Packed Capillary Reversed-Phase Liquid Chromatography with High

Yufeng Shen, Rui Zhao, Mikhail E. Belov, Thomas P. Conrads, Gordon A. Anderson, Keqi Tang,. Ljiljana Pasˇa-Tolic´ , Timothy D. Veenstra, Mary S. Lip...
9 downloads 0 Views 218KB Size
Download PDF
Anal. Chem. 2001, 73, 1766-1775

Packed Capillary Reversed-Phase Liquid Chromatography with High-Performance Electrospray Ionization Fourier Transform Ion Cyclotron Resonance Mass Spectrometry for Proteomics Yufeng Shen, Rui Zhao, Mikhail E. Belov, Thomas P. Conrads, Gordon A. Anderson, Keqi Tang, Ljiljana Pasˇa-Tolic´, Timothy D. Veenstra, Mary S. Lipton, Harold R. Udseth, and Richard D. Smith*

Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, Washington 99352

In this study, high-efficiency packed capillary reversedphase liquid chromatography (RPLC) coupled on-line with high-performance Fourier transform ion cyclotron resonance (FTICR) mass spectrometry has been investigated for the characterization of complex cellular proteolytic digests. Long capillary columns (80-cm) packed with small (3-µm) C18 bonded particles provided a total peak capacity of ∼1000 for cellular proteolytic polypeptides when interfaced with an ESI-FTICR mass spectrometer under composition gradient conditions at a pressure of 10 000 psi. Large quantities of cellular proteolytic digests (e.g., 500 µg) could be loaded onto packed capillaries of 150-µm inner diameter without a significant loss of separation efficiency. Precolumns with suitable inner diameters were found useful for improving the elution reproducibility without a significant loss of separation quality. Porous particle packed capillaries were found to provide better results than those containing nonporous particles because of their higher sample capacity. Twodimensional analyses from the combination of packed capillary RPLC with high-resolution FTICR yield a combined capacity for separations of >1 million polypeptide components and simultaneously provided information for the identification of the separated components based upon the accurate mass tag concept previously described. The availability of whole genome sequence databases allows the entire spectrum of potential proteins (i.e., the proteome) that may be expressed under various conditions to be identified.1-3 Rapid and quantitative analysis of proteins expressed by microorganisms and cells in response to an environmental “perturbation” can provide a powerful tool for deducing the functions of key proteins and, more importantly, allowing the complex molecular pathways and networks involved in cellular responses to (1) Wilkins, M. R., Williams, K. L., Appel, R. D., Hochstrasser, D. F., Eds. Proteome Research: New Frontiers in Functional Genomics; Springer: Berlin, Germany, 1997. (2) Devine, K. M.; Wolfe, K. Trends Genet. 1995, 11, 429-431. (3) http://www.ebi.ac.uk/research/cgg/genomes.html.

