Fluorescence nanoscopy at the sub-10 nm scale

Abstract

Fluorescence nanoscopy represented a breakthrough for the life sciences as it delivers 20-30 nm resolution using far-field fluorescence microscopes. This resolution limit is not fundamental but imposed by the limited photostability of fluorophores under ambient conditions. This has motivated the development of a second generation of fluorescence nanoscopy methods that aim to deliver sub-10 nm resolution, reaching the typical size of structural proteins and thus providing true molecular resolution. In this review, we present common fundamental aspects of these nanoscopies, discuss the key experimental factors that are necessary to fully exploit their capabilities, and discuss their current and future challenges.

Keywords: Molecular resolution; Single-molecule localization; Super-resolution microscopy.

Fig. 1

Examples of the performance of some sub-10 nm resolution nanoscopy methods using DNA origami (a–e) and biological calibration standards (f-l). a MINFLUX images of DNA origami structures showing ~ 1–2 nm resolution (adapted from (Balzarotti et al. 2017)). b p-MINFLUX localizations and lifetime data of a DNA origami structure (adapted from (Masullo et al. 2021)). c DNA-PAINT images of a DNA origami structure used to demonstrate 5-nm resolution (adapted from (Strauss and Jungmann 2020)). d Images of DNA- origami structures obtained using ROSE, demonstrating 5 nm resolution (adapted from (Gu et al. 2019)). e GET + DNA-PAINT images of a cubic DNA origami structure, where ~ 3 nm axial resolution is proved (adapted from (Kamińska et al. 2021)). f Schematic of the NPC molecular structure. g 3D DNA-PAINT images of the Nup96-HALO in a whole cell nucleus and selection of single NPCs (adapted from (Schlichthaerle et al. 2019)). 12 nm distances between neighboring Nup96 are resolved. h MINFLUX images of NPCs in a U-2 OS cell (adapted from (Gwosch et al. 2020)). i 3D MINFLUX images of an individual NPCs in U-2 OS cells (adapted from (Schmidt et al. 2021)). j Schematic of the microtubule molecular structure. k SIMPLER + DNA-PAINT images of microtubules in COS7 cells, where sub-10 nm-axial resolution allows fully resolving the cross-sections of single microtubules (adapted from (Szalai et al. 2021)). l 4Pi-SMLM image of a microtubule at ~ 10 nm resolution shows the hollow structure clearly resolved (adapted from (Huang, et al. 2016)). Scale bars: a 10 nm, b 5 nm, c 20 nm, d 50 nm, e 30 nm, g 2 µm (left) and 50 nm (right), h 500 nm (top) and 50 nm (bottom), k 1 µm (left) and 50 nm (right), l 300 nm

Fig. 2

Key experimental factors for sub-10 nm resolution nanoscopy. a Drift correction can be performed either actively or by post-acquisition processing. The former strategy keeps the sample in a fixed position with nanometer or sub-nanometer precision, while the latter lets the sample drift freely (often hundreds of nanometers) and corrects the position of each detected molecule during the data analysis, considering the trajectory of the fiducial markers. b Labeling strategies are critical in the final achieved resolution. Bulky labels, such as the combination of primary and secondary antibodies, prevent reaching sub-10 nm resolution. In contrast, small labels (e.g. nanobodies, affimers, aptamers, self-labeling enzymes) introduce minimal linkage errors (3–4 nm). c Detecting with low noise, low dark-counts, single-photon detectors such as avalanche photodiodes (blue) produces data with only Poisson noise. Devices such as EMCCD or sCMOS cameras add significant readout noise to the measurement (orange), increasing the uncertainty in the position estimation