Cloud-Point Extraction and Delipidation of Porcine Brain Proteins in

Cloud-point extraction with the nonionic surfactant Triton X-114 was used to simultaneously extract hydrophobic and hydrophilic proteins from porcine ...
0 downloads 9 Views 3MB Size
Download PDF
Cloud-Point Extraction and Delipidation of Porcine Brain Proteins in Combination with Bottom-Up Mass Spectrometry Approaches for Proteome Analysis Ganna Shevchenko, Marcus O. D. Sjo ¨ din,† David Malmstro ¨ m,† Magnus Wetterhall, and Jonas Bergquist* Department of Physical and Analytical Chemistry, Analytical Chemistry, Uppsala University, Uppsala, Sweden Received February 6, 2010

In this study, temperature-induced phase fractionation also known as cloud-point extraction (CPE) with the nonionic surfactant Triton X-114 was used to simultaneously extract hydrophobic and hydrophilic proteins from porcine brain tissue. Various protein precipitation/delipidation procedures were investigated to efficiently remove lipids and detergents while retaining maximum protein recoveries. The best performing delipidation method was then used in combination with CPE to compare three different mass spectrometry (MS) based “bottom-up” proteomic approaches for protein analysis of the porcine brain. In the first approach, the intact proteins were initially separated by one-dimensional (1D) gel electrophoresis. The excised protein bands were digested with trypsin, and the peptides were separated by reversed phase nanoliquid chromatography (RP-nanoLC) followed by electrospray ionization (ESI) tandem mass spectrometry (MS/MS) analysis. The other bottom-up proteomic approaches were based on first enzymatical digestion of the proteins followed by RP-nanoLC separation in combination with matrix assisted laser desorption/ionization time-of-flight tandem mass spectrometry (MALDI-TOF/TOF MS) or on the combination of in-solution isoelectric focusing (IEF) with ESI-nanoLC-MS/MS of the IEF separated peptides. In total, we found and unambiguously identified 331 unique proteins. The overlap between different techniques was about 10%, showing that the use of multiple proteomic approaches is beneficial to yield a better coverage of the proteome. Furthermore, the overlap between the CPE extracted hydrophilic and hydrophobic proteins was rather small (9-16%), indicating an efficient sample preparation technique to extract and separate hydrophilic and hydrophobic proteins from brain tissue. The percentage of identified membrane proteins was 27%, which is in accordance to the fact that about one-third of all genes in various organisms encode for this class of proteins. The results indicate that cloud point extraction is a promising sample preparation tool, which allows simultaneous in depth studies of brain derived membrane proteins as well as hydrophilic proteins. This technique can be very useful when studying human central nervous system (CNS) tissue or animal models of neurological diseases. Keywords: cloud-point extraction (CPE) • delipidation • central nervous system (CNS) • brain • bottomup proteomics • membrane proteins (MPs) • mass spectrometry (MS)

1. Introduction Proteomic technologies are receiving increased attention in neuroscience research since the analysis of proteins expressed in different regions of the brain can provide vital information for the understanding of brain functions and the etiology of brain diseases.1-4 In recent years, rapid progress has been made for the analysis of soluble hydrophilic proteins originating from the central nervous system (CNS). However, membrane proteins (MPs) still remain an under-represented subset of the studied CNS proteins, despite their biological and biomedical * To whom correspondence should be addressed. Prof. Jonas Bergquist, Department of Physical and Analytical Chemistry, Analytical Chemistry, Uppsala University, P.O Box 599, SE-751 24 Uppsala, Sweden. E-mail: [email protected]. Fax: +46 18 4713692. † These authors contributed equally to this manuscript. 10.1021/pr100116k

 2010 American Chemical Society

importance. Membrane proteins carry out many essential cellular functions. They play a pivotal role in regulating cell-cell interactions, recognition, migration, adhesion, and signal transduction. Currently, more than 60% of all major pharmaceutical drug targets are membrane proteins.5 The importance of membrane proteins is highlighted by the fact that about onethird of all genes in various organisms encode for this class of proteins, emphasizing their crucial cellular role.6 Alteration of MPs can result into the dysfunction of the CNS.7-9 For instance, molecular genetic studies of patients suffering from genetic forms of early onset Alzheimer disease (AD) have identified three genes and their protein products as being involved in the etiology of AD. The three proteins are all integral membrane proteins. One of them, the beta-amyloid precursor protein Journal of Proteome Research 2010, 9, 3903–3911 3903 Published on Web 06/30/2010

research articles (APP), is the precursor of the beta-amyloid found in the characteristic neuritic plaques present in the brains of AD patients. The separation and structure analysis of membrane proteins remains a considerable challenge due to problems associated with their inherent amphipathic character and thus correspondingly poor solubility in aqueous solutions.10-14 This is a major issue given that the majority of the proteomic technologies are developed for hydrophilic proteins in waterbased solutions. Apart from the hydrophobic nature of MPs, they are also often low in abundance, which puts further demands on the analysis and detection used. Therefore, involving multiple fractionation and enrichment strategies are beneficial to increase the detection of the low abundant proteins. Several approaches have successfully been deployed to enrich MPs, in which hydrophobic proteins solubilized in either an organic acid, organic solvents, or an aqueous detergent solutions are subjected to digestion. For instance, 90% formic acid was used for the solubilization of membrane proteins which then were cleaved with cyanogen bromide.15 In another approach, membranes were solubilized with 60% methanol and the extracted MPs were digested with trypsin.16,17 Also, nonionic and ionic detergent solutions have been used for the solubilization of crude membranes, followed by digestion of MPs with endoproteinase Lys-C.18,19 We et al. proposed a digestion of nonsolubilized MPs using proteinase K in high salt buffer.20,21 These approaches have shown to be effective, but there is a drive to explore more generally applicable methods. Cloud-point extraction (CPE) is a detergent based enrichment method for hydrophobic analytes.22,23 Due to the low cost, easy removal of surfactants, and low toxicity, it has been proposedasaconvenientalternativetoconventionalliquid-liquid based extractions. The idea behind CPE is that nonionic polyoxyethylene detergents form clear micellar solutions in water at concentrations above the critical micellar concentration. Upon an increase in the temperature, the solution becomes turbid (reach the “cloud point”) followed by the separation into two distinct phases called surfactant-rich and surfactant-poor phases. The inverse relationship between temperature and solubility is due to a reversible dehydration of the polar ethylene oxide head groups of the detergent, which leads to the formation of a water insoluble phase that precipitates out of solution.22,23 An important advantage of this method is that MPs can be easily separated and enriched from hydrophilic peripheral proteins in biological matrices. In this paper, we report the use of CPE fractionation to simultaneously extract and enrich both the membrane and hydrophilic proteins from porcine brain. Five different delipidation methods were evaluated regarding protein yield and reproducibility. The best performing delipidation protocol was then used in combination with CPE to compare three mass spectrometry-based “bottom-up” proteomic approaches for protein separation and identification. In the first approach, intact proteins in both the hydrophilic and hydrophobic CPE fractions were separated using one-dimensional gel electrophoresis, followed by in-gel digestion in combination with nanoliquid chromatography (LC) separation and MS detection. In the other approaches, the proteins in the CPE fractions were first enzymatically digested and then analyzed by nano LC-MS/ MS or a combination of in-solution isoelectric focusing (IEF) followed by nanoLC-MS/MS of the generated peptides. 3904

