Anal. Chem. 1999, 71, 2385-2389
High-Efficiency DNA Separation by Capillary Electrophoresis in a Polymer Solution with Ultralow Viscosity Futian Han,†,‡ Bryan H. Huynh,† Yinfa Ma,*,† and Bingcheng Lin‡
Truman State University, 100 East Normal Street, Kirksville, Missouri 63501, and Dalian Institute of Chemical Physics, Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023, P.R. China
The viscosities of some polymer solutions for DNA separation in capillary electrophoresis are generally very high, which makes them hard to pump into the capillaries. We have developed a novel sieving buffer, based on lowmolecular-weight hydroxypropylmethylcellulose, to separate DNA fragments. The viscosity of this sieving matrix was at least 1 order of magnitude lower than that of traditional buffers with similar sieving effect. The influence of additives such as urea and mannitol was investigated. It was found that the double-stranded DNA (ds DNA) fragments began to denature in 3.5 M urea, and 7 M urea can denature the ds DNA completely. The presence of mannitol will decrease the overlap threshold of the polymer solution (the concentration at which the polymer molecules begin to entangle with each other), which makes it possible to separate DNA fragments in a polymer solution of relatively low concentration. The influence of the electrical field was also investigated, and it was found that the mobility of DNA fragments up to 2000 bp in length did not change greatly with different electric fields. This phenomenon implies that the DNA fragments at this range do not change their conformation with the increase of electric field as was previously believed. The possible mechanism for the separation of DNA fragments is also discussed. The charge densities of DNA fragments are independent of molecular size. Therefore, the mobilities of DNA fragments of different length are almost constant, which makes it impossible to separate DNA fragments in free zone capillary electrophoresis. Although some investigators have explored using free-solution electrophoresis to separate molecule-tagged DNA fragments,1 molecular sieving must be applied to DNA separation by capillary electrophoresis in most cases. Originally, a capillary electrophoresis system with various gels (capillary gel electrophoresis, CGE) was used for the separation of biopolymers such as oligonucleotides, nucleic acids, and proteins. However, an easy-to-operate and reproducible CGE system is difficult to obtain and the cross* Corresponding author: (phone) 660-785-4084; (fax) 660-785-4045; (e-mail)
[email protected]. † Truman State University. ‡ Dalian Institute of Chemical Physics. (1) Heller, C.; Slater, G. W.; Mayer, P.; Dovichi N.; Pinto, D.; Viovy, J.; Drouin, G. J. Chromatogr., A 1998, 806, 113-121. 10.1021/ac990160x CCC: $18.00 Published on Web 05/13/1999
© 1999 American Chemical Society
linked gel-filled capillaries do not have a long lifetime. In recent years, entangled polymer solutions, such as poly(ethylene oxide),2 poly(ethylene glycol), poly(vinyl alcohol),3 glucomannan,4 uncross-linked polyacrylamide,5 and cellulose and its derivatives,6,7,8 were more and more frequently employed as media to achieve molecular sieving of DNA fragments. Several models, such as the Ogston model,9 the reptation model,10 and the biased reptation model,11 were employed to explain the behavior of DNA fragments in these kinds of sieving matrixes. But there still does not exist a complete understanding of electrophoretic migration of DNA fragments in polymer solutions. Therefore, it is hard to predict the resolution of the migration and to design a sieving system that satisfies one’s separation requirements. Generally, polymer solutions with a wide range of concentrations were used for DNA separation. The sieving ability varies greatly according to the polymer concentration and chemical components. For sieving matrixes prepared from cellulose and its derivatives, concentrated solutions are generally required in order to get satisfactory separation of smaller DNA fragments ( 0.9990). The relationship can be described by
1/t ) K1E + K2
(2)
where t is the migration time of the DNA fragments, K1 is a constant related to the size of a DNA fragment, K2 is another constant related to the system (temperature, capillary inner diameter, sieving buffer, etc.), and E is the electric field. The mobility of a DNA fragment is defined by
µ ) L/(tE)
(3)
