Anal. Chem. 2004, 76, 1817-1823
Proteomic Analysis of Integral Plasma Membrane Proteins Yingxin Zhao, Wei Zhang, Yoonjung Kho, and Yingming Zhao*
Department of Biochemistry, University of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75390-9038
Efficient methods for profiling proteins integral to the plasma membrane are highly desirable for the identification of overexpressed proteins in disease cells. Such methods will aid in both understanding basic biological processes and discovering protein targets for the design of therapeutic monoclonal antibodies. Avoiding contamination by subcellular organelles and cytosolic proteins is crucial to the successful proteomic analysis of integral plasma membrane proteins. Here we report a biotindirected affinity purification (BDAP) method for the preparation of integral plasma membrane proteins, which involves (1) biotinylation of cell surface membrane proteins in viable cells, (2) affinity enrichment using streptavidin beads, and (3) depletion of plasma membraneassociated cytosolic proteins by harsh washes with highsalt and high-pH buffers. The integral plasma membrane proteins are then extracted and subjected to SDS-PAGE separation and HPLC/MS/MS for protein identification. We used the BDAP method to prepare integral plasma membrane proteins from a human lung cancer cell line. Western blotting analysis showed that the preparation was almost completely devoid of actin, a major cytosolic protein. Nano-HPLC/MS/MS analysis of only 30 µg of protein extracted from the affinity-enriched integral plasma membrane preparation led to the identification of 898 unique proteins, of which 781 were annotated with regard to their plasma membrane localization. Among the annotated proteins, at least 526 (67.3%) were integral plasma membrane proteins. Notable among them were 62 prenylated proteins and 45 Ras family proteins. To our knowledge, this is the most comprehensive proteomic analysis of integral plasma membrane proteins in mammalian cells to date. Given the importance of integral membrane proteins for drug design, the described approach will expedite the characterization of plasma membrane subproteomes and the discovery of plasma membrane protein drug targets.
The plasma membrane provides a physical boundary between the cell and its environment, playing important roles in many fundamental biological processes such as cell-cell interactions, signal transduction, and material transport. Proteins involved in * To whom correspondence should be addressed. E-mail: biochem.swmed.edu. Fax: (214) 648-2797. Tel: (214) 648-7947. 10.1021/ac0354037 CCC: $27.50 Published on Web 02/28/2004
yzhao@
© 2004 American Chemical Society
the functions of the plasma membrane may be “integral” (i.e., anchored to the lipid bilayer directly through a transmembrane region or lipid modification) or “associated” (i.e., anchored to the lipid bilayer through noncovalent interactions with integral membrane proteins or other membrane-associated proteins). Plasma membrane-associated proteins contain neither transmembrane domains nor lipid modifications. The plasma membrane has been extensively targeted for drug design; plasma membrane proteins account for ∼70% of all known drug targets (e.g., HER2- and G protein-coupled receptors).1 Identification of overexpressed plasma membrane proteins in disease cells would provide protein targets for the design of either therapeutic monoclonal antibodies or small-molecule drugs. Most proteomic analyses are based on classical 2D-gel/mass spectrometry,2 “shotgun” mass spectrometry,3-5 or ICAT/mass spectrometry.6 At best, these methods are able to identify a few thousand of the most abundant proteins in cells.7 Therefore, prior enrichment of the plasma membrane fraction is desired for efficient proteomic analysis of plasma membrane proteins. Analysis of plasma membrane subproteomes has proven to be challenging due to the difficulty of preparing a clean plasma membrane fraction and the large size and hydrophobic nature of the integral plasma membrane proteins. Preparation of a clean plasma membrane fraction prior to proteomic analysis is essential to reduce the complexity of the sample, to simplify the proteomic analysis, and to optimize the efficiency of the analysis of integral plasma membrane subproteomes.8-12 (1) Hopkins, A. L.; Groom, C. R. Nat. Rev. Drug. Discuss. 