Mass spectrometry analysis of soybean seed proteins: optimization of gel-free quantitative workflow

Abstract

For high-throughput quantitative mass spectrometry (MS) analysis of soybean seed proteins, a method that avoids gel fractionation is advantageous. We developed and optimized a workflow from protein isolation to MS-based quantitation without polyacrylamide matrices. The objective was to quantitatively compare extraction methods to reproducibly arrive at the highest yield and proteome coverage. Beginning with mature soybean seed, we compared four proteinextraction methods, employing either TCA/acetone, urea, urea/thiourea, or phenol. Soybean proteins were extracted, quantified for total protein content, and comparatively visualized by Coomassie–SDS-PAGE. The phenolextraction method yielded protein concentrations 2 to 7-fold higher than other extraction methods. Comparison of trypsin to protein ratios (1 : 25, 1 : 50, 1 : 75, and 1 : 100) revealed a near linear increase in spectral counts by MS with increasing trypsin levels. In-solution digestion procedures were also compared to determine optimal resuspension and digestion conditions for peptideextraction and quantitation. A resuspension buffer that contained 50 mM Tris–HCl of pH 8.0 and 5 M urea showed the highest spectral counts and protein identifications. The results of this study show the time-honored phenolextraction method consistently and unequivocally yielded the highest amounts of protein from mature soybean seed, and that buffered urea is sufficient for optimal resuspension of precipitated proteins for tryptic digestion and mass spectrometry.

---

Introduction

Two-dimensional gel electrophoresis (2-DGE) is a useful tool for protein profiling, however, this method is limited by problems such as poor sensitivity, low throughput, and low reproducibility.¹ Several technical advancements such as the development of sensitive protein stains and advanced detection software have helped to address some of these limitations.^2,3 Despite these advances, 2-DGE is still biased against proteins with extreme biochemical or biophysical features (e.g. integral membrane and high/low pIproteins) and therefore limited at capturing an entire proteome.^1,4–6 Alternative techniques, including GeLC and gel-free approaches, have been shown to overcome the biases associated with 2-DGE.^7–9 Moreover, with the development of faster, broader dynamic range mass spectrometers and the coupling of these instruments to reproducible, nanoflow liquid chromatography systems, gel-free proteomics is becoming more feasible.¹⁰

Gel-free methods have been proven to be more reproducible and less biased than gel-based methods.^11–17 Nagaraj et al.¹³ introduced a simple and highly reproducible method for membrane proteomics that allows use of detergents. Kim et al.¹⁴ reported a method for in-solution digestion of a protein mixture in the absence of denaturants. van Midwoud et al.¹⁵ improved reproducibility and detection limits by including 25% dimethyl sulfoxide and 5% formic acid in the LC-MS/MS analysis of hydrophobic peptides. The in-solution digestion efficiency was improved using an excess amount of trypsin.¹⁸ Additionally, digestion of nanogram protein quantities was improved using organic-aqueous solvents or offline two-step sample cleanup.^16,17 These methods proved to increase digestion speed, trypsin efficiency and peptide recovery even when starting with low nanogram amounts of protein.^16,17 Nevertheless, maximizing protein yield or proteome coverage remains challenging and is frequently organism-dependent. Thus, further studies aimed at optimization of sample preparation for improved proteinextraction, solubilization, digestion, and identification are required, especially for recalcitrant plant tissues such as dry seed.

Proteomic analysis has been performed to analyze soybean seed using high-resolution 2-DGE in combination with mass spectrometry.^7,19,20 These proteomic studies have successfully established a 2-DGE-based workflow from protein isolation to protein assignment. Routine examination of changes in seed protein composition could benefit from a simple and automated proteomic workflow. Nanjo and coworkers²¹ investigated flooding stress responses of early stage soybean seedlings by gel-based and gel-free proteomics methods. Natarajan and coworkers²² compared four proteinextraction methods (urea, thiourea/urea, phenol, and TCA/acetoneprecipitation) to determine their efficacy in separating soybean seed protein by 2-DGE. At present, most studies have focused on gel-based quantitative proteomics as opposed to gel-free quantitative workflows, which have the greatest potential for high-throughput analyses. An optimized gel-free workflow can aid high-throughput quantitative analysis by reducing labor and improving sensitivity.