Journal of Proteome Research • Vol. 9, No. 8, 2010

Shevchenko et al.

2. Experimental Section 2.1. Chemicals and Reagents. Acetonitrile (ACN), methanol, (MeOH), acetic acid (HAc), formic acid (FA), ammonium bicarbonate (NH4HCO3), tri-n-butylphosphate (TBP), and sodium chloride (NaCl) were obtained from Merck (Darmstadt, Germany). Triton X-114, acetone, ethylenediaminetetraacetic acid tetrasodium salt dihydrate (EDTA), protease inhibitor cocktail, phosphate buffered saline (PBS), Tris-HCl, formic acid, and trifluoroacetic acid (TFA) were purchased from Sigma Aldrich (St. Louis, MO). For tryptic digestion, iodoacetamide (IAA), urea, and dithiothreitol (DTT) were obtained from Sigma Aldrich and trypsin (sequencing grade from bovine pancreas 1418475; Roche diagnostic, Basel, Switzerland) was used. Ultrapure water was prepared by Milli-Q water purification system (Millipore, Bedford, MA). Other organic solvents and reagents were of the highest purity commercially available. 2.2. Brain Samples. Porcine brains were obtained from the local slaughterhouse, and collected on ice directly at dissection. The frontal cortex parts of the brains were dissected out and homogenized in liquid nitrogen. The brain powder was stored at -80 °C prior to analyses. 2.3. Cloud Point Extraction of Membrane and Hydrophilic Proteins with Triton X-114. Commercially available Triton X-114 preparations are not completely homogeneous. Therefore, with the aim of elimination of the more hydrophilic species, a precondensation procedure of Triton X-114 was performed according to the method of Bordier.23 The brain powder (100 mg) was homogenized for 60 s in a blender (POLYTRON PT 1200, Kinematica) with 1 mL of Triton lysis buffer (10 mM Tris-HCl pH 7.4, 0.15 M NaCl, 1 mM EDTA and PBS containing 1% (v/v) Triton X-114). To prevent any protein degradation, 10 µL of Protease Inhibitor Cocktail was added during the sample preparation. Then, the sample was incubated for 1 h at 4 °C with mild agitation. The cell lysate was clarified by centrifugation for 30 min (10 000× g at 4 °C). The clear supernatant was then transferred directly onto 100 µL of sucrose cushion buffer and the tube was incubated at 37 °C for 5 min. Clouding of the solution occurred. The tube was centrifuged for 3 min (400× g at 37 °C). Two phases formed, aqueous on the top and detergent at the bottom. The aqueous phase was transferred to a fresh tube and incubated on ice. The detergent phase was resuspended in 500 µL of cold PBS and phase separation was repeated again. The second detergent depleted aqueous phase was pooled with the first and kept on ice. The detergent-rich fraction, which contained hydrophobic membrane proteins, was resuspended in 1.5 mL of cold PBS. The pool of detergent-depleted aqueous phase was re-extracted by adding 50 µL of 11.4% Triton X-114 stock solution, incubated in a 37 °C water bath for 3 min, and centrifuged for 3 min (400× g at 37 °C). The aqueous phase contained hydrophilic watersoluble proteins. 2.4. Delipidation and Protein Precipitation. Numerous delipidation and protein precipitation procedures have been reported.24-26 A few of these were evaluated to find an optimal delipidation and detergent removal method that also yields high recoveries and good repeatability both for the detergentrich and the detergent-depleted phases. The best performing delipidation protocol was then exclusively used to delipidate the CPE fractions for the comparison of three different mass spectrometry-based “bottom-up” proteome analysis methods. 2.4.1. Protein Precipitation Using a Centrifugal Filter Device (Spin Filter). For the washing of centrifugal filter device

