a clean, more efficient method for in-solution digestion of protein mixtures without detergent or...

A Clean, More Efficient Method for In-Solution Digestion of Protein

Mixtures without Detergent or Urea

Sung Chan Kim, Yue Chen, Shama Mirza, Yingda Xu, Jaeick Lee, Pingsheng Liu, andYingming Zhao*

Department of Biochemistry, University of Texas Southwestern Medical Center at Dallas,Dallas, Texas 75390-9038

Received July 11, 2006

Abstract: Proteolytic digestion of a complicated proteinmixture from an organelle or whole-cell lysate is usuallycarried out in a dilute solution of a denaturing buffer, suchas 1-2 M urea. Urea must be subsequently removed byC18 beads before downstream analysis such as HPLC/MS/MS or complete methylation followed by IMAC isolationof phosphopeptides. Here we describe a procedure fordigesting a complicated protein mixture in the absenceof denaturants. Proteins in the mixture are precipitatedwith trichloroacetic acid/acetone for denaturation and saltremoval and resuspended in NH4HCO3 buffer. After trypsi-nolysis, the resulting peptides are not contaminated byurea or other nonvolatile salts and can be dried in aSpeedVac to remove NH4HCO3. When this protocol wasapplied to an extract of A431 cells, 96.8% of the trypticpeptides were completely digested (i.e., had no missedcleavage sites), in contrast to 87.3% of those producedby digestion in urea buffer. We successfully applied thisdigestion method to analysis of the phosphoproteome ofadiposomes from HeLa cells, identifying 33 phosphoryl-ation sites in 28 different proteins. Our digestion methodavoids the need to remove urea before HPLC/MS/MSanalysis or methylation and IMAC, increasing throughputwhile reducing sample loss and contamination fromsample handling. We believe that this method should bevaluable for proteomics studies.

Keywords: in-solution digestion • proteomics • IMAC • phos-phopeptides • adiposomes

Introduction

Tandem mass spectrometry (MS/MS) combined with eitherreversed phase HPLC or multiple dimensional LC has becomean indispensable tool for identifying and quantifying proteinsin protein mixtures.1-3 These techniques have been used toprofile protein complexes, organelles, and whole-cell lysates.

To be successful, this approach to analyzing complex proteinmixtures requires efficient digestion of the sample by a proteasesuch as trypsin. In a typical experiment, the protein mixtureof interest is denatured in 8 M urea, alkylated, and then diluted

to 1-2 M urea. The resulting sample is digested with trypsin.This strategy has been used to digest highly complicatedprotein mixtures, such as yeast whole cell lysate.2 However, saltsand urea present in the resulting digest complicate subsequentanalysis.

Several different approaches for in-solution digestion avoid-ing use of chemical denaturants such as urea or SDS to increasedigestion efficiency with simple protein mixtures or cell or-ganelles have been reported. These methods include the useof thermal denaturation of proteins prior to in-solutiondigestion,4-5 trypsin digestion in organic solvents,6 and the useof microwave irradiation.7 Other in-solution digestion methodsusing CNBr/trypsin or surfactant such as SDS or n-octylglucoside (n-OG) have been developed to efficiently digesthighly hydrophobic samples such as membrane proteins orlipid rafts.8-11 These methods have successfully shown fast,efficient digestion of simple protein mixtures and are usefulfor MALDI-MS analysis.

However, these methods might be not suitable for large-scaleprotein identification or global identification of post-transla-tionally modified peptides, especially when highly complexprotein mixtures such as whole cell lysates are applied. Becausestrong surfactants such as urea or SDS are commonly used toprepare whole cell lysates, these MS-unfriendly contaminantsmust be subsequently removed by C18 beads before down-stream analysis such as HPLC/MS/MS or complete methylationfollowed by IMAC isolation of phosphopeptides.

Here we report a cleaner procedure for in-solution digestionof complicated protein mixtures such as whole-cell lysates. Thismethod involves precipitating the protein mixture of interestwith trichloroacetic acid/acetone (TCA/acetone), which dena-tures the proteins and removes all the salts in the sample. Theproteins are then partially resolubilized in NH4HCO3 buffer anddigested with trypsin. After reduction of the disulfide bondsand alkylating cysteine residues, more trypsin is added toensure complete digestion. We describe analysis of two com-plex mixtures using this procedure: whole cell lysate from A431cells, and the phosphoproteome of HeLa cell adiposomes. Weexpect this digestion procedure will find applications in shotgunproteomics and large-scale analysis of protein phosphorylationby IMAC.

