Mechanism of Peptide Separations by Solid Phase Extraction

silica capillary and subsequently released for free zone electrophoresis by injection of an organic elutant. Unlike previous designs and commercially ...
0 downloads 0 Views 296KB Size
Download PDF
+

+

Anal. Chem. 1996, 68, 306-314

Mechanism of Peptide Separations by Solid Phase Extraction Capillary Electrophoresis at Low pH Michael A. Strausbauch,† James P. Landers,‡ and Peter J. Wettstein*,†

Department of Surgery and Immunology, Clinical Capillary Electrophoresis Facility, and Department of Laboratory Medicine and Pathology, Mayo Clinic/Foundation, Rochester, Minneapolis 55905

A device for on-line extraction and concentration of peptides from a dilute sample matrix prior to direct capillary electrophoretic analysis is described. The technique, termed solid phase extraction capillary electrophoresis (SPE-CE), can facilitate analysis of peptides in the low nanograms per milliliter range. Peptides from a sample matrix are adsorbed on a reversed phase resin (C-8 or C-18) cartridge in-line with an uncoated fusedsilica capillary and subsequently released for free zone electrophoresis by injection of an organic elutant. Unlike previous designs and commercially available packed-inlet capillaries, the device is easily constructed from common laboratory materials and is applicable to a wide range of conventional instrumentation and methods. This device and method has been developed for use in our laboratory as a stand-alone preparative technique, specifically to provide a second-dimensional orthogonal separation of biologically derived HPLC fractions of peptides in a single analysis. To this end, extensive effort was required in both device construction and method development to attain the successful separations which are reported in this study. Extractions of dilute peptide mixtures from sample injections exceeding, but not limited to, 20 times (48 µL) the capillary volume with apparent recovery greater than 80% are shown. The selectivity of extraction of individual components of a very dilute peptide mixture (31 ng/mL with 280 µL of sample injected) is presented. The ability to efficiently extract the individual peptides from the sample was found to be concentration-dependent for the individual peptide components over a 1600-field dilution of a common calibration mixture of nine model peptides. Varying the injected volume of elution buffer demonstrated the importance of minimizing the amount of buffer used to desorb peptides to maximize the resolution of individual peptides. This study highlights implementation for direct SPE-CE for peptide analysis and discusses the SPE tip-induced mechanism through which reversal in electroosmotic flow occurs. Capillary electrophoresis (CE), originally pioneered by Hjerte´n1 in the late 1960s, was revitalized into a viable analytical technique by Mikkers et al.2 and Jorgenson and Lukacs3 about a decade †

Departments of Surgery and Immunology. Department of Laboratory Medicine and Pathology. (1) Hjerte´n, S. Chromatogr. Rev. 1967, 9, 122-219. (2) Mikkers, F. E. P.; Everaerts, F. M.; Verheggen, T. P. E. M. J. Chromatogr. 1979, 169, 11-20. (3) Jorgenson, J. W.; Lukacs, K. D. Anal. Chem. 1981, 53, 1298-1302. ‡