CPE and Delipidation of Porcine Brain Proteins 300 µL of MQ water was placed into a Microcon (3 kDa membrane) sample reservoir (Millipore, Bedford, MA), which was inserted into collection vial and spun for 20 min (14 000× g at room temperature). The water was removed from the collection vial. Aliquots (100 µL) of the detergent-depleted aqueous and detergent-rich phases were placed in 1.5 mL Eppendorf Protein LoBind Tubes (Eppendorf AG, Hamburg, Germany) and mixed with 100 µL of 100% ACN and 300 µL of MQ water. Washed Microcon sample reservoirs were inserted into collection vials and sample mixtures were loaded onto the sample reservoirs. The Microcons were spun for 1 h 40 min (14 000× g at room temperature). Afterward, 500 µL of 20% ACN was added to the sample reservoirs and Microcons were spun again for 1 h 40 min (14 000× g at room temperature). Sample reservoirs were placed upside down in new vials and spun for 3 min at 1000× g to transfer protein concentrate to a vial. Then the protein pellets were air-dried. 2.4.2. Delipidation and Protein Precipitation by Acetone. Aliquots (100 µL) of the detergent-depleted aqueous and detergent-rich phases were mixed with 500 µL of ice-cold acetone and incubated at -20 °C overnight. The precipitate was pelleted by centrifugation for 30 min (15 000× g at 4 °C), washed twice with 500 µL of ice-cold 50% acetone, and centrifuged for 30 min (15 000× g at 4 °C). The supernatant was removed and the protein pellets were air-dried according to Tantipaiboonwong et al.24 2.4.3. Delipidation and Protein Precipitation by Chloroform/ Methanol/Water. Aliquots (100 µL) of the detergent-depleted aqueous and detergent-rich phases were mixed with 400 µL of methanol, vortexed, and centrifuged for 30 s at 9000× g. Then chloroform (200 µL) was added and the samples were vortexed and centrifuged again (30 s at 9000× g). For phase separation, 300 µL of water was added, and the samples were vortexed vigorously and centrifuged for 1 min at 9000× g. The upper water phase was carefully removed and discarded. A further volume of 300 µL methanol was added to the rest of the lower chloroform phase and the interphase with the precipitated protein. The samples were mixed and centrifuged again for 2 min at 9000× g to pellet the protein. The supernatant was removed and the protein pellets were air-dried according to Wessel et al.25 2.4.4. Delipidation and Protein Precipitation by Acetone/ Methanol (8:1). Aliquots (100 µL) of the detergent-depleted aqueous and detergent-rich phases were mixed with 1.4 mL of ice-cold acetone/methanol (8:1) mixture and incubated at 4 °C for 90 min. The precipitate was pelleted by centrifugation for 15 min (2800× g at 4 °C), washed sequentially with 1 mL of acetone followed by methanol, and then air-dried according to Mastro et al.26 2.4.5. Delipidation and Protein Precipitation by Tri-nbutylphosphate/Acetone/Methanol (1:12:1). Aliquots (100 µL) of the detergent-depleted aqueous and detergent-rich phases were mixed with 1.4 mL of ice-cold tri-n-butylphosphate/ acetone/methanol mixture (1:12:1) and incubated at 4 °C for 90 min. The precipitate was pelleted by centrifugation for 15 min (2800× g at 4 °C), washed sequentially with 1 mL of TBP, acetone and methanol, and then air-dried according to Mastro et al.26 2.5. One-Dimensional Gel Electrophoresis. The performance and repeatability of the different protocols were visually evaluated by 1D gel electrophoresis. The protein pellets were redissolved in 25 µL of XT Sample Loading Buffer (BioRad Laboratories) and 55 µL of Milli-Q water. A volume of 10 µL of

research articles 45 mM DTT was added and the samples were heated at 95 °C for 5 min to reduce the disulfide bonds. The samples were cooled to ambient temperature and 10 µL of 100 mM IAA was added and the mixtures were incubated for 15 min in darkness at room temperature to carabamidomethylate the cysteines. A volume of 30 µL of sample was loaded into each lane of a 18well, 4-12% Bis-Tris Criterion XT Precast Gel (BioRad Laboratories). The separation was performed with MES-buffer at 200 V constant for 45 min. The proteins were fixed for 1 h in 40% methanol containing 10% acetic acid and stained with colloidal Coomassie blue R-250 (Bio-Rad) for 2 h. Finally the gels were destained 40% methanol containing 10% acetic acid and scanned with a HP scanner (HP Scanjet G3010 Photo Scanner). 2.6. Protein Quantification. The total protein content of delipidated proteins according to section 2.4.5 was determined using the DC Protein Assay Kit (BioRad Laboratories, Hercules, CA), which is based on the modified Lowry method with bovine serum albumin as standard.27 The DC assay was carried out according to the manufacturer’s instructions in 96-well microtiter plates. The samples with hydrophobic and hydrophilic protein pellets (after the delipidation of 100 µL of the detergentdepleted aqueous and detergent-rich phases according to 2.4.5) were redissolved in 100 µL of 2 M urea and 6% SDS, respectively. Five microliters of samples (triplicates) and protein standards were pipetted into separate microtiter wells. Twentyfive microliters of reagent A (alkaline copper tartrate solution) and 200 µL of reagent B (dilute Folin Reagent) were added to each well. The samples were mixed on a horizontal shaker. After 15 min incubation, the absorbance was measured at 750 nm using a microtiter plate reader model 680 (BioRad Laboratories). The concentrations of proteins were 1.22 ( 0.06 mg/mL for the detergent-rich phase and 1.66 ( 0.06 mg/mL for the detergent-depleted aqueous phase. 2.7. Separation of Intact Proteins Followed by Digestion, Separation, and MS/MS Analysis. 2.7.1. One-Dimensional Gel Electrophoresis. Aliquots (100 µL) of the detergent-depleted aqueous and detergent-rich phases were delipidated by the protocol described in section 2.4.5. The protein pellets were redissolved and separated by 1D gel electrophoresis according to section 2.5. 2.7.2. In-Gel Tryptic Digestion of Proteins. For in-gel digestion, selected bands from the 1D gels were cut and placed into 1.5 mL Eppendorf Protein LoBind Tubes (Eppendorf AG, Hamburg, Germany). A volume of 150 µL of 50:50 ACN/100 mM NH4HCO3 destain solution was added to the gel pieces and the samples were agitation for 20 min at room temperature and the liquid was then discarded. This step was repeated 2 times until most Coomassie stain was removed. Afterward, 150 µL 100% ACN was added and the samples were incubated for 5 min and the liquid was then discarded. The gel pieces were then dried in 50 C for 1 h. Protein digestion was performed by redissolving 25 µg trypsin in 1 mL 50 mM NH4HCO3 and adding 8 µL of the trypsin solution to each tube and incubating the samples at +8 C for 1 h. Then, 30 µL of 50 mM NH4HCO3 was added and the samples were digested at 37 °C overnight in darkness. After digestion, the samples were sonicated for 5 min and the liquid was collected into new 0.5 mL Eppendorf low protein binding tubes. The digests were dried down under vacuum using a Speedvac system ISS110 (Thermo Scientific, Waltham, MA). The samples were redissolved in 20 µL 2.5% acetic acid and desalted on ZipTip C18 columns (Millipore) according to a procedure described in detail elsewhere.28 Briefly, the tip was first wetted in 5 × 10 µL 100% ACN and Journal of Proteome Research • Vol. 9, No. 8, 2010 3905