2003, 1, 727-730. (2) Hanash, S. Nature 2003, 422, 226-232. (3) Wu, C. C.; MacCoss, M. J.; Howell, K. E.; Yates, J. R. Nat. Biotechnol. 2003, 21, 532-538. (4) Goshe, M. B.; Blonder, J.; Smith, R. D. J. Proteome Res. 2003, 2, 153-161. (5) Han, D. K.; Eng, J.; Zhou, H.; Aebersold, R. Nat. Biotechnol. 2001, 19, 946-951. (6) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Nat. Biotechnol. 1999, 17, 994-999. (7) Aebersold, R.; Mann, M. Nature 2003, 422, 198-207. (8) Zhao, Y.; Zhang, W.; White, M. A.; Zhao, Y. Anal. Chem. 2003, 75, 37513757. (9) Shin, B. K.; Wang, H.; Yim, A. M.; Le Naour, F.; Brichory, F.; Jang, J. H.; Zhao, R.; Puravs, E.; Tra, J.; Michael, C. W.; Misek, D. E.; Hanash, S. M. J. Biol. Chem. 2003, 278, 7607-7616. (10) Adam, P. J.; Boyd, R.; Tyson, K. L.; Fletcher, G. C.; Stamps, A.; Hudson, L.; Poyser, H. R.; Redpath, N.; Griffiths, M.; Steers, G.; Harris, A. L.; Patel, S.; Berry, J.; Loader, J. A.; Townsend, R. R.; Daviet, L.; Legrain, P.; Parekh, R.; Terrett, J. A. J. Biol. Chem. 2003, 278, 6482-6489. (11) Simpson, R. J.; Connolly, L. M.; Eddes, J. S.; Pereira, J. J.; Moritz, R. L.; Reid, G. E. Electrophoresis 2000, 21, 1707-1732. (12) Santoni, V.; Kieffer, S.; Desclaux, D.; Masson, F.; Rabilloud, T. Electrophoresis 2000, 21, 3329-3344.
Analytical Chemistry, Vol. 76, No. 7, April 1, 2004 1817
Figure 1. Strategy for affinity enrichment of integral plasma membrane proteins. The plasma membrane fraction is prepared by cell surface biotinylation and affinity enrichment using streptavidin beads. The integral plasma membrane fraction is obtained through removal of cytosolic proteins from the plasma membrane fraction by washing with high-salt and high-pH buffers.
A few advances in the study of plasma membrane subproteomes were recently described. Yates and colleagues used multidimensional protein identification technology for the identification of plasma membrane proteins and plasma membraneassociated proteins, as well as their modification sites.3,13 Isolation of plasma membrane proteins by affinity enrichment methods was described by our group and Hanash’s group.8,9,14 This strategy involves cell surface biotinylation and affinity enrichment using immobilized beads for the isolation of either plasma membrane sheets8,14 or plasma membrane proteins.9 Hanash and colleagues have applied the approach, in combination with 2D-gel separation and quantification, to globally profile the cell surface proteomes in a few cancer cell lines.9 Our group demonstrated, by both Western blotting analysis and comprehensive proteomic analysis, that the affinity enrichment dramatically reduced contamination of subcellular organelles in the plasma membrane fraction.8,14 Nevertheless, a large number of plasma membrane-associated cytosolic proteins were retained in the plasma membrane fraction, reducing the quantification efficiency for integral plasma membrane proteins.8 Here we report a method for the profiling of integral plasma membrane proteins exposed to the cell surface. The method involves affinity purification of the plasma membrane fraction, removal of membrane-associated cytosolic proteins by washes with high-salt and high-pH buffers, and nano-HPLC/MS/MS analysis for protein identification (Figure 1). Since our method for the (13) Wu, C. C.; Yates, J. R. Nat. Biotechnol. 2003, 21, 262-267. (14) Zhang, W.; Zhou, G.; Zhao, Y.; White, M. A.; Zhao, Y. Electrophoresis 2003, 24, 2855-2863.
