Sven Baumann, Uta Ceglarek, Georg Martin Fiedler, Jan Lembcke, Alexander Leichtle, and Joachim Thiery *

Clinical Chemistry 5:6 97 9 (5) Proteomics and Protein Markers Standardized Approach to Proteome Profiling of Human Serum Based on Magnetic Bead Separation and Matrix-Assisted Laser Desorption/Ionization Time-of-Flight Mass Spectrometry Sven Baumann, Uta Ceglarek, Georg Martin Fiedler, Jan Lembcke, Alexander Leichtle, and Joachim Thiery * Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics, University Hospital, Leipzig, Germany. These authors contributed equally to this work. *Address correspondence to this author at: Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics, University Hospital Leipzig, Liebigstrasse 7, Leipzig, Germany. Fax 9--979; e-mail thiery@medizin.uni-leipzig.de. Received December, ; accepted April, 5. Previously published online at DOI:.7/clinchem..7 Background: Magnetic bead purification for the analysis of low-abundance proteins in body fluids facilitates the identification of potential new biomarkers by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS). The aims of our study were to establish a proteome fractionation technique and to validate a standardized blood sampling, processing, and storage procedure for proteomic pattern analysis. Methods: We used magnetic bead separation for proteome profiling of human blood by MALDI-TOF MS (mass range, Da) and studied the effects on the quality and reproducibility of the proteome analysis of anticoagulants, blood clotting, time and temperature of sample storage, and the number of freeze thaw cycles of samples. Results: The proteome pattern of human serum was characterized by 5 signals in the mass range of Da. The proteome profile showed timedependent dynamic changes before and after centrifugation of the blood samples. Serum mass patterns differed between native samples and samples frozen once. The best reproducibility of proteomic patterns was with a single thawing of frozen serum samples. Conclusion: Application of the standardized preanalytical blood sampling and storage procedure in combination with magnetic bead-based fractionation decreases variability of proteome patterns in human serum assessed by MALDI-TOF MS. 5 American Association for Clinical Chemistry Proteomic pattern analysis by mass spectrometry (MS) is one of the most promising new approaches for the identification of potential blood biomarkers to assess health and disease (, ). The discovery of biomarkers in body fluids has been advanced by the recent introduction of MS-based screening methods such as surface-enhanced laser desorption/ionization time-of-flight (SELDI-TOF) MS and matrix-assisted laser desorption/ionization timeof-flight (MALDI-TOF) MS (). Techniques such as -dimensional gel electrophoresis and multidimensional liquid chromatography after MS analysis can remove signal-suppressing components of complex biological mixtures such as serum or cell extracts ( ), but these techniques are laborious and time-consuming. In a new pretreatment method, serum peptides are fractionated and concentrated on surface-modified targets with specific protein-capture properties. These targets can be used for TOF MS directly (SELDI-TOF MS) (9). The first clinical investigations using SELDI-TOF MS for different cancer types (e.g., ovarian, prostate, bladder, breast, lung, and brain) revealed high diagnostic sensitivities and specificities ( 6). However, for the same cancers, different patterns were identified by various research groups using the same types of biological specimens and the same analytical platform (7). These dis- Nonstandard abbreviations: MS, mass spectrometry; SELDI-TOF, surface-enhanced laser desorption/ionization time of flight; MALDI, matrixassisted laser desorption/ionization; MB-HIC, magnetic bead hydrophobic interaction chromatography; S/N, signal-to-noise; and apo, apolipoprotein. 97

