Anal. Chem. 2004, 76, 23-30
Open Tubular Capillary Electrochromatography of Synthetic Peptides on Etched Chemically Modified Columns Joseph J. Pesek,* Maria T. Matyska, and G. Brent Dawson
Department of Chemistry, San Jose State University, San Jose, California 95192 Jenny I-Chen Chen, Reinhard I. Boysen, and Milton T. W. Hearn
ARC Special Research Centre for Green Chemistry, Monash University, Clayton, Victoria, 3800, Australia
Two sets of peptides, each having structurally similar amino acid sequences, have been investigated by capillary electrochromatography (CEC) using etched chemically modified capillaries as the separation medium. In comparison to gradient RP-HPLC, the resolving power of the described CEC methods has been found to be superior. A number of variables have been examined with respect to optimization of the separation of these closely related peptides with several different etched chemically modified capillaries. These experimental variables included the nature of the bonded moiety, the pH, the organic modifier type, and the amount of organic modifier in the buffer electrolyte. Systematic variation of these parameters results in significant changes in the migrational behavior of the investigated peptides and provides important insight into the underlying molecular separation processes that prevail in open tubular CEC. Moreover, under optimized conditions, efficient separations characterized by highly symmetrical peaks were achieved. In addition, this study has permitted the long-term stability as well as the short-term and long-term reproducibility of the etched chemically modified capillaries to be documented. Capillary electrochromatography (CEC), a hybrid technique of high-performance liquid chromatography (HPLC) and capillary zone electrophoresis (CZE), was first introduced by Pretorius et al.1 in 1974 for the analysis of nonpolar molecules. In CEC, an electric field is used as the driving force to achieve the migration of solutes. The electroosmotic flow produced by the electric field creates a pluglike flow profile that results in reduced band broadening and higher efficiencies compared to the parabolic profile obtained with pressure-driven flow systems (as in HPLC). Separation of solutes in CEC occurs by a combination of differential electrophoretic mobility and stationary-phase interaction. Depending on the chemical and physical nature of the stationary phase, CEC can be classified into two main categories: open * To whom correspondence should be addressed: (e-mail)
[email protected]. (1) Pretorius, V.; Hopkins, B. J.; Schieke, J. D. J. Chromatogr. 1974, 99, 2330. 10.1021/ac0302253 CCC: $27.50 Published on Web 11/25/2003
© 2004 American Chemical Society
tubular CEC (OTCEC)2,3 or packed-column CEC (PC-CEC).2,4 In OTCEC, the stationary phase is attached to the inner wall of the capillary, while in PC-CEC, the capillary is packed with a solid material confined between two frits. A third type of CEC system, based on a monolithic or continuous bed,5 has been classified by some investigators within the PC-CEC group. In this latter case, an in situ polymerization step is performed, forming a rodlike structure that fills the capillary.6 The format most widely used in CEC consists of fused-silica capillaries packed with reversed-phase materials based on spherical particles (1.5-10 µm in diameter) similar to those used in HPLC. Only part of the capillary, prior to the detection window, is packed to avoid losses in detectability due to light scattering by the particles,7-9 with the packing material kept in place by retaining frits. The general application of packed capillaries can, however, present some problems: namely (i) it is necessary to keep the packing material in place with frits without causing a change in the flow profile; (ii) bubble formation in the frits and around the packing material can occur; (iii) without specialist knowledge and packing equipment, the reproducible preparation of packed beds can present difficulties due to the small diameter of particles and capillaries; and (iv) the stationary phases typically used in PC-CEC are not end-capped, so strong electrostatic interactions can occur between the residual silanols and basic compounds making their analysis difficult. Some of the problems associated with the use of frits and limitations experienced with silica-based particles can be solved using the OTCEC approach. The main drawbacks of OTCEC are the low sample capacity due to the low surface area within the capillary and the relatively long axial distances that molecules have to migrate to interact with the stationary phase bonded to the surface of the inner capillary (2) Colon, L. A.; Burgos, G.; Maloney, T. D.; Cintron, J. M.; Rodriguez, R. L. Electrophoresis 2000, 21, 3965-3993. (3) Jinno, K.; Sawada, H. TrAC, Trends Anal. Chem. 2000, 19, 664-675. (4) Pursch, M.; Sander, L. C. J. Chromatogr., A 2000, 887, 313-326. (5) Svec, F.; Peters, E. C.; Sykora, D.; Frechet, J. M. J. J. Chromatogr., A 2000, 887, 3-29. (6) Legido-Quigley, C.; Marlin, N. D.; Melin, V.; Manz, A.; Smith, N. W. Electrophoresis 2003, 24, 917-944. (7) Yan, C.; Dadoo, R.; Zhao, H.; Zare, R. N.; Rakestraw, D. J. Anal. Chem. 1995, 67, 2026-2029. (8) Behnke, B.; Grom, E.; Bayer, E. J. Chromatogr., A 1995, 716, 207-213. (9) Rebscher, H.; Pyell, U. Chromatographia 1996, 42, 171-176.