1766 Analytical Chemistry, Vol. 73, No. 8, April 15, 2001

be unraveled.1,4,5 Proteome applications would benefit from experimental methods that can reliably identify and quantify changes in the expression of thousands of proteins in a single experiment. Most proteome-level analyses are currently based upon twodimensional polyacrylamide gel electrophoresis (2D-PAGE).1,6-8 The impressive separation power (resolving >1000 protein “spots”) provided by 2D-PAGE combined with mass spectrometric (MS) analysis makes this strategy very attractive for protein identification.9-14 The fact that the number of spots observed in 2D-PAGE displays is often of the same magnitude as estimates of the number of proteins predicted for a given organism or tissue sample has led some to suggest that a substantially complete view of proteomes could be achieved. However, there are well-known reasons why 2D-PAGE displays are incomplete (e.g., the limited solubility of hydrophobic and membrane proteins, difficulties in focusing highly basic and acidic proteins). The complexity of spot patterns arises not only from posttranslationally modified forms of proteins but also from degradation intermediates, alternative expression products (alternative splicing of mRNAs, frame shifts during translation etc.), and likely other reasons that include artifacts of the 2D-PAGE separation itself (e.g., reaction with residual acrylamide monomer, partial retention of some protein structure). Aebersold and co-workers have recently shown that (4) VanBogelen, R. A.; Greis, K. D.; Blumenthal, R. M.; Tani, T. H.; Matthews, R. G. Trends Microbiol. 1999, 7, 320-328. (5) Anderson, L.; Anderson, N. G. Electrophoresis 1998, 19, 1853-1861. (6) Go ¨rg, A.; Obermaier, C.; Boguth, G.; Harder, A.; Scheibe, B.; Wildgruber, R.; Weiss, W. Electrophoresis 2000, 21, 1037-1053. (7) Harry, J.; Wilkins, M. R.; Herbert, B. R.; Packer, N. H.; Gooley, A. A.; Williams, K. L. Electrophoresis 2000, 21, 1071-1081. (8) Patton, W. F. Electrophoresis 2000, 21, 1123-1144. (9) Henzel, W. J.; Billeci, T. M.; Stults, J. T.; Wong, S. C.; Grimley, C.; Watanabe, C. Proc. Natl. Acad. Sci. U.S.A. 1993, 30, 5011-5015. (10) Shevchenko, A.; Jensen, O. N.; Podtelejnikov, A. V.; Sagliocco, F.; Wilm, M.; Vorm, O.; Mortensen, P.; Shevchenko, A.; Boucherie, H.; Mann, M. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 14440-14445. (11) Alam, S. L.; Atkins, J. F.; Gesteland, R. F. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 14177-14179. (12) Gygi. S. P.; Han, D. K. M.; Gingras, A-.C.; Sonenberg, N.; Aebersold, R. Electrophoresis 1999, 20, 310-319. (13) Gauss, C.; Kalkum, M.; Lo¨we, M.; Lehrach, H.; Klose, J. Electrophoresis 1999, 20, 575-600. (14) Gevaert, K.; Vandekerckhove, J. Electrophoresis 2000, 21, 1145-1154. 10.1021/ac0011336 CCC: $20.00

© 2001 American Chemical Society Published on Web 03/16/2001

only the higher abundance proteins are observed by the combined 2D-PAGE MS strategy that now dominates the practice of proteomics, and for these reasons, more than half of all yeast proteins are not accessible to study by global 2D-PAGE MS strategies.15 Clearly, the number of spots visualized provides no assurance that broad proteome coverage has been achieved, nor is it a measure of the intrinsic complexity of a given proteome. In recognition of these challenges, some efforts are already underway to develop advanced proteomic technologies that are not based upon 2D-PAGE. Promising approaches include those using combinations of liquid chromatography (LC) with various forms of MS. For example, Yates and co-workers16 are developing methods for analyzing global protein digests with 2D-capillary LC coupled with tandem mass spectrometry (i.e., MS/MS). This approach should more fully exploit the sensitivity achievable with conventional mass spectrometers (roughly ∼10-16 mol as opposed to ∼10-14 mol in conjunction with 2D-PAGE) allowing many additional proteins to be identified. In our laboratory, we are evaluating new approaches capable of analyzing global proteolytic digests using a single dimension of high-resolution capillary LC combined with Fourier transform ion cyclotron resonance (FTICR) mass spectrometry.17 The ability to analyze highly complex polypeptide (e.g., hundreds of thousands) mixtures by combining high-resolution separations with the high-sensitivity (better than 10-17 mol), high-resolution, and, of particular utility, high-mass measurement accuracy (MMA) of FTICR MS permits exploitation of the new concept of “accurate mass tags”.18-21 In this strategy, measurements for many peptides are obtained in each spectrum, providing a methodology that enables high dynamic range measurements of proteomes in a high-throughput manner (i.e., without the need for routine MS/MS). Such developments, along with the use of global stable isotope-labeling strategies19,22 offer the promise of sensitive, rapid, and quantitative global measurements of proteomes that are far more comprehensive than those achievable using 2D-PAGE. The various on-line couplings of capillary column separations with MS can greatly improve ESI-MS sensitivity by decreasing the solvent flow to microliter per minute levels and below.23,24 Enhanced separation efficiency or resolution also makes it possible to detect low-abundance components by resolving them from highabundance components. Among various liquid-phase capillary separation/MS technologies such as capillary zone electrophoresis (15) Gygi. S. P.; Corthals, G. L. Zhang, Y. Rochon, Y.; Aebersold, R. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 9390-9395. (16) Link, A. J.; Eng, J.; Schieltz, D. M.; Carmack, E.; Mize, G. J.; Morris, D. R.; Garvik, B. M.; Yates, J. R. Nat. Biotechnol. 1999, 17, 676-682. (17) Comisarow. M. B.; Marsall, A. G. Chem. Phys. Lett. 1974, 25, 282-283. (18) Bruce, J. E.; Anderson, G. A.; Wen, J.; Harkewicz, R.; Smith, R. D. Anal. Chem. 1999, 71, 2595-2599. (19) Pasˇa-Tolic´, L.; Jensen, P. K.; Anderson, G. A.; Lipton, M. S.; Peden, K. K.; Martinovic´, S.; Tolic´, N.; Bruce, J. E.; Smith, R. D. J. Am. Chem. Soc. 1999, 121, 7949-7950. (20) Masselon, C.; Anderson, G. A.; Harkewicz, R.; Bruce, J. E.; Pasˇa-Tolic´, L.; Smith, R. D. Anal. Chem. 2000, 72, 1918-1924. (21) Conrads, T. P.; Anderson, G. A.; Veenstra, T. D.; Pasˇa-Tolic´, L.; Smith, R. D. Anal. Chem. 2000, 72, 3349-3354. (22) Gygi. S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Nat. Biotechnol. 1999, 17, 994-999. (23) Hopfgartner, G.; Bean, K.; Henion, J.; Henry, R. J. Chromatogr. 1993, 647, 51. (24) Hopfgartner, G.; Wachs, T.; Bean, K.; Henion, J. Anal. Chem. 1993, 65, 439-446.