97 Baumann et al.: Magnetic Bead Separation and MALDI-TOF MS of Serum crepancies could be attributable to poorly defined preanalytical protocols. For accurate MS analysis, the proteome fractionation procedure and the preanalytical conditions for proteome mapping must be carefully assessed (). The aim of this study was to establish a well-defined protocol for precise proteome analysis and preanalytical blood sample collection, storage, and processing procedures to minimize the variability in data obtained for the proteome pattern. We used a novel sample pretreatment procedure with surface-modified magnetic beads before identification of human serum proteome profiles using MALDI-TOF MS (9, ). Materials and Methods chemicals and calibrators Gradient-grade acetonitrile and ethanol were obtained from J.T. Baker; p.a. trifluoroacetic acid and acetone were purchased from Sigma-Aldrich. The peptide calibrator containing angiotensin II, the protein calibrator, and -cyano--hydroxycinnamic acid were purchased from Bruker Daltonics GmbH. For magnetic bead preparations, we used.-ml polypropylene tubes (-tube strips) from Biozym. Multifly needle sets and polypropylene Monovettes with and without anticoagulants (EDTA, heparin, and citrate) were obtained from Sarstedt. blood samples Blood samples from healthy volunteers ( females and males; age range, 6 9 years) were collected in serum, EDTA, heparin, and citrate Monovettes. Serum and plasma aliquots were stored in -ml polypropylene tubes (NeoLab). The blood samples were processed according to a standardized protocol. After sample collection, the Monovettes were incubated at room temperature (5 C) for min and centrifuged at g for min. The plasma and serum samples were immediately frozen in aliquots of L at C; for proteome fractionation, samples were thawed at room temperature for 5 min and processed immediately. Variations in sample pretreatment are described separately below. For precision and mass accuracy analyses, serum and plasma samples were pooled from individual samples (from females and males). proteome fractionation Proteome fractionation of the samples was performed with ClinProt purification reagent sets from Bruker Daltonics. Magnetic particles (particle size m) with defined hydrophobic surface functionalities [magnetic bead hydrophobic interaction chromatography (MB- HIC) C, C, and C], a mean pore size of nm, and a specific surface area of cm /g were used. As recommended in the manufacturer s protocol, we diluted 5 L of sample with L of a binding solution added to the bead slurry (5 L) in a.-ml polypropylene tube, mixed thoroughly, and incubated the tube for min. After magnetic bead separation and washings, we eluted the proteome fraction from the magnetic beads with 5 L of an acetonitrile water mixture (: by volume). We prepared targets (co-crystallization) by spotting L of a mixture containing L of. g/l -cyano-- hydroxycinnamic acid in ethanol acetone (: by volume) and L of the eluted proteome fraction on the Anchor- Chip TM target (Bruker Daltonics). ms For the proteome analysis, we used a linear MALDI-TOF mass spectrometer (Autoflex; Bruker Daltonics) with the following settings: ion source, kv; ion source,.5 kv; lens, 9. kv; pulsed ion extraction, ns; nitrogen pressure,.5 5 Pa. Ionization was achieved by irradiation with a nitrogen laser ( 7 nm) operating at 5 Hz. For matrix suppression, we used a high gating factor with signal suppression up to 5 Da. Mass spectra were detected in linear positive mode. Mass calibration was performed with the calibration