Anal. Chem. 1999, 71, 1668-1673
Separation of Double- and Single-Stranded DNA Restriction Fragments: Capillary Electrophoresis with Polymer Solutions under Alkaline Conditions Yu Liu and Werner G. Kuhr*
Department of Chemistry, University of California, Riverside, California 92521
Capillary electrophoresis in buffers containing hydroxyethyl cellulose (HEC) was used to separate double- and single-stranded DNA restriction fragments under neutral and alkaline conditions in epoxy-coated capillaries. It was found that better resolution was achieved using highly entangled HEC solutions for a narrow range of DNA fragment sizes, while lower resolution was obtained over a wide separation range using diluted HEC solutions. Optimal resolution of these DNA fragments was obtained using buffers containing 0.5% HEC at pH 11 with plate numbers exceeding 3 × 106 plates/m. It was also found that the diffusion coefficients and electrophoretic mobilities of DNA fragments decreased with increasing pH. This may indicate a more extended DNA conformation and, therefore, enhancement of transient entanglement coupling between DNA and HEC polymers under alkaline condition. At pH 12, ss-DNA were well separated in entangled HEC solutions; however, the resolution of ssDNA was significantly decreased in diluted polymer solution.
Capillary electrophoresis (CE) has been shown to be an attractive separation technique for DNA fragments due to its high efficiencies and speed compared with conventional gel electrophoresis.1-3 Most DNA separations by CE are performed in crosslinked gels or un-cross-linked polymer solutions that act as a molecular sieving matrix. Although high-resolution separations of DNA were first achieved using gel-filled capillaries,4,5 the preparation of homogeneous gels is quite difficult due to polymerization-induced shrinkage and the appearance of bubbles inside the capillaries. Alternatively, the use of un-cross-linked polymer solutions provided a promising protocol for rapid and efficient DNA separations.6-21 Compared to gels, the un-crosslinked polymer solutions have a dynamic pore structure and are (1) Cohen, A. S.; Najarian, D.; Smith, J. A.; Karger, B. L. J. Chromatogr. 1988, 458, 323-29. (2) Camilleri, P. Capillary Electrophoresis: Theory and Practice; CRC Press: Boca Raton, 1997. (3) Heller, C. J. Chromatogr., A 1995, 698, 19-31. (4) Guttman, A.; Cohen, A. S.; Heigeeer, D. N.; Karger, B. L. Anal. Chem. 1990, 62, 137-40. (5) Chen, Y.; Hoeltje, J. V.; Schwarz, U. J. Chromatogr., A 1994, 680, 63-71. (6) Chiari, M.; Nesi, M.; Righetti, P. G. J. Chromatogr. 1993, 652, 31-9. (7) Ruiz-Martinez, M. C.; Salas-Solano, O.; Carrilho, E.; Kotler, L.; Karger, B. L. Anal. Chem. 1998, 70, 1516-27.
1668 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
more flexible, so that shorter analysis time and higher reproducibility can be obtained. Many entangled polymer solutions, such as linear polyacrylamide (LPA),6-8 modified cellulose,9-12 and poly(ethylene oxide) (PEO),13 have been successfully used in DNA sizing and sequencing. Furthermore, it was shown that ultradilute (HEC) polymer solutions are also effective in the separation of DNA fragments.14-21 The diluted polymer solutions have very low viscosity and therefore provide a relatively rapid separation of DNA. Barron et al. suggested that the separations occur due to the transient entanglement between DNA and modified cellulose molecules, which increases with the length of DNA chain.14,15 DNA fragments up to 23 kbp were successfully separated using this ultradiluted HEC solution under constant field strength.15 Theoretically, the separation of double- (ds-) and singlestranded (ss-) DNA by gel electrophoresis should be strictly according to DNA chain length. This is general true for linear, double-stranded DNA. However, it was found that the electrophoretic mobility of ds-DNA may be influenced by base composition or DNA sequencing.22,23 Single-stranded DNA, on the other hand, will generally fold on itself to different extents to form some secondary structure (i.e., compression).24 Such folding will depend on pH, temperature, and ionic strength, and single strands of the same length may fold in different ways and therefore migrate at (8) Ruiz-Martinez, M. C.; Salas-Solano, O.; Carrilho, E.; Kotler, L.; Karger, B. L. Anal. Chem. 1998, 70, 1528-35. (9) Mitnik, L.; Salome, L.; Viovy, J. L.; Heller, C. J. Chromatogr., A 1995, 710, 309-21. (10) Issaq, H. J.; Chan, K. C.; Muschik, G. M. Electrophoresis 1997, 18, 11538. (11) MacCrehan, W. A.; Rasmussen, H. T.; Northrop, D. M. J. Liq. Chromatogr. 