Proteomic Analysis by Multidimensional Protein Identification Technology

Abstract

Multidimensional chromatography coupled to mass spectrometry is an emerging technique for the analysis of complex protein mixtures. One approach in this general category, multidimensional protein identification technology (MudPIT), couples biphasic or triphasic microcapillary columns to high-performance liquid chromatography, tandem mass spectrometry, and database searching. The integration of each of these components is critical to the implementation of MudPIT in a laboratory. MudPIT can be used for the analysis of complex peptide mixtures generated from biofluids, tissues, cells, organelles, or protein complexes. The information described in this chapter will provide researchers with details for sample preparation, column assembly, and chromatography parameters for complex peptide mixture analysis.

1 Introduction

The integration of multidimensional chromatography and mass spectrometry as a proteomic tool has developed into a mature technology. Numerous two-dimensional (2D) chromatographic methodologies have been developed in order to resolve peptide mixtures prior to analysis by mass spectrometry (reviewed in ref. 1). One technique, multidimensional protein identification technology (MudPIT), has proven itself a powerful method to study highly complex protein samples. In MudPIT, a bi- or triphasic microcapillary column packed with reversed-phase (RP) and strong cation exchange (SCX) high-performance liquid chromatography (HPLC)-grade materials is loaded with a complex peptide mixture generated from a biological sample (2–5). The microcapillary column is interfaced with a quaternary HPLC pump coupled to a tandem mass spectrometer, and acts as the ion source. Peptides are directly eluted off of the microcapillary column, ionized, and analyzed in the tandem mass spectrometer, which is capable of fragmenting peptides in a predictable fashion, which allows for the computational determination of the peptide sequence by database-searching algorithms like SEQUEST (6). Utilizing MudPIT, researchers are able to measure protein levels of whole proteomes (7–9), differential protein expression in response to variable growth conditions (10), membrane proteins (11), and large multi-protein complexes (12,13)

2 Materials

2.1 Sample Preparation

• 1. Benzonase (Sigma, St. Louis, MO).

• 2. Trichloroacetic acid (TCA).

• 3. Digestion buffer: 100 mM Tris-HCl (pH 8.5) (stored at 4°C), 8 M urea (added fresh).

• 4. Tris(2-carboxylethyl)-phosphine hydrochloride (TCEP) (1 M stock, stored at −20°C, diluted 1/10) (Pierce, Rockford, IL).

• 5. Iodoacetamide (IAM), 0.5 M stock (Sigma, St. Louis, MO).

• 6. Endoproteinase LysC (Roche Applied Science, Indianapolis, IN).

• 7. Calcium chloride, 500 mM stock.

• 8. Trypsin, modified sequencing grade: 0.1 μg/μL stock in water at −20°C (Roche Applied Science, Indianapolis, IN).

• 9. 90% formic acid.

• 10. Elastase stock solution: 10 μg/μL in 10 mM Tris-HCl (pH 8.5), stored at −20°C (Calbiochem, San Diego, CA).

• 11. Subtilisin A stock solution: 10 μg/μL in 10 mM Tris-HCl (pH 8.5), stored at −20°C (Calbiochem, San Diego, CA).

• 12. Proteinase K stock solution: 15 μg/μL in water, stored at −20°C (Roche Applied Science, Indianapolis, IN).

2.2 Microcapillary Column Construction and Sample Loading

• 1. 100 μm i.d. × 365 μm o.d. and 250 μm i.d. × 365 μm o.d. polyimide-coated fused silica (Polymicro Technologies, Phoenix, AZ).

• 2. Model P-2000 Laser Puller (Sutter Instrument Co. Novato, CA).

• 3. Stainless steel pressurization device (Brechbuehler, Inc., Houston, TX, or MTA for blueprints available by request from John Yates, Scripps Research Institute, La Jolla, CA).

• 4. Five-micrometer C₁₈ Aqua Reversed Phase Packing Material (Phenomenex, Torrance, CA).

• 5. 5-μm Partisphere SCX packing material (Whatman, Florham Park, NJ).

• 6. M-520 Inline Micro Filter Assembly and F-185 Microtight 0.0155 × 0.025 Sleeves (UpChurch Scientific, Oak Harbor, WA).

2.3 Multidimensional Chromatography and Tandem Mass Spectrometry

• 1. Buffer A: 5% acetonitrile, 0.1% formic acid, made with HPLC-grade water.

