Anal. Chem. 1996, 68, 1895-1904
Macromolecule Analysis Based on Electrophoretic Mobility in Air: Globular Proteins Stanley L. Kaufman,* Jeffrey W. Skogen, Frank D. Dorman, and Fahimeh Zarrin†
TSI Incorporated, 500 Cardigan Road, St. Paul, Minnesota 55126 Kenneth C. Lewis
Department of Chemistry, University of North Carolina at Chapel Hill, Venable Hall, CB 3290, Chapel Hill, North Carolina 27599-3290
Globular proteins ranging in molecular mass from 5.7 to 669 kDa were separated and analyzed using an aerosol technique based on the electrophoretic mobility of singlycharged molecular ions in air. The ions were produced by electrospraying and drying 100-nm-diameter droplets of a liquid suspension of the proteins, using ionized air to remove the droplet charge due to the spray process. The electrophoretic mobility was measured using a modified commercial continuous-flow differential mobility analyzer operated near atmospheric pressure. An unmodified commercial condensation particle counter was used for detection. The concentrations analyzed ranged from 0.02 to 200 µg of protein/mL of buffer, with a liquid sample flow rate of approximately 50 nL/min. Sampling time of 3 min was used for each complete distribution measured. The electrophoretic mobilities measured were determined entirely from air flow rates, apparatus geometry, and applied potentials. Results were expressed as electrophoretic mobility equivalent diameters using a Millikan formula. Methods for Size Characterization of Macromolecules. Many methods are currently used for characterizing the size and molecular mass of macromolecules in the range from 10 kDa upwards.1 Of these, size-exclusion or gel permeation chromatography (SEC) is one of the most common. SEC is a relatively slow batch process and suffers from column aging and poor columnto-column reproducibility, requiring careful handling of the columns and frequent calibration. Gel electrophoresis, in plates or columns, also yields size information when a relation between size and mobility in solution can be assumed. Although simple and inexpensive to implement, this method is slow and labor intensive and must be calibrated, usually by the use of markers. Capillary electrophoresis is fast and has extremely high resolution, but sensitivity at low macromolecule concentrations is limited unless molecule-specific labeling methods are implemented or unless it is combined with mass spectrometry (MS). MS using electrospray ionization2 (ESI-MS) or laser desorption and timeof-flight3 (MALDI-TOF) methods can deliver very accurate mo† Present address: 3312 W. Burgundy Ct., Mequon, WI 53092. (1) See, e.g.: Cantor, C. R.; Schimmel, P. R. Biophysical Chemistry, Part 2: Techniques for the Study of Biological Structure and Function; W. H. Freeman and Co., New York, 1980. (2) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64.
S0003-2700(95)01128-0 CCC: $12.00
© 1996 American Chemical Society
lecular masses through perhaps 200 kDa from very small samples but at relatively high capital and operating cost. Multiple peaks observed in ESI-MS, due to the broad envelope of charge states present for a single analyte molecule in this size range, produce a rich spectrum which has been found useful for identifying fragments and validating peak assignments; nevertheless, the complexity of the spectrum of pure analytes somewhat limits the ability of the ESI-MS technique to analyze mixtures.4 Fragmentation of macromolecules is also a feature of MALDI-TOF.5,6 MS determines the ratio of mass to charge and is not directly sensitive to the size or shape of the molecules, although techniques have been devised to infer this type of information from MS measurements.7-9 Other macromolecule size characterization methods include low-angle light scattering, quasi-elastic light scattering, osmometry, and viscometry. These methods require significant sample quantities and relatively high concentrations and are limited to low resolution or, in some cases, to the measurement of a single moment of the size distribution. When multiple components are present or suspected, these methods generally deliver ambiguous results. We describe here a new method for