Key words: phosphopeptide • two-dimensional liquid chromatography (2D-LC) • Ettan MDLC • LC-MS/MS • reversed-phase chromatography (RPC) • strong cation exchange (SCX) chromatography • linear ion trap mass spectrometer
A 2D–LC-MS method was developed to analyze phospho-peptides in mouse brain tissue. The trypsin-digested tissue was separated by strong cation exchange chromatography (SCX), followed by reversed-phase chromatography (RPC) using Ettan™ MDLC. The detection was performed by mass spectrometry using neutral loss of phosphoric acid to selectively detect the phosphorylated peptides. Several phosphorylation sites were found, and a strategy for confident assignment of these was developed.
One of the most important post-translational modifications is phosphorylation of serine, threonine, or tyrosine residues. Phosphorylated proteins play important roles in a wide range of biological processes, such as signal transduction, apoptosis, and cell cycle control. Detection of phosphorylation sites by mass spectrometry in proteins extracted from biological material is hampered by the low abundance, low stoichiometry, and poor ionization of phosphopeptides (1).
In this work, a biocompatible nanoscale liquid chromatography (LC) system, Ettan MDLC, was used for separating phosphopeptides. No metal ions that can chelate phosphate groups are present in the fluid pathway of the LC system, resulting in highly sensitive analyses (2).
Separation of the tryptic peptides was performed in two dimensions, SCX followed by RPC. A Finnigan™ LTQ™ linear ion trap mass spectrometer was used for detecting phosphopeptides by fragmenting all peptides that exhibited a neutral loss of phosphoric acid.
Ettan MDLC 18-1176-44, 11-0008-41
NAP™ 10 Columns 17-0854-01
PlusOne™ DTT 17-1318-01
PlusOne Tris 17-1321-01
Trypsin, sequencing grade 17-6002-75
Other products required
Finnigan LTQ mass spectrometer (Thermo Electron)
TurboSEQUEST™ protein identification software (Thermo Electron)
BioBasic™ SCX, 2.1 x 250 mm (Thermo Electron)
Zorbax™ 300-SB C18 trap column, 300 µm i.d. x 5 mm, 3 µm (Agilent)
Zorbax 300-SB C18 analytical column, 75 µm i.d. x 150 mm, 3 µm (Agilent)
Ammonium bicarbonate (Merck)
Citric acid (Fluka)
Formic acid, ultrapure (Merck Suprapur™)
Iodoacetic acid (Merck)
Acetonitrile, HPLC grade
Water, HPLC grade
Approximately 0.5 mg of mouse brain tissue was prepared according to the following procedure. The proteins were dissolved by adding 1 ml of 9 M urea with 50 mM DTT to the tissue and allowing it to incubate for 60 min at 20 °C. A 1-ml aliquot consisting of 8 M urea, 250 mM TrisHCl, and 125 mM iodoacetamide, pH 8.8, was added. The mixture was allowed to incubate for 60 min at 20 °C. A 1-ml aliquot of this mixture was buffer exchanged with 20 mM ammonium bicarbonate, pH 7.8, on a NAP-10 desalting column. The protein sample was digested with trypsin (concentration ratio of 50:1) for 2–24 h. The trypsin was then inactivated by adding formic acid to the sample.
Using the offline 2D-LC configuration of the Ettan MDLC, 40 µg of trypsin-digested mouse brain sample was injected onto a 2.1 x 250-mm BioBasic SCX column and eluted at 100 µl/min with a linear salt gradient of 0–30% B over 40 min (A: 20 mM citric acid, pH 2.5, 25% CH3CN; B: A + 1 M NH4Cl). Fractions were collected twice every minute (Fig 1). The fractions were injected onto a 0.3 x 5-mm Zorbax RPC trap column, where they were desalted. RPC separation was performed on a Zorbax 0.075 x 150-mm analytical column at 250 nl/min with a linear gradient of 0–60% B over 50 min (A: 0.1% formic acid; B: 84% CH3CN and 0.1% formic acid), a step up to 100% B, and then holding at 100% B for 10 min. To increase throughput, one pair of columns was equilibrated while the other pair was used for analysis, a standard preprogrammed method on the Ettan MDLC.
A Finnigan LTQ linear ion trap was used. The MS method consisted of a cycle combining one full MS scan with three MS/MS events (25% collision energy) followed by an MS3 event (35% collision energy) that was triggered upon detection of -98, -49, or -32.7 Da from the precursor (neutral loss of phosphoric acid, charge states 1+, 2+, and 3+). Dynamic exclusion duration was set to 30 s. The MS/MS and MS3 spectra from all the runs were searched using TurboSEQUEST protein identification software. Modifications were set to allow for the detection of oxidized Met (+16); carboxyamidomethylated Cys (+57); phosphorylated Ser, Thr, and Tyr (+80); and dehydrated Ser and Thr (-18).
