Anal. Chem. 2005, 77, 5072-5075
On-Line Coupling of Gel Electrophoresis and Inductively Coupled Plasma-Sector Field-Mass Spectrometry for the Determination of dsDNA Fragments Wolfram Bru 1 chert and Jo 1 rg Bettmer*
Institute of Inorganic Chemistry and Analytical Chemistry, University of Mainz, Duesbergweg 10-14, D-55099 Mainz, Germany
The on-line coupling of gel electrophoresis (GE) and inductively coupled plasma-mass spectrometry (ICPMS) is described for the first time. The new method combines the separation power of GE for large biomolecules and the high sensitivity and elemental selectivity of ICPMS. This coupling has been achieved by means of gels housed in glass tubes (length 2.5-20 cm; i.d. 0.5-5 mm). The gel is fixed by glass frits permeable for the analytes. After the electrophoretic separation at voltages up to 500 V, the analytes are transferred to the nebulizer of the ICPMS by an eluent stream, which is separated from the electrode chamber by a membrane. This filter blocks molecules with molecular masses larger than 500 Da. Using doublestranded DNA fragments, commercially available standard solutions were analyzed with on-line GE-ICPMS for the first time by monitoring 31P+ with a double-focusing mass spectrometer at a mass resolution of 4000. Although liquid chromatography has shown to be an efficient and alternative method for separating large biomolecules,1 gel electrophoresis (GE) is still the most widely used separation technique in biochemistry, cell biology, and biotechnology.2 This is justified by the wide applicability and the unique variation range of GE concerning the separation mechanisms, which result in excellent performance. The most important types of GE are isoelectric focusing for proteins, isotachophoreses, sodium dodecyl sulfate-polyacrylamide GE, and 2-D GE carried out either in a slab gel or in a capillary (CGE).2 Several methods are available for the detection of the separated analytes and are routinely applied, such as dye staining, blotting, or radiolabeling.2 To obtain further molecular information on the separated species, off-line coupling to soft ionization mass spectrometry has been achieved and becomes more and more significant. Particularly, matrix-assisted laser desorption and ionization (MALDI) has gained importance for studying biomol* To whom correspondence should be addressed: (e-mail)
[email protected]. (1) Premstaller, A.; Oberacher, H.; Huber, C. G. Anal. Chem. 2000, 72, 43864393. (2) Andrews, A. T. Electrophoresis. Theory, Techniques, and Biochemical and Clinical Applications, 2nd ed.; Oxford University Press: New York, 1988.
5072 Analytical Chemistry, Vol. 77, No. 15, August 1, 2005
ecules separated by GE.3,4 On-line coupling of the electrophoretic separation to a detection system only has been achieved for CGE including, e.g., fluorescence, electrochemical detection, and rarely electrospray ionization mass spectrometry (ESI-MS).2,5-7 In 1998, McLeod and colleagues introduced the off-line coupling of slab gel electrophoresis and laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS) for the analysis of cobalt associated with biomolecules in serum.8 This technique combines the high resolution power of GE for large biomolecules and the excellent detection limits and multielement capability of ICPMS for elemental analysis, especially metals.9-11 Recognizing the potential of this hybrid system for detecting elements associated with proteins in biological systems, further applications have been focused on the analysis of selenoproteins12 and the