(CZE)/MS,25,26 capillary isoelectric focusing (CIEF)/MS,27,28 capillary isotachophoresis (CITP)/MS,29,30 capillary LC/MS,31,32 and capillary electrochromatography (CEC)/MS,33,34 high peak capacities (e.g., >100) can most readily be achieved for peptide separations using mobile-phase composition gradient reversedphase liquid chromatography (RPLC). Additionally, relatively large amounts of sample can be introduced to capillary RPLC columns without a significant loss of separation efficiency. In this paper, practical considerations for obtaining highefficiency capillary RPLC separations of global cellular tryptic digests have been investigated. We demonstrate that extremely high resolution 2D separations of complex cellular proteolytic digests can be obtained by the on-line coupling of capillary RPLC with FTICR mass spectrometry. EXPERIMENTAL SECTION Cellular Protease Digest Sample Preparations. Deinococcus radiodurans cells were grown in TGY medium to an OD600 of ∼1.2 and harvested by centrifugation at 10000g at 4 °C. Prior to lysis, cells were resuspended and washed three times with 100 mM ammonium bicarbonate and 5 mM EDTA (pH 8.4). Cells were lysed by bead beating (0.3-mm beads) using three 1-min cycles, allowing a 5-min cooling on ice between cycles. The supernatant containing soluble cytosolic proteins was recovered by centrifugation at 15000g for 15 min. Proteins were denatured and reduced by addition of guanidine hydrochloride (6 M) and DTT (1 mM), respectively, and boiled for 5 min. Prior to trypsin digestion, the protein sample was desalted using a 5000 Da molecular mass cutoff D-Salt gravity column (Pierce, Rockford, IL) equilibrated in 100 mM ammonium bicarbonate (pH 8.4). Proteins were enzymatically digested using sequencing grade modified trypsin (Promega, Madison, WI) (trypsin/protein, 1:50, w/w) at 37 °C for 16 h. Packed Capillary RPLC Experiments. Packed capillary columns were made “in-house”. One end of the fused-silica capillary (150 µm i.d. × 360 µm o.d., Polymicro Technologies, Phoenix, AZ) was connected to a zero dead-volume union (∼16 nL of void volume, Valco Instruments, Houston, TX) using PEEK tubing (380-µm i.d., Upchurch Scientific, Oak Harbor, WA) to position a steel screen having 2-µm pores (Valco). The other end of the capillary was connected to a steel vessel, in which bonded packing materials was introduced. Using a HPLC grade (Aldrich, Milwaukee, WI) solvent mixture of acetonitrile (ACN)/H2O (90: 10, v/v), the particles in the steel vessel were packed into the (25) Smith, R. D.; Olivares, J. A.; Nguyen, N. T.; Udseth, H. R. Anal. Chem. 1988, 60, 436-441. (26) Moseley, M. A.; Deterding, L. J.; Tomer, K. B.; Jorgenson, J. W. Anal. Chem. 1991, 63, 109-114. (27) Yang, L.; Lee. C. S.; Hofstadler, S. A.; Smith, R. D. Anal. Chem. 1998, 70, 4945-4950. (28) Jensen, P. K., Pasˇa Tolic´, L., Anderson, G. A., Horner, J. A., Lipton, M. S., Bruce, J. E.; Smith, R. D. Anal. Chem. 1999, 71, 2078-2084. (29) Udseth, H. R.; Loo, J. A.; Smith, R. D. Anal. Chem. 1989, 61, 228-232. (30) Smith, R. D.; Fields, S. M.; Loo, J. A.; Barinaga, C. J.; Udseth, H. R.; Edmonds, C. G. Electrophoresis 1990, 11, 709-717. (31) Abian, J.; Oosterkamp, A. J.; Gelpı´, E. J. Mass Spectrom. 1999, 34, 244254. (32) Huang, P.; Wall, D. B.; Parus, S.; Lubman, D. M. J. Am. Soc. Mass Spectrom. 2000, 11, 127-135. (33) Ding, J.; Vouros, P. Anal. Chem. 1997, 69, 379-384. (34) Que, A. H.; Konse, T.; Baker, A. G.; Novotny, M. V. Anal. Chem. 2000, 72, 2703-2710.