1818
Analytical Chemistry, Vol. 76, No. 7, April 1, 2004
preparation of the plasma membrane fraction relies on biotinylation and affinity enrichment, we term it biotin-directed affinity purification (BDAP). We describe the use of this method to analyze a human lung cancer cell line. To the best of our knowledge, this work represents the most comprehensive analysis of an integral plasma membrane proteome in mammalian cells yet reported. EXPERIMENTAL PROCEDURES Materials. RPMI medium 1640, fetal bovine serum (FBS), trypsin, EDTA, penicillin/streptomycin, and Colloidal Blue Staining Kit were from Life Technologies, Inc. (Carlsbad, CA). EZLink Sulfo-NHS-SS-Biotin [Sulfosuccinimidyl-2-(biotinamido) ethyl1,3-dithiopropionate] was from Pierce (Rockford, IL). Streptavidin beads mTRAP Maxi was purchased from Active Motif, Inc. (Carlsbad, CA). Bovine serum albumin and dithiothreitol (DTT) were bought from Fisher Scientific Corp. (Pittsburgh, PA). Immobilon transfer membranes (PVDF) were from Millipore (Bredford, MA). Western Lightning Chemiluminescence Reagent Plus was from Perkin-Elmer Life Science (Boston, MA). Protease Inhibitor Cocktail tablets were purchased from Roche Molecular Biochemicals (Indianapolis, IN). Bio-Rad DC protein assay kit was bought from Bio-Rad Laboratories (Hercules, CA). Anti-actin IgG (mAb) was a gift from Dr. Pingsheng Liu (UT Southwestern Medical Center, Dallas, TX). Horseradish peroxidase-conjugated anti-mouse IgG was from Sigma-Aldrich (St. Louis, MO). Precast polyacrylamide 4-20% gradient minigels were from Continental Lab Products (San Diego, CA).
Methods. Cell Culture and Cell Surface Biotinylation. NCIH1299 cells were grown at 37 °C in RPMI 1640 medium with 10% FBS and 1× penicillin/streptomycin until approaching confluency (∼90%). Dishes (10 cm) of cells were washed with prewarmed (37 °C) PBS three times, and then 10 mL of PBS and 10 µL of 1000 × EZ-Link Sulfo-NHS-SS-Biotin stock solution (100 mg/mL, freshly prepared with DMSO) were added. After the cells were incubated at room temperature for 15 min, the reaction solution was removed and 1 mL of lysine solution (1 mg/mL) was added to quench the reaction. The cells were washed with PBS two times. Preparation of Integral Plasma Membrane Fraction. Cells were washed with ice-cold PBS twice and scraped into ice-cold PBS. Cells were collected by centrifugation at 350g for 10 min, resuspended into 1 mL of ice-cold hypotonic buffer (10 mM HEPES, pH 7.5, 1.5 mM MgCl2, 10 mM KCl, 1×protease inhibitor cocktail, 1 mM NaF, and 1 mM Na3VO4), and incubated on ice for 15 min. The cells were broken by dounce homogenization (50 passes, homogenizer from Kontes Glass Co., Vineland, NJ). Unbroken cells and nuclei were pelleted from the cell homogenate by centrifugation at 1000g for 10 min at 4 °C to generate a postnuclear supernatant (PNS). The KCl concentration in the PNS fraction was adjusted to 150 mM. A 300-µL aliquot of suspended streptavidin magnetic beads (10 mg of beads/mL, prewashed with ice-cold hypotonic buffer four times before use) was added to the PNS fraction; the resulting suspension was rotated at 4 °C for 1 h. The beads were collected using a magnetic plate. The resulting preparation was called the “plasma membrane fraction”. The plasma membrane fraction was subjected to washing with 1 mL of ice-cold 1 M KCl (high-salt wash) three times, 1 mL of ice-cold 0.1 M Na2CO3, pH 11.5 (high-pH wash) three times, and then icecold hypotonic buffer once to obtain the integral plasma membrane fraction. Protein Extraction from Affinity-Purified Plasma Membrane. Streptavidin beads bearing the integral plasma membrane fraction (∼25-µL bead volume) were suspended in 50 µL of 2× SDS sample buffer containing 100 mM DTT, vortexed at room temperature for 5 min and centrifuged at 10600g for 3 min. The supernatant was collected, and the extraction was repeated once. The supernatants were combined, and proteins were precipitated with TCA/acetone. Typically more than 60 µg of the proteins could be extracted from five dishes (10 cm) of cells by this procedure. The protein pellet was redissolved in 1× SDS sample buffer prior to SDS-PAGE. Electrophoresis and Western Blotting Analysis. The protein sample was heated at 55 °C for 20 min prior to separation in a 4-20% gradient minigel. The gel was stained using Colloidal Coomassie Blue. For immunoblotting analysis, the proteins on the gel were transferred to a PVDF membrane. The membrane was blocked with 5% dry milk in 25 mM Tris pH 7.5, 150 mM NaCl, 0.1% Tween-20 (TBST) for 1 h at room temperature and then incubated with the anti-actin monoclonal antibody (1:10 000 dilution) in the same solution for 1 h at room temperature. After washing with TBST, the membrane was incubated with horseradish peroxidase-conjugated anti-mouse IgG (1:5000 dilution in TBST/5% dry milk) for 1 h at room temperature. After another TBST wash, the blot was developed using the Western Lightning chemiluminescence reagent.
