Article pubs.acs.org/ac
Comprehensive Multidimensional Separations of Peptides Using Nano-Liquid Chromatography Coupled with Micro Free Flow Electrophoresis Matthew Geiger, Nicholas W. Frost, and Michael T. Bowser* Department of Chemistry, University of Minnesota, 207 Pleasant Street SE, Minneapolis, Minnesota 55455, United States ABSTRACT: The throughput of existing liquid phase twodimensional separations is generally limited by the peak capacity lost due to under sampling by the second dimension separation as peaks elute off the first dimension separation. In the current manuscript, a first dimension nanoliquid chromatography (nLC) separation is coupled directly with a second dimension micro free flow electrophoresis (μFFE) separation. Since μFFE performs continuous separations, no complicated injection or modulation is necessary to couple the two techniques. Analyte peaks are further separated in μFFE as they elute off the nLC column. A side-on interface was designed to minimize dead volume in the nLC × μFFE interface, eliminating this as a source of band broadening. A Chromeo P503 labeled tryptic digest of BSA was used as a complex mixture to assess peak capacity. 2D nLC × μFFE peak capacities as high as 2,352 could be obtained in a 10 min separation window when determined according to the product of the first and second dimension peak capacities. After considering the orthogonality of the two separation modes and the fraction of separation space occupied by peaks, the usable peak capacity generated was determined to be 776. The 105 peaks/min generated using 2D nLC × μFFE was nearly double the previously reported maximum peak capacity production rate achieved using online LC × LC.
C
sional separations to be applied to a wider range of mixtures.5−7 Online 2D separations capture fractions as they elute off a first dimension separation, which are then directly injected onto a second column for further separation. While these hyphenated approaches have allowed a wider range of analytes to be separated, they often do not satisfy the criteria necessary for achieving the ideal peak capacity described in eq 1. To account for nonideal orthogonality and sampling, eq 1 can be modified to
hromatography and electrophoresis separations are powerful techniques for the analysis of mixtures. Coupling a well resolved separation with a sensitive detection strategy provides unmatched selectivity and limits of detection. Unfortunately, the increasingly complex mixtures encountered in modern biomedical and environmental research often overwhelm the peak capacity achievable using traditional onedimensional separations. Two dimensional (2D) separations provide an attractive approach for dramatically increasing peak capacity. As described by Giddings, the peak capacity of a 2D separation (nc,2D) approaches the product of the peak capacities from the first (1nc) and second (2nc) dimension separations:1 nc ,2D = 1nc 2nc
nc ,2D = 1nc 2nc × f ×
(2)
where f is the fraction of the available separation space where chromatographic peaks can be found and 1/β is the fraction of the first dimension peak capacity (1nc) retained through the interface to the second dimension separation.8 The fractional coverage ( f) accounts for losses in peak capacity due to correlations in the two separation modes chosen. In order to attain the maximum 2D peak capacity described in eq 1, the separation mechanisms in each dimension must be completely orthogonal.9 In this ideal case, peaks would be spread randomly across the entire separation space since the elution time in the first dimension would not be predictive of the elution time in the second dimension.
(1)
To achieve the ideal 2D peak capacity described in eq 1, two criteria must be met: (1) the separation mechanisms of the two dimensions must be orthogonal and (2) there is no remixing or under sampling at the interface between the two separations.1 2D gel electrophoresis is a classic example of a well-executed multidimensional separation, often achieving peak capacities of several thousand.2 Unfortunately, 2D gel electrophoresis is slow, difficult to operate, and limited to a specific class of analytes. Bushey and Jorgenson were the first to introduce fully automated online instrumentation for both LC × LC3 and LC × CE4 separations. Since then, comprehensive, online 2D separations have been developed using a variety of first and second dimension separation methods, allowing multidimen© 2014 American Chemical Society
1 β
Received: March 13, 2014 Accepted: April 17, 2014 Published: April 17, 2014 5136
dx.doi.org/10.1021/ac500939q | Anal. Chem. 2014, 86, 5136−5142
Analytical Chemistry
Article
performing LC × LC separations on a reasonable time scale. Similarly, 70% of peak capacity is lost for a 15 min analysis when ts is 20 s. For comparison, the fastest LC × LC cycle times reported to date are 12−21 s.14−16 It should be emphasized that the second dimension separation time cannot generally be reduced without impacting the peak capacity generated in the second dimension. The result is an optimum sampling time that cannot be reduced without a corresponding reduction in 2D peak capacity.17,18 Various separation modes of capillary electrophoresis (CE) have also been utilized in 2D separations of proteins or peptides in either CE × CE19−21 or microfluidic platforms22−26 Micellar electrokinetic chromatography (MEKC) coupled with CE on-chip has been used to separate tryptic digests of proteins with nc,2D as high as 4200 in 12 measurements were taken across the 6.1 s average baseline width of a nLC peak. On the basis of these values, eq 3 predicts that 99% of the nLC peak capacity is retained through the interface, eliminating under sampling as a factor limiting the observed 2D peak capacity. It should be noted that the CCD camera could easily be operated at much faster sampling rates to accommodate narrower peaks or faster separations in the first dimension. The minimum sampling time of the second dimension will ultimately be limited by the sensitivity of the CCD and the exposure time required to record a signal above the LOD. Figure 6d shows a μFFE linescan extracted from the full 2D separation. The average baseline width for peaks in the μFFE
Figure 4. 1D nLC chromatogram of a Chromeo P503 labeled BSA tryptic digest.
