Anal. Chem. 2010, 82, 755–761
Affinity-Trap Polyacrylamide Gel Electrophoresis: A Novel Method of Capturing Specific Proteins by Electro-Transfer Chihiro Awada,† Takashi Sato,‡ and Toshifumi Takao*,† Laboratory of Protein Profiling and Functional Proteomics, Institute for Protein Research, Osaka University, Yamadaoka 3-2, Suita, Osaka 565-0871, Japan, and Department of Applied Biochemistry and Food Science, Saga University, 1-Honjo, Saga 840-8502, Japan A method for the affinity capture of specific proteins from a complex mixture using a polyacrylamide gel technique is described. The approach is based on the orthogonal electro-transfer of proteins separated by ordinary polyacrylamide gel electrophoresis (PAGE) to a ligand-coupled polyacrylamide gel (Li-PAG), which is placed under the PAGE gel. Upon electro-transfer, the proteins orthogonally migrate from the PAGE into the Li-PAG, based on the net charge. During migration to the Li-PAG, proteins that specifically interact with a ligand can be transiently trapped in the Li-PAG, while those that do not interact with a ligand pass through it. This method permits the separation of the proteins that can specifically interact with a ligand, even when present in a complex mixture. The method is demonstrated by applying it to the onestep isolation of a trypsin inhibitor from a crude extract of soybean flour. Polyacrylamide gel electrophoresis (PAGE), a widely used technique, is nearly universally used for the separation and detection of proteins. Two-dimensional PAGE, in conjunction with mass spectrometry, triggered the proteomics era1 and allowed for the facile separation of proteins as function of size, charge, etc. Moreover, the use of additives to the polyacrylamide gel, such as sodium dodecyl sulfate (SDS), urea, pigments, proteins, sugars, metal ions, etc., permit such separations to be enhanced or modulated. The electrophoretic mobility of an eluate is in accordance with the net charge, molecular size, and molecular form as well as the permeability of the polyacrylamide gel. The additives associate or interact with the eluate molecules to a greater extent during electrophoresis and, as a result, modulate the physical properties of the eluate and shift its electrophoretic mobility. Taking the advantage of this mobility shift in electrophoresis, a wide variety of separation or detection techniques, commonly known as affinity electrophoresis,2-5 have been developed. Among * Corresponding author. E-mail:
[email protected]. Phone/Fax: +8166879-4312. † Osaka University. ‡ Saga University. (1) Shevchenko, A.; Wilm, M.; Vorm, O.; Mann, M. Anal. Chem. 1996, 68, 850–858. (2) Horejsı´, V. Anal. Biochem. 1981, 112, 1–8. (3) Cann, J. R. Anal. Biochem. 1996, 237, 1–16. (4) Heegaard, N. H. Electrophoresis 2009, 30, S229–239. 10.1021/ac902290q 2010 American Chemical Society Published on Web 12/28/2009
them, lectin-affinity electrophoresis,6 which is based on lectincarbohydrate interactions, has been widely used not only in biology and the biotechnology fields but also in diagnostics. The lectin, which has an affinity for a specific carbohydrate structure, is immobilized in a gel or preincubated with a sample. Upon electrophoresis, the migration of the sample bands with an affinity for the lectin is retarded relative to those observed in the gel without the lectin. Thus, the interacting molecules can be readily detected as shifted bands. The method allows for the rapid assessment of the target carbohydrate or its conjugate in the sample. One of the most successful applications of lectin-affinity electrophoresis is the AFP (R-fetoprotein)-L3 assay,7 which is a useful prognostic factor in patients with hepatocellular carcinomas. AFP-L3, a fraction containing one of the glycoforms of AFP separated by the lectin Lens culinaris agglutinin (LCA), is characterized as a glycoform with additional R1-6 fucose residues at the reducing termini of the carbohydrate moieties, which is formed by activation of fucosyltransferase-88,9 that is uniquely associated with hepatocellular carcinoma. The above method is based on an in situ interaction between analytes and an additive during electrophoresis. On the other hand, Englert et al.10 recently reported on a method referred to as “layered expression scanning”, which is based on the out-ofgel capture and detection of specific proteins. This method is based on the following rationale. Analyte proteins or mRNA in an electrophoresis gel or tissue section are transferred to a layered array of capture membranes, which are coupled with antibodies or DNA oligomers. Species that interact with the antibodies or DNAs embedded in the individual membranes are specifically captured and detected on the membranes while those without any affinities pass directly through it. However, in general, the membrane itself has a low permeability and obviously displays nonspecific interactions with proteins, especially, in the case of a complex protein mixture. We report here on a new approach for the affinity capture of specific proteins within a polyacrylamide gel which have previously been separated by PAGE or a related technique. The method (5) (6) (7) (8) (9) (10)