later. It has since been shown to be useful for the analysis of a diverse array of molecules including ions,4 small organic molecules,5 sugars,6,7 peptides,8 proteins,9-13 DNA,14,15 and even intact chromosomes.16 As a result of the microscale dimensions of the fused-silica capillaries used for CE, Joule heat generated under extremely high field strengths (30 kV; up to 1000 V/cm] can be efficiently dissipated, allowing for rapid analysis times (second to minute time scale) in comparison with gel electrophoresis and high-performance liquid chromatography (HPLC). The small capillary dimensions essential for CE provide both advantages and limitations. Total capillary volumes on the microliter scale induce the requirement of extremely small injected volumes of sample (1-30 nL) to avoid overloading. This is advantageous when only small volumes of sample are available for analysis. Provided that the sample concentration is adequate (typically >10 µg/mL for peptides), 5 µL is adequate for replicate analyses using detection in the ultraviolet-visible (UV-vis) spectrum and allows for ∼99% of sample to be analyzed by other means. In contrast, when the concentration of desired analytes in the sample is low, the capillary dimensions become a severe limitation in terms of the feasibility of CE for analysis. The sensitivity limit using UV-vis detection is typically thought to be approximately 10-6 M. The development of sensitivity enhancement techniques such as field-amplified stacking17 and on-line isotachophoresis18 has had a profound effect on the applicability of CE for the analysis of low-concentration analytes in the research and clinical laboratory. However, these methods for increasing sample loading in CE are still limited to a sample volumes less than the total capillary volume, i.e., 500 nL to 2.5 µL. In the present study, we describe the development and use of an on-line concentration technique for the analysis of low(4) Jones, W. R.; Jandik, P.; Pfeifer, R. Am. Lab. 1991, 40-46. (5) Carney, S. L.; Osborne, D. J. Anal. Biochem. 1991, 195, 132-140. (6) Liu, J.; Shirota, O.; Wiesler, D.; Novotny, M. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 2302-2306. (7) Liu, J.; Shirota, O.; Novotny, M. Anal. Chem. 1991, 63, 413-417. (8) Oda, R. P.; Madden, B. J.; Morris, J. C.; Spelsberg, T. C.; Landers, J. P. J. Chromatogr. A 1994, 680, 341-351. (9) Rush, R. S.; Cohen, A. S.; Karger, B. L. Anal. Chem. 1991, 63, 1346-1350. (10) Guttman, A.; Horvath, J.; Cooke, N. Anal. Chem. 1993, 65, 199-203. (11) Landers, J. P.; Oda, R. P.; Madden, B. J.; Spelsberg, T. C. Anal. Biochem. 1992, 205, 115-124. (12) Wiktorowicz, J. E.; Colburn, J. C. Electrophoresis 1990, 11, 769-773. (13) Mazzeo, J. R.; Krull, I. S. BioTechniques 1991, 10, 638-645. (14) Heiger, D. N.; Cohen, A. S.; Karger, B. L. J. Chromatogr. 1990, 516, 3348. (15) Ulfelder, K. J.; Schwartz, H. E.; Hall, J. M.; Sunzeri, F. J. Anal. Biochem. 1992, 200, 260-267. (16) Guszczynski, T.; Chrambach, A. Biophys. Res. Commun. 1991, 179, 482486. (17) Chien, R. L.; Burgi, D. S. Anal. Chem. 1992, 64, A489-496. (18) Foret, F.; Szoko, E.; Karger, B. L. Electrophoresis 1993, 14, 417-428.

