From Cells to Peptides: “One-Stop” Integrated Proteomic Processing

Feb 9, 2013 - From Cells to Peptides: “One-Stop” Integrated Proteomic Processing Using ... Citation data is made available by participants in Cros...
0 downloads 0 Views 331KB Size
Download PDF
Technical Note pubs.acs.org/jpr

From Cells to Peptides: “One-Stop” Integrated Proteomic Processing Using Amphipols Zhibin Ning, Deeptee Seebun, Brett Hawley, Cheng-Kang Chiang, and Daniel Figeys* Ottawa Institute of Systems Biology, Department of Biochemistry, Immunology and Microbiology, University of Ottawa, Ontario, Canada S Supporting Information *

ABSTRACT: In proteomics, detergents and chaotropes are indispensable for proteome analysis, not only for protein extraction, but also for protein digestion. To increase the protein extraction efficiency, detergents are usually added in the lysis buffer to extract membrane proteins out of membrane structure and to maintain protein in solutions. In general, these detergents need to be removed prior to protein digestion, usually by precipitation or ultrafiltration. Digestion often takes place in the presence of chaotropic reagents, such as urea, which often need to be removed prior to mass spectrometry. The addition and removal of detergents and chaotropes require multiple steps that are time-consuming and can cause sample losses. Amphipols (APols) are a different class of detergents that have physical and solubilization properties that are distinct from conventional detergents. They have primarily been used in protein structure analysis for membrane protein trapping and stabilization. Here, we demonstrate a simple and rapid protocol for total and membrane proteome preparation using APols. We demonstrate that APols added for cell lysis help maintain the proteome in solution, are compatible with protein digestion using trypsin, and can readily be removed prior to mass spectrometry by a one-step acidification and centrifugation. This protocol is much faster, can be performed in a single tube, and can readily replace the conventional detergent/chaotrope approaches for total and membrane proteome analysis. KEYWORDS: one-stop, proteomics, amphipols, sample preparation, membrane protein

■

INTRODUCTION The study of the proteome by mass spectrometry requires the lysis and recovery of proteins from cells and tissues. Cells and organelles are enclosed by membranes, which need to be disrupted in order to release the intracellular or intraorganelle proteins. In addition, many proteins are associated with membrane, either through transmembrane domains, interactions, or anchoring side groups. Often, transmembrane proteins are hydrophobic and difficult to extract into aqueous solution. Typically, chemical and/or physical means are used to disrupt membranes. One important class of chemical used to disrupt membranes is detergent, also called surfactant, a class of amphiphilic chemicals with both hydrophobicity and hydrophilicity moieties.1 Detergents have the ability to disrupt biological membrane structure, to denature protein’s tertiary structure and help keep hydrophobic proteins in a soluble state in aqueous phase. Sodium dodecyl sulfate (SDS), NP-40, Tween-20, Triton X-100, and sodium-deoxycholate2 are often used for that purpose. Unfortunately, the same detergents at concentration used to disrupt membranes also interfere with proteolytic enzymes, such as trypsin, employed in proteomics to transform proteins into peptides for mass spectrometric analysis. Therefore, steps must be taken to remove the © 2013 American Chemical Society

detergents prior to the use of a proteolytic enzyme. Furthermore, most of detergents interfere with the chromatography and mass spectrometry of peptides by high-performance liquid chromatography/electrospray ionization tandem mass spectrometry (HPLC−ESI−MS). In many instances, removing the detergent would lead to protein loss due to insolubility. This would be particularly the case for hydrophobic proteins. Therefore, chaotrophic agents are added to the solution in order to maintain the proteins in solution for the process of reduction, alkylation, and proteolysis. These chaotropes generally include compounds such urea and guanidine chloride, which are not hydrophobic, but have a strong ability of interfering with intra- and intermolecular interactions mediated by noncovalent forces such as hydrogen bonds, van der Waals forces, and hydrophobic effects.3 Fortunately, the activity of proteolytic enzymes is typically tolerant of low concentration of chaotropes. However, most chaotropic agents are not suitable for MS and, therefore, have to be removed following proteolytic digestion. Received: November 11, 2012 Published: February 9, 2013 1512

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519

Journal of Proteome Research

Technical Note

solution during protein processing and does not interfere with trypsin activity. Moreover, we also demonstrated that APols are also suitable for the membrane proteome digestion. Therefore, there is no need to employ other detergents or chaotropic agents. Although APols are neither cleavable, nor volatile, they are readily removed by simple acidification and centrifugation. On the basis of the properties of APols, we developed a simple and rapid “one-tube” proteomic sample preparation strategy, which employs APols as the only detergent. Compared to classical in-solution digestion, the usage of APols removes the need for protein precipitation or ultrafiltration and sample desalting. Our results show that the APols strategy can identify more proteins by greatly reducing sample loss compared with conventional approaches.