Takeo, K. J. Chromatogr., A 1995, 698, 89–105. Bøg-Hansen, T. C. Scand. J. Immunol. 1983, 17, 243–253. Shimizu, K.; et al. Clin. Chim. Acta 1996, 254, 23–40. Uozumi, N.; et al. J. Biol. Chem. 1996, 271, 27810–27817. Noda, K.; et al. Cancer Res. 2003, 63, 6282–6289. Englert, C. R.; Baibakov, G. V.; Emmert-Buck, M. R. Cancer Res. 2000, 60, 1526–1530.
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
755
is based on the electro-transfer of proteins from a PAGE gel to a ligand-coupled polyacrylamide gel (Li-PAG), followed by the in situ interaction of the proteins with the ligand. The proteins that have an affinity for the ligand are retained in the Li-PAG while those without any affinity pass through it. The Li-PAG displays negligible nonspecific adsorption of proteins in comparison with a conventional ligand-coupled affinity particle11,12 or affinity membrane.10,13 Moreover, the present method utilizes, for the first time, the orthogonal capture of proteins within a polyacrylamide gel by electro-transfer. It thereby enables the rapid screening of target proteins from protein bands or spots obtained by one- or two-dimensional PAGE, which could then be readily subjected to ordinary proteomic analyses. In the context of various methodologies for capturing interacting proteins that are currently under consideration, this capability is especially useful, since it allows for the one-step isolation of the interacting proteins from a biological sample. In this study, the performance and applicability of the method was evaluated using trypsin (as a ligand) and soybean flour (as a biological sample), which contains the protein (trypsin inhibitor) that specifically interacts with trypsin. MATERIALS AND METHODS Materials. Trypsin from porcine pancreas, ribonuclease A from bovine pancrease, soybean flour, leupeptin, o-dianisidine bis(diazotized) zinc double salt (Fast Blue B salt), NR-benzoyl-Larginine 4-nitroanilide hydrochloride (L-BAPNA), Kunitz trypsin inhibitor (KTI) (Glycine max), R-cyano-4-hydroxy cinnamic acid (CHCA), and N-acetyl-DL-phenylalanine β-naphthyl ester were obtained from Sigma (St. Louis, MO). Acrylamide and N,N′methylenebisacrylamide (Bis) were purchased from Wako Pure Chemical Industries (Osaka, Japan). N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP) was from Dojin Kagaku (Kumamoto, Japan). An ultrafiltration membrane filter (Amicon Ultra-4, 10-kDa MW-cut) and PVDF membrane (Immobilon-P, Immobilon -FL) were purchased from Millipore (Billerica, MA). PD-10 columns were obtained from GE Healthcare (Buckinghamshire, UK). HF Bond Elut-C18 columns were purchased from Varian (Walnut Creek, CA). Tris [2-carboxyethyl] phosphine hydrochloride (TCEP) was purchased from Promega (WI, Madison). The dialysis membrane (Spectra/Por3 Membrane, 3.5-kDa MW-cut) was from Spectrum Laboratories, Inc. (Rancho Dominguez, CA). All other reagents were of analytical grade and were obtained from Nacalai Tesque (Kyoto, Japan). The amounts of the proteins (trypsin, KTI, RNase A) were estimated from the absorbance at 280 nm measured by a UV/vis spectrophotometer (Ultrospec 1100 pro, GE-Healthcare), whose molar extinction coefficients (ε280) were given as 34 670, 17 210, and 9440 M-1cm-1, respectively, by the method previously described.14 Extraction of a Crude Trypsin Inhibitor. The preparation of a crude soybean trypsin inhibitor, the Kunitz trypsin inhibitor (KTI), was carried out as described by Shibata et al.15 Soybean (11) Urh, M.; Simpson, D.; Zhao, K. Methods Enzymol. Burgess, R. R.; Deutscher, M. P., Eds.; Academic Press: New York, 2009, Vol. 463, pp 417-438. (12) Roque, A. C.; Lowe, C. R. Methods Mol. Biol. Zachariou, M., Ed.; Humana Press: Totowa, 2008, Vol. 421, pp 1-21. (13) Boi, C. J. Chromatogr., B 2007, 848, 19–27. (14) Pace, C. N.; Vajdos, F.; Fee, L.; Grimsley, G.; Gray, T. Protein Sci. 1995, 4, 2411–2423.
