DNA NANOBRUSHES FOR ENZYMATIC REACTIONS, PROTEIN RECOGNITION, AND BIOMARKER DETECTION VIA ATOMIC FORCE MICROSCOPY

DNA nanotechnology is an emerging field that provides tools to design and create programmable, spatially–resolved, functional nano-scale devices with potential application ranging from synthetic biology to theranostics. The first part of this thesis addresses the fundamental question of whether it is possible to study the function and the recognition of proteins in an in-vitro milieu that is similar to intracellular conditions, in terms of crowding, confinement and compartmentalization, without the use of crowding agents and obtain information that are triggered by localized and nanoscale crowding. To this end, we design and construct surface-bound functional, sequence-dependent DNA nanoreactors with distinct inherent heterogeneity and homogeneity in their molecular structure, using two approaches: an atomic force nanolithography (AFM)-based technique termed nanografting and the spontaneous formation of self-assembled monolayers of pre-hybridized double stranded (ds)DNA molecules. These nanoreactors then serve as test-bed for understanding the functions and interactions of DNA with protein. With such a tool, we studied the recognition and the specificity of a highly stringent, site-specific endonuclease called BamHI, an enzyme with high specificity for its cognate site and well-defined crystallography structure that is routinely used as protein model and for cloning engineering. Our results show an unprecedented digital behaviour of BamHI triggered by nanoscale crowding within highly dense non-cognate spatially resolved DNA nanostructure. The second part of this thesis aims at developing a computation-based strategy to optimise cyclic peptides with inherent affinity and selectivity that mimic those of a single loop of an antibody (µM range affinity). Using a combination of molecular dynamics simulations and Monte Carlo algorithm (tagged algorithm 1), our theoretical collaborators have improved the binding affinities of peptides designed to recognize Beta-2-microglobulin (β2m), a biomarker over-expressed in ovarian cancer type 1 (OVA 1). The affinity of the designed peptide to the target of interest is further characterized by developing a label-free experimental validation, which utilizes DNA-directed immobilization (DDI) combined with nanografting to form nanopatches of spatially-oriented DNA-cyclic peptide able to recognize the solvent-exposed binding site on β2m. Using AFM height measurement, we obtained a micro-molar range binding affinity towards β2m. Subsequently, our theoretical collaborators, Dr Miguel Soler and Dr Sara Fortuna, employed a new computer-based protocol, which allows the identification of possible binding sites on 2m (tagged algorithm 2). Following the generation of peptides, they screen the peptides based on their predicted binding energies, distance between 2m and peptides, and the hydrophobic surface areas. With such computation strategies, pep331 was selected for experimental validation. Afterwards, we selected one peptide (pep381) from the pool of peptides designed by our collaborators using algorithm 1, and the pep331 generated using algorithm 2, and we set up an AFM based experimental validation, to determine the sensitivity of both peptides. Our results show that both peptides bind 2m with sensitivity of ~7 µM. These results give us a benchmark preliminary data to design surface-bound synthetic bidentate peptides with possible application in biomarker detection and discovery.