• 2. Buffer B: 80% acetonitrile, 0.1% formic acid, made with HPLC-grade water.

• 3. Buffer C: 500 mM ammonium acetate, 5% acetonitrile, 0.1% formic acid, made with HPLC-grade water.

• 4. 50 μm i.d. × 365 × μm o.d. polyimide-coated fused silica (Polymicro Technologies, Phoenix, AZ).

• 5. P-775 MicroTee Assemblies and F-185 Microtight 0.0155 × 0.025 Sleeves (UpChurch Scientific, Oak Harbor, WA).

• 6. Gold wire (Scientific Instrument Services, Inc., Ringoes, NJ).

• 7. Agilent 1100 series G1379A degasser, G1311A quaternary pump, G1329A autosampler, G1330B autosampler thermostat, and G1323B controller (Agilent Technologies, Palo Alto, CA).

• 8. LCQ DECA-XP^plus tandem mass spectrometer (Thermo Electron, San Jose, CA).

• 9. Nano electrospray stage (MTA for blueprints available by request from John Yates, Scripps Research Institute, La Jolla, CA). Other options include the Thermo Electron Nanospray II ion source or PicoView Sources from New Objective (Woburn, MA).

3 Methods

3.1 Sample Preparation

A wide variety of samples have been analyzed via MudPIT, including protein complexes (13), organelles (14), whole proteomes (7,8), and biofluids (15). The isolation of protein mixtures from any of these samples is beyond the scope of this article. It is important to have a good idea what the total amount of protein in a sample is, however, in order to add the proper amount of proteases for digestion. In general, all of these samples from disparate sources will be analyzed in the fashion outlined in this chapter, with the only variation being that if a large amount of proteins is in the starting sample, it will be split into 200- to 300-μg aliquots and each fraction will be TCA precipitated independently and then recombined after resuspension for protease digestion. The standard approach for the detection and identification of proteins is the endoproteinase LysC/trypsin digestion protocol. For the identification of posttranslational modifications, high sequence coverage of individual proteins is needed. In this case, samples will be split into at least three fractions and independently digested with endoproteinase LysC/trypsin, elastase, and subtilisin in a similar fashion to the protocol published by MacCoss et al. (16). The highpH proteinase K digestion protocol can also be used to generate high sequence-coverage data from complex protein mixtures for post-translational modification analysis (14).

3.1.1 TCA Precipitation of Proteins From Solutions (see Notes 1–4 )

• 1. Bring the final sample solution of a cellular extract or protein complex, for example, to 400 μL with 100 mM Tris-HCl (pH 8.5). If a larger volume of final sample solution is available, split the solution into multiple 200-μL aliquots followed by the addition of 100 mM Tris-HCl (pH 8.5) to 400 μL. At the resuspension step ( Subheading 3.1.2. , step 1), the resolubilized precipitated proteins may be recombined.

• 2. Add 100 μL TCA (100%) to cold sample; mix well to give a final TCA concentration of 20%. Reaction should be carried out on ice and the sample left overnight at 4°C.

• 3. Spin at 20,800g for 30 min at 4°C; aspirate the supernatant with gel loading tip, leaving 5 μL in the tube so as not to disturb the pellet.

• 4. Wash with 2 × 500 μL of cold acetone. After each wash, spin for 10 min at 20,800g.

• 5. Dry using a speed vac for 5 min.

3.1.2 Protein Denaturation, Reduction, and Alkylation

• 1. Add 100 mM Tris-HCl (pH 8.5), 8 M urea to TCA precipitated proteins; vortex.

• 2. Bring solution to 5 mM TCEP with the 0.1 M TCEP solution (1 M stock diluted 1/10); incubate at room temperature for 30 min.

• 3. Bring solution to 10 mM IAM with 0.5 M stock; incubate at room temperature for 30 min in dark.

3.1.3 Endoproteinase LysC/Trypsin Digestion

• 1. Add endoproteinase LysC at 1 μg/μL (1:100) to the denatured, reduced, and carboxymethylated proteins; incubate at 37°C for at least 6 h.

• 2. Dilute to 2 M urea with 100 mM Tris-HCl (pH 8.5).

• 3. Add CaCl2 to 2 mM (stock at 500 mM).

• 4. Add trypsin at 0.1 μg/μL (1:100); incubate at 37°C overnight while shaking.

