|Home | About | Journals | Submit | Contact Us | Français|
We have developed a method for on-membrane direct identification of phosphoproteins, which are detected by a phosphate-binding tag (Phos-tag) that has an affinity to phosphate groups with a chelated Zn2+ ion. This rapid profiling approach for phosphoproteins combines chemical inkjet technology for microdispensing of reagents onto a tiny region of target proteins with mass spectrometry for on-membrane digested peptides. Using this method, we analyzed human epidermoid carcinoma cell lysates of A-431 cells stimulated with epidermal growth factor, and identified six proteins with intense signals upon affinity staining with the phosphate-binding tag. It was already known that these proteins are phosphorylated, and our new approach proved to be effective at rapid profiling of phosphoproteins. Furthermore, we tried to determine their phosphorylation sites by MS/MS analysis after in-gel digestion of the corresponding spots on the 2DE gel to the rapid on-membrane identifications. As one example of use of information gained from the rapid-profiling approach, we successfully characterized a phosphorylation site at Ser-113 on prostaglandin E synthase 3.
Phosphorylation is a major post-translational modification of proteins, and a large number of proteins in living cells are regulated by phosphorylation by kinases, or dephosphorylation by phosphatases.1 These modifications are essential for cellular events including signal transduction, DNA transcription, protein synthesis, cell-cycle progression, and cell metabolism.2–4 In order to elucidate relationships between regulation mechanisms by phosphorylation/dephosphorylation of proteins and such cellular events, it is necessary to investigate the phosphorylation status of proteins, identifying the specific sites at which they have been modified. This is still a challenging task, even with recent advances in the proteomics field, which includes powerful methods to analyze proteins by mass spectrometry. Several approaches using different strategies have been developed to investigate protein phosphorylation in cells and tissues. Immunostaining or immunoprecipitation approaches using specific phosphoserine, phosphothreonine, and phosphotyrosine antibodies have been utilized for detection or enrichment of phosphoproteins.5 Traditional metabolic labeling methods involving incorporation of the radioactive isotope 32P have also frequently been used in the phosphoproteomics field.6 Furthermore, analyses of phosphorylations by kinases and dephosphorylations by phosphatases in vitro have been effectively used to estimate dynamic movements of target proteins in phosphorylation/dephosphorylation systems.7 However, characterization of phosphorylation sites on target proteins using these approaches is technically challenging. Analysis by mass spectrometry, especially tandem mass spectrometry, has been relied on most often for characterization of phosphoproteins and their phosphorylation sites.8 Database searches are used to analyze MS/MS spectra from tryptic digested phosphoproteins, and these can indicate the phosphorylation sites of proteins. In general, phosphoproteins comprise less than 10% of cellular proteins in typical mammalian cells. Therefore, a method for enrichment of phosphoproteins from cells and tissues is required prior to MS analysis.9 Selective isolation and enrichment of phosphoproteins and -peptides can be carried out using antibodies against phosphorylated amino acids, or by β-elimination chemistry performed on phospho-Ser/Thr residues.10 However, it has been known that these methods have several drawbacks. For example, false-positive enrichment can happen when using antibodies against phospho-Ser/Thr because these antibodies have a low specificity in comparison with a phospho-Tyr antibody. In the β-elimination method, unwanted side reactions can occur, and reproducibility may be problematic. On the other hand, the enrichment of phosphoproteins and -peptides by immobilized metal ion affinity chromatography (IMAC), where chelated metal ions having an affinity for phosphate groups are used (such as Fe3+, Ga3+), is a fast, easy to use, and economical procedure.11 Ficarro et al. analyzed proteins extracted from yeast lysates that were digested with trypsin, and the resulting phosphopeptides were then identified by LC-MS/MS analysis after enrichment of phosphopeptides by IMAC.12 Furthermore, they have shown that methyl esterification of carboxylic groups decreased nonspecific binding of these peptides to the IMAC column.12 A phosphate-binding