A simple and sensitive method for determining relative protein abundance has been introduced by Liu and cowokers.²³ The number of identified total tandem mass spectrometry (MS/MS) spectra (spectral count) for each protein showed strong linear correlation with relative protein abundance.²⁴ Spectral counting can be used as a reliable index for relative quantification and has been successfully applied in different biological contexts.^25,26

The objective of this study was to produce a gel-free proteomics method that reproducibly produced the highest protein yield (from fresh tissue) or proteome coverage (once protein was extracted) from mature soybean seed. An emphasis on maximal proteome coverage, or spectral counts, is critical as most technical variation ultimately arises from low signal to noise. During this investigation we found that many steps in the workflow from protein isolation to protein quantitation (e.g.protein isolation, trypsin digestion, etc.) required optimization. Here we report an optimized protein isolation and digestion procedure for gel-free quantitative proteomic analysis of mature soybean seed proteins. Four proteinextraction methods were compared to identify the method that reproducibly and consistently produced the highest protein yield. Of the methods compared, the phenolproteinextraction method proved to be the most efficient and reproducible, and therefore protein extracted by this method was used to optimize the in-solution digestion method. We investigated different resuspension media to determine optimal resuspension and digestion conditions for optimal gel-free analyses. We expect this procedure will find applications in gel-free quantitative proteomics for high-throughput profiling of prominently expressed soybean seed proteins, most notably allergens.

Materials and methods

Materials

Mature soybean seed [Glycine max L. Merr] of cultivar Williams 82 was purchased from Missouri Crop Improvement Association (Columbia, MO, USA). Acetone, acetonitrile (HPLC-grade), acrylamide, bis-acrylamide, bovine serum albumin (BSA), ethanol, formic acid (sequencing-grade), hexane, methanol, phenol, sodium dodecyl sulfate (SDS), thiourea, ethylenediaminetetraacetic acid (EDTA), iodoacetamide, trichloroacetic acid (TCA) and β-mercaptoethanol were purchased from Fisher Scientific (Houston, TX, USA). Ammonium acetate, CHAPS, dithiothreitol (DTT), n-octylglucoside, sucrose, Tris, and urea were purchased from Research Products International Corporation (Mt Prospect, IL, USA). Sequence-grade modified trypsin was purchased from Promega (Madison, WI, USA). Bradford assay reagent was from Bio-Rad (Hercules, CA, USA).

Plant material

Mature soybean seed were pulverized into a fine powder using a standard, bladed coffee grinder. The fine power was aliquoted into 100 mg samples representing one biological replicate. Aliquots were analyzed in biological quadruplicate using four different proteinextraction methods.

Urea extraction

Mature soybean seed were solubilized according to Berkelman and coworkers.²⁷ Briefly, 500 µL of lysis buffer (50 mM Tris–HCl, pH 8.0, 8 M urea, and 2% (w/v) CHAPS) were added to 100 mg of dried soybean powder. Extracts were vortexed briefly, sonicated for 40 min at room temperature, and finally centrifuged at 13 000g for 10 min at 4 °C. Supernatants were collected and stored at −80 °C.

Thiourea/ureaextraction

In this method as described previously,²⁸ soybean seed powder was defatted twice with hexane, and dried. Defatted soybean seed powder (100 mg) was incubated with 1.5 mL of extractionbuffer (50 mM Tris–HCl, pH 8.0, 4% (w/v) CHAPS, 5 M urea, 2 M thiourea, and 65 mM DTT) for 5 min at room temperature. Homogenates were centrifuged at 13 000g for 5 min at 4 °C and supernatants were collected and stored at −80 °C.

