Abstract
Biomolecules in living cells experience complex multi-directional mechanical forces that regulate their structure, dynamics, and function. However, most single-molecule techniques primarily exert force along a single axis, thereby failing to emulate the mechanical cellular environment. Here, we present Multi-Axial Entropic Spring Tweezer along Rigid DNA Origami (MAESTRO), a molecular platform that applies up to 9 pN forces from up to four directions simultaneously using programmable ssDNA entropic springs anchored to a rigid DNA origami scaffold. Combining MAESTRO, single-molecule Förster resonance energy transfer (smFRET), and Bayesian non-parametric FRET (BNP-FRET) enables a high-throughput study of biomolecules under different complexities of multi-axial tension forces. We applied MAESTRO to Holliday junctions (HJs), four-way DNA intermediates that experience multi-directional tension during homologous recombination. Counterintuitively, we discovered ≥5× slower kinetics of the HJ conformations under multi-axial tension than under tension-free conditions, enabling direct observations of previously hidden HJ conformational intermediates. Most remarkably, we discovered that multi-axial tension restores quasi-ergodicity to HJ dynamics by overcoming the rugged energy landscape, enabling direct observation of kinetic class interconversion within individual molecules-a phenomenon previously thought impossible-that reveals the conformational landscape is far more interconnected than understood and fundamentally challenges existing models. Furthermore, we demonstrated that this conformational control regulates T7 endonuclease I cleavage site selection, directly linking mechanical environments and molecular mechanics to enzymatic function. By overcoming single-axis limitations, MAESTRO opens new frontiers in molecular mechanobiology, revealing how physiologically relevant multi-directional forces access expanded conformational landscapes and can serve as master regulators for biomolecular function through mechanisms inaccessible to conventional single-axis approaches.