1992, 15, 1063-80 (12) Cheng, J.; Mitchelson, K. R. Anal. Chem. 1994, 66, 4210-4. (13) Fung, E. N.; Yeung, E. S. Anal. Chem. 1995, 67, 1913-9. (14) Barron, A. E.; Soane, D. S.; Blanch, H. W. J. Chromatogr. 1993, 652, 3-16. (15) Barron, A. E.; Blanch, H. W.; Soane, D. S. Electrophoresis 1994, 15, 597615. (16) Barron, A. E.; Sunada, W. M.; Blanch, H. W. Electrophoresis 1996, 17, 74457. (17) Barron, A. E.; Sunada, W. M.; Blanch, H. W. Electrophoresis 1995, 16, 6474. (18) Kim, Y.; Morris, M. D. Anal. Chem. 1994, 66, 3081-5. (19) Kim, Y.; Morris, M. D. Anal. Chem. 1995, 67, 784-6. (20) Shi, X.; Hammond, R. W.; Morris, M. D. Anal. Chem. 1995, 67, 1132-8. (21) Hammond, R. W.; Oana, H.; Schwinefus, J. J.; Bonadio, J.; Levy, R. J.; Morris, M. D. Anal. Chem. 1997, 69, 1192-6. (22) Mertz, J. E.; Berg, P. Proc. Natl. Acad. Sci. U.S.A. 1974, 71, 4879-83. (23) Maniatis, T.; Jeffrey, A.; van deSande, H. Biochemistry 1975, 14, 378794. (24) Rosenberg, A. H.; Studier, F. W. Biopolymers 1969, 7, 765-74. 10.1021/ac9814492 CCC: $18.00
© 1999 American Chemical Society Published on Web 04/03/1999
different rates. Consequently, considerable error may occur in DNA sizing or sequencing due to this secondary structure. Capillary electrophoresis under conditions where the single strands are completely unfolded and are unable to renature should eliminate these problems. Denaturing capillary gel electrophoresis in the presence of urea or formamide has been used for this purpose.25,26 Recently, it was shown that operation of gels at elevated temperatures could also reduce the secondary structures and generate a sequence that is free of compression.27 It seems that capillary electrophoretic separations of DNA under alkaline conditions might be an alternative to these systems, such that the DNA is linear and completely denatured. Alkaline gel electrophoresis was first introduced by McDonell et al. in 1977 and has been widely used in analyzing DNA for size determination, single-strand breaks, and depurinations.28,29 It was found that the DNA band became sharper in alkaline gels than that in neutral gels,28 indicating an increase in DNA resolution under alkaline conditions. No DNA separations have been reported in CE under alkaline conditions, presumably because of the low stability of capillary wall coatings, which makes it difficult to operate at alkaline pH. In this work, we report the separation of DNA restriction fragments by CE with polymer solutions under alkaline conditions. Epoxy-coated capillaries, which are very stable and can eliminate EOF over pH range of 7-12, were used to improve the quality and reproducibility of the separation. Both entangled and ultradiluted hydroxyethyl cellulose solutions were used to compare the resolution of DNA at neutral and alkaline conditions. The pH dependence of the DNA diffusion coefficient in polymer solutions was investigated, and the separation mechanism of DNA in entangled and ultradiluted polymer solutions is also discussed. EXPERIMENTAL SECTION Instrumentation and Materials. Electrophoretic separation of DNA fragments were performed on a Bio-Focus 3000 electrophoresis system with UV detection at 260 nm (Hercules, CA). Negative polarity was applied during the run by using coated capillaries. A ΦX174-HaeIII restriction digest was obtained from Gibco/ BRL (Bethesda, MD) at a concentration of 500 µg/mL. The DNA digest was diluted to 250 µg/mL with distilled, deionized water and electrophoretically injected at 300 V/cm for 4 s or less. Mesityl oxide was used as a neutral marker to measure electroosmotic flow (Aldrich Chemical Co., Milwaukee, WI). The buffer used in this experiment was 89 mmol/L Tris, 89 mmol/L boric acid, and 5 mmol/L EDTA (termed TBE buffer). A measured amount of HEC, molecular weight of 105 000 (Polysciences, Warrington, PA), was added to the buffer solution at 60 °C under mechanical stirring to obtain a homogeneous stock solution with a concentration of 1%. The HEC stock solution was then diluted with TBE buffer to yield semi- and ultradiluted HEC solutions, while 2 M sodium (25) Luckey, J. A.; Smith, L. M. Anal. Chem. 1993, 65, 2841-50. (26) Ruiz-Martinez, M. C.; Belenkii, A.; Foret, F.; Mille, A. W.; Karger, B. L. Anal. Chem. 1993, 65, 2851-8. (27) Zhang, J.; Fang, Y.; Hou, J. Y.; Ren, H. J.; Jiang, R.; Roos, P.; Dovichi, N. J. Anal. Chem. 1995, 67, 4589-93. (28) McDonell, M. W.; Simon, M. N.; Studier, F. W. J. Mol. Biol. 1977, 110, 119-46. (29) Freeman, S. E.; Blackett, A. D.; Monteleone, D. C.; Setlow, R. B.; Sutherland, B. M.; Sutherland, J. C. Anal. Biochem. 1986, 158, 119-29.