Experimental Section

Materials. Fetal bovine serum (FBS), trypsin, and Dulbecco’smodified Eagle’s medium (DMEM) were from Life Technologies

* Corresponding author. E-mail: [email protected],Fax: (214) 648-2797; Tel: (214) 648-7947.

3446 Journal of Proteome Research 2006, 5, 3446-3452 10.1021/pr0603396 CCC: $33.50 2006 American Chemical SocietyPublished on Web 11/10/2006

Inc. (Carlsbad, CA). Dulbecco’s phosphate buffered saline (PBS)was from Sigma (St. Louis, MO). Urea, thiourea, CHAPS,ammonium bicarbonate, and dithiothreitol were from FisherScientific Corp. (Pittsburgh, PA). Colloidal Blue Staining Kit wasfrom Invitrogen (Carlsad, CA). Sequencing-grade trypsin wasfrom Promega (Madison, WI). µC18 ZipTips were from MilliporeCorp. (Bedford, MA), and protease inhibitor mixture was fromRoche (Indianapolis, IN).

Preparation of Cell Lysate from A431 Cells. A431 cells weregrown in DMEM supplemented with 10% FBS and 1% penicil-lin/streptomycin. After they reached confluence, the cells werewashed twice with cold Dulbecco’s PBS, and then 400 µL oflysis buffer (7 M urea, 2 M thiourea, 4% CHAPS, 50 mM Tris-HCl, pH 8.5, and protease inhibitor mixture) was added. Thecell lysate was harvested and sonicated 3 times for 10 s eachwith 30-s intervals between sonications. The lysate was cen-trifuged at 22000g for 1 h. The pellet was discarded, and 5 mgof the cellular lysate was precipitated with TCA/acetone todenature whole proteins. Briefly, 1 volume of protein samplewas mixed with a solution containing 1 volume of TCA and 8volumes of acetone. The resulting solution was mixed and keptat -20 °C for 2 h. The protein pellet was recovered bycentrifugation at 22000g for 10 min. The pellet was rinsed twicewith cold acetone to remove residual salts.

Isolation of HeLa Cell Adiposomal Proteins. HeLa cells werecultured in DMEM (high glucose) with 10% fetal bovine serum,1% penicillin/streptomycin. HeLa cells were incubated with 100µM oleate overnight prior to harvesting. Lipid droplets/adipo-somes were purified by a modification of the method of Liu etal.12 Briefly, confluent cells from 40 × 150 mm plates werescraped into ice-cold PBS with proteinase inhibitor PMSF,resuspended in buffer A (25 mM tricine, pH 7.6, 250 mMsucrose, 0.2 mM PMSF), and homogenized by N2 cavitation (450psi for 15 min on ice). The postnuclear supernatant (PNS)fraction (11 mL) was obtained by centrifugation at 1000g andloaded into a SW41 tube. The sample was centrifuged at274000g for 1 h at 4 °C. The white band containing lipiddroplets at top of gradient was collected in 0.5 mL andresuspended in 6 mL of buffer A in a SW41 tube. The droplet/adiposome fraction was overlaid with 4 mL of buffer B (20 mMHEPES, pH 7.4, 100 mM KCl, 2 mM MgCl2), and the samplewas centrifuged at 274000g for 1 h at 4 °C. The droplet band atthe top of gradient was collected in 0.5 mL. Then the samplewas centrifuged at 20000g for 4 min and the procedure repeatedfour times. An additional wash in 1 mL of buffer B using a265000g centrifugation in a TLA 100.3 tube for 5 min wasapplied to remove contaminating membranes. The adiposomeproteins were precipitated by TCA/acetone.

Digestion in the Presence of Urea or SDS. Five miligramsof the TCA/acetone-precipitated and denatured pellet wasresolubilized in 1 mL of 100 mM NH4HCO3 (pH 8.0)/8 M ureaor 100 mM NH4HCO3 (pH 8.0)/2% SDS. The resolubilizedproteins were reduced with 5 mM dithiothreitol (50 °C, 30 min)and then alkylated with 15 mM iodoacetamide at roomtemperature for 30 min in the dark. The reaction was quenchedwith 15 mM cysteine at r.t. for 30 min. Before tryptic digestion,100 mM ammonium bicarbonate buffer was added to reducethe concentration of urea or SDS to 1 M or 0.1%, respectively.For in-solution digestion, trypsin was added to the proteinmixture at an enzyme-to-substrate ratio of 1:50 (w/w). Afterincubation at 37 °C for 16 h, additional trypsin (1:100, w/w)was added to the sample and incubation was continued for 3h to ensure complete digestion. Forty micrograms of the

resulting tryptic peptides or undigested proteins was subjectedto SDS-PAGE followed by colloidal blue staining. µC18 ZipTipswere used to wash the tryptic peptides according to themanufacturer’s directions before nano-HPLC/MS/MS analysis.