Phenol extraction

This procedure was carried out according to a previously described method.^29,30 Ten mL of phenolextractionbuffer (0.9 M sucrose, 100 mM Tris–HCl, pH 8.0, 10 mM EDTA, 0.4% (v/v) β-mercaptoethanol, 50% (v/v) phenol) were added to 100 mg soybean powder aliquots. Extracts were homogenized for 20 s by homogenizer at room temperature, placed on a Nutator agitator for 30 min at 4 °C, and centrifuged at 4000g for 15 min at 4 °C. After centrifugation, the phenol phase was transferred to a new tube and combined with 5 volumes of cold 0.1 M ammonium acetate in 100% methanol to precipitate proteins, vortexed, and incubated at −20 °C overnight. Precipitates were collected by centrifugation (4000g, 10 min, 4 °C), and the pellet was washed with 0.1 M ammonium acetate in methanol, ice-cold 80% (v/v) acetone, and finally cold 70% (v/v) ethanol. Protein pellets were stored in 1 mL of 70% (v/v) ethanol at −20 °C. Fifty µL of precipitated protein were collected by centrifugation at 13 000g at 4 °C. Pellets were resuspended in 200 µL of resuspension buffer (50 mM Tris–HCl, pH 8.0, 7 M urea, 2 M thiourea, and 2% (w/v) CHAPS) by pipetting and vortexing, and finally centrifuged at 13 000g for 10 min at 4 °C. Supernatants were used collected and stored at −80 °C.

TCA/acetoneprecipitation

This method was performed as described previously.²² Mature soybean seed powder aliquots (100 mg) were added with 5 mL of a solution containing 10% (w/v) trichloroacetic acid (TCA) in acetone with 0.07% (v/v) β-mercaptoethanol. Extracts were homogenized for 20 s by homogenizer and precipitated overnight at −20 °C. Pellets were washed twice with acetone containing 0.07% (v/v) β-mercaptoethanol. Then pellets were dried and resuspended in 1 mL of resuspension buffer (50 mM Tris–HCl, pH 8.0, 8 M urea, 2% (w/v) CHAPS and 1% (w/v) DTT), followed by sonication on ice for 30 min. Extracts were centrifuged at 13 000g for 20 min at 4 °C. Supernatants were collected and stored at −80 °C.

Protein quantification and SDS-PAGE

Protein extracted by the various isolation methods was quantified in triplicate using the Coomassie dye binding assay (Bio-Rad) employing BSA as standard.³¹Protein qualitative comparisons were confirmed by SDS-PAGE. Thirty µg of protein extracted by each method were separated by 12% SDS-PAGE and visualized by staining with colloidal Coomassie blue G-250³² to confirm accurate quantitation. For post-tryptic digestion analysis, 1 µg of the resulting tryptic peptide mixture was resuspended in 0.1% (v/v) formic acid and analyzed using 15% SDS-PAGE followed by silver staining.

Protein resuspension

Precipitated protein was dissolved in the following resuspension buffers: (1) 50 mM Tris–HCl, pH 8.0, 7 M urea, or buffer 1 plus one of the following (2) 2 M thiourea, (3) 2 M thiourea, 0.1% (w/v) CHAPS, (4) 2 M thiourea, 0.5% CHAPS, (5) 2 M thiourea, 2.0% CHAPS, (6) 2 M thiourea, 0.1% (w/v) n-octylglucoside, (7) 2 M thiourea, 0.5% n-octylglucoside or (8) 2 M thiourea, 2.0% n-octylglucoside. Addition of these resuspension buffers was followed by gentle vortexing to resuspend the pellets. Extracts were centrifuged at 13 000g for 10 min at 4 °C. Supernatants were collected and stored at −80 °C.

In-solution digestion

As an internal standard, 500 ng of BSA were added to 50 µg of soybean protein, obtained from the phenolextraction method. To break and inhibit the formation of disulfide bonds, proteins were incubated with 10 mM DTT at 25 °C for 1 h. Reduced cysteines were alkylated with 50 mM iodoacetamide at 25 °C in the dark for 1 h and quenched by adding additional DTT to 40 mM and incubating at 25 °C for 30 min. The protein mixture was diluted 1 : 10 with 50 mM ammonium bicarbonate to reduce the concentration of denaturants. Trypsin was reconstituted to 0.02 µg µL⁻¹ and added to proteins obtained from the phenolextraction method at ratios of 1 : 25, 1 : 50, 1 : 75, and 1 : 100 (w/w) (trypsin : protein). Finally, digestions were incubated at 37 °C for 16 h and lyophilized to dryness.