overwhelms the peak capacity of the 1D separation with numerous unresolved peaks observed across the 10 min separation window. Over 30 distinguishable peaks were observed; however, only 5 were baseline resolved. Seventeen peaks were selected, and the integration method in Peak Finder was used to estimate a 1D nLC peak capacity of 101. The last three well resolved peaks had an observed average efficiency of 96,000 theoretical plates. 2D nLC × μFFE. Figures 5 and 6 demonstrate different views of a 2D nLC × μFFE separation of the same fluorescently labeled BSA digest analyzed using 1D nLC in Figure 4. It is important to note that the nLC gradient and separation time did not need to be adjusted to accommodate the 2D separation
Figure 5. 2D nLC × μFFE separation of the Chromeo P503 labeled BSA tryptic digest. 2D contour plots and 1D linescans of this same data set are shown in Figure 6. 5140
dx.doi.org/10.1021/ac500939q | Anal. Chem. 2014, 86, 5136−5142
Analytical Chemistry
Article
Figure 6. 2D nLC × μFFE separation of the Chromeo P503 labeled BSA tryptic digest. (a) Top view and (b) expanded view of region enclosed in the dotted box in a) (i.e., 13.2 to 17.2 min in the 1st dimension and 3 to 7 mm in the 2nd dimension). (c) Extracted chromatogram from the 1st dimension measured at 4 mm on the μFFE device (shown as the vertical dashed line in a). (d) Line scan across the 2nd dimension taken at 14.0 min in the 1st dimension (shown as the horizontal dotted line in a). A 3D representation of this same data set is shown in Figure 5.
nLC and μFFE to maximize orthogonality is expected to improve f and realized 2D peak capacity. The use of Chromeo P503 is important for both of these approaches since it allows the native charge and isoelectric point of the peptides to be maintained through the labeling protocol, retaining the chemical diversity of the sample.54
dimension was estimated to be 0.41 mm using the same 29 peaks used to determine the nLC dimension peak capacity. Considering the 10 mm width of the separation channel, the μFFE separation is able to generate a peak capacity of 24 (2nc). Combining the nLC (1nc) and μFFE (2nc) peak capacities according to eq 1 gives rise to a 2D nLC × μFFE peak capacity (nc,2D) of 2,352 in a 10 min separation window. For comparison, Mellors et al. recently reported a peak capacity of 1,400 in a 40 min separation window using UPLC coupled with microchip CE using a gated interface.53 The 2D nLC × μFFE separations reported here yield a peak capacity production rate of 105 peaks/min, nearly double the previously reported rate using online LC × LC.16 As described in eq 3, the fractional coverage of separation space should also be taken into account to determine a more representative measure of the usable peak capacity. Peaks are well distributed across the high density peak region shown in Figure 6b, suggesting that excellent orthogonality is achieved between the nLC and μFFE separations in this region. Examining the full 2D separation window demonstrates that there are significant areas that are void of peaks, suggesting that peak capacity in these regions is inaccessible for the current combination of separation modes and sample. For example, as shown Figure 6a, no analytes were deflected by more than 1 mm toward the cathode, while analytes were observed that were deflected almost 5 mm toward the anode. Using a minimum convex hull,11 the fraction of separation space where peaks were observed ( f) was determined to be 33%, resulting in a realized 2D peak capacity of 776. Optimization of μFFE conditions to make better use of the full separation channel would dramatically improve f and consequently the realized 2D peak capacity. Similarly, further exploration of combinations of
■
CONCLUSIONS