Analytical Chemistry, Vol. 73, No. 8, April 15, 2001

1767

Figure 1. Packed capillary LC arrangement used in this study. The system was operated at constant pressure and the mobile-phase composition gradient is generated by switching mobile phase from A to B. The inset shows the gradient speeds with split flow rates of 20 and 40 µL/min. Details for operation of the high-pressure (>8000 psi) sample injection: when injecting the sample with a syringe, close V1A and open V1B, 2, 3; after injection, close V1B, V2 and open V1A for injection of the sample into the separation capillary (the injected volume depends the flow rate and injecting time); open V2 to wash away the remaining sample and close valve 3. Both separation capillary column and split line are protected using guard columns (4 cm × 1.58 mm o.d. × 381 µm i.d. stainless steel tubing containing the same bonded phase as used in the separation capillary). For separation pressures e8000 psi, switching valves were used for the mobile-phase gradient and sample introductions.

capillary by gradually increasing the pressure from 500 to 10 000 psi using an Isco LC pump (model 100DM, Isco, Lincoln, NE). The packed capillary was conditioned in an ultrasonic bath for ∼5 min and then depressured overnight. Packed capillary LC separations were carried out at constant pressure using the arrangement schematically illustrated in Figure 1. The mobile phases were stably delivered at the maximal pressure (10 000 psi) of the two Isco pumps (model 100DM, Isco) and were combined in a steel mixer (∼2.8 mL) containing a magnetic stirrer before entering the separation capillary. Fusedsilica capillary flow restrictors (30-µm i.d. with various lengths) were used to control the gradient development speed. The inset to Figure 1 shows the gradient speeds at split flow rates of 20 and 40 µL/min. A simple “off-line” sample introduction arrangement was used for sample loading at 10 000 psi (see the inset to Figure 1) while a switch valve injector (Valco) was used for operation pressures of e8000 psi. For packed capillary RPLC/ UV experiments, the separation was monitored at 215 nm (Spectra 100 UV/visible detector, Spectra-Physics, San Jose, CA). A detection window was made in a 75 µm i.d × 360 µm o.d. fusedsilica capillary by burning off the outer polyimide coating. The distance between detection window and column outlet was ∼2.5 cm. The dead volume between the column outlet and the detection 1768

