Anal. Chem. 1997, 69, 320-326
Open-Tubular Capillary Electrochromatography with an On-Line Ion Trap Storage/Reflectron Time-of-Flight Mass Detector for Ultrafast Peptide Mixture Analysis Jing-Tao Wu, Peiqing Huang, Michael X. Li, Mark G. Qian, and David M. Lubman*
Department of Chemistry, The University of Michigan, Ann Arbor, Michigan 48109-1055
In this work, a novel open-tubular column (OTC) capillary electrochromatography (CEC) system has been coupled to an on-line ion trap storage/reflectron time-of-flight mass spectrometer for ultrafast peptide mixture analysis. Reversed-phase OTCs prepared by the sol-gel process were coated with an amine, which greatly enhanced the electroosmotic flow in an acidic buffer solution and considerably reduced nonspecific adsorption between the peptides and the column wall. A six-peptide mixture could be separated to baseline within 3 min on this system. A full mass range detection speed of 8 Hz was used in all these experiments, which was sufficiently rapid to maintain the high efficiency of ultrafast separations. Because of the high duty cycle of the mass spectrometer and the column path length-independent concentration-sensitive feature of the electrospray ionization process, high-quality total ion chromatograms could be obtained with injections of only 1-2 fmol of peptide samples. A concentration limit of detection of 1 × 10-6 M was also achieved due to the preconcentration capability of CEC. In addition, a novel gradient CEC device was demonstrated which did not result in a pressure-driven flow. A tryptic digest of horse heart myoglobin was successfully separated on the gradient CEC system within 6 min. The use of the mass spectrometer increased the resolving power of this system by clearly identifying coeluting components. Capillary electrochromatography (CEC)1-10 is a developing technique in which the electroosmotic flow (EOF) is used to drive the mobile phase solvent in a reversed-phase liquid chromatography column without the use of pressure. In contrast to the parabolic Poiseuille flow of a pressure-driven system, the flow
profile of the EOF is flat, which remarkably reduces band broadening in CEC. A separation efficiency at least twice as high as that observed in conventional HPLC has been reported.8,11 The high separation efficiency of CEC makes it a potentially ideal technique for ultrafast separations. Since the first paper on this technique published in 1974, CEC has been slow to develop. This is mainly due to the difficulty in obtaining stable flow conditions. In order to perform a chromatographic separation with high efficiency and within a reasonable time, a high flow rate of EOF is essential. These conditions require the use of an extremely high voltage applied across the column, which results in an unstable flow due to the Joule heating effect and the electrochemical reactions on the electrodes under high voltage. Recently, it was demonstrated that this problem could be overcome by applying a supplementary pressure on the column.12 However, unless this pressure is applied at both ends of the column, which is sometimes very difficult, separation efficiency is sacrificed due to the introduction of a pressure-driven flow. In the development of CEC, both packed-column1-7 and opentubular column (OTC)8-10,13 configurations have been reported. Compared with packed columns, OTCs have several advantages. OTCs with inner diameters around 10 µm have been demonstrated to have a smaller plate height due to the lack of band broadening effects associated with the existence of packing particles and endcolumn frits. High concentration sensitivity is also another advantage of OTCs since columns with extremely small dimensions are used. Also, the small diameters of the capillaries allow the use of a higher voltage in CEC without significant Joule heating. Furthermore, OTCs provide much more rapid separations than packed columns by eliminating intraparticulate diffusion, which is the dominant limitation for ultrafast separations in packed columns. In spite of these advantages, OTCs are not widely used in HPLC separations, mainly because of the difficulties associated with sample injection, detection, and column preparation. The injection volume for OTCs is in the low nanoliter or even picoliter range. For such a small injection volume, a split device must be used, which results in a waste of sample and complexity of instrumentation. In the CEC mode, however, samples can be injected electrokinetically, thus dramatically decreasing sample
(1) Pretorius, V.; Hopkins, B. J.; Schieke, J. D. J. Chromatogr. 1974, 99, 2330. (2) Jorgenson, J. W.; Lukacs, K. D. J. Chromatogr. 1981, 218, 209-16. (3) Knox, J. H.; Grant, I. H. Chromatographia 1991, 32, 317-28. (4) Behnke, B.; Bayer, E. J. Chromatogr. 1994, 680, 93-8. (5) Yamamoto, H.; Baumann, J.; Erni, F. J. Chromatogr. 1992, 593, 313-9. (6) Yan, C.; Dadoo, R.; Zhao, H.; Zare, R. N.; Rakestraw, D. J. Anal. Chem. 1995, 67, 2026-9. (7) Tsuda, T. Anal. Chem. 1988, 60, 1677-80. (8) Bruin, G. J. M.; Tock, P. P. H.; Kraak, J. C.; Poppe, H. J. Chromatogr. 1990, 517, 557-72. (9) Pfeffer, W. D.; Yeung, E. S. J. Chromatogr. 1991, 557, 125-36. (10) Guo, Y.; Colo´n, L. A. Anal. Chem. 1995, 67, 2511-6.