mixture of peptides and proteins in a mass range of Da. We measured MALDI preparations (MALDI spots) from each magnetic bead fraction. For each MALDI spot, spectra were acquired ( laser shots at different spot positions). To increase the detection sensitivity, we removed excess matrix with 6 shots at a laser power of 5% before data acquisition at 5%. All signals with a signal-to-noise (S/N) ratio in a mass range of Da were recorded with use of the AutoXecute tool of the flexcontrol acquisition software (Ver..; Bruker Daltonics). We used the ClinProTools bioinformatics software (Ver..; Bruker Daltonics) for proteome pattern recognition. evaluation of the reproducibility of proteomic profiling using ms To investigate the efficiency and recovery of magnetic bead fractionation, we mixed the peptide calibration solution containing angiotensin II (75 g/l) with g/l albumin and with pooled serum. The MALDI-TOF MS (S/N ratio ) detection limits were determined for analysis of angiotensin II directly and after magnetic bead fractionation using MB-HIC C, C, and C beads. To evaluate the analytical recovery and variability, we analyzed bead preparations of the peptide/protein calibration mixture and the pooled serum samples on day (within-day variability) and on consecutive days (dayto-day variability), using MB-HIC C, C, and C beads. To determine the variation in the ratios, we compared relative and absolute peak intensities of selected signals of the pool serum with the signals obtained from the peptide/protein calibration mixture. evaluation of preanalytical conditions of blood sampling and storage The proteomic patterns of serum and plasma (EDTA, citrate, and heparin) were investigated in the pooled serum and plasma samples after fractionation with MB- HIC C, C, and C beads. We investigated the influence

Clinical Chemistry 5, No. 6, 5 975 of the blood collection tube material (polymeric polypropylene) by incubating the Monovettes with ml of an aqueous solution containing. mol/l NaCl for h at room temperature with gentle rocking. To investigate the influence of the blood clotting on the proteomic pattern, we incubated single blood samples at room temperature (5 C) for,, 6,, and min before centrifugation. One aliquot was immediately prepared after each time point for MALDI-TOF MS; the other aliquots were kept frozen at C until analysis. The effects of 5 freeze thaw cycles on the proteomic pattern were investigated. The effect of time and temperature on the proteomic pattern was investigated in the pooled serum samples. After centrifugation according to the standard protocol, serum aliquots were incubated at, 5, and C for, 6,, and min each. The samples were divided into aliquots; aliquot was immediately fractionated and analyzed by MALDI-TOF MS and the other was frozen at C and analyzed week later. Results reproducibility and mass accuracy of maldi-tof ms In the blood samples, magnetic bead fractionation using MB-HIC C (Fig. B), C (Fig. C), and C (Fig. D) beads allowed identification of 5 signals with a S/N ratio. Direct analysis of the pooled serum sample gave MALDI- TOF MS spectra without discrete mass signals (Fig. A). The reproducibility of mass spectrum generation was determined with respect to the relative peak intensities. Within-day CVs for direct MS analysis of the peptide/ protein calibration mixture (without magnetic bead fractionation) were % %, and between-day CVs were 5% % (data not shown). Shown in Table are the within- and between-day CVs of the relative peak intensities for signals of the peptide/protein calibration mixture and of the pooled serum samples representing low-, medium-, and high-abundance peptides over the whole mass range after magnetic bead fractionation with MB-HIC C, C, and C beads. Signals with relative signal intensities % gave within-day CVs % and between-day CVs 6%. We evaluated the mass accuracy of the MS measurements with magnetic bead preparations (MB-HIC C, C, and C) of the peptide/protein calibration mixture and the pooled serum sample, measured in run and on consecutive days. Maximum mass shifts over the range Da were.