Introduction

The advent of proteomic technologies is a major step forward in identifying and characterizing proteins by protein chemical methods rather than by immunochemical techniques, which are dependent on antibody availability and specificity (Box 1)¹.

Advanced proteomics technologies, offering the generation of reference databases, partial proteomes from tissues and cells, and protein profiling, are effective methods for concomitant identification and determination of proteins in a given sample^{2, 3, 4, 5, 6, 7, 8, 9, 10}. Although current technology allows protein identification in the high-throughput mode, only relatively small protein profiles of individual samples are published, and identification rates are far from 100%.

There are many reasons for the limited identification rates. One serious limitation that hampers high-throughput identification is that one MS method is either not able to identify all proteins from a gel because of inherent limitations of the method's principle or sensitivity, or it is too time consuming for identification of large sample size. This analytical problem holds even for highly soluble proteins; the method proposed herein is for the identification of hydrophilic and soluble proteins, mainly detecting proteins from metabolic pathways, the cytoskeleton, proteins from several signaling cascades, chaperones, protein synthesis and degradation (e.g., proteasomal structures⁶) (Fig. 1).

Figure 1: Functional classes of 992 protein spots identified from a map of the rat hippocampus.

Identified proteins were categorized into metabolic, signaling, neuronal and the like, according to their assignment in the SWISS-PROT database (http://expasy.org/) and literature (http://pubmed.com/).

Full size image (36 KB)

Protein classes identified by using this protocol in rat hippocampus, as described in this protocol, could be observed in all other tissues as well and would demonstrate similar results with the exception of tissue- or cell-specific proteins (e.g., synaptosomal, neuronal proteins). However, specific protein patterns would be expected to be generated from individual tissues. In contrast to immunochemical techniques, the method used permits detection of hypothetical proteins (Supplementary Table 1), such as structures that have so far only been described at the nucleic acid level; thus, their existence can be proven at the protein level¹. A series of proteins with unknown function that are not detected by other approaches (Supplementary Table 1) can be demonstrated in most partial proteomes using MS methods.

Major advantages of the described combination of methods include fair protein chemical identification, generation of high protein identification rates, concomitant analysis of the identical spot in a Coomassie-stained 2D gel by two independent MS principles, and de novo sequencing of interesting spots, in particular if sequence conflicts with databases occur³. The de novo sequencing method used in our laboratory is not described as proteins were fairly identified but can be applied to the proposed method when required. Although even small spots were identified, larger spots can be re-picked and MS identification carried out successfully from the gel stored at 10 °C for several months. The method is also suitable for determination of many post-translational modifications, but description of that analysis would exceed the goal of this protocol¹¹.

Partial proteomes resulting from protein profiling can be used for quantitative comparative proteomics. Of course, appropriate pre-fractionation, by ultracentrifugational compartmentalization of subcellular fractions¹², chromatography¹³, solution isoelectric focusing (IEF)^{9, 14}, Blue native SDS-PAGE¹⁵ and preparative gel electrophoresis⁷, to name a few, must be carried out before running the gels used for the experiments.

A flow diagram of the protocol is shown in Figure 2.

Figure 2

Flow diagram of protein profiling by the combination of two independent mass spectrometry techniques.

Full size image (30 KB)