+

concentration samples. The utility of this technique, which we have termed solid phase extraction capillary electrophoresis (SPECE), is demonstrated for the analysis of peptides at concentrations below the detection limit with standard CE. The extrapolation of this technique to the peptide analysis of entire HPLC fractions and comparison to standard CZE,19 forensic detection of sulfonylurea drugs,20,21 and the preparative analysis of biologically derived peptides by fraction collection22 using SPE-CE has been accomplished. This technique is simple in design and involves incorporation of a concentrator device containing a particulate solid phase (on-line) with a CZE capillary. Advances using a packedinlet capillary have previously been reported by Kasicka and Prusik,23 who attached an “adsorption element” to an isotachophoretic apparatus, and Guzman et al.,24,25 who constructed the “analyte concentrator”, which was coupled to a CZE capillary. Both of these devices utilized a packed bed of an immunoaffinity resin for specific concentration of analytes. Debets et al.26 described as “microprecolumn” (reversed-phase C-8) retained within a valve that could be switched on- and off-line to a CZE capillary. Merion and Swartz27,28 showed the utility of the Accusep C/PRP capillary (Waters, Milford, MA), which is similar in construction to the analyte concentrator containing a C-18 reversed-phase packing instead of an immunoadsorbent. Most recently, and of particular relevance to this study, Beattie et al.29 reported a device similar in construction and implementation to the SPE concentrator described here. The SPE-CE method presented in this study follows the concentration strategies of these studies, and the modification to the standard CE capillary is simple in design and can be constructed from commercially available materials with minimal effort. The complete assembly of the SPE-CE tip is usually accomplished in less than 10 min with minimal equipment requirements, allowing for rapid production with relatively little cost. Improved sample handling capacity, allowing exceedingly large sample volumes greater than 200 µL as well as recognizing, understanding, and overcoming the effects of a reversed EOF, stands to revolutionize CE for analytical and preparative purposes. MATERIALS AND METHODS Materials. Acetic acid, ammonium acetate, acetonitrile (ACN; HPLC-UV spectral grade), and hydrochloric acid (HCl) were purchased from Fisher Scientific (Pittsburgh, PA). Twenty-five millimolar phosphate buffer, pH 2.5, was purchased from Scientific Resources Inc. (Eatontown, NJ). C-18 material was removed from a SPE column purchased from J.T. Baker (Phillipsburg, NJ). Glass fiber was removed from the bed support of a common disposable “spin column”. Polyethylene tubing (0.38 mm i.d., IntraMedic No. 7405) was purchased from Clay Adams (Parsip(19) Strausbauch, M. A.; Madden, B. J.; Wettstein, P. J.; Landers, J. P. Electrophoresis 1995, 16, 541-548. (20) Landers, J. P. Clin. Chem. 1995, 41, 495-509. (21) Strausbauch, M. A.; Xu, S. J.; Ferguson, J. E.; Nunez, M. E.; Machacek, D.; Lawson, G. M.; Wettstein, P. J.; Landers, J. P. J. Chromatogr., in press. (22) Nevala, W. K.; Wettstein, P. J. J. Immunol., submitted. (23) Kasicka, V.; Prusik, Z. J. Chromatogr. 1983, 273, 117-128. (24) Guzman, N. A.; Trebilcok, M. A.; Advis, J. P. J. Liquid Chromatogr. 1991, 14 (5), 997-1015. (25) Guzman, N. A. U.S. Patent 5,045,172, Sept 3, 1991. (26) Debets, A. J. J.; Mazereeuw, M.; Voogt, W. H.; Van Iperen, D. J.; Lingeman, H.; Hupe, K.-P.; Brinkman, U. A. Th. J. Chromatogr. 1992, 608, 151-158. (27) Swartz, M. E.; Merion, M. J. Chromatogr. 1993, 632, 209-213. (28) Fuchs, M.; Merion, M. U.S. Patent 5,246,577, Sept 21, 1993. (29) Beattie, J. H.; Self, R.; Richards, M. P. Electrophoresis 1995, 16, 322-328.

+

Figure 1. Simplified diagrammatic representation of solid phase concentrator and its attachment to the inlet of the CZE separation capillary.

pany, NJ). Methanol, ethanol, 2-propanol, hexachloro-2-propanol, acetonitrile, and acetone were spectral or HPLC grade reagents. Synthetic peptides LSPFPFDL (P2CL), HIYEFPQL (P1), and TEGYMYTDNS (P2) were synthesized by FMOC chemistry and provided HPLC-purified by the Mayo Protein Core Facility. A mixture of nine peptide standards was purchased from Bio-Rad (containing (bradykinin, angiotensin II, R-melanocyte hormone, thyrotropin releasing hormone, luteinizing hormone releasing hormone, leucine enkephalin, bombesin, methionine enkephalin, and oxytocin). This mixture was diluted with separation buffer/ water (1:9) to yield a stock concentration of 50 µg/mL for each peptide. SPE-CE Buffer Preparation. Separation buffers were (1) 20 mM ammonium acetate in 1% acetic acid made with doubly deionized water or (2) 50 mM phosphate (pH 2.5) and were used without modification. Elution buffers ACN/1% HCl, 1:1 (initial experiments), 8:1, or 9:1 (v/v) were made with doubly deionized water. All aqueous buffers were filtered through a SepPak column (Waters) to remove organic contaminants and then filtered through a 0.2 µm filter (Gelman, Ann Arbor, MI) before use. CE Instrumentation. HPCE separation was carried out on a Beckman P/ACE System 5510 equipped with a monochromatic UV detector. An IBM Model 60 computer utilizing System Gold software (v 8.1) was used for instrument control and data collection. All peak information (migration time, integrated peak areas, and height) was obtained through the System Gold software. SPE-CE Capillary Construction. The SPE capillary is a hybrid design consisting of two parts, a replaceable concentrator tip and a separation capillary as previously described.19,21 Briefly, the concentrator tip typically contained bed of reversed phase (C18) material (∼50-100 nL) retained in a polyethylene sleeve between two glass fiber frits so that two short sections of capillary could be slid into the sleeve to retain the entire assembly. The device is resistant to blockage due to the glass fiber bed support that acts to both prefilter the sample and prevent fragments of the solid phase from fouling the separation capillary (Figure 1). Concentrator tips constructed in this manner have never failed due to blockage of the packed tip. The bed support component of the concentrator is critical to the design, and without it the capillary may fail immediately due to blockage. The high flow rate and minimal restriction resulting from the presence of the concentrator allows the assembly to remain in-line with the separation capillary throughout the entire analysis. The separation capillary is common uncoated fused silica with dimensions noted