Analytical Chemistry, Vol. 76, No. 1, January 1, 2004 23
wall. These drawbacks are partially solved when an etching process is applied to the inner wall of the capillary. The etching process results in the overall surface of the inner wall being increased by a factor up to 1000.10 Moreover, the etching process creates a new chemical composition at the inner wall of the capillary, and this affects both the electroosmotic flow properties and the adsorptive properties of the capillary.11 While the stationary phase for OTCEC is most often prepared by either bonding or coating the capillary inner wall with an alkyl moiety or a hydrophobic polymer, it is also possible to create a dynamically adsorbed stationary phase including a surfactant in the electrolyte.12 This mode is called CEC with a dynamically modified stationary phase. As described elsewhere,13,14 adsorption of the surfactant onto the inner capillary wall depends on its characteristics and the wall charge. In the present investigation, the CEC migration and resolution behavior of two closely related sets of synthetic peptides have been investigated using OTCEC systems based on etched and bonded surfaces. The migrational behavior of these peptides has been interpreted in terms of the structural characteristics, with the results throwing new light on the basis of peptide resolution in OTCEC vis-a`-vis the use of complementary separation procedures based on reversed-phase HPLC. EXPERIMENTAL SECTION Preparation of the OTCEC Capillaries. The OTCEC capillaries used in this study (n-butylphenyl (BP), n-octadecyl (C18), and cholesterol (CL)) were fabricated by etching and chemical modification of Polymicro Technologies (Phoenix, AZ) fused-silica capillaries, i.d. ) 50 µm. Bonding of the n-butylphenyl and n-octadecyl moiety onto the etched inner wall of the capillaries was achieved using the silanization/hydrosilation process adapted to the capillary format.15-19 Attachment of the cholesteryl-10-undecenoate moiety to the hydride-modified capillary was based on procedures described in more detail previously.20 Before use, each capillary was conditioned overnight with methanol and then washed with Milli-Q water (using a syringe), and finally, several volumes of the buffer electrolyte (degassed with helium for 30 min) were forced through the modified capillary. The buffers used in the mobile phase were as follows: pH 2.14, 60 mM phosphoric acid and 38 mM Tris; pH 3.00, 30 mM citric acid and 25 mM β-alanine; pH 4.41, 30 mM acetic acid and 375 mM γ-butyric acid; pH 7.0, 50 mM Tris and 125 mM boric acid. Capillary dimensions: BP-capillary, total length 46.7 cm and distance to detection window 38.2 cm; CL-capillary, (10) Onuska, F. I.; Comba, M. E.; Bistricki, T.; Wilkinson, R. J. J. Chromatogr. 1977, 142, 117-125. (11) Pesek, J. J.; Matyska, M. T. J. Chromatogr., A 2000, 887, 31-41. (12) Zou, H.; Ye, M. Electrophoresis 2000, 21, 4073-4095. (13) Hult, E. L.; Emmer, A.; Roeraade, J. J. Chromatogr., A 1997, 757, 255262. (14) Emmer, A.; Roeraade, J. J. Liq. Chromatogr. 1994, 17, 3831-3846. (15) Chu, C. H.; Jonsson, E.; Auvinen, M.; Pesek, J. J.; Sandoval, J. E. Anal. Chem. 1993, 65, 808-816. (16) Sandoval, J. E.; Pesek, J. J. Anal. Chem. 1991, 63, 2634-2641. (17) Pesek, J. J.; Matyska, M. T.; Sandoval, J. E.; Williamsen, E. J. J. Liq. Chromatogr. Relat. Technol. 1996, 19, 2843-2865. (18) Pesek, J. J.; Matyska, M. T. J. Chromatogr., A 1996, 736, 255-264. (19) Marciniec, B. Handbook of Hydrosilylation; Pergamon Press: Oxford, U.K., 1992. (20) Matyska, M. T.; Pesek, J. J.; Katrekar, A. Anal. Chem. 1999, 71, 55085514.