Materials

Reagents

• Rats
  Caution Comply with relevant institutional and national guidelines regarding treatment of rats.
• HPLC-grade water (Millipore)
• CHAPS (Bio-Rad; cat. no. 161-0460)
• DTT (Bio-Rad; cat. no. 161-0611)
• EDTA (Sigma; cat. no. 431788-100G)
• Urea (Sigma; cat. no. 431788-100G)
• Thiourea (Sigma; cat. no. T8656-500G)
• PMSF (Sigma; cat. no. 78830-5G)
• Protease inhibitor mixture Complete (Roche; cat. no. 11697498001)
• Immobiline pH 3–10 nonlinear IPG buffer (Amersham Bioscience; cat. no. 17-6000-88)
• Tributylphosphine (TBP) (Sigma; cat. no. T49484-5ML)
• SDS (Sigma; cat. no. L6026-250G)
• Acrylamide/piperazine-di-acrylamide (PDA) (Bio-Rad; cat. no. 161-0108/161-0202)
  Caution Acrylamide and PDA are toxic. When handling these chemicals, wear gloves and use a pipetting aid.
• Bromophenol blue stock
• Agarose
• Acetone
• Glycerol (Sigma; cat. no. G8773-1L)
• Iodoacetamide (Bio-Rad; cat. no. 163-2109)
• Precision Plus Protein Standards (Bio-Rad; cat. no. 161-0363)
• Colloidal Coomassie blue staining kit (Invitrogen; cat. no. LC6025)
• Trifluoroacetic acid (TFA) (Sigma; cat. no. 431788-100G)
• Ammonium bicarbonate (Sigma; cat. no. 40867-50G-F)
• Sequencing-grade trypsin (Promega; cat. no. V5111)
• Octyl -D-glucopyranoside (OGP) (Sigma; cat. no. 75083-5G)
• n-heptane (Sigma; cat. no. IRMM441)
• -cyano-4-hydroxy-cinnamic acid (CHCA) (Sigma; cat. no. C8982-10X10MG)
• Peptide standard for calibration (Bruker Daltonics)
• CHROMASOLV grade isopropanol (Sigma; cat. no. 4959-1L)
• LC-MS CHROMASOLV acetonitrile (ACN) (Sigma; cat. no. 34967-1L)
  Caution ACN is toxic. When handling, wear gloves and use a pipetting aid.
• LC-MS CHROMASOLV methanol (Sigma; cat. no. 34966-1L)
  Caution Methanol is toxic. When handling, wear gloves and use a pipetting aid.
• HPLC-grade formic acid (Sigma; cat. no. 27001-500ML-R)
• Cytochrome c digest (Dionex; cat. no. 161089)

Equipment

• Sonicator (Bransonic; cat. no. 55AOE-MTH)
• Immobilized pH 3–10 nonlinear gradient strips (Amersham Bioscience; cat. no. 17-1235-01)
• Amicon Ultra-4 Centrifugal Filter Unit (Millipore ; cat. no. UFC8 010 24)
• Ettan IPGphor II IEF system (Amersham Bioscience)
• Protean electrophoresis system (Bio-Rad)
• PROTEINEER spII and PROTEINEERdp (Bruker Daltonics)
• AnchorChip 600/384 target (Bruker Daltonics)
• UltraFlexII matrix-assisted laser desorption/ionization–time of flight/time of flight mass spectrometer (MALDI-TOF/TOF) (Bruker Daltonics)
• Flexcontrol software (Bruker Daltonics)
• Flexanalysis software (Bruker Daltonics)
• BioTools 2.3 software (Bruker Daltonics)
• Mascot software (Matrix Science Ltd.)
• Microsoft PowerPoint (Microsoft)
• Protein lobind tube (Eppendorf; cat. no. 0030 108.094)
• Eppendorf SpeedVac Concentrator (Eppendorf; cat. no. 5301 000.210)
• 250-l HPLC vial (Dionex; cat. no. 6820.0029)
• Ultimate 3000 nano HPLC system (Dionex)
• Chromelon 6.7 software (Dionex)
• Qstar XL system (Applied Biosystems)
• AnalystQS 1.1 software (Applied Biosystems)
• Mascot.dll 1.6b21 software (Matrix Science Ltd.)
• C18 PepMap100 solid-phase extraction -Precolumn cartridge (particle size 5 m, pore size 100 Å, 300 m inner diameter) (Dionex; cat. no. 160454)
• Nano-column PepMap C18 reversed-phase material (particle size 3 m, pore size 100 Å, 75 m inner diameter) (Dionex; cat. no. 160321)
• Emitter needle tip (no coating, 20 m inner diameter) (New Objective; cat. no. FS360-20-10-N-20-C10.5)