(11) Dittmann, M. M.; Wienand, K.; Bek, F.; Rozing, G. P. LC-GC 1995, 13, 800-14. (12) Smith, N. W.; Evans, M. B. Chromatographia 1994, 38, 649-57. (13) Tsuda, T.; Nomura, K.; Nakagawa, G. J. Chromatogr. 1982, 248, 241-7.
320 Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
S0003-2700(96)00990-0 CCC: $14.00
© 1997 American Chemical Society
consumption and avoiding the complexity of split injection required for OTC HPLC. Although the extremely small inner diameters of OTCs make optical detection methods difficult, they are compatible with a concentration-sensitive detector such as an electrospray ionization mass spectrometer (ESI/MS),14 whose response is independent of the optical path length of the column. Thus, this major disadvantage of OTCs may be overcome in a CEC-ESI/MS configuration. Indeed, the use of MS as an on-line detector provides additional molecular weight and structural information on the analytes. In addition to the problems associated with injection and detection, the preparation of OTCs has also been a problem due to the difficulty in obtaining a sufficiently large and reproducible stationary phase ratio. Recently, Guo and Colo´n reported the preparation of open-tubular CEC columns by a one-step sol-gel method which greatly simplified the column preparation procedure.10 Most of the papers reported thus far in CEC research have been focused on the separation of small organic molecules. The separation of biopolymers such as peptides has not been extensively studied among researchers for several reasons, including the fact that the interaction between the negatively charged free silanol groups on the CEC column and the positively charged peptide ions results in band broadening, thus reducing separation efficiency.15 During conventional HPLC operation, the free surface silanol groups are suppressed to a minimum by either endcapping or adding buffer additives. Separations are usually performed at a very low pH, which further reduces the ionization of surface silanol groups. In CEC, however, a relatively large number of free ionized silanol groups are essential in producing a fast and reproducible EOF. This relatively large number of negatively charged silanol groups interact with peptides which carry positive charges under the operation conditions normally used and, therefore, considerably decrease the separation efficiency. This phenomenon provides a challenge for the separation of peptides in CEC. Ultrafast separations are very useful techniques for situations that require a high throughput, such as in combinatorial chemistry and clinical laboratories. In addition to the short analysis time, another major advantage for an ultrafast separation is its high sensitivity resulting from the sharp, high concentration peak bands. This, however, also provides a challenge to a mass spectrometer when it is used as an on-line detector. In the work reported here, a typical peak width is around 2-3 s. In order to maintain this high separation quality, a rapid detection speed is essential. In previous work, an ion trap storage/reflectron timeof-flight mass spectrometer (IT/reTOFMS) has been used as a rapid and sensitive detector with the speed required for on-line separations.16,17 This instrument uses a quadrupole ion trap as a front-end storage device, prior to mass separation and identification by a reTOFMS. Ions over a wide mass range are first accumulated in the ion trap using an rf-only voltage applied on the ring electrode, and after a delay the ions are simultaneously ejected by a dc pulse applied to the end cap into the reTOF for mass analysis. The ion trap thus serves as a means of converting a continuous electrospray beam into a pulsed beam for analysis (14) Kebarle, P.; Tang, L. Anal. Chem. 1993, 65, 972A-86A. (15) Nahum, A.; Horvath, C. J. Chromatogr. 1981, 203, 53-63. (16) Qian, M. G.; Lubman, D. M. Anal. Chem. 1995, 67, 2870-7. (17) Wu, J.-T.; Qian, M. G.; Li, M. X.; Liu, L.; Lubman, D. M. Anal. Chem. 1996, 68, 3388-96.