quantification of phosphorus in phosphoproteins.13 Recent papers demonstrated that the combination of 2-D gel electrophoresis and LA-ICPMS is a powerful tool for the determination of the protein phosphorylation in human τ-protein.14 However, beside the great potential of this coupled technique, there are some major drawbacks: (i) high blank concentration of elements can occure (3) Jonsson, A. P.; Aissouni, Y.; Palmberg, C.; Percipalle, P.; Nordling, E.; Daneholt, B.; Jo ¨rnvall, H.; Bergman, T. Anal. Chem. 2001, 73, 5370-5377. (4) Marvin, L. F.; Roberts, M. A.; Fay, L. B. Clin. Chim. Acta 2003, 337, 1121. (5) Freudemann, T.; von Brocke, A.; Bayer, E. Anal. Chem. 2001, 73, 25872593. (6) Von Brocke, A.; Freudemann, T.; Bayer, E. J. Chromatogr., A 2003, 991, 129-141. (7) Barry, J. P.; Muth, J.; Law, S.-J.; Karger, B. L.; Vouros, P. J. Chromatogr., A 1996, 732, 159-166. (8) Neilsen, J. L.; Abildtrup, A.; Christensen, J.; Watson, P.; Cox, A.; McLeod, C. W. Spectrochim. Acta 1998, 53 B, 339-345. (9) Ma, R.; McLeod, C. W.; Tomlinson, K.; Poole, R. K. Electrophoresis 2004, 25, 2469-2477. (10) Che´ry, C. C. In Handbook of Elemental Speciation: Techniques and Methodology; Cornelis, R., Caruso, J., Crews, H., Heumann, K., Eds.; John Wiley & Sons Ltd.: Chichester, 2003. (11) Lustig, S.; Lampaert, D.; De Cremer, K.; De Kimpe, J.; Cornelis, R.; Schramel, P. J. Anal. At. Spectrom. 1999, 14, 1357-1362. (12) Che´ry, C. C.; Gu ¨ nther, D.; Cornelis, R.; Vanhaecke, F.; Moens, L. Electrophoresis 2003, 24, 3305-3313. (13) Wind, M.; Feldmann, I.; Jakubowski, N.; Lehmann, W. D. Electrophoresis 2003, 24, 1276-1280. (14) Becker, J. S.; Boulyga, S. F.; Becker, J. S.; Pickhardt, C.; Damoc, E.; Przybylski, M. Int. J. Mass Spectrom. 2003, 228, 985-997. 10.1021/ac050425+ CCC: $30.25
© 2005 American Chemical Society Published on Web 06/28/2005
Table 1. Instrumental Parameters for the ICPMS System ICP System instrument rf power auxiliary gas flow coolant gas flow nebulizer gas flow sampler cone skimmer cone dwell time mass resolution (R) isotopes monitored nebulizer spray chamber flow rate
Element 2 1300 W 0.8 L min-1 16 L min-1 0.8 L min-1 Pt, 1.0-mm orifice Pt, 0.7-mm orifice 150 ms 4000 (medium) 31P, 103Rh
Sample Introduction PFA µ-flow Scott type 100 µL min-1 Figure 1. Experimental setup of the GE-ICPMS coupling.
in gel,15 (ii) accurate quantification is still difficult in LA-ICPMS,16 and (iii) after the destructive ablation process, either directly from the gel or after blotting, the analyte itself is not available anymore. Thus, for complementary investigations, e.g., MALDI-MS, a second reproducible and time-consuming separation is necessary. To overcome these disadvantages, the development of an on-line coupling is desirable, which has not yet been published to our knowledge. In this paper, we attempted the on-line hyphenation of GE and ICPMS, which offers alternative possibilities to those explored until now with the available chromatographic and electrophoretic techniques. For practical considerations, the gel was placed in glass tubes and fixed by membranes to allow adequate manipulation, similar to the liquid chromatographic columns. The unique separation power of GE for DNA fragments in combination with the elemental selectivity and sensivity of ICPMS as detector could serve as a new tool for the qualitative and quantitative analysis of large heteroatom-containing and metal-containing biomolecules. In particular, these preliminary investigations are focused on the separation and detection of dsDNA fragments. EXPERIMENTAL SECTION