Protein In-Gel Digestion and Nano-HPLC/MS/MS Analysis. Protein in-gel digestion with trypsin was performed as described previously.8 HPLC/MS/MS analyses were performed in an LCQ DECA XP Plus ion trap mass spectrometer (ThermoFinnigan, San Jose, CA) coupled on-line to a nano-HPLC system (Agilent 1100 Nano Pump, Agilent Technologies, San Jose, CA) and nanospray source. Two microliters of the peptide solution in buffer A (5% acetonitrile/94.9% water/0.1% acetic acid (v/v/v)) was manually injected and separated in a nano-HPLC column (50 mm length × 75 µm inner diameter, 5-µm particle size, 100-Å pore diameter) packed in-house with Luna C18 resin (Phenomenex, St. Torrance, CA). The peptides were eluted from the column with a linear gradient of 25-80% buffer B (90% acetonitrile/9.9% water/0.1% acetic acid (v/v/v)) in buffer A over 30 min. The eluted peptides were electrosprayed directly into the LCQ mass spectrometer. The MS/MS spectra were acquired in a data-dependent mode. The four strongest ions in each MS spectrum were automatically selected for fragmentation. The resulting spectra used to identify protein candidates in the NCBI nonredundant protein sequence database with the MASCOT search engine (Matrix Science Ltd., London, U.K.). RESULTS Preparation of Integral Plasma Membrane Fraction. In the BDAP method, preparation of the integral plasma membrane fraction involves the affinity enrichment of the biotinylated, homogenized plasma membrane and removal of cytosolic proteins by washing with high-salt and high-pH buffers (Figure 1). In this paper, we call the sample affinity-enriched from homogenized cells the “plasma membrane fraction” and that after removal of cytosolic proteins the “integral plasma membrane fraction”. Our previous work showed that the plasma membrane fraction affinity-enriched from a biotinylated cell surface contained a significant proportion of cytosolic proteins.8 Since then we have determined that these proteins may be removed by extensive washing using high-salt and high-pH buffers. High-salt buffer has long been used to abolish noncovalent protein-protein associations and was found to wash away a significant proportion of the plasma membrane-associated cytosolic proteins from our plasma membrane fraction. Alkaline buffer is known not only to depolymerize actin bundles but also to disrupt noncovalent proteinprotein interactions while retaining the association of integral plasma membrane proteins with the lipid bilayer.15,16 After being washed with high-salt and high-pH buffers, the highly abundant cytosolic protein actin was almost completely removed from our plasma membrane fraction (Figure 2A). Proteins specific to the mitochondria (NADH-ubiquinol oxydoreductase 39) and endoplasmic reticulum (GRP 94) were not detected by Western blotting analysis when 30 µg of integral plasma membrane protein was resolved in an SDS-PAGE gel, suggesting almost complete removal of these organelles from our preparation (data not shown). In our previously published method, preparation of the plasma membrane fraction involved biotinylation and homogenization of the plasma membrane, preparation of a crude plasma membrane (15) Galkin, V. E.; Orlova, A.; Lukoyanova, N.; Wriggers, W.; Egelman, E. H. J. Cell Biol. 2001, 153, 75-86. (16) Taylor, R. S.; Wu, C. C.; Hays, L. G.; Eng, J. K.; Yates, J. R., 3rd; Howell, K. E. Electrophoresis 2000, 21, 3441-3459.