tag molecule, which is a novel type of IMAC utilizing the chelated Zn2+ ion and has a strong affinity to phosphate groups, has been recently developed.13,14 In the present study, we describe an approach consisting of both microscale identification of phosphoproteins on-membrane by using a Phos-tag molecule and characterization of their phosphorylation sites by MS/MS after enrichment of in-gel digests. The Phos-tag molecule is an alkoxide-bridged dinuclear metal complex (i.e., 1,3-bis[bis(pyridine-2-ylmethyl)amino]pro-pan-2-olato dizinc(II) complex). This method also utilizes on-membrane peptide mass fingerprinting (PMF) analysis of a microscale region using inkjet technology, followed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS).15 To achieve the process described above, a chemical inkjet printer instrument is used to capture the sample image, and to micro-dispense reagents onto the visually designated positions. Trypsin is microdispensed by chemical inkjet printer onto the membrane used for affinity staining, and then the matrix solution is dispensed directly onto the same position prior to MALDI-TOF MS analysis, which is carried out directly from the membrane surface. This approach has the advantage that PMF analysis of the visualized proteins is performed on the same membrane, without the necessity to extract proteolytic peptides after digestion. Furthermore, this microdispensing function of reagents using piezoelectric inkjet technology could be used to improve on-membrane PMF analysis in the microscale region of the protein spots without cross-contamination between proximate proteins. As a result, reagents at sub-nanoliter volume levels can be microdispensed, allowing digestion of only a tiny region within a protein spot. By using phospho-enriched whole-cell lysates of A-431 cells that had been stimulated with epidermal growth factor (EGF), we could successfully identify six proteins with intense signals in affinity staining with biotin-pendant Phos-tag on a 2DE blot.16 However, we could not identify the phosphorylation sites on these peptides by on-membrane MS analysis because of ion suppression by negative charges in a number of the nonphosphorylated peptides that were produced by tryptic digestion. Therefore, we further performed enrichment of phosphopeptides using Phos-tag agarose from in-gel digested peptides.
Firstly, we characterized the phosphopeptides from ovalbumin digests to confirm the detection limit for the MS approach after enrichment by Phos-tag agarose. Phosphopeptides were enriched using Phos-tag agarose, and a few hundred femtomoles were analyzed by SDS-PAGE/MALDI-MS as described above. Subsequently, we carried out enrichment of phosphopeptides followed by MALDI-MS analysis for 2DE-separated proteins of phospho-enriched whole-cell lysates of A431 cells stimulated with EGF. A phosphorylation site in a phosphopeptide of prostaglandin E synthase 3, one of the proteins identified by on-membrane PMF analysis, was characterized by MS/MS analysis to be at Ser 113. In this experiment a phosphorylation site of prostaglandin E synthase 3 was directly characterized, although phosphorylation sites of this protein were predicted to be at Ser 113 and Ser 118 by mutation analysis, including measurements of enzyme activity.17 Here we describe an effective approach for analysis of phosphoproteins consisting of both rapid on-membrane identification using chemical inkjet technology by detection with biotin-pendant Phos-tag and selective enrichment using Phos-tag agarose followed by MS/MS analysis to determine the specific phosphorylation sites.
Phospho-enriched whole-cell lysates of A-431+EGF/PE were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Ovalbumin, tributylphosphine, iodo-acetamide, ProteoPrep Total Extraction Sample Kit, polyvinylpyrrolidone (PVP-40), Direct Blue 71, α-cyano-4-hydroxy-cinnamic acid (CHCA), and 2,5-dihydroxy benzoic acid (2,5-DHB) were obtained from Sigma-Aldrich (St. Louis, MO). Trypsin was obtained from Promega (Madison, WI) and the Immobilon-FL PVDF membrane was purchased from Millipore (Bedford, MA). Streptavidin, Alexa Fluor 633 conjugate, was obtained from Invitrogen (Carlsbad, CA). Pharmalyte (pH 3–10) and Immobiline DryStrip (pH 3–10 NL, 13 cm) were purchased from GE Healthcare Bio-Sciences (Piscataway, NJ). 1-O-n-Octyl-β-D-glucopyranoside was obtained from Nacalai Tesque (Kyoto, Japan). Phos-Tag Agarose was purchased from MANAC, Inc. (Hiroshima, Japan), and biotin-pendant Phos-tag ligands were obtained from the Phos-tag consortium (http://www.phos-tag.com/english/index.html).