Analytical Chemistry, Vol. 73, No. 8, April 15, 2001

point (including the union and the connection tubing) is ∼1.5% of the column void volume, assuming a column dimension of 150 µm i.d × 80 cm with a column porosity of ∼0.5. Packed Capillary LC/ESI FTICR MS. The present study used a 3.5-T FTICR mass spectrometer with an ESI interface composed of a heated metal capillary inlet coupled to an electrodynamic ion funnel assembly with a downstream quadrupole for collisional ion focusing.35,36 The ESI interface was followed by an interface for external selective ion accumulation, an electrostatic ion guide, and a cylindrical dual-cell combination. The external accumulation interface incorporates three quadrupoles, referred to as “ion guiding”, “selection”, and “accumulation” quadrupoles, respectively.37 The accumulation quadrupole was segmented to provide an axial electric field for prompt ion ejection. To maximize the duty cycle of LC-FTICR operation (i.e., spectrum acquisition rate), the accumulation quadrupole exit plate was maintained at high trapping potential to continuously accumulate ions from the (35) Belov, M. E.; Gorshkov, M. V.; Udseth, H. R.; Anderson, G. A.; Tolmachev, A. V.; Prior, D. C.; Harkewicz, R.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2000, 11, 19-23. (36) Belov, M. E.; Gorshkov, M. V.; Anderson, G. A.; Udseth, H. R.; Smith, R. D. Anal. Chem. 2000, 72, 2271-2279. (37) Belov, M. E.; Nikolaev, E. N.; Anderson, G. A.; Auberry, A. J.; Harkewicz, R.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2001, 12, 38-48.

Figure 2. Packed capillary RPLC sample capacity evaluation for D. radiodurans tryptic digests. Conditions: 100 cm × 150 µm i.d. fused-silica capillary packed with 5-µm C18-bonded particles (300-Å pores, Jupiter); constant pressure operation at 8000 psi; sample injection volume of 10 µL with concentrations of (A) 5, (B) 12.5, and (C) 50 mg/mL; after eluting the large unretained peak (e.g., due to polymeric contaminants), start gradient (marked in the figure) from mobile phase A (H2O, 0.1% TFA, v/v) to 75% B (H2O/ACN, 10:90, 0.1% TFA, v/v) in 180 min; UV detection at 215 nm. Vertical axis is relative UV adsorption magnitude. SG: point to start gradient.

ESI source, except for a short period of time (∼400 µs) required to eject ions to the FTICR cell. Given the duration for each spectrum acquisition of 2.5 s, the duty cycle for ion accumulation throughout the separation was close to 100%. Ions ejected from the accumulation quadrupole were dynamically trapped in the FTICR cell. A tapered 2.5-cm fused-silica capillary tubing (75 µm i.d. × 198 µm o.d. fixed with 203-µm-i.d PEEK tubing, Upchurch) was used as connection tubing between the ESI source and the separation capillary column outlet. The distance between the ESI tip (emitter) and the inlet capillary was ∼1.5 mm, and a voltage of 2 kV was used for ESI. RESULTS AND DISCUSSION Sample Capacity of Packed Capillary RPLC for Cellular Proteolytic Digests. Of importance to proteomic applications is

Figure 3. Influence of guard column inner diameter on capillary RPLC separation efficiency for D. radiodurans digests. Conditions: 5-µm particle (C18, 300-Å pores, Jupiter) packed 4-cm-length stainless steel tubing with inner diameter of (A) 508- and (B) 381-µm i.d. Other conditions are the same as in Figure 2A.

