Steric Communication between Dynamic Components on DNA nanodevices

Abstract

Biomolecular nanotechnology has helped emulate basic robotic capabilities such as defined motion, sensing, and actuation in synthetic nanoscale systems. DNA origami is an attractive approach for nanorobotics, as it enables creation of devices with complex geometry, programmed motion, rapid actuation, force application, and various kinds of sensing modalities. Advanced robotic functions like feedback control, autonomy, or programmed routines also require the ability to transmit signals among sub-components. Prior work in DNA nanotechnology has established approaches for signal transmission, for example through diffusing strands or structurally coupled motions. However, soluble communication is often slow and structural coupling of motions can limit the function of individual components, for example to respond to the environment. Here, we introduce a novel approach inspired by protein allostery to transmit signals between two distal dynamic components through steric interactions. These components undergo separate thermal fluctuations where certain conformations of one arm will sterically occlude conformations of the distal arm. We implement this approach in a DNA origami device consisting of two stiff arms each connected to a base platform via a flexible hinge joint. We demonstrate the ability for one arm to sterically regulate both the range of motion as well as the conformational state (latched or freely fluctuating) of the distal arm, results that are quantitatively captured by mesoscopic simulations using experimentally informed energy landscapes for hinge-angle fluctuations. We further demonstrate the ability to modulate signal transmission by mechanically tuning the range of thermal fluctuations and controlling the conformational states of the arms. Our results establish a communication mechanism well-suited to transmit signals between thermally fluctuating dynamic components and provide a path to transmitting signals where the input is a dynamic response to parameters like force or solution conditions.

Accelerating the Characterization of Dynamic DNA Origami Devices with Deep Neural Networks

“…Therefore, the time saved for the entire experiment labor work would be dependent on the target throughput. For example, it typically takes more than 300 particles for sampling dynamic structures with a single condition [9], [35], [36] . Doing the manual annotation for a dataset of ~300-500 particles to directly provide the ground truth could be completed within several hours, which is comparable to the time required to obtain datasets for training and testing DNN performance.…”

“…Ideally, the Resnet50 DNN pose estimation could be applied to a variety of dynamic DNA origami devices To test the robustness of the approach, we applied the same Resnet50 architecture for two other previously obtained particle image datasets: 1) a dataset of hinge devices with incorporated nucleosome (i.e. DNA wrapped around a histone protein core, which is the base packing unit of genomic DNA in eukaryotes) where the nucleosome position is of interest [9], and 2) a dynamic devices with two fluctuating arms where the angular conformation of both arms are of interest [35]. Using the same training protocol, we trained Resnet50 models to predict specified critical features of the device s. e conformations.…”

“…Figure 4B shows the error between the Resnet50 predicted nucleosome position and the manually determined position (ground truth), yielding an average error distance of 3.3 nm for nucleosome. In the second example we used a dynamic device with two fluctuating arms, which we refer to as the SteriDyn [35] due to the steric dynamic interactions of the arms. In this example, we are interested in the conformation of two moving components on a single device (i.e., the angle of each arm relative to the base platform) .…”

“…We first establish the approach using a 'Hinge' nanostructure, which is representative of dynamic nanodevices that are widely used for biophysical measurements [9], biosensing [33], and controlling biomolecular interactions [34]. Secondly, we demonstrated the robustness and versatility of this pipeline by applying it to other dynamic DNA origami device characterizations including a 'Hinge-Nucleosome' system [9] and a three-arm devices designed to exhibit steric interactions between the arms [35]. Our results suggest that the DNN algorithm can be used to overcome the bottlenecks that require excessive labor work for post-processing of micrographs to characterize dynamic DNA nanodevices, which can greatly facilitate the design, experiment, and development cycle.…”

Optimizing Binding among Bimolecular Tethered Complexes

“…42 An example of such a system can be inferred from Cao et al, 11 where DNA origamis are grafted to a surface by means of long single stranded tethers. As in most DNA origami applications, we expect binding between different molecules to be reaction-limited 23,43,44 and thus binding rates will scale as κQ(ϵ) with ϵ roughly the diameter of the molecules involved. By introducing placement techniques such as lithography 45−47 to further control the position of these tethered structures, the methodology and results here introduced can rapidly inform the selection of the optimal relations between tether lengths and locations to strategically maximize (or minimize) assembly times between different nanocomponents at relatively low computational cost, facilitating precise control over the assembly process.…”