Comparison of MS/MS Methods for Protein Identification from 2D-PAGE Giorgio Arrigoni,†,‡ Celine Fernandez,§ Cecilia Holm,§ Michaela Scigelova,| and Peter James*,† Department of Protein Technology, Lund University, Sweden, Department of Biological Chemistry, University of Padova, Italy, Department of Molecular Endocrinology, Lund University, Sweden, and Thermo Electron, Hemel Hempstead, United Kingdom Received March 30, 2006
We have compared the use of a low resolution MALDI-Ion Trap MS/MS and a high-resolution ESITOF-MS/MS for the analysis of spots from 2D gels. The main criteria were speed and accuracy of protein identification. The results obtained using the MALDI-MS/MS system are comparable to those from the LC-MS/MS system in terms of accuracy, but less low-level proteins are identified while the time required for the analysis is dramatically reduced. Keywords: two-dimensional electrophoresis • MALDI • ESI • HPLC-MS • protein identification
Introduction Over the past decade, mass spectrometry has become the main tool for the identification and analysis of proteins and peptides. The introduction of soft ionization methods such as MALDI and ESI have made routine analysis of proteins and peptides possible.1-6 Mass spectrometry has numerous advantages over the previous chemical degradation methods, notably speed, automation, and very high sensitivity. The standard approach for protein identification in a complex mixture requires an initial protein separation such as 2DPAGE or HPLC for peptides. If a protein separation approach is taken, the proteins are then cleaved with a site-specific protease and analyzed by mass spectrometry. The two main methods that are in use for the protein identification are peptide mass fingerprinting (PMF) and MS/MS peptide fragmentation. MALDI-TOF instruments are usually chosen in order to perform PMF. The MALDI source allows the ionization of the peptides predominantly in the singly charged state, so no charge state deconvolution step is required. However, though PMF is probably the most used and fastest method for protein identification, it also has some limitations. In particular, protein mixtures (>3) cannot be reliably analyzed with this method. In these cases, a different approach for the identification must be used and in particular, sequence-specific peptide fragmentation (MS/MS) is the method of choice to overcome these problems. In this way also very complex mixtures can be successfully analyzed and in a normal LC-MS/MS run even hundreds of peptides can be detected, fragmented and identified. The drawback of this kind of approach is basically the * To whom correspondence should be addressed. Protein Technology, Wallenberg Laboratory II, Lund University, P.O. Box 7031, SE-220 07 Lund, Sweden. Fax: +46 46 222 1495. E-mail:
[email protected]. † Department of Protein Technology, Lund University. ‡ Department of Biological Chemistry, University of Padova. § Department of Molecular Endocrinology, Lund University. | Thermo Electron.
2294
Journal of Proteome Research 2006, 5, 2294-2300
Published on Web 08/11/2006
time required for the analysis: usually considering the time needed for sample loading, sample elution and column equilibration a single analysis can require more than 1 h. In this study we have compared the performances of a low resolution vMALDI-LTQ with a high resolution LC-ESI-Q-TOF instrument on protein samples previously isolated by 2D gel electrophoresis. We chose only those proteins which could not be identified by PMF for MS/MS analysis.
Experimental Procedures Materials. The chemicals for sample preparation and gel electrophoresis were all from GE Amersham, (Uppsala, Sweden). Trypsin was obtained from Promega (Falkenberg, Sweden). Solvents and reagents for HPLC were from Sigma-Aldrich (Buchs, Switzerland). Protein Separation and Digestion. A 800-µg portion of mouse liver protein, as determined by the Lowry method (Total protein kit micro Lowry from Sigma, Buchs, Switzerland), was first separated on 24 cm-long IPG strip pH 3-10 using rehydration loading. The proteins were then reduced with DTT, alkylated with iodoacetamide, and separated on a 24 × 14 cm 12.5% acrylamide 2nd dimension polyacrylamide gel. The gel was stained with colloidal Coomassie and the image analyzed with Melanie software (GE Amersham Biosciences) and spots (2 mm diameter) were cut out and digested with trypsin (Promega, Falkenberg, Sweden) using the Ettan spot handling workstation (GE Amersham Biosciences). The peptides extracted from the gel were dried and then resuspended in 10 µL of 0.1% formic acid. All samples (300) were analyzed by MALDI-TOF (M@LDI-HT, Waters) and the spectra used for automated peptide mass fingerprinting with MASCOT (Matrix Sciences, London, UK). Those samples (64) that did not give a clear identification (i.e., with a P value