by TOFMS. A major advantage of this device is that, since it is a nonscanning instrument, it can reach a very high duty cycle (>99%) and do so at a sufficiently rapid rate to accurately capture an on-line separation. Also, the storage property of the ion trap provides ion integration of low-intensity signals, thus enhancing the sensitivity for detection. In this work, we report the use of OTC CEC interfaced to mass spectrometry for ultrafast separation and detection of peptide mixtures including a protein digest. To our knowledge, this is the first report on coupling OTC CEC with mass spectrometry in the analysis of peptide mixtures. The work described includes new improvements in OTC CEC column preparation, such that, after preparing a CEC column with a C-8 stationary phase, the inner surface of the capillary wall is coated with an amine. This coating has two purposes: First, it significantly enhances the EOF in acidic buffer solutions, so that a large EOF flow rate is obtained without using a very high voltage, which in turn results in a stable flow and provides the capability for kinetically optimizing a separation. Second, the surface silanol groups are covered by the amine groups which carry positive charges, so that nonspecific adsorption between the peptide sample and the inner surface is greatly reduced. Furthermore, we have added the capability for performing a novel gradient CEC/MS, which does not require or result in a pressure-driven flow in the CEC column. EXPERIMENTAL SECTION Column Preparation. Fused silica capillaries (Polymicro Technologies, Phoenix, AZ) with an inner diameter of 9 µm and an outer diameter of 150 µm were used to prepare the OTC CEC columns. The column length was typically between 30 and 40 cm. The capillary was first pretreated with 1 N sodium hydroxide, followed by water, and then dried at 120 °C for 20 h under a nitrogen flow. The C-8 stationary phase was loaded onto the inner wall of the capillary in a method similar to that reported by Guo and Colo´n.10 Briefly, a sol was prepared by the hydrolysis of two precursors, n-octyltriethoxysilane and tetraethoxysilane, with a mole ratio of 0.3. The sol was pumped through the pretreated capillary for 10 min, and the column was then flushed by nitrogen for about 30 min before it was placed into a GC oven at 120 °C overnight under nitrogen flow. After preparing the stationary phase, the column was flushed with toluene for 15 min followed by a 5% (v/v) (3-aminopropyl)trimethoxysilane (APS) in toluene solution for 6 h. Finally, the column was flushed by toluene, acetonitrile, and the running buffer, sequentially. CEC Apparatus and CEC/MS Interface. A schematic diagram of the instrument setup is shown in Figure 1. For isocratic experiments, a ∼20% (v/v) acetonitrile/water solution with 0.05% trifluoroacetic acid (TFA) in 5 mM ammonium acetate was used as the running buffer. Slight adjustment of the percentage of acetonitrile was necessary to achieve optimum separations. For gradient experiments, a small buffer vial containing solvent A (5 mM ammonium acetate in water with 0.05% TFA) was placed on a ministirrer (Model HI190M, Thomas Scientific, Swedesboro, NJ), which stirred the solvent in the buffer vial during a separation. Solvent B (5 mM ammonium acetate in acetonitrile/ water (80:20) with 0.05% TFA) was delivered by a syringe pump (Model 22, Harvard Apparatus, South Natick, MA) directly into the buffer vial at a rate predetermined by the gradient. In these experiments, a 0-35% acetonitrile gradient in 6 min was used. To calculate the delivery pump flow rate for this gradient, the total volume V (mL) of solvent B required for a 35% acetonitrile/water Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
321
Figure 1. Schematic diagram of the gradient CEC/MS configuration. 1, Syringe pump; 2, ministirrer; 3, buffer vial; 4, delivery capillary; 5, CEC column; 6, CEC power supply; 7, ESI power supply; 8, nanospray needle; 9, inlet capillary; 10, atmospheric pressure interface; 11, skimmer; 12, ion trap; 13, reflectron; 14, microchannel plate detector.