Efforts have been made to addresses some of the drawbacks of detergents and chaotropes in proteomic analysis. In particular, an acid labile surfactant (ALS) was introduced by Yu et al.4 This surfactant minimally affects trypsin activity and, when required, can be broken into hydrophilic and hydrophobic molecules by reducing the pH, which eliminates the need of surfactant removal prior to MS analysis. However, ALS has not gain widespread use in the proteomic community. Moreover, it was shown that the hydrophobic byproduct molecules from ALS deplete hydrophobic peptides, which can only be recovered by extensive washing using isopropanol.5 Another type of volatile surfactant, called perfluorooctanoic acid (PFOA), was recently proposed as a replacement for SDS during protein extraction and trypsin digestion.6 Trypsin has a relatively high tolerance limit for PFOA concentration. In this approach, the reduction and alkylation reagents are also substituted by volatile chemicals. Unfortunately, PFOA is difficult to remove following proteolytic digestion. In reality, removal of PFOA requires several reconstitution and overnight SpeedVac.7 PFOA was also found not to be as effective as conventional surfactant to solubilize proteins. The ammonium salt of PFOA, ammonium perfluorooctanoate (APFO) was suggested to be more suitable.7 Furthermore, both PFOA and APFO are not very MS friendly, and have some potential safety hazards. Amphipols (APols) are a relatively newer class of polymer aimed at stabilizing membrane proteins in aqueous phase. APols were introduced by Jean-Luc Popot in 1996.8 An extensive review of the APols properties was published.1 Briefly, there are many types of APols, but the most comprehensively studied APols are called A8-35. A8-35 consists of about 35 acrylate residues to which the carboxylates have been randomly grafted with octylamine (∼9 of 35) or isopropylamine (∼14 out of 35). The average molecular weight (MW) is ∼4.3 kDa, and ∼9 to 10 molecules can form a particle. The free carboxylate group makes the polymer very water-soluble, up to 200 mg/ mL. However, at low pH, the carboxylates are protonated and APols readily aggregate. Interestingly, the critical micelle concentration (CMC) of A8−35 is very low, about 0.002 mg/mL. Generally, ionic detergents, such as SDS, can lead to protein denaturation and on this basis are classified as strong detergent. Nonionic detergents, such as TritonX, or zwitterionic detergents such as CHAPS, can only compromise the membrane structure and facilitate native protein extraction, but do not readily cause protein denaturing and are often referred to as “mild detergents”. APols are milder than other surfactant, but have higher affinity for proteins.9 The working concentration for membrane protein stabilization is generally in the order of 0.05−0.5 mg/mL depending on proteins.8 APols have multiple hydrophobic chains in a single molecule and therefore have excellent affinity for membrane proteins. So far, APols have been used in protein crystallization, NMR analysis,9 and as universal molecular adapters to immobilize membrane proteins onto solid supports.10 The most recent usage has extended to MALDI analysis of model transmembrane proteins,11 in which the APols were used to trap the membrane introduced directly onto a MALDI plate. Here, we tested whether APols could be used to develop a simple and rapid protein preparation protocol. We found that in terms of protein extraction APols is as efficient as radioimmunoprecipitation assay buffer (RIPA) buffer, which is recognized as a harsh buffer for general cell lysis.12,13 Furthermore, APols can be used to maintain proteins in

■

EXPERIMENT

Chemical and Materials

Amphipol A8-35 was bought from Affymetrix. Urea, dithiothreitol (DTT), iodoacetamide (IAA), ammonium bicarbonate (ABC), formic acid (FA), citric acid, Triton X-114, Triton 100, CHAPS, sodium dodecyl sulfate (SDS), and N-α-benzoyl-Larginine ethyl ester hydrochloride (BAEE) were obtained from Sigma Aldrich (St. Louis, MO). RapiGest SF Surfactant was bought from Waters Corporation (Milford, MA). Water and acetonitrile (ACN) for HPLC were obtained from JT Baker, Phillipsburg NJ. Trypsin was purchased from Worthington Biochemical Corp. Bio-Rad protein assay kit II (500-0002) and DC protein assay kit II (500-0112) were from Bio-Rad. All of the chemicals were of analytical purity grade except ACN and FA, which were of HPLC grade. All the water used in the experiment was prepared using a Milli-Q system (Millipore, Bedford, MA). Trypsin Activity Assay

Trypsin activity was measured by monitoring the absorbance at 253 nm, which reflects the hydrolysis of trypsin substrate, N-αbenzoyl-L-arginine ethyl ester (BAEE), to N-α-benzoyl-Larginine(BA) on a DU-640B spectrophotometer from Beckman.4,14 The BAEE hydrolysis follows zero-order kinetic, and therefore, the slope is linear to the trypsin activity. BAEE (0.2 mM) was prepared in 50 mM ABC (pH 7.8) and spiked with 1 μg of trypsin. UV absorbance was recorded in 20s intervals; the first 4 min of reaction was used to calculate the slope. Trypsin activities in the presence of different concentrations of APols, urea, and SDS were normalized to control one without any other additives. Sample Preparation

