanol. A MALDI-TOF/TOF spectrum of the surface after cleaning is reported as Figure S2 in Supporting Information. After 20 enrichments each, the spots were compared for their ability to detect phosphopeptides from 0.5 pmol of the peptide mixture produced by tryptic digestion of β-casein. The results (Figure S3 in Supporting Information) confirmed that the titanium dioxide surface is still capable of enriching phosphopeptides even after being used repeatedly. Recovery of Phosphopeptides from the Titanium Dioxide Surface. Not only binding but also recovery of phosphopeptides from the titanium dioxide coated plate was evaluated in our study. This could be useful if, after on-target purification of the phosphopeptides, a sample would be analyzed by other techniques. To evaluate the recovery of a phosphorylated peptide, the relative signal intensity for a standard phosphopeptide before and after on-plate enrichment was compared after elution with 0.5% NH4OH from the T-plate and deposition of the sample onto a normal MALDI plate (see Experimental Procedures and Figure S4, Supporting Information). After averaging 3 measurements, recovery efficiency was estimated around 90%, confirming the possibility to elute enough sample from the T-plate for the following analysis. However, the observed high recovery percentage could be facilitated by the ideal experimental conditions used for the standard peptides. With real and more complex samples, the degree of binding efficiency and selectivity (that also influences the recovery value) depends on the type of sample (competition between phosphopeptides and other peptides) and on loading and washing solutions used (that can contain more DHB when increasing sample complexity). The possibility to recover phosphopeptides from more complex samples was then tested. 1940

Journal of Proteome Research • Vol. 8, No. 4, 2009

Peptides originating from the tryptic digest of β-casein were enriched directly on-plate and analyzed as previously described. After analysis, the phosphopeptides bound to the plate surface were eluted with 0.5 µL of 0.5% NH4OH. The eluted sample was then redeposited on a conventional stainless steel MALDI plate and left to dry. The sample was then covered with 0.5 µL of the previously optimized matrix and the spectrum was acquired. As shown in Figure 7, signals originating from the same sample analyzed in Figure 2B, and now after elution, correspond to phosphopeptides and are characterized by a high intensity, especially in the case of the 3122 m/z ion. We think that this behavior could be due to the fact that the peptide contains 4 phospho-groups, which bind tightly to the TiO2 of the coated plate. This tight binding could make ionization more difficult when working with the modified plate, requiring therefore higher ionization energies. In addition, as shown in Figure 7, it is quite clear that after elution from the coated plate some contaminating nonphosphorylated peptides disappeared. The possibility to analyze the eluted sample with LC-MS/ MS instead of redepositing it on a stainless steel plate was also tested. The feasibility of the method was proven by loading 1 pmol of a β-casein digest onto the T-plate. After LC-MS/MS analysis, the Mascot search identified the most abundant monophosphorylated peptide at m/z 2061 (data not shown); 0.5 pmol of a 5 proteins mixture, prepared as previously described, was also analyzed with LC-MS/MS before and after enrichment. As shown in Table S1, Supporting Information, when analyzing samples before enrichment, all of the 5 proteins of the mixture (and also some contaminants) were identified by the Mascot database search; after the enrichment, search results confirmed that the procedure removed most of the nonphosphorylated peptides (eliminating completely peptides from myoglobin and RNase) and eluted from the T-plate a highly simplified peptide mixture composed mainly of phosphorylated peptides. This procedure led to the identification of more phosphopeptides than before the enrichment, proving the feasibility of the method.

On Target Analysis of Phosphopeptides

Conclusions In this article, we present a new MALDI target, called T-plate, produced by exploiting pulsed laser deposition of a nanostructured titanium dioxide thin film onto a stainless steel plate. The potential of PLD to produce nanostructured films with tailored properties and functionality is here demonstrated. The active surface was tested for standard analysis of protein digests and capacity to bind and enrich phosphopeptides from complex mixtures and make them detectable for a standard MALDIMS or MALDI-MS/MS analysis. All of the experimental steps required for enrichment and analysis are simple and fast. The phosphopeptides can be directly analyzed on the T-plate or alternatively eluted by using NH4OH and analyzed by ESI-LCMS/MS or redeposited on a stainless steel MALDI target. The advantages of using the T-plate involve practical use, fast experimental steps and the possibility of using the same active surface many times. The compatibility with a MALDI-TOF/TOF instrument could then open the perspective of using it for the identification of phosphosites in complex biological samples, exploiting the high mass accuracy in the MS/MS mode and the possibility of coupling an LC device for the separation and automated deposition of sample fractions.

