Anal. Chem. 1999, 71, 642-647
Characterization of Proteins from Human Cerebrospinal Fluid by a Combination of Preparative Two-Dimensional Liquid-Phase Electrophoresis and Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry Pia Davidsson,* Ann Westman, Maja Puchades, Carol L. Nilsson, and Kaj Blennow
Department of Clinical Neuroscience, Unit of Neurochemistry, Go¨ teborg University Sahlgrenska University Hospital, Mo¨ lndal, Sweden
To purify and characterize low-abundance proteins in complex biological mixtures, we used a novel strategy that combined preparative two-dimensional liquid-phase electrophoresis (2D-LPE) and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDITOF MS). Preparative 2D-LPE is based on the same isoelectric focusing and gel electrophoresis principles as the widely used analytical 2D gel electrophoresis, except that analytes remain in solution throughout separation. This novel approach shows many improvements compared to analytical 2D gel electrophoresis for the separation of proteins in biological fluids. For example, larger volumes/amounts of samples can be loaded, yielding sufficient amounts of low-abundance proteins for further characterization. Since proteins remain in liquid phase during the entire procedure, extra steps such as electroelution, extraction, or transfer to membranes from the gels prior to mass spectrometric analysis are obviated. We report the usefulness of 2D-LPE combined with MALDITOF MS for the purification and characterization of cystatin C and β-2 microglobulin in human cerebrospinal fluid. This method should be applicable to a wide range of biological fluids, such as cerebrospinal fluid, serum, tissue extracts, cell media, whole cells, and bacterial lysates. Electrophoresis and mass spectrometry are powerful methods that can provide purification and characterization of proteins and peptides in complex biological samples. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) is a widely used approach for separating proteins from complex mixtures.1 It can be performed in a one- or twodimensional (2D) configuration. For less complicated protein preparation, one-dimensional SDS-PAGE is preferred over 2D gels, because it is simpler. However, SDS-PAGE often results in migrating or overlapping protein bands due to its limited resolving * Corresponding author: (tel) +46 31 862415; (fax) +46 31 862421; (e-mail)
[email protected]. (1) Patterson, S. D.; Aebersold, R. Electrophoresis 1995, 16, 1791.
642 Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
power. What appears to be a single band may actually be a mixture of different proteins. 2D gel electrophoresis incorporates isoelectric focusing (IEF) in the first dimension and SDS-PAGE in the second dimension, leading to a separation by charge and size.2 2D PAGE is a powerful technique for separating very complex protein preparations, resolving up to 10 000 proteins from mammalian tissues and other complex proteins.3-5 Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), introduced in 1987 by Karas et al.6 is now a widely employed technique for mass spectrometric analysis of biological samples, because of its precise (better than 0.1%) mass determination of proteins, high mass range (up to at least 1 000 000 u), high sensitivity, and relative tolerance to common buffer components. Sequence information is especially important for the identification of proteins and may be obtained by MALDI mass spectrometry analysis, together with the use of enzymatic digestion followed by postsource decay (PSD) of the resulting peptides, and database searching.7 Different methods have been developed to characterize SDSPAGE-separated proteins by MALDI-MS.1 The combination of SDS-PAGE with mass spectrometry has some difficulties, because it involves many steps. To determine the mass of the whole proteins, they can be electroeluted or extracted from SDS-PAGE gels.8,9 Proteolytic digestion of the proteins prior to MALDI-MS is widely used, either directly in the gel10-14 or after electroelution (2) O’Farrel, P. H. J. Biol. Chem. 