Membrane protein preparation procedure was modified from reference.15 Briefly, HuH7 cells were grown to 80% confluence in 10 cm dish. Following their harvesting, they were lysed using 1 mL of 2% TritonX-114 in PBS, vortexed for 5 min, then centrifuged at 13 000g for 10 min to remove in-soluble part at 4 °C. The supernatant was transferred to a new tube to perform phase separation at 37 °C for 10 min, and then centrifuged again at room temperature at 13 000g for 10 min. The lower detergent phase was washed twice by vortexing at 0 °C and centrifugation at room temperature. Proteins in lower detergent phase were treated as membrane protein fraction, while the hydrophilic proteins were in the upper aqueous phase. Proteins were precipitated in cold acetone overnight and washed twice. For classical protein extraction, RIPA buffer was added to lyse cells, and the mixture was vortexed for 10 min, followed by 1513

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519

Journal of Proteome Research

Technical Note

performed with internal recalibration (“Lock Mass”).18 On the Orbitrap, the charge state rejection function was enabled, with single and “unassigned” charge states rejected.

sonication. Then, 6-fold volume of cold acetone was added to precipitate proteins and remove detergents. Urea (8 M) in 50 mM ABC was used to reconstitute proteins. Reduction and alkylation were done by adding DTT to a final concentration of 20 mM at 56 °C for 30 min followed by 60 mM IAA at room temperature. The solution was then diluted 10 times by adding 50 mM ABC. Trypsin was added to achieve a protein−enzyme ratio of 20:1. Digestion was performed at 37 °C overnight, with continuous shaking. Digested peptides were then desalted on Sep-Pak column (Waters) and dried down in SpeedVac (ThermoFisher Scientific, San Jose, CA). Dried peptides were reconstituted in 0.5% (v/v) FA, subjected to SCX fractionation or directly analyzed by MS. SCX fractionation was performed on StageTip column,16 eluted by pH buffers, adjusted to pH 3.0,4.0, 5.0, 6.0, 8.0, and 10.0 by adding ammonium hydroxide to 20 mM citric acid. For APols protein extraction, 1 mg/mL APols in 50 mM ABC was used to lyse cells. The solution was also vortexed and sonicated; however, no protein precipitation ultrafiltration was performed. The protein solution was boiled for 5 min prior to DTT/IAA reaction. Trypsin was directly added to the solution to perform digestion. After digestion, 5% FA (v/v) was added to lower the pH to 3 and the solution became cloudy. The aggregated APols were spun down at 12 000 rpm for 5 min and the recovered supernatant was either dried down or used directly for analysis. SCX fractionation was done the same way as above. To test the amount of peptides depleted in the APols precipitates, the spun-down APols were washed using 80% (v/ v) ACN and isopropyl alcohol. The solution was then diluted twice or more with 0.5% (v/v) FA to let APols precipitate. For the digestion of precipitated membrane proteins, aliquot of membrane protein pellet were reconstituted in 8 M urea, 2% SDS, 1 mg/mL APols or 1 mg/mL ALS. Samples in APols and ALS were sonicated to help to dissolve the pellet. FASP preparation was done as previously described17 using 30K filter. ALS digestion was done as previously described.4 The APols and urea digestion are the same as above.

Database Search and Data Analysis

The raw files generated by the LTQ-Orbitrap were processed and analyzed using MaxQuant, Version 1.3.0.519 using the Uniprot protein fasta database (2012, July version), including commonly observed contaminants. The following parameters were used: cysteine carbamidomethylation was selected as fixed modification; methionine oxidation; protein N-terminal acetylation; and enzyme specificity was set to trypsin. Up to two missing cleavages of trypsin were allowed. Precursor ion mass tolerances were 7 ppm, and fragment ion mass tolerance was 0.8 Da for MS/MS spectra. If the identified peptide sequences from one protein were equal to or contained within another protein’s peptide set, then the proteins were grouped together and reported as one protein group. The false discovery rate (FDR) for peptide and protein was set at 1% and a minimum length of six amino acids was used for peptides identification. GO analysis was done in Perseus, which comes with Maxquant. Raw files from LTQ were searched using Mascot version 2.220 against an homemade database of SGD plus frequently used standard proteins for BSA analysis, with the same setting for amino acid modifications as in Maxquant, with the exception of 0.5 Da for precursor ions. The .dat files generated by Mascot were parsed and filtered by Buildsummary,21 with a peptide FDR of 1%. GRAVY value22 calculation and statistic were done using tools also provided by BuildSummary.21 All Comparisons between samples or methods were always based on the same criteria.