Clean Digestion in NH4HCO3 Buffer. Five miligrams of theTCA/acetone-precipitated and denatured pellet was resus-pended in 1 mL of 100 mM NH4HCO3 (pH 8.0), resulting inonly partial solubilization of the peptides. The protein pelletresuspended by NH4HCO3 was ground by a round glass rod ina microcentrifuge tube and sonicated three times for 20 s eachwith 30-s intervals between sonications to make a homoge-neous protein suspension. For in-solution digestion, trypsinwas added to the protein mixture at an enzyme-to-substrateratio of 1:50 (w/w). After incubating at 37 °C for 16 h, the trypticpeptides were reduced and alkylated as described above.Additional trypsin (1:100 w/w) was added, and the mixture wasincubated at 37 °C for 3 h to ensure complete digestion. Fortymicrograms of the resulting tryptic peptides or undigestedproteins were subjected to SDS-PAGE followed by colloidal bluestaining. µC18 ZipTips were used to clean the tryptic peptidesbefore nano-HPLC/MS/MS analysis.

Methylation of Adiposomal Proteins. Methylation of car-boxylic groups of acidic amino acid residues (D and E) andthe C-termini of peptides was carried out using a procedurepreviously described.13 Briefly, 100 µg of tryptic digest was driedin a SpeedVac (Thermo Savant, San Jose, CA) for 6 h. Then 50µL of 2 M methanolic HCl was added to the dried peptides,and the reaction was carried out for 2 h on a shaker. Theanhydrous methanol used for the preparation of methanolicHCl had been distilled against CaH2 by a standard method. Thepeptide mixture was thoroughly dried, and the methylationreaction was repeated to ensure complete conversion ofcarboxylic acid groups to their corresponding methyl esters.

Batchwise IMAC. Batchwise IMAC was carried out accordingto an optimized procedure. First, Poros 20 MC beads wereactivated according to the manufacturer’s instructions. A 10-µL bed volume of activated beads was equilibrated with 200µL of loading solution [acetonitrile/methanol/water (1:1:1, v/v/v) in 0.1% acetic acid] prior to sample loading. One hundredmicrograms of methylated adiposomal tryptic peptides wasdissolved in 20 µL of acetonitrile/methanol/water (1:1:1, v/v/v), and the pH was adjusted to 2.5-3.0 with 2 M NH4HCO3

(pH 8.0). The peptide mixture was then centrifuged at 100000gfor 20 min to remove insoluble particles. The resulting super-natant was mixed with IMAC beads and incubated at roomtemperature with shaking for 30 min. After incubation, thesuspension was centrifuged in a microcentrifuge at 13000g for1 min, and the supernatant was removed. The beads werewashed once with 200 µL of washing buffer I [acetonitrile/methanol/water/acetic acid (75:10:14:1, v/v/v/v) to which wasadded NaCl to 100 mM], followed by two washes with 200 µLof washing buffer II [acetonitrile/water/acetic acid (85:14:1, v/v/v)]. Bound phosphopeptides were eluted three times from theIMAC beads with 50 µL of elution solution [acetonitrile/water/trifluoroacetic acid (45/50/5, v/v/v)].

Nano-HPLC/MS/MS Analysis. HPLC/MS/MS analysis wasperformed in an LCQ DECA XP ion-trap mass spectrometer(ThermoFinnigan, San Jose, CA) equipped with a nano-ESIsource. The electrospray source was coupled online with anAgilent 1100 series nano flow LC system (Agilent, Palo Alto, CA).Two microliters of the peptide solution in loading buffer (2%acetonitrile/97.9% water/0.1% acetic acid (v/v/v)) was manuallyinjected and separated in a nano-HPLC column (50 mm length

technical notes Kim et al.