research articles equilibrated with 5 × 10 µL 1% acetic acid. The tryptic peptides were adsorbed to the media using 30 repeated cycles of sample loading. The tip was washed using 5× 10 µL of 1% acetic acid and finally the peptides were eluted in 2 × 10 µL 50% ACN, 1% acetic acid. This procedure was repeated twice. The desalted tryptic peptide eluates were dried down under vacuum and then redissolved in 10 µL of 0.1% (v/v) trifluoroacetic acid prior to LC-ESI-MS/MS analysis. 2.7.3. NanoLC-ESI-MS/MS Analysis. The separation was performed using a Tempo nanoLC-1D plus (Applied Biosystems, USA) system. A volume of 3 µL of each sample was injected into a 10 µL loop in injection pickup mode. The samples were loaded onto a Peptide CapTrap (Michrom Bioresources, CA) for 5 min at 15 µL/min isocratically (solvent 1A, H2O/ACN/FA 97.9/2/0.1 v/v %). The peptides were then eluted onto a BioBasic-C18 column 5 µm 15 cm × 75 µm (Thermo) using a stepwise gradient (solvent 2A, H2O/ACN/FA 97.9/2/0.1 v/v %; solvent 2B, H2O/ACN/FA 2/97.9/0.1 v/v %) at a flow rate of 350 nL/min. The gradient was as follows: 2% B during 20 min, 2% f 8% B in 5 min, 8% f 32% B in 86 min, 32% f 40% B in 5 min and finally 40% f 80% B in 1 min. After washing for 4 min, the column was equilibrated for the next sample by going from 80% f 2% B in 5 min and kept at 2% B during 14 min giving a total run time of 140 min. The samples were kept in the autosampler at 10 °C. Data was collected in positive ESI mode on a 3200 QTrap hybrid triple quadrupole/linear ion trap (Applied Biosystems/MDS Sciex, Canada) equipped with a MicroIonSpray II head with an uncoated 10 µm i.d. PicoTip (New Objective, MA, USA) as electrode. The source parameters were: ion spray voltage, 2500 V; curtain gas, 20 psi; gas 1, 20 psi; interface heater temperature, 150 °C. Spectra were acquired in automated mode using Information Dependent Acquisition (IDA). Precursors were selected by using an Enhanced Mass Scan (EMS) as a survey scan (fill time 25 ms, m/z 400-1400, 4000 amu/sec). On the five most intense precursor ions, an Enhanced Resolution scan (ER) was employed for charge state determination (dynamic fill time, m/z range 30 amu, 250 amu/sec) as well as an Enhanced Product Ion scan (EPI) for MS/MS data (fill time 200 ms, m/z 100-1400, 4000 amu/sec). The precursor ions were fragmented by collision induced dissociation (CID) in the Q2 collision cell. Collision voltages were automatically adjusted based on the ion charge state and mass using rolling collision energy. Generated fragments ions were captured and their masses analyzed in the Q3 linear ion trap. The IDA data acquired was processed in Analyst 1.4.2 and subjected for automatic protein identification using Mascot Daemon (version 2.2.2, Matrix Science, U.K.) against the NCBI pig (sus scrofa) fasta database. The search parameters were set to Enzyme: Trypsin, Fixed modifications: Carbamidomethyl (C), Variable modifications: Oxidation (M), Peptide tolerance: 1.5 Da, MS/ MS tolerance: 0.6 Da and maximum 1 missed cleavage site. Proteins were considered to be positively matched using the more stringent MudPIT scoring criteria (p e 0.05) in the MASCOT search engine.29 2.8. In-Solution Digestion Followed by Separation and MS/MS Analysis. 2.8.1. In-Solution Tryptic Digestion of Proteins. Aliquots (100 µL) of the detergent-depleted aqueous and detergent-rich phases were delipidated by the protocol described in section 2.4.5. For in-solution digestion, the samples from the hydrophilic CPE protein pellets were redissolved in 100 µL of digestion buffer consisting of 8 M urea and 400 mM NH4HCO3. The samples from the hydrophobic CPE 3906

Journal of Proteome Research • Vol. 9, No. 8, 2010

Shevchenko et al. protein pellets were redissolved in 100 µL of digestion buffer consisting of 50% ACN and 50 mM NH4HCO3. A volume of 10 µL of 45 mM DTT was added to all samples and the mixtures were incubated at 50 °C for 15 min to reduce the disulfide bridges. The samples were cooled to room temperature and 10 µL of 100 mM IAA was added and the mixtures were incubated for an additional 15 min at room temperature in darkness to carabamidomethylate the cysteines. Finally, trypsin solution was added to the samples to yield a final trypsin/ protein concentration of 5% (w/w). The tryptic digestion was performed at 37 °C overnight in darkness. A small volume of 30 µL of each sample was desalted on ZipTip C18 columns (Millipore) as described in section 2.7.1. However, before desalting the hydrophobic CPE fraction, the sample was dried down under vacuum to remove the acetonitrile and then redissolved in 30 µL 2.5% HAc. After desalting, the eluates were vacuum centrifuged to dryness and redissolved in 20 µL of 0.1% (v/v) trifluoroacetic acid prior to LC-MALDI-TOF/TOF MS analysis. 2.8.2. NanoLC-MALDI-TOF/TOF MS Analysis. The reversed phase liquid chromatography separation was performed with a 1100 nanoflow LC system (Agilent Technologies, Waldbronn, Germany), equipped with a fraction collector for direct fractionations onto a MALDI target plate. A volume of 10 µL digestion products was injected into a 10 µL sample loop. For separating the peptides, a 15 cm × 180 µm, C18 column (Thermo) with 5 µm particle size and an H2O/ACN/TFA solvent system (H2O, 0.1% TFA mobile phase [A]; ACN, 0.1% TFA mobile phase [B]) was used. A flow rate of 2 µL/min was applied, starting with isocratic elution at 2% B during 20 min, followed by gradient elution from 2 to 8% B during 5 min, then from 8 to 32% B within 86 min, then from 32 to 40% B during 5 min, and finally from 40 to 80% B during 1 min. The online fractionation onto an MALDI target was performed with four fractions per minute for 96 min within the elution period from 20 min (2% B) and 116 min (40% B) resulting in 384 fractions. For optimal MS results, disposable prespotted anchorchip targets (PAC-targets, Bruker Daltonics, Bremen, Germany) were chosen. The targets were washed with 10 mM NH4H2PO4/0.1% TFA prior to MALDI-TOF/TOF MS analysis. Mass data were acquired with an Ultraflex II MALDI-TOF/TOF MS (Bruker Daltonics) in reflector positive mode. A mass range of 700-4000 Da was analyzed with a sum of 300 shots/spot and 50 shots/ position, respectively, in a hexagonal pattern. The laser frequency was set to 100 Hz. MALDI-TOF/TOF tandem MS analysis was performed in LIFT mode with 30% increased laser energy to give the fragmentation spectra. Post LIFT mother ion suppression was applied to deflect the precursor and elevate fragment ion intensity. Peptide monoisotopic signals were analyzed using the SNAP algorithm implemented in the FlexAnalysis software (Bruker Daltonics). The spectra were calibrated externally using the prespotted calibrants adjacent to the sample spots. For final protein identification, all collected MS/MS data were run in a comprehensive MS/MS ion search using the Mascot search engine version 2.2.2. Acquired MS/ MS spectra were evaluated with the Matrix Science MASCOT database against the NCBI pig (sus scrofa) fasta database. The search parameters were set to Enzyme, Trypsin; Fixed modifications, Carbamidomethyl (C); Variable modifications, Oxidation (M); Peptide mass tolerance, ( 50 ppm; Fragment mass tolerance, ( 0.8 Da; and maximum 1 missed cleavage site. Proteins were considered to be positively matched if at least