the ability to inject a significant amount of sample onto the capillary column to improve detection of low-abundance species. In isocratic packed capillary LC separations, limitations for sample volume introduction result from the degradation of separation efficiency.38 Under composition gradient conditions, however, the introduction of a relatively large sample volume can be conducted without a significant loss of separation efficiency. This is because the sample can be “focused” on the head of column inlet in the weakly solvating mobile phase used for sample introduction. A limitation, however, arises from the extended time required for loading large sample volumes on the separation column. For example, introducing a 10-µL sample with optimum LC linear velocities (∼0.1 cm/s) into a 150-µm-i.d. capillary column requires ∼20 min. This required sample loading time increases proportionally with sample volume or with the inverse square of column inner diameter. Using high mobile-phase linear velocities can shorten the sample loading time; however, the use of long, small particle packed capillaries for achieving high separation efficiency limits the maximum mobile-phase flow rate that current LC pumps can supply due to pressure limitations. In practice, relatively small (38) Scott, R. P. W. Liquid Chromatography Column Theory; Wiley & Sons: New York, 1992; p 115.

Analytical Chemistry, Vol. 73, No. 8, April 15, 2001

1769

Figure 4. Reproduciblity of the separations of D. radiodurans tryptic digests using capillary RPLC with a guard column. Conditions are the same as in Figure 3B: (A) run 1, (B) run 14, and (C) run 15.

volume injections having high sample concentrations are used to reduce analysis time. For the D. radiodurans tryptic digests used in this work, sample concentrations up to ∼50 mg/mL were successfully loaded. More concentrated samples, however, were generally too viscous to be effectively manipulated with a syringe. The allowable sample mass capacity depends on the column dimensions, chromatographic packing material properties, and analyte properties.39-41 A capacity of 20-200 µg/g of packing material can be estimated for reversed-phase LC from previous studies, although different sample capacities have been reported. For long (e.g., ∼1 m) packed capillaries relevant to this work (e.g., 150 µm i.d.), the sample loading capacity can be estimated as 0.15-1.5 µg for a high-abundance analyte. In global cellular (i.e., proteomic-wide) proteolytic digest samples, hundreds of thousands of polypeptides potentially exist, having large differences in both quantity and physiochemical properties, and the sample capacity for such applications must be experimentally determined. Capillaries (100 cm × 150 µm i.d.) containing 5-µm C18 particles (300-Å pore size, Jupiter, Phenomenex, Torrance, CA) were used for this purpose because these capillaries provide a similar sample capacity to those used for high-efficiency separations (80 cm × 150 µm i.d. containing 3-µm C18; see below discussion). With these columns, an accurate switching valve injector can be used for quantitative sample introduction at 8000 psi, and mobile-phase linear velocities of up to 0.1 cm/s (typical optimum value for RPLC). Figure 2 shows the capillary RPLC/UV separations for a D. radiodurans tryptic digest when loading 50, 125, and 500 µg of sample. For sample loadings of up to 500 µg, only limited degradation of the separation quality was observed. However, it (39) Done, J. N. J. Chromatogr. 1976, 125, 43-57. (40) Andreolini, F.; Borra, C.; Novotny, M. Anal. Chem. 1987, 59, 2428-2432. (41) Novotny, M. Anal. Chem. 1988, 60, 500A-510A.

