Unlocking precision medicine using DNA nanotechnology

Abstract

Significant advances in bioinformatics have led to major breakthroughs in our understanding and treatment of many complex diseases. Molecular biomarkers from across the -ome spectrum (genomics, proteomics, metabolomics, etc.) can now be used to help inform disease diagnosis, monitoring and medication. As a direct result of this, healthcare systems have begun the transition to ‘precision medicine’, where patients are provided with personalized treatments selected for maximum efficacy against their unique biomarker profile. However, precision medicine can only be achieved if the methods for detecting and quantifying biomarkers are inexpensive and efficient enough to be applied at scale in clinical settings rather than in limited-throughput laboratories, as is currently the case.

DNA nanotechnology is a potential solution to improving patient access to reliable and inexpensive biomarker monitoring. The flexibility and utility of DNA-based assays have already been demonstrated with various readouts (including electrochemistry, pH, fluorescence, etc.) and affinity reagents (aptamers, anti- bodies, etc.). In this thesis, I propose and experimentally validate techniques for employing DNA nanotechnology for precision-medicine oriented diagnostics.

First, I present a methodology for multiplexed point-of-care biomarker detection. The assay is based on variable-length DNA payload chains, which are systematically disassembled in the presence of specific biomolecular targets, leading to fragments of different sizes that yield characteristic band patterns in gel electrophoresis. I validate the principle with the detection of various nucleic acids and steroids, both individually and in a multiplexed assay. Furthermore, I demonstrate how automated hardware (using capillary electrophoresis) and/or intelligent software (based on machine learning) could be applied to streamline this assay, making it highly amenable for integration in healthcare settings.

I also propose a new paradigm for highly specific and sensitive detection of complex biomarkers such as proteins. Using an ‘affinity zone’, a 3D space where ligands are accurately positioned using DNA origami structures, multivalent interactions can be tuned to specifically bind to a single target, akin to aptamers or antibodies. I present the design and experimental validation of a DNA origami framework that can precisely tune the position of ligands within an affinity zone. If combined with molecular dynamics simulations, the concept could become the ideal testbed for multivalency, potentially paving the way to new computationally-designed affinity reagents.