+

Figure 2. Elution of adsorbed peptides under nonelectrophoretic (pressure only) conditions. (A) Elution of peptides P1 and P2 from the packed SPE-CE tip by pressure injection (20 psi for 1 min) of elution buffer (50% ACN in 1% acetic acid plus additional acetic acid added to match adsorbance). (B) Elution of a sample blank (no peptides) under the same conditions. Separation buffer, 1% acetic acid; capillary, 75 µm × 57 cm; temp, 20 °C.

in figure legends, typically 75-100 µm i.d. × 370 µm o.d. × 57 cm (L). The concentrator tip is mated to the inlet of the separation capillary by joining the two sections with a second polyethylene sleeve (Figure 1). SPE-CE Separation Methods. The capillary is rinsed with three column volumes of elution buffer and then reequilibrated with 10 column volumes of separation buffer prior to sample injection. After sample injection by pressure at the inlet, (1) the SPE capillary is refilled with separation buffer, (2) concentrated sample in elution buffer is pushed by positive pressure into the separation capillary by a second pressure injection of separation buffer. This step ensures that the peptide components contained in the elution buffer are transported past the capillary joints so that analyte stacking and electrophoresis may be carried out in an intact separation capillary section and are unaffected by any mismatch of diameters introduced by the concentrator section. Electrophoretic separations are then carried out in a normal fashion. The SPE capillaries are rinsed with elution buffer as a postseparation treatment to regenerate the reversed phase and to ensure sample is not carried over between consecutive separations. Capillary temperature is maintained at 20 or 25 °C, and detection is by adsorbance at 200 or 214 nm. RESULTS AND DISCUSSION Peptide Loading and Desorption with Hydrostatic Pressure-Driven Flow. The first step in examining the utility of solid phase packing as a matrix for on-line peptide extraction was to confirm that peptides in a volume exceeding that of the capillary could be bound under aqueous conditions and eluted in a controlled manner with an organic buffer. Two synthetic peptides (P1 and P2) at a concentration of 10 µg/mL were loaded onto the SPE capillary under high pressure (20 psi), and the capillary was rinsed with several column volumes of separation buffer. The peptides were subsequently desorbed (eluted) from the packing by continuous positive pressure of 20 psi with elution buffer (50% ACN/0.1% acetic acid) for 1 min. For this experiment, additional acetic acid was added to the elution buffer so that its UV absorbance would approximate that of the separation buffer. UV profiles given in Figure 2 show the elution of the peptide mixture from the reversed-phase packing of the SPE-CE capillary in