756
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
flour (10 g) was suspended in 100 mL of 100 mM phosphate buffer (pH 7.0) and stirred for 12 h at 4 °C. The seed debris was removed by centrifugation (19 500g for 15 min at 4 °C). The total protein concentration of the supernatant (81 mL) was estimated using the Micro BCA Protein Assay Reagent Kit (Pierce, Rockford, IL). An aliquot (equivalent to 126 µg protein) of the supernatant was dissolved in 10 µL of 50 mM Tris-HCl buffer (pH 6.8) containing 0.1% SDS, 0.1% bromophenol blue, and 10% glycerol, heated at 95 °C for 5 min, and applied to SDS-PAGE (Figure 4). Preparation of Ligand-Coupled Polyacrylamide Gel. Trypsin (76.7 nmol, 1.8 mg) was first incubated at room temperature with a 4.7-fold excess of leupeptin (359 nmol, 166 µg), a potent protease inhibitor, in 100 mM phosphate buffer (pH 7.5) containing 100 mM NaCl. The resultant trypsin/leupeptin conjugate was reacted with SPDP (429 nmol, 134 µg) at room temperature for 30 min. The reaction mixture was subjected to a PD-10 column to remove the excess reagent. The 3-(2-pyridyldithio)propionyl trypsin/leupeptin, eluted with 20 mM 2-(Nmorpholino)ethanesulfonic acid (MES) buffer (pH 6.0) containing 1 mM EDTA, was incubated with 1 mM TCEP for 10 min and subjected to PD-10 to remove the excess reagent. The 3-mercaptopropionyl trypsin/leupeptin, eluted with 20 mM MES buffer, was reacted with a 2500-fold excess of Bis in 100 mM Tris -HCl buffer (pH 8.2). The resulting solution was concentrated 10-fold with Amicon Ultra-4 (10-kDa MW-cut, Millipore) and subjected to a PD-10 column to remove the excess Bis. The stoichiometry of Bis linked to trypsin was examined by electrospray ionization mass spectrometry (ESI-MS) (see Figure 2B). The Bis-modified trypsin (9.27 nmol) in MES buffer was mixed with 6 mL of 7.5% acrylamide/Bis (29:1) in 375 mM Tris-HCl buffer (pH 8.8). As a control, intact trypsin (9.27 nmol) in 1 mM HCl was mixed with 6 mL of 7.5% acrylamide/Bis solution (29:1) containing 2 mM CaCl2. Then, into these mixtures, 24 µL of 10% ammonium persulfate and 3.6 µL of N,N,N′,N′-tetramethylenediamine were added to initiate free radical copolymerization. The trypsin/ leupeptin-coupled polyacrylamide gel, thus obtained, could be stored at 4 °C for a week without loss of enzyme activity. The leupeptin and unreacted trypsin were removed prior to use by exhaustive washing with 25 mM Tris-192 mM glycine buffer (pH 8.4) containing 0.1% SDS in a Mini-Trans-Blot electrophoretic transfer cell (Bio-Rad, Richmond, CA) at a constant current of 100 mA for 1 h at 4 °C. Then, the trypsin-coupled polyacrylamide gel (TP-gel) was washed with water to remove SDS. The final concentration (1.52 pmol per 1 mm3 block of the gel) of trypsin coupled to the TP-gel was estimated on the basis of the plot of intensity of a stained gel (gel densitometry) vs concentration of protein (Yint ) 32.18Xconc + 54.92, R2 ) 0.997), which was obtained over the range 0.77 µM < X < 1.92 µM (trypsin). The coupling yield of trypsin to polyacrylamide gel, as a result, was calculated to be 98.7%, based on the initial concentration (1.54 µM) of Bis-modified trypsin. The RNase A-coupled polyacrylamide gel (RNase-gel) was prepared similarly to the above, which, as a result, contained 1.51 pmol of RNase A per 1 mm3 block of the gel. (15) Shibata, H.; Hara, S.; Ikenaka, T.; Abe, J. J. Biochem. 1986, 99, 1147– 1155.