CPE and Delipidation of Porcine Brain Proteins one peptide MS/MS spectrum passed the MudPIT scoring criteria (p e 0.05) in the MASCOT search engine.29 2.8.3. IEF Fractionation of Tryptic Peptides. Aliquots (500 µL) of the detergent-depleted aqueous and detergent-rich phases were delipidated by the protocol described in section 2.4.5, yielding 830 µg of hydrophilic proteins and 610 µg of hydrophobic proteins, respectively. The obtain pellets digested in-solution according to the previously described section 2.8.1. The digests from both CPE fractions were diluted up to 1 mL in 2.5% HAc and then desalted on a Isolute C18(EC) (1 mL, 50 mg capacity, Biotage, Uppsala, Sweden) SPE column using the following schedule: The column was first wetted in 300 µL of 100% acetonitrile and equilibrated with 5 × 1 mL 1% acetic acid. The tryptic peptides were adsorbed to the media using 5 repeated cycles of sample loading. The column was washed using 5 × 1 mL of 1% acetic acid and finally the peptides were eluted in 250 µL 50% acetonitrile, 1% acetic acid. After desalting, the eluate was vacuum centrifuged to dryness. The tryptic peptides were separated based on in-solution isoelectric focusing using a 3100 OFFGEL Fractionator system (Agilent Technologies). The IEF was conducted with the OFFGEL pH 3-10 12-well setup kit according to the manufacturer’s protocol. The desalted samples were reconstituted with 1.8 mL of rehydration solution and 150 µL sample was loaded into each well. The IEF was conducted at maximum 50 µA/strip until 20 kVh was reached (approximately 12 h). After the isoelectric focusing, the solution was removed from each well and dried down and redissolved in 2.5% acetic acid. Each fraction with the focused peptides were desalted on ZipTip C18 columns (Millipore) as described earlier and the eluates were vacuum-dried followed by resuspension in 10 µL of 0.1% (v/v) trifluoroacetic acid prior to LC-ESI-MS/MS analysis as described in section 2.7.3. 2.9. Data Analysis. The theoretical isoelectric point (pI) was defined by the algorithm used in the Compute MW/pI program in Expasy (www.expasy.ch). Molecular weight (MW) values of proteins were computed through an average amino acid weight table. The grand average hydrophobicity (GRAVY) values were calculated using the ProtParam software (available at http:// expasy.org/cgi-bin/protparam). Proteins with positive GRAVY values are considered to be hydrophobic, and those with negative values, hydrophilic. The subcellular location and function of the identified proteins were elucidated by collection of information from the Uniprot database and pie charts for these classifications of the identified porcine brain proteins were created. For the creation of Venn diagrams, the free share program Venn Diagram Plotter was used (http://omics.pnl.gov/ software/VennDiagramPlotter.php).

3. Results and Discussion 3.1. Evaluation of Delipidation and Protein Precipitation Methods. An important sample preparation step concern specific to MPs is the removal of lipids that are always present in biological extracts. Numerous proteins are complexed with lipids. This interaction reduces their solubility, alters protein migration during electrophoresis, or interferes with enzymatic digestion and might affect the pI and MW.30 The most common methods of protein enrichment and purification rely on selective precipitation of proteins using different organic solvents such as cold acetone either alone24 or in combination with other organics,26 methanol/chloroform,25 resulting in the selective solubilization of the lipid component. Other alternatives are the use of a number of commercially available affinity precipitation kits24 or centrifugal filter device (spin filter) for

research articles

Figure 1. 1D gel electrophoresis of CPE extracted proteins of porcine brain. (A) Lanes 1 and 2, hydrophobic and hydrophilic proteins after spin-filtration, respectively; lanes 3, 7, and 11, Triton X-114 solubilized proteins before phase separation after spin-filtration, after delipidation with chloroform/methanol/water and after delipidation with TBP-acetone-methanol (1:12:1) respectively; lanes 4, 8, and 12 - proteins in detergent-free lysis buffer after spin-filtration, after delipidation with chloroform/ methanol/water, and after delipidation with TBP-acetone-methanol (1:12:1), respectively; lanes 5 and 6, hydrophobic and hydrophilic proteins after delipidation with chloroform/methanol/ water, respectively; lanes 9 and 10, hydrophobic and hydrophilic proteins after delipidation with TBP-acetone-methanol (1:12:1), respectively. (B) Lanes 1 and 3, hydrophobic proteins and lanes 2 and 4, hydrophilic proteins after delipidation with acetone/ methanol (8:1), respectively; lanes 5 and 6, hydrophobic and hydrophilic proteins after delipidation with TBP-acetone-methanol (1:12:1), respectively; lanes 7 and 8, Triton X-114 solubilized proteins before phase separation and proteins in detergent-free lysis buffer after delipidation with TBP/acetone/methanol (1:12: 1), respectively; lanes 9 and 11, hydrophobic proteins; lanes 10 and 12, hydrophilic proteins after delipidation with the acetone, respectively.