Analytical Chemistry, Vol. 76, No. 7, April 1, 2004
1819
Figure 2. Affinity-enriched membrane proteins from H1299 cells. (A) Western blotting analysis of actin with anti-actin monoclonal antibody. Lane 1, whole cell lysate; lane 2, plasma membrane fraction without washing with high-salt and high-pH buffers; lane 3, integral plasma membrane after washing with high-salt and high-pH buffers. Five micrograms of protein was loaded in each lane. (B) SDS-PAGE separation of the integral plasma membrane fraction. Thirty micrograms of protein was resolved in a 4-20% gradient minigel and visualized by Colloidal Coomassie Blue staining.
fraction by ultracentrifugation, and affinity enrichment of the biotinylated plasma membrane sheets using streptavidin beads.8 We observed that a precipitate tended to form during affinity enrichment in the resuspended, crude plasma membrane fraction obtained after ultracentrifugation. The protein precipitate would result in contamination of the final preparation by cytosolic proteins despite the high-salt and high-pH washes. To avoid this protein precipitation, the ultracentrifugation step was omitted in our new protocol; the PNS was used directly for affinity enrichment. Identification of Integral Plasma Membrane Proteins by Nano-HPLC/MS/MS. To determine the extent of contamination of the integral plasma membrane fraction by cytosolic proteins, 1820 Analytical Chemistry, Vol. 76, No. 7, April 1, 2004
Figure 3. Typical HPLC/MS/MS analysis. (A) Total ion current (TIC) chromatogram of capillary-HPLC/MS/MS of the tryptic peptides from gel slice 35 (mass range 27-28 kDa in SDS-PAGE); (B) MS spectrum at retention time of 39.76 min; and (C) MS/MS spectrum of peptide VQNATLAVANITNADSAT, unique to erythrocyte band 7 integral membrane protein.
and to comprehensively analyze the integral plasma membrane proteins, the proteins were extracted from the integral plasma membrane fraction and then precipitated. Thirty micrograms of the resulting protein extract was resolved in a 4-20% SDSpolyacrylamide gradient minigel (Figure 2B). Forty-nine gel slices of equal size (∼1.2 mm) in the mass range between 10 and 500 kDa were excised and subjected to in-gel digestion with trypsin. The resulting peptides were extracted from each gel slice and analyzed by nano-HPLC/MS/MS for protein identification. Figure 3 shows HPLC/MS analysis of tryptic peptides from gel slice 35 (see Table 1). The MS/MS spectra were searched against the NCBI nonredundant protein sequence database using the MASCOT algorithm. This analysis resulted in identification of 38 proteins, including 21 proteins containing more than one hydrophobic helix domain based on the Sosui prediction algorithm (http://sosui.proteome.bio.tuat.ac.jp/cgi-bin/sosui.cgi?/ sosui_submit.html). Proteins containing such hydrophobic domains were considered integral membrane proteins.
Table 1. Proteins Identified from Gel Slice 35a protein ID
Gi no.
Integral Plasma Membrane Proteins similar to putative transmembrane protein gi|24308364 Implantation-associated protein gi|14149775 CGI-31 protein gi|7705726 transmembrane 4 superfamily member 1 gi|21265101 stearoyl-coA desaturase gi|2117667 secretory carrier membrane protein 3 gi|16445419 KIAA0152 gene product gi|7661948 similar to thioredoxin domain-containing gi|22209028 CD44 antigen precursor gi|2507241 5′-nucleotidase, ecto gi|4505467 sideroflexin 1 gi|23618867 4F2 heavy chain antigen gi|177216 VAMP-associated protein of 33 kDa gi|20070156 emerin gi|4557553 steroid dehydrogenase homologue gi|7705855 KIAA1181 protein gi|6330243 dehydrogenase/reductase 2 gi|7705905 androgen-regulated short-chain dehydrogenase/reductase 1 gi|20070798 erythrocyte band 7 integral membrane protein gi|114823 unknown gi|16877878 syntaxin 7 gi|20532414 ADP, ATP carrier protein T2 gi|86755 reductase,NADH cytochrome b5 gi|352335 voltage-dependent anion channel 2 gi|4507881 voltage-dependent anion channel 1 gi|4507879 guanine nucleotide-binding protein LIM and senescent cell antigen-like domains 1 Cc3 phosphotidylinositol transfer protein, β repressor of estrogen receptor activity coronin, actin binding protein, 1C hypothetical protein MGC10084 unknown scaffold protein Pbp1 holocytochrome c synthase Simlilar to ribosomal protein S6 unnamed protein product cell division cycle 2 protein isoform 1
Other Proteins gi|11055998 gi|13518026 gi|2618733 gi|6912594 gi|6005854 gi|7656991 gi|24308456 gi|13278939 gi|1916850 gi|4885401 gi|15342049 gi|22760412 gi|4502709
MW (kDa)
peptide no.
TMD no.