solution was calculated according to the following equation:
80% V ) 35% VA + V
(1)
where VA is the initial volume of solvent A in the buffer vial before the gradient. In these experiments, VA was 2 mL, and thus V was calculated to be 1.5 mL. The flow rate of the delivery pump was determined by dividing V by the gradient time, which was 6 min in these experiments. Thus, the delivery flow rate for this gradient was calculated to be 250 µL/min. During the experiments, one end of the column was placed in the buffer vial, where a negative high voltage was applied. The other end of the column was modified and used directly as a nanoelectrospray needle. The column end was first burned to remove the polymer coating and then etched using hydrofluoric acid to produce a sharp tip. The tip was then coated with silver according to a method described in a previous paper.17 This nanoelectrospray needle could operate at a flow rate of less than 100 nL/min. Mass Spectrometer. An IT/reTOFMS was used in these experiments as previously reported.16,17 This device consists of a reflectron time-of-flight mass analyzer (Model D1450) interfaced to a quadrupole ion trap (Model C-1251, R. M. Jordan Co., Grass Valley, CA). Ions generated from the nanoelectrospray source were introduced into an atmospheric pressure interface through a heated stainless steel capillary (0.5 mm i.d., 140 °C). The ions traversing the interface were focused by a coaxial cylindrical lens (+100 V) and subsequently passed through a skimmer orifice (325 µm) into the high-vacuum chamber. The ions were then focused through an Einzel lens into the ion trap. The ions were stored in the ion trap under a preset rf voltage of 1250 V on the ring electrode for a period of 125 ms, corresponding to a detection speed of 8 Hz in these experiments. The ions were then ejected by a dc pulse on the endcap of the ion trap into the time-of-flight device for analysis. The ions were mass separated by the reflectron TOF device and detected by a 25 mm triple-microchannel plate detector (Model C-2501, R. M. Jordan Co.). Data System. The data system used in these experiments has been previously reported.18 It was based on a 250 MHz highspeed transient recorder (Model 9846, Precision Instruments Inc., Knoxville, TN) embedded in a Pentium 66 MHz PC-compatible computer (Model P5-66, Gateway 2000, North Sioux City, SD), (18) Qian, M. G.; Wu, J.-T.; Parus, S.; Lubman, D. M. Rapid Commun. Mass Spectrom. 1996, 10, 1209-14.
322
Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
and data processing was also performed on this computer using a user-written program. This system is capable of conducting data acquisition and storage at rates over 10 Hz for a minimum of 30 min, of individual mass spectra containing 16 000 data points with 10 ns resolution. A sampling time window width of 160 µs, which corresponds to a mass range from 0 to ∼1500 Da, was used for all studies in this work. Materials. All peptide and protein samples, ammonium acetate, and TFA were obtained from Sigma Chemical Co. (St. Louis, MO) and used without further purification. Trypsin was purchased from Promega (Madison, WI). Water used to prepare CEC separation buffers was generated with a Milli-Q water purification system (Millipore Corp., Bedford, MA). For tryptic cleavage, a 100 µg aliquot of horse heart myoglobin was incubated for 24 h at 37 °C with a protein-to-enzyme ratio of 50:1 (w/w) in 50 mM NH4HCO3 solution at pH 8.2. The digested materials were then vacuum-dried to remove the salt and reconstituted in solvent A of the CEC running buffer to a concentration of 1 × 10-5 M of original protein. RESULTS AND DISCUSSION A key goal of ultrafast separations is to achieve the optimum number of theoretical plates within a minimum analysis time. CEC in particular has been shown to have a higher separation efficiency both theoretically and experimentally than that of pressure-driven HPLC, mainly because of its unique flat EOF flow profile. Therefore, the same number of theoretical plates could be obtained with a much shorter column in CEC, resulting in fast analysis. The use of OTCs further increases separation speed by eliminating intraparticulate diffusion, which is the major limitation of separation speed in packed columns. In addition, the optimization of separation kinetics is also an important factor for achieving ultrafast separations, and a good starting point is to study the relationship between flow velocity and plate height. According to Martin and Guiochon,19 the plate height in an opentubular CEC column can be expressed by the following equation:
H)
2Dm 2k′df2u k′2dc2u + + u 3(1 + k′)2Ds 16(1 + k′)2Dm
(2)
where H is the plate height, u is the linear flow velocity, k′ is the capacity factor, df is the thickness of the stationary phase layer, Ds and Dm are the diffusion coefficients in the stationary phase (19) Martin, M.; Guiochon, G. Anal. Chem. 1984, 56, 614-20.