24
Analytical Chemistry, Vol. 76, No. 1, January 1, 2004
total length 34.5 cm and distance to detection window 26 cm; C18capillary, total length 34.0 cm and distance to detection window 26 cm. Peptide Samples. The peptides 1-17 used in this study were prepared by solid-phase peptide synthesis21 based on Fmoc chemistry.22-25 The generic methods employed for the purification of these synthetic peptides have been described previously.24 The molecular masses of the purified peptides were determined by ESI-MS using a Micromass platform (II) quadrupole MS with an electrospray source with the Masslynx NT Version 3.2 software (Micromass, Cheshire, U.K.). The peptides in 1:1 (v/v) acetonitrile-water containing 3% (v/v) formic acid were infused into the instrument at a speed of 10 µL/min. The ESI-MS spectra of the peptides were acquired at 70 °C at 55 V/50 V over the mass/ charge (m/z) range of 200-2000. HPLC Conditions. Peptides were analyzed on a TSK-ODS120 T column (150 × 4.6 mm i.d., 120 Å, 5 µm, end-capped) at a flow rate of 1 mL/min with a 25-min gradient from 100% A (0.1% TFA in water) to 85% B (60% acetonitrile-0.09% TFA), pH 2.1. MALDI TOF Mass Spectrometry. The peptides were analyzed with MALDI-TOF mass spectrometry (Voyager DE-STR, Applied Biosystems) with a 337-nm nitrogen laser using 3-nsduration pulses with a maximum firing rate of 20 Hz using the Voyager DeSTR Biospectrometry Workstation software. All samples were spotted onto 100-well stainless steel plates. Spectra in the reflector mode were obtained with delayed extraction using a delay time of 250 ns with positive polarity. The mass acquisition mass range was 0-10 000 Da, with a low-mass gate at 500 Da. Each spectrum was obtained with 100 laser shots, whereby five spectra were accumulated. The resulting spectra were processed with the Data Explorer Software Version 4.0.0.0, baseline corrected, noise filtered/smoothed and deisotoped with the generic formula C6H5NO. OTCEC Instrumentation. A Beckman P/ACE 5510 Series capillary electrophoresis instrument (Beckman, Fullerton, CA) or Agilent Technologies GmbH HPCE-3D (Waldbronn, Germany) capillary electrophoresis instrument was used. Electrophoretic data were processed using Agilent Technologies (Palo Alto, CA) Chemstation software. The applied voltage was varied between +10 to +30 kV depending on the application. The temperature was held at 30 °C. The hydrodynamic injection mode was used for all samples. Direct detection was performed at 210 nm for the synthetic peptides. The oven used for etching and modifying the capillaries was adapted from a Hewlett-Packard 5890 gas chromatograph. The HPLC system was an Agilent Technologies model 1050 liquid chromatograph. Computational Procedures. The characteristics of the synthetic peptides 1-17 in terms of their monoisotopic masses, mi, and their net charge values q, calculated using the HendersonHasselbalch equation26 at pH 2.14, 3, 4.41, and 7.0, with the pKa (21) Fields, G. B.; Noble, R. L. Int. J. Pept. Protein Res. 1990, 35, 161-214. (22) Keah, H. H.; Allen, N.; Clay, R.; Boysen, R. I.; Warner, T.; Hearn, M. T. W. Biopolymers 2001, 60, 279-289. (23) Boysen, R. I.; Hearn, M. T. W. J. Pept. Res. 2001, 57, 19-28. (24) Boysen, R. I.; Jong, A. J. O.; Hearn, M. T. W. Biophys. J. 2002, 82, 22792292. (25) Yang, Y.; Boysen, R. I.; Chen, J. I. C.; Keah, H. H.; Hearn, M. T. W. J. Chromatogr., A 2003, 1009, 3-14. (26) Skoog, B.; Wichman, A. Trends Anal. Chem. 1986, 5, 82-83.