Reagent setup

• Sample buffer 20 mM Tris, 7 M urea, 2 M thiourea, 4% CHAPS (wt/vol), 10 mM DTT, 1 mM EDTA, 1 mM PMSF and 1 tablet Complete in 2 ml sample buffer
• Rehydration buffer 8 M urea, 4% CHAPS (wt/vol), 10 mM DTT, 0.5% IPG buffer (wt/vol)
• Equilibration buffer 6 M urea, 2% SDS (wt/vol), 0.5 M Tris HCl (pH 8.8), 20% glycerol (wt/vol)
• Solution A 0.1% formic acid (wt/vol) in water
• Solution B 80% ACN–0.08% formic acid (wt/vol) in water

Equipment setup

• Nano-LC-MS/MS system An Ultimate 3000 nano-LC system is connected to Qstar XL to set up the nano-LC-MS/MS system. The flow rate of the loading pump is 20 l/min, and the flow rate of the micro pump is 300 nl min^{- 1}. Load sample to a Pepmap100 C18 Precolumn from 0 min to 4 min, and then separate on PepMap100 C18 nano-column with a gradient of 4% solution B to 60% from 0 min to 30 min, 90% solution B constant from 30 min to 35 min, and 4% solution B from 35 min to 60 min. Ion spray voltage is 3200 V, and gas is set at 10. The mass spectrum is recorded from 10 min to 50 min for a 7-s cycle with one 1-s MS [following three 2-s MS/MS fragmentation of highest intensity precursor ions.]
  Caution Acrylamide, PDA, ACN and methanol are toxic. When handling these chemicals, wear gloves and use a pipetting aid.