Figure 2. Theoretical plate height as a function of linear flow rate in OTC CEC. The parameters used are Dm ) 1 × 10-5 cm/s, and dc ) 9 µm, and stationary phase mass transfer resistance is ignored.
and mobile phase, respectively, and dc is the inner diameter of the column. According to eq 2, a plot of plate height versus linear velocity for a 9 µm i.d. column used in our experiments is shown in Figure 2. For analytes with a typical k′ value between 0.1 and 1, the plate height increases very slowly until the linear flow velocity exceeds ∼3 mm/s. Thus, an optimum linear flow velocity should be in the region close to this value. Although a linear flow velocity of 3 mm/s is easy to achieve in a pressure-driven HPLC system, it is difficult to generate such a high EOF velocity in CEC, especially for the separation of peptides. Peptides are charged species in an aqueous solution, and these charges reduce the hydrophobic interactions between the peptides and the stationary phase, resulting in insufficient retention. In CEC, these charges also affect the retention on the peptides by adding an electrophoretic separation mechanism and complicate the separation. To reduce these charge effects, an ion-pairing agent such as TFA should be used, which results in an acidic buffer solution with a pH between 2 and 3. At this pH value, the ionization of free silanol groups on the inner capillary wall is largely suppressed, and as a result, the EOF velocity is very small. In order to obtain an EOF velocity of ∼3 mm/s with this acidic buffer solution, an extraordinarily high voltage (over 1000 V/cm column) would need to be used, which would inevitably result in bubble formation in CEC. In our work, the column inner wall was modified with APS, which replaces the charge carrier group from a silanol to an amine. Attempts were made to combine the APS modification and the stationary phase preparation into one step by the hydrolysis of n-octyltriethoxysilane, tetraethoxysilane, and APS together. However, the high ionic and strong basic properties of APS resulted in a rate of silane hydrolysis that was too fast to prepare a reproducible stationary phase. Therefore, a two-step procedure was used in these experiments, in which APS was chemically coated onto the column inner surface after the preparation of the stationary phase. The amine groups attached to the column wall are fully protonated in an acidic solution, and as a result, a much higher EOF velocity is observed. In our experiments, an EOF velocity of ∼3 mm/s was achieved with an electrical field as low as 350 V/cm. Taking advantage of this relatively low voltage,
Figure 3. OTC CEC separation of a six-peptide mixture using a column with APS coating. Separation conditions are column length, 30 cm (25 cm to detector); separation voltage, -12 kV; injection, -2000 V × 3 s; sample concentration, 1 × 10-5 M; UV detection at 214 nm. Six peptides are 1, methionine enkephalin; 2, bradykinin; 3, angiotensin III; 4, methionine enkephalin-Arg-Phe; 5, substance P; and 6, neurotensin.