the removal of lipids and detergents. Therefore, five precipitation methods using different organic solvents (pure acetone; acetone/methanol (8:1); tri-n-butylphosphate/acetone/methanol (1:12:1); chloroform/methanol/water) and the use of spin filter to remove lipids and detergents were compared. The resulting 1D gel electrophoresis images of brain proteins samples prepared by different preparation methods after the CPE fractionation are demonstrated in Figures 1 and 2. The gel images of both the hydrophilic and membrane/hydrophobic Journal of Proteome Research • Vol. 9, No. 8, 2010 3907

research articles

Shevchenko et al.

Figure 2. 1D gel electrophoresis of CPE extracted proteins of porcine brain. Lanes 1, 3, hydrophobic proteins and lane 2, hydrophilic proteins after spin-filtration respectively; lanes 4, 6, and 8, hydrophilic proteins and lanes 5, 7, and 9, hydrophobic proteins after delipidation with TBP/acetone/methanol (1:12:1), respectively.

brain proteins prepared by the treatment with tri-n-butylphosphate/acetone/methanol (1:12:1) (section 2.4.5) showed the best resolution and the largest number of protein bands. Furthermore, the gels demonstrate that hydrophilic and hydrophobic proteins can readily and efficiently be separated from each other under mild extraction conditions with good repeatability. Therefore, the protein purification and delipidation method described in section 2.4.5 using tri-n-butylphosphate/acetone/methanol (1:12:1) was extensively used throughout the further experiments in this study. 3.2. 1D Gel Protein Separation Followed by Digestion, Separation, and MS/MS Analysis. The first approach, based on 1D gel electrophoresis separation of the intact proteins followed by in-gel digestion in conjunction with nano-LC-MS/ MS, yielded an overall identification of 280 unique proteins. The identified proteins are listed in Supplementary Table 1 (Supporting Information). In the detergent-rich fraction, 145 proteins could be identified and 181 proteins could be in the detergent-depleted aqueous phase. The two fractions overlapped with 46 proteins (16%), which is illustrated in Figure 3A. The rather low overlap between the two fractions indicates a good sample preparation technique in terms of separating between hydrophilic and hydrophobic proteins extracted from the CNS tissue. The obtained results accentuate the need to study both the hydrophilic and hydrophobic phases in order to cover a wider range of the proteome. This is especially important when any screening studies of tissue is planned to find potential biomarkers for a certain disease. 3.3. In-Solution Digestion Followed by Separation and MS/MS Analysis. The nanoLC-MALDI-TOF/TOF MS bottom up approach gave 98 unique proteins as shown in Table 1. The hydrophobic fraction contained 53 unique proteins, the hydrophilic fraction 55 proteins and 10 proteins overlapped (10%), Figure 3B. The IEF-LC-ESI-MS/MS bottom up approach gave 116 unique proteins with 42 proteins in the detergent-rich phase, 84 proteins in the detergent-depleted phase and an overlap of 10 proteins (9%), see Figure 3C. It is not so surprising that the nanoLC-MALDI-TOF/TOF MS bottom up approach gave fewer matched proteins considering that no additional 3908

Journal of Proteome Research • Vol. 9, No. 8, 2010

Figure 3. Overlap between CPE extracted hydrophobic and hydrophilic proteins from porcine brain identified by the three different proteomic methods: (A) 1D PAGE-LC-ESI-MS/MS, (B) nanoLC-MALDI-TOF/TOF MS, and (C) IEF-LC-ESI-MS/MS.

prefractionation step was utilized, as in the case of the 1D gel electrophoresis or in-solution IEF. However, the results from the in-solution IEF did not yield a considerable larger number of identified proteins. The detergent-rich in-solution IEF experiments gave a somewhat discouraging number of proteins. One reason for this could be difficulties with dissolving the hydrophobic proteins prior to digestion, insufficient sample loading of the tryptic peptides on the IEF strip or lower distribution of the peptides into solution after IEF. Additionally, one could argue about the necessity of performing CPE prior to the bottom up approach. There are successful studies reported on CNS tissue, where the sample has simply been digested without any prefractionation.20,21 However, it is not straightforward to find conditions suitable for all proteins in the sample and it is therefore a risk that the outcome might be biased. There is no common “golden standard” for protein extraction from CNS tissue. However, the utilized cloud point extraction/prefractionation yields good possibilities to simultaneously extract and study both membrane/hydrophobic and hydrophilic proteins from fatty tissue, such as brain. 3.4. Data Analysis of the Combined Results. The analysis of porcine brain revealed 331 unique proteins (Table 1). Among all 331 identified proteins 19 (5.7%) were identified by only 1 peptide, while 312 proteins (94.3%) were identified with 2 or more peptides. From the 312 proteins identified by more than 1 peptide, 58 proteins (17.5%) were identified by 2 peptides,

CPE and Delipidation of Porcine Brain Proteins

Figure 4. Overlap between CPE extracted (A) hydrophobic and (B) hydrophilic proteins identified by three different proteomic methods: nanoLC-MALDI-TOF/TOF MS, IEF-LC-ESI-MS/MS, and 1D PAGE-LC-ESI-MS/MS.