Mass spectrometry analysis

This procedure was carried out according to previously described methods.³³ Briefly, dried peptides were reconstituted in 0.1% (v/v) formic acid to a final concentration of 100 ng µL⁻¹. Ten µL injections were analyzed on an LTQ ProteomeX linear ion trap LC-MS/MS instrument (Thermo Fisher, San Jose, CA). LC separation was performed using fused silica nanospray needles, 10 cm length (360 µm outer diameter, 150 µm inner diameter; Polymicro Technologies, Phenix, AZ) that were packed with “magic C18” (100 Å, 5 µm particle; Michrom Bioresources) in 100% methanol. Samples were analyzed in the data-dependent positive acquisition mode using normal scan rate for precursor ion analysis, and dynamic exclusion enabled. Following each full scan (400–2000 m/z), data-dependent triggered MS/MS scan for the three most intense parent ions was acquired. The C₁₈ column was washed between each set of biological replicates.

Database search and spectra analysis

RAW files were searched against the soybean genome (January 2008, http://www.phytozome.net/) using SEQUEST (Thermo Finnigan, San Jose, CA, USA; version 2.7) batch search in Bioworks version 3.3.1 to assign peptides. SEQUEST was searched with a fragment ion mass tolerance of 1.0 amu and a peptide tolerance of 1000 ppm. Iodoacetamide derivation of cysteine (+57) was specified in SEQUEST as a fixed modification while oxidation of methionine (+16) was a variable modification. Matching peptides were filtered according to cross-correlation score (XCorr at least 1.5, 2.0 and 2.5 for +1, +2 and +3 charged peptides, respectively).

Search result files were imported into Scaffold version 2.2.1 (Proteome Software, Portland, OR, USA). Scaffold was used for spectral counting and to validate MS/MS based peptide and protein identification. Peptide and protein identifications were accepted if they could be established at greater than 95% and 99% probability, respectively. A minimum of two unique peptides was used as a cutoff for protein assignment. False discovery rates as determined by mining a randomized, decoy database were less than 0.02.

Results and discussion

Numerous MS-based quantitative proteomics methods have been reported.¹ These methods include in vivo isotopic labeling using ¹⁵N/¹⁴Nmetabolites and stable isotope labeling by amino acid in cell culture (SILAC). In vitrolabeling has also been demonstrated by ¹⁸O-labeling during trypsin digestion and N-terminal peptidelabeling using isobaric tags (iTRAQ).^1,34 These approaches have been widely employed in proteomic studies. However, labeling-based quantification has several limitations including complexity of sample preparations, incomplete labeling, poor labeling efficiency, high cost of reagents, and the potential for artifacts.^10,26 Label-free quantitation overcomes many of these limitations making it an attractive alternative to isotope labeling methods. Label-free, relative quantitation methods analyze ion intensity (peak integration) or frequency of peptide fragment spectra during LC-MS/MS (spectral counting). In comparison to peak integration, spectral counting is a more simple approach for determining relative differences in protein abundance.^26,33 Spectral counting was therefore employed in the current gel-free quantitative proteomic study of mature soybean seed proteins, in order to optimize a simple and reproducible proteinextraction and preparation method.

Phenol protein extraction protocol is higher yielding and more reproducible than other extraction protocols

One objective of this study was to definitively identify a proteinextraction method that consistently produces the highest protein yield from mature soybean seed. In this study, the protein yield of four extraction methods, TCA/acetone, urea, urea/thiourea, and phenol was compared. To begin, soybean seed proteins were extracted by each method in biological quadruplicate and quantified by Bradford assay. Of the four methods, the phenolextraction method yielded protein concentrations 2-, 5-, 7-fold higher than TCA/acetone, urea, and urea/thioureaextractions, respectively (Fig. 1A). To investigate qualitative characteristics of protein extracts, mature soybean seed proteins were visualized by Coomassie–SDS-PAGE (Fig. 1B). The four extraction methods yielded samples that showed similar resolution of the protein bands, although slight differences were observed. Our data indicate the phenolextraction method was the most successful at retrieving protein from soybean seed powder. In prior publications it was noted that phenolextraction could be biased against low molecular weight proteins³⁵ or that alternative extraction protocols were superior.²² However, in none of these publications was protein yield compared or were doping or loading controls used to verify these conclusions; instead, qualitative comparison by SDS-PAGE was performed. In the absence of a “perfect” soybean seed protein extract for comparison by SDS-PAGE, we chose to focus on protein yield as the single most important parameter by which to ascertain efficacy of the proteinextraction procedure. Using this metric the phenol protocol was markedly more effective than the other three extraction methods and could therefore be considered the least biased proteinextraction procedure for soybean seed.

