Anal. Chem. 2000, 72, 2148-2153
Identification of Protein Vaccine Candidates from Helicobacter pylori Using a Preparative Two-Dimensional Electrophoretic Procedure and Mass Spectrometry Carol L. Nilsson,*,† Thomas Larsson,† Elisabet Gustafsson,† Karl-Anders Karlsson,† and Pia Davidsson‡
Institute of Medical Biochemistry, Go¨teborg University, Box 440, SE-405 30 Go¨teborg, Sweden, and Department of Clinical Neuroscience, Experimental Neuroscience section, Go¨teborg University, Sahlgrenska University Hospital/Mo¨lndal, Mo¨lndal, Sweden
Helicobacter pylori is an important human gastric pathogen for which the entire genome sequence is known. This microorganism displays a uniquely complex pattern of binding to complex carbohydrates presented on host mucosal surfaces and other tissues, through adhesion molecules (adhesins) on the microbial cell surface. Adhesins and other membrane-associated proteins are important targets for vaccine development. The identification and characterization of cell-surface proteins expressed by H. pylori is a prerequisite for the development of vaccines designed to interfere with bacterial colonization of host tissues. However, identification of membrane proteins is difficult using a traditional proteomics approach employing 2D-PAGE. We have used a novel approach in the identification of microbial proteins that employs a rapid preparative two-dimensional electrophoretic separation followed by mass spectrometry and database searches. No pre-enrichment of bacterial membranes was required. The entire process, from sample preparation to protein identification, can be completed in less than 18 hours, and the presence of proteins can be monitored after both the first- and second-dimensional separations using mass spectrometry. We were able to identify 40 proteins from a detergent-solubilized H. pylori preparation; over onethird of these were membrane or membrane-associated proteins. A functionally characterized low-abundance membrane protein, the Leb-binding adhesin, was found in this group. The use of this rapid 2D electrophoretic separation in proteomic studies of H. pylori is expected to speed up the identification of expressed virulence proteins and vaccine targets in this and other microbial pathogens. Helicobacter pylori is an important causative agent for gastrointestinal disorders, such as chronic gastritis, peptic ulcers, and *
[email protected], FAX: +46-31-41 61 08. † Institute of Medical Biochemistry, Go ¨teborg University. ‡ Department of Clinical Neuroscience, Experimental Neuroscience section, Go ¨teborg University.
2148 Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
stomach cancer,1 for which the entire genome sequence is known.2 It is estimated that approximately one-half of the world’s population is infected with this organism, which possesses a large number of virulence proteins for colonizing human gastric mucosa, host defense evasion, and tissue damage.3 There is an urgent need for the development of an effective vaccine against this pathogen, to prevent disease manifestations, eliminate the need for antibiotic therapy, and prevent reinfection. Adhesion of the bacteria to host cell tissues is a requirement for chronic infection to occur. H. pylori strains have been demonstrated to have a highly