Journal of Proteome Research • Vol. 5, No. 12, 2006 3447

× 75 µm ID, 5 µm particle size, 300 Å pore diameter) packedin-house with Luna C18 resin (Phenomenex, St. Torrance, CA).Peptides were eluted from the column with a gradient of 5%to 80% buffer B (90% acetonitrile/9.9% water/0.1% acetic acid(v/v/v)) in buffer A (2% acetonitrile/97.9% water/0.1% aceticacid (v/v/v)) over 10 min. Eluted peptides were electrosprayeddirectly into the mass spectrometer. MS/MS spectra wereacquired in a data-dependent mode, in which the two strongestions in each MS spectrum were selected for fragmentation.

Mass Spectrometry and Protein Sequence Database Search-ing To Identify Phosphoproteins. Nano-HPLC/LTQ massspectrometry was carried out as previously described14 exceptthat a 9-min gradient of 6-90% B buffer (90% acetonitrile,9.95% water, 0.05% acetic acid) in A buffer (97.95% water, 2%acetonitrile, 0.05% acetic acid) at a low flow rate of 0.1 µL/minwas used. The LTQ was operated in a data-dependent modewhere one full MS scan was followed by four pairs of MS2/MS3 scans. MS3 was automatically triggered when a neutralloss peak of 98, 49 or 32.7 ((2) m/z was detected among thetop eight most intense peaks in the MS2 spectrum. MS3fragmentation on the dominant neutral loss ions generated arich backbone fragmentation pattern for correct peptide iden-tification and accurate phosphorylation site determination.Normalized collision energy was set to 22% during MS2acquisition and 35% during MS3 acquisition.

Protein Sequence Database Searching and Manual Verifi-cation. Tandem mass spectra were used to search the NCBI-nr database with the Mascot search engine (version 2.0, MatrixScience, London, UK). Mass tolerance was set to (4 Da forparent ion masses and (0.6 Da for fragment ion masses.Peptides with Mascot scores above 40 were considered poten-tial positive identifications and were manually verified. Strictmanual analysis was applied to validate protein identificationresults, using the following criteria. y, b, and a ions as well astheir water loss or amine loss peaks were considered. Fordoubly charged or triply charged ions, all the major isotopicallyresolved peaks matched fragment masses of the identifiedpeptide. The isotopically resolved peaks were emphasizedbecause single peaks can come from electronic sparks and areless likely to be relevant to peptide fragments. The majorisotopically resolved peaks for doubly charged or triply chargedions were defined as (1) those isotopically resolved daughterions with m/z higher than parent m/z and intensity higher than5% of the maximum intensity or (2) those isotopically resolvedpeaks with intensities higher than 20% of the maximumintensity and m/z values between one-third of the parent m/zand the parent m/z. Typically, more than seven independent(amine and water loss not counted), isotopically resolved peakswere matched to theoretical masses of the peptide fragments.

For manual validation of singly charged parent ions, theMascot score was required to be equal to or above the identityscore threshold of the peptide. For peptides ending witharginine or lysine residues, at least four consecutive aminoacids in the peptide sequence had to be confirmed with theisotopically resolved b-ion and the y-ion series.

Results

A Clean Method To Digest Proteins in NH4HCO3 Buffer.Tryptic digests are usually performed in a reaction buffer suchas NH4HCO3 to produce tryptic peptides without contaminationfrom detergents, urea, or salts. For example, NH4HCO3 bufferis routinely used for in-gel digestion. However, NH4HCO3 bufferis intrinsically unable to extract proteins from either organelles

or cells. Instead, proteins are typically extracted with buffercontaining detergent or urea, which has the additional effectof denaturing the proteins. The resulting solution may thenbe diluted 4- to 8-fold with buffer to reduce the denaturantconcentration to a level at which trypsin is active.

A problem with this procedure is that the presence of ureaor detergents is not desirable for HPLC/mass spectrometricanalysis, or for methylation of carboxylic acids (as is commonlydone when IMAC is used to isolate phosphopeptides). Inprinciple, this problem can be solved by precipitating theprotein extract with TCA/acetone to remove denaturants andresuspending in buffer amenable to subsequent analysis.Unfortunately, NH4HCO3 buffer, the buffer of choice for trypticdigests, does not completely solubilize TCA/acetone proteinpellets. To our knowledge, NH4HCO3 buffer has not been usedto resuspend protein pellets for in-solution digestion of proteinmixtures before proteomics analysis.