■

RESULTS AND DISCUSSIONS

Whole Proteome Extraction from Cells Using APols

The study of the membrane proteome by mass spectrometry requires the disruption of the cells, and the solubilization and maintenance in solution of membrane proteins. Typically, this is achieved through the combination of mechanical means, detergents, and chaotropic agents. Unfortunately, many of the common detergents and chaotropic agents are not compatible with proteolysis or mass spectrometry. Instead, we proposed to replace detergents and chaotropic agents with APols. APols are different from conventional detergents as they are polymers to which hydrophobic and hydrophilic moieties have been grafted.1 When interacting with proteins, APols form threedimensional functional region with steric hydrophobic regions tightly interacting with the hydrophobic section of membrane proteins. In contrast, the hydrophilic regions of APols help maintain the complex in aqueous solution. In general, because of their mild detergent properties, APols are thought to be unsuitable to directly extract proteins from membranes.8,23 However, this observation still needs to be qualified.1 We postulated that sonication could help compensate for the milder detergent properties of APols. We first tested whether APols could lyse cell pellets as well as conventional lysis buffers. Briefly, three different lysis buffers were used to compare the extraction efficiency: classical RIPA buffer, which has a high concentration of detergent, as a positive control; PBS as a negative control; and APols buffer of 1 and 10 mg/mL in PBS. Aliquot cell pellets were lysed using the three cell lysis buffers, and then subjected to pulse sonication. The resulting solutions became clear except for the aliquot dissolved in PBS. After spinning down the insolubilized

MS Analysis

All MS analyses were done by HPLC−ESI−MS/MS. The system consisted of an Agilent 1100 micro-HPLC system (Agilent Technologies, Santa Clara, CA) coupled with an LTQ or LTQ-Orbitrap mass spectrometer (ThermoFisher Scientific, San Jose, CA) equipped with a nano-electrospray interface operated in positive ion mode. The mobile phases consisted of 0.1% (v/v) FA in water as buffer A and 0.1% (v/v) FA in ACN as buffer B. Peptide separation was performed on a 75 μm × 100 mm analytical column packed in-house with reverse phase Magic C18AQ resins (3 μm; 120-Å pore size; Dr. Maisch GmbH, Ammerbuch, Germany). Briefly, the sample was loaded on the column using 98% buffer A at a flow rate of 1.5 μL/min for 15 min. Then, a gradient from 5% to 30% buffer B (20− 50% for APols depletion test) was performed in 60, 120, or 180 min at a flow rate of ∼300 nL/min obtained from splitting a 20 μL/min through a restrictor. On the Orbitrap MS and LTQ MS, the method consisted of one full MS scan from 300 to 1700 m/z followed by data-dependent MS/MS scan of the 5 most intense ions, a dynamic exclusion repeat count of 2, and a repeat duration of 90 s. In addition, for the experiments on the Orbitrap MS, the full MS was performed in the Orbitrap analyzer with R = 60 000 defined at m/z 400, while the MS/MS analysis were performed in the LTQ MS. To improve the mass accuracy, all the measurements in Orbitrap mass analyzer were 1514

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519

Journal of Proteome Research

Technical Note

mL APols buffer can be rapidly estimated by measuring the absorbance at 280 nm using the NanoDrop technology.

cell debris, samples were then subjected to gel-analysis (Figure 1). The results from Figure 1A indicate that APols buffers are as

APols Are Fully Trypsin Compatible

A dichotomy in proteome processing is the need to keep the proteins denatured to expose possible cleavage sites, while the proteolytic enzyme needs to be in its native state to remain active. Therefore, detergents that could accommodate both requirements would be favorable. Recently, some detergents were reported to be trypsin compatible.4,6,7 In addition, APols have been previously demonstrated to be suitable to maintain enzymatic activity in solution.25 Therefore, we tested whether trypsin activity could be maintained in the presence of APols. Briefly, trypsin activity was tested in buffer containing different concentrations of APols, urea, and SDS. The kinetics of trypsin digestion of BAEE has been previously shown to be linear.4 Hence, the initial slope of the kinetic curve can be used to represent the relative activity. We determined that the impact of APols on the activity of trypsin was minimal (Figure 2, less than 10% reduction in activity) and was independent of APols concentration.

Figure 1. Extraction efficiency test of APols and RIPA buffer for total proteome. (A) 1D SDS-PAGE display of cell lysate extracted by PBS, 1 mg/mL APols in PBS, 10 mg/mL APols in PBS, and RIPA buffer. The samples were extracted from the same number of cells. (B) Comparison between APols and RIPA after precipitation. Equal aliquots of cells were extracted using 1 mg/mL APols in 50 mM ABC and RIPA buffer. The RIPA extracted proteins were precipitated in cold acetone overnight, and reconstituted in the same volume of 8 M urea as initially.

good as RIPA buffer for whole proteome extraction. From the number and densitometry of the gel bands, it is clear that APols and RIPA can extract much more proteins than PBS, especially in the high molecular weight region, where large transmembrane proteins appear. Moreover, during the process of lysis, PBS could only disperse the cells, whereas both APols and RIPA could immediately lyse the cells and released the DNA rendering the solution viscous. It appears that, for cultured cells, APols based buffers are able to lyse the plasma and nuclear membranes. In addition, the results from Figure 1A indicate no obvious difference in extraction performance between 1 and 10 mg/mL APols buffers. The critical micelle concentration (CMC) of APols is as low as 0.002 mg/mL.24 Detergents generally function above their CMC; therefore, we decided to employ the concentration of 1 mg/mL in order to reduce cost, and reduce the protein loss during the APols removal. We also tested whether APols are compatible with protein quantitation by three different methods. Briefly, we performed protein quantitation using a NanoDrop at 280 nm, classical BioRad Bradford assay at 595 nm, and detergent compatible (DC) method at 750 nm. We observed no significant difference in protein quantitation between the blank, 1 and 10 mg/mL of APols, for the Bradford and DC methods. Moreover, there was also no significant difference of absorbance at 280 nm between BSA in water and BSA in APols. There are no functional groups such as aromatic in APols that can absorb at 280 nm (Figure S1). At 1 mg/mL, APols’ absorbance at 200 nm is negligibly elevated compared to the background absorbance of water, but does not interfere with absorbance measurement at 280 nm. Therefore, the concentration of proteome dissolved in 1 mg/