characterizing macromolecules and demonstrate it for globular proteins. The method is based on measuring the electrophoretic mobility of singly-ionized molecules in air. The system measures a complete distribution over a size range corresponding to molecular masses from below 10 kDa to above 600 kDa in a few minutes. Range changes are performed by changing only the applied potentials, so that wideor narrow-range scans may be selected at will using the same hardware. The resolution of the system is comparable to that of SEC, without the use of conventional columns and separation media. The system’s calibration is inherently stable and depends only on readily measured quantities. The total amount of liquid sampled during a typical measurement is 250 nL of liquid, at (3) McKeown, P. J.; Johnston, M. V.; Murphy, D. M. Anal. Chem. 1991, 63, 2069. (4) Cheng, X.; Gale, D. C.; Udseth, H. R.; Smith, R. D. Anal. Chem. 1995, 67, 586. (5) Karas, M.; Bahr, U.; Strupat, K.; Hillenkamp, F.; Tsarbopoulos, A.; Pramanik, B. N. Anal. Chem. 1995, 67, 675. (6) Zhu, L.; Parr, G. R.; Fitzgerald, M. C.; Nelson, C. M.; Smith, L. M. J. Am. Chem. Soc. 1995, 117, 6048. (7) Suckau, D.; Shi, Y.; Beu, S. C.; Senko, M. W.; Quinn, J. P.; Wampler, F. M. III; McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 790. (8) Wood, T. D.; Chorush, R. A.; Wampler, F. M., III; Little, D. P.; O’Connor, P. B.; McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 2451. (9) Cox, K. A.; Julian, R. K., Jr.; Cooks, R. G.; Kaiser, R. E., Jr. J. Am. Soc. Mass Spectrom. 1994, 5, 127.
Analytical Chemistry, Vol. 68, No. 11, June 1, 1996 1895
typical protein concentrations between 0.1 and100 µg/mL of buffer, depending on the molecular mass of the sample. The detector used here counts single analyte molecules, thus the firstorder response is to number concentration rather than mass concentration of the analyte molecules. Detection and Electrophoretic Mobility Characterization of Nanometer Aerosol Particles. The flow-through condensation particle counter (CPC)10 detects aerosol particles in the submicrometer range with efficiencies approaching unity. This instrument and its successors have been described and characterized elsewhere.11-15 Briefly, an aerosol sample flow is merged with a flow of saturated vapor, which is cooled to produce a welldefined supersaturation. Particles larger than a threshold size, the Kelvin radius16 at this value of supersaturation, nucleate the condensation of droplets. The droplets are allowed to grow until they are large enough to be counted efficiently by an optical method. By careful control of the condenser conditions and the sample flow introduction, Stoltzenburg14,15 extended the size limits of this aerosol detection method down to diameters around 3 nm, and this extension has been implemented commercially. Submicrometer aerosol particles are routinely characterized by measuring their electrophoretic mobility in air using the differential mobility analyzer (DMA).17,18 This device has also been modified, described, and characterized extensively.15,19-24 Briefly, a flow of aerosol is introduced through a slit in the outer cylinder of a coaxial electrode pair. A laminar axial flow of particlefree air is established between the electrodes. A potential applied to the center electrode creates a radial electric field, causing charged particles to migrate across the flow toward the center rod. A gap opening into the center rod allows particles whose mobilities fall within a narrow range to exit to the particle detector. The resolution depends in first order only on the ratio between the sample flow and the larger axial flow. Limits on resolution are set by flow eddies, at high flow by the onset of turbulence, and for the smallest particles by diffusion. The absolute volume flow rates of the sample and axial flows, the dimensions of the electrodes, and the applied potential uniquely determine the electrophoretic mobility of particles transmitted through the DMA to the CPC detector. With care, all of these parameters can be determined to better than 1%, implying the possibility of an (10) Agarwal, J. K.; Sem, G. J. J. Aerosol Sci. 1980, 343. (11) Bartz, H.; Fissan, H.; Helsper, C.; Kousaka, Y.; Okuyama, K.; Fukushima, N.; Keady, P. B.; Kerrigan, S.; Fruin, S. A.; McMurry, P. H.; Pui, D. Y. H.; Stoltzenburg, M. R. J. Aerosol Sci. 1985, 443. (12) Kesten, J.; Reineking, A.; Porstendoerfer, J. Aerosol Sci. Technol. 