Figure 1. General scheme of affinity-trap polyacrylamide gel electrophoresis. (A) SDS or native PAGE, (B) electro-transfer of proteins from a sample gel to a Li-PAG gel, and (C) electro-blotting of proteins trapped in a Li-PAG gel onto a PVDF membrane. The Li-PAG was prepared as described in Materials and Methods.
Polyacrylamide Gel Electrophoresis (PAGE). PAGE was performed as described by Laemmli.16 Protein samples were dissolved in 50 mM Tris-HCl buffer (pH 6.8) containing 1% SDS, 0.1% bromophenol blue, and 20% glycerol and applied to 12.5% or 15% acrylamide gel (9 cm × 6 cm × 0.1 cm) with a running buffer (25 mM Tris-192 mM glycine buffer (pH 8.4) containing 0.1% SDS). PAGE was carried out at a constant current of 20 mA for 86 min (Figure 3) or 45 min (Figure 4). Electro-transfer and Electro-blotting. After PAGE, the unstained sample gel was washed with water to remove SDS and mounted on a Mini-Trans-Blot electrophoretic transfer cell (BioRad). The TP-gel was placed in between the sample gel and an Immobilon-P PVDF membrane (Figure 1B). Proteins in the sample gel were electro-transferred into the TP-gel or RNase-gel with 25 mM Tris-192 mM glycine buffer (pH 8.4) at a constant current of 100 mA for 300 min (Figure 3) or at 50 mA for 30 or 210 min (Figure 4) at 4 °C. The resulting gels and PVDF membranes were stained with Coomassie Brilliant Blue G-250 (CBB); the TP- and RNase-gels were subjected to trypsin activity staining (see below) (Figures 3 and 4). For protein identification, the TP-gel was removed from the transfer cell, washed with water, and equilibrated with 100 mM CAPS buffer (pH 11). The TP-gel was then heated at 90 °C for 10 min in the same buffer and placed on an Immobilon-FL PVDF membrane (Millipore) without any staining. The trapped proteins in the TP-gel were then electro-blotted using the above cell onto the membrane with 10 mM CAPS buffer (pH 11) containing 0.1% (16) Laemmli, U. K. Nature 1970, 227, 680–685.