Instrumentation. ICPMS System. The Element 2 inductively coupled plasma-sector field-mass spectrometer (ICP-SF-MS) (Thermo Finnigan MAT GmbH, Bremen, Germany) was equipped with a µ-flow nebulizer (Elemental Scientific, Inc., Omaha, NE) and a Scott-type spray chamber (Glass Expansion, Pocasset, MA). Measuring at medium resolution mode was necessary because spectral interferences such as 15N16O+ (30.995 u) and14N16O1H+ (31.005 83 u), all with nominal mass-to-charge ratio of 31, overlapped the 31P+ signal (30.9738 u) significantly. The required resolution to resolve the 31P+ signal from its adjacent 15N16O+ is calculated to be 1461 using m/∆m (10% valley value).17,18 Medium resolution (R ) 4000) was sufficient to separate the abovementioned spectral interferences. Gas flows and rf power of the ICPMS system were daily optimized for optimal 31P+ detection by continuous injection of 20 µg‚L-1 PO43- standard solution in (15) Becker, J. S.; Zoriy, M.; Becker, J. S.; Pickhardt, C.; Przybylski, M. J. Anal. At. Spectrom. 2004, 19, 149-152. (16) Durrant, S. F. J. Anal. At. Spectrom. 1999, 14, 1385-1403. (17) Helfrich, A.; Bettmer, J. J. Anal. At. Spectrom. 2004, 19, 1330-1334. (18) Stu ¨ wer, D.; Jakubowski, N. J. Mass Spectrom. 1998, 33, 579-590.
0.09 mol‚L-1 Tris-Borate-EDTA (TBE) 1× buffer. Afterward, the capillary of the µ-flow nebulizer was directly connected to the GE system. Details of the operating conditions used throughout this work are given in Table 1. GE System. The principal setup of the developed system is demonstrated in Figure 1. The agarose gel of different concentrations is fixed in glass tubes of variable length (2.5-20 cm) and inner diameter (0.5-5 mm) and fixed outer diameter (8 mm). These tubes are set between the Pt electrode chambers. The upper one (cathode) is directly connected to the separation unit and allows the manual sample injection via a HPLC syringe onto the top of the gel. The construction of the buttom electrode chamber (anode) differs. Due to the necessity of the analyte’s transport to the plasma, an additional tube (i.d. 1 mm) is integrated orthogonal to the gel tube. This tube is separated from the electrode chamber by a special membrane (cellulose acetate, BioRad, Munich, Germany) with a molecular mass cutoff of 500 Da (larger molecules are blocked). This ensures that the analyte cannot pass the membrane, while the electrical connection between the electrodes is still guaranteed. To keep the ionic strength within the electrode chambers constant, the electrode buffers are continuously exchanged using a peristaltic pump at a flow rate of 100 µL‚min-1. The transport of the separated and eluted analytes is performed by a self-aspirating µ-flow nebulizer (flow rate, 100 µL‚min-1). Reagents. All solutions were prepared using ultrapure water (Milli-Q water purification system, Millipore, Bedford, MA). Buffer solutions (TBE 1×) were daily prepared by dissolving 10.9 g of tris(hydroxymethyl)aminomethane (Tris), 5.57 g of boric acid, and 0.37 g of ethylenediaminetetraacetic acid disodium salt (all DNAse, RNAse, protease free, from Acros, Geel, Belgium) in 1 L of water. The elution buffer contains in addition 10 µg‚L-1 Rh (rhodium AA-standard solution 1000 µg‚mL-1 Rh in 20% HCl, Alfa, Karlsruhe, Germany) as internal standard in order to monitor the sample transport and the nebulization. Phosphate standard solution (Specpure 1000 µg‚mL-1, Alfa) and adenosine 5-triphosphate disodium salt hydrate (ATP) 98% (Acros) were used for preliminary studies. DNA standards (DNA QuantLadder, DNA Reverse QuantLadder, and 100-bp DNA Ladder) and agarose (SeaKam LE and MetaPhor agarose) were all purchased from Cambrex (Rockland, MA). For the preparation of polyacrylamide gels, the following chemicals are used: acrylamide, N,N′-methylenebisacrylAnalytical Chemistry, Vol. 77, No. 15, August 1, 2005