1975, 250, 4007. (3) Celis, J.; Gromov, P.; Ostergaard, M.; Madsen, P.; Honore, B.; Dejgaard, K.; Olsen, E.; Vorum, H.; Kristensen, D.; Gromova, I.; Haunso, A.; Damme, J. V.; Puype, M.; Vandekerchhove, J.; Rasmussen, H. FEBS Lett. 1996, 129. (4) Klose, J.; Kobaltz, U. Electrophoresis 1995, 16, 1034. (5) Yan, J. X.; Tonella, L.; Sanchez, J. C.; Wilkins, M. R.; Packer, N. H.; Gooley, A. A.; Hochstrasser, D. F.; Williams, K. L. Electrophoresis 1997, 18, 497. (6) Karas, M.; Bachmann, D.; Bahr, U.; Hillenkamp, F. Int. J. Mass Spectrum. Ion Processes 1987, 78, 53. (7) James, P.; Quadroni, M.; Carafoli, E.; Gonnet, G. Protein Sci. 1994, 3, 1347. (8) Cohen, S. L.; Chait, B. T. Anal. Biochem. 1997, 247, 257. (9) Ehring, H.; Stro ¨mberg, S.; Tjernberg, A.; Nore´n, B. Rapid Commun. Mass Spectrom. 1997, 11, 1867. (10) Dunphy, J. C.; Busch, K. L.; Hettich, R. L.; Buchanan, M. V. Anal. Chem. 1993, 65, 1329. 10.1021/ac980672w CCC: $18.00
© 1999 American Chemical Society Published on Web 12/18/1998
of the proteins from the gels.15-17 In another technique, SDSPAGE-separated proteins are electroblotted onto poly(vinylidene difluoride) (PVDF) membranes for molecular weight determinations by MALDI-MS,18-22 and for in situ proteolytic fragmentation prior to MALDI-MS analysis,23,24 or onto nitrocellulose membranes.25 Direct desorption from SDS gels has recently been reported with the use of ultrathin PAGE gels.26,27 Analytical 2D electrophoresis combined with MALDI-MS is a powerful procedure that allows high resolution of proteins and rapid identification. This has been demonstrated in cell cultures28,29 and tissue extracts.3-5 However, analytical 2D gel electrophoresis procedures are seldom capable of supplying sufficient amounts of low-abundance proteins for further characterization. It is often necessary to recover proteins from several gels for mass spectrometry or sequence analysis.16,30 Body fluids, such as cerebrospinal fluid (CSF) or serum, pose a particular problem due to the presence of certain proteins in high concentrations; notably albumin and immunoglobulins, which together comprise more than 80% of the total proteins. The high protein load severely limits the volume of CSF or serum that can be analyzed by analytical 2D gel electrophoresis. The preparative electrophoresis method circumvents the obstacles to analysis that exist in analytical 2D gel electrophoresis. Therefore, we have used a novel combination of methods, preparative two-dimensional liquid-phase electrophoresis (2D-LPE) and MALDI-TOF MS in order to characterize (11) Mortz, E.; Sareneva, T.; Haebel, S.; Julkunen, I.; Roepstorff, P. Electrophoresis 1996, 17, 925. (12) Otto, A.; Thiede, B.; Muller, E.; Scheler, C.; Wittmann-Liebold, B.; Jungblut, P. Electrophoresis 1996, 17, 1643. (13) Nakayama, H.; Uchida, K.; Shinkai, F.; Shinoda, T.; Okuyama, T.; Seta, K.; Isobe, T. J. Chromatogr., A 1996, 730, 279. (14) Shvechenko, A.; Wilm, M.; Vorm, O.; Mann, M. Anal. Chem. 1996, 68, 850. (15) Clauser, K. R.; Hall, S. C.; Smith D. M.; Webb, L. E.; Andrews, L. E.; Tran, H. M.; Epstein, L. B.; Burlingame, A. L. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 5072. (16) Hall, S. C.; Smith, D. M.; Masiarz, F. R.; Soo, V. W.; Tran, H. M.; Epstein, L. B.; Burlingame, A. L. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 1927. (17) Haebel, S.; Jensen, S.; Andersen, O.; Roepstorff, P. Protein Sci. 1995, 4, 394. (18) Eckerskorn, C.; Strupat, K.; Karas, M.; Hillenkamp, F.; Lottspeich, F. Electrophoresis 1992, 13, 664. (19) Vestling, M. M.; Fenselau, C. Anal. Biochem. 1994, 66, 471. (20) Vestling, M. M.; Fenselau, C. Biochem. Soc. Trans. 1994, 22, 547. (21) Strupat, K.; Karas, F.; Hillenkamp, C.; Eckerkorn, C.; Lottspeich, F. Anal. Chem. 1994, 66, 464. (22) Blais, J. C.; Nagnan-Le-Meillour, P.; Bolbach, G.; Tabet, J. C. Rapid Commun. Mass Spectrom. 1996, 10, 1. (23) Rasmussen, H. H.; Mortz, E.; Mann, M.; Roepstorff, P., Celis, J. E. Electrophoresis 1994, 15, 406. (24) Mortz, E.; Vorm, H.; Mann, M.; Roepstorff, P. Biol. Mass Spectrom. 