To test if protein pellets that have been partially solubilizedin NH4HCO3 buffer can be digested efficiently, 100 mM NH4-HCO3 (pH 8.0) was added to the denatured protein pelletobtained from TCA/acetone precipitation of the whole lysatefrom A431 cells. As expected, the denatured protein pellet couldnot be completely dissolved; small protein particles remainedvisible in a cloudy suspension (Figure 1A), suggesting incom-plete dissolution. During the course of tryptic digestion, thesuspended particles gradually disappeared (Figure 1A). Becausetryptic peptides are likely to be more soluble than the originalproteins, this observation suggests that proteolysis was pro-

Figure 1. Digestion of suspended proteins by trypsin. (A)Photographs of protein sample during the course of trypticdigestion at 0, 1, and 16 h. Five milligrams of the TCA/acetone-precipitated and denatured protein from A431 whole-cell lysatewas resuspended in 1 mL of 100 mM NH4HCO3. The resultingsuspension was digested with trypsin at an enzyme-to-substrateratio of 1:50 (w/w). (B) SDS-polyacrylamide gel, stained withcolloidal blue, showing protein samples before and after 16-hdigestion with trypsin (enzyme-to-substrate ratio 1:50, w/w). Fortymicrograms of protein or tryptic peptides was loaded onto a 12%SDS-PAGE gel. U: trypsinization in the presence of 1 M urea; S:trypsinization in the presence of 0.1% SDS; N: trypsinization inthe presence of 100 mM NH4HCO3.

Clean In-Solution Digestion of Protein Mixtures technical notes

3448 Journal of Proteome Research • Vol. 5, No. 12, 2006

ceeding. After 16 h of digestion, the solution was transparent(Figure 1A). This result is consistent with previous report thataggregated protein pellets generated by thermal denaturationwere easily digested and disappeared after trypsin digestionfor 3 h and resulted in more specific cleavage and generationof sufficient numbers of tryptic peptides for protein identifica-tion.4 This result suggests that the denatured protein particlesproduced by TCA/acetone precipitation can be easily digestedby trypsin in aqueous solution without any detergents as shownin Figure 1.

In a parallel experiment, the same amount of protein pelletfrom A431 cells was solubilized in either urea or SDS buffer,diluted, and subsequently digested with trypsin. When samplesfrom digestion reactions containing 1 M urea, 0.1% SDS, or100 mM NH4HCO3 were analyzed by SDS-PAGE, proteins werenearly undetectable using Coomassie blue staining (Figure 1B),indicating that trypsin could digest proteins in all threesolutions.

To check the degree of digestion, tryptic peptides fromsamples digested in the presence of urea or NH4HCO3 wereanalyzed exhaustively by nano-HPLC/LCQ mass spectrometry(Figure 2). Tryptic peptides from the digestion containing SDSwere not analyzed by mass spectrometry due to the difficultyof efficiently removing SDS from the peptides.

Mass spectrometric analysis of tryptic peptides from NH4-HCO3-buffered digestion in combination with automated pro-tein sequence database searching and manual verification ledto the identification of 652 tryptic peptides, of which 631(96.8%) were completely digested, 20 (3.1%) contained 1 missedcleavage site, and 1 (0.1%) contained 2 missed cleavage sites(Figure 3). Analysis of peptides digested in the presence of ureaidentified 664 peptides, of which 580 (87.3%) were completelydigested, 81 (12.1%) contained 1 missed cleavage site, and 3(0.5%) contained 2 missed cleavage sites (Figure 3).

Our results suggest that proteins can be completely digesteddespite only partial solubilization in NH4HCO3 buffer. It was

Figure 2. Nano-HPLC/MS/MS analysis of tryptic peptides from digestion of A431 whole-cell lysate in the presence of urea or NH4HCO3.(A and D) Total ion chromatograms; (B and E) mass spectra obtained at retention times of 55.81 and 29.29 min, respectively; (C andF) tandem mass spectra of peaks at m/z 803.7 (at retention time 55.81 min) and 480.8 (at retention time 29.29 min), identified as peptidesVVLAYEPVWAIGTGK from triosephosphate isomerase 1 and LNVTEQEK from enolase 1, respectively.



surprising to observe that digestion of suspended proteins inNH4HCO3 buffer was more complete than digestion of fullydissolved proteins in dilute urea buffer. The difference in degreeof digestion might arise from the renaturing of proteins in 1 Murea; the resulting folded proteins could be more resistant toenzymatic digestion. Alternatively, 1 M urea might lower theenzymatic activity of trypsin due to low-level inhibition15 ordenaturation of the enzyme.