stable flow conditions could be easily obtained without any supplementary pressure. In Figure 3 is shown the separation of a six-peptide mixture on an APS-coated OTC CEC column with UV detection at 214 nm. Because of the high flow velocity of the EOF, the separation was completed within 3 min. All the six peptides were separated to baseline, demonstrating the high separation efficiency of this novel OTC CEC. Since the APS coating is stable only in solutions with a pH greater than 2.5, 5 mM ammonium acetate was added to the running buffer to slightly adjust the pH above this value. The accurate amount of peptide sample injected in a CEC separation is difficult to calculate using conventional methods that deal with electrokinetic injection in capillary zone electrophoresis (CZE). This is because peptides have different migration rates during the injection period and the separation period. During the injection period, the peptide migration rate is determined only by electrophoretic mobility and EOF, while during the separation period, the migration rate is determined by partitioning, electrophoretic mobility, and EOF. A good approximation is to ignore the electrophoretic mobilities during injection and use the EOF flow rate to calculate the injection amount. This is a reasonable approximation since the EOF is relatively high and the charges on the peptides are greatly reduced by ion pairing with TFA. The EOF velocity in this experiment was determined to be 3.4 mm/s by injection of a plug of ethanol before running a separation. The peptide sample injected is thus estimated to be 1-2 fmol in this experiment. This high sensitivity mainly results from the ultrafast separation and the extremely small dimensions of the OTC CEC columns. Another advantage of modifying the inner capillary wall with APS is to reduce the nonspecific adsorption between the positively Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
323
Figure 4. TIC of the separation of the same six-peptide mixture. Mass detection speed, 8 Hz. All other conditions were the same as in Figure 3.
charged peptide ions and the negatively charged surface silanol groups. The separation of the same six-peptide mixture by an OTC CEC column without APS modification has also been performed (data not shown). As expected, the separation required a much longer time to complete due to the slow EOF velocity, and a dramatically decreased separation efficiency was also observed. The EOF flow rate in this experiment was measured to be 0.68 mm/s. According to Figure 2, there should not be a significant increase in the plate height compared to that at a higher flow rate, shown in Figure 3. The reduction in column efficiency was mainly caused by the interaction between the peptides and the column wall. By replacing the negatively charged silanol groups with positively charged amine groups, nonspecific adsorption between the peptides and the column wall is minimized, resulting in an improved efficiency. The coupling of packed-column CEC with electrospray ionization mass spectrometry has recently been demonstrated by several groups using quadrupole mass spectrometers.20-23 Although quadrupole mass analyzers can provide reasonable mass spectra and sensitivity for many fast separations,24,25 the quality of ultrafast separations may be sacrificed due to the relatively slow scanning speeds of quadrupoles when used as on-line detectors in the fullscan mode. In previous work, we have demonstrated the use of (20) Hugener, M.; Tinke, A. P.; Niessen, W. M. A.; Tjaden, U. R.; van der Greef, J. J. Chromatogr. 1993, 647, 375-85. (21) Schmeer, K.; Behnke, B.; Bayer, E. Anal. Chem. 1995, 67, 3656-8. (22) Gordon, D. B.; Lord, G. A.; Jones, D. S. Rapid Commun. Mass Spectrom. 1994, 8, 544-8. (23) Lane, S. J.; Boughtflower, R.; Paterson, C.; Morris, M. Rapid Commun. Mass Spectrom. 1996, 10, 733-6. (24) Figeys, D.; van Oostveen, I.; Ducret, A.; Aebersold, R. Anal. Chem. 1996, 68, 1822-8. (25) Wahl, J. H.; Goodlett, D. R.; Udseth, H. R.; Smith, R. D. Electrophoresis 1993, 14, 448-57.