Direct analysis on the PVDF membrane was performed using a MALDI-TOF MS instrument, AXIMA-CFR plus (Shimadzu Corporation, Kyoto, Japan, and Kratos Analytical, Manchester, UK), which was operated in positive-ion mode by using an internal calibration method with trypsin autodigest and ACTH (18–39) (m/z 842.51, 2465.20). For MS/MS of enriched phosphopeptides, a MALDI-QIT TOF MS instrument, AXIMA-QIT (Shimadzu Corporation, Kyoto, Japan, and Kratos Analytical, Manchester, UK), was used, with external calibration using angiotensin II and ACTH (18–39) (m/z = 1046.54 and 2465.20, respectively). For on-membrane digestion on a microscale region of protein spots, the chemical inkjet printer (Shimadzu Corporation, Kyoto, Japan) was used for microdispensing the reagents onto blotted protein spots, as previously reported.15
Ovalbumin (5, 2, 1, 0.5, 0.2 pmol) was separated with SDS-PAGE (10–20%) and visualized with Coomassie brilliant blue (CBB) staining. The gel pieces were excised and washed in 50 mM NH4HCO3/50% (v/v) acetonitrile for 10 min. After reducing with 10 mM dithiothreitol at 56°C for 60 min, the protein in the gel was alkylated with 55 mM iodoacetamide for 60 min at room temperature. The following in-gel digestion was performed according to the protocol in the Experimental section, In-Gel Trypsin Digestion, as described below. The digested peptides were utilized for both PMF analysis and enrichment of phosphopeptides for MS analysis.
The proteins from phospho-enriched whole-cell lysates of A-431+EGF/PE (200 μg), were recovered by TCA precipitation and then dissolved in 200 μL Protein Extraction Reagent Type 3 solution (from ProteoPrep Sample Extraction Kit). Protein solutions from the cell lysates of A-431+EGF were prepared according to the manufacturer’s protocol for isoelectric focusing. Solubilized proteins were reduced with 5 mM tributylphosphine for 60 min at room temperature and then alkylated with 15 mM iodo-acetamide for 60 min at room temperature. Pharmalyte (pH 3–10) was added to a final concentration of 0.2% (w/v) and a trace of bromophenol blue (BPB) was also added. The protein solution was centrifuged at 15,000 g for 20 min at 20°C and the supernatant was used for rehydration of IPG strips. Amersham IPG strips (pH 3–10, 13 cm) were rehydrated for 8 h with the prepared sample solution (200 μL) and focused on a Protean IEF Cell apparatus (Bio-Rad, Hercules, CA) for 100 kV/h at a maximum of 8 kV. The focused IPG strips were equilibrated for 10 min with equilibration buffer including 20 mg/mL dithiothreitol, and the second dimensional SDS-PAGE (10–20%) was then performed for these strips. Separated gels were supplied to the following electroblotting for detection of phosphoproteins using a phosphate-binding tag, and they were also utilized for the in-gel digestion technique to enrich phosphopeptides using affinity-tag agarose.