• 5. On the next day, add 90% formic acid to 5%.

• 6. Store sample at −80°C.

3.1.4 Elastase Digestion

• 1. Dilute the denatured, reduced, and carboxymethylated proteins to 2 M urea with 100 mM Tris-HCl (pH 8.5).

• 2. Add elastase to an enzyme-to-substrate ratio of 1:50 (w/w); incubate at 37°C for 6 h while shaking.

• 3. Add 90% formic acid to 5%.

• 4. Store sample at −80°C.

3.1.5 Subtilisin A Digestion

• 1. Dilute the denatured, reduced, and carboxymethylated proteins to 4 M urea with 100 mM Tris-HCl (pH 8.5).

• 2. Add subtilisin A to an enzyme-to-substrate ratio of 1:50 (w/w); incubate at 37°C for 2–3 h while shaking.

• 3. Add 90% formic acid to 5%.

• 4. Store sample at −80°C.

3.1.6 High-pH Proteinase K Digestion

• 1. Resuspend TCA-precipitated proteins in 100 mM sodium carbonate (pH 11.5).

• 2. Add solid urea to 8 M urea; vortex.

• 3. Bring solution to 5 mM TCEP with the 0.1 M TCEP solution (1 M stock diluted 1/10); incubate at room temperature for 30 min.

• 4. Bring solution to 10 mM IAM with 0.5 M stock; incubate at room temperature for 30 min in dark

• 5. Add proteinase K at 0.25 μg/μL, at an enzyme-to-substrate ratio of 1:100 (w/w); incubate at 37°C for 4 h while shaking.

• 6. Add 90% formic acid to 5%.

• 7. Store sample at −80°C.

3.2 Microcapillary Column Construction and Sample Loading

Currently, all samples are desalted on-line using columns similar to the three-phase microcapillary columns described in McDonald et al. (5). These columns contain reversed-phase material, followed by strong cation exchange material, followed by reversed-phase material. In an abbreviated fashion, they are RP/SCX/RP columns. By using these columns, one does not need to carry out additional sample cleanup and buffer exchange prior to loading. For sample quantities of 400 μg or less, the triple-phase fused-silica microcapillary column is used, and for samples containing more than 400 μg, the split triple-phase fused-silica microcapillary column is used. In addition, the split triple-phase column is used for samples that originally contained detergents, allowing for more extensive washing after sample loading.

3.2.1 Pulling Columns

• 1. Make a window in the center of approx 50 cm of 100 μm × 365 μm fused-silica capillary by holding it over an alcohol flame until the polyimide coating has been charred. The charred material is removed by gently wiping the capillary with a tissue soaked in methanol.

• 2. To pull a needle, place the capillary into the P-2000 laser puller. Position the exposed window of the capillary in the mirrored chamber of the puller. Arms on each side of the mirror have grooves and small vises, which properly align the fused silica and hold it in place. Our four-step parameter setup for pulling approx 3-μm tips from a 100 μm i.d. × 365 μm o.d. capillary is as follows with all other values set to zero: Heat = 290, Velocity = 40, and Delay = 200 Heat = 280, Velocity = 30, and Delay = 200 Heat = 270, Velocity = 25, and Delay = 200 Heat = 260, Velocity = 20, and Delay = 200

3.2.2 Triple-Phase Fused-Silica Microcapillary Column

• 1. Pull tip with laser puller as described under Subheading 3.2.1.

• 2. Place approx 20 mg of Aqua RP packing material into a 1.7-mL microfuge tube, add 1 mL of MeOH, and place the tube into a stainless steel pressurization vessel. Secure the pressurization vessel lid by tightening the bolts.

• 3. The lid has a Swagelok® fitting containing a 0.4-mm Teflon ferrule. Feed the fused-silica capillary (pulled end up) down through the ferrule until the end of the capillary reaches the bottom of the microfuge tube. Tighten the ferrule to secure the capillary.

• 4. Apply pressure to the pressurization vessel by first setting the regulator on the gas cylinder to approx 400–800 psi, then opening a valve on the pressurization vessel to pressurize it. The packing material will begin filling the pulled needle capillary. If it does not flow right away, gently open the tip using a capillary scriber. Pack the capillary with 9 cm of RP packing material (see Note 5 ).