+

comparison with the identical elution of a sample blank (no applied voltage). The response observed at ∼15 s during analysis of the sample blank (Figure 2B) was thought to have resulted from a change in the refractive index at the interface of the elution and separation buffers. The results of these “pressure only” experiments demonstrated that (1) peptides diluted in an aqueous buffer can be concentrated on the SPE-CE packing, (2) the immobilized peptides can be retained on the packing while the solid phase is rinsed with several column volumes of an aqueous electrophoresis buffer, and (3) the peptides can be eluted in a controlled manner by the introduction of an organic solvent. Reversal of EOF with Initial Implementation of SPE-CE Peptide Separation. Having established conditions for the binding and controlled elution of peptides, on-line concentration, elution and CZE separation of the P2CL peptide was attempted. The peptide was loaded onto the SPE capillary as described above and rinsed with several column volumes of separation buffer. The peptide was eluted from the packing by injection of 340 nL of elution buffer (∼10% of total capillary volume; all buffer injection volumes were determined by timed injection at 20 psi), which functioned to both elute the peptides from the packing and transport them into the separation region of the SPE capillary. Following application of a voltage for standard electrophoretic separation (inlet as anode and outlet as cathode), no peptides were detected by UV absorbance; even during electrophoresis for 120 min, the absorbance remained at baseline. Injection of a 340 nL plug of elution buffer and application of a positive pressure at the inlet resulted in a detector response consistent with that observed in the previous “pressure only” experiment. One hypothesis predicts that the peptide was eluted from the solid phase but was rapidly readsorbed. By this hypothesis, described in Figure 3, the peptide is (1) extracted from the sample matrix, (2) eluted from the solid phase by injection of the elution buffer, and (3) carried into the separation region of the capillary, where, upon application of voltage to the system, electrophoretic stacking of the peptides occurs at the cathodic interface of the elution and separation buffers. The peptide electrophoretically migrates into the aqueous separation buffer, while a reversed EOF allows the peptide, now in an aqueous environment, to readsorb onto the solid phase. This hypothesis was tested by repeating the previous experiment with the polarity reversed (anode ) outlet; cathode ) inlet). Under these conditions, the elution buffer (which is essentially a neutral marker) and the slower migrating peptide were detected at 8 and 14 min, respectively (Figure 4A). This indicates that the EOF was anodic in direction, i.e., reversed, under normal polarity. The reversed EOF was found to persist and could not be restored to a conventional cathodic direction even after extensive regeneration of the solid phase packing with an acetonitrile-containing elution buffer (Figure 4B). In an attempt to determine the source of the reversed EOF, additional SPE-CE separations were performed with a new SPECE capillary. A small injection of water served as a neutral marker as well as an indicator of the magnitude and direction of the bulk flow. Extensive testing of all system components [separation and elution buffers (acetic acid, ammonium acetate, and acetonitrile)] showed no significant impact on the conventional EOF. Since none of these buffer changes had an effect on the reversed EOF but did complicate interpretation of the electropherogram (ITP “steps”), we chose to utilize a less complex separation buffer through the use of dilute acetic acid or sodium phosphate in

+

+

Figure 3. Reversed EOF and rebinding of peptides to reversed phase. Diagrammatic representation of SPE-CE capillary, buffer and peptide sample. (A) Capillary with packet inlet before sample loading. (B) Capillary after sampling loading, peptides extracted from sample matrix and absorbed to solid phase. (C) Elution of peptides by 100 nL injection of elution buffer, followed by 200 nL injection of separation buffer to push the elution buffer (containing the released peptides) into the separation capillary (D) Start of electrophoresis with analyte stacking at cathodic interface of elution and separation buffers. (E) Peptides have migrated out of elution zone and are now in aqueous phase. (F) Peptides reentering the SPE concentrator tip under the influence of a reversed EOF. (G) Peptides reimmobilized on the solid phase, while the elution buffer has exited at the inlet.