1994, 23, 249. (25) Liang, X.; Bai, J.; Liu, Y. H.; Lubman, D. Anal. Chem. 1996, 68, 1012. (26) Ogorzalek Loo, R. R.; Stevenson, T. I.; Mitchell, C.; Loo, J. A.; Andrews, P. C. Anal. Chem. 1996, 68, 1910. (27) Ogorzalek Loo, R. R.; Mitchell, C.; Stevenson, T. I.; Martin, S. A.; Hines, W. M.; Juhasz, P.; Patterson, D. H.; Peltier, J. M.; Loo, J. A.; Andrews, P. C. Electrophoresis 1997, 18, 382. (28) Matsui, N. M.; Smith, D. M.; Clauser, K. R.; Fichmann, J.; Andrews, L. E.; Sullivan, C. M.; Burlingame, A. L.; Epstein, L. B. Electrophoresis 1997, 18, 409. (29) Li, G.; Waltham, M.; Anderson, N. L., Unsworth, E.; Treston, A.; Weinstein, J. N. Electrophoresis 1997, 18, 391. (30) O′Connell, K. L.; Stults, J. T. Electrophoresis 1996, 18, 349. (31) Blennow, K.; Fredman, P.; Wallin, A.; Gottfries, C. G.; Karlsson, I.; Långstro¨m, G.; Skoog, I.; Svennerholm, L.: Wikkelso¨, C. Eur. Neurol. 1993, 33, 129. (32) Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. J. J. Biol. Chem. 1951, 193, 265.
low-abundance proteins in CSF. In 2D-LPE, the first step fractionates proteins into defined pH ranges by liquid-phase IEF. Proteins are enriched in liquid fractions, up to 500 times, at their respective isoelectric points. In the second purification step, individual proteins are isolated on the basis of their size differences in liquid phase by SDS-PAGE. Preparative 2D-LPE allows high protein loads (up to 1 g) and large volumes (up to 55 mL), thus yielding sufficient amounts of low-abundance proteins for further characterization by mass spectrometry. Also, the proteins remain in liquid phase during the entire procedure, making it possible to monitor proteins both in the first and in the second dimension and obviating the extra steps for extraction, electroelution, or blotting of the proteins prior to mass spectrometric analysis. We used preparative 2D-LPE in combination with MALDI-TOF MS for purification and characterization of low-abundance proteins in cerebrospinal fluid. The usefulness of this new strategy is illustrated by purification and characterization of two CSF proteins, β-2 microglobulin and cystatin C. EXPERIMENTAL SECTION Cerebrospinal Fluid. CSF samples were obtained from the Clinical Neurochemical Laboratory, Sahlgrenska University Hospital. The investigation was performed on pooled CSF samples from patients, on whom lumbar puncture was performed to exclude infectious disorders of the CNS. The inclusion criteria were normal white cell count, blood-brain barrier function, and absence of intrathecal IgG and IgM production.31 Lumbar puncture was performed in the lateral decubitus position, in the L4-L5 vertebral interspace. The first 12 mL of CSF was collected and gently mixed to avoid possible gradient effects. The CSF samples were centrifuged at 2000g (+4 °C) for 10 min to eliminate cells and other insoluble material and kept at +4 °C until analysis. Sample Preparation. A 15-mL volume of CSF was treated with 10% trichloroacetic acid (TCA) for 1 h on ice in order to precipitate the proteins. Protein pellets were washed twice with ether/ethanol (1:1 v/v), centrifuged (2000g, 10 min), and dried. The TCA precipitate was brought to a volume of 12 mL with 6 M urea, 20 mM dithioerithritol, and 2% BioLyte ampholytes (BioRad Laboratories, Hercules, CA), pH range 3-10. Liquid-Phase Preparative IEF. This was performed using a Rotofor apparatus (Bio-Rad Laboratories), following the instructions given by the manufacturer. In short, electrolytes in the anode and the cathode chambers were 0.1 M H3PO4 and 0.1 M NaOH, respectively. In the first separation step, the CSF protein sample (12 mL) was loaded into the Rotofor cell. Constant power (12 W) was applied for 4 h or until the voltage had stabilized. The initial conditions were 560 V and 18 mA, and at equilibrium, the values were 2700 V and 3 mA. Twenty fractions were rapidly harvested and their pH values were determined immediately. The total protein concentration was determined in each fraction by a modification of the Lowry method32 using the DC Protein Assay (Bio-Rad Laboratories). The Rotofor fractions were also analyzed by 12% SDS-PAGE, followed by silver staining using the Xpress silver staining kit (Novex, San Diego, CA) and by western blotting using antibodies against cystatin C or β-2 microglobulin (Dakopatts, Glostrup, Denmark), respectively. Continuous-Elution SDS-PAGE. Rotofor fractions containing cystatin C and β-2 microglobulin as determined by immunoblotting were combined and concentrated by vacuum centrifugaAnalytical Chemistry, Vol. 71, No. 3, February 1, 1999