Figure 1. Effect of buffer pH on electroosmotic mobility in an epoxycoated capillary. Conditions: 35 cm (30 cm to detector) × 50 µm i.d. epoxy-coated capillary; 89 mmol/L Tris, 89 mmol/L boric acid, and 5 mmol/L EDTA as running buffer and 2 mol/L sodium hydroxide added to obtain desired pH. Mesityl oxide was used as neutral marker; applied voltage, 10 kV UV detection at 260 nm.
hydroxide was added to the buffer to obtain the desired pH prior to electrophoresis. Final HEC concentrations ranged between 0.5 (semidilute) and 0.09% (ultradilute). Capillary Preparation. Fused-silica capillaries (50 µm i.d., 375 µm o.d.) were purchased from Polymicro Technologies, Inc. (Phoenix, AZ) and were coated with epoxy as described previously with a few modifications.30 Briefly, the capillary was rinsed first with acetone for 15 min and dried in an oven at 100 °C for 1 h under a nitrogen pressure of 20 psi. Epoxy 314 ND (Epo-Tek, Billerica, MA) was dynamically coated onto the capillary surface by aspirating a solution of epoxy mixture in acetone. Residual solvent was removed from the epoxy coatings by flushing with nitrogen at room temperature for 30 min. The epoxy coating was then cross-linked at 80 °C for 30 min and 150 °C for 2 h under nitrogen pressure of 20 psi. The coated capillaries were washed with buffer for 30 min prior to use. RESULTS AND DISCUSSION To obtain high efficiency and reproducible CE separations under alkaline conditions, it is essential to use a very stable capillary inner coating that can minimize EOF over a wide pH range. A previous study has shown the epoxy coatings are very stable over pH range of 2-12.30 Figure 1 shows the effect of variations in the pH of the buffer on EOF in an epoxy-coated capillary. DMSO was used as a neutral marker and sodium hydroxide was added to tris-borate buffer to obtain the desired pH. Below pH 7, an anodic EOF was observed due to the positive charge of tertiary amine groups in the epoxy coating.30 Above pH 7, the epoxy surface was neutralized and EOF could not be detected after running for more than 2 h in the pH range of 7-12. The elimination of EOF over a wide pH range, especially under alkaline conditions, makes it possible to change buffer pH freely without influencing the overall mobilities of DNA. Separation of ds-DNA with HEC Solution at Neutral pH. Double-stranded DNA fragments were separated using HEC solutions (Mn 105 000) in 89 mM TBE buffer at pH 8.2. Figure 2 shows the CE separation of an HaeIII digest of ΦX174 in an epoxy(30) Liu, Y.; Fu, R.; Gu, J. J. Chromatogr., A 1996, 723, 157-67.
Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
1669
Figure 2. Electrophoretic separation of ΦX174-HaeIII DNA restriction fragments in semi- and ultradiluted HEC solution at pH 8.2: (a) 0.5, (b) 0.4, (c) 0.15, and (d) 0.09% HEC solution. Conditions: 35 cm (30 cm to detector) × 50 µm i.d. epoxy-coated capillary; 6 kV applied voltage, negative polarity; UV detection at 260 nm; electrokinetic injection at 3 kV for 4 s. Peak identification: (1) 72, (2) 118, (3) 194, (4) 234, (5) 271, (6) 281, (7) 310, (8) 603, (9) 872, (10) 1078, and (11) 1353 bp. Table 1. Dependence of Separation Efficiency of DNA Fragments on Buffer pH and Ionic Strengtha theoretical plate numbers (plates/m) at addition of DNA fragments (bp) 72 118 194 234 271 281 310 603 872 1078 1353
0 mM NaOH NaCl pH 8.2 pH 8.2 533 000 602 000 567 000 733 000 *b * 767 000 140 000 26 000 26 000 26 000
533 000 602 000 567 000 733 000 * * 767 000 140 000 26 000 26 000 26 000
15 mM NaOH NaCl pH 9.0 pH 8.2 506 000 267 000 767 000 1 633 000 * * 1 900 000 1 637 000 1 533 000 1 467 000 1 367 000
633 000 602 000 765 000 703 000 * * 867 000 247 000 73 000 130 000 133 000
28 mM
32 mM
34 mM
NaOH pH 10
NaCl pH 8.2
NaOH pH 11
NaCl pH 8.2
NaOH pH 12
NaCl pH 8.2
2 333 000 2 267 000 2 667 000 2 333 000 1 900 000 1 600 000 2 467 000 1 733 000 1 633 000 1 567 000 1 300 000
1 033 000 730 000 767 000 633 000 * * 900 000 313 000 167 000 126 000 53 000
3 110 000 2 226 000 2 300 000 2 533 000 3 667 000 3 400 000 2 600 000 2 067 000 1 900 000 1 967 000 1 010 000
1 067 000 867 000 800 000 602 000 * * 867 000 400 000 230 000 183 000 90 000
1 310 000 1 900 000 1 930 000 2 233 000 3 030 000 3 100 000 2 600 000 2 060 000 1 910 000 1 560 000 1 630 000
1 010 000 760 000 800 000 630 000 * * 430 000 400 000 240 000 190 000 110 000
a Conditions: Sodium chloride and sodium hydroxide were added to 0.5% HEC in TBE buffer to obtain the same ionic strength respectively. Other conditions are the same as in Figure 3. b *, not detectable due to coelution.
coated capillary with HEC solutions ranging in concentrations from 0.09 to 0.5% (w/w). The resolution of small DNA fragments (72-310 bp) significantly increased with an increase in HEC concentration, but the 271/281-bp fragments can only be partial resolved in the HEC solution which was above the entanglement limit. On the other hand, the separation range for DNA fragments decreased with increasing HEC concentration. DNA fragments larger than 872 bp could not be separated and eluted as one broad peak with a 0.5% HEC solution. Barron et al. showned an improvement in the separation of large-size DNA using ultradilute HEC solutions, far below the entanglement threshold.14,15 Similar results were observed in this work. A wide range of DNA separations was obtained at diluted HEC solutions, and DNA fragments as large as 1350 bp were well separated. However, the 271/281-bp fragments were not well resolved over all of the HEC concentrations used in neutral pH. 1670 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
Separation of ss-DNA Fragments under Alkaline Conditions. Under alkaline conditions, guanine and thymine each lose a hydrogen and become negatively charged, leading to the denaturation of DNA. It was found that denaturation was usually complete by pH 11.5;31 however, partial denaturation may occur at even lower pH. In this work, the separation of ds-DNA and ss-DNA samples was compared over the pH range of 8-13 using 0.5% HEC solutions (Figure 3). DNA fragments eluted in the same order with increasing chain length in both neutral and alkaline conditions. However, the resolution of DNA fragments was dramatically improved under alkaline conditions. The larger DNA fragments (above 872 bp) were nicely separated above pH 9.0, and the 271/281-bp fragments were nearly baseline resolved with a resolution of 1.5 at pH 11. Since the net charge of DNA is (31) Johnson, P. H.; Grossman, L. I. Biochemistry 1977, 16, 4217-24.
Figure 4. Comparison of ss-DNA separations at pH 12 using HEC solutions of (a) 0.5, (b) 0.4, (c) 0.15, and (d) 0.09%. Other conditions, same as in Figure 3.