subsequent experiments. However, injection of the synthetic peptide to resin in the SPE tip did have a dramatic effect on the direction of the EOF (data not shown). The phenomenon was reproducible and persistent without further injection of the peptide mixture, since extensive regeneration of the reversed-phase packing did not reestablish a conventional EOF. Replacement of the SPE tip with an unused SPE tip did restore the initial (cathodic) EOF direction. Therefore, the aberrant EOF behavior of the SPE capillary at pH 2.7 appeared to be codependent on both the presence of the SPE concentrator tip and exposure of the solid phase packing to a peptide-containing sample. To determine whether the reversed EOF tendency of the SPE capillary was a phenomenon linked to a design flaw in the SPE tip, a commercial, packed-inlet capillary (Accusep C/PRP, Waters) was tested under comparable conditions. This commercial capillary has a 75 µm i.d. (57 cm length) with a packed resin (C-18) space length of e3.0 mm near the inlet of the capillary. The capillary used for this experiment was estimated to contain less

than half the maximum bed volume after microscopic inspection of the packed tip. The separation of a sample “blank” (H2O) was compared with the separation following injection of a two-peptide mixture (P1 and P2, Figure 5). The EOF was in the conventional (cathodic) direction for the sample blank, as evidenced by the detection of the elution buffer at 11.9 min. However, a subsequent injection and separation of these peptides resulted in a net EOF that remained in the conventional direction, but strongly suppressed, as evidenced by the 89 min needed for the detection of the elution buffer. It is noteworthy that the capillary fill time (measured inlet-to-detector at 20.0 psi) observed during the reconditioning of the capillary between analyses remained unchanged from the initial separation and constant throughout these experiments. This indicated that capillary blockage, shifting, or compression of the solid phase packing was not the cause of the suppressed EOF. Successful separation of the two peptides (at 35 and 68 min) demonstrated that they could be concentrated on the packing, eluted, and electrophoretically resolved. Suppression

+

Figure 4. Reversed polarity with an unmodified capillary. Demonstrates the reversed EOF of the SPE-CE capillary after exposure to peptides. (A) Separation of synthetic octapeptide (LSPFPFDL) with elution buffer as EOF marker. (B) Subsequent separation of sample blank with elution buffer as EOF marker. Conditions: 50 µm × 47 cm capillary, 370 mm × 3 mm C-18 packing; separation buffer, 20 mM ammonium acetate in 1% acetic acid, pH 3.4; elution buffer, 50% ACN/50% separation buffer. Sample injection: 100 nL of a 0.1 µg/ mL octapeptide, followed by 4 column volumes of separation buffer. Elution by injection of 340 nL of elution buffer (∼10% of total capillary volume), followed by second injection of 340 nL of separation buffer. Injections at 20 psi; flow rate, ∼2 µL/min. Electrophoresis at 20 kV, ∼18 µA at 20 °C. Detection at 214 nm. Polarity reversed with inlet as cathode and outlet as anode.

Figure 5. Effect of peptide loading on Waters C/PRP and migration time for neutral marker. Before loading peptide, EOF was measured at 11.9 min; after peptide loading, the EOF marker was detected at 89 min. Capillary flow (measured by inlet to detector times for a marker, +20 psi at inlet) remained constant, indicating that the increased EOF time is not due to physical blockage of the capillary. Capillary, Waters C/PRP 75 µm × 57 cm; sample, HIYEFPQL and TEGYMYTDNS (P1 and P2), 10 ng/µL injected at 0.1 min at 20 psi; elution buffer, 90% ACN in 0.1% HCl, 0.5 min at 0.5 psi, followed by separation buffer push, 2.0 min at 0.5 psi; flow rate, ∼9 µL/min; separation buffer, 1% acetic acid; separation voltage, 20 kV. Capillary maintained at 20 °C.

of the EOF by an unknown mechanism was related directly to peptide loading, as first observed with the SPE capillaries. As with the SPE capillaries described in the current study, extensive washing with elution buffer did not restore the behavior of the C/PRP capillary prior to analysis of the peptides. However, unlike the SPE capillary, the packed inlet of the C/PRP capillary is constructed as an integral part of the capillary and could not be