The proteins separated by 2DE were blotted onto the Immobilon-FL membrane using the semi-dry electro-blotting method described previously.18 The blotted membrane was rinsed with water and air dried. Dried PVDF membranes were rewetted by dipping into 100% methanol for 10 sec and then were air dried on a filter paper for 15 min. Subsequently, the blot was dried in a vacuum chamber for 30 min. The blot was incubated at 37°C for 30 min and then was air dried for 2 h. According to the method reported by Kinoshita et al. with some modifications, phosphoprotein spots were detected with Phos-tag biotin and streptavidin labeled with Alexa Fluor 633.16 In order to form the zinc(II)-bound Phos-tag molecule, 10 μL of 0.1 M Phos-tag biotin in MeOH was incubated with 20 μL of 10 mM Zn(NO3)2, 2 μL of streptavidin conjugate, and 468 μL 10 mM Tris-HCl, 100 mM NaCl containing 0.25% (w/v) PVP-360, pH 7.4 (TBS-PVP), for 30 min at room temperature. After desalting on a filter unit (molecular weight cut off = 10 kDa), this active Phos-tag solution was diluted into 50 mL TBS-PVP buffer. The blot membrane was completely dried according to a slightly modified rapid immunodetection approach, as reported previously.19 The blot was immersed into the active Phos-tag solution and incubated for 30 min. The blot was then washed with TBS buffer for 5 min twice and then rinsed with water. The fluorescent image was detected with an excitation wavelength of 635 nm using the FLA-5000 analyzer (Fujifilm, Tokyo, Japan). After image acquisition, the Phos-tag molecule was removed by incubation in 1 N aqueous NH3 for 15 min three times. This was followed with Direct Blue 71 staining of the blot membrane.20 A direct on-membrane peptide mass fingerprinting (PMF) approach was followed for the dark spots corresponding to spots detected on fluorescent images.
The blot membrane was cut and adhered to the stainless steel plate using 3M electrically conductive tape 9713 (St. Paul, MN). A visualized image of the adhered blot was acquired with a scanner in the chemical inkjet printer, and the target protein spots were selected on the basis of the scanned images. Subsequently, the reagents for on-membrane digestion were printed onto microscale regions of the protein spots.15 Ten nL of 0.1% (w/v) PVP solution in 60% (v/v) MeOH was printed to pre-wet the membrane, and then 50 nL of trypsin at 40 μg/mL in 10 mM NH4HCO3 containing 10% (v/v) 2-propanol was micro-dispensed to each target position. On-membrane digestion was performed for 16 h at 30°C in a humidified chamber. After digestion, 100 nL of 5 mg/mL α-cyano-4-hydroxy-cinnamic acid in 0.1% (v/v) trifluoroacetic acid (TFA) containing 50% (v/v) acetonitrile was printed onto each position on the membrane. The blot was then subjected to on-membrane MS analysis using the AXIMA-CFR plus instrument. Positional information for the printed region was transferred to the mass spectrometer as an output file from the chemical inkjet printer, and MS analysis was performed for the digested region on the basis of this information. A database search from the obtained MS spectrum was conducted using the Sprot database with the aid of Mascot software (Matrix Science, MA), with fixed parameters including a tolerance of 0.3 Da for the MS analysis, and one missed cleavage site.
Relatively dark protein spots on a CBB-stained 2DE gel corresponding to the affinity-stained spots were excised and washed in 50 mM NH4HCO3 containing 50% (v/v) acetonitrile for 10 min. The gel pieces were washed with 100 mM NH4HCO3 for 10 min and then washed twice in 50 mM NH4HCO3/acetonitrile 50% (v/v) for 10 min. After removing supernatant, the gel pieces were dried in vacuo. An aliquot of 10 μg/mL trypsin in 100 mM NH4HCO3 containing 0.1% (w/v) 1-O-n-octyl-β-D-glucopyranoside was added and the pieces were incubated on ice for 15 min. Next, 6 μL of 100 mM NH4HCO3 was added and pieces were incubated at 37°C overnight. Digested peptides were then extracted with 0.1% (v/v) TFA/66% (v/v) acetonitrile. Half of the extract was used for purification of phosphopeptides using Phos-tag agarose, and the other half was used for general peptide mass fingerprinting.