Systematic Analysis of Incompletely Digested Peptides. Weanalyzed the amino acid residues surrounding the non-C-terminal arginine and lysine residues in tryptic peptidescontaining missed cleavage sites. A high proportion of arginineand lysine residues at which cleavage was missed were adjacentto an acidic residue (38% for urea digestion and 41% for NH4-HCO3 digestion, Figure 4). Adjacent acidic residues mightinteract with the basic side chains, interfering with the bindingof trypsin to lysine and arginine. Amino acid residues withpotential to sterically hinder trypsin binding, such as valine

and isoleucine, were also commonly observed adjacent tomissed cleavage sites.

Application of the Clean Digestion Protocol to Subpro-teome Analysis. A clean digestion procedure can streamlinesample preparation. Eliminating the need for desalting andbuffer exchange increases throughput while reducing sampleloss and contamination associated with sample handling. Thenet result should be increased sensitivity and improved sampleintegrity at the LC-MS/MS stage. To test our optimized diges-tion procedure, we applied it in conjunction with phospho-peptide enrichment through batchwise IMAC to the analysisof the adiposome phosphoproteome of HeLa cells.

The adiposome is a metabolically active cellular lipid-storageorganelle.12 This dynamic organelle has been implicated in suchprocesses as lipid and sterol metabolism and membranetransport pathways. Consistent with these proposed functions,our proteomic screen of adiposomal phosphoproteins identi-fied 7 categories of proteins, including those involved in fattyacid metabolism and signal transduction as well as protein andvesicle trafficking. In all, 33 phosphorylation sites were identi-fied in 32 peptides representing 28 proteins (Table 1).

Several observed phosphorylation sites were within knownconsensus sequence motifs. Two sites conformed to thecalcium/calmodulin-dependent kinase (CAMK) motif. Thisresult is reasonable given the central role of calcium in theregulation of signaling cascades and enzymatic activities withinthe cell. We also identified two proteins phosphorylated withinputative PKC recognition motifs. PKC isoforms are involved intransducing signals received in the form of second messengerssuch as diacylglycerol, calcium, and inositol triphosphate,which result from activation of receptor tyrosine kinases andG protein-coupled receptors by extracellular ligands. Becausethe changing extracellular environment is expected to influencecellular metabolic needs, one might predict that the adiposomeis subject to regulation by PKC. Phosphorylation was identifiedat one site corresponding to the recognition sequence of Akt,a protein kinase implicated as a mediator of extracellular andintracellular signaling events. The effects of Akt range frominhibition of apoptosis to transduction of insulin signalingcascades and regulation of the cell cycle. As an energy storage

Figure 3. Distribution of identified tryptic peptides from proteinsresolubilized by urea or resuspended by NH4HCO3 buffer. (A)Total number of identified peptides and proteins from digestionin the presence of urea and NH4HCO3. Tryptic peptides from 2µg of protein from A431 whole-cell lysate were analyzed by nano-HPLC/mass spectrometry. (B and C) Pie charts showing thedistribution of tryptic peptides with 0, 1, and 2 missed cleavagesites.

Figure 4. Nature of amino acids surrounding the non-C-terminalarginine and lysine residues in peptides with missed cleavagesites shown in Figure 3.



depot and active metabolic organelle, one might expect theadiposome to be influenced by the action of Akt.

The discovery of such a contingent of adiposomal phos-phoproteins is novel and suggests that the functions of theorganelle are actively regulated by phosphorylation events.Furthermore, the results demonstrate that the clean digestionprocedure is amenable to downstream sample enrichmentstrategies without the need for desalting or buffer exchange.

Discussion

We developed a clean method for in-solution digestion of aprotein mixture. The resulting tryptic peptides can be dried inSpeedVac to remove solvents and buffer, producing trypticpeptides devoid of detergents, urea, and salts. The results fromexhaustive identification of peptides from A431 whole-celllysate suggest that the protocol produces almost completelydigested tryptic peptides (∼96.8% peptides with 0 missedcleavage sites), and more complete digestion than obtained inthe presence of dilute (1 M) urea (87.3% peptides with 0 missedcleavage sites). It is likely that the clean digestion methoddescribed here will be applicable to proteolytic digestion usingother enzymes. Using the clean digestion method eliminatesthe need for additional steps to remove salts, urea, or detergents

before HPLC/MS/MS analysis, or before methylating carboxylicacids prior to IMAC.