SDS at a constant current of 200 mA for 1 h at room temperature. The resulting membrane was stained with SYPRO Ruby protein blot (Bio-Rad). A strip of the membrane (protein band) was incubated at room temperature for 1 h in 0.5 M Tris-HCl buffer (pH 8.5) containing 6 M guanidine hydrochloride, 0.3% EDTA, 4% acetonitrile, and 1 mg dithiothreitol and then treated with iodoacetamide (3.0 mg in 10 µL of 1 M NaOH) for 20 min in the dark. The resulting membrane was washed with water and aqueous 2% acetonitrile, then treated with trypsin (1 pmol) in 50 µL of 20 mM Tris-HCl buffer (pH 8.2)/aqueous 70% acetonitrile at 37 °C overnight. The supernatant was diluted with 450 µL of 0.1% aqueous TFA and applied to a HF Bond Elut-C18 column (Varian) for desalting. The eluate with 0.1% TFA/aqueous 60% acetonitrile was concentrated under vacuum and applied to nanoflow liquid chromatography (LC)/matrix-assisted laser desorption/ionization (MALDI)-MS (see below). Enzyme Activity Assay. Trypsin activity was assayed by the following two methods. (1) For the activity-staining method,17 the TP-gel was preincubated in 50 mL of 0.1 M sodium phosphate buffer (pH 7.8) for 15 min at 37 °C; it was then incubated in a solution (60 mL) containing N-Ac-DL-Phe-β-naphthyl ester (10 mL, 2.5 mg/mL in N,N-dimethylformamide) and Fast Blue B salt (50 mL, 1 mg/mL in 50 mM phosphate buffer, pH 7.8) that was conjugated with β-naphthol and exhibited a purple color, for 5 min at 25 °C. (2) The enzyme activities of the TP-gel and in-gel intact trypsin (see above) were measured as described by Erlanger et al.18 with minor modifications. Both gels were preincubated in 50 mM Tris-HCl buffer (pH8.0) containing 20 mM CaCl2 at 37 °C for 5 min. A piece of the gel (0.5 cm3), excised from each gel, was placed in 0.2 mM L-BAPNA solution (5 mL of 50 mM Tris-HCl, 20 mM CaCl2, pH 8.0) and incubated at 37 °C. Aliquots of the reaction mixture were withdrawn at 8, 16, 24, 32, 40, and 48 min; the absorbance of the liberated p-nitroaniline (ε ) 8800 M-1cm-1) at 410 nm was measured using a UV/vis spectrophotometer (Ultrospec 1100 pro, GE-Healthcare). The specific activities (mol/min per mg) were determined by the following equation: specific activity ) (∆Abs/min (Figure S-1) × volume)/(ε × mg of protein) RNase activity was assayed as described by Wolf et al.19 with minor modifications. The RNase-gel was incubated with 1 mg/ mL of torula yeast RNA (Sigma) in 100 mM Tris-HCl buffer (pH 7.0) at 25 °C for 2 h. The cleaved RNA was removed by washing with 100 mM Tris-HCl buffer (pH 7.0). The unreacted long-chain RNA in the gel was stained with 0.2% toluidine blue O in 10 mM Tris-HCl buffer (pH 7.0). Nano-Flow Liquid Chromatography and Mass Spectrometry. The trypsin digest was injected into an Ultimate nano-LC system (Dionex, Idstein, Germany), where the peptides were first concentrated with a C18 trapping column (0.3 mm × 1 mm, Dionex) at a flow rate of 30 µL/min and then separated using a C18-Pepmap column (0.075 × 150 mm, Dionex). A linear gradient of solvent A (0.1% trifluoroacetic acid in (17) Uriel, J.; Berges, J. Nature 1968, 218, 578–580. (18) Erlanger, B. F.; Kokowsky, N.; Cohen, W. Arch. Biochem. Biophys. 1961, 95, 271–278. (19) Wolf, G. Experientia 1968, 24, 890–891.
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
757
Figure 2. Preparation of a trypsin-coupled polyacrylamide gel (TP-gel) (A) and ESI mass spectra of an intact trypsin (upper) and the Bismodified leupeptin-free trypsin (lower) (B). The peaks marked as Val and Ile in the spectrum were assigned to variants of porcine trypsin (Val12 and Ile12). The theoretical masses of Bis-modified trypsin are Val12: 23 691.9 (1 × m), 23 934.2 (2 × m), and 24 176.5 (3 × m); Ile12: 23 706.0 (1 × m), 23 948.3 (2 × m), and 24 190.6 (3 × m).