• 5. Slowly release the pressure from the pressurization vessel so as to not cause the packed RP material to unpack. Open the stainless steel pressurization vessel and remove the microfuge tube containing the RP material in MeOH.

• 6. Place approx 20 mg of Whatman Paritshpere SCX packing material into a 1.7-mL microfuge tube, add approx 1 mL of MeOH, and place the tube into the pressurization vessel. Secure the pressurization vessel as described.

• 7. Apply pressure as described in Subheading 3.2.2. , step 4. Pack the capillary with 3 cm of the SCX material and then slowly release the pressure as described.

• 8. Place the tube with Aqua RP particles in methanol back into the pressurization vessel and add 2–3 cm of RP material after the SCX material.

• 9. Wash with methanol for at least 10 min.

• 10. Equilibrate with buffer A (5% ACN, 0.1% formic acid) for at least 30 min.

• 11. To get rid of any particulates (which could clog the microcapillary column), spin down the sample to be loaded for 30 min at 20,800g, and transfer the supernatant to a new 1.7-mL microfuge tube.

• 12. Sample can then be loaded by placing the 1.7-mL microfuge tube into the pressurization vessel.

• 13. After sample is loaded, 500 μL of buffer A is added to the microfuge tube that contained the sample, and the loaded column is washed until installed onto the mass spectrometer (at least 1 h).

3.2.3 Split Triple-Phase Fused-Silica Microcapillary Column

• 1. The first step in this process is to prepare a double-phase column out of 250-μm fused silica microcapillary. Place approx 2 × 15 cm of 250 μm i.d. × 365 μm o.d. fused silica capillaries on both sides of an M-520 Inline Micro Filter Assembly.

• 2. As described under Subheading 3.2.2. , first pack the column with 3-4 cm of Whatman Partisphere SCX packing material by leading the 250-μm capillary connected to the fritted side of the inline micro-filter assembly into the pressurization vessel. The capillary connected to the other end of the inline assembly serves as a waste line flowing into an empty microfuge tube.

• 3. Then pack with 2–3 cm of Phenomenex Aqua RP packing material.

• 4. Wash with methanol for at least 10 min.

• 5. Equilibrate in buffer A for at least 30 min.

• 6. Sample can then be loaded by placing the 1.7-mL microfuge tube into the pressurization vessel.

• 7. After sample is loaded, 1.5 mL of buffer A is added to the microfuge tube that contained the sample, and the loaded column is washed until it is installed onto the mass spectrometer. Because the split three-phase columns are used for larger sample amounts and for samples that contained detergent, longer washing is necessary. Ideally, all 1.5 mL of buffer A is used for washing.

• 8. Next, a single-phase RP column, to be added to the other end of the filtered union assembly, is prepared by first pulling a tip using 100 μm i.d. × 365 μm o.d. fused silica with the laser puller as described under Subheading 3.2.1. and packing 9 to 10 cm of RP material.

• 9. This single-phase RP column is equilibrated with buffer A (5% ACN, 0.1% formic acid) for at least 30 min.

• 10. When both parts of the split column have been washed, the single-phase 100-μm RP column can be connected to the loaded 250-μm double-phase capillary using the filtered union.

3.3 Multidimensional Chromatography and Tandem Mass Spectrometry

The MudPIT system we use is a combination of the Agilent1100 quaternary pump stack with thermostatted autosampler, LCQ DECA-XP^plus tandem mass spectrometer, and a nanoelectrospray stage that interfaces the two systems. Furthermore, the hand-made microcapillary columns described above are single use, and the fused silica portion of each column is discarded after every analysis (see Note 6 ).

3.3.1 Setup of the Nanoelectrospray Stage

• 1. A detailed schematic of the nanoelectrospray stage is shown in ( Fig. 1 ). The triphasic microcapillary column or split three-phase column should be attached as shown to the P-775 MicroTee Assemblies.

• 2. One connection point of the cross contains the transfer line from the HPLC pump. This consists of a piece of 100 μm i.d. × 365 μm o.d. polyimide-coated fused silica.

• 3. A second connection point contains a length of 50 μm i.d. × 365 μm o.d. fused silica capillary that is used as a split/waste line. This split line allows a majority of the flow to exit through the split; therefore, very low flow rates can be achieved through the packed capillary micro-column. The size and length of this section of capillary depend on the flow rate from the pump and the length of the micro-column. A good starting point is to use a 12-in section of 50 μm i.d. × 365 μm o.d. polyimide-coated fused silica for the split line.