while 56 (16.9%) were identified by 3 or 4 peptides and another 198 (59.9%) proteins were identified by 5 or more peptides. Interestingly, the overlap between the different approaches was rather low for both CPE fractions, as shown in Figure 4A and B. The total number of identified hydrophobic proteins was 185. Of those, 107, 22, and 14 were exclusively identified in the 1D gel approach, nanoLC-MALDI-TOF/TOFMS and IEFESI-MS/MS methods respectively (Figure 4A). Only 14 hydrophobic proteins (9%) were identified in all three preparations. For the hydrophilic fraction, 205 proteins were identified in total, with 89 proteins exclusively found in the 1D gel approach, 14 using nanoLC-MALDI TOF/TOFMS and 8 using IEFESI-MS/MS method (Figure 4B). The overlap for all three methods was 21 (10%) of the proteins. The obtained results emphasize the need to use multiple approaches in order to better cover the proteome. The identified porcine brain proteins (Supplementary Table 1, Supporting Information) were classified according to their main biological function as well as their subcellular localization. The most abundant class of the proteins (32%) had enzyme activity; 25 and 14% of proteins were transport and structural proteins and 9% were involved with signal transduction, as shown in Figure 5A. The assignment of subcellular localization of the identified proteins revealed that 30% of them are predicted to be located in the cytoplasm and a considerable fraction (27%) was annotated as membrane proteins (Figure 5B). The obtained percentage of identified membrane proteins are close to the expected 1/36 and demonstrates the importance of using combined ap-

research articles

Figure 5. Classifications according to (A) biological function and (B) subcellular localization of CPE extracted porcine brain proteins as stated by Uniprot.

proaches to yield a representative picture of the proteome. The percentage of proteins annotated as mitochondrial proteins was also high (16%). Nuclear (3%) and extracellular (3%) proteins that could be involved in gene expression control and cellular communication, respectively, were also represented in the sample. Of the remainder, with a subcellular annotation, 6% were annotated as cytoskeleton, 6% as secretory vesicle, 3% as endoplasmatic reticulum and 2% as Golgi apparatus. To assess any possible analytical bias in the protein identification, the molecular weight (MW) and isoelectric point (pI) of each protein was calculated based on its primary amino acid sequence after excluding signal and propeptide chains. The 331 porcine brain proteins identified have a wide range of pI (4.25-11.28) as well as large molecular weight distribution (3.5-300.3 kDa). The level of hydrophobicity of the identified proteins was analyzed on the basis of the calculated average GRAVY values. The GRAVY values of the proteins ranged from -1.311 to +0.724. In total, 44 (13.3%) of these proteins had positive GRAVY values, indicating a considerable number of highly hydrophobic proteins were detected in the study. The weighted GRAVY index (GRAVYw) was calculated for the hydrophobic and hydrophilic CPE fractions according to eq 1. Protein score values is an approximation of the protein quantity and therefore the weighted values of these were used in the equation. Journal of Proteome Research • Vol. 9, No. 8, 2010 3909

research articles

Shevchenko et al.

GRAVYW )

References

GRAVY1 · Score1 + GRAVY2 · Score2 + ... + GRAVYn · Scoren Score1 + Score2 + ... + Scoren (1) The GRAVY index was -0.19 ( 0.02 for the hydrophobic fraction and -0.31 ( 0.03 for the hydrophilic CPE fraction. The obtained results emphasize that the hydrophobic fraction contains more hydrophobic proteins than the hydrophilic fraction.

Conclusions The brain is considered to consist of at least 50% lipid membranes. In these membranes, including highly specialized intracellular compartments, proteins controlling the cellular functions are situated. These proteins play important roles in various cellular processes, such as signal transduction, ion transport, metabolism, etc. There are today a number of methodologies developed for enrichment of membrane bound proteins, but all of them are more or less selective in their nature. In this study, we have developed a more versatile method for the simultaneous extraction and enrichment of both MPs and hydrophilic proteins by using cloud-point extraction. The outcome of the CPE was evaluated with a complementary set of proteomic analysis methodologies. These approaches for protein analysis were used to study the porcine brain proteome. The combined results led to the significant identification of 331 unique proteins from porcine brain tissue. Of these, 27% were integral membrane or membrane-associated proteins with relevant biological functions. The low protein overlap between the CPE extracted hydrophilic and hydrophobic fractions (9-16%) indicates that CPE is an efficient method for extracting and disintegrating brain derived proteins based on their hydrophobicity. The small overlap (10%) between three different techniques shows that it is very beneficial to combine multiple approaches to yield a better coverage of the proteome. The sample preparation method presented in this study may have broad applications for comprehensive proteomic analyses of animal models of human brain and CNS diseases or for the direct study of clinical human CNS tissue.

Acknowledgment. This research was supported by Uppsala Berzelii Technology Centre for Neurodiagnostics, with financing from the Swedish Governmental Agency for Innovation Systems and the Swedish Research Council P29797-1 JB grants (621-2005-5379, 621-2008-3562). The Swedish Institute is also gratefully acknowledged for financial support. Supporting Information Available: Table 1. List of identified proteins in porcine brain tissue using the three different proteomic methods: nanoLC-MALDI-TOF/TOF MS, IEF-LC-ESI-MS/MS and 1D PAGE-LC-ESI-MS/MS. MT, microtubule; C, cytoplasm; M, mitochondria; ER, endoplasmatic reticulum; S, secretory vesicle; GA, Golgi apparatus; N, nucleus; IMP, integral membrane protein; HC, hemoglobin complex; GPC, G-protein complex; IFC, intracellular ferritin complex; GRAVY, grand average of hydrophobicity. This material is available free of charge via the Internet at http:// pubs.acs.org. 3910