1770 Analytical Chemistry, Vol. 73, No. 8, April 15, 2001

was found that the mobile-phase flow rate (or linear velocity) decreased with increasing sample loadings to >2 mg/mL protein content for 10-µL injections due to flow restriction at the column inlet arising from overloading. Use of Precolumns in Packed Capillary LC. To improve the mobile-phase flow rate stability, adopting a relatively large inner diameter precolumn allowed use of reduced sample concentrations during injection, and providing crude sample fractionation prior to the high-efficiency separation capillary. To be effective, the precolumn should be packed with the same packing material as the separation capillary to maintain reproducible solute elution order. Separations shown in Figure 3 illustrate that the precolumn inner diameter significantly affected the separation efficiency. The low mobile-phase linear velocity in the precolumn (linear velocity is inversely proportional to the square of column inner diameter) and the “unswept” dead volume of the connection tubing at the junction of the different inner diameter columns are factors potentially affecting the separation efficiency. Compared with the results obtained without the use of a precolumn (Figure 2A), similar separation efficiencies are obtained using a precolumn having 381-µm inner diameter. Using this precolumn, highly reproducible elution patterns can be achieved for relatively concentrated (e.g., 5 mg/mL) proteolytic digests, as shown in Figure 4. It should be noted that few peaks in the chromatogram contain single components (based upon MS results; see below) and the peak shapes (or separation patterns) are sensitive to changes in the chromatographic conditions. From the results in Figure 4, it can be seen that the peak shapes were reproducible and all recognizable peaks have an average elution time deviation of 50 000 psi for improving separation efficiency.42,43 In the limited effort of this study, we used a pressure of 10 000 psi with commercial pumps for capillary LC separations. A simplified, volume-change flexible sample introduction arrangement (see the inset to Figure 1) was used for sample loading having the same functionality as previously reported.42 For 3-µm C18 particle (120 Å pore size, Phenomenex) packed capillaries, experiments showed that a pressure of 10 000 psi could drive the mobile phase through an 80-cm column with a linear velocity of up to ∼0.1 cm/s (based upon the first eluted peak). Figure 5A shows a separation of the same sample as in Figure 4 using this packed capillary column (data collection was initiated after a 10-min delay). Compared with results obtained using a 5-µm C18 particle packed capillary column (Figure 4), a significant improvement in separation efficiency was observed. The earlyeluting peaks (small hydrophilic peptides) are relatively broad, resulting from poor “focusing” of these components on the column inlet during the large-volume (10 µL) sample loading. A base peak width of ∼8 s was typically observed for highly resolved components. The peak capacity (resolution of unity44) of this separation is estimated to be as high as ∼1100 [1.5 × (120 - 20) × 60/8, time period used for the estimation: 20-120 min]. In comparison, for the 5-µm particle packed capillary column (Figure 4), a peak capacity of ∼600 is estimated. Previous studies have indicated that nonporous particles are attractive packing materials for capillary chromatography.45 The most attractive aspects of nonporous particles include the small resistance to the mobile-phase flow (high column permeability), the fast mass transfer (small C term in van Deemter equation), and the small volume flow rate at a specific liner velocity (small column porosity). For the purpose of high-efficiency separations, the operational linear velocity can be set close to the optimum linear velocity, where the mass-transfer influence on separation efficiency (or plate height, H) is limited. However, the small column resistance to mobile-phase flow in nonporous particle packed capillaries is desirable for operating long packed capillaries at their optimum linear velocities with limited pressure. Furthermore, the low flow rate at a specific linear velocity in nonporous (42) MacNair, J. E.; Lewis, K. C.; Jorgenson, J. W. Anal. Chem. 1997, 69, 983989. (43) MacNair, J. E.; Patel, K. D.; Jorgenson, J. W. Anal. Chem. 1999, 71, 700708. (44) Giddings, J. C. United Separation Science; John Wiley & Sons: New York, 1991; p 105. (45) Shen. Y.; Yang. Y.; Lee, M. L. Anal. Chem. 1997, 69, 628-635.

Figure 5. Capillary RPLC operated at a pressure of 10 000 psi for D. radiodurans tryptic digests. Conditions: 80 cm × 150 µm i.d. fusedsilica capillary packed with 3-µm (A) porous (120-Å pores) and (B) nonporous C18-bonded particles; data were collected after 10 min of gradient; other conditions are the same as in Figure 2A.

particle packed capillary columns also favors achieving high MS detection sensitivity when connecting LC to MS using ESI. The remaining problem with using nonporous particle packed capillaries is their low sample capacity, resulting from a very small specific surface on nonporous particles. In this study, the sample capacity supplied by nonporous particle packed capillaries was directly examined by loading a moderate amount (50 µg) of a global cellular tryptic digest. The same capillary column dimensions (including length, diameter, and particle size) were used for nonporous particles (3-µm C18, Micra Scientific, Northbrook, IL) as for porous particles. The resulting RPLC/UV chromatogram result is shown in Figure 5B. Compared with results obtained using the porous particle packed capillary (Figure 5A), poorer resolution was obtained using the capillary containing the same size nonporous particles as porous particles, which we ascribe to sample overloading. This suggests that nonporous particle packed capillary columns provide insufficient sample capacity for larger injections (e.g., >50 µg) of cellular proteolytic digests. However, the use of nonporous particles in narrow packed capillaries (e.g., 1000 for our 11.5-T FTICR MS). It is noteworthy that accurate mass measurement data are simultaneously acquired for each of the proteolytic polypeptides during the 2D separation. These accurate mass data can be utilized for confident protein identification based upon our accurate mass tag concept.21 Finally, we note that stable isotopeAnalytical Chemistry, Vol. 73, No. 8, April 15, 2001

1773

Figure 8. A single spectrum from a capillary RPLC-FTICR analysis of D. radiodurans tryptic polypeptides. Conditions are the same as in Figure 6.