% for the peptide/protein calibration mixture and.5% for the pooled serum sample. These results correlate with a variation of Da in the absolute mass shift for signals 5 Da. The detection limit for angiotensin II after direct analysis of the peptide/protein calibration mixture was 6. ng/l. After magnetic bead fractionation (MB-HIC C, C, and C), the detection limit for angiotensin II in the calibration mixture containing g/l albumin was 6 Intensity x......5...5 * 6.. (B) 9. 66.6 6.5 7767.7 9.79 597.6 * (A) 6.. (C) 9. 66.6 6.5 7767.7 9.79 597.6. * 6.. (D) 66.6 9. 6.5 7767.7. 597.6 9.79.5. 5 7 9 Fig.. Comparison of serum proteome patterns. (A), direct analysis; (B), fractionation with MB-HIC C; (C), fractionation with MB-HIC C; (D), fractionation with MB-HIC C. ng/l. The detection limit increased to ng/l for angiotensin II after magnetic bead preparation of the pooled serum with angiotensin II added. evaluation of preanalytical processing of blood sampling and storage Compared with EDTA-, heparin-, and citrate-plasma samples (Fig., A C), the MS spectrum of serum (Fig., D and D) showed the highest number of signals with the highest intensities. MS spectra of the plasma samples were dominated by signals at., 7., and 7. Da (signals not shown). In addition, several signals (e.g., 66.6 and 9. Da) were observed in plasma and serum. The mass spectrum of heparin plasma was characterized by various signals in the small mass range Da. The number of signals in the mass spectra increased in platelet-rich citrate plasma (centrifugation at g for min; data not shown). Interferences from blood collection tube polymeric polypropylene materials (EDTA, heparin, and citrate tubes) were excluded. After magnetic bead separation, no

976 Baumann et al.: Magnetic Bead Separation and MALDI-TOF MS of Serum Table. Relative intensities and precision of signals from the peptide/protein calibration solution and serum pool (n ) in the mass range Da, obtained by MALDI-TOF MS after magnetic bead preparation. Peptide/protein calibration solution Pooled serum Within run Between runs Within run Between runs Signal, Intensity a CV, % Intensity a CV, % Signal, Intensity a CV, % Intensity a CV, % 7. 5.6. 5. 5.6 7.9 6. 5. 7 97.5...7.7 5.6 6. 5. 6. 5.6.66 7. 9.. 67.5. 7.9. 6.9.. 5.6 5.. 6. 7. 5.7 6.7 9.9 56..7 6.6 9.6.5.9 6 9.6... 66.7.5 9.5 7 9.7.7.7. 9.6 7. 6. 7. 95..6..5 6. 6...5.6 57.56..7.9 66.6.. 6 565.9 9. 9. 9.79. 5. 9 a The relative intensities (%) were calculated from the total sum of all intensities in the mass spectra between and Da. interfering signals with S/N ratios were obtained in the MALDI-TOF mass spectra in the mass range Da (data not shown). The relative intensities of the serum proteome signals changed with time after clotting (Fig. ). A loss of intensity with time was observed for low molecular-mass Intensity x.6..5. 9.6....5.6..6..57.5 67. 755.9 66.9 755.9 755.9 (A) (B) (C) Intensity x..........5 6.5 66.6 (A) 9.79 7.5 9.5 6.9 766. 767.6 6.5 66.6 6.9 767.6 6.5 66.6 6.9 767.6 (B) 9.79 7.5 9.5 766. (C) 9.79 7.5 9.5 766.. (D) 556.59 6. 6.9.6 67.. 77.5.67 9.6. 67.5 9.. 6. 6.... 6.5 66.6 756.5 7767.7 766. 99.5 6. (D) 9.79 65 7 75 5 9 95 Fig.. Mass spectra of plasma obtained with different anticoagulants after MB-HIC C bead fractionation and MALDI-TOF analysis. (A and A), citrate; (B and B), EDTA; (C and C), heparin; (D and D), serum.

