The Importance of Structural Rigidity

In the recent work of Yang and co-workers,⁴ not only the size and geometry of the phiX174 bacteriophage was synthetically recreated (Figure 1a, b), their DNA nano construct was subsequently loaded with the original phiX174 genome, and surprisingly showed retained infective potential in Escherichia coli (E. coli) bacterial cells. While literature on DNA-nanomaterial uptake in mammalian cells is abundant,^{5, 6} exploring bacterial delivery has been lagging. The authors employed a wireframe design approach,⁷ which allows for a relatively facile engineering of elaborate geometric architectures. However, wireframe structures are notoriously flexible, which negatively impacts a robust capsid structure. Furthermore, nanoscale rigidity is being linked to enhanced selectivity of interactions,⁸ a developing direction within DNA nanotechnology research and applications.

To overcome flexibility hurdles, the icosahedral edges were designed on a four-helix square lattice, where two parallel helices form a foundation layer (FL, Figure 1c) and 2 parallel helices on an additional layer (AL, Figure 1c) rigidify each edge. Interestingly, experiments with more flexible architectures, lacking 10 edges, showed no bacterial infection, suggesting structural rigidity is not only key for spatially-defined functionalization but also crucial for cellular uptake, while changes in structural integrity cannot be excluded but were not reported.

The Capsid as Enthalpy Trap

As functional cargo, the original phiX174 bacteriophage genome was loaded into the DNA capsid. In the natural protein-based capsid, the long viral single-stranded DNA (ssDNA) genome is tightly packed into a tiny space during the last step of the viral life cycle. To neutralize charges, the genome coordinates with 60 copies of a positively charged protein inside the capsid. When loading the cargo into a DNA-based capsid, the repulsion of negative charges inherent to DNA causes an additional challenge. The choice of an open wireframe structure is not only interesting from a material consumption and molecular accessibility design point of view, it also limits the density of repulsive charges toward the cargo.

Specific packaging was achieved through complementary base-pairing with anchoring strands inside the capsid, in a balancing game of entropy versus enthalpy (Figure 2a). In order to force a long, flexible ssDNA molecule (high in entropy) to enter and stay inside (low entropy) the small, repulsive cavity of the DNA origami icosahedral capsid, 30 ssDNA edge-protruding handles were designed to function as directional affinity trap with the complementary region on the plasmid. Using fewer handles (e.g. 18 or 6) did not fully entrap the genome, losing the entropy battle (Figure 2b). For the entrapment to be sufficiently dynamic to allow for release, enzymatic amplification, but also for being sufficiently stable to guarantee the continued residence of the plasmid, a trade-off of 20 nucleotides (nt) length was selected as the perfect enthalpy trap, after 10 nt was found to be too dynamic and 30 nt too strong (Figure 2c). Elegantly, these handles had a dual function as primers for rolling circle amplification (RCA) of the plasmid cargo. The careful screening of affinity (length) versus valency (number) of handles to entrap the phage genome allows to understand the kinetic limits of subsequent functional assays, and to design the most adequate artificial capsid to balance cargo entrapment, uptake, and release.

Regarding uptake, the yield of this passive infection in non-competent cells was orders of magnitude weaker than the infectivity of a natural phage. Roughly, a million artificial “phages” were needed for infection, while for phiX174 almost every virus is assumed to cause infection. This stark difference suggests a diverging internalization path, which opens new potential for artificial delivery routes into bacteria, and competition with natural phage infection. Interestingly, the uptake efficiency seemed irrelevant to various types of oligolysine-PEG surface coating or addition of targeting molecules on the exterior of the artificial capsid. As the bacterial culture milieu differs from that of mammalian cells, likely other coating strategies should be investigated.

Targeting Bacteria

This study is standing at the beginning of a branch in DNA nanotechnology where applications in bacterial-centered biotechnology are found. Being able to engineer a central step in the viral life-cycle, without a priori presenting a viral particle, is innovative. While the fundamental mechanisms of how infectivity progresses are currently unknown, its functional demonstration opens several new opportunities. With bacteria being mass producers of both recombinant proteins and circular ssDNA, which served as the scaffold for DNA origami, a future development could envision the excretion of the origami-own DNA. This could be either loaded as cargo, or perhaps be a tool for the “undressing ”of the staples to release the scaffold constitutively present in the DNA origami particle, and liberate it as a template for reproduction. To control or reduce infectivity, a split plasmid approach, as recently discussed by Behler et al.,⁹ could be incorporated. Lastly, with bacterial delivery being a fresh target of DNA nanotechnology, many fundamental parameters including size, stability and geometry of the DNA-based architecture, which likely affect uptake yield and kinetics, are left to be explored. The doors toward bacterial targeting have been opened, and an exciting new road awaits.

Conflict of interest

The authors declare no conflict of interest.