• 4. Another connection contains a section of gold wire. This is to allow the solvent entering the needle to be energized to 2400 V, thus allowing electrospray ionization to occur.

• 5. Place the packed, loaded, and washed column into a MicroTee on a stage, which in this case is designed for the ThermoFinnigan DECA-XPplus series mass spectrometer. This stage performs a threefold purpose: to support the MicroTee and hold it in place along with the connections, to electrically insulate the MicroTee from contact with its surroundings when it is held at high voltage potential, and to allow for fine position adjustments of the micro-column with respect to the entrance of the mass spectrometer (heated capillary) by using an XYZ manipulator.

• 6. Measure the flow from the tip of the capillary micro-column, using graduated glass capillaries. To do this, set the flow rate of the Agilent1100 to 0.1 mL/min from the controller. The target flow rate at the tip should be approx 200-300 nL/min and a back pressure on the Agilent1100 of between 30 and 50 bars. If the flow rate is too high, cut off a portion of the split line capillary. This will cause more of the flow to exit out of the split and cause less flow through the micro-column. If the flow is too low, a longer piece of 50-μm capillary or a section with a smaller inner diameter can be used to force more flow through the micro-column. Measuring the flow rate and adjusting the split line may have to be repeated a number of times until the target flow rate is reached (see Notes 7 and 8 ).

• 7. Prior to initiating a run, position the micro-column using the XYZ manipulator so that the needle tip is within 5 mm from the orifice of the mass spectrometer’s heated capillary.

Fig. 1.

Three-phase multidimensional protein identification technology (MudPIT) column setup. In the MudPIT system, a triphasic microcapillary column with both strong cation exchange (SCX) and reversed-phase (RP) packing materials is prepared and loaded off-line. Upon insertion of the system, a fully automated analysis can be run with salt bumps moving fractions from the RP to the SCX by a first step reversed-phase gradient and to the to the RP using a series of salt bumps detailed in Tables 1-3. Upon the application of RP gradients, peptides elute into the mass spectrometer. In this approach, a volatile salt must be used, like ammonium acetate or ammonium formate.

3.3.2 Instrument Method Design Description (itsee) Note 9 )

Data-dependent acquisition of tandem mass spectra during the HPLC gradient is also programmed through the LCQ Xcalibur™ software. Here we provide guidance for setting the parameters for data-dependent acquisition using the DECA-XPplus series mass spectrometer. The following settings are for a typical data-dependent MS/MS acquisition analysis. The method consists of a continual cycle beginning with one scan of MS (scan one), which records all of the m/z values of the ions present at that moment in the gradient, followed by three rounds of MS/MS. Full MS spectra are recorded on the peptides over a 400 to 1600 m/z range. Dynamic exclusion is activated to improve the protein identi- fication capacity during the analysis.

• 1. In the main Xcalibur software page, select “Instrumental Setup.”

• 2. In the next window, select the button labeled “Data Dependent MS/MS.”

• 3. The general settings for any given method is as follows: “segments” is set to 1; “start delay” is set to 0; “duration” is 117 min; “number of scan events” is 4; “scan event details” are first, MS, second, MS/MS of most intense ion, third, MS/MS of second most intense ion, fourth, MS/MS of third most intense ion.

• 4. For scan event one, highlight the bar showing “Scan Event 1.” Below this, check “Normal Mass Range”; “Scan Mode” is MS, and “Scan Type” is full. Set “m/z range” to 400–1400, “Polarity” to positive, and “Data Type” to centroid.

• 5. The “Tune Method” box specifies the path for a file containing the parameters for the electrostatic lenses and ion trap. These parameters are established through the “LCQ Tune” as described in the “ThermoFinnigan LCQ Getting Started” manual. The capillary temperature is 200°C and the electrospray voltage is 2.4 kV.

• 6. The data-dependent settings are as follows, with only these options engaged: “default charge state” is 2, “default isolation width” is 3, “normalized collision energy” is 35, “activation time” is 30, “minimal signal required” is 100,000, “minimal MSn signal required” is 5000, “exclusion mass width low” is 0.80, and “exclusion mass width high” is 2.20. Dynamic exclusion is enabled, with the following settings: a repeat count of 2, a repeat duration of 0.5, an exclusion list of 50, an exclusion duration of 5.00 min, an exclusion mass width low of 0.80, and an exclusion mass width high of 2.20.