Clinical Chemistry 5, No. 6, 5 977 relative intensity % 6 min min 6 min min min 5.6 6.. Fig.. Variation in relative intensities for characteristic signals of the pool serum sample after MB-HIC C fractionation depending on clotting time. signals (e.g., 5. Da). Signals with masses Da (e.g.,. Da) increased in relative intensity with time, but signals with lower masses (e.g., 6. Da) did not change in relative intensity. In samples allowed to clot for min at room temperature, the background noise increased considerably (mass range Da). We compared the mass spectra of unfrozen serum aliquots with the spectra of the same samples after freezing. Unfrozen serum showed a high degree of variation in relative signal intensities, and reproducibility was poor. As shown in Fig., freezing thawing markedly decreased the intensities of proteins/peptides at 5.5, 55.5, and 56. Da (panel A, no freezing; panel B, freeze thaw cycle; panel C, freeze thaw cycles; panel D, 5 freeze thaw cycles). The same decrease was observed for signals in the mass range Da (data not shown). Freezing thawing did not affect signals for proteins/peptides 5 Da (e.g., 67. Da). In contrast, signals above 9 Da (e.g., 96., 9., and 9. Da) increased in relative peak intensity after only freeze thaw cycle (Fig. A). Recurrent freeze thaw cycles further decreased the intensity of low molecular-mass signals (e.g., 5.5, 55.5, and 56. Da) as shown in panels B, C, and D of Fig.. However, proteins/peptides with high molecular masses (96., 9., and 9. Da) were remarkably stable and were not influenced by recurrent freeze thaw cycles (Fig., panel B, freeze thaw cycle; panel C, freeze thaw cycles; panel D, 5 freeze thaw cycles). Intensity x 67.5 67.5 67.5 67.5 9. 5.67 55. 5.6 56. 55. 5.6 9. 5.67 56. 5.6 5.67 9. 56. 5.6 5.67 56. (A) (B) (C) (D) 6 5 5 5 56 5 6 Intensity x 6 767.5 9.79 6 767.5 9.79 6 767.5 9.79 6 767.5 9.79 9.79 99.5 99.5 99.5 99.5 9.79 (D) 9.79 9 9 9 (A) (B) (C) 9.79 Fig.. Influence of freeze thaw cycles on the signal intensities in the pooled serum after MB-HIC C bead fractionation and MALDI-TOF analysis. (A and A), unfrozen; (B and B), freeze thaw cycle; (C and C), freeze thaw cycles; (D and D), 5 freeze thaw cycles.

97 Baumann et al.: Magnetic Bead Separation and MALDI-TOF MS of Serum The influence of time and temperature on the proteomic pattern is shown Fig. 5. The pattern was stable and showed little degradation after min at C and did not change for up to min at 5 C. At C, however, there was significant loss of intensity for all signals in the proteome profile after only min. During a 7-day storage period at C, the relative peak intensities and the respective S/N ratios decreased after h for masses Da. Signals at 9. and 6. Da were also affected. In contrast, we observed no change for signals with a molecular mass Da. Discussion Protein profiling is a promising tool for the discovery and subsequent identification of proteins and peptides associated with various diseases (, ). However, knowledge is still limited regarding the sensitivities and reproducibilities of various techniques for analyzing the wide dynamic range of peptides and proteins in blood (). The aim of our study was to establish a valid, standardized approach to preanalytical blood sample processing. We found that Intensity x 67.5 6. 565.99 7767.7 555.5. 66.6 (B) 67.5 6. 565.99 7767.7 555.5. 66.6 (C) 67.5 6. 565.99 7767.7 555.5. 66.6 (A) (D) 6. 565.99 7767.7 67.5. 66.6 555.5 5 7 9 Fig. 5. Influence of storage temperature on signal intensities after MB-HIC C bead fractionation and MALDI-TOF analysis of the pooled serum sample. (A), treated according to the standard protocol; (B), min at C; (C), min at 5 C; (D), min at C. protein profiling by MALDI-TOF MS after proteome fractionation with magnetic beads is a robust, precise, and rapid technique for the investigation of complex blood samples. We found 5 signals in human serum in the mass range Da. Previous studies identified several peptides and protein fragments by use of electrospray ionization tandem MS and MALDI-TOF, and these methods may help to assign probable identities to the proteins in