Figure 3. Comparison of separation of DNA fragments over pH 8.2-13 using a 0.5% HEC solution. Conditions: sodium hydroxide was added to HEC solution to obtain desired pH; others are the same as in Figure 2.
unchanged over a wide pH range (pH 6-10), we believe that the enhanced resolution for DNA fragments is due to an increase in separation efficiency rather than in the selectivity. This could be primarily attributed to the increase in the persistence length of the DNA fragments or of the HEC molecules (through ionization of hydroxy groups) as the pH becomes more alkaline. In either case, the increase in stiffness in the molecules would result in increased transient interactions between them, leading to better resolution. This assumption was strongly supported through the calculation of the number of theoretical plates at different pHs (Table 1). The separation efficiency of DNA fragments significantly increased with increasing pH and over 3 × 106 plates were obtained for the separation of the 271/281-bp DNA fragments at pH 11. However at pH 13, the pH commonly employed in alkaline Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
1671
Figure 5. Dependence of electrophoretic mobility of DNA fragments on pH using an HEC solution of 0.5%. Other conditions, same as in Figure 3.
agarose gel electrophoresis, the resolution of these DNA fragments decreased seriously and only three broad peaks were observed. The reason for this was unknown, but it is possible that it is due to the destruction of epoxy coatings on the capillary surface. The resolution of DNA fragments using ultradilute HEC solutions was also increased with increasing pH. Optimal separations were obtained at pH 11 for all the HEC concentrations used. However, the 271/281-bp fragment was only partially resolved at HEC solutions below the entanglement limit due to the relative low efficiency obtained (9 × 105 plates/m). It should be noted that, unlike highly entangled HEC solutions, the resolution of completely denatured ss-DNA above pH 11.5 in ultradiluted HEC solution was significantly decreased. Figure 4 shows the separation of ss-DNA at pH 12 using HEC solution above and below the entanglement limit. The resolution of ss-DNA was seriously decreased at ultradilute HEC concentrations, especially for small DNA fragments, indicating a different separation mechanism in entangled and diluted polymer solution. Barron et al. suggested that the separation of DNA with ultradilute polymer solutions was due to a transient entanglement coupling between DNA and HEC molecules which increased with the length of the DNA chain.14,15 However, the dynamic and structure of DNA and HEC complex are still unknown.20 We assume that the transient entanglement complex of DNA and HEC is held together via a “linkage” that was enhanced by hydrogen bonding. This assumption is reasonable because both DNA and HEC molecules easily form hydrogen bonds in solution. Above pH 11.5, the hydrogen bond between ss-DNA and HEC becomes weak due to the complete denaturation of DNA. Therefore, decreasing transient entanglement coupling and consequently, decreased resolution of DNA fragments would result. Alternatively, when a 0.5% entangled HEC solution was used, ss-DNA was well separated with high efficiency. This is because the separation of DNA under these entangled conditions is mainly due to size exclusion mechanisms which depend on dynamic pore formation by HEC, although transient entanglement coupling between DNA and HEC may still occur.20 1672 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
Figure 6. Dependence of electrophoretic mobility of 604-bp DNA fragment on pH using HEC solutions with different concentrations. Other conditions, same as in Figure 3.
Since the negative charge of DNA was slightly increased under alkaline conditions due to the ionization of guanine and thymine, an increased electrophoretic mobility should be predicted at higher pH. Figures 5 and 6 show the effect of pH on electrophoretic mobility of DNA fragments in HEC solutions both above and below the entanglement limit. The electrophoretic mobility of the DNA fragments was found to decrease with increasing pH for all HEC concentrations used. Previous studies showed that the electrophoretic mobility of DNA in gel electrophoresis depends on net charge and molecule weight as well as on topology and conformation.31,32 Open circular ds-DNA usually has lower mobility than supercoiled DNA.32 Under alkaline conditions, the DNA conformation may become more open and extended due to the denaturation of local hydrogen-bonded regions, resulting in lower electrophoretic mobility. Furthermore, the extended DNA conformation may also enhance the transient entanglement coupling (32) Mickel, S.; Arena, V.; Bauer, W. Nucleic Acids Res. 1977, 4, 1466-82.