Acknowledgment. This work was partially supported by the EU Network of Excellence “Targeting Cell Migration in Chronic Inflammation” (MAIN). We thank Dr. Emanuele Alpi at the University of Pisa and Dr. Dietmar Waidelich and Dr. Dietrich Merkel at Applied Biosystems (Darmstadt, Germany) where measurements with the 4800 MALDI-TOF/ TOF were performed. Supporting Information Available: MALDI MS/MS fragmentation spectra of the m/z 2061 phosphopeptide obtained before (A) and after (B) enrichment, MALDI MS spectrum of the T-plate surface after cleaning from the analysed peptide mixture, MALDI MS spectrum of the β-casein phosphopeptides by using a titanium surface after 10 cycles of enrichment and cleaning, MALDI-TOF MS spectra of the standard phosphopeptide AKAVDGpYVKPQIKQ (m/z 1624.84) obtained before (A) and after (B) enrichment, following elution from the T-plateand addition of angiotensin I (m/z 1296.48) as reference standard, and Mascot search results after LC-MS/ MS analysis of the 5 proteins mixture sample before and after enrichment onto the T-plate.This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Hubbard, M. J.; Cohen, P. On target with a new mechanism for the regulation of protein phosphorylation. Trends Biochem. Sci. 1993, 18 (5), 172–177. (2) Bodenmiller, B.; Mueller, L. N.; Mueller, M.; Domon, B.; Aebersold, R. Reproducible isolation of distinct, overlapping segments of the phosphoproteome. Nat. Methods 2007, 4 (3), 231–237. (3) Blume-Jensen, P.; Hunter, T. Oncogenic kinase signalling. Nature 2001, 411 (6835), 355–365. (4) Cohen, P. Protein kinases--the major drug targets of the twentyfirst century. Nat. Rev. Drug Discovery 2002, 1 (4), 309–315. (5) Collins, M. O.; Yu, L.; Choudhary, J. S. Analysis of protein phosphorylation on a proteome-scale. Proteomics 2007, 7 (16), 2751–2768. (6) Rossignol, M. Proteomic analysis of phosphorylated proteins. Curr. Opin. Plant Biol. 2006, 9 (5), 538–543. (7) Mann, M.; Jensen, O. N. Proteomic analysis of post-translational modifications. Nat. Biotechnol. 2003, 21 (3), 255–261. (8) Obenauer, J. C.; Cantley, L. C.; Yaffe, M. B. Scansite 2.0: Proteomewide prediction of cell signaling interactions using short sequence motifs. Nucleic Acids Res. 2003, 31 (13), 3635–3641.

