DNA-PAINT is a super-resolution imaging technique that relies on the transient binding of short fluorescent DNA “imager” strands to complementary “docking” strands attached to the target structure. Each binding event produces a localized burst of fluorescence that can be precisely detected and accumulated to reconstruct the image at nanometer resolution.

However, one major limitation remains: imager strands that are not bound continue to diffuse in the sample and emit fluorescence, creating background signal. This prevents researchers from using high imager concentrations and significantly slows down the acquisition process.

To overcome these limitations, a research team led by Yves Mely at the Laboratory of Bioimaging and Pathology (Strasbourg University) in collaboration with a team led by Alain Burger (Nice Institute of Chemistry) developed a new approach that incorporates a dark donor dye into the imager strand. A dark donor is a dye that remains almost non-fluorescent on its own but can transfer its energy to a nearby fluorescent acceptor when the two are brought together. In this system, the modified nucleobase X acts as the dark donor: it stays essentially dark in solution, but when the imager hybridizes with the docking strand labelled with ATTO 647N, X activates the acceptor’s fluorescence. As the signal appears only during true binding events, this fluorogenic behaviour markedly reduces background noise and enables the use of higher imager concentrations.

Single-molecule experiments confirm that the system maintains binding kinetics compatible with DNA-PAINT, and that fluorescence increases roughly 50-fold upon duplex formation. The method was then applied to fixed HeLa cells: microtubules were reconstructed in around 30 seconds, with a resolution of ~50 nm and a median localization precision of 18 nm. By comparison, classical DNA-PAINT required 30 minutes to reach a similar result.

When compared to FRET-PAINT, a variant of DNA-PAINT in which fluorescence is generated through energy transfer between a donor and an acceptor dye brought together during hybridization, the dark-donor strategy showed a clear advantage. FRET-PAINT can suffer from signal leakage, as the donor dye may emit light in the acceptor detection channel. In contrast, the dark-donor system produced far less leakage, leading to cleaner images while preserving a similar acquisition speed.

The main limitation of this first-generation system lies in the photobleaching of ATTO 647N, which shortens the usable imaging time. The authors suggest possible improvements, including the use of more photostable acceptor dyes or the development of new donor–acceptor pairs with enhanced brightness to support longer and higher-resolution acquisitions.

Overall, this work provides the first proof of concept that dark-donor DNA-PAINT can deliver fast, low-background super-resolution imaging and could become a valuable addition to the growing set of DNA-based nanoscopy tools.