The described digestion method is highly amenable tophosphoproteomic analysis. Previously, digestion of cell lysateshas almost exclusively been performed in 1-2 M urea. Phos-phopeptides are lost with this method because of the need todesalt tryptic peptides before methylation. When peptides areproduced by our digestion procedure, desalting is necessaryonly after methylation and IMAC, resulting in improvedrecovery of phosphopeptides because methylated phospho-peptides have higher hydrophobicity. The absence of agentssuch as urea also facilitates complete removal of water, animportant factor for complete methylation. Thus, the improveddigestion procedure should be valuable due to improvedrecovery of phosphopeptides and more complete methylationof carboxylic acids. The advantages of the method are il-lustrated in the phosphoproteomic analysis of HeLa celladiposomes, in which 33 phosphorylation sites were identifiedamong 32 peptides from 28 proteins.

Acknowledgment. Y.Z. is supported by The Robert A.Welch Foundation (I-1550) and NIH (CA 107943).

Table 1. Adiposomal Phosphoproteinsa

protein name GI no. peptide sequence Mascot score

Signal TransductionMel transforming oncogene 16933567 LEGNSPQG pSNQGVK 45Mel transforming oncogene 16933567 LEGN pSPQG pSNQGVK 36RET tyrosine kinase (cAMP dependent) 337358 EDEI pSPPPPNPVVK 48phospholipase A2, group IVA 23943920 HIVSNDSSD pSDDESHEPK 70patatin-like phospholipase domain containing 2 3005717 RVQ pSLPSVPLSCAAYR 66patatin-like phospholipase domain containing 2 3005717 NNLpSLGDALAK 38RAB5C, member RAS oncogene family 54695838 ADLA pSK 32progesterone membrane binding protein 5453916 LLKPGEEPSEY pTDEEDTK 33calnexin 422774 AEEDEILNR pSPR 37calnexin 422774 QK pSDAEEDGGTV pSQEEEDRKPK 41

Transcription/Translationcentrosomal protein 1 34535303 KI pSEAGK 28ribosomal protein P1 isoform 1 4506669 KEESEE pSDDDMGFGLFD 45unnamed protein product (hypothetical protein FLJ10005) 10437063 DGQDAIAQ pSPEKESK 31DNA topoisomerase (ATP-hydrolyzing) 105857 IKNENTEG pSPQEDGVELEGLK 60

Protein Transportvacuolar sorting protein 4 5381417 G pSDSDSEGDNPEKK 31sortilin precursor 84028263 SGYHDD pSDEDLLE 29translocation protein 1 4507525 SDSEEK pSDSEK 60

Structuralmyosin, heavy polypeptide 9, non-muscle 34526505 KGAGDG pSDEEVDGKADGAEA KPAE 96erythrocyte membrane protein band 4.1-like 1 2224617 SEAEEGEVR pTPTK 37stromal interaction molecule 1 precursor 17368447 AEQ pSLHDLQER 45Lamin Adel10 14290259 LRL pSPSPTSQR 39

ChaperoneHsp70-interacting protein 4928064 LGAGGG pSPEKSPSAQELK 43HSPCB protein 33987931 IEDVG pSDEEDDSGKDKK 66heat shock 90kDa protein 1, alpha 32486 DKEV pSDDEAEEK 48novel DnaJ domain protein 11125675 SL pSTSGESLYHVLGLDK 56

MetabolismNADH dehydrogenase (ubiquinone) Fe-S protein 1, 75kDa 7770127 VAV pTPPGLAR 37acetyl-CoA carboxylase 1 33112873 FIIGSVSEDN pSEDEISNLVK 86acetyl-CoA carboxylase 1 33112873 SSM pSGLHLVK 57ubiquitin specific protease 33 5689531 LSA pSPPK 38ancient ubiquitous protein 1 isoform 1 6912260 LRPQSAQSSFPPSPGP pSPDVQLATLAQR 46

Unknownhypothetical protein 12053153 DPLLFKSASD pTNLQK 29tumor differentially expressed 2 11282574 pSDGSLEDGDDVHR 36

a Adiposomes from HeLa Cells were isolated, and proteins were extracted with 8 M urea, precipitated with TCA/acetone, and resuspended in 100 mMNH4HCO3. The suspension was digested with trypsin and subjected to IMAC as described in Materials and Methods, and the resulting peptides were analyzedby nano-HPLC/mass spectrometry. Identified phosphoproteins are grouped according to their functional annotation. For peptide sequences, a ‘p’ in front ofS Or T indicates the phosphorylation site.