our spectra. Plasma protein fragments at (albumin, 5 5) and 67 (fibrinogen A, 6 6) were identified in human hemofiltrate signals after ion-exchange chromatography and MALDI-TOF MS by Richter et al. (). Bondarenko et al. () investigated the potential of MALDI and electrospray ionization MS for screening of polymorphisms in the apolipoproteins associated with VLDL after solid-phase extraction (C modification) of the isolated and delipidated lipoproteins. They identified the major apolipoprotein C (apoc) classes in the VLDL (apoc-i, 6.; apoc-i, 66.6; apo procii, 9.9; and apoc-iii, 9.). We detected signals with the same ratios in our human serum samples (Fig., B D; Fig., D). The variability of relative signal intensities in our study was in the range of the previously published data from Zhang et al. (9). They selected high-, medium-, and low-abundance signals in the mass range Da and calculated that the variability of each peak area was % 5% (9). The intraassay variability in our study was % %, and interassay variability was 6% %. The higher reproducibility in our study was achieved by -fold MALDI spotting of each magnetic bead fractionation product. We determined mass accuracy by analyzing serum samples after magnetic bead fractionation on day (n ) and on consecutive days. The mean mass shift of.5% led to absolute mass differences between.5 and.5 Da in the mass range Da. These absolute mass differences are in accordance with the known mass accuracy of MALDI-TOF MS (). We solved the problem of mass shift by using the most prominent peaks to recalibrate the spectra with the ClinProTools bioinformatics software (9). These findings support the robustness and precision of the magnetic-bead based pretreatment technology. Semmes et al. (5) investigated the mass accuracy and reproducibility of serum protein profiling, using SELDI-TOF MS in an interlaboratory study (n 6). They found mean intra- and interlaboratory mass accuracy differences of.% and a 5% 6% variation in relative signal intensities for the most prominent mass signals with S/N ratios. Maximum interlaboratory mass differences of Da were obtained for a mean mass signal of 997 Da. The high sensitivity and reproducibility of the novel magnetic-bead based platform for proteome profiling in our study is supported by a recent study from Villanueva et al. (). They reported that the use of porous particles for sample pretreatment is more sensitive than surface

Clinical Chemistry 5, No. 6, 5 979 capture on chips because spherical particles have larger combined surface areas than small-diameter spots. Using the magnetic bead technology, they found different proteome signals in the mass range 5 Da in serum. For our study, we used comparable beads with the same specifications, but the mass range was smaller ( Da) and produced 5 proteome signals. In our study, because of the high adsorptive capacity of the magnetic beads, the detection limit for the low molecularmass peptide angiotensin II in serum was ng/l after magnetic bead fractionation. The different sensitivities for the detection of angiotensin II added to serum and to aqueous albumin solution could be caused by competition effects during the magnetic bead fractionation of additional low and high molecular-mass components in serum. In addition, ionization suppression effects of other serum components may influence the ionization of angiotensin II added to serum. Despite these limitations, the application of magnetic bead fractionation in combination with MALDI-TOF is appropriate for the detection of low concentrations of proteins and peptides in serum (6). Use of plasma entails the possibility of residual platelet contamination, which could influence the mass spectrum. To achieve reproducible proteome profiling and accurate data interpretation, we evaluated preanalytical conditions of blood sampling and storage (). We found no interfering signals from the polymeric polypropylene materials in the blood collection tubes (EDTA, heparin, citrate, and serum tubes) in the mass range Da. In contrast, Drake et al. (7) described interfering signals in the mass range Da from