+

removed to accommodate comparison with the EOF associated with a new, unused packed tip. Nevertheless, we propose that the same mechanism suppresses conventional EOF in the C/PRP capillary and in the SPE capillary. This undefined mechanism, which suppresses or even reverses the bulk flow, may explain the lack of reproducible electrokinetic sample and buffer injections reported by Merion and Swartz27 with the C/PRP capillary at pH below 4, which is the pH limit for the C/PRP capillary. It should be noted, however, that the apparent resolution obtained with this capillary was much better than that of the early SPE capillaries when similar methods were employed. Reversal of EOF in SPE-CE at Low pH. Two major characteristics that are unique to the SPE-CE capillary when it is used for electrophoretic separation of peptides at low pH are (1) the tendency for a suppressed or reversed EOF and (2) the fact that field strength (V/cm) is not constant throughout the capillary due to the presence of the elution buffer plug, which creates a discontinuous buffer system (Figure 6). The strong, reversed EOF resulting from the irreversible adsorption of peptides to the reversed-phase packing can be eliminated and the conventional cathodic EOF restored by replacement of the SPE tip. This observation suggests that the vectorial direction of the EOF is not constant along the entire length of the capillary, and, apparently, the generation of the localized reversed EOF is associated with the region occupied by the SPE tip. Since it had been determined that the reversed EOF produced in the region of the packing is directly related to peptide loading, we propose that initial unrecoverable peptide losses to the packing alter the polar character of the reversed-phase surface. At low pH, these peptide losses may provide a cationic charge on the surface of the solid phase and produce a localized reversal of EOF by the same mechanism commonly associated with coating capillaries with a cationic substance.30 Interestingly, the observed effects of the reversed EOF are greater than would be expected given the short length of capillary occupied by the SPE packing (32 min) and responses exceeding the limits of the detector (>0.2 AU). No loss of linearity was detected with the largest sample injection, indicating that the retentive capacity of the phase was not reached, a point at which competition for adsorptive sites would be evident.21,27 Therefore, a total mass of 780 ng of peptide in 288 µL (delivered in 32 min) was bound and found to be below

the retentive capacity of the reversed-phase packing, assuming (1) a conservative extraction efficiency of 60% and (2) a flow rate of 9.0 µL/min at 20.0 psi. Migration times for individual peaks during this linearity study also exhibited a high degree of reproducibility (n ) 7) (Table 1). While the variability in migration time was greater than that usually observed with standard analytical CZE methods (0.5-0.10% CV), it is still adequate for most applications and, in particular, preparative-scale CZE. Elution Buffer Volume Alters Electrophoretic Resolution and Detection of Peptides. The use of an organic solvent to introduce analytes into the separation region of the SPE capillary would be expected to affect the electrophoretic separation, i.e., the efficiency and selectivity of the peptide separation. The effects of varying the volume of elution buffer on the overall separation of a mixture of model peptides was investigated (Figure 9). It is clear that peptides did not elute from the packing with the lowest elution volume (Figure 9A); this result may indicate dilution of the organic phase by diffusion, compounded by the parabolic flow of the elution buffer injection. Increasing the elution volume improved the efficiency of analyte desorption, with a maximum response observed at 130 nL, or 2.5 times the packing volume (Figure 9D); this volume optimized both peak shape and resolution. Increasing the volume did not significantly enhance the elution efficiency beyond that obtained with 130 mL but did result in decreased resolution and extended analysis time (Figures 9F). These results highlight the importance of identifying the correct elution buffer volume for obtaining efficient peptide elution as well as optimal resolution and peak shape. It should also be noted that the use of the positive (+0.5 psi) pressure applied at the inlet is not absolutely required to detect peptides with low mobilities. When the elution buffer injection does not exceed 130 nL, effective separation can be accomplished using a CE system lacking the capacity for simultaneous pressure and voltage application. Discontinuous Buffer Systems in SPE-CE. Inherent to the SPE-CE methodology is the unavoidable use of a discontinuous buffer system, in which an elution buffer plug containing analytes is flanked on both sides by separation buffer. Accordingly, regions within the capillary are characterized by different varying field strengths and buffer zones of dramatically different composition (refer to Figure 6). While this, in itself, is beneficial to some CZE separations, e.g., inducing analyte stacking and/or transient ITP, the use of an excessive elution buffer volume is clearly not beneficial to the SPE-CE process. The results of this study suggest that the discontinuous buffer system allows for very