research articles (9) Andersson, L.; Porath, J. Isolation of phosphoproteins by immobilized metal (Fe3+) affinity chromatography. Anal. Biochem. 1986, 154 (1), 250–254. (10) Sun, X.; Chiu, J. F.; He, Q. Y. Application of immobilized metal affinity chromatography in proteomics. Expert Rev. Proteomics 2005, 2 (5), 649–657. (11) Kweon, H. K.; Hakansson, K. Selective zirconium dioxide-based enrichment of phosphorylated peptides for mass spectrometric analysis. Anal. Chem. 2006, 78 (6), 1743–1749. (12) 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. (13) Larsen, M. R.; Thingholm, T. E.; Jensen, O. N.; Roepstorff, P.; Jorgensen, T. J. Highly selective enrichment of phosphorylated peptides from peptide mixtures using titanium dioxide microcolumns. Mol. Cell Proteomics 2005, 4 (7), 873–886. (14) Chen, C. T.; Chen, W. Y.; Tsai, P. J.; Chien, K. Y.; Yu, J. S.; Chen, Y. C. Rapid enrichment of phosphopeptides and phosphoproteins from complex samples using magnetic particles coated with alumina as the concentrating probes for MALDI MS analysis. J. Proteome Res. 2007, 6 (1), 316–325. (15) Jensen, S. S.; Larsen, M. R. Evaluation of the impact of some experimental procedures on different phosphopeptide enrichment techniques. Rapid Commun. Mass Spectrom. 2007, 21 (22), 3635– 3645. (16) Qiao, L.; Roussel, C.; Wan, J.; Yang, P.; Girault, H. H.; Liu, B. Specific on-plate enrichment of phosphorylated peptides for direct MALDITOF MS analysis. J. Proteome Res. 2007, 6 (12), 4763–4769. (17) Mohammed, S.; Kraiczek, K.; Pinkse, M. W.; Lemeer, S.; Benschop, J. J.; Heck, A. J. Chip-based enrichment and nanoLC-MS/MS analysis of phosphopeptides from whole lysates. J. Proteome Res. 2008, 7 (4), 1565–1571. (18) Liang, S. S.; Makamba, H.; Huang, S. Y.; Chen, S. H. Nano-titanium dioxide composites for the enrichment of phosphopeptides. J. Chromatogr., A 2006, 1116 (1-2), 38–45. (19) Chen, C. T.; Chen, Y. C. Fe3O4/TiO2 core/shell nanoparticles as affinity probes for the analysis of phosphopeptides using TiO2 surface-assisted laser desorption/ionization mass spectrometry. Anal. Chem. 2005, 77 (18), 5912–5919. (20) Eason, R. Pulsed Laser Deposition of Thin Films; Wiley: New York, 2007. (21) Domon, B.; Aebersold, R. Mass spectrometry and protein analysis. Science 2006, 312 (5771), 212–217. (22) Garcia, B. A.; Shabanowitz, J.; Hunt, D. F. Analysis of protein phosphorylation by mass spectrometry. Methods 2005, 35 (3), 256– 264. (23) Areces, L. B.; Matafora, V.; Bachi, A. Analysis of protein phosphorylation by mass spectrometry. Eur. J. Mass Spectrom. (Chichester, England) 2004, 10 (3), 383–392. (24) Dunn, J. D.; Watson, J. T.; Bruening, M. L. Detection of phosphopeptides using Fe(III)-nitrilotriacetate complexes immobilized on a MALDI plate. Anal. Chem. 2006, 78 (5), 1574–1580. (25) Blacken, G. R.; Volny, M.; Vaisar, T.; Sadilek, M.; Turecek, F. In situ enrichment of phosphopeptides on MALDI plates functionalized by reactive landing of zirconium(IV)-n-propoxide ions. Anal. Chem. 2007, 79 (14), 5449–5456. (26) Chrisey, D.; Hubler, G. Pulsed Laser Deposition of Thin Films; Wiley & Sons: New York, 1994. (27) Willmott, P.; Huber, J. Pulsed laser vaporization and deposition. Rev. Mod. Phys. 2000, 72 (1), 315. (28) Geohegan, D.; Puretzky, A.; Duscher, G.; Pennycook, S. Timeresolved imaging of gas phase nanoparticle synthesis by laser ablation. Appl. Phys. Lett. 1998, 72 (23), 2987–2989. (29) Lowndes, D.; Rouleau, C.; Thundat, T.; Duscher, G.; Kenik, E.; Pennycook, S. Silicon and zinc telluride nanoparticles synthesized by low energy density pulsed laser ablation into ambient gases. J. Mater. Res. 1999, 14 (2), 359–370. (30) Bailini, A.; DiFonzo, F.; Fusi, M.; Casari, C. S.; LiBassi, A.; Russo, V.; Baserga, A.; Bottani, C. E. Pulsed laser deposition of tungsten and tungsten oxide thin films with tailored structure at the nanoand mesoscale. Appl. Surf. Sci. 2007, 253, 8130–8135. (31) Fusi, M.; Russo, V.; Casari, C. S.; LiBassi, A.; Bottani, C. E. Titanium oxide nanostructured films by reactive pulsed laser deposition. Appl. Surf. Sci. [Online early access]. DOI: 10.1016/j.apsusc.2008.06.206. (32) Kokubu, M.; Ishihama, Y.; Sato, T.; Nagasu, T.; Oda, Y. Specificity of immobilized metal affinity-based IMAC/C18 tip enrichment of phosphopeptides for protein phosphorylation analysis. Anal. Chem. 2005, 77 (16), 5144–5154.

Journal of Proteome Research • Vol. 8, No. 4, 2009 1941

research articles (33) Nawrocki, J.; Dunlap, C.; McCormick, A.; Carr, P. W. Part I. Chromatography using ultra-stable metal oxide-based stationary phases for HPLC. J. Chromatogr., A 2004, 1028 (1), 1–30. (34) Thingholm, T. E.; Larsen, M. R.; Ingrell, C. R.; Kassem, M.; Jensen, O. N. TiO(2)-based phosphoproteomic analysis of the plasma membrane and the effects of phosphatase inhibitor treatment. J. Proteome Res. 2008, 7 (8), 3304–3313. (35) Kjellstrom, S.; Jensen, O. N. Phosphoric acid as a matrix additive for MALDI MS analysis of phosphopeptides and phosphoproteins. Anal. Chem. 2004, 76 (17), 5109–5117. (36) Laugesen, S.; Roepstorff, P. Combination of two matrices results in improved performance of MALDI MS for peptide mass mapping and protein analysis. J. Am. Soc. Mass Spectrom. 2003, 14 (9), 992– 1002.

1942

Journal of Proteome Research • Vol. 8, No. 4, 2009

Torta et al. (37) Chen, C. T.; Chen, Y. C. A two-matrix system for MALDI MS analysis of serine phosphorylated peptides concentrated by Fe3O4/ Al2O3 magnetic nanoparticles. J. Mass Spectrom. 2008, 43 (4), 538– 541. (38) Schulz, E.; Karas, M.; Rosu, F.; Gabelica, V.; Influence of the matrix on analyte fragmentation in atmospheric pressure, MALDI. J. Am. Soc. Mass Spectrom. 2006, 17 (7), 1005–1013. (39) Gorman, J. J.; Ferguson, B. L.; Nguyen, T. B. Use of 2,6-dihydroxyacetophenone for analysis of fragile peptides, disulphide bonding and small proteins by matrix-assisted laser desorption/ionization. Rapid Commun. Mass Spectrom. 1996, 10 (5), 529–536.

PR8008836