The continuous nature of μFFE separations facilitates a unique solution to the under sampling limitation of existing liquid phase 2D separations. Coupling with μFFE provided a 24-fold improvement in peak capacity with no increase in analysis time or restrictions on the nLC separation conditions. Peak capacities as high as 2,352 were observed in a 10 min separation window. The CCD used for fluorescence measurements in the μFFE separation recorded an image of the detection zone every 500 ms. Much faster CCD sampling rates are possible, suggesting that faster 2D separations are also feasible using this approach. To date, μFFE applications have largely been limited to LIF detection. In the future, we anticipate the development of a range of μFFE detectors that make use of absorbance, electrochemical sensors, and possibly even mass spectrometry. The described nLC × μFFE interface is simple with no complicated valving or timing required. Our μFFE device was coupled with commercially available nLC instrumentation indicating widespread compatibility of the approach. It should also be recognized that many combinations of nLC stationary phases and μFFE separation modes are available, suggesting that 2D nLC × μFFE could be successfully applied to a wide range of analyte mixtures. 5141
dx.doi.org/10.1021/ac500939q | Anal. Chem. 2014, 86, 5136−5142
Analytical Chemistry
■
Article
(28) Raymond, D. E.; Manz, A.; Widmer, H. M. Anal. Chem. 1996, 68, 2515−2522. (29) Zhang, C. X.; Manz, A. Anal. Chem. 2003, 75, 5759−5766. (30) Kohler, S.; Weilbeer, C.; Howitz, S.; Becker, H.; Beushausen, V.; Belder, D. Lab Chip 2011, 11, 309−314. (31) Albrecht, J. W.; Jensen, K. F. Electrophoresis 2006, 27, 4960− 4969. (32) Xu, Y.; Zhang, C. X.; Janasek, D.; Manz, A. Lab Chip 2003, 3, 224−227. (33) Jezierski, S.; Belder, D.; Nagl, S. Chem. Commun. 2013, 49, 904− 906. (34) Fonslow, B. R.; Bowser, M. T. Anal. Chem. 2005, 77, 5706− 5710. (35) Wang, P. L.; Zhang, L. H.; Shan, Y. C.; Cong, Y. Z.; Liang, Y.; Han, B.; Liang, Z.; Zhang, Y. K. J. Sep. Sci. 2010, 33, 2039−2044. (36) Lu, H.; Gaudet, S.; Schmidt, M. A.; Jensen, K. F. Anal. Chem. 2004, 76, 5705−5712. (37) Kohlheyer, D.; Besselink, G. A. J.; Schlautmann, S.; Schasfoort, R. B. M. Lab Chip 2006, 6, 374−380. (38) Kohlheyer, D.; Eijkel, J. C. T.; Schlautmann, S.; van den Berg, A.; Schasfoort, R. B. M. Anal. Chem. 2007, 79, 8190−8198. (39) Fonslow, B. R.; Barocas, V. H.; Bowser, M. T. Anal. Chem. 2006, 78, 5369−5374. (40) Kohlheyer, D.; Eijkel, J. C. T.; Schlautmann, S.; van den Berg, A.; Schasfoort, R. B. M. Anal. Chem. 2008, 80, 4111−4118. (41) Mazereeuw, M.; de Best, C. M.; Tjaden, U. R.; Irth, H.; van der Greef, J. Anal. Chem. 2000, 72, 3881−3886. (42) Han, B.; Wang, P. L.; Zhu, G. J.; Zhang, L. H.; Qu, F.; Deng, Y. L.; Zhang, Y. K. J. Sep. Sci. 2009, 32, 1211−1215. (43) Wen, J.; Wilker, E. W.; Yaffe, M. B.; Jensen, K. F. Anal. Chem. 2010, 82, 1253−1260. (44) Kostal, V.; Fonslow, B. R.; Arriaga, E. A.; Bowser, M. T. Anal. Chem. 2009, 81, 9267−9273. (45) Moritz, R. L.; Ji, H.; Schutz, F.; Connolly, L. M.; Kapp, E. A.; Speed, T. P.; Simpson, R. J. Anal. Chem. 2004, 76, 4811−4824. (46) Xie, H. W.; Onsongo, G.; Popko, J.; de Jong, E. P.; Cao, J.; Carlis, J. V.; Griffin, R. J.; Rhodus, N. L.; Griffin, T. J. Mol. Cell. Proteomics 2008, 7, 486−498. (47) Frost, N. W.; Bowser, M. T. Lab Chip 2010, 10, 1231−1236. (48) Fonslow, B. R.; Bowser, M. T. Anal. Chem. 2006, 78, 8236− 8244. (49) Iki, N.; Yeung, E. S. J. Chromatogr., A 1996, 731, 273−282. (50) Shackman, J. G.; Watson, C. J.; Kennedy, R. T. J. Chromatogr., A 2004, 1040, 273−282. (51) Filgueira, M. R.; Castells, C. B.; Carr, P. W. Anal. Chem. 2012, 84, 6747−6752. (52) Turner, E. H.; Dickerson, J. A.; Ramsay, L. M.; Swearingen, K. E.; Wojcik, R.; Dovichi, N. J. J. Chromatogr., A 2008, 1194, 253−256. (53) Mellors, J. S.; Black, W. A.; Chambers, A. G.; Starkey, J. A.; Lacher, N. A.; Ramsey, J. M. Anal. Chem. 2013, 85, 4100−4106. (54) Craig, D. B.; Wetzl, B. K.; Duerkop, A.; Wolfbeis, O. S. Electrophoresis 2005, 26, 2208−2213.