Procedure

1. Points from here (point 1) up to and including point 11 are related to 2D gel electrophoresisTiming: 1 weekKill rats, dissect the hippocampus within 1.5 min, put the samples in liquid nitrogen and keep at - 80 °C. Powderize the rat hippocampus in a mortar in liquid nitrogen, and suspend about 2 mg powder in 2 ml sample buffer.
2. Sonicate the suspension five times on ice for 30 s, and centrifuge at 15,000 g for 60 min at 12 °C. Transfer the supernatant to an Ultrafree-4 centrifugal filter tube.
   Critical step STEP When transferring protein solution to the filter tube after centrifugation at 15,000 g for 60 min at 12 °C, be sure not to include any surface lipids or bottom pellets, because that will interfere with the 2DE gel pattern.
3. Centrifuge at 2,500 g until the remaining volume is 500 l.
4. To desalt, add 1 ml rehydration buffer, and centrifuge at 2,500 g repeatedly until the eluted volume is 4 ml. Concentrate the sample by centrifuging until the sample volume is 200 l.
5. Determine the protein concentration of the supernatant by a Bradford assay¹⁶.
6. Use rehydration buffer to dilute the sample to 750 g protein in 346.5 l, add 3.5 l 200 mM TBP (5 l TBP + 95 l isopropanol) and apply to 18-cm immobilized pH 3–10 nonlinear gradient strips.
7. Start focusing at 200 V, increase the voltage gradually to 8,000 V at 4 V min^{- 1} and keep constant for an additional 3 h (150,000 V h in all).
8. Equilibrate IPG strips in equilibration buffer with 2% DTT for 15 min and then in equilibration buffer with 2.5% iodoacetamide (wt/vol) instead of DTT for another 15 min.
9. Put the equilibrated IPG strips on the top of a 10–16% gradient SDS-PAGE. Seal and fix the strips using boiled agarose solution comprising 1% agarose and 0.5% bromophenol blue stock in electrophoresis buffer. After running at 50 V for 30 min, keep the running voltage at 200 V for 7 h until the dye front line reaches the edge of the gel. Run standard protein markers ranging from 10 kDa to 250 kDa to determine molecular masses.
   Critical step Fix the parameters of the gel casting system to get a reproducible gradient. No air bubbles should exist between the strip and gel surface to obtain high-quality gel patterns.
10. Fix for 12 h in 50% methanol (vol/vol) and 10% acetic acid (vol/vol).
11. Stain gels with colloidal Coomassie blue for 8 h, and wash away excess dye with distilled water.Pause Point Gels can be kept in 0.02% sodium azide (wt/vol) and stored for several months at 10 °C by being sealed in a plastic bag.Troubleshooting
12. Points from here (point 12) up to and including point 13 are related to Spot excision and in-gel digestionTiming: 3 dExcise spots using a spot picker PROTEINEER spII, and place punched gel pieces into 96-well microtiter plates with pierced well bottoms.
13. Perform in-gel digestion by an automated procedure using PROTEINEER dp. Briefly, wash spots with 10 mM ammonium bicarbonate and 50% ACN in 10 mM ammonium bicarbonate. After washing, shrink gel plugs by adding ACN, and dry by blowing out the liquid through the pierced well bottom. Re-swell the dried gel pieces with 40 ng l^{- 1} trypsin in digestion buffer (consisting of 5 mM OGP and 10 mM ammonium bicarbonate), and incubate for 4 h at 30°C. Extract peptides from gel plugs with 15 l of 1% TFA (vol/vol) in 5 mM OGP for 30 min.Pause Point Samples in a 96-well plate can be kept at - 20 °C for several weeks if wrapped tightly with parafilm after extraction.
    Critical step Pick strong spots initially to get a better identification rate using MALDI-TOF/TOF, because the laser energy parameter in the automatic run is designed to obtain a high-quality spectrum for either strong or weak spots at the same time.
14. Points from here (point 14) up to and including point 20 are related to MALDI-TOF/TOF identificationTiming: 1 weekWipe the AnchorChip target with acetone and n-heptane first, then sonicate in isopropanol for 15 min, and finally in HPLC-grade water for another 15 min.
15. Prepare the matrix thin layer by dragging a 5-l matrix solution (50 mg CHCA in 20 l 0.1% TFA (vol/vol) and 1,980 l acetone) droplet over 12 spots.
16. Directly apply 4 l extracted peptides onto the target, and incubate for 3 min.
17. Use 10 l 0.1% TFA (vol/vol) to wash spots for desalting.
18. Apply 2 l peptides standard onto the target for calibration, and remove after 3 min.
