New reconfigurable DNA nanorobots compute, store data and deliver stuff

A new class of autonomous nanorobots built from reconfigurable DNA origami arrays that can compute, store energy, and deliver nano sized cargo has been developed by researchers from Ludwig-Maximilians-Universität München, Emory University, and the Georgia Institute of Technology.

The nanorobots are based on networks of connected, two-state DNA units. These are arrays that can be programmed to respond to environmental signals.

Authors of the study mention that these reconfigurable arrays were first introduced in 2017 by Yonggang Ke’s lab. Their potential as functional robotic systems has grown through years of research into how their junctions transform.

“I expressed the idea that it would be nice to observe the rules of transformation by visualizing the status of single junctions in the array using dye quencher pairs and single-molecule spectroscopy,” Co-senior author Philip Tinnefeld said.

Detailed in Science Robotics, this work on molecular-scale robotics advances programmable, multi-step tasks powered by energy stored directly within the DNA structures themselves.

In recent years, the collaborating teams analyzed how the DNA arrays change shape and how these transformations could be controlled. They also studied the role of the sequences at the junctions, publishing two key papers in 2024 and 2025 that helped establish a detailed understanding of nanoarray transformation.

The breakthrough that led to the new nanorobot came from Ph.D. students Fiona Cole and Martina Pfeiffer at Ludwig-Maximilians-Universität München. They realized that the origami arrays created by Ke could be treated as a programmable hardware system, where each junction could serve as an independent unit.

“Fiona and Martina realized the unique potential of the reconfigurable arrays for multistep functions,” Tinnefeld said.

They proposed viewing each junction as a component that could be modified with “locks, time delaying units, signalling units, or cargo release units,” which would act as a form of software controlling the DNA-based hardware. This extended previous work, where DNA origami systems typically operated in only two states.

The researchers also showed that these arrays could be pre-loaded with trigger DNA strands that store energy as molecular strain. Ke described the result as “a nanoscale, ‘battery-powered’ machine,” adding, “Another unique feature of this reconfiguration array is that it can be pre-loaded with energy so that they can work autonomously without further energy supply. This is a bit similar to a wind-up car that stores energy as strain.”

The nanoarrays contain dozens of interconnected “anti-junctions,” each of which can be modified to carry out specific tasks. Because the units are connected, they can also communicate, enabling complex cascades of signals and structural changes. The researchers demonstrated that each junction could operate independently, triggering signals or releasing small cargo on command.

“By testing and developing every function unit and combining them on the nanoarray, this autonomous nanobot with programmable functions materialized,” Ke said. The team compared the system to a field-programmable gate array (FPGA), where hardware can be dynamically configured using a set of programming tools.

Although the nanorobot is still at a research stage, the system offers features that could make it suitable for medical uses. It can interact with a wide range of molecules, proteins, and even light, unlike many previous DNA-based nanotechnologies that were limited to nucleic acid interactions. The arrays’ ability to store energy and operate autonomously, powered by allosteric molecular processes, further enhances their potential.

The researchers plan to adapt the nanorobot for different environments and explore alternative energy sources.

“We also plan to address the energy supply problem of nanorobots with the concepts of Brownian DNA computing,” Tinnefeld and Ke said, adding that future versions may use light for operation or expand the design from a flat 2D platform to a fully 3D system.