Technologies for high throughput single molecule DNA sequencing

Next-generation DNA sequencing is rapidly accelerating biological research by permitting the inexpensive and routine analysis of genomes, transcriptomes, and interactomes. Commercial instruments that sequence single DNA molecules are now capable of generating 20--30 gigabases of sequence data per run, but technological advances are required to further reduce costs, improve error rates, and increase throughput. This dissertation focuses on developing the underlying technologies to address these needs.Single molecule sequencing-by-synthesis approaches employ a mesophilic DNA polymerase to sequentially incorporate fluorescently-labeled nucleotides into a surface-tethered primer-template. One major bottleneck is the time required to image thousands of fields of view after each nucleotide incorporation cycle. To maximize the amount of data generated it is therefore critical to pack as many resolvable templates on the surface as possible. Random deposition can at best achieve a density of &ap2 resolvable templates per square micron, so two new simple and scalable approaches were developed using nanoparticle arrays and colloidal epitaxy to pattern surfaces at up to 6-fold higher densities.A comprehensive understanding of how DNA polymerases behave under different conditions is also critical to optimize read length, coverage, and error rate. Single molecule measurements were used to make a detailed characterization of DNA replication as a function of the template's secondary structure and the sequence context. These data enable the measurement the intrinsic "speed limit" of DNA polymerase for the first time by separating the burst synthesis rate from sequence-dependent pausing.Finally, the ability to use a thermophilic polymerase for single molecule sequencing would offer a number of key advantages: improved enzyme heat stability, better ability to incorporate nucleotide analogs, and the capacity to melt templates that are GC-rich or have a high degree of secondary structure. To achieve this, colloidal lenses were used to overcome the temperature limits of oil-immersion microscope objectives by incorporating a focusing element in immediate proximity to an emitting fluorophore. The optical system was completed by a low numerical aperture optic which can have a long working distance and low light collection ability. As proof of principle, colloidal lenses were used to measure real-time single molecule mesophilic and thermophilic DNA polymerase kinetics at 23°C and 70°C using a 20X 0.5 NA air objective.