643
tion (Savant Speed Vac Concentrator, model SC 210A, Savant Instruments, Inc., Farmingdale, NY). The final sample, corresponding to Rotofor fraction numbers 9-11, containing ∼50 mg of total proteins dissolved in 1 mL of SDS sample buffer (0.06 M Tris-HCl pH 6.8 containing 2% SDS, 3% dithioerithritol, 10% glycerol, and 0.025% bromophenol blue), was boiled for 5 min and applied to the model 491 Prep cell (Bio-Rad Laboratories), for purification according to the instruction manual. The gel composition was 17% T/2.67% C, with a height of 10 cm and a gel tube size of 28 mm (internal diameter). The stacking gel composition was 4% T/2.67% C with a height of 2.0 cm. T is the weight percentage of total monomer including cross-linker (N,N′-methylenebisacrylamide) and C is the proportion of cross-linker as a percentage of total monomer. The Prep cell was set up for electrophoresis by using a running and elution buffer containing 25 mM Tris, 0.192 M glycine, and 0.1% SDS (pH 8.3). The apparatus was run at 12 W constant power with continuous, recirculating cooling. Elution of fractions commenced when the bromophenol blue indicator band reached the base of the separating gel (5 h). Fractions of 2.5 mL were collected with a peristaltic pump set at 0.7 mL/min. In total, 100 fractions (2.5 mL) were collected. Fraction 1 was the first fraction containing visible amounts of the bromophenol marker dye. To locate the fractions containing β-2 microglobulin and cystatin C, 40-µL aliquots from every fifth fraction were concentrated and analyzed using 15% analytical Tris-glycine ready gels (Bio-Rad Laboratories), followed by immunoblotting. Removal of SDS in Prep Cell Fractions. SDS Extraction. After preparative gel electrophoresis in the Prep cell, selected Prep cell fractions were dried to 200 µL. The final concentration in the 200 µL sample was Tris (125 mM), glycine (0.96 M), and SDS (0.5%). SDS in the samples was extracted by a modification of the procedure of Wessel and Flu¨gge.33 An 800-µL aliquot of methanol was added to 200 µL of the protein sample, and the samples were mixed for total collection of the sample. Then 200 µL of chloroform was added, and the samples were mixed again. For phase separation, 600 µL of water was added and the samples were mixed continuously for 30 min and then centrifuged (1 min; 14000g). The upper phase was carefully removed and discarded. An additional 600 µL of methanol was added to the rest of the lower chloroform phase and the interphase with the precipitated protein. The samples were mixed and centrifuged again for 5 min at 14000g in order to pellet the protein. The supernatant was carefully removed, and the protein pellet was dried under a stream of air. SDS was quantified in the samples by a colorometric method.34 Ten DG Desalting Columns. One selected Prep cell fraction (2.5 mL), containing β-2 microglobulin was applied to the column (Econo-Pac 10 column, Bio-Rad Laboratories). The columns are packed with Bio-Gel P-6 desalting gel. The desalting procedure was performed according to the manufacturer’s manual for general desalting or buffer exchange. Six fractions (1.0 mL) were collected, and SDS was determined in the samples. (33) Wessel, D.; Flu ¨ gge, U. I. Anal. Biochem. 1984, 138, 141. (34) Arand, M.; Friedberg, T.; Oesch, F. Anal. Biochem. 1992, 207, 73. (35) Laemmli, U. K. Nature 1970, 227, 680. (36) Leary, J. J.; Brigati, D. J.; Ward, D. C. Proc. Natl. Acad. Sci. U.S.A. 1983, 80, 4045. (37) Westman, A.; Nilsson, C. L.; Ekman, R. Rapid Commun. Mass Spectrom. 1998, 12, 1092-1098.