1991, 107. (13) Caldow, R.; Wiedensohler, A.; Hansson, H.-C. Proceedings of the Thirteenth International Conference on Nucleation and Atmospheric Aerosols, Salt Lake City, UT, 24-28 August, 1992. (14) McDermott, W. T.; Ockovic, R. C.; Stolzenburg, M. R. Aerosol Sci. Technol. 1991. (15) Stoltzenburg, M. Ph.D. Thesis, Mechanical Engineering Department, University of Minnesota, Minneapolis, MN, 1988. (16) Hinds, W. C. Aerosol Technology; Wiley-Interscience: New York, 1982; Chapter 13. (17) Knutson, E. O.; Whitby, K. T. J. Aerosol Sci. 1975, 6, 443. (18) Knutson, E. O.; Whitby, K. T. J. Aerosol Sci. 1975, 6, 453. (19) Ahn, K.-H.; Liu, B. Y. H. J. Aerosol Sci. 1990, 21, 249. (20) Ahn, K.-H.; Liu, B. Y. H. J. Aerosol Sci. 1990, 21, 263. (21) Kousaka, Y.; Okuyama, K.; Adachi, M.; Mimura, T. J. Chem. Eng. Jpn. 1986, 19, 401. (22) Winklmayr, W.; Reischl, G. P.; Lindner, A. O.; Berner, A. J. Aerosol Sci. 1991, 22, 289. (23) Pourprix, M. U.S. Patent 5,117,190, 1992. Pourprix, M. U.S. Patent 5,150,036, 1992. (24) Zhang, S.-H.; Akutsu, Y.; Russell, L. M.; Flagan, R. C.; Seinfeld, J. H. Aerosol Sci. Technol. 1995, 23, 357.
1896
Analytical Chemistry, Vol. 68, No. 11, June 1, 1996
absolute electrophoretic mobility measurement accurate to within a few percent or better without direct calibration using molecular standards. One of the aims of the present work was to explore this possibility experimentally. Generation of Macromolecule Aerosols. Ideally, chargeneutral sprayed droplets of an absolutely pure liquid containing only dissolved or suspended macromolecules, on evaporation, would leave a residue consisting of only the macromolecules in aerosol form. This ideal picture is considerably modified by the presence of dissolved small-molecule solutes and by electrical charge if present. With practical liquid purities, it is necessary to spray extremely small droplets to avoid generating particles from the nonvolatile residue which are large enough to be detected by the CPC. Electrospray (ES) nebulizers are capable of generating droplets well below 1 µm in diameter when operated appropriately, but they produce highly charged droplets. When an electrosprayed droplet evaporates, it quickly reaches the Rayleigh limit where Coulombic repulsion overcomes surface tension, and the droplet disintegrates.25 In the process, multiplycharged analyte ions leave the droplet. Regardless of the details of this process, it is this path which is exploited as the means of ion generation in ESI-MS.2 We have constructed a neutralized ES nebulizer in which the high charge of the primary droplets is removed rapidly, by using a source of R radiation to ionize the air surrounding the droplets. The air ions, having high mobility and diffusivity, reach the charged droplets at a sufficient rate to prevent Rayleigh disintegration. Thus in our system, the primary droplets lose their liquid by evaporation, rather than by ejecting daughter droplets or ions, leaving free analyte molecules, possibly with some residual bound solvent and nonvolatile solute. Lewis et al.26 have used the same neutralized ES method with a CPC as a detector of macromolecules in the effluent from a packed capillary size exclusion chromatography column. Chen et al.27 have studied a nearly identical neutralized ES nebulizer and report conditions for producing highly monodisperse droplets ranging from 40 nanometers to 1.8 µm in diameter. The conditions for obtaining a calculable steady-state distribution of charge on aerosols by means of radioactive ionizers have been analyzed by Liu and Pui,28,29 and the steady-state charge distribution on aerosols in ionized gases was first discussed by Fuchs.30 The Fuchs distribution predicts small probabilities of charging for particles in the nanometer range; for example, at 3 nm diameter, the probability of single charge is 1.1% for positive and 1.2% for negative charge, falling to negligible values (