31.1 38.0 34.0 21.6 27.5 38.2 32.2 31.7 81.5 63.3 35.5 58.0 27.3 28.9 34.3 36.8 32.9 35.3 31.7 34.9 29.7 32.9 31.2 31.5 30.7
2 2 1 1 1 2 7 4 1 2 4 1 2 6 4 3 3 4 3 3 3 2 3 5 5
5 5 4 4 4 4 3 3 2 2 2 1 1 1 1 1 1 1 1 1 1 0 0 0 0
37.5 37.2 27.0 31.5 33.2 53.2 34.9 41.1 32.3 30.5 28.6 39.9 34.0
1 1 1 1 1 2 2 2 2 2 4 3 3
0 0 0 0 0 0 0 0 0 0 0 0 0
a The protein names, theoretical masses, numbers of peptides identified, NCBI protein accession numbers, and numbers of predicted transmembrane domains (TMD) are listed.
The tryptic peptides from the other 48 gel slices were analyzed in a similar manner. This analysis led to identification of 898 proteins, of which 46% were identified by more than two MS/MS spectra. The following search parameters were used in all MASCOT searches: maximum of one missed trypsin cleavage, a maximum 4-Da error tolerance in the MS, and 0.8-Da MS/MS data. Those protein hits with more than three peptides matched (peptide score >35) were considered as true protein identification. All other hits were manually analyzed to ascertain the accuracy of protein identification. In the manual analysis, the criterion used for a true identification is that the masses of all the major peaks (typically more than 7 peaks) in a MS/MS spectrum had to match those of the theoretically calculated fragment ions. All the redundant proteins were removed in the protein identification list. Summary of the Identified Proteins. Proteins integral to the plasma membrane were identified based on NCBI annotation, the presence of a consensus sequence for protein lipid modifications, or the presence of hydrophobic R-helical domains. Among 898 identified proteins, 781 were annotated and 117 were unclassified. Among the annotated proteins, 526 were integral to the plasma membrane proteins and 118 were cytosolic (Figure 4). Integral Plasma Membrane Proteins. Because hydrophobic R-helical domains are a major feature of integral plasma membrane
proteins, we searched for such domains within the identified proteins using the Sosui prediction algorithm. We found that 417 of the identified proteins contained more than one hydrophobic R-helical domain; all of these proteins were considered to be integral plasma membrane proteins. The existing literature was used to identify integral membrane proteins that were anchored to the lipid bilayer through domains other than an R-helix, as described below. Lipid-Modified Proteins. A second group of integral plasma membrane proteins consisted of lipid-modified proteins, including prenylated, myristoylated, palmitoylated, and glycosylphosphatidylinositol-anchored proteins. The 62 prenylated proteins were predicted by the presence of the C-terminal sequences CAAX, CC, or CXC, which have been shown to be modified by either farnesylation or geranylgeranylation.17 Notable among these proteins were 45 members of the Ras family. To our knowledge, this is the most extensive proteomic analysis of Ras proteins and prenylated proteins. Myristoylated proteins were predicted by the MTR Predictor (http://mendel.imp.univie.ac.at/myristate/SUPL predictor.htm), which is based on the myristoylation consensus sequence as (17) Fu, H. W.; Casey, P. J. Recent Prog. Horm. Res. 1999, 54, 315-342.
Analytical Chemistry, Vol. 76, No. 7, April 1, 2004
1821
Figure 4. Pie diagram of the distribution of the 898 identified proteins. Integral plasma membrane proteins represent 67.3% of the annotated proteins.