Phos-tag agarose (30 μL of swelled gel) was added to a sample reservoir (a SUPREC-01 centrifugal filter unit). The storage buffer was removed by centrifugation (2000 g, 20 sec, 20°C) and then binding/washing buffer (0.1 M Tris-CH3COOH, 1.0 M CH3COONa, pH 7.5) was added to the sample reservoir. After centrifugation at 2000 g for 20 sec, the balancing buffer (0.1 M Tris-CH3COOH, 1.0 M CH3COONa, 10 μM Zn(CH3COO)2, pH 7.5) was added to form the active zinc (II)-bound Phos-tag agarose. The solution was incubated for 20 min at room temperature and the filtrate was removed by centrifugation (2000 g, 20 sec, 20°C). This washing by centrifugation was repeated with binding/washing buffer (100 μm) three times. Subsequently, the digested peptides was solubilized in 50 μL binding/washing buffer (0.1 M Tris-CH3COOH, 1.0 M CH3COONa, pH 7.5) and placed in the sample reservoir. The filter unit was centrifuged at 2000 g for 20 sec after incubation for 20 min at room temperature. The filter unit was washed with binding/washing buffer three times and the filtrate was discarded. Elution buffer (1 N aqueous NH3) was placed in the sample reservoir and incubated for 5 min at room temperature. Elutant was recovered by centrifugation (2000 g, 20 sec, 20°C). This elution step was repeated three times and the elutant was used for MS analysis after a desalting operation.
The elutant was dried in vacuo and then resolved in 0.1% (v/v) trifluoroacetic acid. A desalting operation was then performed with ZipTip according to the manufacturer’s protocol. Aliquots of 0.5 μL were dispensed onto a stainless steel plate with a micropipette and then an equal volume of 5 mg/mL CHCA in 50% (v/v) acetonitrile/0.1% (v/v) trifluoroacetic acid (TFA) was dispensed for PMF analysis and MS analysis for enriched phosphopeptides. After air drying, the MS operation was performed in positive mode with a MALDI-TOF MS instrument. PMF analysis was then carried out in reflectron mode, and MS operation in linear mode was used for MS analysis of enriched phosphopeptides. MS/MS analysis for phosphopeptides was performed with an AXIMA-QIT instrument with 5 mg/mL 2,5-DHB in 50% (v/v) acetonitrile/0.1% (v/v) TFA. A database search for identification of proteins from their MS spectrum was conducted as described in the section on-membrane PMF analysis. MS/MS ion search of enriched phosphopeptides was performed using the Sprot database with the aid of Mascot software, with fixed parameters including a tolerance of 0.3 Da for the MS analysis, and one missed cleavage site allowed.
Proteins extracted from phospho-enriched whole-cell lysates of A-431 stimulated with EGF were separated by 2DE and electroblotted onto a PVDF membrane. Detection of phosphoproteins using Phos-tag molecules was performed according to a modified rapid immunodetection procedure, as reported previously.16,21 This method permits Western blotting without a blocking procedure, providing the advantage that direct on-membrane MS analysis is unaffected by blocking proteins such as bovine serum albumin, casein, or skim milk. In this study we used a complex of the biotin-pendant Phos-tag and fluorescence-conjugated streptavidin for visualization of phosphoproteins. The method might facilitate the removal of the complexes from the membrane just by incubation in 1 N aqueous NH3 without using a special stripping buffer containing detergents or reducing agents, as described previously.16 This improvement would induce the decrease of the loss of the target proteins on the membrane, conferring a great benefit to subsequent MS analysis. Phosphoproteins on the blot membrane were specifically detected using the complexes. Figure 1b shows a fluorescent image of detected phosphoproteins. After detection of phosphoproteins, Phos-tag molecules bound to phosphoproteins were removed by washing three times with 1 N aqueous NH3. The blot membrane was stained with Direct Blue 71 (Figure 1a). It has been previously reported that the limit for detection of proteins with Direct Blue 71 staining is approximately 10 ng of protein.20 Most of the proteins detected by affinity staining with the biotin-pendant Phos-tag did not overlap spots on the Direct Blue 71–stained image, and most protein spots with intense signals in Figure 1b were estimated to be present at less than 10 ng. Comparing a Direct Blue 71–stained image and an affinity-stained image in Figure 1, we see that intense signals by the biotin-pendant Phos-tag were particularly observed in the molecular-weight range of approximately 24–30 kDa, as described previously (region 1 and region 2 in Figure 2). 22 These proteins were not also observed in the Direct Blue 71–stained image of Figure 1a and are suggested to be present at a very small amount. This result indicates that Phos-tag molecules are highly sensitive in detecting phosphoproteins on membrane.
