Control mechanisms over the RNA-guided self-assembly of tobacco mosaic virus-based nanotemplates

Abstract

Both size and arrangement of well-defined regions of different surface chemistries on individual nanoparticles are limited by their respective fabrication processes. Tobacco mosaic virus (TMV), a plant virus that is nonpathogenic to humans, can be produced inexpensively and in large quantities in a greenhouse. Its self-assembly from single-stranded ribonucleic acid (RNA) of 6395 nt with 2130 identical copies of its coat protein (CP) makes TMV an outstanding and versatile tool for bionanotechnology. The formation of TMV-like particles (TLPs) can be easily induced in vitro – even from synthetic RNAs containing the TMV origin of assembly (OAs) and with genetically modified TMV CP types, of which, moreover, different variants can also be incorporated together into single particles. To be of interest for a wide range of applications, such as imaging and diagnostic tools, templates for microarrays, electronic devices, and biosensors, a large number and variety of possible binding types between CPs and functional molecules, as well as well-defined coupling sites, are advantageous. In this work, new strategies were developed to generate spatially well-defined, selectively addressable CP domains on a single TMV-like nanotube. To this end, dynamic DNA nanotechnology was applied to drive RNA-mediated self-assembly of TMV CP. The established "stop-and-go technique" uses DNA oligonucleotides as "stoppers" that hybridize sequence-specifically at predefined sites in the TMV RNA, blocking its packaging into TMV CP at the hybrid duplex position. A non-binding overhang ("toehold") allows subsequent restart of assembly using a fully reverse-complementary "release" DNA oligonucleotide (as "fuel"), so that an adjacent CP domain with different binding sites can grow after displacing the stopper strand and replacement of the first CP variant by a second one. With this method, for the first time, differently substructured TMV CP exposing three-domain TLPs were investigated for spatially defined assembly of biomolecules on the nano- to mesoscale, with a view to applications as artificial multienzyme complexes or in multiplex biosensors. Additionally, software-based simulations were performed and pave the way towards precise prediction of accessible RNA sequences for binding DNA blocking elements.