Studies of DNA dynamics in slit-like nanochannel confinement

Abstract

The ability to visually observe single DNA molecules has greatly improved our understanding of polymer physics, from gel electrophoresis to the theology of dilute (and even concentrated) polymer solutions. The use of DNA in these general studies, though, resulted in a depth of specific knowledge concerning a particular polymer of major interest in biology. Researchers have taken advantage of this wealth of knowledge to develop new, faster, cheaper, and more direct methods of extracting the information, at a coarse level, embedded in the sequence of basepairs along the DNA backbone. Further development, though, is now limited by the ability to control and manipulate the position and conformation of single DNA molecules. It was recognized long ago that confinement of polymer molecules in geometries with dimensions on the order of the polymer size would greatly affect the physical behavior of that polymer. These physical changes were later hypothesized to be of use to control single molecules of DNA. However, until recently, the confinement theories and their use stood untested due to a lack of techniques to reliably and controllably construct micro- (and nano-) devices with such small feature sizes. It is the focus of this thesis to investigate these confinement effects in an ideal, nanofabricated geometry and their use in the manipulation and control of single DNA molecules. In this thesis, we present a series of single-molecule visualization studies aimed at elucidating polymer behavior in confinement and methods of possible use in the manipulation and control of the polymer conformation. In particular, confinement in a slit was shown, both experimentally and through scaling analysis, to diminish long length scale polymer-induced solvent flow sufficiently enough to render those effects negligible in the behavior of the confined molecule. We also demonstrate that confinement also alters the diffusion and relaxation time of the DNA, and we compare their dependence on channel height and molecular weight to existing theories.

(cont.) De Gennes' blob theory is found to describe the molecular weight scalings quite well, but predictions of the scalings with channel height are plagued by an oversimplified description of short length scale polymer-solvent interaction used in the theory. Thus, empirical knowledge is needed to adequately predict the scaling of DNA transport coefficients in confinement. We also investigate aspects of polymer deformation in confinement. We observe, for the first time, two slow modes of polymer relaxation. The two modes are found to govern polymer behavior based on the polymer's extension, a phenomenon unique to confinement in polymer physics. A simple, physical model is developed to explain the origin of the two governing time scales, to explain their scaling with channel height and molecular weight, and to predict the extension at which the exchange between the two timescales occurs. We also examine the effects of these two characteristic time constants on the steady-state stretch of molecules in confinement. We find that the second-longest relaxation time determines the deformation rate needed to unravel the coil, unlike bulk polymer deformation. Interestingly, details of this unraveling change significantly in confinement, highlighting the need for further work in this area. In larger channels, we demonstrate that microfabrication techniques in the form of an obstacle array with dimensions smaller than the polymer size can aid polymer stretching. While a polymer will often fold or kink during stretching, we find the use of a collision event to "precondition" the polymer conformation for stretching makes these folds and kinks less likely, and therefore, stretching occur more rapidly. The efficiency of the device depends strongly on the probability of a collision event, and results from single molecule/single post experiments are used to demonstrate the capability of a second-generation device. The impact of this thesis will be two-fold. Our fundamental results have and will continue to serve as a basis of comparison and a springboard for more complicated studies of confined polymer physics.

(cont.) These studies provide detailed information on DNA transport coefficients in geometries widely utilized in microfabricated devices. We also directly display the effects of confinement on DNA manipulation. Non-equilibrium polymer dynamics are found to be highly nontrivial, exemplified by the importance of a new timescale of polymer motion. Importantly, it is this new timescale that is of concern for applications such as gene mapping where large scale polymer deformation is required. Lastly, we demonstrate the success of a unit-operation-like approach to the design of polymer manipulation devices.