Fig. 1 Average protein yield from four different extraction methods. (A) Protein was extracted from mature soybean seed in biological quadruplicate using four extraction methods. Total extracted protein was quantified by Coomassie dye binding assay. Values are the average of biological replicates and standard deviations are shown. (B) Four protein isolation methods were analyzed by Coomassie–SDS-PAGE. Thirty µg of protein were loaded in each lane, separated by 12% SDS-PAGE and visualized by colloidal Coomassie Brilliant Blue staining.

To further investigate the efficacy of the phenolextraction protocol, we performed the first phenolextraction plus two back extractions with 5 mL and 10 mL phenolextractionbuffer volumes (Fig. 2). We found that the first extract and back extractions 1 and 2 produced approximately 80, 15, and 5% of total extractable protein, respectively. These data indicate that first phenolextraction produced the highest total protein yield from soybean seed. Comparison of protein from first, second, and third phenolextractions by SDS-PAGE revealed no major differences in the band pattern between these protein samples (data not shown).

Fig. 2 Comparison of phenolextraction efficiency. Protein was extracted from mature soybean seed with 5 mL and 10 mL phenolextractionbuffer volumes. Data are presented as the percentage of isolated protein of the first phenolextraction plus two back extractions from three biological replicates. Error bars denote standard deviation from three biological replicates.

Buffered urea is sufficient for protein resuspension and is optimal for MS analysis

We compared in-solution digestion procedures to determine optimal resuspension media for gel-free analyses. Both urea and thiourea have been used previously to increase protein solubility.^22,36 Nonionic and zwitterionic detergents have also been utilized to solubilize hydrophobic proteins.³⁷ These detergents are commonly used in preparing protein samples.³⁸ To investigate the efficacy of chaotropes such as urea and thiourea, and various detergents such as CHAPS and n-octylglucoside, at resuspending precipitated soybean seed proteins we compared various resuspension media containing different concentrations of these reagents. Proteins obtained from the phenolextraction method were resuspended in eight different resuspension media (see Protein resuspension section in Materials and methods), in triplicate. Addition of these resuspension buffers was followed by gentle vortexing to resuspend the pellets. Following trypsin digestion, peptides were subjected to LC-MS/MS analysis. Among these resuspension buffers, the minimal buffer that contained simply 50 mM Tris–HCl, pH 8.0 and 7 M urea showed the highest average spectral counts (Fig. 3). Furthermore, average spectral counts gradually reduced with increasing concentrations of zwitterionic and nonionic detergents such as CHAPS and n-octylglucoside, respectively. Elimination of all detergents post protein isolation improved the frequency of assigned tandem mass spectra as much as 3-fold.

Fig. 3 Comparison of resuspension media for gel-free protein digestion. Proteins obtained from phenolextraction method were resuspended in eight different buffers, subjected to mass spectrometry analysis and the spectral counts obtained. Error bars are standard deviation among three biological replicates. Precipitated protein was dissolved in the following resuspension buffers: (1) 50 mM Tris–HCl, pH 8.0, 7 M urea, or buffer 1 plus one of the following (2) 2 M thiourea, (3) 2 M thiourea, 0.1% CHAPS, (4) 2 M thiourea, 0.5% CHAPS, (5) 2 M thiourea, 2.0% CHAPS, (6) 2 M thiourea, 0.1% n-octylglucoside, (7) 2 M thiourea, 0.5% n-octylglucoside or (8) 2 M thiourea, 2.0% n-octylglucoside.

Because the minimal resuspension buffer produced the most peptides, possibly due to reduced ion suppression,^39–42 we focused further on the urea concentration of this buffer to further optimize digestion conditions. We compared resuspension media containing 50 mM Tris–HCl, pH 8.0 and urea concentrations ranging from 2 M to 7 M urea. Extracted and precipitated soybean proteins were resuspended in one of the seven urea resuspension buffers and visualized. Precipitated proteins were completely dissolved in 6 M and 7 M urea but not below 5 M urea (Fig. 4A). Resuspended samples were digested, separated by SDS-PAGE and silver stained (Fig. 4B). Most of the digestions showed little to no protein bands detectable by silver staining suggesting the protein digestion was efficient and complete. Only one sample, the non-reduction/alkylation 7 M urea, showed multiple protein bands. This result suggests that the efficiency of digestion was enhanced by reduction/alkylation procedures prior to digestion, supporting a previous observation.⁴³

