Temperature gradient focusing for microchannel ... - Springer Link

3 downloads 0 Views 149KB Size Report
Nov 11, 2006 - Electrophoresis). Anal Bioanal Chem (2007) 387:155–158. DOI 10.1007/s00216-006-0913-4. J. G. Shackman .M. S. Munson .D. Ross (*).
Anal Bioanal Chem (2007) 387:155–158 DOI 10.1007/s00216-006-0913-4

TRENDS

Temperature gradient focusing for microchannel separations Jonathan G. Shackman & Matthew S. Munson & David Ross

Published online: 11 November 2006 # Springer-Verlag 2006

Introduction Microfluidic devices are currently evolving beyond a concept technology into a mature field, with examples already evident in commercial applications. Microchips employing free solution electrophoresis have benefited from over 20 years of development in capillary electrophoresis (CE) aimed at enhancing selectivity and sensitivity. Like conventional CE, these devices suffer from relatively poor detection limits because of the extremely small quantities of analyte that are injected onto the separation column and the very short (10–100 μm) optical detection path lengths. These issues become exacerbated in microfluidics, which can have integrated injection schemes allowing for even smaller injection plugs than traditionally afforded, as well as exceedingly small cross-sectional channel areas (for a review, see [1]). Additionally, many of the microchip substrate materials, whether glass-based or polymeric, can have high backgrounds for fluorescence or other optical detection methods and so degrade detection limits relative to fused silica employed in conventional CE. Sensitivity enhancements have been realized through instrumental detection improvements, novel detection geometries, and detection methods other than fluorescence or absorbance. Several improvements in sample preconcentration prior to the separation step have also evolved to improve detection limits, including sample stacking methJ. G. Shackman : M. S. Munson : D. Ross (*) Analytical Chemistry Division, National Institute of Standards and Technology, 100 Bureau Dr., MS 8394, Gaithersburg, MD 20899-8394, USA e-mail: [email protected]

ods, sweeping methods, and isotachophoresis, as well as chromatographic methods. Another group of analyte concentration methods have been termed equilibrium gradient focusing methods, with the most prevalent example in microchannels being isoelectric focusing, where the separation arises from variations in analyte isoelectric points along a pH gradient. Gradient methods combine concentration and separation steps by forcing analytes to a unique equilibrium point along the separation axis; analytes focus at their unique point of null velocity, with the separation based upon analytes having differing equilibrium points. Due to the nature of the focusing, peaks become both narrower and more concentrated throughout the separation, allowing for high resolution and sensitivity. A novel new equilibrium method was introduced for proteins by Koegler and Ivory using a variable electric field [2, 3]. Electric field gradient focusing (EFGF) has been further developed in recent years by the groups of Lee and Woolley [4, 5], as well as by Myers and Bartle [6]; at present the method has been successfully integrated into microfluidic devices [7–9]. As an alternative to pH or electric field gradients, a temperature gradient can be employed, as has been shown with temperature gradient focusing (TGF), which is based upon balancing analytes’ electrophoretic mobilities against a bulk flow containing both hydrodynamic and electroosmotic flow components (Fig. 1). TGF has been applied to a wide variety of analytes, including proteins, DNA, small dye molecules, and amino acids, in both capillary and microfluidic formats (Fig. 2; [10–16]; Kamande MW, Ross D, Locascio LE, Warner IM, 2006, submitted to Anal Chem; Huber DE, Santiago JG, 2006, submitted to Electrophoresis).

156

Theory and implementation of TGF Counter-flow gradient focusing methods such as EFGF and TGF work by balancing the electrophoretic motion of analyte molecules against the bulk flow of solution through a separation channel. In EFGF, the electrophoretic velocity gradient is created by manipulating the electric field in the channel through an arrangement of electrodes and semipermeable membranes. In TGF the electrophoretic velocity gradient is established by applying a temperature gradient along the length of the channel (Fig. 1). Ross and Locascio initially outlined a theory for TGF based upon earlier work on EFGF [10]. The key insight was that a buffer with a temperature-dependent ionic strength would be required for TGF; otherwise the primary temperature-dependent parameter would be the viscosity. Although a change in viscosity will have an effect on the electrophoretic mobility of an analyte, it will have an equal and opposite effect on the electric field in the separation channel (assuming constant current). Consequently, the velocity, a product of the mobility and the field, will be unaffected by changes in the viscosity. A more complete description of how to select a buffer for TGF was later