644
Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
Analytical Procedures. The presence of β-2 microglobulin and cystatin C in the Rotofor and Prep cell fractions was determined by western blotting using antibodies against cystatin C or β-2 microglobulin. The samples (10 µL) were electrophoresed through a 15% Trisglycine gel (Mini-ProteanII cell; Bio-Rad Laboratories) using the buffer systems of Laemmli.35 The proteins were transferred from the gel to a PVDF membrane (Millipore, Bedford, CA), by means of the semidry technique using the NovaBlot System (Pharmacia, Uppsala, Sweden) at 0.8 mA/cm2 for 30 min, and blocked with 5% milk powder in phosphate-buffered saline (PBS, 58 mM Na2HPO4‚2H2O, 17 mM Na2P04‚H2O, 68 mM NaCl, pH 7.4), containing 0.05% Tween-20. The membranes were then incubated overnight with rabbit anti-serum against cystatin C or β-2 microglobulin (Dakopatts, Glostrup, Denmark), diluted to 1 µg/mL. After washing, the membrane was incubated with alkaline phosphatase conjugated goat anti-rabbit Ig (Bio-Rad Laboratories) diluted 1/1000 for 1 h. The color reaction was developed with 0.015% 5-bromo-4-chloro-3-indolyl phosphate (BCIP) and 0.030% nitro-blue tetrazolium in 0.1 M carbonate buffer containing 1.0 mM MgCl2.36 MALDI Mass Spectrometry. Materials. Calibrant peptides (angiotensin II, ACTH 18-39) were purchased from the Sigma Chemical Co. (St. Louis, MO), and R-cyano-4-hydroxy-cinnamic acid (CHCA) was purchased from Aldrich Chemie GmbH (Steinheim, Germany). CHCA, 15 g/L (saturated), was dissolved in 0.1% TFA in acetonitrile/water (1:1 v/v). All solvents used were of HPLC quality. Sample Preparation. Samples were prepared with the so-called seed layer method.37 First, a matrix seed layer was created by depositing a droplet (0.5 µL) of a 1 g/L solution of matrix dissolved in acetonitrile on the highly polished, stainless steel, sample probe. Thereafter, equal volumes (5 µL) of the matrix and analyte solutions were mixed in a test tube, and a droplet (0.5 µL) of matrix/analyte mixture was deposited on the matrix seed layer. The sample was then left to dry completely in air. Apparatus. All MALDI analyses were performed with a Reflex MALDI-TOF mass spectrometer (Bruker-Franzen Analytik GmbH, Bremen, Germany). Samples were irradiated with a nitrogen laser (VSL-337, Laser Science Inc., Cambridge, MA). A circular gradient neutral density filter (28650, Oriel Co.) permitted continuous attenuation of the laser beam down to 1% of the laser’s output energy. The ion source and flight tube were evacuated by turbo pumps to a pressure of less than 3 × 10-8 Torr. The instrument was equipped with a gridless two-stage electrostatic reflectron and a delayed extraction (time-lag focusing) ion source. All spectra were acquired in reflectron mode at an accelerating voltage of 20 kV. Mass spectra were analyzed using Bruker software on a Sun Sparcstation. The sample potential/first electrode potential ratio was optimized to achieve optimal resolution for the studied peptides. Because the spectrometer was equipped with a video camera, visual inspection of the sample inside the Bruker Reflex II MALDI-TOF mass spectrometer was possible. All spectra shown were calibrated using external calibration. Enzymatic Digestion. Prep cell fractions 61/100 for cystatin C and 55/100 for β-2 microglobulin, which showed the presence of cystatin C/β-2 microglobulin by western blotting, were dried and dissolved in 25 µL of digestion buffer containing 0.1 mM CaCl2
Figure 2. Immunoblot analysis of Rotofor fractions, containing cystatin C and β-2 microglobulin. The CSF protein sample (corresponding to 15 mL of CSF) was loaded onto the Rotofor cell as described in the users’ manual and run at 12 W constant power. Twenty fractions were harvested; SDS-PAGE (15%) followed by immunoblotting of the Rotofor fractions. (a) shows immunoblotting of cystatin C b/β-2 microglobulin.
