Improvement of Automatic In-Gel Digestion by in Situ Alkylation of Proteins

Abstract

We have recently improved the automation of an in-gel digestion system, DigestPro 96, using in situ alkylation of proteins with acrylamide, conducted during one-dimensional (1D) SDS-PAGE. The improved method included the processes of destaining, dehydration, trypsin digestion, and extraction but excluded the reduction and alkylation steps following staining of proteins with CBB. The extracted peptide mixtures were directly loaded onto a micro C18 LC column of the mass spectrometer. The resultant spectra were processed with “Mascot” search engine to estimate the sequence coverage of the bovine serum albumin (BSA). The original method, designed for Laemmli 1D SDS gel applications, consisted of reduction and post-alkylation with iodoacetamide, which produced carboxyamidemethyl (CAM; –S–CH₂CONH₂) derivatives. The original method also included a desalting step essential for mass spectrometry, especially matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. We compared the original and improved methods using BSA (3 pmol loaded to the gel, one third of digested peptide mixture injected into LC-MS). The original method yielded both CAM and propionicamide (PAM; –S–CH₂CH₂CONH₂) derivatives. The source of PAM derivatives is the unpolymerized acrylamide formed during electrophoresis. The sequence coverage of CAM derivatives of BSA by the original method was 10% with desalting and 19% without desalting. The sequence coverage of PAM derivative by the improved method was 32%. Our results clearly show the advantage of our improved automated in-gel digestion method for in situ PAM alkylated protein with respect to peptide recovery, compared with the original method with CAM post-alkylation.

In-gel digestion of protein is a widely used procedure in proteomics studies. However, the procedure is tedious and susceptible to contamination. Automation of the procedure^1,2 has simplified the assay and reduced potential contamination. The automated protocol comprises several steps including destaining, reduction, alkylation (using –S–CH₂CONH₂ [carboxyamidemethyl; CAM]), digestion and extraction of the tryptic peptides. Among the steps of the automated protocol, reduction and alkylation are performed so that the cysteinyl peptide yields CAM derivatives. The cysteinyl residues, however, could also react with unpolymerized acrylamide monomer present in the electrophoresis gel. The reaction occurs during the electrophoresis run and produces propionic amide (PAM) derivatives, which are stable byproducts. The formation of PAM sometimes hinders fragment search procedure that expects CAM derivatives to be present. Mineki et al.³ have considered this by-product (PAM derivative) to be the major and sole “S”-containing peptide for mass spectrometry (MS) analysis. The digestion method described by Mineki et al.³ does not include reduction, alkylation, and desalting procedures. In this article we report improvement of the automated in-gel digestion protocol.

MATERIALS AND METHODS

Material

Molecular marker proteins of a wide molecular mass range (6.5–200 kDa) were obtained from Bio-Rad (Hercules, CA). The marker was run with one-dimensional sodium dodecyl sulfate (1D SDS) electrophoresis so that a band of 3 pmol of bovine serum albumin (BSA) was present in one lane.

In Situ Alkylation and 1D SDS-PAGE

Except for testing the protocol recommended by the supplier of the instrument, the protein was in situ alkylated during the electrophoresis run using the method described by Mineki et al.³ Briefly, sample were added to sample buffer containing 10% glycerol, 2% SDS, 0.1 M dithioerythritol (DTE) and 0.0025% bromphenol blue. The sample was then incubated at 95°C for 3 min. After cooling the sample to room temperature, a 4.7-μL aliquot of 30% acrylamide solution (20 μmol) was added. Then, the sample solution was loaded onto a gel and run.

A 12-lane, 130 × 130 × 1-mm SDS gel was prepared with 10% separation gel and 4% stacking gel acrylamide containing 0.87% and 0.35% piperazine diacrylamide, respectively. The gel was allowed to stand overnight at 4°C, and then pre-run at 20 mA for 1 h before loading the sample. After loading, proteins were concentrated in the stacking gel at 5 mA for 50 min and were resolved in the separation gel at 20 mA for 2 h. Staining was performed using CBB R-350 (Pharmacia & Upjohn Diagnostics, Uppsala, Sweden) or Plus One silver nitrate reagent free of glutardialdehyde (Pharmacia & Upjohn) according to the method of Mineki et al.³

In-Gel Digestion

The BSA band was cut and processed by an automatic in-gel digestion instrument (DigestPro96, Intavis AG, Koeln, Germany). The original method supplied by the manufacturer is shown in Figure 1. A gel run by Laemmli’s method⁴ was used for this protocol. The improved method for CBB-stained gel is shown in Figure 2, and the modified method of silver-stained gel in Figure 3. The manual digestion method employed was the one described by Mineki et al.³

FIGURE 1.

Cysteinyl residues of the protein were reduced and alkylated with CAM before it was digested with trypsin. The original protocol (Digest1.xcp) includes reduction and alkylation processes. The method is for CBB-stained gel.

FIGURE 2.

Cysteinyl residues of protein were alkylated in situ with acrylamide during electrophoresis. In this automated protocol (improved method) for proteins stained by CBB, the reduction and alkylation processes were omitted. The desalting step was also omitted because of the deletion of the alkylation process.

FIGURE 3.

