Computational DNA Droplets Recognizing miRNA Sequence Inputs Based on Liquid–Liquid Phase Separation

Phase-separated biomolecular droplets are formed in cells to regulate various biological processes. This phenomenon can be applied to constructing self-assembled dynamic molecular systems such as artificial cells and molecular robots. Recently, programmable phase-separated droplets called DNA droplets have been reported as a possible method to construct such dynamic molecular systems. This study reports a computational DNA droplet that can recognize a specific combination of tumor biomarker microRNAs (miRNAs) as molecular inputs and output a DNA logic computing result by physical DNA droplet phase separation. A mixed DNA droplet consisting of three DNA nanostructures with orthogonal sticky-end sequences and two linker DNAs to cross-bridge the orthogonal DNA nanostructures is proposed. By the hybridization of miRNAs with the linkers, the cross-bridging ability is lost, causing the phase-separation of the mixed DNA droplet into three DNA droplets, resulting in executing a miRNA pattern recognition described by a logical expression ((miRNA-1 ∧ miRNA-2) ∧ (miRNA-3 ∧ ¬miRNA-4)). This experimentally demonstrates that the computational DNA droplets recognize the above specific pattern of chemically synthesized miRNA sequences as a model experiment. In the future, this method will provide potential applications such as diagnosis and therapy with integration to biomolecular robots and artificial cells.