Figure 1. Flow chart of the general strategy, the combination of 2D-LPE and MALDI-TOF MS, employed for purification and characterization of low-abundance CSF proteins.
and 0.1 M NH4HCO3 in water. Porcine trypsin, 1 g/L (Sigma Chemical Co.) dissolved in 1 mM HCl and 0.1 mM CaCl2 in water, was added. A protein to enzyme ratio on the order of 500:1 was used. The samples were incubated for 4 h at +37 °C. Prior to MALDI mass analysis, the samples were dried and reconstituted in 30 µL of 0.1% TFA in water. High-Accuracy Peptide Mass Mapping. The Internet protein resource ”MS-Digest” (http://donatello.ucsf.edu/) was used to compare the predicted monoisotopic m/z values of proteolytic digest fragments from cystatin C and β-2 microglobulin to the experimental values. RESULTS AND DISCUSSION Preparative 2-D Liquid-Phase Electrophoresis. LiquidPhase IEF. The general strategy employed in this study is outlined in Figure 1. We have chosen cystatin C and β-2 microglobulin to illustrate the application of the combined Rotofor/Prep cell/ MALDI-TOF MS procedure. The total protein amount of the sample loaded on the first dimension was 4.5 mg. The pH range of the 20 Rotofor fractions varied between 2.5 and 9.0. For analysis of β-2 microglobulin and cystatin C, a 5-µL aliquot from each Rotofor fraction was subjected to SDS-PAGE and immunoblotting using antibodies against these proteins. Cystatin C was localized in Rotofor fractions 5-20, as a single band at 15 kDa, covering a pH range of 4.50-9.00. (Figure 2a). β-2 microglobulin was localized in Rotofor fractions 5-12, as a single band at ∼11 kDa, with a pH range of 4.5-7.0 (Figure 2b). Preparative isoelectric focusing using the Rotofor cell has previously been used for isolation and characterization of proteins
from serum,38,39 cell cultures,40-42 and various tissues.43,44 The high concentration of certain proteins, such as albumin, limits the volume of CSF that can be loaded onto analytical 2D PAGE gels. To the best of our knowledge, no study has been performed on CSF proteins using liquid-phase IEF, except one by us.45 In that study, the combination of liquid-phase IEF and immunoblotting was used successfully to enrich trace amounts of synaptic proteins (present in ng/L) from human CSF. The results demonstrate that preparative IEF has a unique ability to enrich CSF proteins, present in low concentrations, for further purification using preparative SDS-PAGE and characterization by MALDI-TOF MS. Continuous-Elution SDS-PAGE. Further purification of cystatin C and β-2 microglobulin by molecular weights was performed using the model 491 Prep cell. From Rotofor fractions 8-10, cystatin C was recovered in fractions 60-75 from the Prep cell (Figure 3a), and β-2 microglobulin was collected in the fractions 45-60 (Figure 3b). The presence of contaminants was assessed by inspection of silver-stained SDS-PAGE gels, on which cystatin C and β-2 microglobulin stained as distinctive bands (Figure 3c). No other bands were visible on the gel. As shown, for β-2 microglobulin and cystatin C, proteins with small differences in mass can easily been separated by the Prep cell, making it possible to ascertain the molecular homogeneity of a gel band and determine the structural relationship between protein isoforms. (38) Goldfarb, M. F. Electrophoresis 1993, 148, 1379. (39) Malle, E.; Hess, H.; Muncher, G.; Knipping, G.; Steinmetz, A. Electrophoresis 1992, 13, 422. (40) Ohshima, Y.; Morita, M.; Hirashima, M.; Mori, K. J.; Akutagawa, H.; Katamura, K.; Mayumi, M.; Mikawa, H. Exp. Hematol. (Charlottesville, VA) 1993, 21, 749. (41) Ni, J.; Karpas, A. Cytokine 1993, 5, 31. (42) Peritt, D.; Flechner, I.; Okunev, E.; Yanai, P.; Halperin, T.; Treves, A. J.; Barak, V J. Immunol. Methods 1992, 155, 159. (43) Furster, C.; Zhang, J.; Toll, A. J. Biol. Chem. 1996, 271, 20903. (44) Paliwal, R.; Costa, G.; Diwan, J. J. Biochem. 1992, 31, 2223. (45) Davidsson, P.; Puchades, M.; Blennow, K. Electrophoresis, in press.
Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
645
Figure 3. Immunoblot analysis of every fifth Prep cell fraction. (a) cystatin C and (b) β-2 microglobulin. (c) Aliquots from every fourth Prep cell fraction were analyzed on a silver stained 15% Tris-glycine ready gel. The elution position of cystatin C was fractions 60-75 and for β-2 microglobulin fractions 45-60.
The particular conditions used in the preparative SDS-PAGE here were used to elute proteins between 10 and 16 kDa, while many other proteins in the mixture were excluded from the gel by molecular mass. These conditions were successfully applied in the isolation of β-2 microglobulin and cystatin C. These results show that preparative 2D-LPE yields highly purified preparations of proteins for mass spectrometry or sequencing. The use of continuous-elution electrophoresis to purify and extract larger quantities of a protein from complex mixtures had been earlier attempted,46-48 but not in combination with liquidphase IEF in the Rotofor cell. To the best of our knowledge, this is the first report of the use of 2D LPE in combination with MALDI-TOF MS. Removal of SDS. The presence of ionic detergents, such as SDS, usually severely compromises the MALDI analysis technique. No MALDI spectra could be obtained in the presence of 0.1-0.5% SDS, results that are in agreement with previous studies. In the present study, prior to MS analysis, SDS was removed from one of the cystatin C/β-2 microglobulin fractions using a (46) Dubois, B.; Deloron, P. J. Immunol. Methods 1997, 4 (6), 29. (47) Gambino, R.; Ruiu, G.; Cassader, M.; Pagano, G. J. Lipid Res. 1996, 37, 902. (48) Lo, C. S.; Hughes, C. V. Oral Microbiol. Immunol. 1996, 11 (3), 181. (49) Puchades, M.; Westman, A.; Blennow, K.; Davidsson, P. Rapid Commun. Mass Spectrom., submitted. (50) Rosinke, B.; Strupat, K.; Hillenkamp, F.; Rosenbusch, J.; Dencher, N.; Kruger, U.; Galla, H. J. J. Mass. Spectrom. 1995, 30, 1462. (51) Schively, J. E. In Methods of protein microcharacterization; Schively, J. E., Ed.; Humana Press: Clifton, NJ, 1986; p 41. (52) Koningsberg, W. H.; Henderson, L. Methods Enzymol. 1983, 91, 254. (53) Kawasaki, H.; Suzuki, K. Anal. Biochem. 1990, 186, 264.
646
Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
modification of the Wessel and Flu¨gge method. The amount of SDS was reduced from 1.25% in the runnning-buffer to below 0.0002%. In another study, we demonstrated that the amount of SDS critically influences the sensitivity of protein detection by MALDI-MS.49 Immunoblotting analysis of cystatin C showed protein sample recovery at approximately 50-70%. However, one problem observed with this procedure is that β-2 microglobulin coprecipitated with SDS, leaving only trace amounts of the protein for analysis. An alternative method for the removal of SDS from the fractions containing β-2 microglobulin is the use of 10 DG desalting columns. The β-2 microglobulin protein was collected in the first 3 mL eluted from the column. This procedure resulted in a reduction of SDS to below 0.003%. Immunoblotting analysis of β-2 microglobulin before and after the SDS removal procedures showed protein sample recovery in the 50-70% range. Previously, a variety of methods have been used to permit MS analysis of SDS-separated proteins. However, isolation procedures for proteins from gel bands frequently do not remove SDS efficiently. This may severely affect sample preparation and quench ion formation, degrading the quality of MS analysis. The presence of SDS in blotted proteins samples is a minor problem, but MS analysis of proteins blotted onto membranes has limitations in sensitivity and resolution.50 A number of methods exist for removing SDS, such as protein precipitation using guadinium chloride,51 ion-pairing reagents,52 or reversed-phase HPLC.53 But these methods are usually timeconsuming and cause a significant loss of proteins. When the detergent is removed, many proteins become insoluble. The risk of losing the protein sample is especially high, when low levels