Figure 2. Relative trypsin activity in different buffers. Trypsin activity was determined by the initial slope (4 min) of the kinetic curve (see main text). The average trypsin acidity in 50 mM ABC was normalized to 100%. Trypsin activity in variable concentrations of APols, urea, and SDS is normalized to the relative activity to the control. Each experiment was repeated three times.

We then tested whether APols were compatible with trypsin digestion of whole proteome. We tested the compatibility with the digestion of hydrophilic proteome and membrane protein preparation respectively. To distinguish cell lysis from protein digestion, we pre-extracted hydrophilic and hydrophobic proteomes from cells. These proteomes were then used to compare the performances of digestion in APols and conventional chaotropes systems (urea). We employed Triton X-114 phase partitioning to prepare relatively enriched hydrophilic proteins and hydrophobic membrane protein preparations from HuH7 cells.15 Triton X-114 is water miscible at 4 °C (below the cloud point of 22 °C in the presence of 150 mM NaCl). However, it partitions into two phases (hydrophilic and hydrophobic) at room temperature. Triton X-114 detergents selectively bind to hydrophobic membrane proteins with 90% or more of integral membrane proteins partitioning in the Triton-114 phase, while the hydrophilic proteins partition in the lighter aqueous phase. Briefly, HuH7 cell pellet was treated 1515

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519

Journal of Proteome Research

Technical Note

Table 1. Identification summary (A) APols and Chaotrope (Urea) Effectsa urea

p-valueb

1806/1630/1618 600/588/584 10308/6174/5941 2108/1606/1580

0.0943 0.05824 0.3202 0.3526

APols Hydrophilic proteins preparation, 1D analysis Membrane protein preparation, 2D analysis

Unique Peptide 1700/1464/1379 Protein Group 661/619/611 Unique Peptide 10621/6117/6399 Protein Group 2184/1384/1458 (B) Unique Peptides and Proteinsc

Membrane protein preparation Membrane protein recovered from precipitated APols (8× input, 10× APols concentration, 10× peptide concentration) (C) Comparison of the One-Stop and Classical RIPA/Urea Strategiesd Intensity (E10) Unique Peptide Protein Group

unique peptide

protein group

950 230

337 121

one-stop

RIPA/Urea

p-valueb

8.12/4.55/3.89 10232/9318/9202 2229/2271/2322

5.77/3.1/3.17 6714/6682/6925 1809/1893/1912

0.01454 0.006541 0.0007822

a

APols and chaotrope (Urea) effects on the identification of membrane and hydrophilic proteins. Identical aliquots of hydrophilic and hydrophobic protein extracts from HuH7 cells were digested and processed in the presence of APols and chaotrope by three independent paired repeats. bPvalues are calculated on fold change of paired experiment by t.test command in R. cUnique peptides and proteins observed by 1D HPLC−ESI− MSMS analysis following digestion in APols and recovered from the APols pellet. dComparison of the one-stop strategy and the classical RIPA/Urea strategy for the processing of cells for proteomic analysis by three independent paired repeats. P-values are calculated on fold change of paired experiment by t.test command in R.

Figure 3. The effects on BSA digest with the presence or removal of APols. (Upper panel) Normal base peak of 1 pmol BSA digest. (Middle panel) 1 pmol BSA with 8 mL/mL APols. (Lower panel) The same sample with middle panel, with APols removal. Most of the major peptide’s peaks can be recovered by APols removal (“*” labeled). Peptides peaks labeled by “Δ” in middle panel are those not affected by APols and present in all three analyses.

centrifuged to remove APols (discussed below). The digested membrane fraction was fractionated by pH steps on SCX StageTip column ahead of MS analysis in order to get more identification. The proteome identification in terms of unique peptide and protein group numbers was statistically identical between the samples digested in APols and urea (Table1A) for both hydrophilic and hydrophobic protein analysis.

with Triton X-114 to obtain two protein phases, and then precipitated by cold acetone. Aliquots from the two protein pellets (hydrophobic and hydrophilic) were redissolved in 10 mg/mL of APols or 8 M urea in 50 mM ABC, and were reduced and alkylated. Then, the protein samples were diluted 10 times and subjected to trypsin digestion. The urea sample was desalted, while the APols sample was acidified and 1516

