Engineering Probes and Circuits for Detecting Single Nucleotide Variations in DNA and RNA

Abstract

Even single nucleotide changes in nucleic acid sequence can result in drastic phenotype differences with significant health impact, such as driver mutations for cancer or antibiotics resistance in pathogens. However, reliable detection and quantitation of single nucleotide variants (SNVs) is challenging when the SNV is at low allele-frequency, due to the physical and chemical similarity of the SNV molecules and their corresponding wildtypes. In the past 30 years, researchers typically empirically optimize assay conditions to temperatures and buffer conditions that are conducive to high specificity hybridization of nucleic acids, but such a process is time-consuming, sensitive to small changes in protocol, and ineffective in multiplexed settings where multiple target SNVs need to be simultaneously analyzed. This thesis presents a rational design approach to SNV detection and quantitation to overcome the challenges of conventional analysis, by creating probes and circuits that are capable of robustly detecting SNVs at low allele frequencies. These probes and circuits are designed based on the sequence information of target SNVs, and do not require significant empirical optimization for function. Consequently, this approach is amenable to integrating with existing nucleic acid bioanalytic technologies, and I have started working with collab- orators at Thermo Fisher (Applied Biosystems) to translate these and other technologies to commercial use. Four distinct projects are presented herein. The first describes “toehold exchange” probes for single-stranded DNA and RNA targets that rely on molecular competition by a “protector” oligonucleotide to ensure hybridization specificity; this work enables detection of SNVs across a wide range of temperatures and buffers. The second describes probes for detecting SNVs in double-stranded DNA targets; by utilizing the SNVs present in both the forward and reverse strands of DNA, we are able to significantly improve SNV detection. The third describes construction of reaction networks that use a system’s potential energy to circumvent theoretical limitations to hybridization specificity at pre-equilibrium times. The final work describes the use of multiple circuits to implement a linear combination analysis of the concentrations of two distinct RNA species that differ by a single nucleotide. The recurrent theme of this thesis is that a systematic, engineering-based approach to designing nucleic acid molecules can enable a set of robust behaviors and interactions that is valuable to the field of molecular biology. My work thus far have focused on the single nucleotide variant detection problem, but in principle