6.4 Conclusions

Although force is not the natural control parameter in the minitweezers, we can perform unzipping experiments of DNA at controlled force by using a force feedback. The bandwidth of the force feedback is wide enough to keep the force constant while the DNA molecule undergoes transitions between metastable states. The force feedback, though, cannot follow the fluctuations of the force induced by the thermal motion.

The unzipping FDC$ _f$ differs significantly from the FDC$ _x$. The FDC$ _f$ is a step-like monotonically increasing function and does not exhibit the typical sawtooth pattern of the FDC$ _x$. Even at a very low loading rate, the unzipping and rezipping FDC$ _f$ do not overlap each other. The unzipping/rezipping cycles are not quasistatic at the timescale of the lab and they always exhibit hysteresis. The hysteresis has a minimal value and it increases with the loading rate. The minimal value corresponds to the slowest loading rate that is experimentally feasible in the instrument.

The irreversibility of the pulling experiments at controlled force is due to the height of the energetic barriers of the energy landscape. In a pulling experiment, the system does not have enough time to overpass such energetic barriers, which means that the entire distribution of energy states is not correctly sampled and full equilibrium cannot be reached. The number of metastable states detected in experiments is lower (and the barriers are larger) at controlled force as compared to controlled position. Besides, at controlled force the metastable states observed during unzipping are different from the ones observed during rezipping.

The unzipping at controlled force starts with small and frequent transitions between metastable states. As the force increases and the base-pairs are being disrupted, the transitions become larger and less frequent. This effect can be understood by looking at the free energy landscape. The progressive tilt of the free energy explains these transitions. The rezipping also shows this behavior.

Although the molecule is far from being an extensive thermodynamic system and the experiments are not quasistatic, the measurements verify the prediction of the number of open base-pairs vs. the force exerted.

JM Huguet 2014-02-12