19. Use Flexcontrol to control UltraFlexII. Analyze samples by one peptide mass fingerprinting (PMF) from MALDI-TOF, followed by additional LIFT-TOF/TOF MS/MS analysis of three peptides. Accumulate data from 200 consecutive laser shots to produce PMF and MS/MS spectra. Acceleration voltage should be 25 kV for PMF. For MS/MS, acceleration voltage should be 8 kV in the TOF1 stage to promote metastable fragmentation. After selection of jointly migrating parent and fragment ions in a timed ion gate, lift ions by 19 kV to high potential energy in the LIFT cell. Use nitrogen as the dry gas. Use peptide standard as an external calibration. Exclude autoproteolysis products of trypsin when doing data-dependent LIFT MS/MS. The m/z range is 700–4,000 for PMF and 40–2,560 for MS/MS.
    Critical step The ion source should be cleaned every month to prevent the spectrum quality from declining. When preparing the thin layer, discard the matrix droplet when it becomes turbid.
20. Interpret PMF and MS/MS spectra primarily with the Flexanalysis software. Use the autoproteolysis products of trypsin as internal calibration. Set the signal-to-noise ratio threshold at 3. By using the Mascot search engine, perform database searching against the MSDB 20051115 database (ftp://ftp.ncbi.nih.gov/repository/MSDB/) to identify protein spots, using combined PMF and MS/MS data sets via BioTools 2.3 software (Bruker Daltonics). The searching parameters should be set as follows: a mass tolerance of 25 p.p.m. is allowed for PMF, and in case of MS/MS a tolerance of 0.5 Da is accepted; both allow one missing cleavage site and consider fixed modification of carbamidomethyl (C) and variable modification of oxidation of methionine residues.Troubleshooting
21. Points from here (point 21) up to and including point 24 are related to Nano-LC-MS/MS identificationTiming: 1–2 weeksTransfer the remaining extracted peptides of spots not identified in MALDI-TOF/TOF to 0.5-l protein lobind tubes. Extract those spots two more times with 15 l 0.1% formic acid and 4% ACN, then pool them. Reduce the volume in a SpeedVac to about 8 l. Do not dry to completeness, because this will substantially reduce the yield of peptides.Pause Point After this step, samples can be frozen at - 20 °C for several weeks.
22. Use Chromelon 6.7 to control the Ultimate 3000 nano-LC system. Inject 6 l sample, and separate in a PepMap100 C18 nano-column as shown in the Equipment Setup section.
23. Use AnalystQS 1.1 to control QSTAR XL. Record the MS spectra over the mass range of m/z 350–1,600, and MS/MS spectra in information-dependent data acquisition over the mass range of m/z 50–1,600. Repeatedly, one MS spectrum is generated followed by three MS/MS spectra on the QSTAR XL instrument; the accumulation time is 1 s for MS spectra and 2 s for MS/MS spectra. The collision energy is set automatically according to the mass and charge state of the peptides chosen for fragmentation. Choose doubly or triply charged peptides for MS/MS experiments because of their good fragmentation characteristics.
    Critical step The length of the exposed part of the ion spray tip should be 2 mm. Adjust the position of the ion spray tip, ion spray voltage and gas carefully to obtain a high intensity before analyzing actual samples. Clean the ion source about every 50 samples to prevent the spectrum quality from declining. Make the connection of nano-LC and Qstar XL as short as possible to keep separated peptides from becoming merged before ionization. Make zero dead volume connection for every connection.
24. Use Mascot.dll 1.6b21 in AnalystQS to interpret MS/MS spectra, and search against the MSDB 20051115 database (ftp://ftp.ncbi.nih.gov/repository/MSDB/) to identify proteins. The searching parameters should be set as follows: a mass tolerance of 500 p.p.m. for peptide tolerance, 0.15 Da for MS/MS tolerance, one missing cleavage site, fixed modification of carbamidomethyl (C) and variable modification of oxidation of methionine residues.Troubleshooting
25. Points from here (point 25) up to and including point 26 are related to Data processing and miningTiming: 3 weeksGenerate a list of identified proteins by taking the protein accession number, MS score, matched peptides, MS sequence coverage, MS/MS score, MS/MS sequence coverage, MS/MS peptides from the Mascot result files. Categorize identified proteins into metabolic, signaling, neuronal and the like, according to their assignment in the SWISS-PROT database (http://expasy.org/) and literature (http://pubmed.com/). Proteins' theoretical molecular weight, isoelectric point and GRAVY value can be calculated on the ExPASy Proteomics server http://expasy.org/. The transmembrane domain numbers can be calculated by TMHMM (http://www.cbs.dtu.dk/services/TMHMM/), ConPredII (http:/bioinfo.si.hirosaki-u.ac.jp/~ConPred2/) and TopRed (http://bioweb.pasteur.fr/seqanal/interfaces/toppred.html).
26. Produce a map of identified proteins using Microsoft PowerPoint by annotating the accession numbers on the 2DE gel image referring to spot-picking images.