Anal Bioanal Chem (2007) 387:155–158

presented by Shackman et al., along with demonstrations of TGF in buffers ranging in pH from 3 to 10.5 [14]. The initial theoretical description of TGF was generalized by Ghosal and Horek to include axial gradients in other buffer properties and to take into account the increased band-broadening caused by Taylor dispersion from induced pressure gradients. They also solved for the time evolution of the position and width of an analyte peak as they approach their steady state values [17]. Huber and Santiago solved for the peak asymmetry caused by induced pressure gradients in the Taylor–Aris regime and found excellent agreement between experimental measurements and their theoretical description (Huber DE, Santiago JG, 2006, submitted to Electrophoresis). The initial report on TGF included a demonstration of the focusing of carboxyfluorescein using a temperature gradient induced by internal Joule heating [10]. Kim et al. expanded on this idea by showing that TGF with Joule heating could be used for the focusing and separation of proteins [16]. In addition, they showed that TGF could be performed with sodium phosphate buffer in contrast to the assertion by Ross and Locascio [10] that typical, strongly ionized buffers would not be useful for TGF. One of the limitations of TGF and other counter-flow gradient focusing techniques is limited peak capacity. Generally, only a small number of analyte peaks can be simultaneously focused and separated. A solution to this problem was recently described by Hoebel et al. [15]. Instead of using a constant bulk flow velocity, they varied the bulk flow over time to sequentially focus and elute analytes past a fixed detection point in a method they called “scanning” TGF. In addition to increasing the peak capacity, the use of scanning TGF also allowed for more reproducible and quantitative measurements with TGF.

TGF with recognition molecules

Fig. 1 Schematic of temperature gradient focusing. Top: The apparatus used for external application of temperature gradient and pressure control. Middle: Temperature profile. Bottom: Plot of electrophoretic, bulk, and net analyte velocities

TGF can also be implemented in formats for chiral analyses [11] and DNA hybridization assays [12]. These two methods depend upon the interaction of the target species with a recognition molecule; either the recognition molecule or the target can be focused with TGF while the unfocused component is allowed to flow freely through the channel. The affinity between the target and the recognition species produces a variation in the focusing behavior, allowing quantification of the target species. For chiral investigations, the enantiomeric mixture of compounds was initially focused to an unresolved zone, which was then followed by the introduction of a neutral chiral selector, β-cyclodextran, into the buffer. The preferential interaction of one enantiomer with the selector as it moved through the zone of focused analytes allowed for resolution of the

Anal Bioanal Chem (2007) 387:155–158

enantiomers in the channel (Fig. 2g). Detection of 0.05% enantiomeric impurities was achieved with up to 1000-fold concentration enhancement of amino acids with minimal sample preparation. In a similar manner, sequence-specific detection of DNA was achieved using TGF by focusing target DNA and observing the interaction with labeled peptide nucleic acids (PNA) that had been added to the background buffer. The hybridization of DNA/PNA produces a significant change in the DNA electrophoretic mobility, resulting in a shift in zone location (Fig. 2e). By altering the analysis steps, single nucleotide polymorphisms (SNP) could be examined by prehybridizing the target DNA with the probe PNA and observing the point within the temperature gradient where melting occurred. The focusing step allowed for a 100-fold improvement in sensitivity over traditional UV melting measurements, and the entire analysis could be performed in less than five minutes.

Micellar affinity gradient focusing A similar but mechanistically different approach to TGF was implemented by Balss et al., whereby a temperature gradient was implemented across a separation length filled with buffer and a surfactant capable of supporting micelles [13]. Effectively resulting in a combination of TGF and micellar electrokinetic chromatography, the method was