polymeric materials in blood collection tubes. In the present study, however, we used a blood collection system from a different manufacturer and a different sample pretreatment procedure. Our study results showed that anticoagulants, temperature, and freeze thaw cycles have a significant impact on dynamic alterations of the serum proteome in the mass range Da. In a recent study, Schaub et al. () investigated preanalytical factors that can modify urine proteome profiling by SELDI-TOF MS. They studied the influence of extrinsic (instrument set-up, storage, freeze thaw cycles) and intrinsic factors (blood in urine, urine dilution, first void vs midstream urine) on urine values. They found that up to freeze thaw cycles with freezing at 7 C did not alter the urine protein profile but that variation in intrinsic factors caused significant differences. Urine, however, is characterized by high amounts of polar low molecular-mass components and an excess salt content; therefore, direct comparison of preanalytical variables for urine and blood samples is limited. Proteome profiling using magnetic bead-based proteome fractionation MALDI-TOF MS is a promising screening tool for the identification of specific proteome patterns. It is difficult, however, to identify the proteins and peptides or to quantify the interesting analytes directly by MALDI-TOF MS technology (7). The relative peak intensities can be influenced by ion suppression effects and do not necessarily reflect the concentrations of the detected proteins or peptides. Therefore, additional specific chromatographic separation techniques and mass spectrometric platforms (e.g., liquid chromatography with electrospray ionization tandem MS) are necessary for the further identification and quantification of potential biomarkers (, 7). Our results show that magnetic bead fractionation in combination with MALDI-TOF MS is a highly sensitive and reproducible analytical platform for proteome profiling of human blood in the mass range Da. Use of our evaluated procedure could reduce preanalytical variables; the procedure involves (a) blood sample collection in polypropylene tubes; (b) incubation of the tubes at room temperature for min to allow coagulation; (c) serum separation by centrifugation ( min at g) at room temperature; (d) separation of - L aliquots in small polypropylene tubes for immediate freezing at C (no changes in the proteome pattern were observed for a storage period of at least 6 months); and (e) thawing of frozen aliquots for 5 min at room temperature and immediate processing by MB-HIC proteome fractionation. This work was supported by a grant from the Sächsische Aufbaubank (SAB) and by a grant of the Deutsche Forschungsgemeinschaft, Th 7/-. We thank Drs. Markus Kostrzewa and Thomas Elssner (Bruker Daltonics, Leipzig, Germany) for helpful advice and discussions. References. Etzioni R, Urban N, Ramsey S, McIntosh M, Schwartz S, Reid B, et al. The case for early detection. Nat Rev Cancer ;: 5.. Aldred S, Grant MM, Griffiths HR. The use of proteomics for the assessment of clinical samples in research. Clin Biochem ; 7:9 5.. Banks RE, Dunn MJ, Hochstrasser DF, Sanchez JC, Blackstock W, Pappin DJ, et al. Proteomics: new perspectives, new biomedical opportunities. Lancet ;56:79 56.. Adkins JN, Varnum SM, Auberry KJ, Moore RJ, Angell NH, Smith RD, et al. Toward a human blood serum proteome: analysis by multidimensional separation coupled with mass spectrometry. Mol Cell Proteomics ;:97 55. 5. Washburn MP, Wolters D, Yates JR rd. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat Biotechnol ;9: 7. 6. Pieper R, Gatlin CL, Makusky AJ, Russo PS, Schatz CR, Miller SS, et al. The human serum proteome: display of nearly 7 chromatographically separated protein spots on two-dimensional electrophoresis gels and identification of 5 distinct proteins. Proteomics ;:5 6. 7. Chan KC, Lucas DA, Hise D, Schaefer CF, Xiao Z, Janini GM, et al. Analysis of the human serum proteome. Clin Proteomics ;: 5.. Peng J, Elias JE, Thoreen CC, Licklider LJ, Gygi SP. Evaluation of multidimensional chromatography coupled with tandem mass spectrometry (LC/LC-MS/MS) for large-scale protein analysis: the yeast proteome. J Proteome Res ;: 5. 9. Isaaq HJ, Veenstra TD, Conrads TP, Felschow D. The SELDI-TOF