diverse binding capacity to molecules displayed on host cell surfaces, including N-acetylneuraminyllactose,4 sulfated carbohydrates,5 lactosylceramide,6 polyglycosylceramides,7 gangliotetraosylceramide,8 and the Lewisb (Leb) blood group antigen.9 A strategy for vaccine design aimed at inducing immunity toward colonization factors requires the identity of expressed adhesins contained in the microbial cell envelope. Many microbial proteins can be rapidly profiled using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI(1) Dunn, B. E.; Cohen, H.; Blaser, M. J. Clin. Microbiol. Rev. 1997, 10, 720741. (2) Tomb, J.-F.; White, O.; Kerlavage, A. R.; Clayton, R. A.; Sutton, G. G.; Fleischmann, R. D.; Ketchum, K. A.; Klenk, H. P.; Gill, S.; Dougherty, B. A.; Nelson, K.; Quackenbush, J.; Zhou, L.; Kirkness, E. F.; Peterson, S.; Loftus, B.; Richardson, D.; Dodson, R.; Khalak, H. G.; Glodek, A.; McKenney, K.; Fitzgerald, L. M.; Lee, N.; Adams, M. D.; Hickey, E. K.; Berg, D. E.; Gocayne, J. D.; Utterback, T. R.; Peterson, J. D.; Kelley, J. M.; Cotton, M. D.; Weidman, J. W.; Fuji, C.; Bowman, C.; Watthey, L.; Wallin, E.; Hayes, W. S.; Borodovsky, M.; Karp, P. D.; Smith, H. O.; Fraser, C. M.; Venter, J. C. Nature (London) 1997, 388, 539-547. (3) Mobley, H. L. T. Gastroenterology 1997, 113, S21-28. (4) Evans, D. G.; Evans,D. J., Jr.; Moulds, J. J.; Graham, D. Y. Infect. Immun. 1988, 56, 2896-2906. (5) Saitoh, T.; Natomi, H.; Zhao, W.; Okuzumi, K.; Sugano, K.; Iwamori, M.; Nagai, Y. FEBS Lett. 1991, 282, 385-387. (6) Ångstro¨m, J.; Teneberg, S.; Milh, M. A.; Larsson, T.; Leonardsson, I.; Olsson, B. M.; ¨Olwegård Halvarsson, M.; Danielsson, D.; Ljungh, A° .; Wadstro ¨m, T.; Karlsson, K.-A. Glycobiology 1998, 8, 297-309. (7) Miller-Podraza, H.; Milh, M. A.; Bergstro ¨m, J.; Karlsson, K.-A. Glycoconjugate J. 1996, 13, 453-460. (8) Gold, B. D.; Huesca, M.; Sherman, P. M.; Lingwood, C. A. Infect. Immun. 1993, 61, 2632-2638. (9) Bore´n, T.; Falk, P.; Roth, K. A.; Larson, G.; Normark, S. Science (Washington, D.C.) 1993, 262, 1892-1895. 10.1021/ac9912754 CCC: $19.00
© 2000 American Chemical Society Published on Web 03/25/2000
TOFMS) analysis of intact bacteria10-13 and bacterial extracts.14-18 Two earlier studies have focused on MALDI-TOFMS fingerprinting strains of H. pylori.19,20 High-confidence identification of proteins expressed by H. pylori requires, however, further separation steps. Two-dimensional (2D) separation using isoelectric focusing (IEF) followed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) is a powerful method for separating thousands of proteins in microbial systems and an excellent visualization tool for studies of protein expression.21 Peptide mapping performed on proteins digested in-gel is used routinely to identify spots. However, an important drawback to 2D-PAGE separation of microbial proteins is the underrepresentation of membrane proteins, making their identification difficult. It was this limitation that stimulated our efforts to design a separation method that would yield a selection of microbial proteins in which potential vaccine candidates, of which many are membrane proteins that interact with host molecules, could be identified. We have previously used two-dimensional liquid-phase electrophoresis (2D-LPE) and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOFMS) to identify low-abundance proteins in cerebrospinal fluid (CSF)22 and a pleural exudate.23 Recently, we described a rapid 2D electrophoretic procedure, including liquid-phase IEF in the Rotofor cell and electroelution in the Mini Gel Eluter, to purify several CSF proteins in sufficient amount for MALDI-TOFMS analysis.24 These results encouraged us to adapt our methods for the study of microbial proteomes and, in particular, virulence proteins expressed by H. pylori. The advantages of using this novel procedure for identifying bacterial proteins are speed of analysis, amount of total protein that can be separated in the Rotofor (milligrams), and ability to monitor proteins by mass spectrometry after both the first- and second-dimensional separation. EXPERIMENTAL SECTION Chemicals. Porcine trypsin, horse heart myoglobin, equine cytochrome C, angiotensin II, ACTH (18-39), and Glu-fibrinopep(10) Holland, R. D.; Wilkes, J. G.; Rafii, F.; Sutherland, J. B.; Persons, C. C.; Voorhees, K. J.; Lay, J. O., Jr. Rapid Commun. Mass Spectrom. 1996, 10, 1227-1232. (11) Welham, K. J.; Domin, M. A.; Scannell, D. E.; Cohen, E.; Ashton, D. S. Rapid Commun. Mass Spectrom. 1998, 12, 176-180. (12) Karty, J. A.; Lato, S.; Reilly, J. P. Rapid Commun. Mass Spectrom. 1998, 12, 625-629. (13) Arnold, R. J.; Reilly, J. P. Rapid Commun. Mass Spectrom. 1998, 12, 630636. (14) Krishnamurthy, T.; Ross, P. L.; Rajamani, U. Rapid Commun. Mass Spectrom. 1996, 10, 883-888. (15) Chong, B. E.; Wall, D. B.; Lubman, D. M.; Flynn, S. J. Rapid Commun. Mass Spectrom. 1997, 11, 1900-1908. (16) Wang, Z.; Russon, L.; Li, L.; Roser, D. C.; Long, S. R. Rapid Commun. Mass Spectrom. 1998, 12, 456-464. (17) Haag, A. M.; Taylor, S. N.; Johnston, K. H.; Cole, R. B. J. Mass Spectrom. 1998, 33, 750-756. (18) van Adrichem, J. H. M.; Bo ¨rnsen, K. O.; Conzelmann, H.; Gass, M. A. S.; Eppenberger, H.; Kresbach, G. M.; Ehrat, M.; Leist, C. H. Anal. Chem. 1998, 70, 923-930. (19) Winkler, M. A.; Uher, J.; Cepa, S. Anal. Chem. 1999, 71, 3416-3419. (20) Nilsson, C. L. Rapid Commun. Mass Spectrom. 1999, 13, 1067-1071. (21) VanBogelen, R. A.; Schiller, E. E.; Thomas, J. D.; Neidhardt, F. C. Electrophoresis 1999, 20, 2149-2159. (22) Davidsson, P.; Westman, A.; Puchades, M.; Nilsson, C. L.; Blennow, K. Anal. Chem. 1999, 71, 642-647. (23) Nilsson, C. L.; Puchades, M.; Westman, A.; Blennow, K.; Davidsson, P. Electrophoresis 1999, 20, 860-865. (24) Davidsson, P.; Nilsson, C. L. Biochim. Biophys. Acta 1999, 1473, 391-399.
tide B were obtained from the Sigma Chemical Company (St. Louis, MO), and n-octylglucoside (n-octyl-β-D-glucopyranoside) was obtained from Boehringer Mannheim (GmbH, Germany). The MALDI matrixes R-cyano-4-hydroxy-cinnamic acid (CHCA, Aldrich, Steinheim, Germany) and sinapinic acid (Fluka Chemie AG, Switzerland) were used without further purification. ZipTips were purchased from Millipore (Bedford, MA). All solvents used were of HPLC grade. Bacterial Strains and Growth Conditions. The H. pylori strain 17875 was obtained from the Culture Collection, University of Go¨teborg. The bacteria were stored at -80 °C in soy broth containing 15% glycerol by volume and grown on agar (14 g/L) containing 10% heat-inactivated fetal calf serum, brucella broth (28 g/L; Difco