between DNA and HEC molecules, decreasing the overall electrophoretic mobility and increasing resolution as well. Effect of Ionic Strength and pH on Resolution. A dramatic increase in separation efficiency of DNA fragments was obtained under alkaline conditions. Similar results were also reported by McDonell,28 who found greater sharpness of DNA bands in alkaline gels than in neutral gels; however, the results could not be explained at that time. It is well-known that separation efficiency is strongly dependent on the relative buffer concentration in the separation capillary to the buffer concentration in the sample solution. Sample stacking may occur when the ionic strength of the separation buffer is higher than that in the sample solution, resulting in sharper DNA peaks. In this work, electrophoresis buffers with different pHs were prepared by adding sodium hydroxide to the buffer (89 mM TBE containing HEC). Since the ionic strength greatly increased with increasing pH, sample stacking must play an important role in the observed effect. To investigate the effect of sample stacking on resolution, sodium chloride was added to 89 mM TBE buffer to generate the same ionic strength with those buffers under alkaline conditions. Table 1 compares the theoretical plate numbers obtained using buffers of different pH with those buffers at pH 8.2 with the same ionic strength. It can be seen that separation efficiency does increase with ionic strength; however, this is much lower than the increase in efficiency observed with increasing pH. A separation with over 3 × 106 plates/m was obtained at pH 11, while only 9 × 105 plates/m was observed with buffer containing NaCl at pH 8.2, indicating that the significant increase in separation efficiency was mainly due to the alkaline pH. The diffusion coefficient of DNA fragments was measured at different pHs using a “stop migration technique” to investigate the effect of pH on DNA during these separations.33,34 In this technique, two runs were employed to measure the variance contributed by molecular diffusion. The first run was uninterrupted and completed by applying a voltage of 6 kV. In the second run, the separation was interrupted at the halfway point to allow the solute to diffuse for an additional 60 min, and then the voltage was turned on to complete the separation. The difference in peak variance between two runs should be due solely to molecule diffusion which occurs during the period when the separation was interrupted. Therefore, the diffusion coefficient can be easily calculated by the following equation:
D ) (σ22 - σ12)/2∆t
(1)
where σ12 and σ22 are the peak variances in the first and second runs, respectively, and ∆t is the time period before the second run. Table 2 shows the diffusion coefficients of DNA fragments (33) Walbroehl, Y.; Jorgenson, J. W. J. Microcolumn Sep. 1989, 1, 41-5. (34) Yin, H. F.; Kleemib, M. H.; Lux, J. A.; Schomburg, G. J. Microcolumn Sep. 1991, 3, 331-5.
Table 2. pH Dependence of Diffusion Coefficient of DNA Fragments in 0.4% HEC Polymer Solution diffusn coeff (×107 cm2 s-1) DNA fragment (bp)
pH 8.2
pH 10.0
72 118 194 234 310 603 872
2.85 2.70 2.32 1.56 1.13 1.82 3.65
2.43 2.32 1.45 1.16 1.01 1.16 2.62
measured using 0.4% HEC solution at pH 8.2 and pH 10.0. As we expected, the diffusion coefficient of DNA fragments in a given solution decreases with an increase in molecular weight. However, the diffusion coefficient of DNA began to increase when the fragment was larger than 603 bp. The reason for this was not clear, probably due to the relative low efficiency obtained for the larger DNA fragments that make this method less accurate and effective to measure the diffusion coefficient. A more significant increase in the diffusion coefficient of large DNA fragments was observed in an HEC solution of pH 8.2 than that of pH 10, which may due to serious band broadening of large DNA at pH 8.2. On the other hand, the diffusion coefficient of each DNA fragment dramatically decreased under alkaline conditions (Table 2). The diffusion coefficient of a molecular fragment in a given solution can be described as follows:
D ) kT/6πηr
(2)
where k is the Boltzman constant, T is the absolute temperature, η is the solvent viscosity, and r is the hydrodynamic (Stokes) radius of the solute molecule. In this experiment, the ionic strength of the HEC solutions was maintained the same to obtain a similar viscosity in both buffers. Therefore a decrease in diffusion coefficient may indicate an increase in the hydrodynamic radius of DNA fragments, which indicates a more open and extended DNA conformation under alkaline conditions. The extended DNA conformation may allow more efficient transient entanglement coupling between the DNA and HEC molecules, as a result, improving resolution of the DNA fragment further. ACKNOWLEDGMENT This work was supported by the National Institutes of Health (Grant 1R21HG01828-01). We also thank Eric Mayran of PE/ Applied Biosystems for his donation of DNA and restriction enzymes. Received for review December 31, 1998. Accepted March 10, 1999. AC9814492
Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
1673