• 7. In the “Timed Events” window of the instrument setup, the following sequence should be inserted: Time (min) = 0.00, Settings = Contact 1, Value = Open Time (min) = 0.05, Settings = Contact 1, Value = Closed Time (min) = 0.10, Settings = Contact 1, Value = Open

3.3.3 Gradient Profiles for Complex Mixture Analysis

We use two general gradient profiles. One is for analyzing protein complexes and one is for analyzing organelles or other cellular fractions. The method for analyzing protein complexes takes approx 10 h, and the second method takes approx 20 h. Because we use three-phase columns, the first step in any run is a reversed-phase gradient to move any bound peptide from the first RP to the SCX material inside the column. Then, successive salt bumps are run to move small amounts of peptides from the SCX onto the last RP, followed by a slow reversed-phase gradient to resolve peptides within the last RP before they are eluted off into the mass spectrometer. In the “Gradient Program” window of the instrument setup, the following sequences are steps that can be used or adjusted by a researcher. An example base peak chromatogram is shown in Fig. 2 as an example of the profile that should be seen from one step of a successful sample.

Table 1 Gradient Profile of First Step

This is the general gradient profile that we use for all additional steps where the % of buffer C increases with each step:

Fig. 2.

Chromatographic profile of the fifth step of a multidimensional protein identification technology (MudPIT) analysis of the HeLa cell mediator component CRSP70. As described in Sato et al. (13), the HeLa cell mediator component CRSP70 was FLAG tagged and affinity purified from HeLa cell extracts and analyzed via the six-step MudPIT chromatographic elution profile shown in Table 4 . (bdA) The % buffer composition of buffers A, B, and C over the 117-min step 5, which contains a 70% buffer C salt bump. (bdB) The base peak chromatogram (m/z 400-1400) of this step for the CRSP70 analysis, which is a general representation of the peptides eluting from the biphasic MudPIT column into a tandem mass spectrometer

Table 2 Gradient Profile of X(2–80)% Buffer C Step

This is the last step for any given MudPIT analysis, and is designed to try to remove everything off the column:

Table 3 Gradient Profile of 100% Buffer C Step

In the Xcalibur software main page, check the “sample list” button. In the “Sequence Setup” window, fill in the “Filename,” “Path,” and “Inst Method” lines with the appropriate methods indicated previously. Save the sequence to the same directory of the paths for the *.RAW files. In the table below are the sequences we use for the analysis of a protein complex (6-step MudPIT) and for more complex mixtures like an organelle (12-step MudPIT):

Table 4 Multidimensional Protein Identification Technology (MudPIT) Analyses Methods

• 5. Once everything is set up and the needle is positioned correctly in front of the heated capillary on the mass spectrometer, hit “Actions” and “Run Sequence.” A box will come up that should have the “Agilent1100 Quat,” “Agilent 1100 Thermostatted AS,” and “LCQ DECAXP MS” listed, and “Yes” should be typed under the “Start Instrument” box of the “Agilent 1100 Thermostatted AS.” The “start when ready” box should be checked, as well as “After Sequence Set System to-Standby. ” Both the “pre-acquisition” and “post-acquisition” boxes under “run synchronously” should contain check marks. Hit “OK” and the run will begin. A message will likely come up saying that there are devices that need to be turned on; hit “OK.”

• 6. Upon the completion of a run, the *.RAW files that have been accumulated by the mass spectrometer need to be converted to *.DAT files for SEQUEST analysis. To do this, go to the “Home Page” and hit the “Tools” and “File Converter” buttons. Follow the instructions to select the jobs that need to be converted to *.DAT files, and hit “OK.”

3.4 Data Analysis

For complex peptide mixture analysis via MudPIT, database searching is an important component and requires a computer cluster to analyze tandem mass spectra. SEQUEST (6) clusters are available from IBM and ThermoElectron (San Jose, CA), and Mascot (17) clusters are available from Matrix Science (Boston, MA). In addition, researchers who have the necessary expertise can design their own cluster and purchase licenses for these software platforms to install on their cluster. Lastly, an open-source tandem mass spectral searching algorithm named X!Tandem (18,19) is emerging as an alternative to SEQUEST (6) and Mascot (17). There is growing evidence in the literature that single-peptide hits to proteins from complex mixtures must be considered with great care, but proteins that are identified by at least two good-quality tandem mass spectra are likely to be real identifications (20,21). The manual assessment criteria listed below offer suggestions on how to evaluate spectrum/ peptide matches.