We carried out microscale on-membrane PMF analysis for six proteins that were detected upon Direct Blue 71 staining and that also gave intense signals from detection with Phos-tag molecules (Figure 2). As described previously, trypsin and matrix solution were microdispensed onto a tiny region of the target protein spots using chemical inkjet technology, and the resulting digested peptides were directly analyzed on the membrane with a MALDI-TOF MS instrument.21 This piezoelectric inkjet technology enables to microdispense reagents at sub-nanoliter volume levels onto microscale region of proximate protein spots without cross-contamination. Furthermore, this microscale on-membrane PMF analysis allows detection and identification of phosphoproteins on a single blot membrane. It is a rapid, easy-to-use method that does not require a desalting procedure (e.g., by using a ZipTip). Therefore, this technology is very attractive for rapid on-membrane identification of proteins detected by affinity staining such as antibody, biotin-pendant affinity tag. The digested proteins were identified by a database search on the basis of the obtained MS spectrum. Table 1 shows the results of a database search (protein of spot 1 was also identified by PMF analysis of in-gel digests). All identified proteins were previously reported to be phosphorylated. By combining the Phos-tag molecules detection with on-membrane direct protein identification using inkjet technology, this rapid on-membrane profiling approach can be highly effective for phosphoprotein analysis.
An MS signal derived from phosphopeptides could not be observed upon on-membrane PMF analysis. It appeared that a negative charge on a phosphate group suppressed MS signals of phosphopeptides in a number of non-phospho peptides. Therefore, we tried to enrich phosphopeptides from in-gel digested peptides using Phos-tag agarose to characterize phosphorylation sites of identified proteins. We first demonstrated phosphopeptide isolation for ovalbumin to confirm the detection limit in MALDI-MS analysis after enrichment of phosphopeptides using Phos-tag agarose. Ovalbumin (5, 2, 1, 0.5, 0.2 pmol) was separated by SDS-PAGE (10–20%) and gel pieces corresponding to each protein band were excised. In-gel digestion of the gel pieces was performed, and phosphopeptides were enriched from the resulting tryptic digested peptides by Phos-tag agarose to give MS analysis in a linear mode. Figure 3a shows the result of MS analysis for in-gel digested ovalbumin without enrichment by Phos-tag agarose. A number of non-phosphopeptide signals and a single signal from phosphopeptide (340–359, [M+H]+ = 2088.91) were observed in typical MS spectra of in-gel digested ovalbumin. An MS spectrum of phosphopeptides enriched from tryptic digests of ovalbumin is shown in Figure 3b. Four signals from phosphopeptides are seen, except at the lowest concentration, 0.2 pmol ovalbumin. Furthermore, a number of non-phosphopeptides observed in Figure 3a were hardly detected after enrichment. This result highlights the advantages of detecting phosphopeptides after enrichment by Phos-tag agarose. The observed MS signals correspond to the regions 340–359 (m/z 2088.91), 62–84 (m/z 2525.15), and 59–84 (m/z 2915.34) including a single phosphate group (80 Da), and MS signals after β-elimination (approximately –98 Da) were also confirmed in a spectrum taken in reflectron mode (data not shown). Two phosphopeptides at m/z 2525.15 and m/z 2915.34 were likely modified with acrylamide on a cysteine residue. A signal corresponding to phosphopeptides alkylated with iodoacetamide was also observed at m/z 2511.14 (the region 62–84). In this study we successfully detected phosphopeptides processed by in-gel digestion of ovalbumin at concentrations above 0.2 pmol. Our results show that MS signals from non-phosphorylated ovalbumin peptides were suppressed after enrichment, whereas Phos-tag agarose enriched phosphorylated peptides with high specificity.