324 Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
Figure 5. TICs of CEC separation of a peptide mixture. (a) Sample was prepared in water; (b) sample was prepared in the running buffer. Column length, 35 cm; separation voltage, -14 kV; injection, -3000 V × 10 s. The two peptides in the mixture were 1, angiotensin III, and 2, methionine enkephalin-Arg-Phe.
an IT/reTOFMS as a rapid and sensitive detector for HPLC and CE separations.16,17 This instrument can perform full mass range detection at a speed of over 10 Hz with a nearly 100% duty cycle. Figure 4 shows the total ion chromatogram (TIC) of the separation of the same six-peptide mixture on the OTC CEC-ESI-IT/ reTOFMS system. Full mass range (0-1500 Da) spectra were obtained at a rate of 8 Hz. Figure 4 was qualitatively similar to Figure 3, indicating that the mass spectrometer could accurately maintain the high quality of the CEC separation. It should be noted that the signal/noise ratio in Figure 4 is higher than that in Figure 3, even though the same amount of peptides was injected. This results from the different characteristics of the UV detector and the electrospray mass spectrometer when columns with extremely small dimensions are used. The response of a UV detector follows Beer’s law; i.e., for the same analyte, the response is proportional to the product of the analyte concentration and the optical path length. For a 9 µm i.d. column, the optical path length is very short, thus resulting in a low UV response. In the case of the electrospray ionization method, however, above a minimum flow rate, the response depends only on the analyte concentration and not on the column dimensions. Electrospray ionization mass spectrometry is, therefore, particularly useful as a sensitive detector for OTC CEC. As a hybrid technique, CEC combines some of the best features of HPLC and CZE. The flat EOF flow profile from CZE
Table 1. Comparison of Calculated and Measured Tryptic Fragments of Horse Heart Myoglobin from CEC/MS Analysis
no.
fragment
calcd mass (Da)
1 2 3 6 7 8 10 13 14, 15 15 16 17 18 20 21
1-16 17-31 32-42 48-50 51-56 57-62 64-77 80-96 97-102 99-102 103-118 119-133 134-139 146-147 148-153
1816.0 1606.8 1271.4 396.5 707.8 661.7 1378.7 1854.1 752.9 469.6 1885.2 1501.6 747.9 309.4 649.7
a
measd massa (Da)
sequence
1816.4 1606.3 1271.4 396.4 707.6 661.8 1379.0 1853.6 752.7 469.4 1885.5 1501.2 748.1 309.5 649.6
GLSDGEWQQVLNVWGK VEADIAGHGQEVLIR LFTGHPETLEK HIK TEAEMK ASEDIK HGTVVLTALGGILK GHHEAELKPLAQSHATK HKIPIK IPIK YLEFISDAIIHVLHSK HPGNFGADAQGAMTK ALELFR YK ELGFQG
Average mass of all charge states of the fragment observed.
Figure 6. UV trace (a) and TIC (b) of gradient CEC separation of a tryptic horse heart myoglobin digest. Conditions: 0-35% acetonitrile gradient in 6 min; column length, 40 cm (for UV detection, 35 cm to UV detector); separation voltage, -14 kV; injection, -2000 V × 5 s; UV detection at 214 nm; MS detection speed, 8 Hz.
is maintained in CEC, which reduces band broadening compared with pressure-driven HPLC. A variety of HPLC stationary phases provide CEC with more tunable selectivity compared to CZE. Sample preconcentration, one unique feature from reversed-phase HPLC, is also retained in CEC, which improves the concentration limit of detection. By dissolving the sample in a water solution, the hydrophobic sample will be focused at the beginning of the stationary phase and thus will be preconcentrated. This effect is shown in Figure 5a and b, where a two-peptide mixture with a concentration of 1 × 10-6 M, prepared by dissolving in water (with preconcentration) and the running buffer (without preconcentration), respectively, were subjected to analysis on the CEC/MS system. In Figure 5a, these two peptides could be clearly separated and identified, while in Figure 5b, both the separation and signal/noise ratio were considerably worse. An improved concentration limit of detection is a major advantage of CEC over CZE. Because of the characteristics of the capacity factors of peptides, small changes in isocratic concentrations of the organic component in the mobile phase result in dramatic changes in peptide retention times.26 It is thus difficult to optimize peptide separations using the isocratic mode, especially for complex peptide mixtures such as protein digests. Gradient CEC without a pressure-driven flow was reported recently by Yan et al.27 In
our experiments, a novel gradient elution CEC was used, which did not require or cause a pressure-driven flow in the CEC column. Moreover, the instrumentation employed was very simple. In Figure 6 is shown the UV trace and the TIC of a gradient CEC
(26) Mant, C. T.; Lorne Burke, T. W.; Hodges, R. S. Chromatographia 1987, 24, 565-72.
