Stefan W. Hell | PAS Academician

Molecular-scale resolution in fluorescence microscopy

Optical microscopy has become nanoscopy: The MINFLUX and MINSTED concepts provide truly molecule-size resolution

Following 20th-century textbook wisdom, the resolution of light microscopy is limited by optical diffraction to about half the wavelength of light, which is why conventional light microscopes fail to distinguish object details that are closer together than about 200 nanometers.

Breaking this century-old diffraction barrier appeared as a far-out, perhaps entirely unrealistic goal. And yet, since the early 1990s, physical concepts were put forward and later demonstrated in practice, which resulted in novel fluorescence microscopes featuring diffraction-unlimited spatial resolution (Hell, 2007). The work of my research group laid the foundation of a new scientific field: superresolution fluorescence microscopy, also called fluorescence nanoscopy. The novel superresolution methods are destined to become primary tools for imaging living biological samples ranging from cells and tissues to small animals, with the potential to transform the life sciences.

Prominent methods implementing the first viable concepts to overcome the resolution barrier include STED and RESOLFT microscopy, which were complemented a few years later by the PALM/STORM concept, initially demonstrated with fluorescent proteins switching by a related mechanism. All these methods utilize reversible transitions or switching of fluorescent labels between a bright and a dark state. Since these superresolution concepts fundamentally rely on transitions between molecular states, novel labels are required that can be optically prepared in at least two different states. Consequently, my research groups in Göttingen and Heidelberg also pioneer the chemical synthesis and application of new labeling methods and techniques to improve the performance of the labels’ switching behavior to separate close-by molecules.

Now, in the early 2020s, the outlook for this field is truly exciting. We continue to push the performance of nano-optical molecular imaging in (living) cells and tissues. As the Nobel foundation had put it on their widely distributed posters, “[the 2014 Chemistry laureates] had crossed the [resolution] threshold”. And yet the actual holy grail of the superresolution field, molecule-size resolution, had remained outstanding until recently: With resolution at the ~1-nm scale, MINFLUX was the first concept to push the resolution of fluorescence microscopy to truly molecular dimensions (Balzarotti et al., 2017).

Of all the nanoscopy or super-resolution advances of the last decade, MINFLUX stands out, because it contains a radically new idea for localizing individual fluorophores. In PALM/STORM imaging with typical resolutions of 20-30 nm, the localization of a molecule is based on maximizing the number of detected fluorescence photons on a camera, which is inevitably limited by photobleaching. By contrast, in MINFLUX the molecule is actively localized by probing the molecule’s position in the vicinity of the intensity zero of the donut-shaped excitation beam (Fig. 1). By iteratively refining the probing steps, the molecule’s position ultimately coincides with the position of the donut at which fluorescence emission is minimal. Under ideal conditions, the MINFLUX localization precision doubles with every five photons. By abolishing a strict dependence just on the photon count, MINFLUX releases the ‘square root brake’ – which traditionally makes higher resolutions increasingly more difficult to achieve. The exponential dependence thus makes single nanometer resolutions and below now realistically achievable, including in cellular fluorescence microscopy (Fig. 2).

By fundamentally reducing the required number of fluorescence photons, MINFLUX is able to detect molecular positions and movements of a few nanometers, at temporal sampling speeds of hundreds of microseconds, while maintaining ~2-nm precision (Eilers et al., 2018). MINFLUX has thus opened the door to low-light level optical analysis of tiny objects at true molecular scale resolution (1-5 nm). And lens-based fluorescence microscopy has reached the ultimate resolution limit: the size of the fluorescent molecule itself. For practical biological purposes, the ultimate resolution limit is now merely set by the size and orientation of the molecular tag consisting of the fluorophore plus its specific molecular linker. Moreover, the resolution is attained at relatively high speed, at least 10 times faster than in PALM/STORM.

Further progress in implementations and applications of the concept has been achieved, including nanometer resolution in three dimensions, in living cells and multiple color channels (Gwosch et al., 2020), multicolor-3D imaging of densely packed mitochondrial proteins (Pape et al., 2020), MINSTED – a new super-resolution concept related to the same powerful idea of “optical coordinate injection” as MINFLUX (Weber et al., 2021, see below). We also contributed to the realization of MINFLUX on a conventional microscope stand (Schmidt et al., 2021). The fundamentally strong localization properties of the so-called 4Pi arrangement, collecting and recombining fluorescence emissions from both sides of the sample, continue to be of great interest. This approach, which historically led to the first substantial axial resolution improvements, features a sharp signal modulation along the optical axis due to interference, and the 4Pi illumination and detection principle’s fundamental benefits for MINFLUX are also the subject of ongoing explorations to push molecular imaging and tracking capabilities in 3D to new heights.

To highlight the novel spatiotemporal capabilities created, some of the initial work focused on tracking the movements of various protein complexes. In early tracking experiments, sub-millisecond position sampling had been demonstrated (Balzarotti et al., 2017). In the last couple of years, we improved on this work by advancing the measurement principle to an unprecedented spatiotemporal resolution of 5×10-10 m×s-1 with as few as 40 photons per localization. Current work focuses on tracking the stepping behavior of the molecular motor Kinesin-1. Kinesin-1 moves along microtubules at a rate of approximately 400 nm per second under physiological conditions with a step distance of 16 nm and a sub-step distance of 8 nm, which is an ideal model system for displaying the capabilities of MINFLUX. To date we can track Kinesin with single digit nanometer precision at a temporal resolution below 1 ms (Fig. 3), which allows us to clarify open questions regarding the stepping mechanism of Kinesin-1 (Wolff et al., 2022). In particular, MINFLUX all-optical tracking also extends into the regime of physiological ATP concentration, without the need for a bulky bead marker. Longstanding controversies such as whether ATP binding takes place in the one-head-bound or the two-head-bound state can now be answered with MINFLUX.