Laboratories, Detroit, MI), and 1 mL/L IsoVitale X enrichment vitamins (Becton Dickson Europe, Melyan, France) in a humid (98%) microaerophilic chamber (O2, 5-7%; CO2, 8-10%; N2, 83-87%; and H2, less than 2%) at 37 °C. After 48 h, the cells were scraped off and washed three times in phosphate-buffered saline (PBS). Bacteria from a single culture dish (90 mg, wet weight) were used. Liquid-Phase IEF. Liquid-phase IEF has been described previously in detail.24 Briefly, the bacterial lysate was diluted with 15 mL of 0.1% n-octylglucoside and sonicated for 15 min. Servalyte (40%, pH range 3-10 isodalt, Serva Electrophoresis, GmbH, Germany) was added to the bacterial mixture to a concentration of 2.5%. The solubilized bacterial pellet was then loaded into the Rotofor cell (Bio-Rad, Hercules, CA) for focusing without further treatment. Constant power (10 W) was applied with the system cooled to +4 °C. Initial voltage was 570 V, and a plateau of 1100 V was reached after 4 h. Twenty separate fractions were harvested and their pH values measured. Fifteen microliters of each fraction was reserved for MALDI-MS analysis of proteins. The Rotofor fractions were analyzed by the NuPAGE system (Novex, San Diego, CA) using 10% Bis-Tris gels (15 wells, 7 cm wide, 1-mm thickness) and the 3-(N-morpholino) propane sulfonic acid (MOPS) sodium dodecyl sulfate (SDS) running buffer system, following which they were silver stained using the Xpress silver staining kit (Novex, San Diego, CA). Four Rotofor fractions with different isoelectric points were selected for further separation and electroelution. Electroelution. The electroelution procedure in the Mini Whole Gel Eluter (Bio-Rad, Hercules, CA) has previously been described in detail.24 Briefly, Rotofor fractions 5, 7, 11, and 19 were concentrated by vacuum centrifugation (Savant, Speed Vac Concentrator, model SC 210A, Savant Instruments, Inc., Farmingdale, NY). The Rotofor fractions were dissolved in 200 µL of NuPAGE sample buffer (0.14 M Tris, 0.10 M Tris-HCl, 0.4 mM EDTA, pH 8.5, containing 10% glycerol, 2% lithium dodecyl sulfate (LDS), 0.08% serva blue, 0.025% phenol red, 3% dithioerithritol), boiled for 5 min, and then separated by the NuPAGE system (Novex, San Diego, CA) using 10% Bis-Tris gels, (1well, 7 cm wide, 1-mm thickness) and the NuPAGE MOPS SDS running buffer. The electrophoresis was run 50 min at 200 V. The Mini Whole Gel Eluter was used, according to the manufacturer’s instruction. The Gel Eluter apparatus was filled with elution buffer (25 mM histidine, 30 mM MOPS, pH 6.5), and the gel was transferred to the unit. Electroelution was performed at 100 mA (constant) for 20 min. Fourteen protein fractions of approximately 0.5 mL each Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
2149
Figure 1. The bacterial pellet was diluted with 15 mL of 0.1% n-octylglucoside and 2.5% servalyte, pH range 3-10 isodalt. Solubilized bacteria were loaded onto the Rotofor cell and run at 10 W of constant power. Twenty fractions were harvested and analyzed by the NuPAGE system and following which were silver stained. Two gels are depicted side by side. Lanes 1-20 correspond to Rotofor fraction 1-20, lane 21 the initial bacterial pellet, and lane 22 the molecular weight standard (Mark 12, Novex, San Diego, CA).