water) and solvent B (0.1% trifluoroacetic acid in acetonitrile) was used for the separation, and the peptides were eluted by increasing the concentration of solvent B from 5% to 80% over a period of 60 min at a flow rate of 200 nL/min. The effluent was directly blotted at 1 min intervals onto the flat surface of a stainless steel plate (a MALDI sample plate) over a 96 min period. Thereafter, the matrix solution, the supernatant from 0.1% TFA/aqueous 50% acetonitrile saturated with CHCA, was blotted manually onto each sample spot and then dried. Overall, peptide identification was carried out using a MALDITOF/TOF (4700 proteomics analyzer, Applied Biosystems, Framingham, MA), followed by a database search with Mascot version 2.0 (Matrix Science, Manchester, UK). Ions were generated by irradiating the sample area with a 200 Hz Nd:YAG laser operated at 355 nm. ESI-MS was performed on a Q-TOF II mass spectrometer (Micromass, Manchester, UK) in the positive ion mode. The capillary voltage and a cone voltage were operated at 1.5 kV and 35 V, respectively. A nitrogen counter gas temperature was adjusted to 80 °C. The sample solution was introduced into a glass capillary (Proxeon, Odense, Denmark). All data were acquired and analyzed with the MassLynx software (Micromass). 758
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
RESULTS AND DISCUSSION Principle of Affinity-Trap Polyacrylamide Gel Electrophoresis. The method was developed to permit specific proteins to be captured from a complex mixture in a polyacrylamide gel matrix. It is based on the in-gel interaction of proteins with a ligand, which is covalently coupled to polyacrylamide gel, and their trapping within the gel during the electro-transfer process. It can be efficiently combined with ordinary PAGE such as one- or two-dimensional or native PAGE. The sample gel could be placed on a ligand-coupled polyacrylamide gel (Li-PAG) and the proteins electro-transferred from the sample gel to the Li-PAG. As a result, proteins that have an affinity for the ligand are retained in the Li-PAG while those without any affinity pass through it. The proteins specifically trapped in the LiPAG can be eluted by electro-blotting onto the membrane and identified by a conventional proteomics technique (Figure 1). The method has the advantage of the efficient recovery of specific proteins interacting with a ligand from a complex mixture which has previously been separated by PAGE. Coupling of a Ligand to Polyacrylamide Gel. In order to prepare a ligand-coupled polyacrylamide gel, (1) a bifunctional linker, N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP), was reacted with trypsin (ligand) at a 5.6:1 molar ratio. Of note,
Figure 3. Affinity-trap polyacrylamide gel electrophoresis of Kunitz trypsin inhibitor (KTI). KTI (lane a: 33.6, b: 16.8, c: 8.4, d: 4.2, e: 2.1 pmol) were applied to nondenaturing SDS-PAGE (12.5% polyacrylamide gel); the two pieces of the sample gels were placed in the electrophoretic transfer cell as indicated in (A). The resulting 1st TP-gel, the RNase-gel, and the 2nd TP-gel were then subjected to trypsin activity staining (B).
leupeptin, which reversibly inhibits trypsin activity,20 was conjugated with trypsin throughout preparation of a trypsin-coupled polyacrylamide gel (TP-gel) to avoid autolysis; (2) the resulting 3-(2-pyridyldithio)propionyl trypsin was reduced with TCEP to give 3-mercapto-propionyl trypsin. This treatment turned out not to perturb the intramolecular disulfide bonds from the fact that the observed molecular masses of Bis-modified trypsin agreed well with the theoretical ones (lower spectrum in Figure 2B). (3) The emerging free thiols of the linker moieties readily reacted with either of the two acryl groups of Bis via 1,4-addition (Figure 2A) under the conditions described in Materials and Methods. ESIMS of the product clearly showed that one to three Bis moieties were predominantly incorporated into a trypsin molecule (Figure 2B); (4) the Bis-modified trypsin was copolymerized using