The strength claims in this dossier rest on a two-step idea: fold DNA into a designed shape, then coat it in silica so the glass inherits the geometry (see DNA Origami Templating). This paper, an arXiv preprint (2204.07385) by Martina F. Ober, Anna Baptist, Lea Wassermann, Amelie Heuer-Jungemann and Bert Nickel of Ludwig-Maximilians-Universität München and the Max Planck Institute of Biochemistry, asks two questions about that second step. Does the DNA keep its shape while silica forms? And does the silica just coat the outside, or does it also fill the inside? The authors cite work from Oleg Gang's group but do not report on it; their paper studies single origami objects in solution, not that group's three-dimensional nanolattices.
Why a new measurement was needed
The authors point out that earlier evidence of silica growth on DNA origami came mostly from dried samples examined by electron microscopy, atomic force microscopy and X-ray spectroscopy. "Shell" thicknesses were inferred from images. None of those methods shows what happens inside a helix bundle while the reaction is under way in liquid. Small-angle X-ray scattering (SAXS) can do this. It records how a sample in solution scatters X-rays at small angles, which reveals both the overall shape of an object and repeating spacings within it, such as the distance between neighbouring DNA double helices.
What they did
Two origami designs, both built on a honeycomb arrangement of helices, were silicified with the sol-gel chemistry the paper describes as a modified Stöber reaction. First the positively charged precursor TMAPS binds to the negatively charged DNA backbone. Then TEOS is added and co-condenses onto it to build silica. The designs were a cylindrical 24-helix bundle (24HB) and a flat, cuboid four-layer brick (4-LB). Most measurements used an in-house molybdenum X-ray source. The paper says this delivers less radiation dose than copper radiation of the same intensity, so runs of up to 80 hours were possible. A purpose-built cell tumbled the sample to stop the particles settling. Heated samples were measured at the Austrian SAXS beamline at the ELETTRA synchrotron.
What was measured
Shrinkage. Before any reaction, the 24HB fitted a cylinder of radius 80.1 ± 0.2 Å with an interhelical spacing of 26.2 ± 0.3 Å. With TMAPS alone, after eight hours, the radius had fallen to 73.4 ± 0.4 Å and the spacing to 25.2 ± 0.3 Å. When TEOS was added straight after TMAPS, the shrinkage came faster. The radius reached 74.2 ± 0.5 Å within the first four hours, and the spacing fell to 23.8 ± 0.2 Å by eight hours. The abstract sums this up as "an almost 10 % size reduction." The flat brick showed the same effect. Its thickness fell from 89.9 ± 0.4 Å to a minimum of 80.3 ± 1.3 Å at 56 hours.
Silica inside the structure. The scattering peak produced by the regular helix spacing vanished soon after the reaction began and then came back stronger than before. The authors read this as contrast matching. DNA, water and silica scatter X-rays to different degrees, so the peak disappears when the material between the helices scatters about as strongly as the DNA itself. The paper calculates that this happens once roughly 40 % of the water inside the origami has been replaced by silica. When the peak returned, more silica had filled in. On this basis the authors conclude that silica forms between the helices and does not only coat the outside.
A very thin outer shell. By the end of the reaction the 24HB radius was 80.4 ± 0.1 Å, almost the starting value, while the helix spacing stayed condensed at 24.7 ± 0.05 Å. The authors treat the gap between the final radius and the most condensed radius, 6.2 ± 0.3 Å, as an upper limit on the outer coating. That is under one nanometre.
Reaction time. A model-free measure of total scattering, the Porod invariant, levelled off after about 24 hours for the 24HB. Earlier reports cited by the paper describe reactions lasting up to a week. The authors suggest the gentle tumbling at room temperature may have sped things up.
Heat stability. Bare 24HBs fully melt at 60 °C. A 24HB taken at its most condensed state (radius 74.5 ± 0.4 Å) and held at 60 °C for 30 minutes stayed intact by SAXS and TEM. The authors credit this mainly to the silica inside the structure rather than to an outer shell.
The flat-surface problem
The bricks behaved differently. They clumped into aggregates visible to the naked eye as clouds in solution, and SAXS showed the low-angle intensity rise that signals aggregation. Their scattering invariant levelled off after only about 4 hours and rose only about half as much as for the cylinders. The helix peak vanished but never recovered. The authors propose that the same entropic depletion forces that compress the origami also push flat faces together. Curved cylinders have too little contact area for this to happen. They also suggest that aggregation may block further silica from getting inside. For future protocols they float two ideas: bulky water-soluble silanes that would bind only to outer surfaces, and smaller primary silica particles followed by stepwise TEOS additions.
What this does and does not establish
The paper gives direct, time-resolved evidence that silicification changes the shape of the template. At the stage these authors measured, DNA origami shrank by several ångströms in its dimensions and was then partly re-expanded by deposited silica. Silica grew between the helices as well as on the surface. The authors describe the inner order and overall shape as "well-preserved," so what they report is condensation, not loss of design.
For this dossier, the findings raise two questions. First, how exactly does a silicified structure match its designed dimensions? Second, is a silica-templated strut a hollow shell or a filled composite? These matter when reading the strut thicknesses quoted for the Silica Nanolattice. But the evidence is limited. Only two origami shapes were studied. The paper reports a phosphate:TMAPS:TEOS molar ratio of 1:5:12.5 for the cylinder; the brick protocol differed slightly, with TEOS added 15 minutes after TMAPS and no ratio stated. The paper made no mechanical measurements. It does not show that the Gang group's lattices use the same chemistry or undergo the same shrinkage, and it does not test their strength claims. Its reference list does cite related work from that group, including Michelson and colleagues' 2021 Nano Letters paper on silicon carbide frameworks. That is a different paper from the silica-nanolattice study covered on the Aaron Michelson page. The mechanism the authors propose for condensation is screening by TMAPS of the repulsion between helices, plus a hydrophobic effect as less-polar silica displaces water. They present this as an interpretation, not as a direct observation.
Source notes & attribution
- https://arxiv.org/pdf/2204.07385