References

(1) Aebersold, R.; Mann, M. Mass spectrometry-based proteomics.Nature 2003, 422, 198-207.

(2) Washburn, M. P.; Wolters, D.; Yates, J. R. 3rd. Large-scale analysisof the yeast proteome by multidimensional protein identificationtechnology. Nat. Biotechnol. 2001, 19, 242-247.

(3) Link, A. J.; Eng, J.; Schieltz, D. M.; Carmack, E.; Mize, G. J.; Morris,D. R.; Garvik, B. M.; Yates, J. R. 3rd. Direct analysis of proteincomplexes using mass spectrometry. Nat. Biotechnol. 1999, 17,676-682.

(4) Park, Z. Y.; Russell, D. H. Thermal denaturation: A usefultechnique in peptide mass mapping. Anal. Chem. 2000, 72, 2667-2670.

(5) Park, Z. Y.; Russell, D. H. Identification of individual proteins incomplex protein mixtures by high-resolution, high-mass-accuracyMALDI TOF-mass spectrometry analysis of in-solution thermaldenaturation/enzymatic digestion. Anal. Chem. 2001, 73, 2558-2564.

(6) Russell, W. K.; Park, Z. Y.; Russell, D. H. Proteolysis in mixedorganic-aqueous solvent systems: applications for peptide massmapping using mass spectrometry. Anal. Chem. 2001, 73, 2682-2685.

(7) Pramanik, B. N. et al. Microwave-enhanced enzyme reaction forprotein mapping by mass spectrometry: a new approach toprotein digestion in minutes. Protein Sci. 2002, 11, 2676-2687.

(8) Quach, T. T. T. et al. Development and applications of in-gelCNBr/tryptic digestion combined with mass spectrometry for theanalysis of membrane proteins. J. Proteome. Res. 2003, 2, 543-552.

(9) Zhang, N.; Li, L. Effects of common surfactants on proteindigestion and matrix-assisted laser desorption/ionization mass

spectrometric analysis of the digested peptides using two-layersample preparation. Rapid Commun. Mass Spectrom. 2004, 18,889-896.

(10) Li, N.; Shaw, A. R. E.; Zhang, N.; Mak, A.; Li, L. Lipid raftproteomics: analysis of in-solution digest of sodium dodecylsulfate-solubilized lipid raft proteins by liquid chromatography-matrix-assisted laser desorption/ionization tandem mass spec-trometry. Proteomics 2004, 4, 3156-3166.

(11) Yu, Y. Q.; Gilar, M.; Gebler, J. C. A complete peptide mapping ofmembrane proteins: a novel surfactant aiding the enzymaticdigestion of bacteriorhodopsin. Rapid Commun. Mass Spectrom.2004, 18, 711-715.

(12) Liu, P.; Ying, Y.; Zhao, Y.; Mundy, D. I.; Zhu, M.; Anderson, R. G.Chinese hamster ovary K2 cell lipid droplets appear to bemetabolic organelles involved in membrane traffic. J. Biol. Chem.2004, 279, 3787-3792.

(13) Ficarro, S. B.; McCleland, M. L.; Stukenberg, P. T.; Burke, D. J.;Ross, M. M.; Shabanowitz, J.; Hunt, D. F.; White, F. M. Phos-phoproteome analysis by mass spectrometry and its applicationto Saccharomyces cerevisiae. Nat. Biotechnol. 2002, 20, 301-305.

(14) Chen, Y.; Kwon, S. W.; Kim, S. C.; Zhao, Y. Integrated approachfor manual evaluation of peptides identified by searching proteinsequence databases with tandem mass spectra. J. Proteome Res.2005, 4, 998-1005.

(15) Shannon, J. Using proteinases for Edman sequence analysis andpeptide mapping. In Proteolytic Enzymes: A Practical Approach;Beynon, R., Bond, J. S., Eds., Oxford University Press: New York,NY, 2001; pp 187-210.

PR0603396



a clean, more efficient method for in-solution digestion of protein mixtures without detergent or...

Documents