were harvested from the unit. An aliquot of 100 µL was concentrated 10-fold and analyzed by the NuPAGE system (Novex, San Diego, CA) using 10% Bis-Tris gels (15 wells, 7 cm wide, 1-mm thick gel) and the NuPAGE MOPS SDS running buffer, followed by colloidal blue staining (Novex, San Diego, CA). The remaining part of the selected eluted protein fractions was dried by vacuum centrifugation. SDS Removal and Enzymatic Digestion Prior to Peptide Mapping. Following electroelution, SDS was removed from the fractions using a modification of the method described by Maire et al.25 Briefly, 200 µL of protein sample was mixed with 600 µL of ice-cold acetone, stored for 2 h at -20 °C, and centrifuged (10 min, 14000g). The supernant was carefully removed and discarded, and the protein pellet was dried under a stream of air. To identify the protein fractions collected from the electroelution device, enzymatic digestion was performed in order to produce peptide maps. Gel eluter fractions were dried and dissolved in 25 µL of digestion buffer containing 0.1 mM CaCl2 and 0.1 M NH4HCO3 in water. Porcine trypsin (Sigma Chemical Company, St. Louis, MO), dissolved at 1 g/L in 1 mM HCl and 0.1 mM CaCl2 in water, was added prior to digestion. Proteinenzyme ratios on the order of 500:1 were used. The samples were incubated for 4 h at +37 °C. Mass Spectrometry and Database Searching. Samples were analyzed using a Micromass TofSpecE MALDI-TOF mass spectrometer (Micromass, Manchester, UK) equipped with a pulsed 337-nm nitrogen laser, a delayed extraction ion source, and a reflectron. Rotofor fractions, containing proteins, were treated with ZipTips according to the manufacturer’s instruction prior to analysis. Sinapinic acid (15 mg/mL in acetonitrile/H2O, 1:1) was used as a matrix. Proteins were analyzed in linear mode at an accelerating voltage of 25 kV. External calibration using equine cytochrome C and horse heart myoglobin was performed. Dried tryptic digests were reconstituted in 25 µL of 0.1% TFA and treated with ZipTips containing C18 resin according to the manufacturer’s instruction. A 0.5 µL sample was mixed with 0.5 µL of matrix solution (CHCA 10 mg/mL in acetonitrile/H2O, 1:1) (25) LeMaire, M.; Deschamps, S.; Moller, J. V.; LeCaer, J. P.; Rossier, J. Anal. Biochem. 1993, 214, 50-57.
2150 Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
Figure 2. Bacterial proteins separated by IEF in the Rotofor cell were further separated by the NuPAGE system and the whole-gel eluter. Fourteen fractions were harvested and analyzed by the NuPAGE system and following which were stained by colloidal blue staining. Lanes 1-14 correspond to gel eluter fractions 1-14; lane 15 contains a molecular weight standard (Mark 12, Novex, San Diego, CA). The results from Rotofor fraction no. 5 are shown here.
Figure 3. MALDI-TOFMS analysis of proteins in Rotofor fraction no. 11. Intensity is given in arbitrary units and is magnified 3X in the mass range above 37 000 u. Observed proteins, whose identites were revealed through peptide mapping (Table 1), were matched against calculated mass values. The proteins and their experimentally determined masses are adhesin-thiol peroxidase (TagD, 18 330 u), iron superoxide dismutase (FeSOD, 24 483 u), urease alpha subunit (UreA, 26 675 u), translation elongation factor Ef-Tu (EF-Tu, 43 400 u) and urease beta subunit (UreB, 61 700 u).
directly on the MALDI probe and allowed to dry at ambient conditions. The remainder of ZipTip treated samples were dried and reserved for electrospray (ESI) analysis. Peptide spectra were acquired in reflectron mode at an accelerating voltage of 20 kV and are the sum of 100 laser shots. External calibration using angiotensin II and ACTH (18-39) was used. This procedure typically results in mass accuracies of 50-100 ppm. Resulting values for monoisotopic peaks were used for searches (MS-Fit, http://prospector.ucsf.edu) against the NCBInr database. Helicobacter pylori was selected as the species, but pI and molecular weight ranges were not restricted. A mass window of 200 ppm was used. Confirmation of protein identity for samples with fewer than five matching peptides and low sequence coverage provided by MALDI-TOFMS was obtained through fragment ion data acquired
Table 1. Proteins identified in tryptic digests of gel eluter fractions obtained from Rotofor fractions 5, 7, 11, and 19a Rotofor fraction no.
Gel Eluter fraction no.