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519

Journal of Proteome Research

Technical Note

Particularly for membrane proteins, in both digest methods, about 50% of the proteins were annotated or predicted as membrane proteins (Table S1). Even for membrane proteins, the number of unique peptide and protein identified using the APols method is statistically identical to the chaotrope approach. We also did not find any differences in hydrophobicity and misscleavage site (Figure S2) for the peptides identified and numbers of predicted transmembrane domains (Figure S3) using the two approaches. Therefore, trypsin digestion in APols buffer is as good as the conventional chaotropic based approach. We further compared APols with previously described FASP and ALS (RapiGest) methods using the same amount of membrane pellet without SCX peptide fractionation. We found that FASP identified a statistically higher number of peptides (P = 0.01) by about 14% compared to APols. However, the number of identified proteins were statistically identical (P = 0.2) with a total of 1103 proteins groups by SDS combined FASP and 1081 protein groups by APols (Table S2). Furthermore, no differences were observed in the number of membrane proteins (Table S3), and transmembrane domain prediction (Figure S4) identified. Finally, both APols and FASP statistically outperformed ALS in terms of the number of identified peptides and proteins. All the above suggests that APols perform at least as good as the present sample preparation strategies.

recovered. The signal intensity was also not affected by the addition/removal of APols (Figure S6). It was previously reported that for other approaches, such as ALS, some very hydrophobic peptides can be depleted during the processing of the sample.5 We did not observe any biases using APols during the digestion of BSA (Table S4). Furthermore, we tested whether APols would bias the hydrophobic peptide recovery in complex hydrophobic protein extracts. Briefly, the Triton-X114 extracted membrane proteins from HuH7 cells were subjected to trypsin digestion in the presence of APols followed by the precipitation of APols. The supernatant was removed and the APols precipitated were washed twice with 0.1% (v/v) FA to reduce nonspecific bindings. Then, the APols precipitates were back-extracted using first 80% (v/v) ACN and then 100% isopropyl alcohol. Any losses of peptides in the APols precipitates would be impacted by the initial concentration of APols and the hydrophobicity of the peptides. Very few unique peptides (81) were identified for back-extracted fractions from APols precipitate from 100 μg of protein digested in 1 mg/mL of APols. Moreover, even when the amount of protein was increased 8-fold, the concentration of APols increased 10-fold, and the final peptide concentration increased 10-fold (dissolved in 1/10 volume as previously), only 230 unique peptides were identified (Figure S7). These recovered peptides have on average more misscleavage site and are more hydrophobic (Figure S8). Overall, lower concentration of APols (1 mg/mL) does not appear to bias the proteome processing and recovery.

APols Can Be Easily Removed after Digestion

APols are mild detergents and polymers. Typically, detergents and polymers can cause serious band broadening in HPLC and ion suppression in MS analysis of peptides. As expected, APols’ MS analysis has a typical distribution for detergent/polymer (Figure S5). As well, the impact of APols on the analysis of BSA digest by HPLC-ESI-MSMS is illustrated in Figure 3. Clearly, the presence of APols in the peptide mixture (middle panel) leads to strong ion suppression in the mass spectrometer compared to the peptide mixture without APols (top panel). This is particularly the case during the early stage of the reverse phase gradient (5−35% ACN). This directly impacts the identification of the relatively hydrophilic peptides (Table S4). Interestingly, the ion suppression due to APols can be reduced by decreasing the concentration of APols. In particular, the effect is barely observable when the concentration of APols is lowered to 1 mg/mL which is the working solution. In contrast to other detergents,5 APols does not appear to cause band broadening in the chromatography. APols have very good solubility at neutral and basic pH due to the presence of free carboxylate groups in the structure. The protonation of the carboxylate groups changes the properties of the whole molecule (more hydrophobic) which leads to a reduced solubility and the formation of APols aggregates. This is very similar to sodium deoxycholate.2 This property provides a very convenient way to remove APols prior to MS analysis. Briefly, following protein digestions, APols can be readily removed through acidification of the solution followed by centrifugation. The supernatant contains the peptides free of APols. Formic acid was added, following digestion in APols, until the solution became cloudy. The transition point for aggregation of APols solution is around pH 3 to 4. The APols aggregates can then be readily precipitated by centrifugation. As shown in Figure 3 lower panel, following APols removal in BSA digest, the ion suppression cannot be observed anymore. Additionally, the base peak is similar to the APols free digestion of BSA (upper panel) with most of the peptide peaks

A One-Stop Workflow Using APols

Sample losses are common in complex proteomic processing protocols. Usually, one step required in proteomic protocols is the removal of the detergent used to solubilize the proteome either by precipitation or continuous ultrafiltration, such as FASP.17 Protein precipitation can cause serious sample losses. It is particularly acute when dealing with minute amounts of proteins. Based on our experience, ∼25% or more of the proteins are lost during precipitation and this can be compounded when dealing with lower quantity and hydrophobic samples. Even with FASP, nonspecific absorption and sample-flow through are believed to be an important source of sample losses. Sample loss in FASP was estimated to be 50%.17 We postulated that a one-tube protocol based on APols could be developed for the processing of proteome and that this simplified protocol would greatly reduce sample loss. In this protocol, protein extracts or cells can be directly lysed/ solubilized with APols in a single tube, and then all chemical and enzymatic processing of the proteome can be performed prior to MS in the same single tube. Because the compatibility of APols with trypsin digestion, there is no need for extra steps for detergent removal before protein digestion. As well, the lack of chaotropic agent in the protocol removed the need for desalting prior to mass spectrometry. Moreover, the APols protocol is also faster and less laborious than conventional detergent/chaotrope based protocols. We performed parallel processing (n = 3) of identical amounts of cells using our one-stop protocol and a conventional RIPA/urea protocol to compare sample loss and preparation time (Figure 4). For the classical in-solution strategy, RIPA buffer was used for cell lysis and protein extraction. Then, the cell lysate was precipitated using 6 times volume of cold acetone overnight. Protein pellet was washed twice by acetone and reconstituted in 8 M urea in 50 mM ABC. 1517