Table 1. Amino Acid Sequence, Net Charge Values q, Intrinsic Hydrophobicity, and Monoisotopic Masses of Peptides 1-17 net charge values, q, calculated at different pH values
intrinsic hydrophobicity
mass
peptide code
sequence
pH 2.14
pH 3.0
pH 4.41
pH 7.0
C8
C18
mi [H]+
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
DHDINR WDHDINR SWDHDINR NSWDHDINR HNSWDHDINR HHNSWDHDINR HHHNSWDHDINR HHHNSW QHNHFHR DQHNHFHR QDQHNHFHR HQDQHNHFHR IHQDQHNHFHR NIHQDQHNHFHR TNIHQDQHNHFHR HTNIHQDQHNHFHR HTNIHQD
2.89 2.89 2.89 2.89 3.89 4.89 5.89 3.92 4.92 4.90 4.90 5.90 5.90 5.90 5.90 6.90 2.79
2.39 2.39 2.39 2.39 3.39 4.39 5.39 3.61 4.61 4.50 4.50 5.50 5.50 5.50 5.50 6.50 2.25
0.82 0.51 0.51 0.51 1.48 2.46 3.44 2.99 3.99 3.23 3.22 4.20 4.20 4.20 4.20 5.18 1.21
-0.92 -0.96 -1.23 -1.23 -0.86 -0.76 -0.66 0.24 1.13 0.27 0.13 0.34 0.34 0.06 0.34 0.44 -0.86
-2.05 3.86 2.47 1.2 0.46 -0.28 -1.02 1.03 5.32 2.48 0.79 0.05 4.43 3.16 4.97 4.23 -1.09
0.24 2.53 1.91 2.16 -0.08 -2.32 -4.56 -4.80 -2.21 -2.41 -2.10 -4.34 -0.86 -0.61 0.04 -2.20 0.01
769.3587 955.4381 1042.4701 1156.5130 1293.5719 1430.6308 1567.6897 817.3488 975.4656 1090.4926 1218.5511 1355.6100 1468.6941 1582.7370 1683.7874 1821.8436 864.3959
values of the side chains derived according to Dawson et al.27 and for the N- and C-termini according to Rickard et al.28 were implemented within the Charge subroutines of the Hephaestus software developed at the Monash laboratories and formatted as Microsoft Excel version 5.0 files. Hydrophobicity values of the individual peptides were calculated according to the descriptors developed by Wilce et al.29 RESULTS AND DISCUSSION Over the past several years, capillary electrochromatographicbased techniques have been increasingly applied to the analysis of peptides. These hybrid techniques take advantage of the occurrence of multimodal retention mechanisms, i.e., chromatographic retention, electrophorectic migration, and secondary retention effects, all of which can be optimized by changing the experimental conditions. In previous studies on the application of etched chemically modified capillaries, the OTCEC behavior of a series of synthetic peptides was studied as a function of the bonded material, pH, temperature, and organic solvent composition.30-32 These investigations revealed that some synthetic peptides, which gave single peaks by gradient elution RP-HPLC analyses could be resolved into multiple peaks in OTCEC experiments with etched chemically modified capillaries. In this paper, our emphasis has been to examine the OTCEC behavior of different sets of synthetic peptides (the structural characteristics are shown in Table 1). We were particularly interested to ascertain whether the OTCEC behavior of these closely related sets of peptides could be correlated with their gradient elution RP-HPLC (27) Dawson, R. M. C.; Elliot, D. C.; Elliot, W. H.; Jones, K. M. Data for biomedical research, 3rd ed.; Clarendon Press: Oxford, U.K., 1986. (28) Rickard, E. C.; Strohl, M. M.; Nielsen, R. G. Anal. Biochem. 1991, 197, 197-207. (29) Wilce, M. C. J.; Aguilar, M. I.; Hearn, M. T. W. Anal. Chem. 1995, 67, 1210-1219. (30) Matyska, M. T.; Pesek, J. J.; Boysen, R. I.; Hearn, M. T. W. Electrophoresis 2001, 22, 2620-2628. (31) Matyska, M. T.; Pesek, J. J.; Boysen, R. I.; Hearn, M. T. W. J. Chromatogr., A 2001, 924, 211-221. (32) Matyska, M. T.; Pesek, J. J.; Boysen, R. I.; Hearn, M. T. W. Anal. Chem. 2001, 73, 5116-5125.
