Characterization of Low Abundant Membrane Proteins Using the Protein Sequence Tag Technology Thorsten Prinz, Jo1 rg Mu1 ller, Karsten Kuhn, Ju1 rgen Scha1 fer, Andrew Thompson, Josef Schwarz, and Christian Hamon* Proteome Sciences R&D, Industriepark Ho¨chst, Building G865a, D-65929 Frankfurt am Main, Germany Received April 6, 2004
Abstract: About 25% of open reading frames in fully sequenced genomes are estimated to encode transmembrane proteins that represent valuable targets for drugs. However, the global analysis of membrane proteins has been proven to be problematic, e.g., because of their very amphiphilic nature. In this paper, we show that the recently published Protein Sequence Tag (PST) technology combined with an efficient sample preparation is a powerful method to perform protein analysis of highly enriched membrane fractions. The PST approach is a gelfree proteomics tool for the analysis of proteins, which relies on a “sampling” strategy by isolating N-terminal protein sequence tags from cyanogen bromide cleaved proteins. The identification of these N-terminal PST peptides is based on LC-MS/MS. The effectiveness of the technology is demonstrated for a membrane fraction, which was isolated from crude mitochondria of yeast after alkaline sodium carbonate treatment. The PST approach performed on this fraction analyzed 148 proteins, whereas 84% are identified as membrane proteins. More interestingly, among these membrane proteins 56% are predicted to be of low abundance. These encouraging results are an important step toward the development of a quantitative PST approach (qPST) for the differential display of membrane protein analysis. Keywords: Protein Sequence Tag • membrane proteins • low abundant protein • gel-free approach • proteomics
Introduction Conventional techniques for proteomic investigations such as 2-D gel electrophoresis (2-DE) do not adequately represent hydrophobic proteins.1 However, among these are membrane proteins that represent more than 50% of known drug targets.2 Therefore, proteomics approaches for both qualitative and quantitative analyses of membrane proteins are of great interest for drug discovery programs. Gel-free proteomics approaches using mass spectrometric analysis of peptide digests separated by liquid chromatography3-7 are emerging as powerful complementary tools for the analysis of complex protein mixtures, which has the potential to overcome the limitations of 2-DE * To whom correspondence should be addressed. christian.hamon@ proteomics.com. 10.1021/pr049925u CCC: $27.50
2004 American Chemical Society
based approach with respect to hydrophobic proteins.4,5,8,9 Among them, ‘sampling’ techniques represent valuable methods to reduce complexity and therefore maximize analytical capacity of LC-MS systems while retaining sufficient information about the original sample to identify a majority of its components.5,10,11-13 Thus, the Isotope Coded Affinity Tag (ICAT) procedure was the first introduced chemical derivatization procedure to capture and analyze specifically cysteine residue containing peptides for proteomics studies.6 It was reported that on average 5 cysteine residue containing peptides are available per protein in S. cerevisiae and that 92% of proteins have at least one cysteine residue. Recently, a diagonal method was developed to isolate N-terminal peptides from entire proteins. This theoretically reduces the complexity of the analysis to one peptide per protein for each protein present in the biological sample.14 More recently, we introduced the Protein Sequence Tag technology (PST), which obtains Nterminal peptides from each fragment generated by CNBr cleavage of a polypeptide mixture.9 These N-terminal peptide fragments are then amenable to analysis by two-dimensional chromatography coupled to MS/MS. Thus, this technique samples peptides based on the distribution of methionine. By way of comparison with the above-described techniques, the PST process obtains up to 8 peptides per protein on average from S. cerevisiae and 99% of proteins have at least one peptide available per protein. In that way, almost all the proteins can be analyzed and the moderate degree of redundancy increases confidence in protein identifications and the chance to isolate at least one peptide from the protein compatible for identification in MS/MS analyses. A principal objective of the PST approach is the identification of hydrophobic proteins. Taking advantage of the solubilization step using cyanogen bromide, the procedure allows the study of complex mixtures of hydrophobic proteins, particularly membrane proteins which has been confirmed by a previous study with a crude mitochondrial fraction.9 In the view to further demonstrate the effectiveness of PST toward the protein identification of highly enriched hydrophobic fractions, we analyzed a membrane fraction of crude mitochondria isolated from S. cerevisiae. The mitochondrial membranes contain both R-helical and β-barrel membrane proteins and represent, therefore, optimal materials for such general investigation. Thus, the ability of the PST technology to work on pure membrane fractions is a prerequisite to establish a differential display strategy for membrane protein analysis related to the future development of the quantitative PST (qPST). Journal of Proteome Research 2004, 3, 1073-1081
1073
Published on Web 07/24/2004
PST for Membrane Protein Samples