Timing

Running 12 2DE gels including staining and destaining, requires 5 d, although a full days' work is not necessary on each day. Picking, proteolytic digestion (all by robots) of 1,000 spots, normally well separated on a 2DE gel and therefore representing well-defined spots, can be done by using three targets containing 384 samples each. Three targets can be run within 6 d (MALDI-TOF, MALDI-TOF/TOF) by an automated procedure. Manual re-analysis of unidentified samples, when necessary, requires 6 d for 100 samples. One should allow 10 d to run the sample aliquots for any remaining unidentified proteins by nano-LC-MS/MS. In our experience this is necessary for 20% of all protein spots. The raw data obtained from database searches must be handled; a list of proteins with identification criteria and the like, as well as a map of proteins with accession numbers must be constructed. Preparation of the list and map construction (to refine the map and table into a publishable form) require 2 weeks.

Troubleshooting

There is a multitude of possible mistakes at all phases of the protocol, and these cannot be addressed here. The reader is referred to the competent troubleshooting manuals provided by the suppliers of 2DE, MALDI, LC, Q-TOF and protein databases.

There is a multitude of possible mistakes at all phases of the protocol, and these cannot be addressed here. The reader is referred to the competent troubleshooting manuals provided by the suppliers of 2DE, MALDI, LC, Q-TOF and protein databases.

There is a multitude of possible mistakes at all phases of the protocol, and these cannot be addressed here. The reader is referred to the competent troubleshooting manuals provided by the suppliers of 2DE, MALDI, LC, Q-TOF and protein databases.

Anticipated results

We applied 750 g of rat hippocampal proteins on a pH 3–10 nonlinear IPG strip. After focusing in the first dimension, proteins were separated on a 10–16% gradient SDS-PAGE. The gel was stained by colloidal Coomassie blue with a detection limit of 10 ng. The typical 2DE pattern is shown in Figures 3 and 4 with annotations of identified proteins. We excised 1,056 spots, digested them with trypsin and identified them by MALDI-TOF/TOF first; then we identified those not identified in the MALDI system by nano-LC-MS/MS. As a standard protein, 1 pmol cytochrome c digest was analyzed on nano-LC-MS/MS as shown in Supplementary Figure 1. Typical MALDI-TOF/TOF and nano-LC-MS/MS spectra for 2DE gel spots identification are shown in Supplementary Figures 2 and 3. To significantly determine the confidence in the reported protein identifications, a score criterion of MASCOT score was considered significant at P<0.05 (i.e., score >63 for MS, score >20 for MS/MS). According to this criterion, 992 spots (811 by MALDI and 181 by LC-MS/MS) out of 1,056 picked spots were identified by searching against the MSDB database with Mascot software, yielding a protein identification success rate of 94%. The identified proteins are listed in Supplementary Table 1.

Figure 3: Two-dimensional map of rat hippocampal proteins identified by MALDI-TOF/TOF.

Rat hippocampal proteins were extracted, and 750 g were applied on an immobilized pH 3–10 nonlinear gradient strip, followed by 10–16% linear gradient polyacrylamide gel. Spots, identified by MALDI-TOF/TOF, were labeled with SWISS-PROT accession numbers. Because of readability of 881 spots on one map, we divided it into three maps (a–c).

Full size image (93 KB)

Figure 4: Two-dimensional map of rat hippocampal proteins identified by nano-LC-MS/MS.

Rat hippocampal proteins were extracted, and 750 g were applied on an immobilized pH 3–10 nonlinear gradient strip, followed by 10–16% linear gradient polyacrylamide gel. Spots not identified by MALDI-TOF/TOF were identified by nano-LC-MS/MS and labeled with SWISS-PROT accession numbers.

Full size image (78 KB)

The remaining 6% were not identified, because (i) several proteins were identified in one spot and (ii) no spectra were generated, most probably because of sensitivity (very faint Coomassie-stained spots) or technical shortcomings.

We here present a successful identification method of spots from 2DE gels, forming the basis for advanced proteomic analyses. The improvement of identification success rate is attributed to the greater protein coverage obtained by nano-LC-MS/MS and also to the combination of two different kinds of ionization methods (MALDI and electrospray ionization)^{10, 17}. We use MALDI-TOF/TOF to carry out first identification as a fast screening method, then apply nano-LC-MS/MS to identify those spots that cannot be identified in the first step. The advantage of the protocol is that it is suitable for high-throughput reliable identification of proteins after separation of complex protein mixtures (e.g., hippocampal tissue) in terms of relatively rapid analysis that had not been previously published. By this strategy we combined the high throughput of MALDI-TOF/TOF and the high identification success rate of nano-LC-MS/MS.