(27) Yan, C.; Dadoo, R.; Zare, R. N.; Rakestraw, D.; Anex, D. S. Anal. Chem. 1996, 68, 2726-30.
Figure 7. Extracted ion current profiles of the peak marked by an asterisk in Figure 6. (a) 3.475 min at m/z 636.5, and (b) 3.505 min at m/z 470.4.
Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
325
the isotopic peaks of many components, which could be used to simplify the interpretation of mass spectra containing multiply charged ions. The peak marked by an arrow in both the UV trace and the TIC corresponds to unretained components in the mixture and serves as an approximate marker for the EOF velocity. A further advantage of using the mass spectrometer as an online detector for CEC is that it could identify some partially resolved or unresolved peaks, thus increasing the resolving power of this instrument. This is demonstrated in Figures 7 and 8, in which are shown the extracted ion current profiles and the mass spectra of two coeluting components corresponding to the peak marked by an asterisk (*) in Figure 6b. As indicated by the extracted ion current profiles, even though the elution times of the two peaks differ only by 1.8 s, they are of different identities. The mass spectra in Figure 8 clearly identified them as fragments LFTGHPETLEK and IPIK. Because a mass detection speed of 8 Hz was used, spectra that contained only one species could be easily obtained, which further simplified the mass identification. These mass spectra were acquired by a single pulse-out with a trapping storage time of 0.125 s; nevertheless, they were of high signal/noise ratio, owing to the high sensitivity of the mass spectrometer and the suppression of background noise by TFA.
Figure 8. Mass spectra corresponding to the extracted ion current profiles in Figure 7. Each spectrum contains a single pulse-out with a trapping storage time of 0.125 s. (a) Doubly charged (m/z 636.5, calculated m/z 636.7) and triply charged (m/z 424.9, calculated m/z 424.8) ions of fragment LFTGHPETLEK. (b) Singly charged (m/z 470.4, calculated m/z 470.6) ion of fragment IPIK.
separation of horse heart myoglobin digest. Taking advantage of the gradient elution, over 10 peaks were resolved in both the UV and TIC within 6 min. Among the peaks shown in the TIC, 15 usable mass spectra (including coeluting components) could be obtained to cover about 90% of the amino acid residues in the protein, as listed in Table 1. The calculated and the measured masses were found to be in excellent agreement. In addition, as a result of the nonscanning property of the mass spectrometer, these mass spectra were obtained with high-resolution conditions (∼1500), although they were obtained at a full mass range sampling speed of 8 Hz. This resolution is sufficient to resolve
326 Analytical Chemistry, Vol. 69, No. 3, February 1, 1997
CONCLUSIONS In this work, APS-coated reversed-phase OTC CEC columns were prepared and coupled to an IT/reTOFMS for ultrafast analysis of peptide mixtures. This new column greatly increased EOF velocity and reduced peptide-wall interactions. A six-peptide mixture could be separated to baseline within 3 min with the injection of 1-2 fmol of sample and detected at a full mass range sampling speed of 8 Hz. The concentration limit of detection was also improved due to the preconcentration effect of reversed-phase CEC. In addition, a myoglobin digest was successfully analyzed on this system using a novel gradient design, which eliminated the use of a pressure-driven flow. Coeluting components were unambiguously identified by the mass spectrometer. ACKNOWLEDGMENT We gratefully acknowledge support of this work by the National Institutes of Health under Grant No. 1R01GM49500 and the National Science Foundation under Grants No. BIR-9223677 and BIR-9513878. Received for review September 26, 1996. November 26, 1996.X
Accepted
AC9609900 X
Abstract published in Advance ACS Abstracts, January 1, 1997.
