Enriching Architectures for Biosensing and Motor-Filament Systems Through the Programmability of DNA

Citation

Guareschi, Matteo Michele (2025) Enriching Architectures for Biosensing and Motor-Filament Systems Through the Programmability of DNA. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/fmhp-r892. https://resolver.caltech.edu/CaltechTHESIS:05052025-211734908

Abstract

Since its inception, the field of DNA nanotechnology has focused on studying the fundamental behaviors and capabilities of engineered nucleic acids. A deep understanding of this toolkit has enabled advancements in several fields, for research tools and in translational applications. Together with its programmability and nanometric resolution, the great promise of DNA nanotechnology lies in the incorporation of structure and function in a single molecule. In this work, we show how these advantages can be leveraged to expand the capabilities of two different systems: a sensor for biomarkers and a motor-filament architecture. During our exploration, we also discover and work to overcome some of the less obvious limitations of the technology, shining light on more foundational questions.

We demonstrate an electrochemical biosensor based on a DNA origami that can detect and quantify nucleic acids and proteins in a package easily adaptable to different analytes by simply replacing the binder molecules. Upon target binding, the structure undergoes a large conformational change, bringing a multitude of redox reporters to the electrode surface where an electric current can be measured. The high number of reporter molecules on a single detector results in improved signal gain per binding event, allowing for the detection of low analyte concentrations, while the conformational change yields an unprecedented gain between the off and on state. We demonstrate how the system can be readily adapted to different analyte molecules and reused over several cycles to analyze multiple samples. We then run simulations of the detector molecule to understand structural deformations intrinsic to this design, in order to optimize the number and placement of the redox reporters. We discover and investigate a phenomenon that causes significant curling of the DNA origami, possibly limiting the contribution of many of the reporter molecules. We explore experimental directions to mitigate the issue by changing the configuration of the redox molecules and by designing stiffer sensors.

We then set out to integrate DNA origami-based nanostructures with an engineered dynein protein that can bind to and kick double-stranded DNA instead of tubulin. Motor-filament architectures have been studied as the main mechanism for cellular transport and as a system that can exhibit mesoscopic active matter behaviors in biology, but the relative difficulty of engineering microtubules has hindered the exploration of their properties. The high-resolution programmability of DNA nanostructures makes them prime candidates to overcome this obstacle and this study has been enabled by the recent development of new protein motors where the tubulin binding domain is replaced by a DNA binding domain. We first look at DNA nanotubes, structures that resemble microtubules, but that retain a level of programmability that is typical of DNA nanotechnology. By exploiting the DNA strand displacement technique, we incorporate machinery that enables new behaviors, with a focus on different ways to turn gliding on and off by stopping the DNA nanotubes.

We then turn our focus to more complex gliders designed with DNA origami. We explore the space of DNA origami polymers in order to assemble superstructures that can be detected under light microscopy, encountering again issues of deformations due to the addition of overhangs. We then assess the gliding capabilities of DNA origami, designing ways to incorporate motor binding sequences on them, but we find that DNA origami sticks nonspecifically to the engineered dynein motors. After testing several different hypotheses, we gather evidence that this interaction might be caused by the large sequence variability of the scaffold strand in DNA origami, coupled with the recognition of spurious binding sequences by the motor proteins.

Thesis Files