3.4.1 SEQUEST Analysis

• 1. The software algorithm 2to3 (22) is used to determine charge state and to delete spectra of poor quality.

• 2. SEQUEST (6) is used to match MS/MS spectra to peptides in a database containing protein sequences (typically *.fasta files downloaded from the National Center for Biotechnology Information and updated frequently), complemented with sequences for common contaminants, including human keratin variants and immunoglobulins.

• 3. The validity of peptide/spectrum matches is assessed using the SEQUEST-defined parameters, cross-correlation score (XCorr), and normalized difference in cross-correlation scores (DeltCn). Spectra/peptide matches are retained only if they have a DeltCn of at least 0.08 and minimum XCorr of 1.8 for singly, 2.5 for doubly, and 3.5 for triply charged spectra. In addition, the peptides have to be at least seven amino acids long.

• 4. The program DTASelect (23) is used to select and sort peptide/spectrum matches passing this criteria set. Peptide hits from multiple runs are compared using CONTRAST (23).

3.4.2 Manual Assessment Criteria

• 1. The MS/MS spectrum must be of good quality, with fragment ions clearly above baseline noise. Sometimes a spectrum will consist purely of “sticks.” In general, these are poor-quality spectra. There should be some noise that allows one to define what the baseline noise is.

• 2. There must be some continuity to the b and y ion series. Of ions that are matched by SEQUEST to the original tandem mass spectrum, the matched ions should correspond to signal rather than noise. Furthermore, better matches contain strings of ions rather than isolated ions here and there throughout the sequence.

• 3. The intensity of a b or y ion resulting from a proline should be far more intense than the other ion in the spectrum. If there are multiple prolines present in a peptide, one may see internal fragment ions corresponding to the mass of the portion of the peptide between the two prolines.

• 4. When evaluating peptides generated by nontryptic digestions, be aware that because basic residues can be located anywhere within the peptides, doubly charged precursor ions can lead to doubly charged fragments, which will appear as unmatched intense peaks in most spectrum displays (24).

4 Notes

• 1. For samples that likely contain large amounts of DNA, the addition of 0.1 U of Benzonase to the sample prior to TCA precipitation and incubation at 37°C for 30 min is critical. This will decrease the viscosity of the mixture and prevent column clogging during sample loading, washing, or analysis.

• 2. Using the TCA precipitation protocol as the starting point, the only chemical that cannot be dealt with is glycerol. The TCA precipitation protocol has been useful in detergent removal. When the presence of glycerol cannot be avoided, loading such samples on split columns with extensive washes may prevent column clogging.

• 3. With the TCA precipitation protocol and protein complexes, it is common to not see a pellet. For this reason, a small amount of acetone should be left at the bottom of the tube during washing to be sure to not aspirate and discard the pellet. 4. When handling samples, gloves must be worn at all times in order to prevent the addition of large amounts of human keratin and immunoglobulins to a sample.

• 5. Placing a black three-ring binder about 6 in. behind the loading column will provide the contrast necessary to see the status of column packing.

• 6. The microassemblies used for the split three-phase columns can be reused by first sonicating in 50% methanol:50% water and allowing them to air dry. However, we discard the filters (M-120X filter end fitting from UpChurch Scientific, Oak Harbor, WA) and use a new one for each sample.

• 7. If the Agilent1100 reaches the high-pressure limit (set at 100 bars), the most likely place for clogging is the small piece of 100 μm i.d. × 365 μm o.d. fused-silica tubing that connects the two MicroTees ( Fig. 1 ). If changing the capillary does not alleviate the pressure, removing the MicroTees and sonicating them in 50% methanol:50% water and allowing them to air dry may solve the problem.

• 8. For columns that clog during loading of the sample, washing of the sample, or testing of the flow rate prior to analysis, heating the length of the column with a Varitemp Heat Gun (model VT-750C, Master Appliance Corp, Racine, WI) usually makes it flow again. Once a column starts flowing again, extensive additional washing with buffer A is recommended to prevent subsequent clogging.

• 9. Especially with whole cell extracts and their analysis by MudPIT, it is critical to clean and tune the mass spectrometer according to the manufacturer’s instructions at least every 10 d.