the standard acrylamide/Bis preparation protocol to produce a TPgel (see Materials and Methods). Assessment of a Ligand Coupled to a Polyacrylamide Gel. In order to assess whether or not a protein ligand (trypsin) immobilized to polyacrylamide gel retains its native form, trypsin activity was assayed. The modified trypsin in the TP-gel as well as the in-gel intact trypsin were incubated with the substrate (see Materials and Methods). The absorption of the product was plotted for both preparations (Figure S-1 in the Supporting Information). As a result, modified and intact trypsin showed a similar specific activity (2.57 µmol/min per mg vs 2.70 µmol/min per mg, respectively), indicating that the modification made for the covalent coupling of trypsin to the polyacrylamide gel had no effect on the specific activity, indicating that the modified trypsin retained its native conformation in the TP-gel. It should be noted that such chemical modification of a ligand such as a protein for coupling to a gel matrix might cause a significant change in the
conformational or chemical properties of the molecule. Since the surface of a molecule, which could, most probably, be the modification site, might be involved in interactions with other molecule(s), it is important to examine several ligand-coupling methods or conditions with respect to the interacting potential of a ligand, especially, in the case of a survey of unknown proteins. In-gel Trapping of an Interacting Protein. The trapping capability of a ligand-coupled polyacrylamide gel (the TP-gel in this experiment) was assessed by the experiment illustrated in Figure 3. Authentic soybean KTI, a protein that can form a stable conjugate with trypsin,21 was subjected to standard SDS-PAGE and then electro-transferred into the TP-gel in parallel with the RNase-gel. In order to check the specific trapping of interacting proteins, second TP-gels were placed under both gels (Figure 3A). For visualization of the trapped protein (KTI), the TP-gels were stained, in the presence of a substrate (N-Ac-DL-Phe-β-napthyl ester), with Fast Blue B salt (see Materials and Methods). The trapping capacity of the TP-gel could be estimated by the amount of trypsin coupled to the gel: in this experiment, based on the amount (1.52 pmol/mm3) of trypsin coupled (see Materials and Methods), lane c (6 mm wide × 1.5 mm high × 1 mm thick) in the left panel of Figure 3B was calculated to contain 13.7 pmol of trypsin. Of note, when a larger amount of a ligand is coupled to the gel, a correspondingly larger amount of protein could be trapped, but it becomes difficult to visualize the trapped protein band owing to an increase in the background. Taking into consideration that the present method is intended for the identification of trapped proteins, the amount of a protein obtained by the above trapping capacity was sufficient for characterization by conventional MS-based protein identification methods. In the case of lane a (6 mm wide × 3 mm high × 1 mm thick) in the left
(20) Kuramochi, H.; Nakata, H.; Ishii, S. J. Biochem. 1979, 86, 1403–1410.
(21) Blow, D. M.; Janin, J.; Sweet, R. M. Nature 1974, 249, 54–57.
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
759
Figure 4. One-step isolation of KTI from a crude extract of soybean flour. SDS-PAGE (15% polyacrylamide gel) of the crude extract (134 µg protein) (A); affinity-trap polyacrylamide gel electrophoresis (electro-transfer time: 30 min (left panel) and 210 min (right panel)) of the sample gel obtained in A (B); electro-blotting of the proteins trapped in the TP-gel (b′1) to a PVDF membrane (C); amino acid sequence of soybean KTI and the tryptic peptides (solid bars) identified by nano-flow LC/MALDI MS/MS (D). The gels in A, a, a′, b1, and b′1, and PVDF membranes in c and c′ were stained with CBB; the TP-gels in b2 and b′2 were subjected to trypsin activity staining; the PVDF membrane in C was stained with SYPRO Ruby protein blot (Bio-Rad). The arrowheads indicate the protein band that was identified with KTI.