Cysteinyl residues of protein were alkylated during the electrophoresis run. In the automated protocol (improved method) for proteins stained by silver nitrate, the reduction and alkylation processes were deleted. The desalting step was also omitted because of the deletion of the alkylation process.

Mass Spectrometry Analysis

Mass spectrometry analysis was performed with AB-QSTAR pulsar i hybrid (Applied Biosystems, Framingham, CA) combined with a microliquid chromatograph (Magic 2002; Michrom Bioresources, Abum, CA) equipped with 0.2-mm ID × 50 mm Magic C18 column. In order to evaluate the improved method of DigestPro96, the sequence coverage of BSA by MS/MS was obtained with automatic search using “Mascot” software (Matrix Science, London, UK). The sequence coverage was calculated in the “Mascot” software as (total number of the amino acid residues of identified sequences) / (total number of amino acid residues of BSA).

RESULTS AND DISCUSSION

Comparison of Sequence Coverage of BSA Between the Original and Improved Methods on 1D SDS-PAGE

The BSA sequence coverage analysis by Mascot software was performed using CAM derivatives and also using passively coexisting PAM derivatives. The coverage was 10% for CAM derivatives and 16% for PAM derivatives. PAM derivatives were formed during electrophoresis and are shown in Table 1. The sequence coverage of the method, which did not include desalting with Zip Tip, was 19% with CAM derivatives and 27% with PAM derivatives. These results indicate that the recovery of peptides from the desalting column was 50–59%. Small-sized hydrophilic peptides seemed to pass through the Zip Tip while the large hydrophobic peptides were retained on the column (Figures. 4 and 5). This resulted in a marked decrease in the recovery of peptides.

TABLE 1.

Comparison of Original Method and Improved Method

	Original (Post-electrophoresis alkylation)
	With desalting	No desalting	Improved (In situ alkylation)
CAM derivatives	10%	19%	—
PAM derivatives	16%	27%	32%

Data are sequence coverage of 3 pmol BSA stained with CBB. One third of the extract was injected for mass spectrometry.

FIGURE 4.

Mass spectrometric total ion current of the sample processed through assays that included desalting. The samples were processed by the original method shown in Table 1.

FIGURE 5.

Mass spectrometric total ion current of the sample processed through assays that included no desalting. The samples were processed by the original method shown in Table 1.

With regard to the improved method using in situ alkylation with acrylamide, the sequence of BSA identified by Mascot software was limited to PAM derivative. The method provided 32% coverage of BSA. Compared with the original method, three processes were omitted in the improved method: reduction, alkylation of stained proteins before in-gel trypsin digestion, and desalting before mass spectrometric analysis.

Comparison of Sequence Coverage of BSA Between Manual and Automated Protocol (Performed by DigestPro96)

In the next step, we compared the BSA sequence coverage obtained by the manual protocol with that obtained by the improved (automated) method. Both protocols employed in situ alkylation with acrylamide during electrophoresis. The sequence coverage of the manual protocol was 48% and that of the automated instrument was 32%. The latter represents 66% of the sequence coverage of the manual protocol. Taking into consideration the relatively small amount of protein (3 pmol) that was digested, our results indicate a significant improvement by the automated protocol.

Differences in Sequence Coverage of BSA (3 pmol) Between CBB and Silver Nitrate Staining

Next we compared the sequence coverage of BSA stained with CBB with the coverage obtained with silver nitrate reagent (Table 2). The protocol used in both studies was the automated protocol employing in situ alkylated proteins.

TABLE 2.

Comparison of BSA Sequence Coverage and Detection Limits Between CBB-Stained and Silver-Stained (Without Glutardialdehyde) Gel in the Automated Protocol

	Loaded amount to gel	Loaded amount to MSa	BSA sequence coverage
SEQUENCE COVERAGE
CBB stain (n = 5)	3 pmol	1 pmol	32%
Silver stain (n = 2)	3 pmol	1 pmol	21%
DETECTION LIMIT FOR SILVER STAINED GEL
Silver stain (n = 2)	3 pmol	1.5 pmol	21%
Silver stain (n = 2)	750 fmol	750 fmol	11%
Silver stain (n = 2)	375 fmol	375 fmol	Not detected

^aExpressed as amount of protein loaded to gel.

The sequence coverage of BSA stained with silver nitrate was 21%, which represented 66% of the coverage of BSA stained with CBB. Although the sequence coverage on silver staining was lower, the improved protocol with silver staining was still considered to be usable.

We also analyzed the detection limit of the modified automated protocol for silver staining (Table 2). The detection limit was 750 fmol with a BSA sequence coverage of 11%.

CONCLUSION

In this report, we describe the use of a modified automated in-gel digestion protocol for proteins using in situ alkylation with acrylamide. The sequence coverage of BSA using the modified protocol was 32%, which was threefold higher than that of the original method (10%). The improvement is probably due to PAM derivatization during electrophoresis, elimination of reduction and alkylation processes before in-gel digestion with trypsin, and deletion of the desalting process before mass spectrometric analysis. The sequence coverage of BSA by the improved protocol exhibited 66% value of the corresponding manual protocol in the digestion of the 3-pmol sample. Thus, the automated protocol seems to be useful for routine protein digestion. The protocol of the silver-stained gel showed 21% sequence coverage, which corresponded to 66% of the value of CBB-stained gel. The detection limit of the protocol was 750 fmol.