9 Baumann et al.: Magnetic Bead Separation and MALDI-TOF MS of Serum MS approach to proteomics: protein profiling and biomarker identification. Biochem Biophys Res Commun ;9:57 9.. Diamandis EP. Proteomic patterns in serum and identification of ovarian cancer [Letter]. Lancet ;6:7.. Guillaume E, Zimmermann C, Burkhard PR, Hochstrasser DF, Sanchez JC. A potential cerebrospinal fluid and plasmatic marker for the diagnosis of Creutzfeldt-Jakob disease. Proteomics ; :95 9.. Howard BA, Wang MZ, Campa MJ, Corro C, Fitzgerald MC, Patz EF Jr. Identification and validation of a potential lung cancer serum biomarker detected by matrix-assisted laser desorption/ionization-time of flight spectra analysis. Proteomics ;:7.. Koopmann J, Zhang Z, White N, Rosenzweig J, Fedarko N, Jagannath S, et al. Serum diagnosis of pancreatic adenocarcinoma using surface-enhanced laser desorption and ionization mass spectrometry. Clin Cancer Res ;:6.. Menon U, Jacobs I. Screening for ovarian cancer. Best Pract Res Clin Obstet Gynaecol ;6:69. 5. Petricoin EF, Ardekani AM, Hitt BA, Levine PJ, Fusaro VA, Steinberg SM, et al. Use of proteomic patterns in serum to identify ovarian cancer. Lancet ;59:57 7. 6. Zheng PP, Luider TM, Pieters R, Avezaat CJ, van den Bent MJ, Sillevis Smitt PA, et al. Identification of tumor-related proteins by proteomic analysis of cerebrospinal fluid from patients with primary brain tumors. J Neuropathol Exp Neurol ;6:55 6. 7. Diamandis EP. Mass spectrometry as a diagnostic and a cancer biomarker discovery tool: opportunities and potential limitations. Mol Cell Proteomics ;:67 7.. Omenn GS. The Human Proteome Organization Plasma Proteome Project pilot phase: reference specimens, technology platform comparisons, and standardized data submissions and analyses. Proteomics ;:5. 9. Zhang X, Leung SM, Morris CR, Shigenaga MK. Evaluation of a novel, integrated approach using functionalized magnetic beads, bench-top MALDI-TOF-MS with prestructured sample supports, and pattern recognition software for profiling potential biomarkers in human plasma. J Biomol Tech ;5:67 75.. Villanueva J, Philip J, Entenberg D, Chaparro CA, Tanwar MK, Holland EC, et al. Serum peptide profiling by magnetic particleassisted, automated sample processing and MALDI-TOF mass spectrometry. Anal Chem ;76:56 7.. Pusch W, Flocco MT, Leung S-M, Thiele H, Kostrzewa M. Mass spectrometry-based clinical proteomics. Pharmacogenomics ;:.. Richter R, Schulz-Knappe P, Schrader M, Standker L, Jurgens M, Tammen H, et al. Composition of the peptide fraction in human blood plasma: database of circulating human peptides. J Chromatogr B Biomed Sci Appl 999;76:5 5.. Bondarenko PV, Cockrill SL, Watkins LK, Cruzado ID, Macfarlane RD. Mass spectral study of polymorphism of the apolipoproteins of very low density lipoprotein. J Lipid Res 999;:5 55.. Mengel-Jorgensen J, Sanchez JJ, Borsting C, Kirpekar F, Morling N. MALDI-TOF mass spectrometric detection of multiplex single base extended primers. A study of 7 Y-chromosome single-nucleotide polymorphisms. Anal Chem ;76:69 5. 5. Semmes OJ, Feng Z, Adam BL, Banez LL, Bigbee WL, Campos D, et al. Evaluation of serum protein profiling by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry for the detection of prostate cancer: I. Assessment of platform reproducibility. Clin Chem 5;5:. 6. Tirumalai RS, Chan KC, Prieto DA, Issaq HJ, Conrads TP, Veenstra TD. Characterization of the low molecular weight human serum proteome. Mol Cell Proteomics ;:96. 7. Drake SK, Bowen RA, Remaley AT, Hortin GL. Potential interferences from blood collection tubes in mass spectrometric analyses of serum polypeptides. Clin Chem ;5:9.. Schaub S, Wilkins J, Weiler T, Sangster K, Rush D, Nickerson P. Urine protein profiling with surface-enhanced laser-desorption/ ionization time-of-flight mass spectrometry. Kidney Int ;65:. 9. Hood L, Heath JR, Phelps ME, Lin B. Systems biology and new technologies enable predictive and preventative medicine. Science ;6:6.