5
2 3 4 5 6 7 8 10 11 12
13 14 7
2 3 4 5 6 7 8
9 10 11
12
11
19
13 4 7 8 11 13 14 3 6 8 9 10
protein ID
accession no.
no. of peptides matched
MW(calc)
pI(calc)
aconitase B translation elongation factor EF-G 70 kDa chaperone heat shock protein 60 catalase flagellar export protein outer membrane protein 20 heat shock protein 60 translation elongation factor EF-Tu S-adenosylmethionine synthetase 2 alkyl hydroperoxide reductase predicted coding region HP1479 outer membrane protein 6 70 kDa chaperone 70 kDa chaperone putative outer membrane protein cyclic nucleotide phosphodiesterase outer membrane protein 4 alkyl hydroperoxide reductase aldo-keto reductase alkyl hydroperoxide reductase methyl-accepting chemotaxis protein catalase translation elongation factor EF-Tu D-lactate dehydrogenase iron-regulated outer membrane protein adhesin binding fuc. histo-BG antigen heat shock protein 60 urease beta subunit outer membrane protein 20 catalase predicted coding region HP1089 translation elongation factor EF-Tu predicted coding region HP0486 heat shock protein 60 alkyl hydroperoxide reductase S-adenosylmethionine synthetase 2 transcriptase R chain polar flagellin thioredoxin reductase outer membrane protein 6 70 kDa chaperone urease alpha subunit S-adenosylmethionine synthetase 2 outer membrane protein 4 70 kDa chaperone urease alpha subunit cation efflux system protein curved DNA binding protein A outer membrane 15 flagellar sheath adhesin alkyl hydroperoxide reductase adhesin-thiol peroxidase translation elongation factor EF-Tu conserved hypothetical membrane protein urease R subunit iron superoxide dismutase alkyl hydroperoxide reductase urease beta subunit type III restriction enzyme R putative regulatory protein translation elongation factor EF-Tu predicted coding region HP0486 L-asparaginase II cell binding factor 2 predicted coding region HP0369 cell binding factor 2 type III restriction enzyme R heat shock protein
4155281 2494251 4154608 486851 2493545 2493149 2314050 486851 2494256 3024119 2507172 2314667 2313322 2495351 2495351 4154636 2313187 2313212 2507172 2314358 2507172 2313179 2493545 2494256 2314381 2314692 2791668 486851 137076 2314050 2493545 2314241 2494256 2313598 486851 2507172 3024119 2500600 2313876 3024765 2313322 2495351 137069 3024119 2313212 2495351 137069 2314494 2314166 2313829 2497717 2507172 2313491 2494256 2313380 137069 1174379 2507172 137076 2314541 4156100 2494256 2313598 2313847 2499779 2313472 2499779 2314541 486851
7 9 15 27 10 7 11 16 17 9 5 7 5 6 14 6 8 5 5 5 7 7 6 6 5 7 4 14 6 12 13 11 12 6 4 5 6 4 4 7 5 6 10 7 5 9 9 8 5 5 5 7 4 5 5 9 6 5 12 6 5 7 5 6 7 5 8 11 7
92 742 77 022 67 122 58 272 58 629 47 637 55 932 58 272 43 648 42 362 22 236 96 654 53 064 67 052 67 052 31 855 65 799 31 891 22 236 37 074 22 236 74 466 58 629 43 648 106 279 97 385 77 621 58 272 61 684 55 932 58 629 90 793 43 648 59 417 58 272 22 236 42 362 38 167 13 392 33 539 53 064 67 052 26 540 43 262 31 891 67 052 26 540 115 336 32 908 29 929 29 041 22 236 18 292 43 648 60 228 26 538 24 518 22 236 61 684 113 677 43 226 43 648 59 417 33 539 34 031 27 793 34 031 113 677 58 272
6.0 5.2 5.0 5.4 8.7 6.0 9.2 5.4 5.2 6.0 5.9 6.2 9.2 5.0 5.0 9.6 8.7 9.6 5.9 7.1 5.9 5.8 8.7 5.2 7.0 9.1 8.7 5.4 5.6 9.2 8.7 5.6 5.2 9.4 5.4 5.9 6.0 5.0 5.5 5.9 9.2 5.0 8.5 6.0 9.6 5.0 8.5 9.0 8.5 9.0 8.6 5.9 7.7 5.2 9.2 8.5 6.0 5.9 5.6 8.4 10.0 5.2 9.4 8.6 9.3 9.5 9.3 8.4 5.4
% sequence coverage 11 23 24 54 24 27 29 26 39 22 34 12 12 12 22 25 18 17 22 16 15 16 24 6 11 8 31 17 35 32 19 39 10 10 34 14 12 26 27 13 14 42 29 23 14 42 11 21 16 31 37 31 23 9 39 42 38 41 6 13 24 9 18 25 30 29 11 14
a The measured fractions were 4.2, 4.7, 5.8, and 8.8, respectively. Protein identities with fewer than five matching peptides and low sequence coverage using MALDI-MS were confirmed by MS/MS analysis in the QTof.
Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
2151
Table 2. Fragment Ion from the Tryptic Peptide at m/z 931.6 Obtained through MS/MS Analysis in the QTof m/z value ion type a b-18 b y
H-Y
S
136.1
223.1
E
L
G
N
T
Y
N
S
I
T
T
A
L
S
K-NH2
1482.8
1369.6
1312.6
1198.6
1097.6
934.5
820.5
733.4
620.3
519.3
418.3
347.2
234.2
147.1
362.1 380.1 1698.8
by nanoflow electrospray mass spectrometry in a QTof (Micromass, Manchester, UK). Peptide samples were dissolved in 3 µL of acetonitrile/H2O (1:1) containing 1% acetic acid and sprayed from gold-coated glass capillaries (Micromass). Argon was used as the collision gas. Instrument calibration was performed using fragment ions from Glu-fibrinopeptide B and a fourth-order polynomial fit. Both MALDI- and ESI-MS spectra were analyzed using the MassLynx software in a WindowsNT environment. RESULTS AND DISCUSSION An initial fractionation of the solubilized bacteria was carried out using the Rotofor cell. The separation by liquid-phase IEF yielded 20 different fractions, pH range 2.5-9.0, which were analyzed by the NuPAGE system and silver staining (Figure 1). Four Rotofor fractions (5, 7, 11, and 19; measured pI values 4.2, 4.7, 5.8, and 8.8, respectively) were selected for further purification by molecular weight using the Mini Whole Gel Eluter. The fractions were concentrated, separated by the NuPAGE system, and recovered in the Mini Whole Gel Eluter. SDS-PAGE analysis of the protein fractions obtained from the gel eluter demonstrated that the bacterial proteins had been separated into narrow molecular mass fractions (Figure 2). Separated proteins were found in all eluter fractions, and the fractions contained 1-5 protein bands each. The first dimension IEF is performed under nondenaturing conditions; the entirety of the bacteria were solubilized in n-octylglucoside. This detergent is compatible with MALDITOFMS26 analysis, and mass measurements can be performed on proteins in Rotofor fractions23 using this technique. In Figure 3, protein peaks are distinguished in Rotofor fraction no. 11 at m/z values closely corresponding to the calculated values of proteins identified in Gel Eluter fractions derived from Rotofor fraction no. 11 (Table 1). Mass values obtained for proteins by mass spectrometry are more accurate than those obtained by SDSPAGE and can be useful in confirming protein identity. From Table 1 it is apparent that measured pI does not always correspond to the calculated pI value. This tendency is especially pronounced for some membrane proteins with calculated pI values of 8-9, which were expected to appear in fraction 19 (pI 8.8), but which were instead identified in fractions with much lower pIs. The reason for this unexpected behavior may be the presence of phospholipids or other charged groups associated with the membrane proteins. Also, some proteins could be identified in multiple Rotofor and or Gel Eluter fractions, indicating heterogeneity of pI and molecular weight, respectively. This phenomenon is frequently observed in proteins separated by 2D PAGE. 2152 Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
Figure 4. Product-ion spectrum of the doubly protonated fragment observed at m/z 931.6 from the tryptic digest of Rotofor fraction 7, Gel Eluter fraction 4, confirming the sequence YSELGNTYNSITTALSK previously matched with MALDI-MS data. Through MS/MS analysis, the identity of the protein was confirmed as the Leb-binding adhesin, a low-abundance membrane protein expressed by the H. pylori strain used in this investigation.