Subsequently, we tried to characterize phosphorylation sites on phosphoproteins identified by on-membrane microscale PMF analysis. Intense phosphopeptide signals were obtained upon MS analysis of prostaglandin E synthase 3 when enrichment of phosphopeptides using in-gel digests was carried out for six proteins. Figure 4 shows MS signals of phosphopeptides specifically selected by Phos-tag agarose. The MS signals observed in reflectron mode consisted of the phosphopeptide at m/z 1955.59 (DWEDDSDEDMSNFDR), its oxidized form (+16 Da), and the peptides that estimated to be cleaved with β-elimination. MS spectra of the other proteins after enriching for phosphopeptides did not detect any phosphopeptides (data not shown). There were faint spots for some of the other proteins on a Direct Blue 71–stained image, indicating they were not abundant to detect in MS mode due to loss of the samples during the enrichment procedure. In relation with heat-shock protein beta-1, which is relatively abundant (especially spot 5), some non-phosphorylated peptides observed on PMF analysis were retained after the enrichment procedure, and the phosphopeptides could not be completely enriched (data not shown). It seemed that nonspecifically bound nonphosphorylated peptides could not be removed, and that they inhibited ionization of phosphopeptides on MS analysis after enrichment. However, future refinements of this method should lead to improvements in characterization of phosphorylation sites. Heat-shock protein beta-1 is also known to be phosphorylated on multiple Ser/Thr residues, and identification of highly phosphorylated peptides is more difficult because of ion suppression with multiple phosphates in comparison with singly phosphorylated peptides.23
We performed MS/MS analysis for isolated phosphopeptides to characterize phosphorylation sites in prostaglandin E synthase 3 (Figure 5a). Although the enriched phosphopeptide was confirmed to be the region 108–122, corresponding the sequence DWEDDSDEDMSNFDR by MS/MS ion search, the phosphorylation sites could not be identified from the results shown in Figure 5a. Because this peptide has Ser residues at positions 113 and 118, no fragment ions containing a phosphate group (+80 Da) could be observed in the MS/MS spectra. Therefore, we could not determine which serine residue has a phosphate group by MS/MS ion search of the phosphopeptide. Subsequently, we tried to characterize the phosphorylation site by comparing the MS/MS spectrum for the phosphopeptide ([M+H+] = 1955.59) described above with that of the β-eliminated phosphopeptide ([M+H+] = 1857.60) (Figure 5b). Some fragment ions with β-elimination ( 98 Da) were seen in a series of y ions (y10, y11, y12, y13, and y14) that were observed in both MS/MS spectra. If Ser 118 is phosphorylated, y5, y6, y7, y8 and y9 ions should be signals of fragment ions with β-elimination ( 98 Da) in Figure 5 (b). However, the observed fragment ions with β-elimination ( 98 Da) were y10, y11, y12, y13, and y14 ions, and the phosphorylation site was determined to be at Ser 113 by comparing information about the y10 to y14 ions. These results indicate that it is highly efficient to compare MS/MS for two peptides with/without β-elimination to determine the actual phosphorylation sites of peptides containing multiple predicted phosphorylation sites (Ser/Thr). Phosphorylation sites of prostaglandin E synthase 3 have been previously predicted to be at Ser 113 and at Ser 118 on the basis of mutation analysis.17 Our results confirm a previous report showing phosphorylation of Ser 113, even though no fragment ions derived from phosphorylation at Ser 118 were detected in MS/MS mode.
In conclusion, we report an effective rapid profiling approach consisting of detection with the biotin-pendant Phos-tag and microscale on-membrane protein identification, which utilizes a combination of inkjet printing technology and MALDI-TOF MS. The characterization of phosphorylation sites by enrichment of phosphopeptides using Phos-tag agarose followed by MS/MS analysis could become a powerful approach in the field of phosphoproteomics.
This work was supported by a Grant in Aid for Scientific Research (B) (15390013) from JSPS and a Grant in Aid for Young Scientists (B) (17790034) from MEXT.