Results and discussion
By injecting a large amount of sample and separating it on an analytical scale SCX column, collecting the fractions, and then injecting these onto a nanoscale LC, the peptides of low abundance, such as phosphopeptides, could be detected (3). Forty micrograms of a mouse brain tissue tryptic digest was first separated by SCX using salt gradient elution. The chromatogram is shown in Figure 2.
Thirty of the collected fractions were further analyzed by LC-MS using the neutral loss function. Phosphopeptides were found in one-third of the analyzed SCX fractions as shown in Figure 2. They mainly eluted early in the salt gradient due to their reduced charge state (4). SCX can therefore be used as a phosphopeptide enrichment strategy.
An ion chromatogram and the MS3 events (i.e. where a neutral loss was detected) from one of the fractions are shown in Figure 3.
The phosphopeptides were found and the phosphorylation sites identified by TurboSEQUEST database searches on all MS3 spectra. The results were confirmed manually by studying the raw spectra. It was important to confirm that the charge state of the peptide was correct, that the neutral loss ion dominated the MS/MS spectrum (Fig 4), and that the sequence data was of high quality.
Database searches were then performed on all MS/MS spectra, and the results were used to confirm the MS/MS searches and to find tyrosine phosphorylations. Phospho-tyrosine does not lose phosphoric acid during collision in the ion trap; therefore, sequence data from MS/MS was used to find these phosphorylations (+80). Some possible tyrosine phosphorylations were identified but need further evaluation. Tyrosine-phosphorylated proteins exist in very low abundance in cells and probably fall below the detection limit of the method.
In total, 60 phosphorylated peptides were found originating from 50 proteins. Some of the identified phosphorylation sites are shown in Table 1. The proteins presented in the table are all known phosphoproteins involved in cell growth and cell differentiation.
MS/MS and MS3 spectra of one previously known phospho-peptide originating from Stathmin 1 are shown in Figure 4. A previously unreported phosphorylation site originating from scaffold attachment factor B2 is shown in Figure 5. Scaffold attachment factor B2 has been reported to suppress estrogen receptors and might have an important role in breast cancer (4).
The strategy for analyzing phosphopeptides confidently is summarized here:
1. 2D-LC (SCX/RPC)
2. MS3 on all peptides that show neutral loss of phosphoric acid
3. TurboSEQUEST searches on all MS3 spectra (-18@ST)
4. Manual confirmation of charge state and that neutral loss dominates MS/MS spectra
5. Further confirmation by MS/MS searches (+80@STY)
To confidently assign phosphopeptides in a complex mixture such as a tryptic digest of brain tissue, two-dimensional separations are needed. 2D-LC separated the peptides with high resolution, and the neutral loss MS detection was very selective for phosphopeptides. Care had to be taken when interpreting the data to avoid false positives from the database searches.
To increase the number of identified phosphopeptides, a greater amount of starting material would be needed (5), and possibly another chromatographic enrichment step specific for phosphopeptides, for example using titanium oxide media (6).
1. Ficarro, S. B. et al. Phosphoproteome analysis by mass spectrometry and its application to Saccharomyces cerevisiae. Nat. Biotechnol. 20, 301–305 (2002).
2. Application note: Highly sensitive phosphopeptide analysis using Ettan MDLC and a linear ion trap mass spectrometer, GE Healthcare, 11-0027-38, Edition AA (2005).
3. Beausoleil, S. A. et al. Large-scale characterization of HeLa cell nuclear phosphoproteins. Proc. Natl. Acad. Sci. USA 101, 12130–12135 (2004).
4. Jiang, S. et al. Scaffold attachment factor SAFB1 suppresses ERα-mediated transcription in part via interaction with N-CoR. Mol. Endocrinology 10.1210/me.2005-0100 (29 September 2005).
5. Ballif, B. A. et al. Phosphoproteomic analysis of the developing mouse brain. Mol. Cell. Proteomics 3, 1093–1101 (2004).
6. Pinkse, M. W. et al. Selective isolation at the femtomole level of phosphopeptides from proteolytic digests using 2D-NanoLC-ESI-MS/MS and titanium oxide precolumns. Anal. Chem. 76, 3935–3943 (2004).
back to top