Agarose gel electrophoresis of DNA: A Brownian dynamics simulation

Agarose gel electrophoresis of DNA was investigated with a Brownian dynamics simulation. The terms in the Langevin equation, the equation of motion for the DNA chain, were selected to represent DNA and agarose as realistically as possible. The DNA was modeled as a bead-spring chain, and the agarose was modeled as a three dimensional array of regularly spaced cylindrical obstacles. Some simulations with randomly spaced and oriented obstacles were also completed. The simulation was able to duplicate much of the relevant experimental phenomena. The time averaged velocity in the simulation is a strongly decreasing function of chain length for shorter chains, but this dependence of velocity on chain length disappears for longer chain lengths. Investigation of the dynamic behavior of individual chains reveals that the motion of chains in the gel is characterized by a cycling of the chain between bunched and extended conformations. Moreover, this cycling occurs with a characteristic period which is dependent on the chain's contour length.; The response of chains to a sudden 120 degree change in field direction was investigated. Enhanced separation was predicted to occur in these pulsed field simulations. The increased resolution was due to a forced entanglement after field switching. The disentanglement is characterized by a disentanglement time which varies linearly with chain length. The disentanglement time and the conformational cycling period were found to be closely related with the disentanglement time being approximately half the conformational cycling period. This relationship is due to the similarity between the entanglements which occur during conformational cycling and the entanglements which occur when the electric field direction is changed.; The response of chains to a 180 degree change in field direction (field inversion) was investigated but, in contrast to the 120 degree case, no enhanced resolution was observed. The chains did not become entangled after field inversion in these simulations. These field inversion simulations were conducted in regular arrays of obstacles and it is believed that randomly oriented obstacles may be necessary to achieve enhanced resolution.