Fig. 4 Comparison of urea concentration in minimal resuspension buffer needed for gel-free protein digestion. (A) Appearance of resuspended protein samples with different resuspension buffers. Proteins obtained from phenolextraction method were resuspended in different urea concentration (from 2 to 7 M urea), and under either reduction/alkylation or non-reduction/alkylation procedures. (B) Resuspension efficiency of six different urea concentrations was visualized by SDS-PAGE using silver staining. One microgram of protein or tryptic peptides was loaded onto a 15% SDS-PAGE gel. (C) Spectral counts were obtained in different resuspended samples with six different urea concentration containing 50 mM Tris–HCl, pH 8.0 either reduction/alkylation or non-reduction/alkylation procedures.

To determine optimal digestion conditions, tryptic peptides from samples digested in the presence of different urea concentrations were subjected to LC-MS/MS analysis (Fig. 4C). Among these resuspension buffers, 50 mM Tris–HCl, pH 8.0 and 5 M urea showed the highest average spectral counts. Furthermore, average spectral counts were slightly reduced when concentrations of urea deviated from 5 M, either higher or lower. Digestion of resuspended protein in 50 mM Tris–HCl, pH 8.0 and 5 M ureabuffer showed higher average spectral counts than digestion of proteins in 7 M ureabuffer. The result suggests that a minimal resuspension buffer such as 50 mM Tris–HCl, pH 8.0 and 5 M ureabuffer is optimal for signal detection and sensitivity.

Optimizing trypsin digestion efficiency

After proteinextraction and resuspension, in-solution protein digestion and MS analysis are the two remaining steps for optimization of quantitative gel-free MS. The advantage of trypsin is that it cleaves proteins on the carboxyl side of lysine and arginine residues producing peptides that can be easily predicted and typically within the preferred mass range for sequencing.⁴⁴ Furthermore, tryptic peptides are usually doubly protonated by ESI, ideal for peptide fragmentation by collision-induced dissociation (CID) during tandem mass spectrometry analysis.^39,45 Prior to digestion, Cys disulfide bonds are reduced and then alkylated. These procedures eliminate disulfide bridges enhancing digestion efficiency.⁴³ For complex samples, reduction and alkylation procedures are likely necessary for efficient tryptic digestions.

Trypsin digestion efficiency is a key factor in protein digestion.¹⁶ This aspect was investigated using different trypsin to protein ratios in the in-solution digestion of the total proteins with three biological replicates. Precipitated proteins using the phenolextraction method were resuspended with 50 mM Tris–HCl, pH 8.0 and 7 M urea to determine optimal trypsin levels for digestion. Tryptic peptides were then analyzed by LC-MS/MS to obtain spectral counts from the different digestion reactions. Comparison of trypsin to protein ratio (w/w) (1 : 25, 1 : 50, 1 : 75, and 1 : 100) showed a near linear increase in the average spectral counts with increasing trypsin levels and revealed the 1 : 25 ratio produced the highest average spectral counts (Fig. 5). This result indicated that the digestion efficiency was proportional to trypsin concentration. The increase in peptides was not due to trypsin autolysis as the database search was confined only to soybean proteins. Trypsin is the most expensive reagent for gel-free MS, and cost must be considered in the development of any high-throughput workflow. From a cost-benefit perspective, a two-fold increase in trypsin produced only a 10% gain in spectral counts, when comparing the results from the 1 : 50 and 1 : 25 ratios. Therefore, a 1 : 50 ratio (trypsin : protein) was employed to in-solution digestion of mature soybean seed proteins.

Fig. 5 Comparison of trypsin concentration for gel-free protein digestion. Proteins obtained from phenolextraction method were resuspended in 50 mM Tris–HCl, pH 8.0, 7 M ureabuffer and compared for digestion efficiency as a function of trypsin concentration. Spectral counts were obtained from in-solution digestions at different trypsin to protein ratio (w/w, 1 : 25 to 1 : 100). Error bars are standard deviation from three biological replicates.