Figure 9. A dense area of a 2D display for a capillary RPLC-FTICR analysis of D. radiodurans tryptic polypeptides. The packed capillary RPLC conditions are the same as in Figure 6, and the ESI FTICR conditions are described in the text.

labeling strategies provide the basis for simultaneous precise quantitation of peptide abundances. The number of detected isotopic distributions or polypeptides in the 2D displays was found to depend on both sample loading and separation efficiency. When loading 5 µg of sample on the same column, only 8840 isotopic distributions were observed. In comparison, 20 531 isotopic distributions were detected when loading a 50-µg sample on a low-efficiency column (30 cm packed with 5-µm particles). For high-efficiency packed capillaries (80 cm packed with 3-µm particles) with high sample loading (50 µg), the number of detected polypeptides was ∼3-fold greater than using low-efficiency packed capillaries (30 cm packed with 5-µm 1774 Analytical Chemistry, Vol. 73, No. 8, April 15, 2001

particles) and nearly 7-fold greater than low sample loading (5 µg). Larger sample loadings consistently provided a wider dynamic range and enabled the detection of more low-abundance polypeptides, although the number of additional polypeptides detected is not necessarily proportional to the amount of increased sample loading. High-efficiency separations are advantageous since they decrease the dynamic range limitations within a single mass spectrum, decrease the possibility of ESI suppression of minor species by coeluting major polypeptide components, and improve the MS detection limits (MS response increases with narrowing of the peak width for a specific amount of analyte). The overall methodology for the D. radiodurans data analysis (e.g., capillary

RPLC-FTICR MS/MS, AMT validation and database development, and quantitation based upon stable isotope labeling) is presently being refined in our laboratory and will be described elsewhere. CONCLUSIONS Small particle (e.g., 3 µm) packed, long (e.g., 80 cm) capillary columns for RPLC can be operated at optimum linear velocities using commercial LC pumps at 10 000 psi to achieve highefficiency separations for complex soluble protein proteolytic digests. These high-efficiency separations can provide a peak capacity of ∼1000 using ESI FTICR MS. Combined packed capillary RPLC with FTICR can provide an effective 2D peak capacity of more than 106 for proteome-wide cellular proteolytic polypeptides. For D. radiodurans tryptic digests, ∼48 600 polypeptides from ∼60 700 isotopic distributions have been detected in a single capillary LC/FTICR MS. Larger sample loadings with highefficiency separations enabled the detection of additional lowabundance components. Precolumns with suitable inner diameters are recommended to improve the elution reproducibility in such applications. For the 150-µm-i.d. long capillary columns (µLC), porous particle packed columns provided a total sample capacity of up to 500 µg for cellular tryptic digests, while nonporous particle

packed columns were overloaded for even moderate sample loadings (e.g., 50 µg). The 2D capillary LC-FTICR methodology described here clearly has enormous potential for providing broad highthroughput studies of complex polypeptide mixtures. The application of this approach combined with highly accurate mass measurement data, high resolution, and ultrahigh sensitivity provided by FTICR is expected to constitute a major advance for proteome measurements. ACKNOWLEDGMENT Portions of this research were supported by the Life Sciences Division, Office of Biological and Environmental Research, U.S. Department of Energy and the National Cancer Institute under Grant CA81654. PNNL is operated by Battelle Memorial Institute for the United States Department of Energy through Contract DEACO6-76RLO 1830.

Received for review September 21, 2000. Accepted February 5, 2001. AC0011336

Analytical Chemistry, Vol. 73, No. 8, April 15, 2001

1775