where µ is the mobility and L is the effective length of the capillary. From eqs 2 and 3, we can get
µ ) K1L + K2L/E ) µ0 + K2L/E
(4)
where µ0 was defined as the optimized mobility of a DNA fragment in this buffer, which is the maximum mobility that a DNA fragment can reach because K2 is a negative value of all cases. µ0 can be determined experimentally by plotting the reciprocal migration time as a function of electric field and multiplying the slope by the effective length of the capillary. Differentiation of the above leads to
dµ/dE ) -K2L/E2
(5)
Equation 5 indicates that the mobility variation with the change of electric field is inversely proportional to the square of the electric field. From the experimental data, we can see that dµ/ dE is a very tiny (almost zero) value, which means that the influence of the electric field on mobility is so slight that it can be neglected. As a matter of fact, the second term of eq 4 is at least 2 orders of magnitude less than the first term of eq 4 in all of this study. The above description states that the mobility of the DNA fragments does not change greatly under different electric fields ranging from 80 to 480 V/cm. We also tested a larger fragment from the low-range DNA sample and our calculations show that the mobilities of the DNA fragments up to 2000 bp in length also do not change greatly while the electric field changes. We even obtained the same results from the 2% HPMC-50 undenatured buffer (inset of Figure 6). Therefore, this phenomenon is not the result of using the denaturing buffer, which can eliminate the influence of the secondary or higher order structure of the DNA fragments. This result is not consistent with the biased reptation model,11 which suggests that the DNA molecule may change its conformation when the electrical field increases, because the change in the conformation will lead to great change in the mobility. On the basis of this phenomenon, we suggest that the electric field does not influence the resolution of the DNA fragments as much as what is thought. Figure 7 shows the resolution of the 80/89,184/192, and 540/587 fragment pairs under different electric fields. Resolution was calculated using a traditional chromatographic method,17 and the result shows that there is no significant change in the resolution of 80/89, 184/192, and 540/587 fragment pairs when the electric fields changed from 80 to 320 V/cm. The resolution of these fragment pairs decreased slightly when the electric fields changed from 320 to 480 V/cm, probably due to band-broadening resulted from the Joule heating effect. Therefore, a higher electric field should be used to get faster separation until Joule heating becomes obvious. CONCLUSION The mannitol-modified low-molecular-size HPMC buffer has very low viscosity, which makes it easy for capillary filling, flushing, and refilling. This is a great advantage over those buffers (17) Snyder, L. R.; Kirkland, J. J. Introducation to Modern Liquid Chromatography; John Wiley & Sons: New York, 1979; Chapter 2.
Figure 7. Resolutions of the 80/89, 184/192, and184/192 fragment pairs under different electrical fields. The line was drawn through the data to guide the eye.
with high viscosities. Mannitol can enhance the separation by interacting with HPMC molecules and decreasing the threshold of the polymer solution. Satisfactory resolution of DNA fragments can be achieved in the solution with a concentration even lower than the overlap threshold concentration. A theory used to interpret the DNA separation mechanism in a ultradilute polymer solution is found to fit the data presented in this paper for the low-molecular-weight solution. A concentration of 7 M urea is necessary for the complete denaturing of the ds DNA fragments. The electric field does not have much influence on the mobility of the DNA fragments, which implies that the DNA fragments up to 2000 bp do not change their conformation with a change of electric field. ACKNOWLEDGMENT The authors thank Dr. Kenneth Martin and Dr. Kenneth Fountain of Truman State University for their help on the manuscript. This work was supported by an Internal Faculty Research Grant from Truman State University awarded to Y.M.. We also appreciate the support from the National Natural Science Foundation of China. Received for review February 10, 1999. Accepted April 6, 1999. AC990160X
Analytical Chemistry, Vol. 71, No. 13, July 1, 1999
2389