Conclusions

We report an optimized workflow to support a gel-free quantitative proteomic approach for analyzing protein content in mature soybean seed. This study shows that the phenolextraction procedure is an efficient, cost effective method that consistently yielded the highest amount of protein from powdered seed. Resuspension of precipitated protein (after phenolextraction) in buffered urea was optimal for digestion and spectral counting by MS. Furthermore, assigned spectral counts were increased by elimination of all detergents (e.g. CHAPS and Triton X-100) post-protein isolation. From a cost-benefit perspective, utilization of trypsin at a 1 : 50 ratio is recommended compared to higher dilutions typically employed in high-throughput protocols. The optimized workflow has the potential to quantify more peptides over a broader dynamic range due to the enhanced sensitivity. Besides improved sensitivity, the optimized workflow offers the advantages of simplicity, reduced cost, and better reproducibility. This protocol is suitable for high-throughput, gel-free quantitative proteomic analysis of soybean seed. Quantitative MS methods and the optimized preparation methods described here for soybean may be particularly useful with targeted profiling approaches to measuring specific proteins such as the many allergens present in soybean seed.

Abbreviations

LC-MS/MS

2-DGE	Two-dimensional gel electrophoresis
SDS-PAGE	SDS-polyacrylamide gel electrophoresis
	Liquid
	chromatography
	-
	tandem mass spectrometry

Acknowledgements

This work was supported by a grant from the International Life Science Institute-Health Environmental Science Institute and the National Science Foundation-Plant Genome Grant DBI-0332418.