characteristics or whether the electromigrational and electrodiffusion features inherent to CEC systems could be exploited to further enhance the resolution of peptide mixtures hitherto found difficult to separate by conventional RP-HPLC methods. The current investigations validate that OTCEC when operated with appropriate buffer electrolyte composition can provide an additional level of resolving power for such challenging separation tasks with synthetic peptides. In initial experiments, electrochromatograms of several peptides obtained with the various capillaries examined in this study were compared to the results from gradient elution RP-HPLC analyses of the same samples. Representative of these results, for peptide 6 (Figure 1A), gradient elution RP-HPLC yielded two minor components that were not well resolved from the main peak. However, when the same sample was analyzed using an etched capillary modified with a BP moiety, the OTCEC separation resulted in the two minor components being well resolved from the main peak. In addition, three more components were evident in the OTCEC separation than obtained by gradient elution RPHPLC. When operated under the same buffer and voltage conditions, OTCZE also does not generate the same level of resolution as OTCEC with this group of peptides (unpublished results), with the number of components present in the peptide mixtures resolved by OTCEC exceeding in many cases that of the corresponding OTCZE analysis. This is not an unexpected result, due to the selectivity differences introduced by the multimodal retention mechanisms that operate in OTCEC. Mass spectrometric analysis confirmed that the minor components were due to the presence of deletion peptides, side-chain reaction products, and group I ion adducts produced during the synthesis, purification, or storage. In particular, with the sample of peptide 6, the deletion peptides 6-des-arginine (obtained mass shift ∆m, -156.0878 Da; expected monoisotopic mass shift ∆m, 156.101 11; ∆∆m, 0.0133 Da) and 6-des-histidine (obtained mass shift ∆m, -137.0476 Da; expected monoisotopic mass shift ∆m, 137.058 91; ∆∆m, 0.0113 Da) were identified by MALDI TOF MS. Peptide 6 derivatives related to decarboxylation (obtained mass shift ∆m, Analytical Chemistry, Vol. 76, No. 1, January 1, 2004
25
Figure 1. Comparisons of gradient RP-HPLC analyses of the synthetic peptides with electrophoretic determinations using various etched chemically modified capillaries. (A) Peptide 6 on butylphenyl capillary at pH 2.14; (B) peptide 9 on C18 and butylphenyl capillaries at pH 3.0; (C) peptide 11 on butylphenyl capillary using two different pH electrolytes, pH 2.14 and pH 3.0; and (D) peptide 13 on butylphenyl and cholesterol capillaries at pH 2.14. Gradient HPLC conditions, OTCEC conditions, buffers and capillary dimensions given in Experimental Section. Applied voltage, 25 kV in all cases.
-43.9949 Da; expected monoisotopic mass shift ∆m, 43.989 83; ∆∆m, -0.0051 Da) and dehydration (obtained mass shift ∆m, -18.0104 Da; expected monoisotopic mass shift ∆m, -18.0106; ∆∆m, -0.0002 Da) were also identified. The group I ion adducts obtained for peptide 6 were the H+/Na+ exchange product (obtained mass shift ∆m, +21.9758 Da; expected monoisotopic mass shift ∆m, 21.9819; ∆∆m, 0.0061 Da) and the H+/K+ exchange product (obtained mass shift ∆m, +37.9487 Da; expected monoisotopic mass shift ∆m, 37.9559; ∆∆m, 0.0072 Da). Thus, for the OTCEC separation of this sample of peptide 6, there was a close correlation between the number of components detected in the purified peptide sample by MALDI-TOF mass spectrometry and the number of components resolved by OTCEC. Another interesting difference between the RP-HPLC and OTCEC analyses of the peptide 6 sample was the selectivity relationships between the minor components, reflected as changes in their relative position in the chromatogram/electrochromatogram. Thus, in the RP-HPLC separation of peptide 6 sample, these minor components eluted before the major peak, while several appeared after the major peak with the etched BP-OTCEC capillary. This result confirms that fundamental differences in selectivity can be achieved between the two systems. As well documented,33 in RP-HPLC peptides are retained predominantly by a solvophobic (33) Hearn, M. T. W. Biologicals 2001, 29, 159-178.
26
Analytical Chemistry, Vol. 76, No. 1, January 1, 2004
mechanism while in OTCEC both chromatographic retention and electrophoretic mobility are simultaneously manifested. In particular, deletion peptides with a net charge different from that of the parent peptide, i.e., due to loss of an arginine, will manifest different (q/M2/3) versus pH dependencies and thus show different electromigrational/electrodiffusional behavior. If the structural deletion is associated with only a small change in peptide hydrophobicity, yet a significant change in the magnitude of the q/M2/3 term for a specified buffer electrolyte system, then resolution of the deleted product is expected to be favored in CEC. In Figure 1B, the analogous analysis of a sample of peptide 9, previously purified by RP-HPLC, is shown. The analytical RPHPLC results shown in this case are representative of the experimental data obtained for the other synthetic peptides in these two sets as well as for several previously studied TRAP peptides.32 A single peak was obtained for the gradient elution RP-HPLC analysis of a sample of peptide 9 while multiple peaks were observed for all OTCEC separations. Thus, with either the BP- or C18-OTCEC column, an additional, later-eluting major component was revealed. Analysis of the sample by MALDI-TOF MS confirmed the presence of a peptide component with a mass loss relative to the parent peptide of ∆m 17.0210 Da, corresponding to pyroglutamate (
