Atypical Sieving of Open Circular DNA during Pulsed Field Agarose Gel Electrophoresis

Pulsed field agarose gel (PFG) electrophoresis, originally used to improve the resolution by length of linear DNA [Cantor et al.(1988) Annu. Rev. Biophys. Biophys. Chem. 77, 287-304], is found here to cause atypical sieving of 48.5-97.0-kb open circular DNA. Two procedures of PFG electrophoresis are used: rotating gel electrophoresis with rotation of 2π radians [2πRGE; Serwer, P., & Hayes, S. J. (1989) Appl. Theor. Electrophor. (in press)] and field inversion gel electrophoresis [FIGE; Carle, G. F., Frank, M., & Olson, M. V.(1986) Science 232, 65-68]. During 2πRGE at 6 V/cm, the electrophoretic mobility (μ) of 48.5-kb open circular DNA increases in magnitude as agarose percentage (A) increases from 0.4 to 1.5. The sieving revealed by this μ vs A relationship is highly atypical (possibly unique) for any particle. The extent of this atypical sieving increases as electrical potential gradient, DNA length, and pulse time increase. In some cases a maximum is observed in a plot of μ's magnitude vs A. The μ of open circular λ DNA is smaller in magnitude than the μ of equally long linear μ DNA. A typical sieving has also been observed by use of FIGE. As pulse times used during FIGE decrease below those achievable by 2πRGE, the progressive loss of circular DNA's atypical sieving is accompanied by both a dramatic increase in μ's magnitude at the lower A values and a decrease in μ's magnitude at the higher A values. At the lower A values, open circular DNA sometimes migrates more rapidly than linear DNA of the same length. The sieving of a solid sphere (intact bacteriophage T3) is not modified by use of FIGE; intact T3 has been used as a normalization standard for μ. The atypical sieving observed here is explained by the following hypothesis: (1) Projections from the gel thread, and therefore arrest, open circular DNA during agarose gel electrophoresis. (2) This arrest is relieved by changing the direction of the electrical potential gradient. (3) The gel's threading capability increases as A decreases. Application of this hypothesis to the sieving of linear DNA explains previously found inconsistencies of theory and experiment.