Experimental Procedures Preparation of a Pure Membrane Fraction. The crude mitochondrial fraction from S. cerevisiae strain W303A15 grown in YPD (Yeast Peptone Dextrose) medium at 30 °C was isolated by differential centrifugation essentially as described16,17 except that bovine serum albumin was left out of the homogenization buffer and mitochondria were finally taken up in a buffer containing 20 mM Na2HPO4/NaH2PO4 (pH 7.2) and 250 mM saccharose. Further purification of the membrane fraction by alkaline Na2CO3 treatment was carried out as published before18,19 except that the membrane fraction was collected by centrifugation at 100 000 g and 4 °C for 45 min. To evaluate the successful separation of the membrane fraction from the soluble fraction both fractions were subjected to Tris-tricine SDS-PAGE20 and immunodecoration with antisera directed against DnaK (Hsp70) and mitochondrial porin protein. PST Procedure. The subsequent PST process was performed on the isolated membrane fraction derived from 10.5 mg crude mitochondria. Thus, cyanogen bromide (CNBr; 5 M in acetonitrile) was added to the sample dissolved in 3 mL formic acid (90%, v/v) at a final concentration of 1 mg CNBr per mL. After the cleavage reaction (for 24 h in the dark), the mixture was diluted to 40 mL with distilled water, frozen, and lyophilized. 1.2 mg recovered CNBr cleaved proteins dissolved in 760 µL of a denaturing buffer (2 M urea, 1 M thiourea, 2 M guanidine hydrochloride and 200 mM borate buffer, pH 7.2) were then subjected to reduction and alkylation of Cys residue with tris[2-Carboxyethylphosphine] (TCEP) and iodoacetamide (IAA). Free amino groups were then blocked using our in-house developed Basic Mass Tag (BMT) label (N,N-dimethylglycine N-hydroxysuccinimide ester). BMT (8.7 mg dissolved in 335 µL DMF) was added to the polypeptide mixture and the reaction mixture was stirred at room temperature (RT). After 3 h, the same amount of BMT dissolved in 35µL DMF was added again to the mixture and incubated overnight. The polypeptide mixture was purified by size exclusion centrifugation using centrifugal filter devices with Molecular Weight Cut Off (MWCO) of 5000 (Centricon, Millipore Inc.). The enzymatic digestion was performed by adding 40 µg trypsin (Promega) to the recovered polypeptide solution (750 µg in 700 µL) for 24 h at 37 °C and pH 7.8. The isolation of the N-terminal peptides was finally obtained by a scavenger bead mediated separation by using half of the amount of digested peptides. Thus, the digest peptide mixture was diluted to 600 µL with borate buffer (50 mM) and adjusted to pH 7.2. This solution was added to a 200 mg portion of scavenger beads suspended in 2400 µL DMF. The scavenger beads were prepared directly before use by reacting a carboxy-modified resin (Polystyrene AM COOH resin from Rapp Polymere, Germany) with a solution of 175 mg N-hydroxysuccinimide in DMF and 233 µL N,N′-diisopropylcarbodiimide to the freshly swollen beads in DMF for 3 h at RT. After 18 h vigorous shaking of the mixture at RT, the filtrated peptide-containing solution was transferred to a second 200 mg portion of activated beads. After additional 6 h incubation, the solution was filtered off, and the solvent was evaporated. PST Peptide Fractionation Using Cation Exchanger. Sample fractionation using Strong Cation Exchange (SCX) was introduced prior to the LC-MS/MS analysis. Basically, the remaining residue dissolved in 1500 µL water:acetonitrile 95:5 + 0.1% (v/v) trifluoro acetic acid was loaded onto a SCX column (selfpacking cartridge filled with S-Sepharose “Fast Flow”, Fluka). After washing the column with water:acetonitrile 95:5 + 0.1% 1074
Journal of Proteome Research • Vol. 3, No. 5, 2004
technical notes (v/v) trifluoro acetic acid, the peptides were eluted with an salt buffer (ammonium acetate in water:acetonitrile 95:5, pH 2.0) using a step gradient (50-250 mM salt) to obtain 6 fractions. After lyophilization, each residue was dissolved in 100 µL water: acetonitrile 95:5 and 0.1% (v/v) trifluoro acetic acid. Mass Spectrometry Analysis. A Surveyor HPLC System (ThermoFinnigan) and a LCQ Deca (ThermoFinnigan) was used to perform the analyses. A 20 µL of each SCX fraction were loaded and the peptides were eluted from a reversed-phase C18 column (Dionex, 150 mm × 75 µm, PepMap C18) using a binary solvent gradient (solvent A: acetonitrile, 0.1% formic acid; solvent B: water, 0.1% formic acid). The column was connected to a PEEK Mixing-T in order to split the flow of the HPLC pump to an effective flow rate of 0.18-0.22 µL/min. At the ion source, a spray voltage of 1.0 kV was supplied. The overall time of the LC method is 130 min including conditioning step, elution gradient, and wash step. Tandem Mass Spectral (MS/MS) analysis for peptide identification was performed with dynamic exclusion mode, repeat count set at one. MS and MS/MS spectra were collected as centroided data. Peptide and Protein Identification. Peak List Creation. DTA files were allowed to be created from MS raw data when the Total Ion Current (TIC) of 1e4 was fulfilled with a threshold set at 35 ions. In the grouping criteria, Minimum Allowed Intervening Scans (Group Scan) was set at 5 with a mass tolerance of 1.4 Da. SEQUEST Analyses. SEQUEST21 analyses of the DTA files were carried out as follows: The data were subjected to two separate sets of SEQUEST analysis parameters. 1st run: Cys +57 (static), Lys and His +85 (differential), Nter peptide +85; 2nd run: Cys +57 (static), Lys and His +85 (differential), Nter protein +42. In both runs, cleavage sites for in silico digest were defined after M, K, and R, up to 5 miscleavages were allowed. The allowed mass deviation for precursor ions was set to 3.0 Da and for fragment ions to 0.5 Da. Only 4-fold charged peptides with cross-correlation scores (Xcorr) > 3.0, triply charged peptides with Xcorrs > 2.5, doubly charged peptides with Xcorr > 1.9 and singly charged peptides with Xcorrs > 1.5, all having a ∆Cn g 0.08, were investigated further by a detailed analysis. In special cases, ∆Cn values