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519

Journal of Proteome Research

Technical Note

properties of APols. In particular, APols can lyse cells and extract whole proteome as efficiently as frequently used RIPA buffer. APols are amphiphilic and therefore theoretically have the ability to extract membrane proteins. In addition, because of their mild detergent properties, APols are fully compatible with trypsin digestion, even at concentration as high as 10 mg/mL. Therefore, it does not need to be removed prior to digestion. Following proteolytic digestion, APols can simply be removed by lowering the pH and centrifugation. The precipitation of APols does not significantly deplete hydrophobic peptides. We found that the “one-stop” strategy outperformed the conventional in-solution digestion.

■

ASSOCIATED CONTENT

* Supporting Information S

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: dfi[email protected]. Tel: 613-562-5800, ext 8674, Fax: 613-562-5655.

Figure 4. Comparison of conventional in-solution digestion workflow (left panel) and APols aided one-stop digestion strategy (right panel). The length of the color bar beside the axis represents the length of time for each step. The common steps are listed in the middle. The two steps colored in dark blue are absent in the APols aided one-stop workflow.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS D.F. acknowledges a Canada Research Chair in Proteomics and Systems Biology. Funding for this project was provided by NSERC-Canada. Z.N. acknowledges a postdoctoral scholarship from the CIHR Training Program in Neurodegenerative Lipidomics (TGF-96121).

SPE column was used to remove the urea after digestion. In contrast, in our one-stop protocol, the cells were lysed and protein extracted using the APols buffer. The sample was flash heated, cooled to room temperature, reduced, alkylated, and digested, all in a single tube. Following digestion, the APols were aggregated by lowering the pH to 3. Then, the APols were simply removed by benchtop centrifugation. The peptide samples derived from the classical in-solution strategy and our one-stop protocol were fractionated on SCX and then subjected to MS analysis. As shown in Table 1C, the results from these two strategies are significantly different even though the same amount of cells was used in both approaches. From three replicates, in total, around 20% more protein groups were identified using the one-stop strategy compared to the conventional in-solution digestion. At the peptide level, about 40% more unique peptides were obtained using our one-stop protocol. The total signal intensity from our one-stop protocol is about 30−40% higher than the RIPA-urea strategy; the difference obviously results from the sample loss of from protein precipitation and desalting step. Figure 1B shows the comparison between the APols extracted and RIPA extracted proteins after precipitation. The amount of protein recovered from the two protocols is clearly different based on the density of the two lanes. The characteristic of the peptides observed by mass spectrometry in terms of GRAVY index, misscleavage sites and peptide length are nearly identical between these two strategies (Figure S9).

■

REFERENCES

(1) Popot, J. L.; Berry, E. A.; Charvolin, D.; Creuzenet, C.; Ebel, C.; Engelman, D. M.; Flotenmeyer, M.; Giusti, F.; Gohon, Y.; Hong, Q.; Lakey, J. H.; Leonard, K.; Shuman, H. A.; Timmins, P.; Warschawski, D. E.; Zito, F.; Zoonens, M.; Pucci, B.; Tribet, C. Amphipols: polymeric surfactants for membrane biology research. Cell. Mol. Life Sci. 2003, 60 (8), 1559−74. (2) Zhou, J.; Zhou, T.; Cao, R.; Liu, Z.; Shen, J.; Chen, P.; Wang, X.; Liang, S. Evaluation of the application of sodium deoxycholate to proteomic analysis of rat hippocampal plasma membrane. J. Proteome Res. 2006, 5 (10), 2547−53. (3) Collins, K. D. Charge density-dependent strength of hydration and biological structure. Biophys. J. 1997, 72 (1), 65−76. (4) Yu, Y. Q.; Gilar, M.; Lee, P. J.; Bouvier, E. S.; Gebler, J. C. Enzyme-friendly, mass spectrometry-compatible surfactant for insolution enzymatic digestion of proteins. Anal. Chem. 2003, 75 (21), 6023−8. (5) Yu, Y. Q.; Gilar, M.; Gebler, J. C. A complete peptide mapping of membrane proteins: a novel surfactant aiding the enzymatic digestion of bacteriorhodopsin. Rapid Commun. Mass Spectrom. 2004, 18 (6), 711−5. (6) Kadiyala, C. S.; Tomechko, S. E.; Miyagi, M. Perfluorooctanoic acid for shotgun proteomics. PLoS One 2010, 5 (12), e15332. (7) Vieira, D. B.; Crowell, A. M.; Doucette, A. A. Perfluorooctanoic acid and ammonium perfluorooctanoate: volatile surfactants for proteome analysis? Rapid Commun. Mass Spectrom. 2012, 26 (5), 523−31. (8) Tribet, C.; Audebert, R.; Popot, J. L. Amphipols: Polymers that keep membrane proteins soluble in aqueous solutions. Proc. Natl. Acad. Sci. U.S.A. 1996, 93 (26), 15047−50.