157

termed micellar affinity gradient focusing (MAGF), and it allowed for simultaneous concentration and separation of both ionic and neutral species. Due to the temperature dependence of the critical micelle concentration (CMC) of the surfactant (originally sodium dodecyl sulfate; SDS), an effective partitioning gradient is created along the separation length. When an electric field and hydrodynamic flow are superimposed upon this, analytes exhibit a unique zerovelocity point based upon their micelle retention factor. As MAGF is an electrophoretic-based focusing method that is not dependent upon electrophoretic mobility, the method allows for different selectivity than TGF alone, and also eliminates the need to use a buffer exhibiting a temperaturedependent ionic strength. Initial work in a nonscanning mode allowed for focusing and separation of two similar zwitterionic rhodamine dyes (Fig. 2h). Additionally, neutral anthracene was shown to be concentrated at a rate of over 50-fold per minute. Futher work by Kamande et al. investigated the use of polymeric surfactants as the pseudostationary phase in MAGF (Kamande MW, Ross D, Locascio LE, Warner IM, 2006, submitted to Anal Chem); these covalently linked molecular micelles have several advantages over traditional surfactants (e.g., SDS) such as elimination of monomer–aggregate dynamic equilibriums, lower working concentrations due to the exclusion of a CMC, and the ability to support higher organic content within the media. Using achiral poly(sodium undecenyl sulfate) and a scanning mode of MAGF, quantitative concentration and separation of three coumarin dyes was accomplished, with a focusing rate of 10- to 25-fold per minute within a 2 mm gradient.

Outlook

Fig. 2 Examples of the variety of analytes that have been focused with TGF: a Oregon Green (left band) and Cascade Blue (right band) [10]; b FQ-labeled aspartic acid isomers [10]; c CBQCA-labeled serine (left band) and tyrosine (right band) [10]; d Green Fluorescent Protein isoforms [10]; e DNA mutation detection by separation of DNA/PNA hybrids (left band) from unbound DNA (right band) [12]; f micron-sized particles [10]; g chiral separation of dansyl-labeled D(left band) and L-glutamic acid (right band) [11]; h separation of neutral molecules, rhodamine B (red) and rhodamine 110 (green), with MAGF [13]

TGF has been demonstrated to be a useful tool for sample preparation and analysis. The primary advantage of TGF and other equilibrium methods is the focusing (rather than dispersive) nature of the separation. TGF allows for low detection limits (mid-pmol/L input concentrations; [14]) with relatively low-cost and simplistic detection methods (i.e., arc lamp illumination, focusing times of less than 10 min, uncooled and unfiltered CCD imaging, etc.). Relatively short channels can achieve resolution comparable to other techniques utilizing long separation lengths which, in combination with the absence of an injection scheme, makes TGF well-suited to the development of integrated microfluidic systems. Practical applications of TGF and integration with other preparatory/analytical techniques remain an active field of inquiry. Current work at NIST emphasizes incorporation of the pressure and temperature gradient-generating portions into a single microfluidic device, development of microfluidic TGF in

158

combination with a second dimension of separation (e.g., TGF–capillary gel electrophoresis for DNA sequence analysis), and sample preparation-free TGF assays of “dirty” samples, such as whole blood.

References 1. Dittrich PS, Tachikawa K, Manz A (2006) Anal Chem 78:3887– 3908 2. Koegler WS, Ivory CF (1996) Biotechnol Prog 12:822–836 3. Koegler WS, Ivory CF (1996) J Chromatogr A 726:229–236 4. Humble PH, Kelly RT, Woolley AT, Tolley HD, Lee ML (2004) Anal Chem 76:5641–5648 5. Kelly RT, Woolley AT (2005) J Sep Sci 28:1985–1993 6. Myers P, Bartle KD (2004) J Chromatogr A 1044:253–258

Anal Bioanal Chem (2007) 387:155–158 7. Petsev DN, Lopez GP, Ivory CF, Sibbett SS (2005) Lab Chip 5:587–597 8. Kelly RT, Li Y, Woolley AT (2006) Anal Chem 78:2565–2570 9. Liu J, Sun X, Farnsworth PB, Lee ML (2006) Anal Chem 78:4654–4662 10. Ross D, Locascio LE (2002) Anal Chem 74:2556–2564 11. Balss KM, Vreeland WN, Phinney KW, Ross D (2004) Anal Chem 76:7243–7249 12. Balss KM, Ross D, Begley HC, Olsen KG, Tarlov MJ (2004) J Am Chem Soc 126:13474–13479 13. Balss KM, Vreeland WN, Howell PB, Henry AC, Ross D (2004) J Am Chem Soc 126:1936–1937 14. Shackman JG, Munson MS, Kan C-W, Ross D (2006) Electrophoresis 27:3420–3427 15. Hoebel SJ, Balss KM, Jones BJ, Malliaris CD, Munson MS, Vreeland WN, Ross D (2006) Anal Chem 78:7186–7190 16. Kim SM, Sommer GJ, Burns MA, Hasselbrink EF (2006) Anal Chem (accepted) DOI 10.1021/ac061194p 17. Ghosal S, Horek J (2005) Anal Chem 77:5380–5384