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors gratefully acknowledge Dr. Robert C. Allen for his help in developing and implementing the in-house Matlab codes to analyze data. This research was funded by the National Science Foundation (grant #1152022). μFFE devices were fabricated at the University of Minnesota Nano Fabrication Center, a NNIN funded site.
■
REFERENCES
(1) Giddings, J. C. Anal. Chem. 1984, 56, 1258A−1270A. (2) Friedman, D. B.; Hoving, S.; Westermeier, R. In Guide to Protein Purification, Second ed.; Burgess, R. R., Deutscher, M. P., Eds.; Elsevier Academic Press Inc: San Diego, 2009; Vol. 463, pp 515−540. (3) Bushey, M. M.; Jorgenson, J. W. Anal. Chem. 1990, 62, 161−167. (4) Bushey, M. M.; Jorgenson, J. W. Anal. Chem. 1990, 62, 978−984. (5) Malerod, H.; Lundanes, E.; Greibrokk, T. Anal. Methods 2010, 2, 110−122. (6) Francois, I.; Sandra, K.; Sandra, P. Anal. Chim. Acta 2009, 641, 14−31. (7) Seeley, J. V.; Seeley, S. K. Anal. Chem. 2013, 85, 557−578. (8) Davis, J. M.; Stoll, D. R.; Carr, P. W. Anal. Chem. 2008, 80, 8122−8134. (9) Dugo, P.; Cacciola, F.; Kumm, T.; Dugo, G.; Mondello, L. J. Chromatogr., A 2008, 1184, 353−368. (10) Gilar, M.; Fridrich, J.; Schure, M. R.; Jaworski, A. Anal. Chem. 2012, 84, 8722−8732. (11) Rutan, S. C.; Davis, J. M.; Carr, P. W. J. Chromatogr., A 2012, 1255, 267−276. (12) Murphy, R. E.; Schure, M. R.; Foley, J. P. Anal. Chem. 1998, 70, 1585−1594. (13) Davis, J. M.; Stoll, D. R.; Carr, P. W. Anal. Chem. 2008, 80, 461−473. (14) Stoll, D. R.; Carr, P. W. J. Am. Chem. Soc. 2005, 127, 5034− 5035. (15) Stoll, D. R.; Cohen, J. D.; Carr, P. W. J. Chromatogr., A 2006, 1122, 123−137. (16) Filgueira, M. R.; Huang, Y.; Witt, K.; Castells, C.; Carr, P. W. Anal. Chem. 2011, 83, 9531−9539. (17) Huang, Y.; Gu, H. W.; Filgueira, M.; Carr, P. W. J. Chromatogr., A 2011, 1218, 2984−2994. (18) Horie, K.; Kimura, H.; Ikegami, T.; Iwatsuka, A.; Saad, N.; Fiehn, O.; Tanaka, N. Anal. Chem. 2007, 79, 3764−3770. (19) Kraly, J. R.; Jones, M. R.; Gomez, D. G.; Dickerson, J. A.; Harwood, M. M.; Eggertson, M.; Paulson, T. G.; Sanchez, C. A.; Odze, R.; Feng, Z. D.; Reid, B. J.; Dovichi, N. J. Anal. Chem. 2006, 78, 5977− 5986. (20) Chen, X. G.; Fazal, M. A.; Dovichi, N. J. Talanta 2007, 71, 1981−1985. (21) Dickerson, J. A.; Ramsay, L. M.; Dada, O. O.; Cermak, N.; Dovichi, N. J. Electrophoresis 2010, 31, 2650−2654. (22) Cong, Y. Z.; Zhang, L. H.; Tao, D. Y.; Liang, Y.; Zhang, W. B.; Zhang, Y. K. J. Sep. Sci. 2008, 31, 588−594. (23) Ramsey, J. D.; Jacobson, S. C.; Culbertson, C. T.; Ramsey, J. M. Anal. Chem. 2003, 75, 3758−3764. (24) Rocklin, R. D.; Ramsey, R. S.; Ramsey, J. M. Anal. Chem. 2000, 72, 5244−5249. (25) Shadpour, H.; Soper, S. A. Anal. Chem. 2006, 78, 3519−3527. (26) Osiri, J. K.; Shadpour, H.; Park, S.; Snowden, B. C.; Chen, Z. Y.; Soper, S. A. Electrophoresis 2008, 29, 4984−4992. (27) Raymond, D. E.; Manz, A.; Widmer, H. M. Anal. Chem. 1994, 66, 2858−2865. 5142
dx.doi.org/10.1021/ac500939q | Anal. Chem. 2014, 86, 5136−5142
