Life’s molecular motors — myosin hauling cargo through muscle fibres, kinesin ferrying vesicles along microtubules, the ribosome translating messenger RNA one codon at a time — convert chemical energy into directed mechanical work with a precision and efficiency that synthetic systems have long struggled to approach. Building an artificial equivalent entirely from proteins, rather than from the comparatively tractable chemistries of small organic molecules or DNA origami, has remained one of nanotechnology’s most stubborn unmet goals. A study published in July 2026 in Nature Nanotechnology by Patrik Nilsson, Paul Curmi and an international team spanning Lund University, the University of New South Wales, and collaborating institutions in Sweden, Australia, Canada and the United Kingdom now reports the realization of exactly such a device: an externally controlled protein walker named Tumbleweed.
The scientific challenge underlying this work is essentially one of modular design. Previous synthetic motors built from DNA or small molecules had already demonstrated directional motion, but none had matched the speed, processivity or structural sophistication of naturally evolved protein motors, largely because assembling flexible, information-transmitting protein architectures from scratch remains beyond current de novo design capabilities. Rather than designing an entirely new protein fold, the authors pursued a modular, bottom-up engineering strategy, repurposing three well-characterized bacterial DNA-binding repressor proteins — TrpR, DtxR and MetJ — as interchangeable “feet,” each activated by a distinct small-molecule ligand (tryptophan, cobalt ions and S-adenosylmethionine, respectively) and joined to a central hub through coiled-coil “legs” using the SpyTag/SpyCatcher covalent-assembly system.
To generate directional movement, the team engineered a synthetic DNA track bearing repeating, ordered binding sites for the three feet, then bathed the resulting Tumbleweed protein in a microfluidic device that cycled through pairs of controlling ligands in a defined temporal sequence. Using single-molecule Förster resonance energy transfer, the researchers directly visualized individual Tumbleweed molecules taking discrete, 16-nanometer steps along the track, with both the timing and the direction of stepping precisely dictated by the order in which ligands were introduced — a demonstration that walking behavior can emerge from the coordinated interplay of non-motor protein components rather than requiring a purpose-built motor domain.
Importantly, the study is candid about the system’s current constraints. Individual Tumbleweed molecules took, on average, only a handful of steps before dissociating from the track, and single-molecule traces revealed that the walker “oversteps” onto non-adjacent binding sites in roughly a third of observed transitions, an imprecision the authors attribute to the inherent flexibility of both the DNA track and the protein’s linker regions. Present speeds, on the order of one nanometer per second, remain one to two orders of magnitude slower than natural motors such as kinesin, and the entire system has so far been characterized only in vitro, under externally imposed ligand cycling rather than autonomous chemical fuel consumption.
These limitations notwithstanding, the achievement carries meaningful implications for the future of engineered nanomachines. Because Tumbleweed is built from modular, interchangeable domains rather than a monolithic bespoke structure, its feet, ligands and track sequences could in principle be substituted to program entirely different behaviors, opening a route toward protein-based devices for massively parallel biocomputation, single-molecule diagnostics and eventually autonomous molecular delivery systems that would not depend on external microfluidic control. Whether such devices can be engineered to operate independently, drawing energy from ambient chemical gradients rather than researcher-imposed cycling, remains the central unresolved question — and, quite plausibly, the next major milestone in synthetic protein motor design.
Leave a Reply