The relative contributions of dispersion and diffusion to band spreading (resolution) in gel electrophoresis

DNA of approximately 2 kbp in length was previously found not to diffuse significantly in 1-1.5% agarose gels in the absence of an electric field, but to disperse during electrophoresis (Yarmola, E., Chrambach, A., Electrophoresis 1995, 16, 345-349). Accordingly, a process distinct from diffusion, and responsible for band spreading with migration time in gel electrophoresis, was defined as dispersion. Correspondingly, the diffusion coefficient, D(diff), was distinguished from a dispersion coefficient, D(disp). For DNA of approximately 1, 2 and 3 kbp, D(diff) and D(disp) were measured in agarose gel electrophoresis (1.0% SeaKem GTG). In that order of DNA length, D(disp)/D(diff) was found to increase from 5 to 15 to 45, showing that with increasing DNA length, time-dependent band spreading, and thus resolution in gel electrophoresis, is governed predominantly by dispersion, not diffusion. It is assumed that the essential part of electrophoretic dispersion is due to entanglement of the DNA molecule in the gel. Indirect evidence for such an entanglement derives from the observation of peak asymmetry and its interpretation by the Giddings-Weiss model.