+

+

complished through the field strength “step” gradient between the low-conductivity elution buffer and the higher conductivity separation buffer. The beneficial effects and mechanism associated with this phenomenon have been previously described.17 This effect is further enhanced by a pH-driven focusing, which results from the use of HCl as a polar modifier in the elution buffer. At pH 1.8, the carboxyl groups and, in particular, the carboxyl termini of the peptides would be expected to be more fully protonated in the elution buffer than in the separation buffer, resulting in an overall increase in peptide stacking. However, as demonstrated earlier (Figure 9), an excessive elution buffer volume intolerably extends peptide migration times, and analytes are not adequately resolved when mobilized with continuous low pressure. This may be caused by the low conductivity of the elution buffer and resultant voltage drop across the elution buffer zone (Figure 6D) that result in a concurrent reduction of field strength in the remaining regions of the capillary (Figure 6E). Accordingly, power dissipated across the elution/sample zone during electrophoresis is wasted after analyte stacking insofar as the separation is concerned. Laminar flow at the interface of the two buffer systems, as described by Chien,32 may also produce band broadening, but reduction of this effect by matching EOF velocities produced by the elution and separation buffers has not been attempted.

Figure 9. Effect of elution buffer volume on separation. Separation of a standard peptide mixture containing (1) bradykinin, (2) angiotensin, (3) R-melanin-stimulating hormone, (4) thyrotropin releasing hormone, (5) luteinizing hormone releasing hormone, (6) leucine enkephalin, (7) bombesin, (8) methionine enkaphalin, and (9) oxytocin, all at individual concentrations of 0.5 µg/mL. Sample was pressure injected for 1.0 min at +0.5 psi. Elution buffer, 90%/10% ACN/1% HCl with elution injection varied as noted, followed by a 2 min, +0.5 pressure injection of separation buffer (∼268 nL). Separation was carried out in a 75 µm × 57 cm capillary containing 25 mM phosphate buffer, pH 2.50, at 20 kV, with the actual electrophoresis starting at 5 min, with the capillary thermostated at 25 °C. Dashed line indicates time point separation was continued with the simultaneous application of 20 kV + 0.05 psi at inlet. SPE-concentrator, 0.5 mm × 0.37 mm, 40 µm diameter C-18 phase. Detection at 200 nm. Bar represents 0.010 AU. Data were combined and plotted by export to Deltagraph Each data set was offset by 0.020 AU.

efficient stacking of the peptides at the cathodic interface of the elution and separation buffers. This analyte stacking is ac(32) Chien, R. L.; Burgi, D. S. Anal. Chem. 1992, 64, 1046-1050.

CONCLUSIONS Solid phase extraction capillary electrophoresis provides the capability for the analysis of dilute peptide samples at concentrations well below the detection limit for conventional UV detectors. The experiments described herein identify reversed EOF as a potential obstacle that can be overcome by a simultaneous application of voltage and pressure during analyte separation. Packed-inlet devices such as the SPE capillary described in this study have the potential for increasing the sample handling capacity of standard CZE for micropreparative analyses (i.e., fraction collection) without off-line sample preconcentration or the pooling of fractions from replicate separations. The results clearly show that analysis of samples with analyte concentrations below the direct UV detection limit is now possible. In terms of clinical applications, the SPE-CE approach will overcome the inherent practical barrier in applying standard CE: sample injection is not limited to low nanoliter volumes, allowing direct injection of very dilute peptide samples. We believe these will be among the most important applications of packed-inlet capillaries and the SPE-CE technology. ACKNOWLEDGMENT This work was supported by the National Institutes of Health Contract AI-45197 and the Development Committee of the Department of Laboratory Medicine and Pathology. Received for review March 31, 1995. Accepted November 1, 1995.X AC9503217 X

Abstract published in Advance ACS Abstracts, December 1, 1995.