Journal of Proteome Research • Vol. 9, No. 8, 2010

(1) Zetterberg, H.; Ru ¨etschi, U.; Portelius, E.; Brinkmalm, G.; Andreasson, U.; Blennow, K.; Brinkmalm, A. Clinical proteomics in neurodegenerative disorders. Acta Neurol. Scand. 2008, 118, 1– 11. (2) Drabic, A.; Bierczynka-Krzysik, A.; Bodzon-Kulakowska, A.; Suder, P.; Kotlinska, J.; Silberring, J. Proteomics in neurosciences. Mass Spectrom. Rev. 2007, 26, 432–450. (3) Olsen, J. V.; Nielsen, P.; Andersen, J. R.; Mann, M.; Wis´niewski, J. R. Quantitative proteomic profiling of membrane proteins from the mouse brain cortex, hippocampus, and cerebellum using the HysTag reagent: mapping of neurotransmitter receptors and ion channels. Brain Res. 2007, 1134, 95–106. (4) Morrison, R.; Kinoshita, Y.; Johnson, M. D.; Uo, T.; Ho, J. T.; McBee, J. K.; Conrads, T.; Veenstra, T. Proteomic Analysis in the Neurosciences. Mol. Cell. Proteomics 2002, 1, 553–560. (5) Hopkins, A. L.; Groom, C. R. The druggable genome. Nat. Rev. Drug Discovery 2002, 1, 727–730. (6) Stevens, T. J.; Arkin, I. T. Do more complex organisms have a greater proportion of membrane proteins in their genomes. Proteins 2000, 39, 417–420. (7) Priller, C.; Bauer, T.; Mitteregger, G.; Krebs, B.; Kretzschmar, H. A.; Herms, J. Synapse formation and function is modulated by the amyloid precursor protein. J. Neurosci. 2006, 27, 7212–7221. (8) Vercauteren, F. G.; Bergeron, J. J.; Vandesande, F.; Arckens, L.; Quirion, R. Proteomic approaches in brain research and neuropharmacology. Eur. J. Pharmacol. 2004, 500, 385–398. (9) Hooper, N. M. Roles of proteolysis and lipid rafts in the processing of the amyloid precursor protein and prion protein. Biochem. Soc. Trans. 2005, 33, 335–338. (10) Jia, J.; Lamer, S.; Schuemann, M.; Schmidt, M. Quantitative Proteomics Analysis of Detergent-resistant Membranes from Chemical Synapses. Mol. Cell. Proteomics 2006, 5, 2060–2071. (11) Martosella, J.; Zolotarjova, N.; Liu, H.; Moyer, S. C.; Perkins, P. D.; Boyes, B. E. High recovery HPLC separation of lipid rafts for membrane proteome analysis. J. Proteome Res. 2006, 5, 1301–1312. (12) Behan, A. T.; Foy, M.; Wynne, K.; Clarke, M.; Sullivan, M.; Cotter, D. R.; Maguire, P. B. Analysis of membrane microdomainsassociated proteins in the insular cortex of human brain. Proteomics Clin. Appl. 2007, 1, 1324–1331. (13) Le Bihan, T.; Goh, T.; Stewart, I. I.; Salter, A. M.; Bukhman, Y. V.; Dharsee, M.; Ewing, R.; Wis´niewski, J. R. Differential analysis of membrane proteins in mouse fore- and hindbrain using a labelfree approach. J. Proteome Res. 2006, 5, 2701–2710. (14) Wang, H.; Qian, W. J.; Chin, M. H.; Petyuk, V. A.; Barry, R. C.; Liu, T.; Gritsenko, M. A.; Mottaz, H. M.; Moore, R. J.; Camp, D. G.; Khan, A. H.; Smith, D. J.; Smith, R. D. Characterization of the mouse brain proteome using global proteomic analysis complemented with cysteinyl-peptide enrichment. J. Proteome Res. 2006, 5, 361–369. (15) Washburn, M. P.; Wolters, D.; Yates, J. R. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol. 2001, 19, 242–247. (16) Blonder, J.; Goshe, M. B.; Moore, R. J.; Pasa-Tolic, L.; Masselon, C. D. Enrichment of integral membrane proteins for proteomic analysis using liquid chromatography-tandem mass spectrometry. J. Proteome Res. 2002, 1, 351–360. (17) Blonder, J.; Xiao, Z.; Veenstra, T. D. Proteomic profiling of differentiating osteoblasts. Expert Rev. Proteomics 2006, 3, 483– 496. (18) Nagaraj, N.; Lu, A.; Mann, M.; Wis´niewski, J. R. Detergent-based but gel-free method allows identification of several hundred membrane proteins in single LC-MS runs. J. Proteome Res. 2008, 7, 5028–5032. (19) Chen, P.; Li, X.; Sun, Y.; Liu, Z.; Cao, R.; He, Q.; Wang, M.; Xiong, J.; Xie, J.; Wang, X.; Liang, S. Proteomic analysis of rat hippocampal plasma membrane: characterization of potential neuronal-specific plasma membrane proteins. J. Neurochem. 2006, 98, 1126–1140. (20) Wu, C. C.; MacCoss, M. J.; Howell, K. E.; Yates, J. R. A method for the comprehensive proteomic analysis of membrane proteins. Nat. Biotechnol. 2003, 21, 532–538. (21) Rabilloud, T. Membrane proteins ride shotgun. Nat. Biotechnol. 2004, 21, 508–510. (22) Tani, H.; Kamidate, T.; Watanabe, H. Micelle-mediated extraction. J. Chromatogr., A. 1997, 780, 229–241. (23) Bordier, C. Phase separation of integral membrane protein in Triton X-114 solution. J. Biol. Chem. 1981, 256, 1604–1607. (24) Tantipaiboonwong, P.; Sinchaikul, S.; Sriyam, S.; Phutrakul, S.; Chen, S. T. Different techniques for urinary protein analysis of normal and lung cancer patients. Proteomics 2005, 5, 1140–1149.

research articles

CPE and Delipidation of Porcine Brain Proteins (25) Wessel, D.; Flugge, U. I. A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal. Biochem. 1984, 138, 141–143. (26) Mastro, R.; Hall, M. Protein delipidation and precipitation by trin-butylphosphate, acetone, and methanol treatment for isoelectric focusing and two-dimensional gel electrophoresis. Anal. Biochem. 1999, 273, 313–315. (27) Lowry, O. H.; Rosebrought, N. J.; Farr, A. L.; Randall, R. J. Protein measurement with the Folin fenol reagent. J. Biol. Chem. 1951, 193, 265–275.

(28) Bergquist, J.; Palmblad, M.; Wetterhall, M.; Hakansson, P.; Markides, K. E. Peptide mapping of proteins in human body fluids using electrospray ionization Fourier transform ion cyclotron resonance mass spectrometry. Mass Spectrom. Rev. 2002, 21, 2–15. (29) www.matrixscience.com. (30) Rabilloud, T. Solubilization of proteins for electrophoretic analyses. Electrophoresis 1996, 17, 813–829.

PR100116K

Journal of Proteome Research • Vol. 9, No. 8, 2010 3911