Reference

1. J. J. Thelen and S. C. Peck, Plant Cell, 2007, 19, 3339–3346.
2. A. Görg, W. Weiss and M. J. Dunn, Proteomics, 2004, 4, 3665–3685.
3. S. Chen and A. Harmon, Proteomics, 2006, 6, 5504–5516.
4. S. Wolff, A. Otto, D. Albrecht, J. S. Zeng, K. Bűttner, M. Glűchmann, M. Hecker and D. Becher, Mol. Cell. Proteomics, 2006, 5, 1183–1192.
5. K. Gevaert, P. Van Damme, B. Ghesqière, F. Impens, L. Martens, K. Helsens and J. Vandekerckhove, Proteomics, 2007, 7, 2698–2718.
6. L. Nie, G. Wu and W. Zhang, Crit. Rev. Biotechnol., 2008, 28, 297–307.
7. G. K. Agrawal, M. Hajduch, K. Graham and J. J. Thelen, Plant Physiol., 2008, 148, 504–518.
8. J. E. Froehlich, C. G. Wilkerson, W. K. Ray, R. S. McAndrew, K. W. Osteryoung, D. A. Gage and B. S. Phinney, J. Proteome Res., 2003, 2, 413–425.
9. M. Z. Wang, J. Q. Wu, J. B. Dennison, A. S. Bridges, S. D. Hall, S. Kornbluth, R. R. Tidwell, P. C. Smith, R. D. Voyksner, M. F. Paine and J. E. Hall, Proteomics, 2008, 8, 4186–4196.
10. A. S. Haqqani, J. F. Kelly and D. B. Stanimirovic, Methods Mol. Biol., 2008, 439, 241–256.
11. A. J. Link, J. Eng, D. M. Schieltz, E. Carmack, G. J. Mize, D. R. Morris, B. M. Garvik and J. R. Yates, III, Nat. Biotechnol., 1999, 17, 676–682.
12. M. P. Washburn, D. Wolters and J. R. Yates, III, Nat. Biotechnol., 2001, 19, 242–247.
13. N. Nagaraj, A. Lu, M. Mann and J. R. Wiśniewski, J. Proteome Res., 2008, 7, 5028–5032.
14. S. C. Kim, Y. Chen, S. Mirza, Y. Xu, J. Lee, P. Liu and Y. Zhao, J. Proteome Res., 2006, 5, 3446–3452.
15. P. M. van Midwoud, L. Rieux, R. Bischoff, E. Verpoorte and H. A. Niederlander, J. Proteome Res., 2007, 6, 781–791.
16. M. B. Strader, D. L. Tabb, W. J. Hervey, C. Pan and G. B. Hurst, Anal. Chem., 2006, 78, 125–134.
17. M. Gaspari, V. Abbonante and G. Cuda, J. Sep. Sci., 2007, 30, 2210–2216.
18. J. Norrgran, T. L. Williams, A. R. Woolfitt, M. I. Solano, J. L. Pirkle and J. R. Barr, Anal. Biochem., 2009, 393, 48–55.
19. S. S. Natarajan, C. Xu, H. Bae, T. J. Caperna and W. M. Garrett, J. Agric. Food Chem., 2006, 54, 3114–3120.
20. B. P. Mooney and J. J. Thelen, Phytochemistry, 2004, 65, 1733–1744.
21. Y. Nanjo, L. Skultety, Y. Ashraf and S. Komatsu, J. Proteome Res., 2010, 9, 3989.
22. S. Natarajan, C. Xu, T. J. Caperna and W. M. Garrett, Anal. Biochem., 2005, 342, 214–220.
23. H. Liu, R. G. Sadygov and J. R. Yates, III, Anal. Chem., 2004, 76, 4193–4210.
24. B. Zybailov, M. K. Coleman, L. Florens and P. Washburn, Anal. Chem., 2005, 77, 6218–6224.
25. M. Ferro, S. Brugière, D. Salvi, D. Seigneurin-Berny, M. Court, L. Moyet, C. Ramus, S. Miras, M. Mellal, S. Le Gall, S. Kieffer-Jaquinod, C. Bruley, J. Garin, J. Joyard, C. Masselon and N. Rolland, Mol. Cell. Proteomics, 2010, 9, 1063–1084.
26. W. Zhu, J. W. Smith and C.-M. Huang, J. Biomed. Biotechnol., 2010, 2010, 840518.
27. T. Berkelman, T. Stenstedt, B. Bjellqvist, N. Laird, M. McDowell, I. Olsson and R. Westermeier, 2D Electrophoresis Using Immobilized pH Gradients: Principles and Methods, American Biosciences, Piscataway, NJ, 1998.
28. E. Herman, R. Helm, R. Jung and A. J. Kinney, Plant Physiol., 2003, 132, 36–43.
29. W. Hurkman and C. Tanaka, Plant Physiol., 1986, 81, 802–806.
30. M. Hajduch, A. Ganapathy, J. W. Stein and J. J. Thelen, Plant Physiol., 2005, 137, 1397–1419.
31. M. M. Bradford, Anal. Biochem., 1976, 72, 248–254.
32. U. K. Laemmli, Nature, 1970, 227, 680–685.
33. S. E. Stevenson, Y. Chu, P. Ozias-Akins and J. J. Thelen, J. Proteomics, 2009, 72, 555–566.
34. W. Yan and S. Che, Briefings Funct. Genomics Proteomics, 2005, 4, 1–12.
35. S. C. Carpentier, E. Witters, K. Laukens, P. Deckers, R. Swennen and B. Panis, Proteomics, 2005, 5, 2497–2507.
36. J. Bergquist, M. Palmblad, M. Wetterhall, P. Håkansson and K. E. Markides, Mass Spectrom. Rev., 2002, 21, 2–15.
37. C. Finnie, Plant Proteomics, Annual Plant Reviews, Blackwell Publishing, 2006, vol. 28, pp. 1–32.
38. N. Zhang and L. Li, Rapid Commun. Mass Spectrom., 2004, 18, 889–896.
39. H. Steen and M. Mann, Nat. Rev. Mol. Cell Biol., 2004, 5, 699–711.
40. T. M. Annesley, Clin. Chem., 2003, 49, 1041–1044.
41. F. Moragues and C. Igualada, Anal. Chim. Acta, 2009, 637, 193–195.
42. A. Hirabayashi, M. Ishimary, N. Manri, T. Yokosuka and H. Hanzawa, Rapid Commun. Mass Spectrom., 2007, 21, 2860–2866.
43. B. Q. Tran, H. Loftheim, L. Reubsaet and T. Greibrokk, J. Sep. Sci., 2008, 31, 2913–2932.
44. K. Gevaert, P. Van Damme, B. Ghesquière, F. Impens, L. Martens, K. Helsens and J. Vandekerckhove, Proteomics, 2007, 7, 2698–2718.
45. J. Z. Kiser, M. Post, B. Wang and M. Miyagi, J. Proteome Res., 2009, 8, 1810–1817.

---