panel of Figure 3B, the amount of trypsin within the protein band was calculated to be 27.4 pmol, which was less than the amount of KTI (33.6 pmol) applied to the gel; the amounts of KTI applied to lanes b-e were less than those of trypsin embedded in their respective bands (left panel of Figure 3B). Thus, an excess amount of KTI in lane a was found to pass through the first TP-gel and then become trapped in the second TP-gel. Meanwhile, the same amounts of KTI, which has no affinity for RNase at all, completely passed through the RNase-gel and was detected in the second TP-gel (right panel of Figure 3B), which was almost the same as that obtained in the first TP-gel. The result demonstrates that the present method clearly permitted the affinity capture of a specific protein in a polyacrylamide gel. Isolation of Kunitz Trypsin Inhibitor (KTI) from Soybean Flour. In order to assess the applicability of the method for capturing a specific protein from a complex mixture, we attempted the one-step isolation of KTI from a crude extract of soybean flour (see Materials and Methods). A crude extract (126 µg) was directly applied to conventional SDS-PAGE under nondenaturing 760
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
conditions. The sample gel (Figure 4A) was washed with water to remove SDS, placed upon the TP-gel with a PVDF membrane at the bottom (Figure 4B), and then electro-transferred to the TPgel (transfer time: 30 min (a-c); 210 min (a′-c′) in Figure 4B). Afterward, the original gel, the TP-gel, and the PVDF membrane were stained with CBB, and another TP-gel was subjected to activity staining (see Materials and Methods). The transfer efficiency was in accordance with electro-transfer time and the molecular size or net charge of a protein as well as the acrylamide concentration: the longer transfer time allowed more proteins to migrate through the TP-gel onto the PVDF membrane (see Figure 4A, B-c, B-c′). The high-molecular-weight proteins had a propensity to be resistant to transfer from the original gel (see Figure 4A, B-a, B-a′), which was also correlated with the acrylamide concentration. The resulting TP-gel, obtained by electro-transfer for 210 min, was likely to specifically capture a single protein (Figure 4B-b′1), which exhibited inhibitory activity against trypsin (Figure 4B-b′2). Based on the trapping capacity (1.52 pmol (trypsin)/mm3) of the TP-gel and the supposition
that the captured protein interacts with trypsin on a one to one basis, the maximal amount of the captured protein was estimated to be 32.8 pmol on the basis of the corresponding band volume (ca. 21.6 mm3, Figure 4B-b′2). Taking into account the KTI content (28-32 mg/g) reported for soybean flour,22 the amount of KTI in Figure 4A was estimated to be 287-328 pmol; the overcharged KTI, thus, passed through the TP-gel and was accommodated on the PVDF membrane (arrowhead in Figure 4B-c′). In order to identify the protein trapped in the TP-gel, the resulting whole TP-gel was subjected to electro-blotting. The major protein band at ca. 20.1 kDa, blotted onto the PVDF membrane (Figure 4C), was reduced, alkylated, and then digested with trypsin in situ (Materials and Methods). MALDI-MS/MS of the digested peptides readily revealed the soybean KTI sequence with a sequence coverage of 72.9% (Figure 4D). It should be noted that the protein band obtained in the TP-gel (Figure 4B-b′1) could be directly applied for identification using a protocol similar to the above. However, significant amounts of peptides derived from trypsin embedded in the TP-gel were simultaneously detected in the MS spectrum, which obviously interfered with the identification of the target protein to a considerable extent (data not shown). CONCLUSIONS The present study demonstrates a method for the affinity capture of specific proteins from a complex mixture using a ligand (trypsin in this study)-coupled polyacrylamide gel (Li-PAG). The method is based on ordinary PAGE-based separation followed by (22) Rackis, J. J.; Gumbmann, M. R.; Liener, I. E. Qual. Plant Foods Hum. Nutr. 1985, 35, 213–242.
orthogonal electro-transfer of the separated proteins to the LiPAG, which can be readily prepared by linking a sulfhydrylmodified ligand to Bis via 1,4-addition. Of note, in cases where a ligand contains a sulfhydryl group within the structure, it can be directly coupled with Bis, which, however, might change its functional or structural properties. When a sample protein has a basic pI, it migrates oppositely to the cathode under the electrotransfer conditions used in the present study (see Materials and Methods). However, it could be applied to the same electrotransfer conditions and protocol but with the polarity being reversed. Polyacrylamide gel is one of the most widely used matrixes for separating proteins as a function of molecular size, net charge, etc. This is in large part due to the fact that it displays high permeability with negligible nonspecific interaction with analytes (Figure S-2 in the Supporting Information). This point has been key in developing the method described herein, making it possible to achieve in situ affinity capture of specific proteins from a complex mixture such as a crude biological sample, based on the interaction between an analyte and a ligand coupled to a polyacrylamide gel. This capability could be applied to a ligand-based global survey of interacting proteins in a complex mixture. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.
Received for review December 7, 2009.
October
10,
2009.
Accepted
AC902290Q
Analytical Chemistry, Vol. 82, No. 2, January 15, 2010
761