MINSTED: back to the roots for maximal precision

Considerations of the underlying principles – the paradigm shift that enabled the jump from typically 10s of nanometers to single-digit and even ~1-nm precision – came with the conviction that MINFLUX would not remain the only molecular resolution method, but rather would represent the first member of a new family of techniques with this level of detail. As the name suggests, MINSTED localization and nanoscopy (Figs. 4 and 5) relies on the original STED principle even more than MINFLUX. Like MINFLUX, it achieves molecular resolution, but the resolution can in fact be adjusted almost continuously from confocal-spot dimensions of 200-300 nanometers down to the molecular size.

In conventional STED imaging, only the molecules in the middle of the donut-shaped beam can fluoresce. Thus, the experiment always “knows” where the emitting molecules are. STED microscopy does not typically achieve molecular resolution because in practice the donut-shaped fluorescence inhibition beam cannot be made so strong that only a single fluorophore can fit into the central intensity minimum. For this reason, in MINSTED (Weber et al., 2021 and Weber et al., 2022), the fluorophores are initially isolated by randomly switching them on through an independent photochemical switching process. The fluorescence-preventing STED donut beam is then used to locate the fluorescent molecules individually. Its minimum, the intensity zero, serves as an ‘optically injected’ reference point. If the minimum coincides with the fluorophore, the fluorophore emits most strongly and one can find out precisely where it is, because the exact position of the STED donut beam is always known. The MINSTED experiment therefore gradually approaches the fluorophores in a targeted fashion based on direct feedback from every photon detection event with the donut beam (Fig. 4) and can thus locate the fluorescent molecules with precision and accuracy of 1 to 3 nanometers. In connection with the photochemical on and off switching which samples the ensemble of fluorophores over time and provides them in the on-state one-by-one, the resolution becomes molecular-scale.

An important aspect in practice, the STED light – applied to inhibit spontaneous emission peripheral to the scan coordinate – ensures that any additional fluorophore or fluorophores that happen to be activated inadvertently in the vicinity of the probed position are kept off. The result is a substantial reduction in background noise contributions to the MINSTED single-molecule registrations, and improved signal-to-background ratio in many situations compared to MINFLUX, where the excitation donut excites the fluorescence from all active fluorophores, including peripheral ones that should be avoided.

Entering the Ångström-level localization regime

The exciting developments seem to know virtually no bounds: In recent experiments with a blue-shifted MINSTED implementation, our analysis showed that if 10,000 emissions can be detected from the fluorophore under the same practical background and stability conditions, the precision is estimated to 𝜎 = 2.3 Å, a value that is in fact about 8 times smaller than the size of the fluorophore itself (Weber et al., 2022).

Microscopy on the molecular scale is here to stay. It is to be expected that MINSTED and MINFLUX will become widely used in the life sciences. The inherent confocality of MINFLUX, MINSTED and related concepts should also provide a critical advantage when considering imaging in more dense and three-dimensional specimens, such as brain slices and in-vivo imaging scenarios. With further development of other aspects, including field of view enlargement, etc., MINFLUX and MINSTED are bound to transform the limits of what can be observed in cells and molecular assemblies with light. This should most probably impact cell and neurobiology and possibly also structural biology. Moreover, it should be a great tool for studying molecular interactions and dynamics in a range that has not been accessible so far.


Hell SW (2007) Far-Field Optical Nanoscopy. Science 316: 1153-1158.

Balzarotti F, Eilers Y, Gwosch KC, Gynna AH, Westphal V, Stefani FD, Elf J, Hell SW (2017) Nanometer resolution imaging and tracking of fluorescent molecules with minimal photon fluxes. Science 355: 606-612.

Eilers Y, Ta H, Gwosch KC, Balzarotti F, Hell SW (2018) MINFLUX monitors rapid molecular jumps with superior spatiotemporal resolution. Proc Natl Acad Sci USA 115: 6117-6122.

Gwosch KC, Pape JK, Balzarotti F, Hoess P, Ellenberg J, Ries J, Hell SW (2020) MINFLUX nanoscopy delivers 3D multicolor nanometer resolution in cells. Nat Methods 17: 217-224.

Pape JK, Stephan T, Balzarotti F, Buchner R, Lange F, Riedel D, Jakobs S, Hell SW (2020) Multicolor 3D MINFLUX nanoscopy of mitochondrial MICOS proteins. Proc Natl Acad Sci USA 117: 20607-20614.

Weber M, Leutenegger M, Stoldt S, Jakobs S, Mihaila TS, Butkevich AN, Hell SW (2021) MINSTED fluorescence localization and nanoscopy. Nat Photonics 15: 361-366.

Schmidt R, Weihs T, Wurm CA, Jansen I, Rehman J, Sahl SJ, Hell SW (2021) MINFLUX nanometer-scale 3D imaging and microsecond-range tracking on a common fluorescence microscope. Nat Commun 12:1478.

Wolff JO, Scheiderer L, Engelhardt T, Engelhardt J, Matthias J, Hell SW (2022) MINFLUX dissects the unimpeded walking of kinesin-1. bioRxiv,

Weber M, von der Emde H, Leutenegger M, Gunkel P, Cordes VC, Sambandan S, Khan TA, Keller-Findeisen J, Hell SW (2022) MINSTED nanoscopy enters the Ångström localization range. Nat Biotechnol (in press)