■

CONCLUSIONS APols, although a mild detergent, are well suited for proteome processing. We developed using APols a “one-stop” procedure in which the whole proteome processing is simplified and performed in one tube. This is possible due to specific 1518

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519

Journal of Proteome Research

Technical Note

(9) Zoonens, M.; Catoire, L. J.; Giusti, F.; Popot, J. L. NMR study of a membrane protein in detergent-free aqueous solution. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (25), 8893−8. (10) Charvolin, D.; Perez, J. B.; Rouviere, F.; Giusti, F.; Bazzacco, P.; Abdine, A.; Rappaport, F.; Martinez, K. L.; Popot, J. L. The use of amphipols as universal molecular adapters to immobilize membrane proteins onto solid supports. Proc. Natl. Acad. Sci. U.S.A. 2009, 106 (2), 405−10. (11) Bechara, C.; Bolbach, G.; Bazzaco, P.; Sharma, K. S.; Durand, G.; Popot, J. L.; Zito, F.; Sagan, S. Maldi-tof mass spectrometry analysis of amphipol-trapped membrane proteins. Anal. Chem. 2012, 84 (14), 6128−35. (12) Ngoka, L. C. Sample prep for proteomics of breast cancer: Proteomics and gene ontology reveal dramatic differences in protein solubilization preferences of radioimmunoprecipitation assay and urea lysis buffers. Proteome Sci. 2008, 6, 30. (13) Peach, M.; Marsh, N.; Macphee, D. J. Protein solubilization: Attend to the choice of lysis buffer. Methods Mol. Biol. 2012, 869, 37− 47. (14) Schwert, G. W.; Takenaka, Y. A spectrophotometric determination of trypsin and chymotrypsin. Biochim. Biophys. Acta 1955, 16 (4), 570−5. (15) Brusca, J. S.; Radolf, J. D. Isolation of integral membrane proteins by phase partitioning with Triton X-114. Methods Enzymol. 1994, 228, 182−93. (16) Ishihama, Y.; Rappsilber, J.; Mann, M. Modular stop and go extraction tips with stacked disks for parallel and multidimensional peptide fractionation in proteomics. J. Proteome Res. 2006, 5 (4), 988− 94. (17) Wisniewski, J. R.; Zougman, A.; Nagaraj, N.; Mann, M. Universal sample preparation method for proteome analysis. Nat. Methods 2009, 6 (5), 359−62. (18) Olsen, J. V.; de Godoy, L. M.; Li, G.; Macek, B.; Mortensen, P.; Pesch, R.; Makarov, A.; Lange, O.; Horning, S.; Mann, M. Parts per million mass accuracy on an Orbitrap mass spectrometer via lock mass injection into a C-trap. Mol. Cell. Proteomics 2005, 4 (12), 2010−21. (19) Cox, J.; Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat. Biotechnol. 2008, 26 (12), 1367−72. (20) Perkins, D. N.; Pappin, D. J.; Creasy, D. M.; Cottrell, J. S. Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 1999, 20 (18), 3551−67. (21) Sheng, Q.; Dai, J.; Wu, Y.; Tang, H.; Zeng, R. BuildSummary: Using a group-based approach to improve the sensitivity of peptide/ protein identification in shotgun proteomics. J. Proteome Res. 2012, 11 (3), 1494−502. (22) Kyte, J.; Doolittle, R. F. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 1982, 157 (1), 105−32. (23) Champeil, P.; Menguy, T.; Tribet, C.; Popot, J. L.; le Maire, M. Interaction of amphipols with sarcoplasmic reticulum Ca2+-ATPase. J. Biol. Chem. 2000, 275 (25), 18623−37. (24) Giusti, F.; Popot, J. L.; Tribet, C. Well-defined critical association concentration and rapid adsorption at the air/water interface of a short amphiphilic polymer, amphipol a8−35: a study by forster resonance energy transfer and dynamic surface tension measurements. Langmuir 2012, 28 (28), 10372−80. (25) Picard, M.; Dahmane, T.; Garrigos, M.; Gauron, C.; Giusti, F.; le Maire, M.; Popot, J. L.; Champeil, P. Protective and inhibitory effects of various types of amphipols on the Ca2+-ATPase from sarcoplasmic reticulum: a comparative study. Biochemistry 2006, 45 (6), 1861−9.

1519

dx.doi.org/10.1021/pr301064z | J. Proteome Res. 2013, 12, 1512−1519