5073
Table 2. Operating Parameters of the GE System voltage electrode buffers eluent flow rates sample volume gel length gel i.d. gel materials
250-300 0.09 mol‚L-1 TBE 1×, pH 8.0 0.09 mol‚L-1 TBE 1×, pH 8.0, 10 µg‚L-1 Rh 100 µL‚min-1 5-50 µL 2.5-20 cm 0.5-5 mm agarose, polyacrylamide
amide (Bis) (both for molecular biology, min. 99%, Sigma, Taufkirchen, Germany), N,N,N′,N′-tetramethylethylenediamine (TEMED), electrophoresis grade 99% (Acros), and ammonium persulfate (APS) 98+% (Sigma). Both argon for the ICP and helium, to degas the eluent solutions, consisted of 99.996 vol % (both Westfalen AG, Mu¨nster, Germany). Prior to analysis, all analyzed standards were diluted in a mixture of 0.09 mol‚L-1 TBE 1× and 0.2 g‚mL-1 D(+)-sucrose to the appropriate concentration. Gel Preparation. Before preparing the agarose gels, the inner surfaces of the glass tubes have to be coated with Seakam LE agarose in order to improve fixing of the gel to the glass walls. This step is carried out by soaking the glass tubes in hot agarose solution (1% (w/v) in water) in a vertical position for 5 min and then drying for 2 h at 60 °C. For preparation of the appropriate concentration (e.g., 1.8% (w/v)), agarose (Metaphor) is weighed in exactly and filled to 100 mL with MQ water. This suspension is heated to 90 °C in a water bath to melt the agarose. After complete melting, the solution is cooled to 60 °C, filled into the prepared glass tubes, and afterward both ends are cut. The gel itself is then fixed with cellulose membranes. In the case of the polyacrylamide gels (20% T, 4% C), 0.8 g of Bis and 19.2 g of acrylamide are filled to 100 mL with degassed 0.09 mol‚L-1 TBE 1× buffer, and 50 µL of APS (10% (w/w)) and 5 µL of TEMED are added, vigorously agitated, and then filled into the untreated glass tubes. After complete polymerization, the gel are ready to use. (Safety note: During all steps of gel preparation, gloves and good ventilation at the working place are necessary. Acrylamide and Bis are carcinogens and extremely toxic!) Before using the prepared gels, it is necessary to decontaminate them. For this, the electrophoretic cell is started at a voltage of 250 V for 15 min to elute possible contaminants. These gels stored under buffer solution can be used for several weeks without changing separation characteristics. RESULTS AND DISCUSSION First experiments were directed to evaluate the functional efficiency of the developed system. For this purpose, phosphate and ATP were chosen as model substances to test the system. After optimization of experimental parameters affecting the separation (Table 2), typical electropherograms such as those shown in Figure 2 could be obtained. As noted, relatively broad peaks were obtained (e.g., fwhm for phosphate, 3 min) with a slight tailing. These parameters were mainly influenced by the gel composition and its preparation. But, nevertheless, after peak elution the phosphorus signal reached the baseline level very rapidly, which indicated that the memory effects within the 5074 Analytical Chemistry, Vol. 77, No. 15, August 1, 2005
Figure 2. Polyacrylamide gel electropherogram of phosphate and ATP. (Conditions: PA (20% T, 4% C); length, 2.5 cm; i.d., 3 mm; voltage, 400 V; injection volume, 5 µL).