described.18 Thirty-four proteins were predicted to be myristoylated. Annotation of palmitoylated proteins was based on the existing literature. A significant number of palmitoylated proteins probably went unrecognized due to their unique modification sequence and limited information in the existing literature. Integral membrane proteins anchored to the bilayer by means other than hydrophobic R-helices (e.g., β-sheet), or the modifications described above might have also been missed in our analysis. Our method is biased against integral plasma membrane proteins that have extreme hydrophobicity, as only 62 proteins containing more than 7 R-helix transmembrane domains were identified. And very few G-protein associated receptors were identified. It is likely that these proteins are prone to precipitation due to their high hydrophobicity. They might be difficult to be extracted due to their unique structures. Cytosolic Proteins. We realized that a significant proportion of the identified proteins were probably not anchored to the bilayer through a transmembrane domain or lipid modification, but rather through strong noncovalent interactions with integral membrane proteins. Three washes with high-pH and high-salt buffers might not be enough to completely remove some abundant cytosolic proteins, such as ribosomal proteins. Alternatively, some proteins might precipitate during the affinity enrichment of the plasma membrane fraction. Precipitated protein particles associated with the beads may be difficult to remove and would contaminate the subsequent protein extract. A significant proportion of the identified cytosolic proteins are abundant proteins in the cells. Unclassified Proteins. A total of 117 proteins were categorized as unclassified due to unavailability of information in the literature about their association with the plasma membrane. DISCUSSION There are a few salient features in the BDAP method for the preparation of an integral plasma membrane fraction. The method is able to remove most of the cytosolic protein using harsh washing conditions. The high-pH buffer depolymerized actin (18) Farazi, T. A.; Waksman, G.; Gordon, J. I. J. Biol. Chem. 2001, 276, 3950139504.
1822
Analytical Chemistry, Vol. 76, No. 7, April 1, 2004
bundles, allowing the actin and its associated proteins to be washed away. The technique allows not only for isolation of integral plasma membrane proteins containing transmembrane domains but also those that anchor to the plasma membrane bilayer through lipid modification. Included in these lipid-modified proteins are prenylated proteins, myristoylated proteins, palmitoylated proteins, and glycosylphosphatidylinositol-anchored proteins. Since the experimental procedure relies on the isolation of plasma membrane sheets and subsequent protein extraction, the method allows not only for the isolation of integral plasma membrane proteins that are biotinylated but also for the isolation of those that remain nonbiotinylated. Removal of cytosolic proteins dramatically increases the representation of the integral plasma membrane proteins in the preparation. In the experiment described here, HPLC/MS/MS analysis of only 30 µg of protein from an integral plasma membrane fraction identified 898 proteins. Sixty-seven percent of the annotated proteins were integral plasma membrane proteins. The method presented here is an improvement over our previous method,8 which resulted in a preparation in which only 16% of identified, annotated proteins were integral plasma membrane proteins. In that experiment, the plasma membrane fraction was enriched by a method similar to that described here, but without the high-salt and high-pH washes. It should be noted that, in both experiments, only 30 µg of protein was used. The integral plasma membrane fraction prepared by the procedure described here would provide an excellent sample for various proteomic methods, such as ICAT/mass spectrometry, 1D or 2D-gel/mass spectrometry, or the shotgun approach, allowing highly efficient profiling and quantification of integral plasma membrane proteins. In summary, we report a powerful method for the proteomic analysis of integral plasma membrane proteins, involving affinity purification of a biotinylated plasma membrane, removal of peripheral cytosolic proteins by washing with high-salt and highpH buffers, SDS-PAGE separation, and protein identification by nano-HPLC/MS. Western blotting analysis showed that almost all actin was removed by the harsh washing conditions. HPLC/ MS/MS analysis of the integral plasma membrane fraction
resulted in the identification of 898 proteins, including 781 annotated and 117 unclassified proteins. Among the annotated proteins, at least 67.3% were integral to the plasma membrane, 17.6% were plasma membrane-associated, and 15.1% were cytosolic, suggesting limited contamination of the fraction by cytosolic proteins. Notable among the integral membrane proteins were 62 prenylated proteins and 45 members of the Ras family. Given the importance of integral membrane proteins for therapeutic drug design, the described protocol should expedite the discovery of protein drug targets. Abbreviations: 2D, two-dimensional; DTT, dithiothreitol; ER, endoplasmic reticulum; GPCR, G-protein coupled receptor; ICAT, isotope-coded affinity tag; EZ-Link Sulfo-NHS-SS-Biotin, sulfosuccinimidyl-2-(biotinamido) ethyl-1,3-dithiopropionate; MS, mass
spectrometry; PNS, postnuclear supernatant; TMD, transmembrane domain; DMSO, dimethyl sulfoxide. ACKNOWLEDGMENT Y.-m.Z. is supported by The Robert A. Welch Foundation (I-1550) and NIH (CA 85146). SUPPORTING INFORMATION AVAILABLE Table of protein IDs from 49 gel bands. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review November 26, 2003. Accepted February 9, 2004. AC0354037
Analytical Chemistry, Vol. 76, No. 7, April 1, 2004
1823