Enzymatic digestion, peptide mass mapping and database searches were performed on separated proteins from Gel Eluter fractions from four Rotofor fractions of differing pI values (Table 1). We identified forty different proteins expressed by this strain of H. pylori, fifteen of which are membrane- or membraneassociated proteins. Most of the proteins identified using this method have not been identified previously using 2D-PAGE as a separation method. Several of the identified outer membrane proteins may have binding affinities to host carbohydrates or proteins. The Leb-binding adhesin is the only lectin from H. pylori that has been fully characterized so far.27 The Leb-binding adhesin is a low-abundance protein, with an estimated number of 500 copies per cell.26 Because only four peptides in Rotofor fraction 7, gel eluter fraction 4 matched this protein, MS/MS analysis of two doubly charged peptides, m/z 931.6, YSELGNTYNSITTALSK (Figure 4, Table 2) and m/z 677.4, VPNAQSLQNVVGK was performed in the QTof. The extensive series of a, b, and y-ions obtained confirmed the sequences of both peptides studied and the identity of this protein. CONCLUSIONS The use of an approach for studying detergent-solubilized H. pylori, using liquid-phase IEF and electroelution in the gel eluter prior to mass spectrometry, was found to be a useful tool for (26) Puchades, M.; Westman, A.; Blennow, K.; Davidsson, P. Rapid Commun. Mass Spectrom. 1999, 13, 344-349. (27) Ilver, D.; Arnqvist, A.; O ¨ gren, J.; Frick, I.-M.; Kersulyte, D.; Incecik, E. T.; Berg, D. E.; Covacci, A.; Engstrand, L.; Bore´n, T. Science (Washington, D.C.) 1998, 279, 373-377.
identifying proteins of potential value in vaccine design. More than one-third of the identified proteins were membrane- or membraneassociated, but no pre-enrichment of bacterial membranes or complicated solubilization protocol was required. The nature of any binding affinities of most of the membrane proteins identified is as yet unkown, and will require further investigations. However, the well-characterized Leb-binding adhesin was identifiedsa protein expressed in a very low number of copies (approximately 500 per cell) by the strain studied. Advantages to using this novel approach are speed of analysis, nondiscrimination of hydrophobic proteins, and ability to monitor the presence of proteins by mass spectrometry in both Rotofor and gel-eluter fractions. Also, no pre-enrichment of bacterial membranes prior to IEF is necessary. The method lacks the advantage of analytical 2D PAGE of high-resolution separation; however, membrane proteins are difficult to analyze using 2D PAGE, which is also a more time-consuming procedure. The high proportion of membrane proteins identified in this study demon-
strates the method’s appropriateness in strategies requiring the identification of potential protein vaccine candidates. ACKNOWLEDGMENT Financial support for this study was provided by the Swedish Foundation for Strategic Research (Infection and Vaccinology), the Swedish Society of Medicine, the Go¨teborg Medical Society, the Wilhelm and Martina Lundgren Fund, the Knut and Alice Wallenberg Foundation, the Swedish Medical Research Council (Grant no. 12769, 13121), the Swedish Research Council for Engineering Sciences, the Inga Britt and Arne Lundberg Foundation, Lundbeckstiftelsen, Magnus Bergvalls Stiftelse, Thurings Stiftelse, and Åke Wibergs Stiftelse.
Received for review November 8, 1999. Accepted January 20, 2000. AC9912754
Analytical Chemistry, Vol. 72, No. 9, May 1, 2000
2153