Figure 3. Gel electropherogram of a mixed DNA QuantLadder. (Conditions: agarose (1.8%); length, 3 cm; i.d., 3 mm; voltage, 250 V; injection volume, 10 µL; concentration, 100 ng each fragment).
coupling system were negligible. Control measurements of injecting blank samples after GE separation of the analytes confirmed these findings. To investigate the separation of double-stranded DNA (dsDNA) fragments, agarose gels were used. In slab GE, a typical agarose concentration of 1.8% is often applied for the separation of 1001000-bp dsDNA fragments.2 Therefore, our experiments were conducted under these conditions. Figure 3 represents a typical electropherogram of five DNA fragments obtained at a voltage of 250 V. Although the 100- and 200-bp fragments were well separated, the separation of the larger molecules suffered from a distinct tailing phenomenon. Consequently, the peak resolution decreased with growing size of the DNA fragments. To overcome this disadvantage, further optimization procedures were carried out. It was observed that the applied voltage was one of the most influencing factors. Better separation results could be obtained for increasing voltages. At a constant voltage of 300 V, an electropherogram of a 100-bp DNA ladder standard is shown in Figure 4. The fragments containing 100-700 base pairs were well separated, and the previously observed tailing improved enormously. As a result, the
Figure 4. Gel electropherogram of a 100-bp DNA ladder. (Conditions: agarose (1.8%); length, 3 cm; i.d., 3 mm; voltage, 300 V; injection volume, 10 µL).
peak widths remained constant during the electrophoretic run. But, there were two effects remaining: (i) larger DNA fragments between 800 and 1000 bp could not be separated under these conditions (migration time between 43 and 50 min) and (ii) there was decreasing peak height for larger compounds. This effect can be explained by the inefficient sample transport within the tubes and probably in the spray chamber. With increasing molecular mass (e.g., for a 1000-bp DNA fragment, ∼600 kDa) the adsorption raises up, resulting in a deficient rinsing. This is mainly caused by the pH of the eluent (pH 8.0), which is not most advantageous in sample introduction in ICPMS. However, lower pH of the eluent buffer affected the separation very strongly, leading to unresolved peaks or even to impossible compound elution from the gel. As a consequence of these observations, we kept the pH at 8.0. Under these conditions, the migration times showed good reproducibility between independent electrophoretic runs (RSD 2-3% (n ) 4)). To exploit the excellent capabilities of ICPMS for elemental quantification, a calibration was recorded for the dsDNA fragments under the conditions summarized in Figure 4. As an example, the calibration curve for the 100-bp fragment is given in the range between 0 and 400 ng of DNA versus the peak area of the relative signal intensity (31P+/103Rh+) (Figure 5). It showed a linear range between 10 and 400 ng but can be easily expanded to higher masses. A detection limit of 0.1 ng absolute could be obtained for phosphorus (equivalent to 1 ng of DNA), which is comparable to the values reported about phosphoproteins.13 However, it has to be mentioned that the detection sensitivity declines for larger DNA fragments. But in conjuction with polymerase chain reaction (PCR), this drawback can be overcome. Nevertheless, these
Figure 5. Calibration curve of a 100-bp DNA fragment.
preliminary results confirmed the applicability of ICPMS in DNA quantification and might provide a reliable and precise method of quantification in real-time PCR systems. CONCLUSIONS The proposed design has resulted in the first electropherograms of an on-line coupling of gel electrophoresis and inductively coupled plasma-mass spectrometry. On the example of DNA fragments, it could be shown that a reproducible separation can be combined with the detection of a ICP-SF-MS for quantification purposes in low-concentration ranges. However, these electropherograms represent the first separation of DNA fragments with following phosphorus detection by ICPMS and might offer an introduction of ICPMS in the analysis of DNA. Current investigations are directed to evaluate the rinsing behavior of large DNA molecules in order to improve peak tailing and enhance 31P detection sensitivity. Further optimization of the system parameters should lead to a device with a high potential in (metallo- and phospho-) proteomics and genomics, especially under the aspects of accurate quantification. Furthermore, it might be interesting to attempt to split the eluent stream for partial introduction into ICPMS and parallel fraction collection for further analysis. Moreover, there is the high prospect to couple the developed electrophoresis system to other detectors such as ESIMS offering information about molecular structures. Received for review March 11, 2005. Accepted May 27, 2005. AC050425+
Analytical Chemistry, Vol. 77, No. 15, August 1, 2005
5075
