Fig. 1: miRNA detection using dynamic FRET-FISH.

a Scheme of target miRNA detection. Only the acceptor signal is used for miRNA detection whereas only donor fluorophores (Alexa-488 and Cy3) are excited. In this way, background acceptor fluorescence coming from floating accepter strands is reduced because the acceptor signal is generated only when both donor and acceptor strands simultaneously bind to target miRNAs. b miRNA preparation. Purified RNA is hybridized with biotinylated poly (T) after poly (A) tailing. Then RNA is immobilized on a surface via biotin-streptavidin interaction. c Sequences of the target miRNA (let-7a), Cy5-labeled seed probe, Alexa-488-labeled mid probe, and Cy3-labeled tail probe. d Representative Cy5 fluorescence intensity time traces (red) detected at Alexa-488 excitation (top) and at Cy3 excitation (bottom). The blue and green traces represent the results of hidden Markov modeling of acceptor signals at Alexa-488 and Cy3 excitations, respectively. e Observation time dependency of the let-7a number detected using dynamic FRET-FISH (blue) and Ago-FISH (orange). In dynamic FRET-FISH, we identified as a proper target miRNA if a molecule exhibited an acceptor (Cy5) fluorescence signal at both Alexa-488 and Cy3 excitations during the designated observation time. In Ago-FISH, 5% duty cycle of the three DNA probes was used as a threshold for target identification⁶. The duty cycle is defined as the fraction of the total observation time in which a probe was binding to the target miRNA. For dynamic FRET-FISH, 1200 nM donor and 40 nM acceptor probes were used. For Ago-FISH, 2 nM probes were used.