of the protein being handled. We have demonstrated that the use of the modification of the Wessel and Flu¨gge method and the 10 DG desalting columns are easy and rapid methods for the removal of SDS. Also, the methods provide sufficient reduction of SDS prior to enzymatic digestion and high recovery rates for protein. MALDI-MS. MALDI-MS analysis of the tryptic digests from Prep cell fractions 55 and 61 confirmed the presence of cystatin C (Figure 4a) and β-2 microglobulin (Figure 4b), respectively. The sequence coverage provided by this analysis was 48.3% for cystatin C (Table 1) and 27.3% for β-2 microglobulin (Table 2). CONCLUSION By using the combination of preparative 2D-LPE and MALDITOF MS, characterization of small amounts of proteins in complex biological samples is facilitated. MALDI-TOF MS has a high sensitivity in biological samples, typically 1 pmol/µL. For proteins in the range of 10-100 kDa, this means that the concentration has to be between 20 and 200 ng/µL. In biological samples, such as serum and CSF, few proteins are found in levels above this. Thus, there is a need to be able to load large amounts of total protein in the purification steps before MS. This can be accomplished with preparative 2D-LPE. Preparative 2D-LPE is based on the same isoelectric focusing and gel electrophoresis principles as the widely used analytical 2-D gel electrophoresis. However, this novel approach shows many improvements compared to analytical two-dimensional gel electrophoresis in the investigation of complex biological fluids. For example, it is possible to load large sample volumes/protein amounts of the mixtures and therefore obtain sufficient amounts of low-abundance proteins for further characterization. Also, the proteins remain in liquid phase
Figure 4. MALDI-MS analysis of tryptic digest of (a) Prep cell fraction 61/100. Tryptic peptides from cystatin C are indicated by arrows. (b) Prep cell fraction 55/100. Tryptic peptides from β-2 microglobulin are indicated by arrows. Table 1. Tryptic Digest Observed by MALDI-MS (Cystatin C)a m/z
Table 2. Tryptic Peptides Observed by MALDI-MS (β-2 Microglobulin)a m/z
m/z
sequence
obsd
calcd
sequence
obsd
calcd
sequence
obsd
calcd
9-24 9-24, Met-ox 9-25 9-25, Met-ox 25-36
1645.6 1661.4 1801.7 1816.6 1383.5
1644.7 1660.7 1800.5 1816.9 1382.7
26-36 37-45, Met-ox 54-70 55-70 55-74, Pyro-Glu
1226.4 1096.3 1920.8 1791.5 2382.7
1226.6 1096.4 1921.1 1792.9 2381.2
42-48 49-58 82-91
845.4 1148.5 1122.6
845.4 1148.5 1122.6
a
a
Sequence coverage was 27/99 amino acids or 27.3%.
Sequence coverage was 58/120 amino acids or 48.3%.
during the entire procedure, making it possible to monitor proteins both in the first and in the second dimension, and obviating the extra steps for extraction, electroelution, or blotting of the proteins prior to mass spectrometric analysis. This combination provided enough resolving power to completely separate two proteins with similar pI and a mass difference of 2 kDa. Our findings serve to illustrate the usefulness of this novel combination of methods in the analysis of low-abundance proteins in body fluids. This new strategy might also be useful in proteome studies in tissue extracts, whole cells, bacterial lysates for characterization or sequencing of proteins by MALDI-TOF MS, or production of monoclonal antibodies.
ACKNOWLEDGMENT This work was supported by grants from The Swedish Medical Research Council (Grants 12769, 11560, 12103, 07517, and 12575); State under the LUA agreement (98-280, R. Ekman), Alzheimerfonden, Lund, Sweden; Bohuslandstingets FoU fond, Sweden; Gun and Bertil Stohne′s Stiftelse; Kista, Sweden; Stiftelsen Handlanden Hjalmar Svenssons Forskningsfond, Go¨teborg, Sweden; JanssenCilag AB Sweden; Loo and Hans Osterman fond, Stockholm, Sweden; Lundbecksstiftelsen, Lund, Sweden; Magnus Bergvalls Stiftelse, Stockholm, Sweden; and Åke Wibergs Stiftelse, Stockholm, Sweden. Received for review June 19, 1998. Accepted October 27, 1998. AC980672W Analytical Chemistry, Vol. 71, No. 3, February 1, 1999
647
